EXCURSION MÉTALLOGÉNIQUE
Santiago – San Pedro d’Atacama – Chili 2012 6 au 24 mai
Organisé par le chapitre étudiant
Responsables : Alexis Paulin-Bissonnette Sophie Maltais
Co-responsables :
Denis Côté Lou Millot Sandra Lalancette
Excursion métallogénique Chili 2012 page 2
Organisé par le chapitre étudiant de la Society of Economic Geologist de l’Université du Québec à Chicoutimi
Conseil exécutif : Alexis P.-Bissonnette (Pésident) Sophie Maltais (Vice-présidente) Lou Millot (Secrétaire) Sandra Lalancette (Trésorière)
Comité géologie : Alexis P.-Bissonnette Denis Côté Jacques Carignan
Comité activités et logistique : Sophie Maltais Martine Chabot
Comité commandites : Lou Millot Christophe P.-Doucet Gabrielle Rochefort Jakrapun Khambooruang Benoît M.-Tanguay Jean-David Pelletier Jean-Philippe Arguin Tonny Girard
Comité transport : Sandra Lalancette Géraldine St-Pierre Quentin Duparc
Comité livret guide : Sophie Maltais Christophe P.-Doucet
Excursion métallogénique Chili 2012 page 3
MERCI À TOUS NOS GÉNÉREUX PARTENAIRES
Excursion métallogénique Chili 2012 page 4
ITINÉRAIRE
6 mai 2012 : 17h20 Départ de l’aéroport de Bagotville, transfert à l’aéroport de Montréal et de Toronto
7 mai 2012 : 8h25 Arrivée à l’aéroport international Arturo Merino Benitez, transfert vers l’hôtel Chile Apart, journée libre
8 mai 2012 : 6h15 Départ pour la visite de la mine Andina
9 mai 2012 : 7h00 Départ pour la visite de la mine El Teniente
10 mai 2012 : 7h00 Départ pour la visite de la mine Los Bronces
11 mai 2012 : 7h00 Départ pour la visite de la mine El Soldado
12 mai 2012 : 11h00 Départ pour la visite du vignoble Concha Y Toro
13 mai 2012 : Journée libre
14 mai 2012 : 8h00 Conférence sur la géothermie à l’Université du Chili
15 mai 2012 : 5h30 Départ de l’aéroport Arturo Merino Benitez 7h30 Arrivée à l’aéroport de Calama
16 mai 2012 : 19h00 Visite de l’observatoire SPACE
17 mai 2012 : 6h30 Excursion Vallée de la lune 14h00 Excursion Vallée de l’arc-en-ciel.
18 mai 2012 : 7h30 Excursion Lagunas altiplanicas
19 mai 2012 : 8h30 Excursion Trekking machuca
20 mai 2012 : 4h30 Excursion Geyser del Tatio 15h00 Excursion Ojos del Salar
21 mai 2012 : 8h00 Tour archéologique 14h00 Baignade Termes de Puritama
22 mai 2012 : 20h00 Départ de l’aéroport de Calama 22h00 Arrivée à l’aéroport international Arturo Merino Benitez, transfert vers l’hôtel Chile Apart
23 mai 2012 : 18h20 Départ de l’aéroport Arturo Merino Benitez, transfert à l’aéroport de Toronto et de Montréal
24 mai 2012 : 16h55 Retour à l’aéroport de Bagotville
Excursion métallogénique Chili 2012 page 5
INFORMATIONS VOLS
Vol Bagotville – Santiago : Air Canada Vol Santiago – Calama : LAN Airlines
# Vol Date Départ Heure Arrivée Heure AC8685 6 mai Bagotville 17h20 Montréal 18h33 AC427 6 mai Montréal 20h00 Toronto 21h15 AC92 6 mai Toronto 22h55 Santiago 9h25 LA140 15 mai Santiago 5h00 Calama 7h10 LA157 22 mai Calama 20h00 Santiago 22h05 AC93 23 mai Santiago 18h30 Toronto 5h40 AC486 24 mai Toronto 7h30 Montréal 8h40 AC8684 24 mai Montréal 15h50 Bagotville 16h57
BAGAGES
1 bagage de moins de 23kg, dimension maximale de 203cm (longueur+largeur+hauteur)
1 bagage à main de moins de 8kg, dimension maximale de 55cm*23cm*35cm
1 bagage à main de moins de 8kg, dimension maximale de 43cm*33cm*16cm
Excursion métallogénique Chili 2012 page 6
CHILI - MAI 2012
Dimanche Lundi Mardi Mercredi Jeudi Vendredi Samedi 6 7 8 9 10 11 12 Départ de Arrivée à Mine Mine Mine Mine Vignoble Bagotville Santiago Andina El Teniente Los Bronces El Soldado Concha y Toro 17h20 8h25 8h00 8h00 8h00 8h00 11h00
13 14 15 16 17 18 19 Journée libre Conférence Départ de Arrivée à Vallée de Lagunas Trekking Université du Santiago San Pedro la lune Altiplanicas Machuca à Chili 5h30 14h00 6h30 7h30 Rio Grande 8h30 Arrivée à Vallée de l’arc Antofagasta Observatoire en ciel 16h00 19h00 14h00
20 21 22 23 24 25 26 Geyser del Tour Départ de Départ de Arrivée à Tatio Archéologique l’Hôtel l’Hôtel Bagotville 4h30 8h00 10h00 12h00 16h55
Termes de Départ de Départ de Ojos del Salar Puritama Calama Santiago 15h00 14h00 20h00 18h20
Excursion métallogénique Chili 2012 page 7
Excursion métallogénique Chili 2012
VISITES DE MINES
Excursion métallogénique Chili 2012 page 8
LOCALISATION DES MINES
Excursion métallogénique Chili 2012 page 9
ANDINA 8 Mai 2012
Horaire : Départ de l’hôtel à 6h15 Services : Dîner
LOCALISATION Le gisement se situe à 80 km au nord de Santiago à 2 km de la mine Los Bronces à une altitude de 3 700m à 4 200m.
PROPRIÉTAIRE La division Andina appartient à Codelco.
TYPE DE GISEMENT Cu porphyrique
PRODUCTION Andina extrait présentement 188 494 tonnes métriques par an de concentré de cuivre fin et 2 901 tonnes métriques de molybdène. Les ressources estimées s’élèvent à 114 millions de tonnes de cuivre.
http://www.telegraph.co.uk/finance/newsbysector/industry/mining/8892695/Anglo-American-knocked-in-first-battle- with-Codelco.html
Excursion métallogénique Chili 2012 page 10 HISTORIQUE Les opérations de la division Andina se font depuis 1970 dans le dépôt de Rio Blanco connu depuis 1920. La division exploite du minerai dans la mine souterraine Rio Blanco et dans la mine à ciel ouvert Sur Sur. Un projet d’expansion est en cours afin d’augmenter les activités d’extraction, passant de 70 000 tonnes par jours à 94 000 tonnes par jour.
GÉOLOGIE La géologie d’Andina fait partie du complexe de Los Bronces. Voir références pour plus de détails.
RÉFÉRENCES
Sillitoe, Richard H., Porphyry Copper Systems, 2010, Economic Geology, v. 105, 3-41
Warnaars, F.W., Holmgren, C. & Barassi, S. (1985). Porphyry copper and Tourmaline breccias at Los Bronces, Rio Blanco, Chile. Economic Geology 80, 1544-1565 http://www.codelco.cl/prontus_codelco/site/edic/base/port/andina.html
Excursion métallogénique Chili 2012 page 11 EL TENIENTE 9 Mai 2012
Horaire : Indéterminée Services : Dîner
LOCALISATION La mine est située à 75 km au sud de Santiago à environ 2500m d’altitude.
PROPRIÉTAIRE El Teniente appartient à l’entreprise nationale CODELCO.
TYPE DE GISEMENT Cu-Mo porphyrique
PRODUCTION Sa production en 2010 a atteint 403 616 tonnes de cuivre et 5 617 tonnes de molybdène. Environ 4500 employés qui y travaillent.
https://communismeouvrier.wordpress.com/2011/06/10/chili-lutte-des-ouvriers-a-la-mine-d%E2%80%99el-teniente/
Excursion métallogénique Chili 2012 page 12 HISTORIQUE Cette mine est exploitée depuis 1904 et possède plus de 2 400km de galeries souterraines. Cette entreprise exploite la plus grande mine de cuivre souterraine au monde et a enregistré sa meilleure année de profits en 2007.
GÉOLOGIE Voir les articles Magmatic evolution of the giant El Teniente Cu-Mo deposit, central Chile et Genesis of the Late Miocene to Pliocene copper deposits of central Chile in the context of Andean magmatic and tectonic evolution
RÉFÉRENCES
Sillitoe, Richard H., Porphyry Copper Systems, 2010, Economic Geology, v. 105, 3-41
Skewes, A.; Stern, C.R. 1995. Genesis of the Late Miocene to Pliocene copper deposits of central Chile in the context of Andean magmatic and tectonic evolution. Queen's University, 38-56
Stern, C.R.; Skewes, M.A.; Arévalo, A. 2010. Magmatic evolution of the giant El Teniente Cu-Mo deposit, central Chile. Journal of Petrology. 1-27
Excursion métallogénique Chili 2012 page 13 LOS BRONCES 10 Mai 2012
Horaire : Départ de l’hôtel vers 7h00 Services : Transport + Dîner
LOCALISATION La mine Los Bronces se situe à 65 km de Santiago à 3 500m d’altitude.
PROPRIÉTAIRE La division de Los Bronces appartient à la compagnie Anglo-American.
TYPE DE GISEMENT Cu-Mo porphyrique
PRODUCTION La mine produit annuellement environ 235 800 tonnes de cuivre ainsi que 2 578 tonnes de molybdène. La mine est à ciel ouvert. Environ 1 700 employés travaillent pour cette division.
http://www.anglochile.cl/en/operaciones/br_imagenes.htm#
Excursion métallogénique Chili 2012 page 14 HISTORIQUE Le gisement de Los Bronces fut découvert en 1862. L’exploitation commença seulement en 1916. Anglo-American acheta le gisement en 2002 et commença 3 importants projets d’expansion de la mine pour être capable d’extraite environ 54 000 tonnes par jour.
GÉOLOGIE Voir l’article Genesis of the Late Miocene to Pliocene copper deposits of central Chile in the context of Andean magmatic and tectonic evolution
RÉFÉRENCES
Skewes, A.; Stern, C.R. 1995. Genesis of the Late Miocene to Pliocene copper deposits of central Chile in the context of Andean magmatic and tectonic evolution. Queen's University, 38-56 http://www.angloamerican-chile.cl/our-operations/losbronce.aspx?sc_lang=en
Excursion métallogénique Chili 2012 page 15 MINE EL SOLDADO 11 Mai 2012
Horaire : Départ de l’hôtel vers 7h00 Services : Transport + Dîner
LOCALISATION Le gisement se situe 132 km au nord-ouest de Santiago à 600m d’altitude
PROPRIÉTAIRE La division El Soldado appartient à Anglo-American
TYPE DE GISEMENT Cu porphyrique
PRODUCTION La production annuelle de cuivre tourne autour de 50 000 tonnes de cuivre. 1 400 employés travaillent pour la division El Soldado. Le développement de la mine se fait à la fois à ciel ouvert et par extraction souterrain.
HISTORIQUE Les opérations de la mine ont commencé en 1842. Elle fut exploité par méthodes souterraines jusqu’en 1989 où elle fut aussi transformée en mine à ciel ouvert. Anglo American devient l’actionnaire majoritaire en 2002 et propose un plan d’extension de la mine en 2004. Ce plan d’extension fut concrétisé en 2006 et on estime que la mine pourra continuer ces opérations jusqu’en 2026.
GÉOLOGIE Voir article Organic petrology, chemical composition, and reflectance of pyrobitumen from the El Soldado Cu deposit, Chile.
RÉFÉRENCES
Sillitoe, Richard H., Porphyry Copper Systems, 2010, Economic Geology, v. 105, 3-41
Wilson, N.S.F., 2000. Organic petrology, chemical composition, and reflectance of pyrobitumen from the El Soldado Cu deposit, Chile. International Journal of Coal Geology, 53–82.
Excursion métallogénique Chili 2012 page 16
Excursion métallogénique Chili 2012
SANTIAGO
Excursion métallogénique Chili 2012 page 17
LOCALISATION SANTIAGO
Excursion métallogénique Chili 2012 page 18
SANTIAGO
URGENCES MÉDICALES HÔPITAL CENTRAL Avenue Portugal, 125 +463-38-00
ADRESSES UTILES BIBLIOTECA NACIONAL Angle Miraflores et O’Higgins +360-52-00
Lundi au samedi, 9h à 18h45 avec internet CONSULAT DE FRANCE Condell, 65 +470-80-00
CONSULAT DU CANADA Nueva Tajamar, 481, Torre Norte +652-38-00
MERCADO CENTRAL Angle San Pablo et 21-De-Mayo
Ouvert tous les jours, 7h à 17h, réputé pour ses poissons et fruits de mer Plusieurs petits restos intéressants pour le dîner
OFFICE DU TOURISME Plaza de Armas +713-67-45 Munistgo.cl
Excursion métallogénique Chili 2012 page 19 UNIVERSIDAD DE CHILE O’Higgins, 1058 +978-20-00 Uchile.cl
OÙ DORMIR
CHILE APART (7 au 15 mai, 22 mai) Merced 562, Oficina 818-A, Centro Santiago, 8320148 Chili +562-632-57-86
OÙ MANGER BAR NATIONAL Bandera, 317 +365-33-68
Style casse-croûte, cuisine chilienne entre 7$ et 15$, parfait pour un dîner rapide Tous les jours 8h30 à 22h, sauf dimanche CASA NARANJA Santo Domingo, 528 +639-58-43
À l’allure d’une galerie d’art, bonne cuisine, entre 14$ et 30$, plus cher le soir Tous les jours, sauf dimanche DON VICTORINO Lastarria, 138 +639-52-63
Quartier chaleureux, avec terrasse et mezzanine. Lundi au samedi Menu à la carte, entre 15$ et 30$
Excursion métallogénique Chili 2012 page 20 LES ASSASSINS Merced, 297 B +638-42-80
Lundi au vendredi et samedi soir, jusqu’à 23h, environ 20$ et plus en soirée Petit resto genre bistrot, cuisine style française PATAGONIA Lastarria, 96 +664-38-30
Resto vinothèque chaleureux avec certains plats moins onéreux Ouvert le midi et le soir RESTAURANT R. Lastarria, 307 +664-98-44
Lundi au vendredi et samedi soir, plus de 30$, plus chic en soirée Évoque une chaleureuse chaumière avec terrasse, cuisine internationale SAN MARCO Huerfarnos, 618 +633-68-80
Bon resto italien avec carte très variée, excellents déjeuners servis jusqu’à 15h Tous les jours 12h à 23h sauf dimanche, entre 15$ et 30$
OÙ SORTIR ABARZUA Merced, 337 +638-72-56
Tous les jours 12h à 3h (dimanche 22h), bar style industriel, très tendance BOMBON ORIENTAL Merced, 345 +639-10-69
Tous les jours 10h à 21h, pâtisseries orientales, petite terrasse
Excursion métallogénique Chili 2012 page 21 CAFÉ COLONIA Mac Iver, 161 +639-72-56
Tous les jours 8h à 21h, salon de thé à l’ancienne, délicieux gâteaux allemands
CAFÉ ESCONDIDO Rosal, 346 +632-73-56
Lundi au samedi, 19h à 2h (samedi 4h), bistrot tranquille avec resto
EL DIABLITO Merced, 336 +638-35-12
Tous les jours 18h à 4h30 (dimanche 1h), pub de style vieillot Consommations à prix très abordables et quelques plats, bonne ambiance
LA PIOJERA Face au métro Cal-Y-Canto
Lundi au samedi 10h à 23h, plus ancien café de Santiago et très populaire
QUOI FAIRE CENTRO ARTESANAL SANTA LUCIA Angle O’Higgins et Carmen
Petit marché populaire ouvert tous les jours 10h à 20h30
PLAZA DE ARMAS
Autour de la place, cathédrale, nombreux musées, il s’agit du cœur de la ville, à ne pas manquer
Excursion métallogénique Chili 2012 page 22 AUTRES SUGGESTIONS
Cerro San Cristobal, via le téléphérique pour avoir une vue panoramique sur Santiago Paseo Ahumada, une rue piétonne peuplée de musiciens et de marchands ambulants Palais de la Moneda Musée d'Art précolombien Musée de Santiago National Historique Palacio de Bellas Artes Quartier bella vista, foire d'artisanat, vendredi et samedi après-midi Centre d'Observation des Satellites, construit par la NASA Parc de la Campana Musée d'Art Contemporain Musée National des Beaux Arts
URGENCES POLICE +133
AMBULANCE +131
Excursion métallogénique Chili 2012 page 23
VIGNOBLE CONCHA Y TORO 13 Mai 2012
Horaire : Départ de l’hôtel à 11h00 et retour vers 15h00 Services : Visite + Dégustation
LOCALISATION Situé dans la vallée du Maipo, ce vignoble se trouve à 20km au sud de Santiago dans la région de Pirque. Le domaine Concha y Toro a été fondé en 1883 par Don Melchor.
http://cosmictravel.net/chile/packages/regional-tour-packages/tours/the-chile-wine-route-6-days-and-5-nights.html
DESCRIPTION DE L’ACTIVITÉ La visite du vignoble débute à midi pour une durée d’une heure trente. Après la visite du parc et des caves à vins, une dégustation d’un des meilleurs crus de la maison, soit le Casillero Del Diablo, aura lieu.
RÉFÉRENCE http://www.conchaytoro.com
Excursion métallogénique Chili 2012 page 24
Excursion métallogénique Chili 2012
ANTOFAGASTA
Excursion métallogénique Chili 2012 page 25
LOCALISATION ANTOFAGASTA
Excursion métallogénique Chili 2012 page 26
ANTOFAGASTA Road trip
Départ 15 mai 8h00 Arrivée à l’aéroport El Loa de Calama à 7h30 et prise de possession des voitures à 8h00.
Mine Chuquicamata Cette mine est située à 15km au nord de Calama à 2 870m d’altitude et appartient à Codelco. Elle produit 10 760 tonnes métriques de molybdène par année. Ils produisent également 528 377 tonnes de cathodes à 99,99% de cuivre. Il s’agit de la plus grand mine à ciel ouvert au monde. Visite non disponible en mai en raison de rénovation.
Maria Elena Cette ville fut très prospère à l’époque du salpêtre. Au Chili, le salpêtre (nitrate de sodium) fut très exploité et utilisé dans les années 1920. L’usine pour l’extraction du nitrate de sodium a été inaugurée en 1926. Ce type gisement devint le seul en 1996 qui était en exploitation dans la région d’Atacama. En 2007, un tremblement de terre de magnitude de 7.7 dévasta presque entièrement la ville.
http://www.panoramio.com/photo/24293422
Excursion métallogénique Chili 2012 page 27 Tocopilla Ville se situant à mi-chemin entre Antofogasta et Calama. 24 000 personnes résident à cet endroit. Les activités économiques de la ville sont surtout portées sur l’extraction de salpêtre et sa transformation, ainsi que sur la transformation du cuivre. La pêche y est aussi importante. Ces longues plages attirent les touristes l’été.
Gatico Ancien village abandonné. On peut y apercevoir une seule maison avec le cimetière qui est toujours présent. Paysage magnifique et vue à couper le souffle sur l’océan Pacifique.
http://www.panoramio.com/photo/9355400
Hornitos Situé à 90 km au nord d’Antofagasta, Hornitos est une petite ville côtière. Cette ville balnéaire est très peu habitée et la majorité des résidents ne sont pas permanents en raison des hivers froids et arides.
Mejillones Cette ville portuaire se situe à 60 km au nord d’Antofagasta. Son nom en espagnol signifie moule, qui est d’ailleurs un aliment de base pour les habitants.
Excursion métallogénique Chili 2012 page 28 La Portada Située à 21km au nord d’Antofagasta, La Portada est un énorme rocher sculpté par la mer en forme d’arc de triomphe. Sa base est constituée d'andésite noire autour de laquelle sont arrangées des roches sédimentaires maritimes, une strate de grès jaune et des couches de coquilles fossiles datant de 2 à 35 millions d'années.Il s’agit du symbole de la ville.
http://www.chile365.cl/galeria-fotos-fotos-de-chile-3.php
Mantos Blancos Cette mine à ciel ouvert appartient à Anglo American et se situe à 45 km au nord est d’Antofagasta. Voir article The Mantos Blancos copper deposit: an upper Jurassic breccias style hydrothermal system in the Coastal Range of Northern Chile.
Baquedano Le musé ferroviaire de Baquedano se situe à 72 km d’Antofagasta. Il s’agit plutôt d’un cimetière d’anciennes locomotives à vapeur utilisées à l’époque du salpêtre.
Chacabuco Le site se situe à 106 km au nord d’Antofagasta. Cette ville fantôme était autrefois habitée par plus de 7 000 personnes après la découverte d’un gisement de salpêtre en 1920. Le site fut abandonné en 1940 et nommé monument national en 1971. Ce site a d’ailleurs été utilisé comme camp de prisonniers politiques sous la dictature de Pinochet. Certains édifices sont en restauration dont le théâtre.
Excursion métallogénique Chili 2012 page 29
Excursion métallogénique Chili 2012
SAN PEDRO D’ATACAMA
Excursion métallogénique Chili 2012 page 30
LOCALISATION SAN PEDRO D’ATACAMA
Excursion métallogénique Chili 2012 page 31
SAN PEDRO D’ATACAMA
CONSEILS PRATIQUES
Il existe encore des mines antipersonnelles perdues dans le désert d’Atacama. La plupart ont été retirées, mais soyez vigilant et éviter de sortir des sentiers balisés Il n’y a pas de location de voitures à San Pedro, seulement à Calama Le village est à une altitude d’environ 2 400 m, il faut donc prendre certaines précautions concernant le mal de l’altitude L’eau du robinet est potable, mais contient un taux assez élevé d’arsenic, donc préférer l’eau embouteillée Attention de ne pas gaspiller l’eau ou de prendre des douches longues, l’endroit est l’un des plus secs au monde
URGENCES MÉDICALES ALMACEN FARMACEUTICO LA BOTICA Gustavo Le Paige, 258 +5655-851-169
Pharmacie, tous les jours 10h à 13h30 et 16h30 à 20h, sauf dimanche CONSULTORIO GENERAL RURAL Gustavo Le Paige +5655-851-010
Urgences médicales, si besoins Prendre rendez-vous, tous les jours 8h à 13h et 14h à 17h ADRESSES UTILES CONAF Toconao +5655-851-608
Informations concernant la réserve nationale de Los Flamencos et ses 7 secteurs, leur accès, état des routes, météo et sentiers de randonnée Heures d’ouverture : lundi au vendredi 9h à 13h et 14h30 à 18h30
Excursion métallogénique Chili 2012 page 32 INFOCENTRO Ignacio Carrera Pinto, 547 +5655-851-240
Internet de la bibliothèque donnant un accès wifi gratuit pour 30 minutes Heures d’ouverture : lundi au vendredi 9h à 13h et 14h à 17h45
TAXIS Feria Astesanal +5655-852-105
Compagnie de taxis relativement fiable OÙ DORMIR
NUEVO AMANECER (16 au 22 mai) Los Geiser 161 Alto Jama San Pedro de Atacama, 1410000 Chili +5655-851-34
OÙ MANGER ADOBE Caracoles, 211 +5655-851-132
Service en continu, un peu plus de 14$ Ambiance chaleureuse, privilégié les tapas et quesadillas plutôt que les pizzas, salades et viandes, qui sont supposément trop chers BLANCO Caracoles, 195 +5655-851-164
Tous les jours à partir de 19h, menu à 13$ et plus, resto le plus chic du village Cuisine avec les produits du coin, pâtes fraîches, sushi, escalopes et autres
Excursion métallogénique Chili 2012 page 33 ETNICO Tocopilla, 423 +5655-851-377
Tous les jours 11h à minuit, entre 5,80$ et 7,80$ Tenu par une française, menu peu varié de style bistro LA CASONA Caracoles, 195 +5655-851-004
Même propriétaire que Blanco, donc assez semblable, mais davantage porté sur les grillades et ambiance plus sobre LA ESTAKA Caracoles, 259 +5655-851-201
Tous les jours 9h à 23h30, plus de 13$, ambiance lounge Cuisine de qualité, cannellonis, poissons, viandes, crêpes et autres
LA SOLENA
Ignacio Carrera Pinto, 48
Tous les jours sauf dimanche soir, moins de 7$ Peu de touristes, menu du jour style pot-au-feu, chaleureux
LAS DELICIAS DE CARMEN Calama, 360
Tous les jours 7h30 à 23h, entre 7$ et 14$ Plats traditionnels, empanadas, pizzas, pâtisseries et excellents déjeuners
Excursion métallogénique Chili 2012 page 34 OÙ SORTIR CAFÉ PUB AYLLU Toconao, 479 +5655-592-781
Tous les jours midi à environ 1h Style feux de camp allumés dans une cour avec fauteuils
CAFÉ BAR EXPORT Caracoles, angle Toconao
Un des rares endroits animés le soir, musique forte, alcool et pizzas
QUOI FAIRE RANCHO CACTUS Toconao, 568 +5655-851-506 Rancho-cactus.cl
Centre équestre tenu par une française, les guides parlent un peu l’anglais Environ 30$ pour 2 heures
PISCINE POZO 3
À 4,5 km au sud-est de San Pedro, sur la route de Paso Jama Alimentée en permanence par une source d’eau tiède Tous les jours 8h à 18h30 sauf lundi, entrée 6$
VULCANO EXPEDICIONES Caracoles, 317 +5655-851-023 Vulcanochile.com
Agence offrant soit des descentes de vélo, du sandboard, des ballades à cheval ou encore une excursion près des volcans
Excursion métallogénique Chili 2012 page 35 OBSERVATOIRE 16 Mai 2012
Horaire : Départ de l’hôtel à 19h00 et retour vers 21h30 Services : Transport + Guide + Chocolat chaud
San Pedro de Atacama Celestial Explorations Caracoles, 166 +5655-851-935
http://www.spaceobs.com/fr/tour.php
DESCRIPTION DE L’ACTIVITÉ Le site se trouve à 6 km au sud de San Pedro. SPACE est un observatoire ouvert depuis 2003 par un couple, dont un ingénieur en astronomie.
La tournée comporte trois parties pour une durée d’environ 2h30, transport compris. Première partie: Il s’agit de l’observation du ciel à l'œil nu, à l'aide d'un laser vert pour pointer les étoiles, les constellations sont expliqués, ainsi que la rotation du ciel c’est-à- dire que le ciel change en fonction de la latitude, et ainsi de suite. Deuxième partie: Il s’agit de l’observation avec les dix télescopes mis à votre disposition qui tous ont une orientation différente. Il est maintenant possible d'observer différents types de nébuleuses et amas d’étoiles, avec les explications fournies. Troisième partie: À la fin de la tournée, le partage d'une boisson chaude, habituellement du chocolat chaud est offerte et c'est le temps pour des questions et commentaires.
RÉFÉRENCES http://www.spaceobs.com/fr/tour.php
Excursion métallogénique Chili 2012 page 36 EXCURSIONS SAN PEDRO D’ATACAMA
Excursion métallogénique Chili 2012 page 37 VALLÉE DE LA LUNE 17 Mai 2012
Horaire : Départ 6h30 l’hôtel et retour à San Pedro aux alentours de 12h30 Services compris : Transport + Guide + Petit-déjeuner
LOCALISATION La vallée de la Lune se situe à 13 km à l’ouest de San Pedro d’Atacama. Il s’agit d’une splendide vallée aux allures lunaires possédant de nombreuses formations géologiques sculptées par le vent et l’eau. Cette vallée porte son nom en raison des ses multitudes couleurs et textures ressemblant fortement à la surface de la lune. On peut aussi observer diverses cavernes ainsi que des blocs de sel ayant l’aspect de sculptures.
http://www.valledelalunaonline.com/San%20Juan,%20Valle%20de%20la%20Luna%202.jpg
Excursion métallogénique Chili 2012 page 38 http://www.ranchochago.com http://www.ranchochago.com
DESCRIPTION DE L’EXCURSION L’idée de cette excursion est de vous faire découvrir les énigmatiques vallées du désert d'Atacama au lever du soleil, en commençant par un petit déjeuner sur la Pierre du Coyote, lieu privilégié avec une vue impressionnante sur la Cordillière des Andes. Vous ferez trois belles balades à pied de 30 à 45 minutes chacune. Vous visiterez un canyon de sel, "Las Tres Marias" et vous monterez en haut de la grande dune pour apprécier un panorama à 360 degrés où se découpe une étonnante formation de sel que l'on nomme ici l'Amphithéâtre. Vous visiterez aussi un surprenant canyon, ancien lit d’une rivière disparue. La visite se terminera vers la vallée de la mort. Il est très probable que vous soyez absolument seuls au cours de cette excursion. Dunes, canyons, grottes et cavernes seront les principaux composants de la visite.
http://www.ranchochago.com
RÉFÉRENCES
GLOAGUEN P., DUVAL M., (2012) ; Le guide du routard : Chili et Île de Pâques, Hachette guide tourisme, Octobre 2011, 500p. ISBN 978-2-01245128-5 http://en.wikipedia.org/wiki/Valle_de_la_Luna_(Chile) http://www.ranchochago.com
Excursion métallogénique Chili 2012 page 39 VALLÉE DE L’ARC-EN-CIEL 17 Mai 2012
Horaire : Départ 14h00 l’hôtel et retour à San Pedro aux alentours de 19h30 Services compris : Transport + Guide + Apéritif
LOCALISATION La vallée de l’Arc-en Ciel se situe à seulement quelques kilomètres de San Pedro.
© Laurent Joubert janvier 2006
DESCRIPTION DE L’EXCURSION Vous parcourrez la vallée "hierbas buenas" où vous observerez la plus grande concentration de pétroglyphes de la région. De ces dessins, gravés dans la pierre par les atacameños, peuple originel du désert d'Atacama, nous conterons de leur histoire. La Pachamama, les mythes et légendes chamanes seront expliqués par notre guide spécialisé. Plus loin, à une quinzaine de kilomètres de ce lieu mystique, vous vous promènerez dans la merveilleuse vallée arc-en-ciel. Cette formation géologique aux surprenantes couleurs: rouges, vertes, ocres, jaunes correspond à l'ancienne chaudière d'un volcan disparu depuis des millions d'années. Vous terminerez votre visite par l'authentique petit village de Rio Grande qui a gardé les traces de la colonisation du peuple atacamenien. Vous y verrez les traditionnelles maisons construites en adobe, des cultures en terrasse et une charmante petite église. Avant de reprendre la route, vous dégusterez un bon vin chilien sur la place du village.
RÉFÉRENCE http://www.ranchochago.com
Excursion métallogénique Chili 2012 page 40 LAGUNAS ALTIPLANICAS 18 Mai 2012
Horaire : Départ de l’hôtel à 7h30 du matin et retour à San Pedro vers 17h30 Services compris : Transport + Guide + Dîner + Barbecue
LOCALISATION Le petit village de Socaire se situe à 100 km au sud-est de San Pedro d’Atacama à une altitude de 3 200m. Ce village contient 380 habitants seulement. On peut y observer deux églises en pierre taillée au toit de chaumes et s’entoure de petites parcelles rectangulaires plantées de maïs. Les lagunes Miniques et Miscanti se retrouvent quant à elles à 26 km au sud de Socaire. Situées à 4 000 m d’altitude, on s’y rend à travers les steppes sauvages au pied de volcans enneigés. Et puis soudain, surgissent un, puis deux sublimes lacs couleur bleu lagon, devant les quelques paissent des vigognes. La lagune Chaxa se trouve à un peu plus de 50 km au sud de San Pedro. Au cœur du désert d’Atacama, elle se situe dans l’un des sept secteurs de la Réserva Nacional Los Flamencos tout comme les lagunes Miniques et Miscanti. Trois sortes de flamants sont observables sur le site. Se nourrissant sur le plan d’eau en journée, ils regagnent le soir leur lieu de nidification dans trois autres lacs entièrement préservés de la fréquentation humain. Il est spectaculaire de les voir s’envoler par petits groupes. Sur le site, on peut accéder à un sentier en boucle de 800 m avec panneaux d’interprétation et vidéo en espagnol consacrés au salar. Le village de Tonaco se trouve à 24 km au nord-est de la lagune Chaxa. Quelques 880 habitants vivent dans ce village dans des maisons en pierre blanche volcanique. Sur la place, l’église San Lucas (1970) possède un joli clocher blanc séparé, charpente et escalier en colimaçon sont en bois de cactus.
http://www.photoway.com/voyage/chili/chili-128-photo-socaire-eglise.html http:/www.ranchochago.com
Excursion métallogénique Chili 2012 page 41 DESCRIPTION DE L’EXCURSION Vous sortirez de San Pedro pour vous diriger vers une petite oasis perdue dans le désert où vous prendrez un petit déjeuner. Vous emprunterez une piste de haute montagne qui vous mènera au centre des volcans où se forment les extraordinaires lagunes « Miniques et Miscanti ». Perchées à 4200 mètres d'altitude, ces lagunes portent le nom des volcans qui se reflètent en elles, de couleur bleu turquoise ou vert émeraude, selon l'intensité du soleil. Ces lacs hébergent une faune et une flore unique, une bulle de vie au pied des volcans. Vous continuerez votre route vers la quebrada "Nacimiento", un canyon qui s'est forgé à partir de la lave volcanique du volcan Miniques qui a fait éruption il y a un million d'années. C'est dans ce cadre typique de l'altiplano chilien que vous sera proposé "El asado chilien" : barbecue traditionnel de la cuisine chilienne, composé de viande rouge, viande blanche et chorizos. Puis vous poursuivrez l'excursion jusqu'au petit village de Socaire, qui garde encore l'empreinte du passage du Incas dans cette contrée. Ensuite vous vous dirigerez vers le grand Salar d'Atacama dans le secteur de "Chaxa" : flamants roses, lagunes du désert de sel, autour duquel vous prendrez le temps de vous balader. De retour à San Pedro de Atacama, vous visiterez le village colonial de Toconao.
http://www.entrelacets.fr/index.php/tour-du-monde/42-pendant/142-noel-en-ete
RÉFÉRENCES
GLOAGUEN P., DUVAL M., (2012) ; Le guide du routard : Chili et Île de Pâques, Hachette guide tourisme, Octobre 2011, 500p. ISBN 978-2-01245128-5 http://fr.wikipedia.org/wiki/Laguna_Miscanti http://www.ranchochago.com
Excursion métallogénique Chili 2012 page 42
TREKKING MACHUCA À RIO GRANDE 19 Mai 2012
Horaire : Départ 8h30 l’hôtel et retour à San Pedro aux alentours de 17h30 Services compris : Transport + Guide + Dîner + Ration de marche
LOCALISATION Le petit village de Machuca se situe à un peu de plus de 30 km au nord de San Pedro. Le village de Rio Grande, se situe quant à lui à un peu moins de 10 km à l’ouest de Machuca.
http://www.ranchochago.com/fr/exclusif_fr.html
DESCRIPTION DE L’EXCURSION Cette très belle excursion qui inclut un trekking de 6 heures, accessible à tous, vous mènera au cœur de la cordillère des Andes. Tout le parcours suivra le lit de la rivière Puripika, large vallée verdoyante. Vous traverserez un village fantôme, qui fût le refuge d'anciens éleveurs de lamas. D'ailleurs, durant une grande partie de ce trekking, vous serez entourés par des lamas devenus sauvages. Notre itinéraire vous amènera à traverser des ponts improvisés avec des troncs d'arbres et à vous rafraichir dans les eaux cristallines de la rivière. L'arrivée au petit village de Rio Grande vous permettra d'observer de près des cultures en terrasse aux diverses couleurs. Vous prendrez alors un apéritif à l'ombre des Algarrobos, majestueux arbres du désert.
RÉFÉRENCES http://www.ranchochago.com http://fr.wikipedia.org/wiki/San_Pedro_de_Atacama
Excursion métallogénique Chili 2012 page 43 OJOS DEL SALAR 20 Mai 2012
Horaire : Départ de l’hôtel à 15h00 et retour vers San Pedro à 19h30 Services : Transport + Guide + Cocktail
LOCALISATION La lagune Tebinquinche se situe à 30 km au sud de San Pedro. Juste avant d’y accéder, on peut admirer les Ojos del Salar, qui représentent deux yeux profonds d’eau douce parfaitement circulaire. Les lagunes Cejar sont quant à eux situés à 10 km au nord de la lagune Tebinquinche. Un sentier permet de faire le tour afin d’admirer les concrétions salines et parfois des flamants roses.
http://www.itouringchile.com/fichaLugar.asp?id=101
DESCRIPTION DE L’EXCURSION Cette excursion vous fera découvrir ces mystérieuses formations circulaires qui sont une partie importante de l'écosystème du désert d'Atacama. Ensuite vous vous promènerez sur la croute de sel où se forme la lagune Tebinquinche, véritable miroir naturel de la cordillère des Andes. Vous terminerez cette journée par une baignade dans les eaux ultra- salées des lagunes de Cejar. On y flotte comme dans la mer morte. Vous attendrez le coucher du soleil autour d'un snack.
RÉFÉRENCES
GLOAGUEN P., DUVAL M., (2012) ; Le guide du routard : Chili et Île de Pâques, Hachette guide tourisme, Octobre 2011, 500p. ISBN 978-2-01245128-5 http://www.ranchochago.com
Excursion métallogénique Chili 2012 page 44 TOUR ARCHÉOLOGIQUE 21 Mai 2012
Horaires : Départ depuis votre hôtel à 8h00 et retour à San Pedro de Atacama de 12h30 Services compris : Transport + Guide +Dîner
LOCALISATION Le village de San Pedro est dominé par le volcan du Licancabur (5 916 m d'altitude) et par le volcan de Sairecabur (5 971 m d'altitude). Ces deux volcans sont situés à une trentaine de kilomètres à l'est du village. Le village est à une altitude d'environ 2 400 mètres. Le climat est extrêmement sec et les précipitations n'atteignent que 35 mm par an. On peut y voir le musée Gustavo Le Paige, évoquant l’histoire des populations de l’Atacama, depuis les premiers chasseurs jusqu’au influence tihuanaco (Bolivie), inca et espagnole. Les objets proviennent essentiellement des découvertes du père missionnaire belge Gustavo Le Paige, qui débarqua à San Pedro en 1955 pour servir sa paroisse. À 2,7 km au nord de San Pedro se trouve Pukara de Quitor, des anciennes ruines représentant une forteresse en terrasses dominant San Pedro et construite par les Atacamènes au 12e siècle. Ces bastions émergèrent après la chute de Tihuanaco et l’émergence des pouvoirs locaux. Les conquistadores s’y attaquèrent une première fois en 1536, mais furent repoussés. En 1540, eut une seconde bataille décisive menée par Francisco de Aguirre qui, en guise de victoire, fit décapiter 300 indiens et exposa leurs têtes sur les murs. À 9 km au sud de San Pedro se trouve Aldea de Tulor, un ancien lieu d’habitation où l’on voit les fondations de vieilles maisons circulaires où ont vécu certains Atacamènes entre 800 avant J.-C. et l’an 500. Deux maisons ont été reconstruites pour les visiteurs.
http://louves.canalblog.com/albums/7__chili/photos.html
Excursion métallogénique Chili 2012 page 45 DESCRIPTION DE L’EXCURSION Option réaliser ce tour en bicyclette du village de San Pedro de Atacama. Suivi de la visite du musée Gustavo Le Paige. Pukara de Quitor, site archéologique correspondant à une ancienne forteresse atacamenienne. Aldea de Tulor : Les vestiges d'un village Atacaménien datant de 8000 ans avant JC.
http://www.annuaire-mairie.fr/ville-san-pedro-de-atacama.html
RÉFÉRENCES
GLOAGUEN P., DUVAL M., (2012) ; Le guide du routard : Chili et Île de Pâques, Hachette guide tourisme, Octobre 2011, 500p. ISBN 978-2-01245128-5 http://fr.wikipedia.org/wiki/San_Pedro_de_Atacama http://www.sanpedroatacama.com/ingles/home.htm http://www.ranchochago.com
Excursion métallogénique Chili 2012 page 46 TERMES DE PURITAMA 21 Mai 2012
Horaire : Départ 14h00 l’hôtel et retour à San Pedro aux alentours de 19h30 Services compris : Transport + Guide + Goûter
LOCALISATION Les termes de Puritama se situent à 28 km au nord de San Pedro à 3 500 m d’altitude. Dans un canyon, une rivière d’eau tiède, environ 80-85°F, forme des piscines naturelles au creux des roches et des roseaux. Une passerelle y a été installée ainsi que des sanitaires par un hôtel de luxe de San Pedro. Certains bienfaits médicaux seraient attribués à une baignade dans les termes de Puritama notamment pour contrer le rhumatisme et les douleurs musculaires.
DESCRIPTION DE L’EXCURSION Vous vous relaxerez dans cette merveilleuse oasis aux nombreuses piscines naturelles d’eau thermales.
http://archdoc.mr926.me/tag/chile/page/38/
RÉFÉRENCES http://www.explore-atacama.com/eng/attractions/puritama-hotsprings.htm
Excursion métallogénique Chili 2012 page 47
Excursion métallogénique Chili 2012
RÉFÉRENCES
Excursion métallogénique Chili 2012 page 48
Mineralium Deposita (2003) 38: 787–812 DOI 10.1007/s00126-003-0379-7
ARTICLE
Richard H. Sillitoe Iron oxide-copper-gold deposits: an Andean view
Received: 3 March 2003 / Accepted: 22 July 2003 / Published online: 17 September 2003 Springer-Verlag 2003
Abstract Iron oxide-copper-gold (IOCG) deposits, de- ising the close connection with mafic magmatism. The fined primarily by their elevated magnetite and/or hema- deposits formed in association with sodic, calcic and tite contents, constitute a broad, ill-defined clan related to potassic alteration, either alone or in some combination, a variety of tectono-magmatic settings. The youngest and, reveal evidence of an upward and outward zonation therefore, most readily understandable IOCG belt is lo- from magnetite-actinolite-apatite to specular hematite- cated in the Coastal Cordillera of northern Chile and chlorite-sericite and possess a Cu-Au-Co-Ni-As-Mo-U- southern Peru, where it is part of a volcano-plutonic arc (LREE) (light rare earth element) signature reminiscent of Jurassic through Early Cretaceous age. The arc is of some calcic iron skarns around diorite intrusions. characterised by voluminous tholeiitic to calc-alkaline Scant observations suggest that massive calcite veins plutonic complexes of gabbro through granodiorite and, at shallower palaeodepths, extensive zones of bar- composition and primitive, mantle-derived parentage. ren pyritic feldspar-destructive alteration may be indi- Major arc-parallel fault systems developed in response to cators of concealed IOCG deposits. extension and transtension induced by subduction roll- The balance of evidence strongly supports a genetic back at the retreating convergent margin. The arc crust connection of the central Andean IOCG deposits with was attenuated and subjected to high heat flow. IOCG gabbrodiorite to diorite magmas from which the ore deposits share the arc with massive magnetite deposits, the fluid may have been channelled by major ductile to copper-deficient end-members of the IOCG clan, as well brittle fault systems for several kilometres vertically or as with manto-type copper and small porphyry copper perhaps even laterally. The large, composite IOCG deposits to create a distinctive metallogenic signature. deposits originated by ingress of the ore fluid to rela- The IOCG deposits display close relations to the tively permeable volcano-sedimentary sequences. The plutonic complexes and broadly coeval fault systems. mafic magma may form entire plutons or, alternatively, Based on deposit morphology and dictated in part by may underplate more felsic intrusions, as witnessed by lithological and structural parameters, they can be sep- the ore-related diorite dykes, but in either case the origin arated into several styles: veins, hydrothermal breccias, of the ore fluid at greater, unobserved depths may be replacement mantos, calcic skarns and composite inferred. It is concluded that external ‘basinal’ fluids deposits that combine all or many of the preceding were not a requirement for IOCG formation in the types. The vein deposits tend to be hosted by intrusive central Andes, although metamorphic, seawater, evap- rocks, especially equigranular gabbrodiorite and diorite, oritic or meteoric fluids may have fortuitously contam- whereas the larger, composite deposits (e.g. Candelaria- inated the magmatic ore fluid locally. The proposed Punta del Cobre) occur within volcano-sedimentary se- linkage of central Andean and probably some other quences up to 2 km from pluton contacts and in inti- IOCG deposits to oxidised dioritic magmas may be mate association with major orogen-parallel fault compared with the well-documented dependency of systems. Structurally localised IOCG deposits normally several other magmatic-hydrothermal deposit types on share faults and fractures with pre-mineral mafic dykes, igneous petrochemistry. The affiliation of a spectrum of many of dioritic composition, thereby further emphas- base-metal poor gold-(Bi-W-Mo) deposit styles to rela- tively reduced monzogranite-granodiorite intrusions may be considered as a closely analogous example. Editorial handling: B. Lehmann Keywords Iron oxide-copper-gold deposits Æ R. H. Sillitoe Metallogeny Æ Central Andes Æ Diorite Æ Extensional 27 West Hill Park, Highgate Village, London N6 6ND, UK E-mail: [email protected] tectonics Æ Volcano-plutonic arcs
Excursion métallogénique - Chili 2012 Références page 49 788
One of the best developed, but perhaps rather poorly Introduction appreciated, IOCG provinces is located in the South American Coastal Cordillera and immediately adjoining Iron oxide-copper-gold (IOCG) deposits comprise a areas of northern Chile and southern Peru (latitudes 13– broad and ill-defined clan of mineralization styles which, 33 30¢S; Fig. 1), where it is closely associated with as the name implies, are grouped together chiefly Mesozoic batholiths and major arc-parallel fault sys- because they contain hydrothermal magnetite and/or tems. The origin of IOCG deposits has recently become specular hematite as major accompaniments to chalco- the subject of considerable debate, with both metal- pyrite±bornite (e.g. Ray and Lefebure 2000). Besides bearing magmatic brine (e.g. Hitzman et al. 1992; the copper and by-product gold, the deposits may also Pollard 2000) and external ‘basinal’ brine heated by contain appreciable amounts of Co, U, REE, Mo, Zn, intrusions (e.g. Barton and Johnson 1996; Hitzman Ag and other elements. IOCG deposits currently 2000) being proposed as viable ore-forming fluids. In account for <5 and <1%, respectively, of the world’s view of the fact that the central Andean IOCG province annually mined copper and gold production, much of it is the world’s youngest, is largely unaffected by the derived from Olympic Dam and Ernest Henry in complicating effects of later metamorphism and defor- Australia and Candelaria and Mantoverde in Chile. mation and is relatively well documented geologically, it Notwithstanding their modest economic contributions, provides an ideal example with which to assess the IOCG deposits have become fashionable exploration competing genetic models. Understanding the origin of and research objectives over the past few years.
Fig. 1 Position of the central Andean IOCG belt of northern Chile–southern Peru with respect to the Jurassic–Early Cretaceous magmatic arc and a series of interconnected back- arc basins along its eastern side. Approximate locations of arc segments and intra- and back- arc basins mentioned in the text are shown. Also marked are: two post-Early Cretaceous IOCG deposits located east of the main IOCG belt; axes of the two main belts of ’Kiruna-type’ massive magnetite deposits; the two main concentrations of manto-type copper-(silver) deposits; the area occupied by VHMS deposits; and selected Jurassic and Early Cretaceous porphyry copper-(gold) deposits (1 Tı´a Marı´a; 2 Galenosa-Puntillas; 3 Antucoya-Buey Muerto; 4 Mercedita; 5 Andacollo)
Excursion métallogénique - Chili 2012 Références page 50 789
IOCG deposits is, of course, also fundamental to their and Aguirre 1992), were active during the Mesozoic effective exploration. volcanism and plutonism. Widespread extension This article reviews the geological and metallogenic induced tilting of the volcano-sedimentary sequences. settings of the IOCG province in the Coastal Cordillera Immediately east of the Mesozoic arc terrane of the of Chile and Peru and then the styles and salient features Coastal Cordillera in northern Chile and southern Peru, of the IOCG deposits themselves, with particular sedimentary sequences accumulated in a series of inter- emphasis on smaller, higher-grade deposits as well as the connected, predominantly marine back-arc basins large, better-documented examples like Candelaria and (Mpodozis and Ramos 1990). Mantoverde (Table 1). On balance, the evidence favours Early to mid-Jurassic through mid-Cretaceous vol- a magmatic-hydrothermal origin for central Andean canism and plutonism throughout the Coastal Cordil- IOCG deposits, besides revealing several features lera and immediately adjoining regions are generally and relationships of potential use during deposit- and considered to have taken place under variably exten- district-scale exploration. sional conditions in response to retreating subduction boundaries (slab roll-back) and steep, Mariana-type subduction (Mpodozis and Ramos 1990; Grocott and Geological setting Taylor 2002). Nevertheless, Atherton and Aguirre (1992) questioned the existence of subduction during the Early General features Cretaceous in southern Peru and favoured extension at a passive continental margin. Throughout much of the In the Coastal Cordillera and immediately adjoining Coastal Cordillera of northern Chile and southern Peru, physiographic regions of northern Chile and southern western portions of the Mesozoic arc terrane (and the Peru, major Mesozoic plutonic complexes are emplaced corresponding fore-arc) seem likely to have been into broadly contemporaneous arc and intra-arc volca- removed by subduction erosion or lateral translation nic products and underlying penetratively deformed (Rutland 1971; Dalziel 1986; Mpodozis and Ramos metasedimentary units of Palaeozoic age. Early Prote- 1990) or, at the very least, lie below sea level. rozoic cratonic basement of the Arequipa-Antofalla massif underpins the central segment of the Coastal Cordillera (Shackleton et al. 1979) and the adjoining Volcano-sedimentary rocks Andean Cordillera, between about latitudes 14 and 26 S (Ramos and Aleman 2000). Extensive longitudinal The Middle to Late Jurassic La Negra Formation, up brittle fault systems and/or ductile shear zones, includ- to 5,000–10,000 m of subaerial to locally shallow-sub- ing the Atacama Fault System in northern Chile (e.g. marine basalt, basaltic andesite and andesite lavas, tuffs Scheuber and Andriessen 1990) and deeply penetrating and minor intercalated sedimentary rocks, and correla- faults that localised the Can˜ ete basin in Peru (Atherton tive formations comprise the arc and intra-arc
Table 1 Tonnage and grade of selected IOCG deposits, central Andes
Deposit (Fig. 4) Tonnagea Cu(%) Au(g/t) Ag(g/t) Data source (million tonnes)
Rau´l-Condestable, Peru >25 1.7 0.9 6 de Haller et al. (2002) Eliana, Peru 0.5 2.7 Injoque (2002) Monterrosas, Peru 1.9 1.0–1.2 6 20 Injoque (2002) Mina Justa, Peru 209 0.86 Minor Present Rio Tinto Mining and Exploration Ltd. (unpublished data, 2003) Cobrepampa, Peru 3–5 2–5 Present 15 Injoque (2002) Tocopilla, Chile 2.4 (0.31) 3.1 (16) Present locally Ruiz and Peebles (1988) Montecristo, Chile 15 1.6 0.6 J. Esquivel (personal communication, 2003) Cerro Negro, Chile 249 (49) 0.4 (0.71) 0.15 Atna Resources (press release, 2002) Teresa de Colmo, Chile 70 0.8 Trace Hopper and Correa (2000) Mantoverde, Chile 230 oxide, 0.55 oxide, 0.11 Zamora and Castillo (2001) >400 sulphide 0.52 sulphide Candelaria, Chile 470 0.95 0.22 3.1 Marschik et al. (2000) Punta del Cobre, Chile 120 1.5 0.2–0.6 2–8 Marschik and Fontbote´(2001b) Carrizal Alto, Chile 3 5 Ruiz et al. (1965) Panulcillo, Chile 3 (10.4) 2.7–3.5 (1.45) Up to 0.1 Hopper and Correa (2000) Tamaya, Chile >2 (0.9) 12 (20) Ruiz and Peebles (1988) Los Mantos de Punitaqui, Chile 2 (gold 4 R. Muhr (personal communication, 1998) zone only) El Espino, Chile 30 1.2 0.15 Correa (2003) La Africana, Chile 3.3 2.5 N. Saric (personal communication, 2003) aCumulative production and/or reserves, only approximate for mines active before the 20th century Alternative tonnage and corresponding Cu grade
Excursion métallogénique - Chili 2012 Références page 51 790 successions in northern Chile (Boric et al. 1990; Pic- farther west up to 5,000 m of Early Cretaceous volcanic howiak 1994; Figs. 1 and 2). La Negra lavas overlap the and volcaniclastic sedimentary rocks accumulated in an tholeiitic and calc-alkaline compositional fields (Pic- intra-arc basin formed in response to vigorous extension howiak et al. 1990). The volcanic arc appears to have (Aberg et al. 1984; Mpodozis and Ramos 1990; Ramos been topographically subdued and to have developed 2000). Most of the volcanic rocks range in composition close to sea level (Fig. 2). Late Jurassic to Early Creta- from basalt to andesite and are high-K calc-alkaline to ceous arc volcanism occurred along the eastern side of shoshonitic in composition; parts of the sequence dis- the Coastal Cordillera, at least from latitudes 26–29 S, play compositional bimodality (Levi et al. 1988). where it is represented by up to 3,000 m of basaltic Mesozoic arc rocks in southern Peru include the Rı´o andesite, andesite and dacite volcanic rocks now as- Grande and Chala Formations of mid-Jurassic age, both signed to the Punta del Cobre Group (Lara and Godoy of which comprise basaltic andesite of medium- to high- 1998), host to the Candelaria-Punta del Cobre IOCG K calc-alkaline affinity (Romeuf et al. 1995). The back- district (e.g. Marschik and Fontbote´2001a). arc domain includes the Arequipa basin (Fig. 1) in The Jurassic and Early Cretaceous back-arc domain which up to 1,500 m of Early Jurassic basaltic volcanic between latitudes 21 and 27 S in northern Chile, the rocks belonging to the Chocolate Formation are over- Tarapaca´basin (Fig. 1), is dominated by marine car- lain by several thousand metres of mainly terrigenous, bonate and continental terrigenous sequences, although Middle to Late Jurassic sedimentary rocks (Vicente interbedded andesitic volcanic rocks also occur locally 1990; Sempere et al. 2002a, 2002b). The Can˜ ete basin (Mun˜ oz et al. 1988; Mpodozis and Ramos 1990; Ardill (Fig. 1), the southern portion of the West Peruvian et al. 1998). Evaporite horizons appear locally, especially trough (Wilson 1963), is dominantly Early Cretaceous in in the Late Jurassic (Fig. 2; Boric et al. 1990; Ardill et al. age (Cobbing 1978) and probably best interpreted as a 1998). In the back-arc basin of central Chile (Aconcagua product of advanced intra-arc extension (Ramos and Platform; Fig. 1), south of about latitude 31 30¢S, a Aleman 2000). The Copara´and Quilmana´Formations Jurassic marine carbonate sequence, including a thick in the Can˜ ete basin are dominated by high-K gypsum horizon, is overlain by Late Jurassic continental calc-alkaline to shoshonitic basalt and basaltic andesite, red beds and Early Cretaceous marine carbonates, while although subordinate dacite and rhyolite impart a bi- modal signature to the latter formation (Atherton and Aguirre 1992). These volcanic formations are underlain by a clastic-carbonate succession containing very minor amounts of evaporite minerals (Palacios et al. 1992). The Jurassic and Early Cretaceous arc and intra-arc sequences throughout the Coastal Cordillera are domi- nated by basaltic andesite, appear to possess greater amounts of lava than other volcanic or volcaniclastic products and lack volumetrically important felsic vol- canic rocks. Furthermore, there is little evidence of major volcanic edifices typical of most subduction-related arc terranes, and the volcanic environment may well have been more akin to flood basalt provinces. Low-grade, non-deformative, diastathermal (burial) metamorphism induced by elevated geothermal gradi- ents consequent upon crustal thinning was active during accumulation of all the Mesozoic arc and intra-arc volcanic sequences, with the resulting metamorphic grade commonly attaining the prehnite-pumpellyite facies and, at depth, greenschist facies (Levi et al. 1989; Atherton and Aguirre 1992). This low-grade regional metamorphism is not directly related to pluton emplacement, which gave rise to fairly restricted and easily distinguishable contact aureoles similar to that related to the Tierra Amarilla batholith near the Candelaria IOCG deposit (Tilling 1976; Marschik and Fig. 2 Schematic tectonic sections of the central Andean margin at A latitudes 21–26 S in the Late Jurassic–Early Cretaceous and Fontbote´1996, 2001b). B latitudes 12–14 S in the Early Cretaceous, showing steep subduction at a retreating convergent boundary. Note in A that IOCG deposits occur in a subaerial arc paralleled eastwards by a Plutonic rocks sediment-dominated back-arc basin, whereas in B IOCG deposits occur in a subaqueous intra-arc basin. Approximate ages of evaporites and IOCG deposits (see text) are also shown. Sections The plutonic complexes, ranging in composition from adapted from Ramos and Aleman (2000) primitive early gabbro and diorite through quartz diorite
Excursion métallogénique - Chili 2012 Références page 52 791 and quartz monzodiorite to tonalite and granodiorite In northern Chile and southern Peru, the Jurassic and and, uncommonly, monzogranite were emplaced Early Cretaceous intrusive rocks, most of them horn- throughout the Jurassic and Early Cretaceous as a series blende-bearing, are largely metaluminous and calc- of relatively short pulses, each estimated to last roughly alkaline (Parada 1990; Pichowiak 1994), although early 3 to14 M.Y. where extensively dated between latitudes gabbros are tholeiitic in character (Regan 1985; 25 30¢ and 27 30¢S (Dallmeyer et al. 1996; Lara and Pichowiak et al. 1990). All the intrusive rocks are oxi- Godoy 1998; Grocott and Taylor 2002). Hence, unsur- dised and belong to the magnetite-series (Ishihara and prisingly, multiple ages of any particular intrusive rock Ulriksen 1980). Initial strontium isotope ratios for plu- type occur throughout the Coastal Cordillera; for tonic rocks decrease markedly eastwards across the example, the early gabbros from latitudes 23–24 Sin Coastal Cordillera of northern Chile, in general from northern Chile are dated at 196–185 Ma (Early Jurassic) 0.704–0.705 for the Middle–Late Jurassic to 0.703–0.704 (Pichowiak et al. 1990), whereas those farther north in for the Early Cretaceous rocks (McNutt et al. 1975; Berg the Can˜ ete basin of Peru are clearly assignable to the and Baumann 1985; Pichowiak 1994; Parada et al. mid-Cretaceous (Regan 1985). Plutons are irregular in 1999), a pattern that may be interpreted to imply max- outline but markedly elongate parallel to the orogen, imal extension and crustal thinning and, as a conse- northerly in northern Chile and northwesterly in quence, minimal crustal contamination during the Early southern Peru. Typical plutons exceed 50 km in longi- Cretaceous period. Nevertheless, the mantle wedge re- tudinal dimensions. During the Mesozoic, the locus of mained the main site of magma generation throughout plutonism in northern Chile migrated 50 km or so the Jurassic–Early Cretaceous interval (Rogers and eastwards to reach the eastern border of the Coastal Hawkesworth 1989). Cordillera by the Early Cretaceous (Farrar et al. 1970; The plutonic complexes of the Coastal Cordillera Berg and Baumann 1985; Parada 1990; Dallmeyer et al. between about latitudes 26 and 27 30¢S were emplaced 1996; Lara and Godoy 1998; Fig. 3), and an apparently syntectonically during the Early to Middle Jurassic as similar but still poorly defined progression also took gently inclined, sheet-like bodies up to several kilometres place in southern Peru (Clark et al. 1990). thick, controlled by east-dipping extensional fault sys- Abundant andesite, basaltic andesite and basalt tems (Grocott et al. 1994; Grocott and Taylor 2002), but dykes cut many of the plutons and their host rocks (e.g. thereafter probably as steep, slab-like bodies localised by Pichowiak and Breitkreuz 1984; Regan 1985; Scheuber ductile shear zones (Grocott and Wilson 1997). Roof and Gonzalez 1999; Taylor and Randall 2000; Sempere lifting and floor depression both enabled pluton et al. 2002b). Both synplutonic emplacement features emplacement (Grocott and Taylor 2002). In accord with (Moore and Agar 1985; Regan 1985) and radiometric this intrusion mechanism, the outcropping plutons were dating (Dallmeyer et al. 1996) show that the dykes are emplaced at relatively high crustal levels, <10 km broadly synchronous with host or nearby plutons. between about latitudes 22 and 28 S (Dallmeyer et al. Furthermore, individual dyke swarms tend to be 1996; Scheuber and Gonzalez 1999), and cooled rapidly centred on single plutonic complexes, beyond which as shown by concordance between U-Pb zircon ages and they cannot be traced very far (Taylor and Randall 40Ar/39Ar isotope-correlation ages (Berg and Baumann 2000). 1985; Dallmeyer et al. 1996), as well as by relatively restricted (<4 km) contact-metamorphic aureole deve- lopment. The Coastal Cordillera became amagmatic after 90 Ma, and most of the Late Cretaceous and younger plutonism, including emplacement of the main Coastal Batholith of southern Peru, was restricted to belts far- ther east (Cobbing 1985; Taylor et al. 1998; Grocott and Taylor 2002).
Structural elements
The Atacama Fault System follows the axis of the Coastal Cordillera for >1,000 km between about lati- tudes 20 and 30 S where it is made up of a series of concave-west segments comprising NNW-, N- and NNE-striking ductile and brittle faults which underwent Fig. 3 Generalised spatial and temporal distributions of magmatic variable motion, including sinistral strike-slip (e.g. arc rocks (from Hammerschmidt et al. 1992) and IOCG deposits Herve´1987; Scheuber and Andriessen 1990; Brown et al. (this study) in northern Chile. Note the systematic eastward migration of both the arc and contained mineralization, and the 1993). Transient ductile deformation, charted by marked decline of IOCG mineralization from the Late Cretaceous greenschist and amphibolite facies mylonites (Scheuber onwards and Andriessen 1990; Scheuber et al. 1995), occurred at
Excursion métallogénique - Chili 2012 Références page 53 792 shallow crustal levels (<10 km) in close association with mid-Cretaceous age. In addition to the IOCG deposits Mesozoic pluton emplacement, but gave way to brittle highlighted in this article, ‘Kiruna-type’ massive mag- behaviour during arc cooling (Brown et al. 1993). The netite-(apatite), porphyry copper-(gold), manto-type brittle faults tend to be localised by pre-existing mylonite copper-(silver) and volcanic hosted massive sulphide zones, commonly along pluton margins (Brown et al. (VHMS) zinc-copper-barite deposits are the principal 1993). Fault displacement on the Salado segment of the ore types. Atacama Fault System, between latitudes 25 and 27 S, changed from normal slip to left-lateral transtension at 132 Ma (Grocott and Wilson 1997; Grocott and Magnetite deposits Taylor 2002), as apparently it also did as far north as latitude 22 S (Scheuber and Gonzalez 1999). Neverthe- The massive magnetite deposits occupy the same belt as less, sinistral motion of Jurassic (pre-155 Ma) age has many of the IOCG deposits over a longitudinal distance been interpreted between latitudes 22 and 26 S of nearly 700 km between latitudes 25 and 31 Sin (Scheuber et al. 1995; Scheuber and Gonzalez 1999). The northern Chile and a similar distance in southern Peru, Atacama Fault System is the best documented of three although large deposits there are far more restricted in principal orogen-parallel fault systems in the Coastal latitudinal extent (Fig. 1). Many investigators subscribe Cordillera between latitudes 25 30¢ and 27 S, where it is to a hydrothermal-replacement origin for the magnetite paralleled westwards and eastwards, respectively, by the and minor associated actinolite and apatite (e.g. Ruiz ductile to brittle Tigrillo and Chivato systems (Grocott et al. 1965, 1968) and locally developed clinopyroxene and Taylor 2002; Fig. 5). The three fault systems, in (e.g. Fierro Acarı´, southern Peru; Injoque 2001), concert with the plutonism, young eastwards from although some advocate emplacement mainly as intru- Jurassic–Early Cretaceous in the case of the normal-slip sions and minor extrusions of iron oxide melt (e.g. Tigrillo system to Early Cretaceous for the left-oblique Espinoza 1990; Nystro¨ m and Henrı´quez 1994; Naslund extensional Chivato system. East-side-down displace- et al. 2002). A number of small magnetite deposits occur ment on the fairly shallowly inclined Tigrillo fault ex- as veins in diorite intrusions, which both Me´nard (1995) ceeds 1 km, and only a few kilometres of strike-slip and Injoque (2001) favour as the magmatic-fluid source offset are deduced for the Atacama Fault System for the magnetite deposits in general. Several large (Grocott and Taylor 2002). magnetite deposits, including El Romeral and El In southern Peru, a series of poorly known, orogen- Algarrobo in northern Chile, are steep, lens-like bodies parallel faults exist in the arc terrane, including the within intrusion-bounded screens of Early Cretaceous Can˜ ete intra-arc basin, as well as localising the Arequipa andesitic volcanic rocks along strands of the Atacama back-arc basin. The prominent Treinta Libras fault zone Fault System (Ruiz et al. 1968; Bookstrom 1977). along the eastern margin of the Coastal Cordillera However, the major Marcona deposit in Peru is differ- underwent dextral strike-slip motion in the Jurassic– ent, being a series of strata-bound bodies (mantos) Early Cretaceous, and is marked by a broad dyke swarm replacing principally early Palaeozoic but also Jurassic (Caldas 1978; Injoque et al. 1988). (Rı´o Grande Formation) carbonate horizons west of, During the early Late Cretaceous, transpression rather than within, the major Treinta Libras fault zone triggered by final opening of the Atlantic Ocean basin (Injoque et al. 1988; Injoque 2002). caused tectonic inversion of the formerly extensional Inclusion of the magnetite deposits as end-members back-arc basins (Mpodozis and Ramos 1990; Ladino of the IOCG clan (Hitzman et al. 1992) is supported by et al. 1997). At the same time, the Chivato fault system, the abundance of early-stage magnetite in many IOCG a set of northwest-striking transverse faults throughout deposits, the occurrence of late-stage pyrite, chalcopyrite the Coastal Cordillera and other structural elements and gold in and near some massive magnetite deposits between at least latitudes 18 and 30 S, underwent reac- (e.g. Marcona, El Romeral, Cerro Negro Norte; tivation in the transpressive regime (Taylor et al. 1998; Bookstrom 1977; Injoque et al. 1988; Vivallo et al. 1995) Grocott and Taylor 2002). Positive tectonic inversion and the commonality of certain alteration and gangue also affected southern Peru in the early Late Cretaceous minerals, especially actinolite and apatite, although no- (Benavides-Ca´ceres 1999) and caused demise of the where are the two deposit types observably transitional. Can˜ ete basin (Cobbing 1985). The far more subdued Nevertheless, magnetite veins and lens-like bodies occur deformation in the Coastal Cordillera of northern Chile widely in both Jurassic and Early Cretaceous IOCG vein and southern Peru since the mid-Cretaceous took place districts, including some of the more important ones like in a fore-arc setting. Los Pozos (Mantoverde; Vila et al. 1996), Naguaya´n and Montecristo (Boric et al. 1990). There are also several examples of IOCG deposits located alongside Metallogenic setting major concentrations of massive magnetite (e.g. Mina Justa in the Marcona district; Moody et al. 2003). Thus, The Coastal Cordillera of northern Chile and southern the genetic model for the massive magnetite bodies in the Peru is endowed with iron, copper and subordinate gold, Coastal Cordillera is likely to possess major components silver and zinc resources, all of mainly Early Jurassic to in common with that preferred for the IOCG deposits.
Excursion métallogénique - Chili 2012 Références page 54 793
Porphyry copper deposits Tertiary porphyry copper belts farther east (Fig. 4), and hypogene grades are relatively low (up to 0.4% Cu). Porphyry copper deposits, some relatively enriched in The deposits are related to small stocks of quartz diorite gold, are distributed throughout the Coastal Cordillera to granodiorite porphyry emplaced into arc plutonic or of northern Chile (Fig. 1), where most of them appear to volcanic rocks, and tend to be dominated by potassic be of Early Cretaceous age (135–100 Ma) (Munizaga alteration. et al. 1985; Boric et al. 1990; Perello´et al. 2003). The The porphyry copper and IOCG deposits in the best-known deposit, and the only producer, is Andacollo Coastal Cordillera are readily distinguishable because where a zone of supergene chalcocite enrichment is the potassic alteration and copper-(gold) mineralization exploited; however, several others have been extensively in the former are centred on, and largely confined to, drill tested (e.g. Galenosa-Puntillas, Antucoya-Buey porphyry stocks, which are absent from the latter. Fur- Muerto, Mercedita; Fig. 1). Several prospects also exist thermore, the characteristic quartz veinlets containing all in the Coastal Cordillera of southern Peru (Fig. 1), or part of the chalcopyrite in the porphyry copper where Tı´a Marı´a is likely to be Jurassic in view of deposits as well as the pyrite-dominated, sericite-bor- available radiometric ages for nearby plutonic rocks dered D-type veinlets are also absent from the IOCG (Clark et al. 1990). The deposits are typically much deposits, and the iron oxides that define the IOCG class smaller (<300 million tonnes) than those in the are sparsely represented in the porphyry copper deposits.
Fig. 4 Subdivision of the central Andean IOCG province into western Middle–Late Jurassic and eastern Early Cretaceous belts, showing distribution of different deposit styles discussed in the text. Also marked are axes of Palaeocene– Early Eocene, Late Eocene– Early Oligocene and Late Miocene–Pliocene porphyry copper belts, including locations of principal deposits
Excursion métallogénique - Chili 2012 Références page 55 794
Manto-type copper deposits manifestations of the IOCG type (Vivallo and Henrı´quez 1998; Orrego et al. 2000). However, manto-type deposits Manto-type copper deposits occur as strata-bound are apparently nowhere observed to be directly related or disseminated bodies, as steep hydrothermal breccias transitional to IOCG deposits; while an intimate genetic around barren, finger-like gabbro to diorite plugs and connection cannot be precluded at present, substantive as related veins, mostly within basaltic to andesitic arc geological support is a clear necessity. volcanic sequences of the La Negra Formation between latitudes 22 and 25 S (Fig. 1). However, the largest deposit, Mantos Blancos, is unusual in being VHMS deposits partly hosted by felsic volcanic rocks and plugs (Ramı´rez 1996). Broadly similar copper-silver deposits, Several VHMS deposits of Kuroko type were formed including El Soldado, are widespread in the Early in central and northern Peru during the Early Creta- Cretaceous volcanic and sedimentary rocks of the ceous. The VHMS belt overlaps with the northern Central Chile intra-arc basin (e.g. Fig. 1; Maksaev and recognised limit of the IOCG belt in the Can˜ ete intra- Zentilli 2002). arc basin (Fig. 1; Injoque 2000). The deposits display The highest-grade parts of manto-type deposits, classic massive and stringer types of mineralization and typically controlled by the permeability provided by are particularly noted for their zinc and barite contents faults, hydrothermal breccias, dyke contacts, vesicular (Vidal 1987), although copper besides zinc is important flow tops and flow breccias, are characterised by hypo- at Cerro Lindo, the most southerly deposit (Ly 2000). gene chalcocite and bornite, which grade outwards and These ore textures and metal contents, besides the downwards through chalcopyrite to minor distal con- deficiency of magnetite and hematite, clearly distin- centrations of pyrite. The chalcocite-bornite cores of guish the Peruvian VHMS from central Andean IOCG large deposits commonly abut original redox boundaries deposits. in the host stratigraphic packages and are overlain or flanked by sulphide-deficient zones containing hypogene hematite (Sillitoe 1992; Kirkham 1996). Albite, quartz IOCG deposits and chlorite are the main alteration minerals in the cores of the deposits. Opinion is divided between magmatic- Sites of mineralization hydrothermal (e.g. Holmgren 1987; Wolf et al. 1990) and metamorphogenic (e.g. Sato 1984; Sillitoe 1990, In northern Chile, mainly between latitudes 22 and 31 S, 1992) fluid origins for the manto-type deposits, although most of the IOCG deposits are hosted by the La Negra the latter alternative is favoured by the obvious simi- Formation arc volcanics and their stratigraphic equiva- larities to stratiform, sediment-hosted copper deposits lents farther south as well as by the Late Jurassic and (Kirkham 1996). Nevertheless, emplacement of plutonic Early Cretaceous plutons that intrude them (Table 2). complexes may have been instrumental in causing the Candelaria-Punta del Cobre and some smaller IOCG fluid circulation that resulted in manto-type copper deposits, however, were generated near Early Creta- formation (Maksaev and Zentilli 2002). ceous plutons emplaced near the contact between Late The manto-type deposits comprise a distinctive class Jurassic–Early Cretaceous volcanogenic sequences of copper mineralization uncommon outside the Coastal (Punta del Cobre Group) and Neocomian marine car- Cordillera of northern and central Chile (Sillitoe 1992; bonate sequences. Most IOCG deposits documented Kirkham 1996). Although many manto-type copper from southern Peru, between latitudes 12 30¢ and 14 S, deposits contain albite alteration, calcite and minor are confined to the Can˜ ete intra-arc basin (Fig. 1). The hematite, and some are spatially related to gabbro and copper-bearing Marcona magnetite district, including diorite bodies—features shared with some central An- the Mina Justa IOCG deposit, and several minor mag- dean IOCG deposits (see below)—the manto-type style netite and IOCG deposits farther south pre-date for- appears to be distinguished by its asymmetrical sul- mation of the Can˜ ete basin and occur within the Jurassic phide-oxide zonation and marked deficiency in gold. arc terrane (Fig. 4). Caution is necessary, however, because it will be recalled The latitudinal extent of Mesozoic IOCG deposits in that the breccia-hosted IOCG deposit at Olympic Dam the central Andes is closely comparable with that of in South Australia is also characterised by similar Tertiary porphyry copper deposits (Fig. 4), although asymmetrical sulphide-oxide zonation, with distal pyrite known IOCG deposits are apparently few and relatively giving way through chalcopyrite and bornite-chalcocite minor between latitudes 16 and 22 S where the west- to overlying hematite (Reeve et al. 1990). ernmost part of the IOCG belt may now lie beneath sea Notwithstanding these apparent differences, some level. The Coastal Cordillera IOCG province spans three investigators treat at least selected large manto-type structurally, stratigraphically and metallogenically dis- deposits (e.g. Mantos Blancos) as members of the IOCG tinct tectonic segments of long standing, and is delimited class (Williams 1999; Pollard 2000), include the two de- by fundamental transverse segment boundaries at posit types in a broader manto-type category (Injoque roughly latitudes 13 (the Pisco-Abancay deflection) and 2000) or propose that manto-type deposits are shallow 33 30¢S.
Excursion métallogénique - Chili 2012 Références page 56 Excursion Table 2 Selected geological features of principal IOCG deposits, central Andes
Deposit Host rocks Principal ore Closely related Deposit Deposit style Ore-related alteration Main hypogene Associated Data source(s) (Fig. 4) control intrusive rock(s) age (Ma) opaque minerals metals métallogénique
Rau´l- Andesitic lava, NW, NE faults Diorite body, 115 Veins, mantos, Scapolite, albite, Mg, hem, py, Co, Mo, Zn, Vidal et al. (1990); Condestable tuff, limestone dacite porphyry disseminated actinolite cp, po Pb, As, de Haller et al. dykes bodies LREE (2002) Eliana Gabbrodiorite, Base of sill Gabbrodiorite sill 114–112 Mantos Amphibole, scapolite Mg, py, cp As, Zn, Vidal et al. (1990)
- volcaniclastic Mo, Co
Chili siltstone Monterrosas Gabbrodiorite N70 W fault Gabbrodiorite Vein Actinolite, epidote, Mg, cp, py, po Zn, Co, Vidal et al. (1990)
2012 chlorite, scapolite Mo, Pb Mina Justa Andesitic NE listric normal Andesite porphyry 160–154 Irregular vein-like Actinolite, K-feldspar, Mg, cp, bn, Moody et al. (2003) volcaniclastics, fault dykes, dacite replacement body chlorite, cc, py andesite porphyry dykes clinopyroxene, porphyry sill apatite Cobrepampa Monzonite-diorite NW faults Monzonite- Veins K-feldspar, actinolite, Mg, hem, py, Co, Mo, Injoque (2002) diorite pluton garnet, tourmaline cp, bn Zn, Pb Tocopilla Diorite- N70 E faults Mafic dykes 165 Multiple veins Uncertain Mg, hem, Mo, U, Co, Ruiz et al. (1965) granodiorite and fractures (6 main veins), py, cp Ni, Zn, Sb, local stockworks As Gatico Quartz diorite- N80 W, N70 E Mafic dykes Veins Chlorite Mg, py, cp, bn As, Mo, U, Boric et al. (1990) granodiorite faults and Co, Ni fractures Références Julia Granodiorite N-N10 E fault Diorite-gabbro 164 Veins Chlorite, epidote, Hem, mg, cp, Mo Boric et al. (1990) dyke (30 m wide) albite bn, py Teresa Andesitic volcanics NNW, WNW Diorite body Breccia pipe Albite, chlorite Hem, cp, py Hopper and de Colmo and volcaniclastics faults and dykes Correa (2000) Cerro Negro Andesitic lava NNE faults Diorite pluton Breccia mantos, Sericite, chlorite Hem, mg, cp, Gelcich et al. (1998) and tuff stockworks, veins bn El Salado Andesitic lava NE fault Diorite dykes Veins Biotite, chlorite, Mg, hem, py, Zn Gelcich et al. (1998) sericite, epidote, cp scapolite Las Animas Diorite and N60–90 W faults Microdiorite 162 Veins Actinolite, biotite, Mg, py, cp U, As, Zn Gelcich et al. (1998) metasedimentary and fractures bodies K-feldspar, epidote rocks Mantoverde Andesitic lava N15–20 W fault Diorite dykes 123–117 Vein breccia, K-feldspar, chlorite, Hem, mg, LREE Vila et al. (1996) and volcaniclastics stockworks, sericite py, cp breccia manto Dulcinea Andesitic lava N10 W fault Mafic dyke 65–60 Vein Chlorite, sericite Hem, mg, Mo, Zn, Pb Ruiz et al. (1965) and tuff py, cp Candelaria- Andesitic-basaltic NW, NNW faults Diorite and 116–114 or Mantos, breccias, Biotite, K-feldspar, Mg, hem, cp, Mo, LREE, Marschik and Punta lava and dacite dykes 112–110 veins, stockworks quartz, actinolite/ py, po Zn, As Fontbote´(2001b) del Cobre volcaniclastics chlorite, albite, sericite Carrizal Alto Diorite N50–70 E faults Mafic dykes 150 Veins Chlorite, actinolite, Py, ars, po, Mo, Co, Ruiz et al. (1965) epidote, quartz cp, mg As, U Panulcillo Limestone and NNW fault Diorite intrusion 115 Skarn horizons, Garnet, diopside, Cp, po, sph, As, Mo, Pb, Hopper and andesitic volcanics volcanic-hosted scapolite, amphibole, py, mg, bn Zn, Co, U Correa (2000); lenses albite, K-feldspar, Sugaki et al. (2000) page biotite 795 57 796 (1950); Ruiz et al. (1965) Saric (1978) Data source(s) Hg McAllister et al. Mo, Co Correa (2003) Mo, Ni, As Ruiz et al. (1965) Associated metals cp, py cp, py hem, mg bn, cp, py Hem, mg, Hem, mg, Hem, mg, opaque minerals sphalerite sph pyrite; py Fig. 5 Schematic east–west sections of the Middle–Late Jurassic and Early Cretaceous IOCG belts in northern Chile, showing chlorite, actinolite, sericite Sericite Chlorite, quartz Cp, py, distributions of plutonic, volcanic and sedimentary rocks and main fault systems. Selected IOCG and massive magnetite deposits, pyrrhotite; coded on the basis of deposit style, are projected onto the sections. po Note the eastward migration of plutonic and volcanic rocks and their contained mineralization with time, the structural localisation of some of the deposits and the close spatial association between IOCG and magnetite deposits. Taken with slight modification from Vein Vein Deposit style Ore-related alterationVeins Main hypogene Uncertain
magnetite; Espinoza et al. (1999) and Gelcich et al. (1998), with fault additions schematised from Grocott and Taylor (2002) mg
108 Mantos, veins Albite, epidote, Metallogenic epochs Deposit age (Ma)
hematite; The relatively restricted radiometric age data, provided
hem for gangue (actinolite) or alteration (biotite, sericite) minerals by the K-Ar method unless stated otherwise, suggest that the principal IOCG deposits in northern Chile and southern Peru were mainly generated in diorite intrusion
Closely related intrusive rock(s) Middle–Late Jurassic (170–150 Ma) and Early Creta- chalcopyrite;
cp ceous (130–110 Ma) epochs (Fig. 4), although a few Late Cretaceous and Palaeocene examples are also known (Figs. 1 and 4). The metallogenic epochs W fault Diorite E fault Uncertain
bornite; migrated eastwards in concert with spatially related bn NNE fault Monzodiorite/ N10 N10 Principal ore control plutonic belts (Figs. 3, 4 and 5). The Middle–Late Jurassic deposits are located near the Pacific coast. In northern Chile, they include Tocopilla (165±3 Ma), Guanillos (167±7 Ma), Naguaya´n (153±5 Ma), Montecristo and Julia (164± arsenopyrite; 11 Ma; Boric et al. 1990), Las Animas (162±4 Ma; ars Gelcich et al. 1998) and, based on ages of 150 Ma for sedimentary rocks trachytic volcanics, lutite the host diorite pluton (Moscoso et al. 1982), probably Andesitic volcano- Host rocks Carrizal Alto. In southern Peru, the large Mina Justa deposit and other copper mineralization in the Marcona district (154±4.0 and 160± 4.0 Ma; Injoque et al. (Contd.) 1988) and Rosa Marı´a (ca. 160 or 145 Ma; Clark et al. 1990) are assigned to the same overall metallogenic de Punitaqui Mineral abbrevations: La Africana Diorite Los Mantos El Espino Andesitic volcanics N fault Diorite intrusion Tamaya Andesitic and Table 2 Deposit (Fig. 4) epoch.
Excursion métallogénique - Chili 2012 Références page 58 797
Most of the major IOCG deposits and numerous Tectonic controls smaller examples are located farther east in the Coastal Cordillera and are Early Cretaceous in age (Fig. 4). This Only limited information is available on the detailed epoch includes: Candelaria (116–114 Ma, 40Ar/39Ar and tectonic environments of IOCG formation in the Re-Os; Marschik and Fontbote´2001b; Mathur et al. Coastal Cordillera, although all major deposits were 2002; or 112–110 Ma, 40Ar/39Ar; Ullrich and Clark generated during regional extension or transtension, and 1999; Are´valo et al. 2000; Ullrich et al. 2001), Mant- localised by ductile to brittle faults and fractures of overde (123±3, 121±3 and 117±3 Ma; Vila et al. varied strike (Table 2). At least between latitudes 22 and 1996; Orrego et al. 2000), Galleguillos (121±4 Ma; 27 30¢S, the Late Jurassic deposits were generated in R.H. Sillitoe and M. Orrego, unpublished data, 1999), association with normal fault systems displaying east- Brillador (contiguous pluton dated at 108.5 Ma; side-down displacements, whereas those dated as later Moscoso et al. 1982), Panulcillo (115±3 Ma; R. Ardila than 132 Ma, essentially all the Early Cretaceous in Sugaki et al. 2000) and El Espino (nearby pluton deposits, were localised by sinistral transtensional dated at 108±3 Ma; Rivano and Sepu´lveda 1991) in structures within or related to the Atacama and Chivato northern Chile; and Rau´l-Condestable (116.5–113 Ma, fault systems (Grocott and Wilson 1997; Scheuber and U-Pb on sphene; de Haller et al. 2002; A. de Haller, Gonzalez 1999; Grocott and Taylor 2002), but typically personal communication, 2003) and Eliana (115±5.0 beyond the main north–south splays (Fig. 5). The con- and 113±3.0 Ma; Vidal et al. 1990) in southern Peru. trol of the Mantoverde deposit, for example, has been Additionally, district-wide hydrothermal alteration at interpreted as a strike-slip duplex or side-wall ripout the Productora IOCG occurrence, near latitude 29 S, is (Brown et al. 1993; Taylor et al. 1998) or, alternatively, centred on an albitised diorite intrusion (Ray and Dick as a strike-slip relay ramp breached by the ore-control- 2002) dated at 129.8±0.1 Ma (U-Pb, zircon; Fox 2000), ling Mantoverde fault (C. Bonson in Grocott and Taylor although K-feldspar associated with the IOCG miner- 2002). alization returned an average age of 91 Ma In contrast to the steep attitudes of most IOCG- (40Ar/39Ar; Fox 2000; Fox and Hitzman 2001), proba- controlling faults, a low-angle listric normal fault, bly due to re-setting during subsequent batholith >1 km in down-dip extent but giving rise to little emplacement (G.E. Ray, personal communication, apparent offset, in combination with a series of steep 2003). hanging-wall splays localised the Mina Justa IOCG The largest of the younger IOCG deposits is Dulci- deposit in the Marcona district of southern Peru (Moody nea, situated about 12 km east of the eastern border of et al. 2003). The fault may merge eastwards with the the Coastal Cordillera (Figs. 1 and 4), which is hosted major Treinta Libras strike-slip fault zone (see above). by a diorite-monzodiorite intrusion and andesitic Most, but perhaps not all, of the ductile deformation metavolcanic rocks assigned ages of 65–60 Ma (Iriarte in individual fault systems pre-dated related IOCG et al. 1996). The La Africana deposit, at the southern deposit formation, as clearly observed at Mantoverde extremity of the IOCG belt and also previously mined and elsewhere. Syn-mineralization ductile shearing was, formally, cuts a diorite of presumed Late Cretaceous age however, proposed at the Panulcillo deposit by Hopper (Saric 1978). The El Espino deposit may also be Late and Correa (2000), and is observed to have been active Cretaceous or Palaeocene rather than Early Cretaceous, during early magnetite introduction at several of the as stated above, if the Late Cretaceous–Palaeocene age IOCG deposits. Opinion concerning the tectonic setting favoured for the nearby diorite intrusion by Rivano and of the major Candelaria-Punta del Cobre deposit is Sepu´lveda (1991) is substantiated. A few small and less- divided: Martin et al. (1997) and Are´valo et al. (2000) important IOCG deposits, as well as several small believed that mineralization took place during trans- massive magnetite deposits, of Late Cretaceous or Pal- tension while low-angle ductile shearing was still active aeocene age are also present immediately east of the because of thermal mediation by the nearby plutonic Coastal Cordillera in northern Chile. Minor IOCG vein complex (cf. Dallmeyer et al. 1996; Grocott and Taylor deposits also occur in the coastal batholith of Peru, 2002). Biotite schist is believed to have formed concur- where they are likely to be of Late Cretaceous age (Vidal rently with the early stages of magnetite and chalcopy- 1985). rite precipitation (Are´valo et al. 2000). In stark contrast, The age distribution of IOCG deposits and occur- Marschik and Fontbote´(2001b) and Ullrich et al. (2001) rences in the Coastal Cordillera is certainly more com- proposed a less-likely interpretation that copper-gold plex than the two broad epochs defined above would mineralization post-dated schist formation and coin- suggest, as witnessed by K-Ar ages obtained for cided with initial back-arc basin inversion and con- numerous mineral occurrences, many containing iron comitant uplift. A similar notion, gabbro and diorite oxides, copper and gold, in the Coastal Cordillera be- emplacement and, by association, IOCG generation tween latitudes 26 and 28 S(Dı´az and Vivallo 2001). during initial tectonic inversion, was favoured by Regan Based on the ages, these workers proposed four (1985) and Injoque (2001) for the Can˜ ete basin, although metallogenic epochs: 188–172, 167–153, 141–132 and an extensional setting for the mafic magmatism would 130–98 Ma, which coincide with four eastward-youn- seem to be more reasonable. The relatively minor Late ging plutonic belts of broadly the same ages. Cretaceous and Palaeocene IOCG deposits were formed
Excursion métallogénique - Chili 2012 Références page 59 798 after the early Late Cretaceous tectonic inversion event, replacement mantos besides veins, as observed at Can- during subsequent extensional episodes (e.g. Cornejo delaria-Punta del Cobre (Martin et al. 1997; Marschik and Matthews 2000). and Fontbote´2001b), Mantoverde (Vila et al. 1996; Zamora and Castillo 2001), Cerro Negro and Rau´l- Condestable (Vidal et al. 1990; de Haller et al. 2002). Deposit styles Breccias, both hydrothermal and tectonic in origin, are common components of the composite deposits The IOCG deposits of northern Chile and southern Peru (Table 2), in particular at Mantoverde where they include representatives of most common mineralization comprise the shallower, currently mined parts of the styles, either alone or in varied combinations (Table 2, main fault-controlled vein structure (Vila et al. 1996) Fig. 4). Vein deposits are by far the most abundant, with and the associated Manto Ruso strata-bound deposit many hundreds of them occurring throughout the (Fig. 7; Orrego and Zamora 1991). The predominant Coastal Cordillera belt, especially in northern Chile. breccias at Cerro Negro also comprise strata-bound There, the IOCG veins accounted for Chile’s position as mantos. Composite deposits, including Candelaria- the world’s leading copper producer in the 1860–1870s, Punta del Cobre (Marschik and Fontbote´2001b) and although most of them have not been the focus of Rau´l-Condestable (Vidal et al. 1990), contain bodies of attention over the last 40 years or so because of their dispersed mineralization controlled in part by stratal relatively small size (Table 1) and the fact that many of the mines are severely depleted. The veins, products of both replacement and associated open-space filling, typically occur as swarms of up to 40 occupying areas up to several tens of square kilometres (Fig. 6). The prin- cipal veins are 1–5 km long and 2–30 m wide, with ore shoots worked for at least 500 m down the dip of the veins, and attaining 700 m at Tocopilla and 1,200 m at Dulcinea. In addition to the veins, isolated breccia pipes (Car- rizalillo de las Bombas, Teresa de Colmo) and calcic skarns [San Antonio, Panulcillo, Farola (Las Pintadas)] also occur locally (Fig. 4). The major IOCG deposits, however, are typically composite in style and comprise varied combinations of breccias, stockworked zones and
Fig. 7 Geological sketch of the Los Pozos (Mantoverde) district, showing its confinement to a fault-bounded screen of Jurassic volcanic rocks (La Negra Formation) and tight control by the transtensional Atacama Fault System. Also shown are the two contiguous plutonic complexes and the different styles and relative sizes of IOCG and massive magnetite mineralization that comprise the district, along with their respective radiometric ages. Note the temporal and probable genetic relationship between the Sierra Dieciocho diorite-monzodiorite complex and the Mantoverde Fig. 6 Pluton-hosted IOCG veins and controlling faults in the IOCG deposit based on their age similarities. Map slightly modified Tocopilla district, northern Chile. Principal mined deposits are after Espinoza et al. (1999) and radiometric ages summarised from named. Taken from Boric et al. (1990) sources cited in the text
Excursion métallogénique - Chili 2012 Références page 60 799 permeability provided by fragmental volcanic or volca- niclastic horizons. The large Mina Justa deposit in the Marcona district consists of irregular patches, veinlets and breccia fillings of well-zoned sulphide mineralization within a low-angle fault zone that transgresses the host stratigraphy (Moody et al. 2003). Hornfelsing of volcano-sedimentary host rocks to IOCG deposits is ubiquitous and may have predisposed them to widespread brittle fracturing and consequent permeability enhancement. Typically, however, the thermal effects are difficult to discriminate from meta- somatic products, including widespread and pervasively developed biotite, actinolite, epidote, albite and related minerals. Permeability barriers, especially marbleised or even little-altered carbonate sequences, may have played an important role in the confinement and ponding of hydrothermal fluid in some deposits, such as Candelaria- Punta del Cobre and El Espino (Correa 2003). Never- theless, if fluid penetration is more effective, carbonate rocks may be transformed to skarn and constitute inte- gral parts of some composite deposits (e.g. Rau´l- Condestable; Vidal et al. 1990).
Fig. 8 Spatial relations of IOCG and carbonate-hosted massive magnetite deposits, including the major Candelaria deposit and Intrusion relations related deposits in the Punta del Cobre district, to the Ojancos plutonic complex, in particular the diorite phase. The rest of the In common with many IOCG deposits worldwide (e.g. complex comprises monzodiorite, tonalite and monzogranite. The Ray and Webster 2000), a number of the Andean Farola deposit is a garnet-rich skarn (Ruiz et al. 1965). Although a case may be made for a genetic relationship between the diorite and examples lack clear genetic relations to specific intru- deposits hosted by both volcanic (Punta del Cobre Group) and sions despite being located in close proximity (<2 km) sedimentary (Chan˜ arcillo Group) rocks, it should be noted that the to outcropping plutonic complexes, including early pluton-hosted IOCG veins share faults with pre-ore ‘diabase’ dioritic phases (e.g. Sierra Dieciocho pluton east of dykes, suggesting that the unobserved magmatic source of the dykes may be more closely related genetically to the mineralisation Mantoverde; Fig. 7; Zamora and Castillo 2001; and than the outcropping diorite itself. Compiled from Dı´az et al. Ojancos plutonic complex west of Candelaria; Fig. 8; (1998) and Marschik and Fontbote´(2001b) Marschik and Fontbote´2001b). Mantoverde and Can- delaria-Punta del Cobre are typical examples of deposits where the IOCG mineralization and nearby plutonic (Fig. 4), are hosted by plutons, most of them dioritic in complexes are not observed to be in contact, although composition. Several of the IOCG vein deposits and radiometric dating has shown that the intrusive activity their host intrusions have been shown to possess similar and alteration-mineralization episode overlap tempo- ages, a relationship that is particularly clear at Las rally (Fig. 9). For example, at Mantoverde, K-Ar ages Animas where alteration biotite is dated at 162±4 Ma of 123±3, 121±3 and 117±3 Ma (Vila et al. 1996; (K-Ar) by Gelcich et al. (1998) and the nearby diorite at Orrego et al. 2000) for hydrothermal sericite are 161±4 (K-Ar, biotite), 159.7±1.6 (U-Pb, zircon) and encompassed by U-Pb zircon, whole-rock Rb-Sr iso- 157.6±2.6 Ma (Rb-Sr, whole rock; Fig. 9; Dallmeyer chron, 40Ar/39Ar hornblende and K-Ar ages of 127– et al. 1996). Furthermore, in southern Peru, the 120 Ma for the contiguous Sierra Dieciocho plutonic Monterrosas and Eliana veins are hosted mainly by complex (Figs. 7 and 9; Berg and Baumann 1985; gabbrodiorite (Atkin et al. 1985; Vidal et al. 1990). Dallmeyer et al. 1996; Espinoza et al. 1999). Irrespective of whether host rocks are dioritic or Similarly, preferred ages of 116–114 Ma (Marschik more felsic plutons (e.g. Julia) or their nearby wall rocks and Fontbote 2001b; Mathur et al. 2002) or 112–110 Ma (e.g. Brillador, Tamaya; Table 2), the IOCG veins nor- (Ullrich and Clark 1999; Are´valo et al. 2000; Ullrich mally share localising faults with mafic to intermediate et al. 2001) for copper mineralization at Candelaria fall dykes. They are variously described as andesite, basalt, within the 117.2±1.0- to 110.5±1.7-Ma emplacement dolerite, diabase, diorite, gabbro or simply mafic in span for the contiguous Ojancos plutonic complex composition, and are typically of pre- or syn-ore timing (40Ar/39Ar; Ullrich et al. 2001). (Table 2; Ruiz et al. 1965; Boric et al. 1990; Espinoza In clear contrast, however, most of the principal et al. 1996), but locally mapped as post-ore (e.g. La IOCG vein deposits in northern Chile, such as Toco- Africana; Saric 1978). Additionally, syn- to late-miner- pilla, Gatico, Montecristo, Julia, Las Animas, Ojancos alization diorite dykes occur alongside the volcanic- Nuevo, Carrizal Alto, Quebradita and La Africana hosted Mantoverde vein-breccia deposit (Vila et al.
Excursion métallogénique - Chili 2012 Références page 61 800
Fig. 9 Comparisons of radiometric ages for hydrothermal minerals hydrothermal sphene associated with the IOCG miner- from IOCG deposits and magmatic minerals from host (Las alization (de Haller et al 2002; A. de Haller, personal Animas) or contiguous (Mantoverde, Candelaria) plutonic com- plexes determined using a variety of methods. Note the overall communication, 2003). temporal coincidence of hydrothermal and intrusive events within the error limits of the methods employed. See text for further details. Data compiled from Gelcich et al. (1998), Espinoza et al. Geochemistry and mineralogy (1999), Ullrich and Clark (1999), Orrego et al. (2000), Marschik and Fontbote´(2001b), Ullrich et al. (2001) and Mathur et al. (2002) Long before the iron-oxide-bearing copper deposits of northern Chile were assigned to the IOCG clan, Ruiz 1996; M. Orrego, personal communication, 2002), pre- and Ericksen (1962) and Ruiz et al. (1965) (see also Ruiz and post-breccia diorite dykes are associated with the and Peebles 1988) subdivided them into magnetite- Teresa de Colmo breccia (Correa 2000; Hopper and dominated and (specular) hematite-dominated subtypes. Correa 2000) and diorite dykes are present in the El Most members of their two subtypes are chalcopy- Salado vein district (Browne et al. 2000) and at Punta del rite±bornite-bearing veins, but the hematite-rich sub- Cobre. Typically, the dykes observed by the writer are type also includes the vein breccia at Mantoverde and best described in hand sample as fine- to medium- the veins, breccias and mantos at Punta del Cobre. No grained diorite porphyries. doubt Candelaria would have been assigned to the It is particularly instructive to point out that the magnetite-rich category had it been known at that time! Ojancos Nuevo veins lie within, and the Farola copper Subsequent work has shown that at least some of the skarn abuts, an areally extensive diorite phase of the hematite-rich veins are transitional downwards to the Ojancos plutonic complex, which, as noted above, is magnetite-rich variety (Fig. 10), as observed at Julia <2 km from the Candelaria-Punta del Cobre deposit (Espinoza et al. 1996), Las Animas (Gelcich et al. 1998) and of broadly the same age (Figs. 8 and 9). The and, as a result of recent deep drilling, at both Mant- Panulcillo copper-gold skarn deposit, the southernmost overde (Zamora and Castillo 2001) and El Salado in an 80-km long, north-trending belt of small copper (Browne et al. 2000), in keeping with the generalised skarns extending as far as San Antonio (Fig. 4), lies vertical zonation of IOCG deposits proposed by adjacent to an albitised diorite intrusion (Sugaki et al. Hitzman et al. (1992). A similar upward and outward 2000), as do the breccia mantos at the Cerro Negro change from magnetite to hematite is also documented deposit. Drilling also intersected an albitised diorite at the district scale at Candelaria-Punta del Cobre intrusion containing low-grade chalcopyrite mineraliza- (Marschik and Fontbote´2001b). An appreciable pro- tion about 500 m beneath the Teresa de Colmo chalco- portion of the magnetite in the hematite-rich veins is the pyrite-bearing breccia (Correa 2000; Hopper and Correa mushketovite variety: pseudomorphous after specular 2000). hematite (Ruiz et al. 1965). Late-stage hematite also cuts Notwithstanding the apparently widespread associa- and replaces some of the magnetite. Widespread devel- tion between IOCG deposits and broadly dioritic plu- opment of magnetite after hematite was recently re- tons and minor intrusions, some of them intensely emphasised at Candelaria-Punta del Cobre (Marschik albitised, in the Coastal Cordillera, it should also be and Fontbote´2001b), Rau´l-Condestable (de Haller et al. mentioned that volumetrically minor dacite porphyry 2002) and Mina Justa (Moody et al. 2003). The iron dykes, documented as either syn- or inter-mineral in oxides are typically post-dated by pyrite and copper- timing, occur within the Punta del Cobre (R.H. Sillitoe, bearing sulphides (e.g. Ruiz et al. 1965), although tem- unpublished data, 1992; Marschik and Fontbote´1996; poral overlap is observed locally. Pop et al. 2000), Rau´l-Condestable (de Haller et al. The magnetite-rich veins contain appreciable actino- 2002) and Mina Justa (Moody et al. 2003) deposits. At lite, biotite and quartz, as well as local apatite, clino- Rau´l-Condestable, zircon from the dacite porphyry pyroxene, garnet, hematite and K-feldspar, and possess yields U-Pb ages of 115 Ma, closely similar to that for narrow alteration haloes containing one or more of
Excursion métallogénique - Chili 2012 Références page 62 801
albeit without chalcocite in the central zone, is also de- scribed from Panulcillo (Hopper and Correa 2000). As in many vein deposits, the copper is concentrated in well-defined ore shoots separated by barren or low-grade vein segments. Copper contents, without any influence by supergene processes, tend to diminish in some vein systems at depths of several hundred metres, in response to increasing pyrite (La Africana) or pyrrhotite (Carrizal Alto) contents. Gold contents are higher, but typically undetermined, in the hematite-rich than in the magne- tite-rich deposits (Ruiz et al. 1965). A few of the hematite-rich veins were worked as small, stand-alone gold deposits, including Los Mantos de Punitaqui (Table 2) where the economically dominant metals are uniquely zoned from copper through gold to mercury over a distance of 4.5 km (McAllister et al. 1950; Ruiz et al. 1965). Both IOCG vein subtypes are characterised by highly anomalous amounts of Co, Ni, As, Mo and U (Table 2), as shown by the widespread occurrence of minor amounts of cobaltite, safflorite, danaite (all with Co and As), niccolite, chloanthite (both with Ni and As), molybdenite and uraninite (Ruiz et al. 1965). The Car- rizal Alto veins contain as much as 0.5% Co in places (Ruiz et al. 1965). Arsenic, as arsenopyrite, may also Fig. 10 Idealised section of an IOCG vein in the Coastal Cordillera occur commonly, especially at Tocopilla, and is also showing upward zonation from magnetite to hematite domination, and the possibility of coarse calcite (± silver mineralisation) in its reported at Candelaria-Punta del Cobre (Hopf 1990). top parts and copper-poor massive magnetite at depth. Much of the Cobalt and Mo contents are also anomalously high at magnetite is the mushketovite variety. Hematite zone may display Rau´l-Condestable (Atkin et al. 1985), Candelaria hydrothermal/tectonic brecciation. Note shared fault/fracture (Marschik and Fontbote´2001b) and El Espino (Correa control with pre-vein mafic dyke. Expanded from Espinoza et al. (1996) 2003). Ilmenite is recorded as an ancillary hydrothermal mineral in several deposits, especially in southern Peru (Injoque 2002), although the magnetite from IOCG actinolite, biotite, albite, K-feldspar, epidote, quartz, deposits is typically low in titanium (Hitzman et al. 1992; chlorite, sericite and scapolite (Table 2; Ruiz et al. 1965; G.E. Ray, personal communication, 2003). Minor, typ- Boric et al. 1990; Espinoza et al. 1996; Injoque 2001, ically late-stage Zn and, in some examples, Pb are 2002). In contrast, the hematite-rich veins tend to con- present in several of the vein deposits (e.g. Espinoza et al. tain sericite and/or chlorite, with or without K-feldspar 1996) as well as at Rau´l-Condestable (Vidal et al. 1990; or albite, and to possess alteration haloes characterised de Haller et al. 2002), whereas the anomalously high zinc by these same minerals (Table 2). Tourmaline may be a contents in parts of the Candelaria-Punta del Cobre constituent of either subtype, but is perhaps most com- deposit appear to accompany the final stage of copper mon where hematite is more abundant than magnetite. introduction (N. Pop, personal communication, 1999; Both IOCG subtypes tend to be relatively poor, but by Marschik and Fontbote´2001b). Several hundred parts no means deficient, in quartz, while especially the spec- per million of LREE are reported in parts of the Can- ular hematite-rich variety is commonly associated with delaria-Punta del Cobre deposit and Productora pros- coarse-grained calcite and ankerite, either as early or late pect, at least partly in allanite (Marschik et al. 2000; additions or as a distal equivalent (Fig. 10; Ruiz et al. C. Osterman in Ray and Dick 2002), as well as at Rau´l- 1965). Monomineralic chalcopyrite may be intergrown Condestable (A. de Haller, personal communication, with these carbonate minerals. 2003). Both the magnetite- and specular hematite-rich Alteration related to the large composite deposits is IOCG veins contain chalcopyrite and generally subor- typically complex and rather varied in character dinate pyrite, but in a few cases bornite accompanies the (Table 2). Widespread, early sodic or sodic-calcic alter- chalcopyrite (Table 2). The main Tamaya vein was ation characterised by albite with or without actinolite dominated by bornite to a depth of 400 m (Ruiz et al. occurs in some of the IOCG districts (e.g. Candelaria- 1965). The irregular but broadly vein-like Mina Justa Punta del Cobre; Marschik and Fontbote´1996, 2001b), deposit contains concentrically zoned sulphide assem- but is apparently absent elsewhere (e.g. Mantoverde; blages, with a bornite-chalcocite core grading outwards Vila et al. 1996; Cornejo et al. 2000). Pervasive biotite- through bornite-chalcopyrite and chalcopyrite-pyrite to quartz-magnetite±K-feldspar alteration immediately a broad pyrite halo (Moody et al. 2003). Similar zoning, preceded copper introduction at Candelaria-Punta del
Excursion métallogénique - Chili 2012 Références page 63 802
Cobre, an event associated even more closely with for- basaltic to andesitic sequences were tilted above zones of mation of actinolite (Ullrich and Clark 1999; Are´valo extensional detachment, and subjected to prehnite- et al. 2000; Marschik and Fontbote 2001b). Signifi- pumpellyite and greenschist facies diastathermal (burial) cantly, the same minerals also comprise narrow alter- metamorphism in response to elevated geothermal gra- ation haloes to the IOCG veins within the contiguous dients prior to and possibly also during IOCG ore for- Ojancos plutonic complex (Dı´az et al. 1998). Albite, mation. The IOCG deposits, along with massive chlorite and calcite become predominant in the shal- magnetite, manto-type copper and small porphyry cop- lowest parts of the Punta del Cobre deposit (Marschik per deposits, provide a distinctive metallogenic signature and Fontbote 2001b) as they are in the Teresa de Colmo to the Jurassic and Early Cretaceous Coastal Cordillera breccia pipe (Correa 2000; Hopper and Correa 2000). (cf. Oyarzu´n 1988; Maksaev and Zentilli 2002). Once High-grade mineralization at Mina Justa is intergrown compression, crustal thickening and more evolved with actinolite, clinopyroxene and apatite, and is closely magmas became widespread in response to the early associated with K-feldspar-chlorite-actinolite alteration Late Cretaceous tectonic inversion, IOCG (as well as (Moody et al. 2003). In the Mantoverde vein-breccia, massive magnetite and manto-type copper) deposit for- however, sericite besides K-feldspar and chlorite is clo- mation diminished dramatically in the Late Cretaceous sely associated with copper mineralization, and biotite is and only very locally persisted into the Palaeocene. scarce (Vila et al. 1996; Cornejo et al. 2000). In contrast, In contrast to many IOCG provinces worldwide, at Rau´l-Condestable, potassic alteration is not evident especially those of Precambrian age, the relationship of and early albite, scapolite and a variety of calcic am- the Andean IOCG deposits to intrusive rocks is sub- phiboles are followed by iron oxides, chlorite and seri- stantially clearer. In particular, a number of the deposits cite (Vidal et al. 1990; de Haller et al. 2002). Potassic are hosted by or occur near gabbrodiorite or diorite alteration is also unreported at El Espino where early intrusions. Even where somewhat more felsic plutonic albite is overprinted by epidote, chlorite and lesser phases act as host rocks, broadly contemporaneous amounts of actinolite and sericite (Correa 2003). Pro- diorite dykes commonly share controlling faults with the grade garnet dominates the skarn-type IOCG deposits IOCG veins, implying that relatively primitive magma (Fre´raut and Cuadra 1994) and, at Panulcillo (Table 2), sources existed at depth just before and potentially is observed to be paragenetically equivalent to K-feld- during copper mineralization. Therefore there is strong spar-albite-quartz and biotite-magnetite assemblages in suggestion of an intimate connection between relatively contiguous andesitic volcanic rocks (Hopper and Correa primitive, poorly fractionated and little-contaminated 2000). gabbrodiorite to diorite magmas and the IOCG deposits (Table 2; Fig. 11). In this regard, it should be remarked that Marschik and Fontbote´(1996; but not 2001b) Metallogenic model linked the Candelaria-Punta del Cobre IOCG deposit to nearby diorite of the Ojancos plutonic complex, which, Geological synthesis of the Coastal Cordillera IOCG province in northern Chile and southern Peru at re- gional, district and deposit scales enables construction of a preliminary metallogenic model.
Regional- and district-scale aspects
Most of the IOCG deposits were generated during the early development of the ensialic Andean orogen, when the crust was variably extended and attenuated and unusually hot, and magmatism was relatively primitive. IOCG formation took place during both extensional and transtensional tectonic regimes. The greatest number of IOCG deposits, including some of the largest, were generated during the Early Cretaceous when crustal attenuation attained a maximum. The deposits are controlled principally by brittle faults, although ductile deformation locally overlapped Fig. 11 Cartoon of Jurassic IOCG vein in the La Negra arc of with the early stages of mineralization. The voluminous northern Chile to show possible sources of ore fluid. Vertically tholeiitic to calc-alkaline intrusions that either host or ascendant magmatic fluid supplied from an unobserved diorite occur in proximity to the IOCG deposits possess a magma source at depth is preferred. See text for further discussion. Note Triassic evaporites are very restricted in both volume and dominantly mantle source, lack appreciable crustal extent, whereas those in both the overlying intra-arc basin and contamination and are oxidised, in common with their back-arc basin are too young to have contributed to the ore fluid, thick volcanic-dominated host-rock sequences. These being deposited after IOCG formation
Excursion métallogénique - Chili 2012 Références page 64 803 as noted above, itself hosts IOCG veins and skarns. Similarly, the IOCG deposits in the Can˜ ete basin have been related genetically to gabbrodiorite intrusions (Vidal et al. 1990; Injoque 2001, 2002). Furthermore, it should also be recalled that Me´nard (1995) concluded that the massive magnetite deposits within the Atacama Fault System are also genetically related to diorite intrusions. Notwithstanding the compelling evidence in favour of a gabbrodiorite to diorite intrusive source for the IOCG ore fluids, it should be re-emphasised that very minor volumes of dacitic magma, in addition to the more mafic melt, were also clearly available during copper mineralization in the case of at least three of the deposits (Candelaria-Punta del Cobre, Rau´l-Condesta- Fig. 12 Sulphur isotope values for sulphide minerals, mainly ble, Mina Justa), although not necessarily sourced from chalcopyrite and pyrite, from selected IOCG deposits and the same part of the parental magma chamber as the prospects in northern Chile and southern Peru. Data taken from metalliferous fluid. Vivallo and Henrı´quez (1998), Fox (2000), I. Ledlie in Hopper and Correa (2000), Injoque (2001), Marschik and Fontbote´(2001b) and This inferred genetic association with relatively mafic de Haller et al. (2002). The markedly high and low values at Rau´l- plutonism would nicely explain the Cu-Au-Co-Ni-As- Condestable are attributed to reduction of evaporitic or seawater Mo-U signature, bearing in mind that a similar metal sulphate (Ripley and Ohmoto 1977; de Haller et al. 2002) and suite, albeit subeconomic with respect to copper, cha- biogenic sulphur (de Haller et al. 2002), respectively, whereas values for the other deposits suggest a dominance of magmatic racterises some calcic iron skarns associated with dioritic sulphur intrusions (Einaudi et al. 1981; Meinert 1992; Ray and Lefebure 2000). The Larap magnetite deposit in south- eastern Luzon island, the Philippines, part of a Neogene deposit, which cuts an evaporite sequence (Correa 2000; island arc, provides an instructive example. The mag- Hopper and Correa 2000). Rau´l-Condestable presents netite skarn, developed from both carbonate and non- an apparent exception, however, with both highly posi- carbonate lithologies, is part of a low-grade porphyry tive and highly negative d34S values being interpreted in copper system related to diorite porphyry intrusions, terms of the involvement of seawater or evaporitic sul- and is enriched in Co, Ni and U, besides Cu, Mo and Au phur and biogenic sulphur, respectively (Fig. 12; Ripley (Sillitoe and Gappe 1984). Wang and Williams (2001) and Ohmoto 1977; de Haller et al. 2002). reported a similar Cu-Au-Ni-Co-Te-Se suite from the In porphyry copper deposits, metal-bearing, mag- Mount Elliott skarn deposit in the Cloncurry IOCG matic-hydrothermal fluid is channelled upwards from district of Queensland, Australia. parent magma chambers via steep, typically cylindrical Magmatic-hydrothermal provision of copper and, to porphyry stocks, within and around which much of the a lesser degree, gold in the Coastal Cordillera province is copper and gold are eventually concentrated in re- unequivocally confirmed locally by the existence of the sponse to declining fluid temperature. Alteration and Mesozoic porphyry copper-(gold) deposits in associa- mineralization are, therefore, relatively confined, al- tion with volumetrically restricted albeit somewhat more though zones of >5 km2 may be affected by potassic felsic porphyry stocks. Indeed, the extensional, as op- alteration in giant systems (e.g. El Teniente, Chile; posed to compressive, stress regime prevalent during Skewes et al. 2002). In the case of large composite porphyry copper formation in the Coastal Cordillera is IOCG deposits, such confinement of alteration and believed to be a major factor responsible for the small mineralization is not so apparent, especially where so- sizes and low hypogene grades of these deposits (cf. dic-calcic alteration either presages or accompanies the Sillitoe 1998; Tosdal and Richards 2001). Therefore, the copper mineralization. Alteration in the Candelaria- widely exposed, deeper plutonic complexes, from the Punta del Cobre district, for example, occupies tops of which porphyry copper stocks may have already >30 km2 (Marschik and Fontbote 2001b). Neverthe- been eroded, may also reasonably be expected to have less, the alteration associated with simple IOCG veins, had the capacity to generate broadly similar magmatic- breccia pipes and skarns is generally just as volumet- hydrothermal fluids for IOCG genesis (cf. Oyarzu´n rically restricted as that with non-IOCG deposits of 1988). Available sulphur isotopic results for several of these types. the IOCG deposits fall in a fairly narrow range centred The tendency for alteration and mineralization to be around 0 per mil (Fig. 12; Fox 2000; I. Ledlie in Hopper unusually widespread in many IOCG districts, especially and Correa 2000; Marschik and Fontbote´2001b), en- in association with large composite IOCG deposits, may tirely consistent with a largely magmatic source for the be ascribed to the existence of magmatic-hydrothermal sulphide sulphur; however, an origin by leaching from fluid sources at considerable depth within either the host Mesozoic igneous rocks cannot be entirely ruled out. or contiguous plutonic complexes. Fluid ascent on ap- Sulphur isotopic values consistent with a magmatic proach to ore-forming levels appears to be guided by origin (Fig. 12) even characterise the Teresa de Colmo second- and lower-order splays of the major localising
Excursion métallogénique - Chili 2012 Références page 65 804 fault zones, intrusive contacts and permeable strati- genetic model involving such external fluids is perhaps graphic horizons, and therefore may not be as tightly not unreasonable for IOCG deposits within the Can˜ ete focused as in most porphyry copper deposits. The pre- basin (e.g. Rau´l-Condestable) and along the eastern cise locations of the source intrusions remain to be edge of the Coastal Cordillera (e.g. Teresa de Colmo) clarified, although these could potentially be at consid- and, hence, close to the back-arc environment, it seems erable palaeodepths, perhaps as great as 10 km, given a far less likely possibility for the majority of the inferred depths of pluton emplacement and the associ- deposits occurring within the volcanic arc itself. It ation of the IOCG deposits with crustal-scale ductile to would seem to be especially convoluted to invoke ba- brittle fault zones (e.g. Grocott and Wilson 1997). In the sinal brine access as a means of forming the IOCG case of the vertically extensive veins and other deposit veins, many of which are sealed within sizeable plu- styles hosted by gabbrodiorite and diorite, it may rea- tonic complexes and possess original depth extents of sonably be presumed that the mineralizing fluids were >1 km (Fig. 11). Most of these veins were formed exsolved during final consolidation of the deep, com- immediately following emplacement of their host plu- positionally similar portions of the plutons (Fig. 11). tons (Fig. 9), clearly while they were still hot and However, where diorite dykes and the IOCG veins share therefore even less likely to permit the ingress of controlling faults cutting more felsic plutons, derivation external brine. of both the dyke magma and metalliferous fluid from Provision of brine from a back-arc sedimentary basin deeper, more mafic and less-fractionated parts of the by means of gravity-induced flow is precluded by pal- plutonic complexes may be inferred. Replenishment of aeo-topographic considerations, given that an at least magma chambers by more primitive mantle melts could partially subaerial arc must be higher in elevation than a result in underplating of plutonic complexes by more marine back-arc basin. The extensional setting also mafic material as well as acting as a potential trigger for precludes tectonically induced brine expulsion at the liberation of sulphur- and metal-charged fluid, in the times when most of the IOCG deposits were generated. manner proposed recently by Hattori and Keith (2001). The only other alternative, crustal-scale convection Hypothetically, IOCG ore fluids might be supplied by (Fig. 11; Barton and Johnson 1996, 2000), also seems mafic intrusive phases anywhere within host or nearby implausible, especially in the case of the Middle to Late plutonic complexes, which range from several to perhaps Jurassic La Negra arc, because fluid circulation across 10 km in vertical extent (e.g. Grocott and Taylor 2002), an 50-km width of a pluton-dominated arc would need assuming that they were efficiently tapped by steep, to be invoked. Furthermore, some of the oldest (latest through-going faults. Such a deep origin for IOCG flu- Middle Jurassic) IOCG veins in northern Chile were ids in the central Andes accords with a postulated deeper generated at least 7 M.Y. before evaporite formation at source of magmatic fluids in IOCG than in porphyry <155 Ma in the adjoining back-arc basin (Fig. 2; copper deposits, inferred from higher CO2 contents Ardill et al. 1998). Finally, it is important to note that (Pollard 2001), in keeping with those documented in the Mesozoic IOCG belts of the central Andes span 19 fluid inclusions from Candelaria (Ullrich and Clark of latitude, within and alongside only parts of which 1999). The elevated geothermal gradients that existed in evaporites are documented. the extensional Mesozoic arc terranes of the Coastal Even less likely is an evaporite or formational brine Cordillera would have favoured prolonged ascent and source within or beneath the Mesozoic plutonic com- even lateral flow of the deeply derived magmatic fluid plexes of the IOCG-bearing La Negra arc in northern before cooling was sufficient to cause wholesale metal Chile. Triassic sedimentary sequences locally beneath precipitation. the arc and the thin Jurassic sedimentary intercalations within it are preserved only discontinuously and are volumetrically minor (Fig. 11). Moreover, the only Consideration of alternative fluid sources known potential brine sources would appear to be the extremely limited sabkha facies described locally as part A non-magmatic origin for IOCG ore fluids and their of Triassic rift sequences, but such material would be contained metals in the Andean province gains little restricted to the region between approximately latitudes support from the overall geological settings of many of 24 and 27 S (Sua´rez and Bell 1992, 1994). Palaeozoic the deposits. Hitzman (2000) and Kirkham (2001) fa- meta-sedimentary sequences and older crystalline base- voured chloride-rich basinal brine produced by evap- ment are the most common substrate to the plutonic orite dissolution as perhaps the most likely fluid for complexes and clearly could not have acted as brine copper and gold transport and IOCG deposit forma- sources during the Mesozoic (Fig. 11). Indeed, the host tion in the Coastal Cordillera of northern Chile. plutons for the Las Animas, Carrizal Alto and Que- Evaporites, albeit predominantly sulphates rather than bradita vein deposits directly intrude the metasedimen- halite, are preserved in the Tarapaca´and Aconcagua tary rocks (Ruiz et al. 1965). back-arc basins, as noted above (Fig. 11; Mun˜ oz et al. Descent of brine from overlying sources, proposed in 1988; Mpodozis and Ramos 1990; Ardill et al. 1998) some shield areas (Gleeson et al. 2000) and elsewhere and occur in minor amounts at depth in the Can˜ ete (Haynes 2000), might be invoked as a means of gener- intra-arc basin (Palacios et al. 1992). Although a ating the IOCG deposits within the arc, but widespread
Excursion métallogénique - Chili 2012 Références page 66 805 descent of fluid for at least 1 km down pluton-hosted Deposit-scale aspects faults to generate veins at palaeodepths as great as 5 km or more seems highly improbable (Fig. 11). Indeed, the Pluton-hosted IOCG deposits in the Coastal Cordillera, widespread pseudomorphing of specular hematite by chiefly veins, tend to be localised by minor faults and magnetite, suggestive of thermally prograding hydro- fractures and to be relatively small in size, albeit of thermal systems, not to mention formational tempera- appreciable horizontal and vertical extents (Fig. 13). tures of >500 C for the magnetite (e.g. Marschik and Nevertheless, the large vein districts, like Tocopilla, may Fontbote´2001b), would seem to be more easily expli- be areally extensive (Fig. 6). In contrast, IOCG deposits cable in terms of ascent rather than descent of the ore in volcanic and sedimentary host rocks to plutons are fluid, in keeping with more conventional concepts of vein formation. Furthermore, there are no known Mesozoic sedimentary accumulations that are either capable of copious brine generation or sufficiently widespread to have overlain the numerous IOCG deposits in the Mesozoic arc of northern Chile, bearing in mind that significant copper mineralization was active during both Middle–Late Jurassic and Early Cretaceous epochs. Moreover, the well-known evaporite occurrence in the restricted Coloso basin, at latitude 23 50¢S, is appreciably younger than nearby IOCG deposits formed in the Middle–Late Jurassic epoch (Flint and Turner 1988). A perhaps more reasonable non-magmatic ore fluid would be metamorphic brine of the type believed by some investigators to have been responsible for the manto-type copper deposits (see above). Generation of metamorphic fluid accompanied subsidence of the intra- arc basins transgressed by the central Andean IOCG belts (Aguirre et al. 1999), and such a fluid has been considered as a possible, even the sole, contributor to the IOCG deposits in the Can˜ ete basin of southern Peru (Vidal et al. 1990; Injoque 2000), and perhaps also to those in the Coastal Cordillera of northern Chile (Hitzman 2000). However, ingress of metamorphic fluid to large cooling plutons to generate the IOCG veins confronts some of the same difficulties as those consid- ered above for other externally derived brines. Heated seawater is another fluid that may have been available during pluton emplacement and IOCG for- mation in the intra-arc basin environment, and has been proposed at Rau´l-Condestable in the northern part of Fig. 13 Schematised styles of IOCG deposits in the Coastal the Can˜ ete basin on the basis of the sulphur isotopic Cordillera of the central Andes. Note the fundamental control values (Ripley and Ohmoto 1977; de Haller et al. 2002). imposed by faults, commonly shared with pre-ore mafic (basaltic Certainly, seawater may be inferred to have played a key andesite/diorite) dykes. Large deposits are composite, in the sense of comprising several closely spaced mineralization styles, and role in VHMS formation only slightly later in the same localised by zones of high structural and lithological permeability, part of the Can˜ ete basin (Vidal 1987). possibly confined beneath carbonate or other lithologically On the basis of this discussion of possible external determined aquitards. Vein breccias (and breccia mantos and fluid sources, it is concluded that evaporitic, meta- pipes) tend to occur at relatively shallow palaeodepths and, hence, are typically confined to volcanogenic wallrocks. There is an morphogenic and seawater brines all seem unlikely to upward change in the predominant hydrothermal iron oxide from have been solely responsible for the genesis of the magnetite to specular hematite. The IOCG system may be IOCG deposits in the Coastal Cordillera, although concealed beneath an extensive zone of barren feldspar-destructive their local involvement, along with that of locally de- alteration containing pyrite. A deeply derived magmatic fluid guided upwards along the dyke-filled faults, and possibly sourced rived meteoric water, remains feasible. Given that a from the same magmatic reservoir as the dyke rock itself, is single fluid type rather than different or blended fluids hypothesised. Note that telescoping of alteration types (e.g. would seem to be required to explain the common potassic over sodic-calcic) and mineralisation styles, especially characteristics of the IOCG deposits throughout the within composite deposits, is a possibility locally; however, the 1,700-km-long Coastal Cordillera belt, the magmatic- phenomenon is considered to be far less widely developed in the low-relief extensional arcs of the Coastal Cordillera than it is in the hydrothermal model discussed above is believed to gain highly uplifted compressive arcs that host the Tertiary porphyry further support. copper deposits farther east
Excursion métallogénique - Chili 2012 Références page 67 806 variable in size, but include all the largest deposits. The copper-(gold) deposits (e.g. Sillitoe 2000). Moreover, if largest host-rock deposits appear to be those where this comparison between the upward transition from fault-guided magmatic-hydrothermal fluid permeates sodic-calcic to potassic alteration in some IOCG and one or more porous stratigraphic horizons (Fig. 13), porphyry copper deposits is valid, then similar fluid which in the case of Candelaria span a 350-m-thick rock evolutions, perhaps controlled by declining temperature, package (Ryan et al. 1995; Marschik and Fontbote´ might be invoked in both cases. Zoning in IOCG 2001b). Low-angle fault or shear zones may also deposits of the Coastal Cordillera is still poorly docu- enhance syn-mineralization permeability. Fluid ponding mented, although observations from several vein dis- beneath aquitards, such as marbleised carbonate tricts and Candelaria-Punta del Cobre show that sequences, may also favour the formation of large magnetite-actinolite-apatite is transitional upwards to composite deposits (Fig. 13). Known copper-gold skarn hematite-chlorite-sericite at both the individual vein and deposits in the Coastal Cordillera are small, but clearly district scales (Figs. 10 and 13). Sizeable hydrothermal an integral component of the IOCG spectrum, thereby breccia veins, pipes and mantos appear to be largely rendering redundant any discussion of the generic dif- restricted to this shallower, hematite-dominated IOCG ference between these skarns and other IOCG deposits zone (Fig. 13), where fluid overpressures may develop in the belt. This assertion is amply supported by obser- more readily. vations at Panulcillo, where Hopper and Correa (2000) The close association of magnetite-dominated IOCG charted the equivalence of garnet skarn and potassic and massive magnetite-(apatite) veins containing minor assemblages developed in contiguous andesitic volcanic copper in several districts, especially but not limited to rocks. those of Middle–Upper Jurassic age, may be taken to Marschik and Fontbote´(1996) considered the Punta suggest that the two deposit types are transitional and, del Cobre IOCG deposit to be intermediate in overall furthermore, that copper contents of IOCG veins may style between massive magnetite and porphyry copper decrease downwards, giving rise to massive magnetite deposits, both of which occur fairly close by in the veins (Fig. 10; cf. Espinoza et al. 1999; Naslund et al. Coastal Cordillera (Fig. 1). However, IOCG and por- 2002; Ray and Dick 2002). The same relationship is also phyry copper deposits, as discussed above, are clearly favoured by the tendency of IOCG mineralization to distinct and apparently not directly related; nevertheless, occur alongside some massive magnetite deposits, per- they may display certain features in common, including haps suggestive of a crude zonal relationship (e.g. Mina occurrence of hydrothermal magnetite and/or hematite Justa, Mantoverde). The deeper massive magnetite and potassic, potassic-calcic and/or sodic-calcic alter- bodies, with or without copper, lack hydrothermal bio- ation (cf. Pollard 2000; Lang and Thompson 2001). tite and K-feldspar and are accompanied by sodic-calcic Many gold-rich porphyry copper deposits worldwide alteration, in keeping with the conclusions of several contain abundant hydrothermal magnetite±hematite as previous workers (Hitzman et al. 1992; Pollard 2000; a component of both early, barren sodic-calcic and later, Ray and Dick 2002). The district- and deposit-scale ore-related potassic-(calcic) alteration assemblages (e.g. geological evidence, especially the intimate association Sillitoe 2000). Magnetite contents at Grasberg, for between IOCG and massive magnetite deposits in parts example, attain 15 vol% in parts of the potassic of the Coastal Cordillera, does not support radically alteration zone (MacDonald and Arnold 1994), not different fluid sources for the two deposit types, as re- volumetrically dissimilar to some IOCG deposits. cently proposed on the basis of differences in their Furthermore, in a few porphyry copper deposits, sodic- 187Os/188Os ratios (Mathur et al. 2002). calcic alteration, defined by sodic plagioclase, clinopy- Hydrothermal magnetite in porphyry copper deposits roxene, amphibole and magnetite, rather than the more is normally considered to result from precipitation of normal potassic assemblages directly hosts all or part of iron partitioned directly from the source magma into the copper-gold mineralization (Sillitoe 2000), especially magmatic-hydrothermal brine (e.g. Arancibia and Clark in the case of deposits in the Intermontane belt of British 1996). However, at least part of the iron present as iron Columbia, Canada (Lang et al. 1995). Moreover, two of oxides in some of the massive magnetite (Ruiz et al. the deposits in the Intermontane belt (Afton and Ajax) 1968; Me´nard 1995) and IOCG (Cornejo et al. 2000) constitute late stages of the Iron Mask batholith, which deposits of the Coastal Cordillera may have resulted happens to contain magnetite-apatite veins like those in from leaching by hot hypersaline magmatic fluid of the Coastal Cordillera belt (Cann and Godwin 1983; ferromagnesian minerals in igneous rocks adjoining the Snyder and Russell 1995). sites of mineralization. Zones of mafic-poor, albite- Interestingly, copper mineralization in several central K-feldspar-altered rocks developed in the vicinities of Andean IOCG deposits (e.g. Rau´l-Condestable, El many large massive magnetite and some IOCG deposits, Espino) is exclusively present in zones of sodic-calcic including Candelaria (Marschik and Fontbote´2001b) alteration, although potassic alteration or combinations and Mantoverde (Cornejo et al. 2000), provide the of this with sodic-calcic alteration phases are more supporting evidence. typical hosts. This situation may be a product of vertical The upward extensions of IOCG deposits are even less alteration zoning, not just radically different fluid well known than their roots, although there is limited chemistries, as it is in the case of at least some porphyry observational evidence for occurrence of coarse-grained
Excursion métallogénique - Chili 2012 Références page 68 807 calcite veins (Fig. 10), even immediately above large minor copper skarn occurrences may represent composite deposits like Candelaria-Punta del Cobre. hanging-wall leakage anomalies (Fig. 13). The pos- Small copper-gold skarns located above the Candelaria sible relationship of calcic skarns to IOCG deposits deposit (Ryan et al. 1995) are probably more a reflection should not be overlooked. of the carbonate protolith than uppermost manifesta- 6. Broad, strongly developed contact-metamorphic tions of the entire Candelaria-Punta del Cobre district. (hornfels) and metasomatic (sodic-calcic and/or Ray and Dick (2002) concluded that a 1.5-km-wide, potassic alteration) aureoles to gabbrodiorite or down-faulted block of massive silicified tuff containing diorite intrusions are a favourable indicator for large pyrite, sericite and minor dumortierite represents the composite IOCG deposits. shallowest alteration facies at the Productora IOCG 7. Intense and pervasive hydrothermal alteration is a prospect. This proposal would be in keeping with the prerequisite for large, composite IOCG deposits, al- widespread occurrence of extensive zones of pyritic though the copper-gold mineralization may be feldspar-destructive alteration affecting volcanic se- accompanied by potassic, potassic-calcic or sodic- quences locally throughout the Coastal Cordillera, some calcic assemblages. of them in proximity to IOCG districts. Silicification 8. Mineralized hydrothermal breccia and the predomi- accompanied by sericitic and/or advanced argillic alter- nance of specular hematite over magnetite both sug- ation is commonly recorded. gest relatively shallow palaeodepths and, hence, persistence of IOCG potential at depth (Fig. 13). By the same token, widespread development of magne- Exploration consequences tite and actinolite indicate fairly deep levels in IOCG systems, with less likelihood of encountering eco- nomic copper-gold contents at appreciable depth. The preceding discussion highlights several geological 9. Some, but by no means all, composite IOCG deposits features and relationships of possible use in IOCG have irregularly and asymmetrically developed pyrite exploration in the Coastal Cordillera of the central haloes that may provide useful vectors to ore. Andes and, potentially, in similar extensional environ- 10. Coarsely crystalline calcite or ankerite veins may be ments elsewhere: either the tops or distal manifestations of IOCG 1. Middle–Late Jurassic and Early Cretaceous plutonic deposits. belts in the Coastal Cordillera are more prospective 11. Speculatively, extensive zones of barren feldspar- for IOCG deposits than the younger magmatic arcs destructive alteration, including silicification, seri- farther east. The latter coincide with the principal cite, pyrite and even advanced argillic assemblages, porphyry copper belts of the central Andes (Fig. 4), within volcano-sedimentary sequences may either thereby underlining an inverse correlation between conceal underlying IOCG deposits or intimate their major IOCG and porphyry copper deposits. presence nearby. In essence, such zones are litho- 2. Large IOCG deposits seem more likely to form caps, comparable to those well documented from within major orogen-parallel, ductile to brittle fault the porphyry copper environment (e.g. Sillitoe systems that underwent extension or transtension 2000). than in association with either minor or compres- 12. The distal fringes and immediate surroundings of sional fault structures. massive magnetite deposits may be prospective for 3. Receptive rock packages cut by gabbrodiorite, diorite IOCG deposits if suitable structural preparation and or more felsic plutons containing IOCG veins or volcano-sedimentary host rocks are present. bordered by skarns may be especially prospective for 13. Notwithstanding point 12, districts dominated by large composite IOCG deposits. The intrusive rocks massive magnetite bodies or veins may imply rela- are likely to display at least localised zones of weakly tively deep erosion levels unfavourable for major developed potassic-(calcic) and/or sodic-calcic alter- IOCG deposit preservation. ation. 4. Fragmental volcanic or volcaniclastic host rocks characterised by high intrinsic and/or structurally Concluding remarks imposed permeability favour the formation of large composite IOCG deposits if suitable progenitor This review of the central Andean IOCG province intrusions and deeply penetrating feeder faults are concludes that the most likely ore fluid is of magmatic present. High- or low-angle faults or shears may parentage, although inadvertent participation of non- create the structural permeability. magmatic fluids, of the types generated during low-grade 5. Relatively impermeable rocks, such as massive mar- diastathermal (burial) metamorphism, seawater circula- bleised carbonate units, may be conducive to fluid tion or evaporite dissolution, cannot be ruled out locally ponding and the consequent development of imme- and, indeed, have been proposed at Candelaria-Punta diately subjacent IOCG deposits (Fig. 13). Such del Cobre (Ullrich et al. 2001) and Rau´l-Condestable impermeable units may even still conceal IOCG (Ripley and Ohmoto 1977; Vidal et al. 1990; de Haller deposits and, as at Candelaria (Ryan et al. 1995), et al. 2002). Metals, with the possible exception of some
Excursion métallogénique - Chili 2012 Références page 69 808 of the iron, are also thought most likely to have been subduction-related arcs and intraplate settings, in an provided directly by the same magmatic source, for analogous manner to formation of lithophile-element which a primitive gabbrodiorite to diorite composition enriched gold deposits both along the landward sides of at appreciable depths beneath the deposit sites is pre- Cordilleran arcs and in collisional settings (Thompson ferred. It is salutary to recall that Buddington (1933) et al. 1999). proposed this same genetic relationship between diorite intrusions and veins rich in magnetite, Cu, Au, Co and Acknowledgements This article is an expanded version of a keynote Ni. No evidence for involvement of alkaline magmas, as address presented at the 11th Quadrennial IAGOD Symposium and Geocongress 2002 in Windhoek, Namibia. The Organising implied for IOCG deposits in general by Groves and Committee, and especially its chairman, Roy Miller, are thanked Vielreicher (2001), is present in the central Andean for the invitation to attend and, indirectly, for the impetus to province. prepare this review. Thanks are due to the many companies and Notwithstanding this fundamental genetic conclu- geologists with whom I have had the pleasure of working on IOCG deposits and prospects in Chile, Peru and elsewhere over the last sion, the disparate and poorly defined nature of the three decades. Special acknowledgement is also due to the late IOCG deposit clan does not necessarily imply that all Carlos Ruiz Fuller and his colleagues, and their several generations other iron oxide-rich copper-gold deposits worldwide of successors, at the Chilean Geological Survey (Servicio Nacional are generated in the same or even a similar manner. de Geologı´a y Minerı´a, formerly Instituto de Investigaciones Indeed, a non-magmatic brine origin for some IOCG Geolo´gicas) for pioneering studies of regional geology and IOCG deposits in the Coastal Cordillera. The manuscript was improved provinces, as advocated by Barton and Johnson (1996, as a result of reviews by Constantino Mpodozis, Pepe Perello´, 2000), Haynes (2000) and others, may remain a possi- Gerry Ray, John Thompson and, on behalf of Mineralium Depo- bility. Nevertheless, although the IOCG deposit class is sita, Lluı´s Fontbote´and Peter Pollard. too all-encompassing as presently defined, the obvious similarities between several large IOCG deposits, including Candelaria-Punta del Cobre in Chile, Sossego References in the Caraja´s district, Brazil and Ernest Henry in the Cloncurry district of Queensland, Australia (e.g. Mark Aberg G, Aguirre L, Levi B, Nystro¨ m JO (1984) Spreading-subsi- et al. 2000; Leveille and Marschik 2001), may be taken dence and generation of ensialic marginal basins: an example from the early Cretaceous of central Chile. In: Kokelaar BP, to suggest that broadly similar, probably pluton-related Howells MF (eds) Marginal basin geology: volcanic and asso- hydrothermal systems were operative periodically from ciated sedimentary and tectonic processes in modern and an- the Archaean to the Mesozoic. Indeed, diorite and/or cient marginal basins. Geol Soc Lond Spec Publ 16:185–193 gabbro also abut both the Sossego and Ernest Henry Aguirre L, Fe´raud G, Morata D, Vergara M, Robinson D (1999) Time interval between volcanism and burial metamorphism and orebodies, although current wisdom would consider it to rate of basin subsidence in a Cretaceous Andean extensional possess no direct genetic connection with, and to pre- setting. Tectonophysics 313:433–447 date, the copper-gold introduction. In all three cases, the Arancibia ON, Clark AH (1996) Early magnetite-amphibole-pla- plutonism and deeply penetrating fault zones proximal gioclase alteration-mineralization in the Island Copper por- to ore are products of regional extension, in either phyry copper-gold-molybdenum deposit, British Columbia. Econ Geol 91:402–438 subduction-related arc or intracontinental rift settings Ardill J, Flint S, Chong G, Wilke H (1998) Sequence stratigraphy (cf. Hitzman 2000). of the Mesozoic Domeyko Basin, northern Chile. J Geol Soc If the genetic relationship to oxidised, primitive, Lond 155:71–88 40 39 dioritic magmatism proposed herein for the Coastal Are´valo C, Grocott J, Pringle M, Martin W (2000) Edad Ar/ Ar de la mineralizacio´n en el yacimiento Candelaria, Regio´nde Cordillera province and, possibly, some deposits else- Atacama. Actas 9th Congr Geol Chileno 2:92–96 where is correct, then at least a selection of deposits Atherton MP, Aguirre L (1992) Thermal and geotectonic setting of assigned to the IOCG class would seem to constitute an Cretaceous volcanic rocks near Ica, Peru, in relation to Andean extended clan in the same manner as proposed by crustal thinning. J S Am Earth Sci 5:47–69 Atkin BP, Injoque-Espinoza JL, Harvey PK (1985) Cu-Fe amphi- Thompson et al. (1999) and Lang et al. (2000) for base bole mineralization in the Arequipa segment. In: Pitcher WS, metal-poor, lithophile element (Bi-W-Mo)-enriched Atherton MP, Cobbing EJ, Beckinsale RD (eds) Magmatism gold deposits in association with highly fractionated at a plate edge. The Peruvian Andes. Blackie, Glasgow, and relatively reduced felsic intrusions. In both cases, a pp 261–270 broad range of deposit styles, including pluton-hosted Barton MD, Johnson DA (1996) Evaporitic source model for igneous-related Fe oxide-(REE-Cu-Au-U) mineralization. veins, skarns, breccias and replacement mantos, is Geology 24:259-262 evident. The even broader spectrum of copper and gold Barton MD, Johnson DA (2000) Alternative brine sources for Fe- deposits linked to alkaline magmatism (Jensen and oxide (-Cu-Au) systems: Implications for hydrothermal alter- Barton 2000; Sillitoe 2002) may be cited as yet another ation and metals. In: Porter TM (ed) Hydrothermal iron oxide copper-gold and related deposits: A global perspective. Aus- example of the same fundamental metallogenic influ- tralian Mineral Foundation, Adelaide, pp 43–60 ence exerted by magma type, as detailed by Blevin and Benavides-Ca´ceres V (1999) Orogenic evolution of the Peruvian Chappell (1992) and others for common mineralization Andes: The Andean cycle. In: Skinner BJ (ed) Geology and types like porphyry copper and tin-tungsten deposits. ore deposits of the central Andes. Soc Econ Geol Spec Publ 7:61–107 Acceptance of this proposal might provide an expla- Berg K, Baumann A (1985) Plutonic and metasedimentary rocks nation for the occurrence of IOCG deposits in associ- from the Coastal Range of northern Chile: Rb-Sr and U-Pb ation with petrochemically similar intrusions in both isotopic systematics. Earth Planet Sci Lett 75:101–115
Excursion métallogénique - Chili 2012 Références page 70 ©2010 Society of Economic Geologists, Inc. Economic Geology, v. 105, pp. 3–41
Porphyry Copper Systems*
RICHARD H. SILLITOE† 27 West Hill Park, Highgate Village, London N6 6ND, England
Abstract Porphyry Cu systems host some of the most widely distributed mineralization types at convergent plate boundaries, including porphyry deposits centered on intrusions; skarn, carbonate-replacement, and sediment- hosted Au deposits in increasingly peripheral locations; and superjacent high- and intermediate-sulfidation epi- thermal deposits. The systems commonly define linear belts, some many hundreds of kilometers long, as well as occurring less commonly in apparent isolation. The systems are closely related to underlying composite plutons, at paleodepths of 5 to 15 km, which represent the supply chambers for the magmas and fluids that formed the vertically elongate (>3 km) stocks or dike swarms and associated mineralization. The plutons may erupt volcanic rocks, but generally prior to initiation of the systems. Commonly, several discrete stocks are emplaced in and above the pluton roof zones, resulting in either clusters or structurally controlled alignments of porphyry Cu systems. The rheology and composition of the host rocks may strongly influence the size, grade, and type of mineralization generated in porphyry Cu systems. Individual systems have life spans of ~100,000 to several mil- lion years, whereas deposit clusters or alignments as well as entire belts may remain active for 10 m.y. or longer. The alteration and mineralization in porphyry Cu systems, occupying many cubic kilometers of rock, are zoned outward from the stocks or dike swarms, which typically comprise several generations of intermediate to felsic porphyry intrusions. Porphyry Cu ± Au ± Mo deposits are centered on the intrusions, whereas car- bonate wall rocks commonly host proximal Cu-Au skarns, less common distal Zn-Pb and/or Au skarns, and, beyond the skarn front, carbonate-replacement Cu and/or Zn-Pb-Ag ± Au deposits, and/or sediment-hosted (distal-disseminated) Au deposits. Peripheral mineralization is less conspicuous in noncarbonate wall rocks but may include base metal- or Au-bearing veins and mantos. High-sulfidation epithermal deposits may occur in lithocaps above porphyry Cu deposits, where massive sulfide lodes tend to develop in deeper feeder structures and Au ± Ag-rich, disseminated deposits within the uppermost 500 m or so. Less commonly, intermediate- sulfidation epithermal mineralization, chiefly veins, may develop on the peripheries of the lithocaps. The alteration-mineralization in the porphyry Cu deposits is zoned upward from barren, early sodic-calcic through potentially ore-grade potassic, chlorite-sericite, and sericitic, to advanced argillic, the last of these constituting the lithocaps, which may attain >1 km in thickness if unaffected by significant erosion. Low sulfidation-state chalcopyrite ± bornite assemblages are characteristic of potassic zones, whereas higher sulfidation-state sul- fides are generated progressively upward in concert with temperature decline and the concomitant greater degrees of hydrolytic alteration, culminating in pyrite ± enargite ± covellite in the shallow parts of the litho- caps. The porphyry Cu mineralization occurs in a distinctive sequence of quartz-bearing veinlets as well as in disseminated form in the altered rock between them. Magmatic-hydrothermal breccias may form during por- phyry intrusion, with some of them containing high-grade mineralization because of their intrinsic permeabil- ity. In contrast, most phreatomagmatic breccias, constituting maar-diatreme systems, are poorly mineralized at both the porphyry Cu and lithocap levels, mainly because many of them formed late in the evolution of systems. Porphyry Cu systems are initiated by injection of oxidized magma saturated with S- and metal-rich, aqueous fluids from cupolas on the tops of the subjacent parental plutons. The sequence of alteration-mineralization events charted above is principally a consequence of progressive rock and fluid cooling, from >700° to <250°C, caused by solidification of the underlying parental plutons and downward propagation of the lithostatic- hydrostatic transition. Once the plutonic magmas stagnate, the high-temperature, generally two-phase hyper- saline liquid and vapor responsible for the potassic alteration and contained mineralization at depth and early overlying advanced argillic alteration, respectively, gives way, at <350°C, to a single-phase, low- to moderate- salinity liquid that causes the sericite-chlorite and sericitic alteration and associated mineralization. This same liquid also causes mineralization of the peripheral parts of systems, including the overlying lithocaps. The pro- gressive thermal decline of the systems combined with synmineral paleosurface degradation results in the char- acteristic overprinting (telescoping) and partial to total reconstitution of older by younger alteration-mineral- ization types. Meteoric water is not required for formation of this alteration-mineralization sequence although its late ingress is commonplace. Many features of porphyry Cu systems at all scales need to be taken into account during planning and exe- cution of base and precious metal exploration programs in magmatic arc settings. At the regional and district scales, the occurrence of many deposits in belts, within which clusters and alignments are prominent, is a pow- erful exploration concept once one or more systems are known. At the deposit scale, particularly in the por- phyry Cu environment, early-formed features commonly, but by no means always, give rise to the best ore- bodies. Late-stage alteration overprints may cause partial depletion or complete removal of Cu and Au, but metal concentration may also result. Recognition of single ore deposit types, whether economic or not, in por- phyry Cu systems may be directly employed in combination with alteration and metal zoning concepts to
† E-mail, [email protected] *An Invited Paper
Submitted: April 15, 2009 0361-0128/10/3863/3-39 3 Accepted: November 18, 2009
Excursion métallogénique - Chili 2012 Références page 71 4 RICHARD H. SILLITOE
search for other related deposit types, although not all those permitted by the model are likely to be present in most systems. Erosion level is a cogent control on the deposit types that may be preserved and, by the same token, on those that may be anticipated at depth. The most distal deposit types at all levels of the systems tend to be visually the most subtle, which may result in their being missed due to overshadowing by more promi- nent alteration-mineralization.
Introduction intermediate-sulfidation epithermal Au ± Ag ± Cu orebodies (<1 Mt−>1 Gt). Porphyry Cu systems were generated world- PORPHYRY Cu systems are defined as large volumes (10−>100 km3) of hydrothermally altered rock centered on porphyry Cu wide since the Archean, although Meso-Cenozoic examples are most abundantly preserved (e.g., Singer et al., 2008; Fig. stocks that may also contain skarn, carbonate-replacement, 1), probably because younger arc terranes are normally the sediment-hosted, and high- and intermediate-sulfidation epi- least eroded (e.g., Seedorff et al., 2005; Kesler and Wilkinson, thermal base and precious metal mineralization. Along with 2006; Wilkinson and Kesler, 2009). calc-alkaline batholiths and volcanic chains, they are the hall- Porphyry Cu systems presently supply nearly three-quar- marks of magmatic arcs constructed above active subduction ters of the world’s Cu, half the Mo, perhaps one-fifth of the zones at convergent plate margins (Sillitoe, 1972; Richards, Au, most of the Re, and minor amounts of other metals (Ag, 2003), although a minority of such systems occupies postcol- Pd, Te, Se, Bi, Zn, and Pb). The systems also contain major lisional and other tectonic settings that develop after subduc- resources of these metals as well as including the world’s tion ceases (e.g., Richards, 2009). The deeper parts of por- largest known exploitable concentrations of Cu (203 Mt: Los phyry Cu systems may contain porphyry Cu ± Mo ± Au Bronces-Río Blanco, central Chile; A.J. Wilson, writ. com- deposits of various sizes (<10 million metric tons [Mt]-10 bil- mun., 2009) and Mo (2.5 Mt: El Teniente, central Chile; lion metric tons [Gt]) as well as Cu, Au, and/or Zn skarns (<1 Camus, 2003), and the second largest of Au (129 Moz: Gras- Mt−>1 Gt), whereas their shallower parts may host high- and berg, including contiguous skarn, Indonesia; J. MacPherson,
Galore Creek
Pebble Mt Polley Recsk Almalyk Oyu Tolgoi Highland Valley dist. Rosia Poieni Butte Kounrad Island Copper Bingham Majdanpek Tintic Yerington Reko Diq Lepanto & Guinaoang Cananea Saindak (Mankayan dist.) Copper Canyon Santo TomasII & Chelopech Dizon Nugget Hill Pueblo Viejo Sepon Ray Globe- Boyongan-Bayugo Miami Sar Cheshmeh Mineral Park Morenci Tampakan Mamut Grasberg-Ertsberg Resolution Santa Rita (Superior dist.) (Central dist.) Cerro de Pasco Ok Tedi & Colquijirca Panguna Cerro Colorado Sierrita- Bau Esperanza Yanacocha Choquelimpie Collahuasi dist. Batu Hijau Red Antamina El Abra Mountain Esperanza Bisbee Cabang Kiri Cotabambas Escondida & Chuquicamata dist. Koloula Chimborazo Frieda River dist.(Nena) Toquepala & Gaby Cuajone Wafi-Golpu El Salvador Los Pelambres Taca Taca Bajo Los Bronces- Bajo de la Alumbrera (Farallón Negro dist.) & Agua Rica Río Blanco Potrerillos Northparkes Cadia Nevados del Famatina Marte & El Teniente Caspiche Pascua-Lama & Veladero Andacollo
Principal metals Deposit type Age Cu-Mo Porphyry Bau Miocene-Pleistocene Cu-Mo-Au Porphyry + major skarn/ Gaby Eocene-Oligocene Cu-Au carbonate replacement Ray Late Cretaceous-Paleocene Ag-Pb-Zn-Cu High-sulfidation Bisbee Late Triassic-Early Cretaceous epithermal porphyry± No porphyry known OyuT olgoi Late Devonian-Carboniferous Cadia Ordovician
FIG. 1. Worldwide locations of porphyry Cu systems cited as examples of features discussed in the text along with five ad- ditional giant examples. The principal deposit type(s), contained metals, and age are also indicated. Data mainly from sources cited in the text.
0361-0128/98/000/000-00 $6.00 4
Excursion métallogénique - Chili 2012 Références page 72 PORPHYRY COPPER SYSTEMS 5 writ. commun., 2009). Typical hypogene porphyry Cu de- 72° PERU 68° posits have average grades of 0.5 to 1.5 percent Cu, <0.01 to 18° 0.04 percent Mo, and 0.0× to 1.5 g/t Au, although a few “Au- only” deposits have Au tenors of 0.9 to 1.5 g/t but little Cu ARICA (<0.1 %). The Cu and, in places, Au contents of skarns are typically higher still. In contrast, large high-sulfidation epi- BOLIVIA thermal deposits average 1 to 3 g/t Au but have only minor or no recoverable Cu, commonly as a result of supergene removal. This field-oriented article reviews the geology of porphyry Collahuasi Cu systems at regional, district, and deposit scales. The resul- tant geologic model is then used as the basis for a brief syn- thesis of porphyry Cu genesis and discussion of exploration guidelines. The deposits and prospects used as examples El Abra Chuquicamata throughout the text are located and further characterized in 22° Figure 1. The economically important results of supergene CALAMA-E LT OR oxidation and enrichment in porphyry Cu systems have been addressed elsewhere (Sillitoe, 2005, and references therein).
Regional- and District-Scale Characteristics O Belts and provinces ANTOFAGASTA Porphyry Cu systems show a marked tendency to occur in Escondida linear, typically orogen-parallel belts, which range from a few AR CH tens to hundreds and even thousands of kilometers long, as IBAR exemplified by the Andes of western South America (Sillitoe CA and Perelló, 2005; Fig. 2) and the Apuseni-Banat-Timok- Srednogorie belt of Romania, Serbia, and Bulgaria (Jankovi´c, 26° 1977; Popov et al., 2002). Deposit densities commonly attain El Salvador Potrerillos 15 per 100,000 km2 of exposed permissive terrane (Singer et al., 2005). Each belt corresponds to a magmatic arc of broadly similar overall dimensions. One or more subparallel belts COPIAPÓ constitute porphyry Cu or epithermal Au provinces, several of which give rise to global-scale anomalies for Cu (e.g., north- ern Chile-southern Peru, southwestern North America) or Au (northern Peru; Sillitoe, 2008). Notwithstanding the ubiquity ARGENTINA of porphyry Cu belts, major deposits may also occur in isola- tion or at least as distant outliers of coherent belts and provinces (e.g., Pebble in Alaska, Butte in Montana, and Bingham in Utah; Sillitoe, 2008; Fig. 1). Pueblo Viejo in the LA SERENA Dominican Republic (Fig. 1) is the best example of a major, 30° isolated high-sulfidation epithermal Au deposit, albeit with no currently known porphyry Cu counterpart. Porphyry Cu belts developed during well-defined metallo- Lineament genic epochs, which isotopic dating shows to have typical du- rations of 10 to 20 m.y. Each porphyry Cu epoch is closely Fault linked to a time-equivalent magmatic event. Again, the Andes (Sillitoe and Perelló, 2005), southwestern North America (Ti- Porphyry Cu deposit tley, 1993; Barra et al., 2005), and Apuseni-Banat-Timok- Srednogorie belt (Zimmerman et al., 2008) provide prime ex- 100km amples. Individual porphyry Cu belts are commonly spatially SANTIAGO separate rather than superimposed on one another, reflecting arc migration as a result of steepening or shallowing of sub- ducted slabs between the individual magmatic-metallogenic FIG. 2. A preeminent example of spatial and temporal coincidence be- epochs (e.g., Sillitoe and Perelló, 2005). The processes of sub- tween a porphyry Cu belt and an intra-arc fault zone: the northern Chile part duction erosion and terrane accretion at convergent margins of the central Andean middle Eocene to early Oligocene porphyry Cu belt may assist with land- or trenchward migration of the arcs and and Domeyko fault system (summarized from Sillitoe and Perelló, 2005). The apparent termination of the belt in northernmost Chile is a result of con- contained porphyry Cu belts (e.g., von Huene and Scholl, cealment beneath Miocene volcanic rocks. Approximate positions of the 1991; Kay et al., 2005). Nevertheless, several temporally main arc-transverse lineaments in northern Chile are also shown (after Sal- discrete porphyry Cu-bearing arcs may be superimposed on fity, 1985, in Richards et al., 2001).
0361-0128/98/000/000-00 $6.00 5
Excursion métallogénique - Chili 2012 Références page 73 6 RICHARD H. SILLITOE one another: five since ~45 Ma in the Chagai belt, Pakistan much wider (160 km) Texas lineament of southwestern North (Perelló et al., 2008). America (Schmitt, 1966) being oft-quoted examples. These transverse features, possibly reflecting underlying basement Tectonic settings structures, may facilitate ascent of the relatively small magma Porphyry Cu systems are generated mainly in magmatic arc volumes involved in porphyry Cu systems (e.g., Clark, 1993; (including backarc) environments subjected to a spectrum of Richards, 2000). regional-scale stress regimes, apparently ranging from mod- erately extensional through oblique slip to contractional (Tos- Deposit clusters and alignments dal and Richards, 2001). Strongly extensional settings, typi- At the district scale, porphyry Cu systems and their con- fied by compositionally bimodal basalt-rhyolite magmatism, tained deposits tend to occur as clusters or alignments that lack significant porphyry Cu systems (Sillitoe, 1999a; Tosdal may attain 5 to 30 km across or in length, respectively. Clus- and Richards, 2001). The stress regime depends, among ters are broadly equidimensional groupings of deposits (e.g., other factors, on whether there is trench advance or rollback Globe-Miami district, Arizona; Fig. 3a), whereas alignments and the degree of obliquity of the plate convergence vector are linear deposit arrays oriented either parallel or trans- (Dewey, 1980). verse to the magmatic arcs and their coincident porphyry Cu Nevertheless, there is a prominent empirical relationship belts. Arc-parallel alignments may occur along intra-arc between broadly contractional settings, marked by crustal fault zones, as exemplified by the Chuquicamata district, thickening, surface uplift, and rapid exhumation, and large, northern Chile (Fig. 3b) whereas cross-arc fault zones or lin- high-grade hypogene porphyry Cu deposits, as exemplified by eaments control arc-transverse alignments, as in the Cadia, the latest Cretaceous to Paleocene (Laramide) province of New South Wales (Fig. 3c) and Oyu Tolgoi, Mongolia dis- southwestern North America, middle Eocene to early tricts (Fig. 3d). Oligocene (Fig. 2) and late Miocene to Pliocene belts of the Irrespective of whether the porphyry Cu systems and con- central Andes, mid-Miocene belt of Iran, and Pliocene belts tained deposits define clusters or alignments, their surface in New Guinea and the Philippines (Fig. 1; Sillitoe, 1998; Hill distributions are taken to reflect the areal extents of either et al., 2002; Perelló et al., 2003a; Cooke et al., 2005; Rohrlach underlying parental plutons or cupolas on their roofs. Within and Loucks, 2005; Sillitoe and Perelló, 2005; Perelló, 2006). the clusters and alignments, the distance (100s−1,000s m) be- Large, high-sulfidation epithermal Au deposits also form in tween individual deposits (e.g., Sillitoe and Gappe, 1984) and similar contractional settings at the tops of tectonically thick- even their footprint shapes can vary greatly, as observed in the ened crustal sections, albeit not together with giant porphyry Chuquicamata and Cadia districts (Fig. 3b, c). Cu deposits (Sillitoe and Hedenquist 2003; Sillitoe, 2008). It Clusters or alignments of porphyry Cu systems can display may be speculated that crustal compression aids development a spread of formational ages, which attain as much as 5 m.y. of large mid- to upper-crustal magma chambers (Takada, in the Chuquicamata (Ballard et al., 2001; Rivera and Pardo, 1994) capable of efficient fractionation and magmatic fluid 2004; Campbell et al., 2006) and Yanacocha districts (Longo generation and release, especially at times of rapid uplift and and Teal, 2005) but could be as much as ~18 m.y. in the Cadia erosional unroofing (Sillitoe, 1998), events which may district (Wilson et al., 2007). This situation implies that the presage initiation of stress relaxation (Tosdal and Richards, underlying parental plutons have protracted life spans, albeit 2001; Richards, 2003, 2005; Gow and Walshe, 2005). Changes intermittent in some cases, with porphyry Cu formation tak- in crustal stress regime are considered by some as especially ing place above them at different places over time. favorable times for porphyry Cu and high-sulfidation epi- thermal Au deposit generation (e.g., Tosdal and Richards, Pluton-porphyry relationships 2001), with Bingham and Bajo de la Alumbrera, Argentina, Varied relationships are observed between porphyry Cu for example, both apparently occupying such a tectonic niche systems and precursor plutons, which are typically multi- (Presnell, 1997; Sasso and Clark, 1998; Halter et al., 2004; Sil- phase, equigranular intrusions, commonly of batholithic di- litoe, 2008). mensions and dioritic to granitic compositions; they are not Faults and fault intersections are invariably involved, to only spatially, but also temporally and probably genetically re- greater or lesser degrees, in determining the formational sites lated to porphyry Cu and superjacent epithermal Au forma- and geometries of porphyry Cu systems and their constituent tion (Fig. 4). The precursor plutons may act as hosts to a sin- parts. Intra-arc fault systems, active before as well as during gle deposit, as at Mount Polley, British Columbia (Fraser et magmatism and porphyry Cu generation, are particularly im- al., 1995); an alignment of coalesced deposits, as in the Los portant localizers, as exemplified by the Domeyko fault sys- Bronces-Río Blanco district (Fig. 5a); or clusters of two or tem during development of the preeminent middle Eocene more discrete deposits, as in the El Abra intrusive complex, to early Oligocene belt of northern Chile (Sillitoe and Perelló, northern Chile (Fig. 5b) and Guichon Creek batholith, High- 2005, and references therein; Fig. 2). Some investigators em- land Valley district, British Columbia (Fig. 5c). The precursor phasize the importance of intersections between continent- plutons and porphyry Cu stocks are typically separated by scale transverse fault zones or lineaments and arc-parallel time gaps of 1 to 2 m.y. or less (e.g., Dilles and Wright, 1988; structures for porphyry Cu formation, with the Archibarca Casselman et al., 1995; Mortensen et al., 1995; Dilles et al., and Calama-El Toro lineaments of northern Chile (Richards 1997; Deckart et al., 2005; Campbell et al., 2006). Many por- et al., 2001; Fig. 2), the Lachlan Transverse Zone of New phyry Cu systems, particularly those that are only shallowly South Wales (Glen and Walshe, 1999), comparable features exposed, lack known precursor plutons, probably because in New Guinea (Corbett, 1994; Hill et al., 2002), and the they lie at inaccessible depths (Fig. 4).
0361-0128/98/000/000-00 $6.00 6
Excursion métallogénique - Chili 2012 Références page 74 PORPHYRY COPPER SYSTEMS 7
a 110° 45’ b
510.000 N
N Fault st Diamond H Copper Cities RT
We Pinto Valley 7.540.000 Miami Inspiration Miami Castle East GLOBE Dome MIAMI Chuquicamata Blue Bird Cactus- Oxhide Carlotta
5km 33° 20 Mina Sur
Porphyry Cu deposit Skarn magnetite-Cu-Au deposit
Exotic (supergene) Cu deposit MM Postmineral fault Quetena Genoveva Trend of magmatic arc 7.520.000 Miranda Toki Town Opache 5km CALAMA
c d N 106.85°E Cadia 686000 E Ridgeway Big Ulan Cadia Khud N (prospect)
Cadia Quarry Hugo Dummett Little Cadia Cadia Hill Central 6296000 N 43°N Cadia East - Far East Southwest & South ? Heruga
1km 5km
FIG. 3. Examples of porphyry Cu clusters and alignments of various sizes and at different orientations with respect to the axes of contemporaneous magmatic arcs. a. Globe-Miami district cluster, Arizona within the Late Cretaceous-early Tertiary (Laramide) arc (after Creasey, 1980), with the spatial distribution partially the result of mid-Tertiary extensional tectonism (Wilkins and Heidrick, 1995; Seedorff et al., 2008). b. Chuquicamata district, northern Chile aligned parallel to the middle Eocene-early Oligocene arc axis (after Rivera and Pardo, 2004; S. Rivera, writ. commun., 2009), with the spatial distribution possibly partly the result of postmineral sinistral strike-slip faulting (Brimhall et al., 2006). c. Cadia district, New South Wales, Australia, aligned oblique to the Ordovician arc axis (after Holliday et al., 2002). d. Oyu Tolgoi district, Mongolia aligned nearly perpendicular to the Late Devonian arc axis (after Khashgerel et al., 2008). Porphyry Cu and other deposit outlines projected to surface where unexposed. Note scale difference between c and a, b, and d.
0361-0128/98/000/000-00 $6.00 7
Excursion métallogénique - Chili 2012 Références page 75 8 RICHARD H. SILLITOE
Limit of lithocap Paleosurface
Multiphase porphyry Base of Cu stock degraded volcanic edifice Late-mineral Porphyry Intermineral stock Early Parental pluton + + + Composite precursor + pluton + +
5km v Comagmatic v v volcanic rocks 5km Subvolcanic basement
FIG. 4. Spatial relationships between porphyry Cu stocks, underlying pluton, overlying comagmatic volcanic rocks, and the lithocap. The precursor pluton is multiphase, whereas the parental pluton is shown as a single body in which the con- centric dotted lines mark its progressive inward consolidation. The early, intermineral, and late-mineral phases of the por- phyry Cu stocks, which span the interval during which the porphyry Cu deposits formed, originate from increasingly greater depths in the progressively crystallizing parental chamber. The volcanic sequence is a stratovolcano (but could just as read- ily be a dome complex; Fig. 6) and has been partly eroded prior to porphyry Cu formation. The lithocap affects the volcanic pile as well as uppermost parts of the underlying rocks. Note that subvolcanic basement rocks host much of the porphyry Cu deposit on the left, whereas that on the right is mainly enclosed by two phases of the precursor pluton. Inspired by Sillitoe (1973), Dilles (1987), Tosdal and Richards (2001), Casselman et al. (1995), and Dilles and Proffett (1995).
The precursor plutons are considered as the mid- to upper- products, at least in the general vicinities of the deposits crustal crystallization sites of mafic to felsic magmas that as- themselves. Nevertheless, at a few localities, including the cended from deeper reservoirs before porphyry Cu systems shallowly formed Marte porphyry Au deposit, northern Chile were developed (see Richards, 2003). Outcropping precursor (Vila et al., 1991), a comagmatic andesitic stratovolcano is still plutons normally represent the shallower, earlier consolidated partially preserved, including parts of its unmodified lower parts rather than the magma volumes from which the fluids depositional slopes (or planèze). Notwithstanding their lower for porphyry Cu generation were derived (Fig. 4). These preservation potential, smaller volume volcanic centers— parental magma chambers, also represented by similar flow-dome complexes and maar-diatreme systems (e.g., equigranular to weakly porphyritic plutons, are not exposed in Mankayan district, Philippines and Grasberg; Sillitoe and An- porphyry Cu systems unless postmineralization extensional geles, 1985; MacDonald and Arnold, 1994; I. Kavalieris, pers. tectonism caused profound tilting and dismemberment of the commun., 1999) —may still also be recognizable in the shal- systems, as reconstructed in the Yerington district, Nevada low parts of porphyry Cu systems. Volcanic landforms are ob- (Dilles, 1987; Dilles and Proffett, 1995) and elsewhere (See- viously even better preserved in the shallower high-sulfida- dorff et al., 2008). tion epithermal environment above porphyry Cu deposits (e.g., flow-dome complexes at Yanacocha; Turner, 1999; Volcanic connections Longo and Teal, 2005; e.g., Fig. 6). Porphyry Cu systems may be spatially associated with co- Catastrophically explosive volcanism, particularly ash-flow magmatic, calc-alkaline or, less commonly, alkaline volcanic caldera formation, is normally incompatible with synchronous rocks, typically of intermediate to felsic composition (Sillitoe, porphyry Cu and superjacent epithermal Au deposit forma- 1973; Fig. 4), which are generally erupted subaerially 0.5 to 3 tion, because magmatic volatiles are dissipated during the vo- m.y. prior to stock intrusion and mineralization, as well docu- luminous pyroclastic eruptions rather than being retained and mented in the Bingham (Waite et al., 1997), Farallón Negro, focused in a manner conducive to ore formation (Sillitoe, Argentina (Sasso and Clark, 1998; Halter et al., 2004), Yer- 1980; Pasteris, 1996; Cloos, 2001; Richards, 2005). Neverthe- ington (Dilles and Wright, 1988; Dilles and Proffett, 1995), less, calderas may influence the localization of later, geneti- Tampakan, Philippines (Rohrlach and Loucks, 2005), and cally unrelated porphyry Cu systems (e.g., El Salvador, north- Yanacocha (Longo and Teal, 2005) districts. However, the ern Chile; Cornejo et al., 1997). erosion involved in the unroofing of porphyry Cu deposits There is a strong suggestion that comagmatic volcanism also severely degrades volcanic landforms (e.g., Farallón may be inhibited in some major porphyry Cu belts as a result Negro district) and, commonly, entirely removes the eruptive of their characteristic contractional tectonic settings, as in the
0361-0128/98/000/000-00 $6.00 8
Excursion métallogénique - Chili 2012 Références page 76 PORPHYRY COPPER SYSTEMS 9
a 675000E 680000E c
N Mainly mineralized hydrothermal N breccias
Krain 6335000N South Seas
Bethlehem Río 53° 30 Valley JA Los Bronces Blanco Lornex Highmont
6330000N
Los Piches Los Sulfatos Ag-Pb-Zn- Cu veins 10km 5km
b Major fault
N Late-mineral diatreme complex Porphyry stock and porphyry Cu deposit
El Abra Late felsic phases Precursor Conchi Viejo Early, mainly pluton dioritic phases Host rocks
22° 00
5km 69° 15
FIG. 5. Examples of porphyry Cu deposits within and near precursor plutons. a. Los Bronces-Río Blanco breccia-domi- nated deposit trending across the San Francisco batholith, central Chile (after Serrano et al., 1996; J.C. Toro, writ. commun., 2007). b. El Abra and Conchi Viejo deposits in the El Abra intrusive complex, northern Chile (after Dilles et al., 1997). c. Highland Valley deposit cluster in the Guichon Creek batholith, British Columbia (after Casselman et al., 1995). Note the variable positions of the deposits with respect to the exposed plutons, but their confinement to late felsic phases. Scales are different. middle Eocene to early Oligocene belt of northern Chile, be- Wall-rock influences cause of the tendency for subsurface magma accumulation in the absence of widely developed extensional faulting Porphyry Cu systems are hosted by a variety of igneous, (Mpodozis and Ramos, 1990). The same situation is also ap- sedimentary, and metamorphic rocks (e.g., Titley, 1993), giv- parent in several giant high-sulfidation epithermal Au de- ing the initial impression of wall rocks playing a noninfluen- posits generated in thickened crust during tectonic uplift, tial role. It is becoming increasingly clear, however, that cer- such as Pascua-Lama and Veladero, northern Chile-Ar- tain lithologic units may enhance grade development in both gentina, where the near absence of contemporaneous volcan- porphyry Cu and related deposit types. ism is more certain (Bissig et al., 2001; Charchaflié et al., Massive carbonate sequences, particularly where marble is 2007) given the much shallower erosion level, including par- developed near intrusive contacts, and other poorly fractured, tial paleosurface preservation (see below). fine-grained rocks have the capacity to act as relatively
0361-0128/98/000/000-00 $6.00 9
Excursion métallogénique - Chili 2012 Références page 77 10 RICHARD H. SILLITOE
High-sulfidation epithermal disseminated Au ± Ag ± Cu
Intermediate- sulfidation epithermal Au-Ag
High-sulfidation lode Cu-Au ± Ag Base of lithocap Carbonate-replacement Zn-Pb-Ag ± Au (or Cu)
Distal Au/Zn-Pb skarn
Sediment- hosted distal- Subepithermal disseminated vein Zn-Cu-Pb- Au-As ± Sb ± Hg Ag ± Au Marble front Porphyry Cu ± Au ± Mo Proximal Cu-Au skarn
1km
1km
Late-mineral porphyry LITHOCAP Phreatic breccia
PORPHYRY Intermineral magmatic-hydrothermal breccia Dacite porphyry plug-dome STOCK Intermineral porphyry Lacustrine sediment MAAR- Early porphyry DIATREME Late phreatomagmatic breccia COMPLEX PRECURSOR Early phreatomagmatic breccia PLUTON Equigranular intrusive rock Late-mineral porphyry
Dacite dome
V V V V
V Felsic tuff unit HOST V V ROCKS V V V Andesitic volcanic unit
Subvolcanic basement / carbonate horizon
FIG. 6. Anatomy of a telescoped porphyry Cu system showing spatial interrelationships of a centrally located porphyry Cu ± Au ± Mo deposit in a multiphase porphyry stock and its immediate host rocks; peripheral proximal and distal skarn, car- bonate-replacement (chimney-manto), and sediment-hosted (distal-disseminated) deposits in a carbonate unit and subep- ithermal veins in noncarbonate rocks; and overlying high- and intermediate-sulfidation epithermal deposits in and alongside the lithocap environment. The legend explains the temporal sequence of rock types, with the porphyry stock predating maar- diatreme emplacement, which in turn overlaps lithocap development and phreatic brecciation. Only uncommonly do indi- vidual systems contain several of the deposit types illustrated, as discussed in the text (see Table 3). Notwithstanding the as- sertion that cartoons of this sort (including Fig. 10) add little to the understanding of porphyry Cu genesis (Seedorff and Einaudi, 2004), they embody the relationships observed in the field and, hence, aid the explorationist. Modified from Silli- toe (1995b, 1999b, 2000).
0361-0128/98/000/000-00 $6.00 10
Excursion métallogénique - Chili 2012 Références page 78 PORPHYRY COPPER SYSTEMS 11 impermeable seals around and/or above porphyry Cu de- are hosted by the porphyry intrusions (Camus, 1975, 2003; posits, resulting in high-grade ore formation (e.g., Grasberg; Ambrus, 1977). The distal parts of porphyry Cu systems, be- Sillitoe, 1997). Elsewhere, small-volume porphyry intrusions yond the porphyry Cu deposits, either lack porphyry intru- and the associated magmatic fluids fail to effectively pene- sions or contain only relatively minor dikes (e.g., Virgin dike trate low-permeability rock packages, leading to the appar- in the skarn-dominated Copper Canyon district, Nevada, and ently uncommon development of blind, high-grade deposits, Yerington district skarn Cu occurrences; Wotruba et al., 1988; as at Hugo Dummett in the Oyu Tolgoi district (Kirwin et al., Dilles and Proffett, 1995). 2003, 2005) and Ridgeway in the Cadia district (Wilson et al., The porphyry Cu-related intrusions comprise multiple 2003). High-sulfidation epithermal deposits may be similarly phases (Kirkham, 1971; Gustafson, 1978), which were em- blind, beneath a thick limestone sequence in the case of placed immediately before (early porphyries), during (inter- Pueblo Viejo (Sillitoe et al., 2006). mineral porphyries), near the end of (late mineral por- Ferrous Fe-rich lithologic units also appear to favor high- phyries), and after (postmineral porphyries) the alteration grade porphyry Cu mineralization (e.g., Ray and Mineral and mineralization events (Fig. 6). For example, seven phases Park, Arizona; Phillips et al., 1974; Wilkinson et al., 1982), are mapped at Bajo de la Alumbrera (Proffett, 2003), five at presumably because of their capacity to effectively precipitate Yerington (Proffett, 2009), and four at Bingham (Redmond et Cu transported in oxidized magmatic fluids (see below). It is al., 2001). The immediately premineral, early porphyries and unlikely coincidental that at least half the ore at three of the their contiguous host rocks contain the highest grade miner- highest grade hypogene porphyry Cu deposits is hosted by alization in most deposits although, exceptionally, the earliest such rocks: a gabbro-diabase-basalt complex at El Teniente phase can be poorly mineralized (e.g., Grasberg; MacDonald (Skewes et al., 2002), a Proterozoic diabase sill complex at and Arnold, 1994). Intermineral porphyries are typically less Resolution, Arizona (Ballantyne et al., 2003), and a tholeiitic well mineralized as they become progressively younger, and basalt sequence in the Oyu Tolgoi district (Kirwin et al., late- and postmineral phases are barren. The earlier porphyry 2005). bodies are not destroyed when intruded by later phases but Mineralization elsewhere in porphyry Cu systems may be merely split apart, causing overall inflation of the rock pack- even more profoundly influenced by rock type. Proximal and age as would occur during ordinary dike emplacement. Sev- distal skarn, carbonate-replacement, and sediment-hosted eral criteria, in addition to metal contents and ratios mineralization types are obviously dependent on the presence (Cu/Au/Mo) and intensity of veining, alteration, and mineral- of reactive carbonate rocks, particularly thinly bedded, silty ization, are used to distinguish the relative ages of porphyry units. Large-tonnage, high-sulfidation epithermal deposits intrusions: younger phases truncate veinlets in, are chilled are favored by permeable rock packages, commonly pyroclas- against, and contain xenoliths of older phases (Fig. 7; Sillitoe, tic or epiclastic in origin (e.g., Yanacocha; Longo and Teal, 2000). Commonly, the xenoliths are largely assimilated by the 2005), although disparate lithologic units can also prove re- younger phases, leaving only the contained quartz veinlets, ceptive where extensively fractured (e.g., granitoid at Pascua- chemically more refractory than the host porphyry, as “float- Lama; Chouinard et al., 2005). ing” pieces (Fig. 7). Wall-rock xenoliths in the marginal parts of some porphyry intrusions may be sufficiently abundant to Deposit-Scale Characteristics constitute intrusion breccias. The upper contacts of a few porphyry Cu intrusions are characterized by unidirectional Porphyry stocks and dikes solidification textures (USTs): alternating, crenulate layers of Porphyry Cu deposits are centered on porphyry intrusions quartz and aplite and/or aplite porphyry produced as a result that range from vertical, pluglike stocks (Fig. 6), circular to of pressure fluctuations at the transition from magmatic to elongate in plan, through dike arrays to small, irregular bod- hydrothermal conditions (e.g., Kirkham and Sinclair, 1988; ies. The stocks and dikes commonly have diameters and Garwin, 2002; Lickfold et al., 2003; Cannell et al., 2005; Kir- lengths, respectively, of ≤1 km. However, much larger por- win, 2005). However, USTs are not consistently developed phyry intrusions act as hosts in places, such as the elongate, and, hence, do not provide a reliable means of subdividing 14-km-long stock at Chuquicamata-Radomiro Tomic (e.g., porphyry Cu intrusion phases. Ossandón et al., 2001; Fig. 3b) and the 4-km-long, <50-m- The porphyry intrusions in porphyry Cu deposits are exclu- wide dike at Hugo Dummett (Khashgerel et al., 2008; Fig. sively of I-type and magnetite-series affiliation (Ishihara, 1981), 3d). Mining and deep drilling in a few large porphyry Cu de- and typically metaluminous and medium K calc-alkaline, but posits show that mineralized intrusions have vertical extents may also fall into the high K calc-alkaline (shoshonitic) or al- of >2 km (e.g., Chuquicamata and Escondida, northern kaline fields (see Seedorff et al., 2005, for further details). Chile, and Grasberg) and, based on evidence from the steeply They span a range of compositions from calc-alkaline diorite tilted systems, perhaps ≥4 km (Dilles, 1987; Seedorff et al., and quartz diorite through granodiorite to quartz monzonite 2008; Fig. 6). The size of the stocks does not appear to bear (monzogranite), and alkaline diorite through monzonite to, any obvious relationship to the size of the associated porphyry uncommonly, syenite (e.g., Galore Creek, British Columbia; Cu deposits and their Cu contents (cf. Seedorff et al., 2005). Enns et al., 1995). Mo-rich porphyry Cu deposits are nor- For example, the 12.5-Gt resource at Chuquicamata- mally tied to the more felsic intrusions, whereas Au-rich por- Radomiro Tomic is confined to the 14-km-long stock referred phyry Cu deposits tend to be related to the more mafic end to above (Ossandón et al., 2001; Camus, 2003), whereas per- members, although intrusions as felsic as quartz monzonite haps only roughly 20 percent of the similarly sized El Te- may also host Au-rich examples (e.g., Mamut, East Malaysia; niente deposit and <10 percent of the 1.5-Gt El Abra deposit Kósaka and Wakita, 1978). However, Cu-poor porphyry Au
0361-0128/98/000/000-00 $6.00 11
Excursion métallogénique - Chili 2012 Références page 79 12 RICHARD H. SILLITOE
the component phases (e.g., Ballard et al., 2001; Maksaev et A-veinlet al., 2004; Padilla-Garza et al., 2004; Jones et al., 2007; Perelló et al., 2007; Harris et al., 2008). Furthermore, there seems to Veinlet Chilled be no obvious relationship between the size of porphyry Cu truncation contact deposits and the duration of the intrusive activity, the latter seemingly being the main parameter defining the total hy- B-veinlet drothermal life spans of porphyry Cu systems. Detailed geochronology of the high-sulfidation parts of some porphyry D-veinle Cu systems also suggests extended life spans, 1 to >1.5 m.y. at t Cerro de Pasco and Colquijirca, central Peru (Bendezú et al., Refractory 2008; Baumgartner et al., 2009). However, these life spans quartz veinlet are orders of magnitude longer than the theoretically mod- xenoliths eled times required for consolidation of individual porphyry intrusions (<40,000 yr; Cathles, 1977; Cathles et al., 1997), porphyry Cu ore formation (<100,000 yr; McInnes et al., Flow-aligned 2005), or major potassic alteration events (<2,000 yr; Shino- plagioclase hara and Hedenquist, 1997; Cathles and Shannon, 2007). 5cm phenocrysts Diatremes Late-mineral porphyry - Propylitic alteration Diatremes, upward-flared volcanic vents generated mainly Late intermineral porphyry by phreatomagmatic eruptive activity, are widespread in por- Early intermineral porphyry Potassic ± phyry Cu systems (Sillitoe, 1985), including their shallow, epi- chlorite-sericite Early porphyry thermal parts (e.g., Yanacocha; Turner, 1999; Fig. 6). The di- alteration ≥ VV Wall rock atremes, commonly 1 km in near-surface diameter and up to at least 2 km in vertical extent (e.g., >1.8 km preserved at
FIG. 7. Schematic crosscutting relationships between early (immediately the Braden diatreme, El Teniente; Howell and Molloy, 1960; premineral), intermineral, and late-mineral porphyry phases in porphyry Cu Camus, 2003), can be manifested at the paleosurface by maar stocks and their wall rocks. Veinlet truncation, quartz veinlet xenoliths, volcanoes: ephemerally lake-filled craters encircled by tuff chilled contacts, and flow-aligned phenocrysts as well as textural, grade, and rings (Fig. 6). Diatreme breccias have a distinctive texture, in metal-ratio variations may denote the porphyry contacts, albeit generally not which widely separated, typically centimeter-sized clasts are all present at the same contact. Early A, B, and late D veinlets are explained in the text and Figure 13. Note that early A veinlets are more abundant in the dominated by rock-flour matrix containing an andesitic to early porphyry, less abundant in the early intermineral porphyry, and absent dacitic tuffaceous component (Table 1), the latter commonly from the two later porphyry phases. The late-mineral porphyry lacks veinlets difficult to recognize where alteration is intense. The poorly and displays only propylitic alteration. Modified from Sillitoe (2000). lithified, friable nature and clay-rich matrix of many diatreme breccias give rise to recessive topography and little, if any, deposits appear to occur exclusively in association with calc- surface exposure. A positive topographic expression results alkaline diorite and quartz diorite porphyries (e.g., Vila and only where barren, late- to postmineral porphyry plugs in- Sillitoe, 1991). The porphyry intrusions contain variable trude the diatreme breccias (e.g., Dizon and Guinaoang, amounts of phenocrysts, typically including hornblende Philippines and Batu Hijau; Sillitoe and Gappe, 1984; Sillitoe and/or biotite, and fine-grained, commonly aplitic ground- and Angeles, 1985; Garwin, 2002; Fig. 6). mass, resulting in open to crowded textures. The distinctive Many diatremes are late-stage additions to porphyry Cu aplitic groundmass texture is ascribed to pressure quenching systems, in which they commonly postdate and either cut or during rapid ascent and consequent volatile loss (Burnham, occur alongside porphyry Cu mineralization at depth (Howell 1967). The porphyry phases in some individual deposits may and Molloy, 1960; Sillitoe and Gappe, 1984; Perelló et al., have clear compositional differences (e.g., Bajo de la Alum- 1998; Garwin, 2002) but overlap with high-sulfidation events brera; Proffett, 2003) and/or characteristic igneous textures at shallower epithermal levels (e.g., Dizon; Fig. 6). The dia- (e.g., El Salvador; Gustafson and Hunt, 1975); however, par- tremes, particularly their contact zones, may localize part of ticularly in many porphyry Au and Cu-Au deposits, the dif- the high-sulfidation Au mineralization (e.g., Wafi-Golpu, ferent phases are commonly only subtly different or nearly Papua New Guinea; Fig. 6). In a minority of cases, however, identical. Furthermore, textural obliteration is commonplace diatremes (e.g., Grasberg, Galore Creek, and Boyongan- in the highly altered, early porphyry phases, thereby render- Bayugo, Philippines; MacDonald and Arnold, 1994; Enns et ing them difficult to distinguish from volcanic wall rocks in al., 1995; Braxton et al., 2008) or tuff-filled depressions pre- some deposits (e.g., Galore Creek and Hugo Dummett). sumably fed by one or more subjacent diatremes (e.g., Reso- Isotopic dating, using the U-Pb zircon method, suggests that lution) are early features that act as receptive wall rocks to the the multiphase porphyry intrusions in porphyry Cu systems main alteration and mineralization. can be assembled in as little as 80,000 years (Batu Hijau, In- donesia; Garwin, 2002), but the process commonly takes much Magmatic-hydrothermal and phreatic breccias longer. Emplacement of the porphyry stocks in many central The most common hydrothermal breccias in the deeper Andean deposits took from 2 to 5 m.y., implying that appre- parts of porphyry Cu systems are of magmatic-hydrothermal ciable time (0.5−1.5 m.y.) elapsed between emplacement of type, the products of release of overpressured magmatic
0361-0128/98/000/000-00 $6.00 12
Excursion métallogénique - Chili 2012 Références page 80 PORPHYRY COPPER SYSTEMS 13 itute high grade in pre-existing mineralization barren, but may host porphyry (e.g., Bisbee; Bryant, 1987) Cu/Au/Ag ore Cu or high- sulfidation ore types luzonite high-sulfidation ± sericitic; bornite argillic, or none uncommonly advanced argillic on exposure level None or ad- Locally enargite Commonly type depending supported advanced Clast/matrix Alteration Main Cu-bearing Economic sulfides ± rock flour ± igneous rock (i.e., igneous breccia) subrounded rounded to angular to subrounded quartz-muscovite- tourmaline- Polymict, Muddy rock flourGener ally none Matrix Sericitic, Barren unless rich locally juvenile (magma blob, pumice) clasts porphyry Cu mineralization examples cut by dominated layers any alteration 1. Features of Principal Hydrothermal Breccia Types in Porphyry Cu Systems 1. Features of Principal Hydrothermal Breccia Types timing Clast features Matrix/cement proportions 2) types (Table mineral(s) potential Relative ABLE T bodies (abundance) Form locally around them (ubiquitous) in diam) (10s−100s m and Within Cu deposits Dikes, sills and Late of systems) Position in system (relatively irregular common) lithocaps; Within Irregular Typically Commonly Chalcedony, Clast or matrix Advanced Enargite, May constitute manifestations as (10s−100s m relative to angular to barite, sulfides, eruption breccia (relatively development in diam)common) porphyry Cu and lithocap epithermal envi- scale, ronments; surface downward-manifestations as narrowing subroundedmaar volcanoes conduits (present in ~20% examples known late, but early native S centimeter-sized, juvenile tuff or rounded, and polished; dominated; magma blob vanced argillic, component; early accretionary lapilli in matrix- but early examples with Type Magmatichydrothermal Cu deposits, porphyry Within pipe- Irregular, Typically like bodies Phreatic intermineral (porphyry Commonly Cu level) monomict, Quartz-magnetite- Clast or matrix around porphyry biotite-sulfides/ uncommonly Potassic ± supported Chalcopyrite, chlorite-sericite May const uncommonly ore, commonl y Phreatic (epithermal level) local surface bodies intermineral silicified, quartz, alunite, supported argillic Phreatomagmatic Diatremes span Kilometer- Commonly Polymict, Rock flour with Matrix
0361-0128/98/000/000-00 $6.00 13
Excursion métallogénique - Chili 2012 Références page 81 14 RICHARD H. SILLITOE
fluids (Sillitoe, 1985). Many porphyry Cu deposits contain Quartz-pyrophyllite minor volumes (5−10%) of magmatic-hydrothermal breccia alteration: quartz-pyrite- (Fig. 6); however, even major deposits can be either breccia enargite cement free, as at Chuquicamata (Ossandón et al., 2001), or breccia dominated, as exemplified by >5 Gt of ore-grade breccia at Wall Sericitic alteration: Los Bronces-Río Blanco (Warnaars et al., 1985; Serrano et al., rocks quartz-tourmaline- pyrite cement 1996; Fig. 5a). Magmatic-hydrothermal breccias display a variety of textures (Table 1), which are mainly dependent on Sericitic alteration: clast form and composition, clast/matrix ratio, matrix/cement quartz-tourmaline- constitution, and alteration type. They are distinguished from chalcopyrite cement the phreatomagmatic diatreme breccias by several features (Table 1), particularly the absence of tuffaceous material. The Remanent magnetite breccia clasts may be set in rock-flour matrix, hydrothermal cement, fine-grained igneous material, or some combination of the three. Igneous matrices tend to be more common at Early porphyry depth near the magmatic source, where the term igneous Potassic breccia is appropriately applied (e.g., Hunt et al., 1983; Fig. 8). alteration: Magmatic-hydrothermal breccias, generally occupying biotite-magnetite- chalcopyrite cement steep, pipelike to irregular bodies, are commonly intermin- eral in timing as a result of being generated in close associa- tion with intermineral porphyry phases (Figs. 6, 8). Hence, Igneous breccia many of the breccias overprint preexisting alteration-mineral- ization patterns and veinlet types (e.g., Red Mountain, Ari- zona; Quinlan, 1981), which are incorporated as clasts. Early breccias may display potassic alteration and have biotite, mag- Intermineral netite, and chalcopyrite cements, whereas later ones are com- porphyry monly sericitized and contain prominent quartz, tourmaline, specularite, chalcopyrite, and/or pyrite as cementing miner- 200m als. Sericitized breccia may change downward to potassic-al- Pegmatoidal tered breccia (e.g., Los Bronces-Río Blanco; Vargas et al., patches 200m 1999; Frikken et al., 2005; Fig. 8). The metal contents of some magmatic-hydrothermal breccias may be higher than FIG. 8. Schematic depiction of a large magmatic-hydrothermal breccia those of surrounding porphyry Cu stockwork mineralization, body genetically linked to the apex of an intermineral porphyry intrusion. reflecting their high intrinsic permeability. In contrast to dia- The alteration-mineralization is zoned from advanced argillic (with pyrite- tremes, magmatic-hydrothermal breccias are normally blind enargite) at the top through sericitic (with shallow pyrite and deep chalcopy- rite) to potassic (with magnetite-chalcopyrite ± bornite) at the bottom, where and do not penetrate the overlying epithermal environment, the breccia texture may be almost imperceptible and pegmatoidal pods are whereas downward they gradually fade away as a result of in- commonplace. Injection of fine-grained, igneous matrix defines the igneous creased clast/matrix ± cement ratios, in places accompanied breccia near the base of the body. The entire breccia body originally under- by pods of coarse-grained, pegmatoidal, potassic minerals went potassic alteration prior to partial overprinting by sericitic and, subse- quently, advanced argillic assemblages, as documented by localized occur- representing former sites of vapor accumulation (e.g., Los rence of remanent magnetite and muscovite after coarse-grained biotite in Pelambres, central Chile; Perelló et al., 2007; Fig. 8; see the cement to the sericitized breccia. below). Several types of phreatic (meteoric-hydrothermal) breccia are widely observed in porphyry Cu systems; they may be because of texture obliteration caused by intense advanced simply subdivided into pebble dikes and, uncommonly, larger argillic alteration (e.g., Pascua-Lama). bodies resulting from flashing of relatively cool ground water Phreatic breccias, as exemplified by pebble dikes, normally on approach to magma, typically late-mineral porphyry dikes; contain polymictic clast populations set in muddy, rock-flour and steep, tabular to irregular bodies triggered by vapor-pres- matrices (Table 1). Vertical clast transport may be apprecia- sure buildup beneath impermeable layers, commonly result- ble (e.g., >1 km at Tintic; Morris and Lovering, 1979). The ing from self sealing by silicification (Sillitoe, 1985; Table 1). breccias are typically late-stage features and, hence, unal- Hence, pebble dikes display downward transitions to por- tered and barren. In contrast, clast transport in the phreatic phyry intrusions (e.g., Tintic, Utah and Toquepala, southern breccias produced by fluid confinement in the high-sulfida- Peru; Farmin, 1934; Zweng and Clark, 1995), whereas the tion environment is more restricted, with many of the clasts breccias triggered by fluid confinement do not normally form being locally derived from the seals themselves and, hence, in close proximity to intrusive bodies. The pebble dikes and commonly composed of silicified rock (Table 1). Although related breccias are chiefly restricted to porphyry Cu de- rock flour may occur between the clasts, quartz, chalcedony, posits, including their marginal parts, whereas brecciation in- alunite, barite, pyrite, and enargite are also widely observed duced by the fluid confinement typifies the overlying high- as cementing minerals. These phreatic breccias host ore in sulfidation epithermal environment (Fig. 6). There, distinction many high-sulfidation Au ± Ag ± Cu deposits (e.g., Choque- from phreatomagmatic diatreme breccias may be difficult limpie, northern Chile; Gröpper et al., 1991). In contrast to
0361-0128/98/000/000-00 $6.00 14
Excursion métallogénique - Chili 2012 Références page 82 PORPHYRY COPPER SYSTEMS 15 the magmatic-hydrothermal breccias and pebble dikes at the rock-hosted ore types are replacements of receptive beds, porphyry Cu level, these phreatic breccias in the high-sulfi- commonly beneath relatively impermeable rock units (e.g., dation environment may attain the paleosurface, where hy- Titley, 1996) and, hence, tend to be strata bound, although drothermal eruptions result in subaerial breccia accumulation high- and low-angle fault control is also widely emphasized as aprons around the eruptive vents (e.g., Hedenquist and (e.g., proximal skarns at Ok Tedi, Papua New Guinea, and Henley, 1985; Fig. 6). Antamina, central Peru; Rush and Seegers, 1990; Love et al., 2004). Orebody types and geometries Distal ore formation in porphyry Cu systems is less com- The deeper, central cores of porphyry Cu systems are oc- mon in igneous or siliciclastic wall rocks, within propylitic cupied by porphyry Cu deposits, in which ore-zone geome- halos, where fault- and fracture-controlled, subepithermal tries depend mainly on the overall form of the host stock or Zn-Pb-Cu-Ag ± Au veins of currently limited economic im- dike complex, the depositional sites of the Cu-bearing sul- portance tend to be developed (e.g., Mineral Park; Eidel et fides, and the positions of any late, low- and subore-grade al., 1968 and Los Bronces-Río Blanco; Figs. 5a, 6). Neverthe- porphyry intrusions and diatremes. The forms of many por- less, larger tonnage orebodies may occur where permeable phyry Cu deposits mimic those of their host intrusions; thus, wall rocks exist, as exemplified by the stacked, Au-bearing cylindrical stocks typically host cylindrical orebodies (Fig. 6), mantos in amygdaloidal and brecciated andesitic flow tops at whereas laterally extensive dikes give rise to orebodies with Andacollo, Chile (Reyes, 1991). similar narrow, elongate shapes (e.g., Hugo Dummett; In the lithocap environment—typically located above, are- Khashgerel et al., 2008). Many porphyry Cu deposits are ally more extensive than, and commonly overprinting por- formed as vertically extensive bodies, which become progres- phyry Cu deposits (Fig. 6; see below)—high-sulfidation epi- sively lower grade both outward and downward, whereas oth- thermal Au, Ag, and/or Cu deposits are characteristic; ers have a bell- or cap-like form because little Cu was precip- nevertheless, the preserved parts of many lithocaps are es- itated internally at depth (e.g., Resolution; Ballantyne et al., sentially barren. The deeper level high-sulfidation deposits, 2003). The tops of the orebodies tend to be relatively abrupt the Cordilleran base metal lodes of Einaudi (1982), tend to be and controlled by the apices of quartz veinlet stockworks (see characterized by massive sulfides, commonly rich in the Cu- below). The shape of any porphyry Cu orebody may undergo bearing sulfosalts (enargite, luzonite, and/or famatinite). They significant modification as a result of emplacement of late- to commonly occur as tabular veins overprinting porphyry Cu postmineral rock volumes (e.g., Fig. 5a), as exemplified by the deposits, like those at Butte (Meyer et al., 1968), Escondida low-grade cores caused by internal emplacement of late- (Ojeda, 1986), Chuquicamata (Ossandón et al., 2001), and mineral porphyry phases (e.g., Santo Tomas II, Philippines; Collahuasi, northern Chile (Masterman et al., 2005; Fig. 6). Sillitoe and Gappe, 1984) and, much less commonly, late- Alternatively, for up to several kilometers beyond porphyry stage diatremes (e.g., El Teniente; Howell and Molloy, 1960; Cu deposits, they comprise structurally controlled replace- Camus, 2003). A few deposits, instead of dying out either ments and hydrothermal breccias, either in volcanic rocks as gradually (e.g., El Salvador; Gustafson and Quiroga, 1995) or at Lepanto in the Mankayan district (Garcia, 1991; Heden- fairly abruptly (e.g., H14-H15 at Reko Diq, Pakistan) at quist et al., 1998), Nena in the Frieda River district, Papua depth, have knife-sharp bases as a result of truncation by late- New Guinea (Espi, 1999), and Chelopech, Bulgaria (Cham- mineral intrusions (e.g., Santo Tomas II; Sillitoe and Gappe, befort and Moritz, 2006) or, where lithocaps impinge on car- 1984). Coalescence of closely spaced porphyry Cu deposits bonate rocks, as deposits like Smelter in the Marcapunta sec- enhances size potential (e.g., H14-H15 at Reko Diq; Perelló tor at Colquijirca (Vidal and Ligarda, 2004; Bendezú and et al., 2008) Fontboté, 2009). In contrast, much larger tonnage, dissemi- Development of wall rock-hosted orebodies alongside por- nated Au ± Ag orebodies are more typical of the shallower phyry Cu deposits is most common where receptive carbon- (<500 m) parts of lithocaps (Sillitoe, 1999b), as exemplified ate rocks are present (Fig. 6). Deposit types include proximal by Yanacocha (Harvey et al., 1999) and Pascua-Lama Cu ± Au and, less commonly, distal Au and/or Zn-Pb skarns (Chouinard et al., 2005), although much deeper development (e.g., Meinert, 2000; Meinert et al., 2005); more distal, car- of disseminated Cu-Au deposits is also relatively common bonate-replacement (chimney-manto), massive sulfide bodies (e.g., Tampakan; Rohrlach et al., 1999). dominated by either Cu (e.g., Superior district, Arizona and Intermediate-sulfidation epithermal precious metal de- Sepon district, Laos [Fig. 9c]; Paul and Knight, 1995; Loader, posits, containing Zn-Pb-Ag ± Cu ± Au as well as Mn-bear- 1999) or, more commonly, Zn, Pb, Ag ± Au (e.g., Recsk, Hun- ing carbonates, rhodonite, and quartz, occur alongside litho- gary; Kisvarsanyi, 1988) beyond the skarn front (Fig. 6); and, caps but typically spatially separate from the high-sulfidation uncommonly, sediment-hosted (distal-disseminated; Cox and orebodies, as observed in the case of the Victoria and Teresa Singer, 1990) Au concentrations on the fringes of the systems vein systems at Lepanto (Claveria, 2001; Hedenquist et al., (e.g., Barneys Canyon and Melco, Bingham district; Babcock 2001) and in the Collahuasi district (Masterman et al., 2005; et al., 1995; Gunter and Austin, 1997; Cunningham et al., Fig. 6). Locally, however, the intermediate-sulfidation and 2004; Fig. 9a). Continuity between some of these carbonate both Cordilleran lode and shallow, disseminated high-sulfi- rock-hosted deposits is possible; for example, transitions from dation mineralization types display transitional mineralogic proximal Cu-Au to distal Au skarn in the Copper Canyon dis- relationships, as exemplified by the so-called Main Stage trict (Cary et al., 2000) and distal Zn-Pb-Cu-Ag skarn to car- veins at Butte (Meyer et al., 1968) and the disseminated A bonate-replacement Zn-Pb-Ag at Groundhog, Central dis- and Link Au zones at Wafi-Golpu (Leach, 1999; Ryan and trict, New Mexico (Meinert, 1987). All these carbonate Vigar, 1999). The intermediate-sulfidation epithermal veins
0361-0128/98/000/000-00 $6.00 15
Excursion métallogénique - Chili 2012 Références page 83 16 RICHARD H. SILLITOE
a b Limit of data Au-Ag N Barneys Canyon 30000N N Pb-Zn Melco Au-As
Sa Cu Cu-Mo Pb-Zn cr Tertiary amen volcanic Pyrite-Cu rocks + to Covered area alluvium Fa Cu-Mo ult orebody Cu
Pb-Zn Pb-Zn
10000S 3km Au-Ag
0
Outer limits of: 5km Cu-Mo Pb-Zn Cu Au Pyrite
c Discovery Khanong Au deposit Cu deposit Au-As-Sb Padan quartz veinlet stockwork
Cu Mo-Cu N Cu-Mo Nam Kok Au deposit Cu-Mo Nalou Au deposit Au-As-Sb Thengkham quartz veinlet stockwork
limit of jasperoid Outer 5km
FIG. 9. Examples of well-developed metal zoning centered on porphyry Cu deposits. a. Bingham, Utah, where the por- phyry Cu-Au-Mo deposit is followed successively outward by Cu-Au skarn, carbonate-replacement Zn-Pb-Ag-Au, and distal sediment-hosted Au deposits, the latter formerly exploited at Barneys Canyon and Melco (after Babcock et al., 1995). b. Min- eral Park, Arizona, where the northwest-striking vein system centered on the porphyry Cu-Mo deposit is zoned outward from Cu through Pb-Zn to Au-Ag (after Lang and Eastoe, 1988). c. Sepon, Laos, where two subeconomic porphyry Mo-Cu cen- ters marked by quartz veinlet stockworks are zoned outward through carbonate-replacement Cu to sediment-hosted Au de- posits without any intervening Zn-Pb-Ag zone (summarized from R.H. Sillitoe, unpub. report, 1999). Note the large radii (up to 8 km) of some systems. Scales are different.
0361-0128/98/000/000-00 $6.00 16
Excursion métallogénique - Chili 2012 Références page 84 PORPHYRY COPPER SYSTEMS 17 are the shallow-level (<1 km paleodepth) counterparts of the chlorite-sericite, sericitic, and advanced argillic (cf. Meyer Zn-Pb-Cu-Ag ± Au veins located alongside porphyry Cu de- and Hemley, 1967; Table 2; Figs. 10, 11). Chloritic and propy- posits, but no direct connection between the two types is ev- litic alteration develop distally at shallow and deeper levels, ident (Fig. 6). The massive, high-sulfidation pyrite-enargite respectively (Fig. 10). Equating chlorite-sericite alteration— replacements in carbonate rocks are also locally transitional the abbreviated name used by Hedenquist et al. (1998) for outward through high- to intermediate-sulfidation Zn-Pb-Ag the sericite-clay-chlorite (SCC) of Sillitoe and Gappe (1984) ore, a continuum observed in the Tintic and Colquijirca dis- —with Meyer and Hemley’s (1967) low-temperature inter- tricts (Lindgren and Loughlin, 1919; Bendezú et al., 2003; mediate argillic type (e.g., Hedenquist et al., 1998; Sillitoe, Bendezú and Fontboté, 2009). 2000; Seedorff et al., 2005; Bouzari and Clark, 2006) causes confusion and should probably be discontinued. Phyllic Alteration-mineralization zoning in porphyry Cu deposits (Lowell and Guilbert, 1970) and sericitic are synonyms. Porphyry Cu deposits display a consistent, broad-scale alter- The alteration-mineralization zoning sequence typically af- ation-mineralization zoning pattern that comprises, centrally fects several cubic kilometers of rock (e.g., Lowell and Guil- from the bottom upward, several of sodic-calcic, potassic, bert, 1970; Beane and Titley, 1981), although sericitic and,
Vuggy residual quartz/silicification Steam heated
Quartz- kaolinite Quartz- alunite Intermediate argillic Quartz- pyrophyllite Chloritic tic Serici Decalcification More reduced Weakly skarn altered ite ic er Oxidized Massive -s skarn sulfide te
lori
Ch
Unaltered Potassic
Propylitic 1km
1km Sodic- calcic
FIG. 10. Generalized alteration-mineralization zoning pattern for telescoped porphyry Cu deposits, based on the geologic and deposit-type template presented as Figure 6. Note that shallow alteration-mineralization types consistently overprint deeper ones. Volumes of the different alteration types vary markedly from deposit to deposit. Sericitic alteration may project vertically downward as an annulus separating the potassic and propylitic zones as well as cutting the potassic zone centrally as shown. Sericitic alteration tends to be more abundant in porphyry Cu-Mo deposits, whereas chlorite-sericite alteration de- velops preferentially in porphyry Cu-Au deposits. Alteration-mineralization in the lithocap is commonly far more complex than shown, particularly where structural control is paramount. See text for further details and Table 2 for alteration-miner- alization details. Modified from Sillitoe (1999b, 2000).
0361-0128/98/000/000-00 $6.00 17
Excursion métallogénique - Chili 2012 Références page 85 18 RICHARD H. SILLITOE only Comm Economic but locally ore bearing Barren, except barren, but may their roots for subepither- mal veins contributor constitute ore constitutes ore constitutes bite/oligoclase barren, Normally Al pyrophyllite/dickite, quartz-kaolinite in lithocaps and 1 (designation) selvages Veinlet potential 3 actinolite (M-type) quartz-sulfides ± magnetite (A-type), quartz-molybdenite chalcopyrite ± bornite; ± pyrite chalcopyrite others none, except (central suture; B-type) locally K-feldspar around A- and B-types Typically absentTypically Magnetite ± assemblages (minor) veinlets bornite ± chalcocite, galena) (pyrite-enargite ± sulfides (D-type) tennantite, pyrite- pyrite-sphalerite) Actinolite, epidote, Pyrite-chalcopyrite, albite, carbonate, tourmaline, magnetite Biotite (EB-type), K-feldspar, digenite ± chalcocite bornite, bornite ± EDM-type with sulfides (EDM/T4-type), K-feldspar-andalusite- + disseminated Main ore K-feldspar ± andalusite specularite Possible ancillary Principal sulfide Contemporaneous dumortierite, topaz, pyrite-covellite carbonate, tourmaline, 4 2. Characteristics of Principal Alteration-Mineralization Types in Porphyry Cu Systems 2. Characteristics of Principal Alteration-Mineralization Types ABLE T Key minerals minerals pyrophyllite, specularite dickite, kaolinite specularite) residual, vuggy), zunyite, corundum, pyrite-chalcocite, (includes veins) Deep, including Albite/oligoclase, Diopside, below porphyry Cu deposits (uncommon) actinolite, magnetite porphyry Cu deposits (ubiquitous) K-feldspar epidote, garnet sericite, andalusite, chalcopyrite ± Marginal parts of quartz-biotite-sericite- Chlorite, sericite ± biotite Actinolite, hematite, Pyrite (± sphalerite, contributor Pyrite, epidote Position in system (common) systems, below lithocaps (ubiquitous) carbonate epidote, albite, zones (common, magnetite particularly in Au-rich deposits) hematite (martite, porphyry Cu deposits constitutes lithocaps alunite, intrusions) (ubiquitous, except with alkaline 2 Many veinlets in potassic, chlorite-sericite, and sericitic alteration contain anhydrite, which also occurs as late, largely monomineralic veinlets Many veinlets in potassic, chlorite-sericite, and sericitic alteration contain anhydrite, which also occurs as late, largely monomineralic Alunite commonly intergrown with aluminum-phosphate-sulfate (APS) minerals (see Stoffregen and Alpers, 1987) Excluding those developed in carbonate-rich rocks Arranged from probable oldest (top) to youngest (bottom), except for propylitic that is lateral equivalent of potassic; advance d argillic also forms above potassic early in systems (Fig. 10) 1 2 3 4 Potassic (K-silicate) Core zones of Biotite, Propylitic Sodic-calcic Alteration type (alternative name) (abundance) terminology) Chlorite-sericite (sericite-clay-chlorite [SCC]) porphyry Cu core Upper parts of sericite/illite, Sericitic (phyllic) Chlorite, smectite Upper parts of Carbonate, epidote, Pyrite-chalcopyrite Quartz, sericitein Russian Chlorite ± sericite sulfides Pyrophyllite, Chlorite, sericite/illite Common ore Pyrite ± chalcopyrite Quartz-pyrite ± other Quartz-sericite Advanced argillic (secondary quartzite Cu deposits, Above porphyry Quartz (partly Diaspore, andalusite, Pyrite-enargite, Pyrite-enargite ± Cu sulfides quartz- Quartz-alunite, Locally
0361-0128/98/000/000-00 $6.00 18
Excursion métallogénique - Chili 2012 Références page 86 PORPHYRY COPPER SYSTEMS 19
Steam Vuggy residual SHALLOW Advanced argillic heated quartz/silicification (py) (py-en, py-cv)
Quartz- Quartz- alunite kaolinite Sericitic Chloritic (py, py-cp, py-bn) Sericitic Quartz-
pyrophyllite km Chlorite-
Chlorite- 5 sericite sericite 1. e (cp-py) at st Potassic n Propylitic io (cp-bn) at lfid prop su (py) + ity Potassic Multiphase id porphyry Ac stock 1km DEEP 1km EARLY 0.2– 5 Ma LATE
FIG. 11. Generalized alteration-mineralization zoning pattern for a non- FIG. 12. Schematic representation of generalized alteration-mineraliza- telescoped porphyry Cu system, emphasizing the appreciable, commonly tion sequence in porphyry Cu systems in relation to paleodepth and system barren gap that exists between the lithocap and underlying porphyry stock. life span. The sequence, from potassic with peripheral propylitic (prop) Legend as in Figure 10. through chlorite-sericite and sericitic to advanced argillic, is the result of in- creasing acidity consequent upon the declining temperature of the hy- drothermal fluids. A broadly parallel increase in sulfidation state of the fluids results in changes in the sulfide assemblage from chalcopyrite (cp)-bornite particularly, advanced argillic alteration are much less well (bn), through chalcopyrite-pyrite (py) and pyrite-bornite, to pyrite-enargite (en) or pyrite-covellite (cv), as charted for several deposits by Einaudi et al. developed in porphyry Cu deposits associated with alkaline (2003). Note the absence of Cu-bearing sulfides from the early, high-tem- than with calc-alkaline intrusions (Lang et al., 1995; Sillitoe, perature advanced argillic zone. Modified from Sillitoe (2000). 2002; Holliday and Cooke, 2007), reflecting control of the K+/H+ ratio by magma chemistry (e.g., Burnham, 1979). Spe- cific opaque mineral assemblages are intrinsic parts of each Large parts of many porphyry Cu deposits (e.g., Lowell and alteration type (Table 2; Fig. 12) because of the direct linkage Guilbert, 1970; Titley, 1982), especially deeply formed (e.g., between sulfidation state, the chief control on sulfide assem- Butte; Rusk et al., 2004, 2008a) or relatively deeply eroded blages, and solution pH, a principal control of alteration type examples like El Abra (Ambrus, 1977; Dean et al., 1996) and (Barton and Skinner, 1967; Meyer and Hemley, 1967; Ein- Gaby (Gabriela Mistral), northern Chile (Camus, 2001, audi et al., 2003; Fig. 12). Sulfidation state, a function of S fu- 2003), are made up predominantly of potassic alteration, gacity and temperature, changes from low through interme- which grades marginally into generally weakly developed diate to high as temperature declines (Barton and Skinner, propylitic zones (Fig. 10). Biotite is the predominant alter- 1967; Einaudi et al., 2003). In general, the alteration-miner- ation mineral in relatively mafic porphyry intrusions and host alization types become progressively younger upward (Fig. rocks, whereas K-feldspar increases in abundance in more 12), with the result that the shallower alteration-mineraliza- felsic, granodioritic to quartz monzonitic settings. Sodic pla- tion zones invariably overprint and at least partly reconstitute gioclase may be an accompanying alteration mineral in both deeper ones. settings. Locally, texture-destructive quartz-K ± Na-feldspar Sodic-calcic alteration, commonly magnetite bearing (Table flooding overprints and destroys the more typical potassic as- 2), is normally rather poorly preserved at depth in some por- semblages (e.g., Chuquicamata; Ossandón et al., 2001). The phyry Cu deposits, commonly in the immediate wall rocks to chalcopyrite ± bornite ore in many porphyry Cu deposits is the porphyry intrusions (e.g., Panguna, Papua New Guinea largely confined to potassic zones (Table 2; Fig. 12), with one and El Teniente; Ford, 1978; Cannell et al., 2005), a position or more bornite-rich centers characterizing the deeper, cen- that may give rise to confusion with propylitic alteration (Fig. tral parts of many deposits. In some bornite-rich centers, the 10). Nevertheless, it also characterizes the centrally located sulfidation state is low enough to stabilize digenite ± chal- zones of some porphyry Cu stocks (e.g., Koloula, Solomon Is- cocite (Einaudi et al., 2003; Table 2). Chalcopyrite-bornite lands and Island Copper, British Columbia; Chivas, 1978; cores are transitional outward to chalcopyrite-pyrite annuli, Perelló et al., 1995; Arancibia and Clark, 1996). Sodic-calcic which, with increasing sulfide contents, grade into pyrite alteration is typically sulfide and metal poor (except for Fe as halos, typically parts of the surrounding propylitic zones magnetite) but can host mineralization in Au-rich porphyry (Table 2). Pyrrhotite may accompany the pyrite where re- Cu deposits (e.g., Nugget Hill, Philippines), in some of which duced host rocks are present (e.g., Kósaka and Wakita, 1978; hybrid potassic-calcic (biotite-actinolite-magnetite) assem- Perelló et al., 2003b). The potassic alteration affects the early blages are also commonplace (e.g., Santo Tomas II, Ridgeway, and intermineral porphyry generations (Fig. 7) and many in- and Cotabambas, southern Peru; Sillitoe and Gappe, 1984; termineral magmatic-hydrothermal breccias as well as vari- Wilson et al., 2003; Perelló et al., 2004a). able volumes of wall rocks. The potassic-altered wall rocks
0361-0128/98/000/000-00 $6.00 19
Excursion métallogénique - Chili 2012 Références page 87 20 RICHARD H. SILLITOE range from restricted veneers near the stock or dike contacts quartz veinlet stockworks (see below). These high-sulfidation to kilometer-scale zones, such as those in the mafic host litho- assemblages commonly have higher Cu contents than the for- logic units mentioned previously at El Teniente, Resolution, mer potassic alteration, resulting in hypogene enrichment and Oyu Tolgoi. The potassic alteration generally becomes (Brimhall, 1979), although any Au may be depleted (e.g., less intense from the older to younger porphyry phases, al- Wafi-Golpu; Sillitoe, 1999b). The Cu-bearing sulfides typi- though the late-mineral intrusions postdate it and display a cally occur as fine-grained coatings on disseminated pyrite propylitic assemblage (Fig. 7), albeit of later timing than the grains, which leads to recognition difficulties in deposits that propylitic halos developed peripheral to potassic zones. also underwent supergene Cu sulfide enrichment (e.g., Chlorite-sericite alteration (Table 2), giving rise to distinc- Chuquicamata, Ossandón et al., 2001); indeed, the hypogene tive, pale-green rocks, is widespread in the shallower parts of contribution is commonly not distinguished from the super- some porphyry Cu deposits, particularly Au-rich examples, gene enrichment products (e.g., Taca Taca Bajo, Argentina; where it overprints preexisting potassic assemblages (Figs. 10, Rojas et al., 1999). 11). The alteration is typified by partial to complete transfor- The root zones of advanced argillic lithocaps, commonly at mation of mafic minerals to chlorite, plagioclase to sericite least partly structurally controlled, may overprint the upper (fine-grained muscovite) and/or illite, and magmatic and any parts of porphyry Cu deposits, where the sericitic alteration is hydrothermal magnetite to hematite (martite and/or specu- commonly transitional upward to quartz-pyrophyllite (Fig. larite), along with deposition of pyrite and chalcopyrite. Al- 10), an assemblage widespread in the deep, higher tempera- though Cu and/or Au tenors of the former potassic zones may ture parts of many lithocaps (e.g., El Salvador; Gustafson and undergo depletion during the chlorite-sericite overprints Hunt, 1975; Watanabe and Hedenquist, 2001). Elsewhere, (e.g., Esperanza, northern Chile; Perelló et al., 2004b), metal however, lower temperature quartz-kaolinite is the dominant introduction is also widely recognized (e.g., Leach, 1999; overprint assemblage (e.g., Caspiche, northern Chile). The Padilla Garza et al., 2001; Harris et al., 2005; Masterman et advanced argillic alteration preferentially affects lithologic al., 2005) and, at a few deposits, is considered to account for units with low (e.g., quartz sandstone, felsic igneous rocks) much of the contained Cu (e.g., Cerro Colorado, northern rather than high (mafic igneous rocks) acid-buffering capaci- Chile; Bouzari and Clark, 2006). ties. At several localities, the advanced argillic alteration at Sericitic alteration (Table 2) in porphyry Cu deposits nor- the bottoms of lithocaps displays a characteristic patchy tex- mally overprints and wholly or partially destroys the potassic ture, commonly defined by amoeboid pyrophyllite patches and chlorite-sericite assemblages (Figs. 10−12), although embedded in silicified rock (e.g., Escondida and Yanacocha; sericitic veinlet halos are zoned outward to chlorite-sericite Padilla Garza et al., 2001; Gustafson et al., 2004). However, alteration in places (e.g., Dilles and Einaudi, 1992). The degree the patches may also comprise alunite or kaolinite, suggesting of overprinting is perhaps best appreciated in some magmatic- that the texture may result by either preferential nucleation of hydrothermal breccia bodies in which isolated magnetite ag- any common advanced argillic mineral or even advanced gregates occur as stranded remnants in sericitic or chlorite- argillic overprinting of a nucleation texture developed during sericite zones up to 1 km above the magnetite-cemented, earlier potassic or chlorite-sericite alteration of fragmental rocks potassic parts (e.g., Chimborazo, northern Chile; Fig. 8). (e.g., Hugo Dummett; Khashgerel et al., 2008, and Caspiche). Sericitic alteration may be subdivided into two different types: The vertical distribution of the alteration-mineralization an uncommon, early variety that is greenish to greenish-gray types in porphyry Cu deposits depends on the degree of over- in color and a later, far more common and widespread, white printing or telescoping, the causes of which are addressed variety. In the few deposits where it is recognized, the early, further below. In highly telescoped systems, the advanced greenish sericitic alteration is centrally located and hosts a argillic lithocaps impinge on the upper parts of porphyry low sulfidation-state chalcopyrite-bornite assemblage, which stocks (Fig. 10) and their roots may penetrate downward for is commonly ore grade (e.g., Chuquicamata; Ossandón et al., >1 km. In such situations, the advanced argillic alteration may 2001). The late, white sericitic alteration has varied distribu- be 1 to >2 m.y. younger than the potassic zone that it over- tion patterns in porphyry Cu deposits. It may constitute annu- prints (e.g., Chuquicamata and Escondida; Ossandón et al., lar zones separating the potassic cores from propylitic halos, 2001; Padilla-Garza et al., 2004), reflecting the time needed as emphasized in early porphyry Cu models (Jerome, 1966; for the telescoping to take place. Where telescoping is lim- Lowell and Guilbert, 1970; Rose, 1970), but is perhaps more ited, however, the lithocaps and potassic-altered porphyry common as structurally controlled or apparently irregular re- stocks may be separated by 0.5 to 1 km (Sillitoe, 1999b), a gap placements within the upper parts of chlorite-sericite and/or typically occupied by pyritic chlorite-sericite alteration (Fig. potassic zones (Fig. 10). The sericitic alteration is commonly 11). pyrite dominated, implying effective removal of the Cu (± Au) Where carbonate (limestone and dolomite) instead of ig- present in the former chlorite-sericite and/or potassic assem- neous or siliciclastic rocks host porphyry Cu deposits, calcic blages. However, sericitic alteration may also constitute ore or magnesian exoskarn is generated in proximity to the por- where appreciable Cu remains with the pyrite, either in the phyry intrusions, whereas marble is produced beyond the form of chalcopyrite or as high sulfidation-state assemblages skarn front (Fig. 10). In the case of limestone protoliths, an- (typically, pyrite-bornite, pyrite-chalcocite, pyrite-covellite, pyrite- hydrous, prograde andraditic garnet-diopsidic pyroxene skarn tennantite, and pyrite-enargite; Table 2; cf. Einaudi et al., 2003). forms contemporaneously with the potassic alteration of non- The main development of these bornite-, chalcocite-, and carbonate lithologic units, whereas hydrous, retrograde skarn, covellite-bearing, high-sulfidation assemblages is largely con- commonly containing important amounts of magnetite, acti- fined to white sericitic alteration that overprints now-barren nolite, epidote, chlorite, smectite, quartz, carbonate, and iron
0361-0128/98/000/000-00 $6.00 20
Excursion métallogénique - Chili 2012 Références page 88 PORPHYRY COPPER SYSTEMS 21 sulfides, is the equivalent of the chlorite-sericite and sericitic type), and K-feldspar, and typically lacking alteration sel- assemblages (Einaudi et al., 1981; Meinert et al., 2003). Dis- vages; (2) sulfide-bearing, granular quartz-dominated veinlets tal Au skarns are typically more reduced (pyroxene rich) than with either narrow or no readily recognizable alteration sel- their proximal counterparts (Fortitude, Copper Canyon dis- vages (A and B types); and (3) late, crystalline quartz-sulfide trict; Myers and Meinert, 1991; Fig. 10). A quartz-pyrite as- veins and veinlets with prominent, feldspar-destructive alter- semblage replaces any carbonate rocks incorporated in ad- ation selvages (including D type). Group 1 and 2 veinlets are vanced argillic lithocaps (e.g., Bisbee, Arizona; Einaudi, mainly emplaced during potassic alteration, whereas group 3 1982). Endoskarn tends to be volumetrically minor (Beane accompanies the chlorite-sericite, sericitic, and deep ad- and Titley, 1981; Meinert et al., 2005). The massive sulfide vanced argillic overprints. Narrow, mineralogically complex carbonate-replacement deposits are normally enveloped by quartz-sericite-K-feldspar-biotite veinlets with centimeter- marble. Any sediment-hosted Au mineralization on the scale halos defined by the same minerals (± andalusite ± fringes of carbonate rock-hosted porphyry Cu systems forms corundum) along with abundant, finely disseminated chal- where rock permeability is enhanced by decalcification (Fig. copyrite ± bornite characterize the changeover from group 1 10), including sanding of dolomite, but also locally occluded to 2 veinlets in a few deposits, although they may have been by Au-related jasperoid formation (e.g., Bingham and Sepon confused elsewhere with D-type veinlets because of their districts; Babcock et al., 1995; Smith et al., 2005; Fig. 9a, d). eye-catching halos; they are termed early dark micaceous (EDM) halo veinlets at Butte (Meyer, 1965; Brimhall, 1977; Porphyry Cu veinlet relationships Rusk et al., 2008a) and Bingham (Redmond et al., 2004), and The veinlet sequence in porphyry Cu deposits, first elabo- type 4 (T4) veinlets at Los Pelambres (Atkinson et al., 1996; rated by Gustafson and Hunt (1975) at El Salvador and Perelló et al., 2007). Group 3 also includes uncommon, but widely studied since (e.g., Hunt et al., 1983; Dilles and Ein- economically important massive chalcopyrite ± bornite ± audi, 1992; Gustafson and Quiroga, 1995; Redmond et al., chalcocite veinlets at the high-grade Grasberg (Pollard and 2001; Pollard and Taylor, 2002; Cannell et al., 2005; Master- Taylor, 2002; I. Kavalieris, pers. commun., 1999), Hugo Dum- man et al., 2005), is highly distinctive. In a general way, the mett (Khashgerel et al., 2008), and Resolution deposits as well veinlets may be subdivided into three groups (Table 2; Fig. as elsewhere. 13): (1) early, quartz- and sulfide-free veinlets containing one Many porphyry Cu deposits display single veinlet sequences or more of actinolite, magnetite (M type), (early) biotite (EB that comply with the generalizations summarized above and
a b A A M D A A
B Chlorite halo
D
Quartz- sericite halo M 2cm A A A A B K-feldspar halo VEINLET CHRONOLOGY
Biotite M M Magnetite±actinolite Quartz-magnetite- Granular quartz- A A A A chalcopyrite±bornite chalcopyrite Quartz-molybdenite± A A Quartz-chalcopyrite BBchalcopyrite±pyrite (±suture) Chlorite-pyrite±quartz± chalcopyrite Quartz-pyrite±
DDchalcopyrite LATE EARLY
FIG. 13. Schematic chronology of typical veinlet sequences in a. porphyry Cu-Mo deposits and b. porphyry Cu-Au de- posits associated with calc-alkaline intrusions. Porphyry Cu-Au deposits hosted by alkaline intrusions are typically veinlet poor (Barr et al., 1976; Lang et al., 1995; Sillitoe, 2000, 2002). Background alteration between veinlets is mainly potassic, which is likely to contain more K-feldspar in the Mo-rich than the Au-rich porphyry Cu stockworks. Note the common ab- sence of B- and D-type veinlets from Au-rich porphyry Cu stockworks and M-, magnetite-bearing A-, and chlorite-rich vein- lets from Mo-rich porphyry Cu stockworks. Veinlet nomenclature follows Gustafson and Hunt (1975; A, B, and D types) and Arancibia and Clark (1996; M type).
0361-0128/98/000/000-00 $6.00 21
Excursion métallogénique - Chili 2012 Références page 89 22 RICHARD H. SILLITOE in Figure 13 and Table 2, but repetitions of group 1 and 2 vein- of the A-type generations, but D-type veinlets may also con- lets, for example, early biotite, EDM halo, and A types cut by tain appreciable amounts of molybdenite in some deposits. lesser numbers of later EDM halo and A types (e.g., Bingham; The B-type veinlets are typically absent from Au-rich, Mo- Redmond et al., 2001), occur where time gaps between por- poor porphyry Cu deposits (Fig. 13b). The D-type veinlets, phyry phases are sufficiently large; however, group 2 and 3 far more abundant in porphyry Cu-Mo than Cu-Au deposits veinlets are only uncommonly repeated. Additional complica- (Fig. 13a), may also occur as structurally controlled swarms tions are widely introduced by repetitive veinlet reopening (e.g., El Abra; Dean et al., 1996), a characteristic particularly during subsequent veining events. Much of the metal in many evident in the case of the late-stage, meter-scale, enargite- porphyry Cu deposits is contained in the quartz-dominated, bearing, massive sulfide veins spanning the upper parts of group 2 veinlets and as disseminated grains in the intervening porphyry Cu deposits and lower parts of overlying lithocaps potassic-altered rocks, although some of the late, group 3 (Fig. 6; see above). quartz-sulfide veins and their wall rocks may also be important Magnetite ± actinolite (M-type) and quartz-magnetite (A- contributors. Irrespective of whether the Cu-bearing sulfide type) veinlets are far less common in Mo- than Au-rich por- minerals are coprecipitated with veinlet quartz or, as generally phyry Cu deposits (Fig. 13), the latter typified by particularly seems to be the case, introduced paragenetically later (e.g., elevated hydrothermal magnetite contents, commonly attain- Redmond et al., 2001, 2004), a particularly strong correlation ing 5 to 10 vol percent (Sillitoe, 1979, 2000; MacDonald and exists between quartz veinlet intensity and metal content in Arnold, 1994; Proffett, 2003). The dominant veinlets in most many porphyry Cu deposits, particularly in Au-rich examples Au-only porphyry deposits, as documented in the Maricunga (Sillitoe, 2000). However, the porphyry Cu-Au deposits associ- belt, are distinctly banded and comprise layers of both translu- ated with alkaline rocks, particularly those in British Colum- cent and dark-gray quartz (Vila and Sillitoe, 1991), the color of bia, are largely devoid of veinlets (Barr et al., 1976). Once the latter commonly caused by abundant vapor-rich fluid in- formed, the quartz-bearing veinlets are permanent features clusions (Muntean and Einaudi, 2000). These banded veinlets that are not erased during subsequent alteration overprinting, are ascribed to the shallowness of porphyry Au formation (<1 although their metal contents may be wholly or partially re- km; Vila and Sillitoe, 1991; Muntean and Einaudi, 2000). moved (see above). Therefore recognition of A- and B-type Anhydrite and tourmaline are prominent veinlet, breccia- veinlets in sericitic or advanced argillic zones testifies unam- filling, and alteration minerals in many porphyry Cu deposits biguously to the former presence of potassic alteration. (Table 2), including associated skarns. The anhydrite, attain- The A-type veinlets range from stockworks to subparallel, ing 5 to 15 percent of rock volumes, occurs in small amounts sheeted arrays, the latter particularly common in Au-rich por- in most group 2 and 3 veinlet types as well as in the form of phyry deposits (Sillitoe, 2000). Few, if any, stockworks are disseminated grains in the intervening altered rocks but com- truly multidirectional and one or more preferred veinlet ori- monly also constitutes end-stage, nearly monomineralic vein- entations are the norm. These may reflect district-scale tec- lets. Absence of anhydrite to depths of several hundred me- tonic patterns (e.g., Heidrick and Titley, 1982; Lindsay et al., ters beneath the current surface in many porphyry Cu 1995) or, where concentric and radial arrays predominate, systems is normally due to supergene dissolution (see Sillitoe, control by magma ascent and/or withdrawal in the subjacent 2005). Tourmaline may occur in minor amounts in several parental chambers (e.g., El Teniente; Cannell et al., 2005). veinlet types, even those formed early in deposit histories The quartz veinlet stockworks are most intense in and around (e.g., T4 veinlets al Los Pelambres; Perelló et al., 2007), but the early porphyry intrusions, where they may constitute as it is most abundant with quartz and/or pyrite in D-type vein- much as 90 to 100 percent of the rock (e.g., Ok Tedi and let generations and any associated magmatic-hydrothermal Hugo Dummett; Rush and Seegers, 1990; Khashgeral et al., breccias affected by sericitic alteration (Fig. 8). 2006), and die out gradually both laterally into the wall rocks (e.g., Sierrita-Esperanza, Arizona; Titley et al., 1986) and Advanced argillic lithocaps downward (e.g., El Salvador; Gustafson and Quiroga, 1995); The upper parts of porphyry Cu systems, mainly at shallower however, they tend to have more clearcut upper limits, just a levels than their porphyry intrusions, are characterized by few tens of meters above the apices of the porphyry intru- lithocaps: lithologically controlled zones of pervasive advanced sions, in the few deposits where relevant data are available argillic alteration with structurally controlled components, (e.g., Guinaoang, Wafi-Golpu, and Hugo Dummett; Sillitoe including their subvertical root zones (Figs. 4, 6, 10; Table 2; and Angeles, 1985; Sillitoe, 1999b; Khashgeral et al., 2006). Sillitoe, 1995a). Original lithocaps have areal extents of sev- The quartz veinlets commonly cut proximal prograde ex- eral to >10 and, locally, up to 100 km2 and thicknesses of >1 oskarn (Einaudi, 1982) but do not extend into the more distal km, and hence are much more extensive than the underlying carbonate rock-hosted ore types. Locally, early A-type vein- porphyry Cu deposits. Indeed, two or more porphyry Cu de- lets displaying aplitic centers or along-strike transitions to posits may underlie some large, coalesced lithocaps (Fig. 4), aplite and/or aplite porphyry (vein dikes) are observed (e.g., which, as noted above, may have formed progressively over Gustafson and Hunt, 1975; Heithersay et al., 1990; Lickfold periods of up to several million years (e.g., Yanacocha; et al., 2003; Rusk et al., 2008a). The earliest A-type veinlets Gustafson et al., 2004; Longo and Teal, 2005). Most observed may be sinuous and have nonmatching margins, features as- lithocaps are only erosional remnants, which may either cribed to formation under high-temperature, overall ductile wholly or partially overlie and conceal porphyry Cu deposits conditions, whereas later veinlets are more planar. (e.g., Wafi-Golpu; Sillitoe, 1999b) or occur alongside them and, Much of the Mo in many porphyry Cu-Mo deposits occurs hence, above propylitic rock (e.g., Nevados del Famatina, Ar- in the B-type veinlets, in marked contrast to the Cu dominance gentina, Batu Hijau, and Rosia Poieni, Romania; Lozada-
0361-0128/98/000/000-00 $6.00 22
Excursion métallogénique - Chili 2012 Références page 90 PORPHYRY COPPER SYSTEMS 23
Calderón and McPhail, 1996; Clode et al., 1999; Milu et al., solution in bornite and, to a lesser degree, chalcopyrite (e.g., 2004; Figs. 6, 10). Many lithocaps are vertically zoned, from Arif and Baker, 2002), and Cu are introduced together as the previously described quartz-pyrophyllite at depth to pre- components of centrally located potassic zones; hence, the dominant quartz-alunite and residual quartz—the residue of two metals normally correlate closely (Sillitoe, 2000; Ulrich extreme base leaching (Stoffregen, 1987) with a vuggy ap- and Heinrich, 2001; Perelló et al., 2004b). Gold grades may pearance that reflects the original rock texture—at shallower be up to ~50 percent higher in bornite-rich than chalcopyrite- levels where the causative fluid was cooler and, hence, more dominated potassic assemblages, which has been explained acidic (Giggenbach, 1997; Fig. 10). The roots of lithocaps may by the experimental observation that bornite solid solution is also contain the relatively high temperature species, an- capable of holding up to one order of magnitude more Au dalusite and corundum (>~370ºC; Hemley et al., 1980), as ac- than intermediate solid solution (ISS), the high-temperature companiments to pyrophyllite and/or muscovite (e.g., Cabang precursors of bornite and chalcopyrite, respectively (Simon et Kiri, Indonesia, El Salvador, and Cerro Colorado; Lowder al., 2000; Kesler et al., 2002). The Au grains in some deposits and Dow, 1978; Watanabe and Hedenquist, 2001; Bouzari contain minor amounts of PGE minerals, particularly Pd tel- and Clark, 2006). Where the fluids that cause advanced lurides (Tarkian and Stribrny, 1999). In contrast, Cu and Mo argillic alteration are F rich, topaz, zunyite, and fluorite are correlate less well, with spatial separation of the two metals lithocap minerals (e.g., Hugo Dummett; Perelló et al., 2001; commonly resulting from the different timing of their intro- Khashgerel et al., 2006, 2008, and Resolution). The principal duction (e.g., Los Pelambres; Atkinson et al., 1996). In many borosilicate mineral in lithocaps is dumortierite rather than Au-rich porphyry Cu deposits, Mo tends to be concentrated tourmaline. The more structurally and lithologically confined as external annuli partly overlapping the Cu-Au cores (e.g., components of lithocaps, termed ledges rather than veins be- Saindak, Pakistan, Cabang Kiri, Batu Hijau, Bajo de la Alum- cause they are mainly the products of rock replacement brera, and Esperanza; Sillitoe and Khan, 1977; Lowder and rather than incremental open-space filling, display well-de- Dow, 1978; Ulrich and Heinrich, 2001; Garwin, 2002; Prof- veloped alteration zoning (e.g., Steven and Ratté, 1960; Stof- fett, 2003; Perelló et al., 2004b). The Bingham, Island Cop- fregen, 1987), with cores of vuggy, residual quartz, and asso- per, and Agua Rica, Argentina, porphyry Cu-Au-Mo deposits ciated silicification rimmed outward (and downward) by are exceptions to this generalization because of their deep, consecutive bands of quartz-alunite, quartz-pyrophyllite/dick- centrally located molybdenite zones (John, 1978; Perelló et ite/kaolinite (pyrophyllite and dickite at hotter, deeper levels), al., 1995, 1998). and chlorite-illite/smectite. The Cu ± Mo ± Au cores typically have kilometer-scale Although all these alteration zones are pyritic, the Au-, Ag-, halos defined by anomalous Zn, Pb, and Ag values that reflect and Cu-bearing, high sulfidation-state assemblages (com- lower temperature, hydrothermal conditions (Fig. 9a, b). In monly pyrite-enargite and pyrite-covellite; Table 2; Fig. 12) some systems, Mn (±Ag) is also markedly enriched in the out- tend to be confined to the vuggy, residual quartz and silicified ermost parts of the halos (e.g., Butte; Meyer et al., 1968). rock, the latter normally better mineralized where phreatic These Zn-Pb-Ag ± Mn halos commonly coincide spatially breccias are present (see above). Apart from the massive, with propylitic alteration zones but are invariably best defined commonly enargite-bearing sulfide veins and replacement in the distal skarn environment (e.g., Meinert, 1987; Meinert bodies in the deeper parts of some lithocaps (see above), et al., 2005), beyond which even more distal Au-As ± Sb veins and veinlets are generally only poorly developed, with zones may be developed (e.g., Bingham and Sepon districts; much of the pyrite and any associated sulfides being in dis- Babcock et al., 1995; Cunningham et al., 2004; Smith et al., seminated form. Open-space filling is also uncommon, except 2005; Fig. 9a, c). Peripheral veins cutting propylitic halos may in phreatic breccias and unusual, isolated veins (e.g., La Meji- also be Au rich, and at Mineral Park an outward zoning from cana alunite-pyrite-famatinite vein at Nevados del Famatina; Pb-Zn to Au-Ag is evident (Eidel et al., 1968; Lang and Eas- Lozada-Calderón and McPhail, 1996). Barite and native S are toe, 1988; Fig. 9b). Nevertheless, in some porphyry Cu de- common late-stage components of many ledges. posits, these halo metals, particularly Zn, occur as late-stage These advanced argillic alteration zones extend upward to veinlet arrays overprinting the Cu-dominated cores rather the sites of paleowater tables, which may be defined, if suitable than peripherally (e.g., Chuquicamata; Ossandón et al., 2001). aquifers (e.g., fragmental volcanic rocks) were present, by sub- In a general sense, the broad-scale zoning pattern devel- horizontal, tabular bodies of massive opaline or chalcedonic oped in the deeper parts of porphyry Cu systems persists into silicification, up to 10 m or so thick; the low crystallinity is the overlying lithocap environment where any Cu and Au caused by the low temperature (~100ºC) of silica deposition. (±Ag) commonly occur approximately above the underlying The overlying vadose zones are marked by easily recognizable, porphyry Cu deposits, albeit commonly areally more exten- steam-heated alteration rich in fine-grained, powdery cristo- sively, particularly where structural control is prevalent. The balite, alunite, and kaolinite (Sillitoe, 1993, 1999b; Fig. 10). main geochemical difference between the Cu-Au zones in porphyry Cu deposits and those in the overlying lithocaps is Metal zoning the elevated As (±Sb) contents consequent upon the abun- Metal zoning in porphyry Cu systems is well documented, dance of the Cu sulfosalts in the latter. Nevertheless, the particularly at the deeper, porphyry Cu levels (e.g., Jerome, lithocap mineralization also contains greater albeit trace 1966; Titley, 1993). There, Cu ± Mo ± Au characterize the amounts of Bi, W, Sn, and/or Te (e.g., Einaudi, 1982) as well potassic, chlorite-sericite, and sericitic cores of systems. How- as appreciable Mo. The Cu/Au ratios of lithocap-hosted, high- ever, in Au-rich porphyry Cu deposits, the Au, as small (<20 sulfidation mineralization tend to decrease upward, with the µm) grains of high (>900) fineness native metal and in solid result that most major high-sulfidation Au (±Ag) deposits
0361-0128/98/000/000-00 $6.00 23
Excursion métallogénique - Chili 2012 Références page 91 24 RICHARD H. SILLITOE occur in the shallow parts of lithocaps, commonly—but not The parental magmas need to be water rich (>~4 wt %) and always—with their tops immediately below the former pale- oxidized in order to maximize the metal contents of the re- owater table positions (Sillitoe, 1999b). Nevertheless, super- sultant aqueous phase (Burnham and Ohmoto, 1980; Candela gene leaching commonly masks the original Cu distribution and Holland, 1986; Dilles, 1987; Cline and Bodnar, 1991; pattern. Any intermediate-sulfidation precious metal miner- Candela, 1992; Candela and Piccoli, 2005; Richards, 2005). alization developed alongside the lithocaps contains much High water contents result in magmas becoming saturated higher contents of Zn, Pb, Ag, and Mn than the high-sulfida- with the aqueous phase, into which the ore metals can parti- tion orebodies, in keeping with the situation described above tion efficiently; and high oxidation state suppresses magmatic from the porphyry Cu level. The shallow-level, steam-heated sulfide, such as pyrrhotite, precipitation, a process that may and paleowater-table zones are typically devoid of precious cause sequestration of metals before they can partition into and base metals and As and Sb, unless telescoped onto the the aqueous phase. Nevertheless, resorption of any sulfide underlying mineralization as a result of water-table descent, melt during ascent of oxidized magmatic fluids could make a but commonly have elevated Hg contents (e.g., Pascua-Lama; major contribution to metal budgets (Keith et al., 1997; Hal- Chouinard et al., 2005). ter et al., 2005). The magmas are also exceptionally S rich, as emphasized by recognition of anhydrite as a magmatic min- Genetic Model eral in some porphyry stocks (Lickfold et al., 2003; Audétat et al., 2004; Stern et al., 2007; Chambefort et al., 2008). Addi- Magma and fluid production tion of mafic melt to the parental chambers could be an ef- Porphyry Cu systems typically span the upper 4 km or so fective means of augmenting S and metal budgets (Keith et of the crust (Singer et al., 2008; Figs. 6, 10), with their cen- al., 1997; Hattori and Keith, 2001; Maughan et al., 2002; Hal- trally located stocks being connected downward to parental ter et al., 2005; Zajacz and Halter, 2009). magma chambers at depths of perhaps 5 to 15 km (Cloos, 2001; Richards, 2005; Fig. 4). The parental chambers, tend- Early porphyry Cu system evolution ing to be localized at sites of neutral buoyancy (Cloos, 2001; Porphyry Cu mineralization in the deeply formed (up to 9 Richards, 2005), are the sources of both magmas and high- km) potassic alteration zones at Butte and elsewhere took temperature, high-pressure metalliferous fluids throughout place directly from a single-phase, relatively low salinity system development. (2−10 wt % NaCl equiv), aqueous liquid (Rusk et al., 2004, Field observations and theoretical calculations suggest that 2008a); such a phase may contain several thousand ppm to parental chambers with volumes on the order of 50 km3 may several percent of base metals and several ppm Au, based on be capable of liberating enough fluid to form porphyry Cu de- thermodynamic (Heinrich, 2005) and analytical (Audétat et posits, but chambers at least an order of magnitude larger are al., 2008) observations. However, at the shallower depths typ- needed to produce giant systems, particularly where deposit ical of most deposits (<~4 km), the mineralization is intro- clusters or alignments exist (Dilles, 1987; Cline and Bodnar, duced by a two-phase fluid, comprising a small fraction of 1991; Shinohara and Hedenquist, 1997; Cloos, 2001; Cathles hypersaline liquid (brine) and a much larger volume of low- and Shannon, 2007). The metal-charged aqueous phase is re- density vapor (Fournier, 1999), produced by either direct ex- leased from the cooling and fractionating parental chambers solution from the melt (Shinohara, 1994) or, more typically, as during open-system magma convection as well as later stag- the single-phase liquid decompresses, cools, and intersects its nant magma crystallization (Shinohara and Hedenquist, solvus (e.g., Henley and McNabb, 1978; Burnham, 1979; 1997). Convection provides an efficient mechanism for deliv- Cline and Bodnar, 1991; Webster, 1992; Bodnar, 1995; Cline, ery of copious amounts of the aqueous phase, in the form of 1995). Coexistence of immiscible hypersaline liquid and bubble-rich magma, from throughout the parental chambers vapor has been ubiquitously demonstrated in numerous fluid to the basal parts of porphyry stocks or dike swarms (Candela, inclusion studies (Roedder, 1984), which also show that the 1991; Shinohara et al., 1995; Cloos, 2001; Richards, 2005). In liquid phase is enriched in Na, K, and Fe chlorides, giving rise most systems, any volcanism ceases before porphyry Cu sys- to salinities of 35 to 70 wt percent NaCl equiv (e.g., Roedder, tem formation is initiated, although relatively minor eruptive 1971; Nash, 1976; Eastoe, 1978; Bodnar, 1995), whereas the activity, such as dome emplacement, may be either inter- vapor phase contains acidic volatile species, preeminently spersed with or perhaps even accompany ascent of the mag- SO2, H2S, CO2, HCl, and any HF (e.g., Giggenbach, 1992, matic aqueous phase (e.g., Bingham and Yanacocha; Deino 1997). Fluid inclusion microanalysis and experimental studies and Keith, 1997; Longo and Teal, 2005). reveal that, during phase separation, specific element suites The shallow-level porphyry stocks do not themselves gen- selectively fractionate between the vapor and hypersaline liq- erate the bulk of the magmatic fluid volume, but simply act uid. In many cases, vapor can contain an appreciable amount as “exhaust valves,” conduits for its upward transmission of the Cu, Au, Ag, and S, plus much of the As, Sb, Te, and B, from the parental chambers, perhaps via cupolas on their whereas Fe, Zn, Pb, Mn, and possibly Mo preferentially par- roofs (Fig. 4). This scenario implies episodic but focused tition into the hypersaline liquid (Heinrich et al., 1999; Hein- magma and fluid ascent for as long as ~5 m.y. in the case of rich, 2005; Pokrovski et al., 2005, 2008, 2009; Williams-Jones long-lived porphyry Cu systems, whereas elsewhere the loci and Heinrich, 2005; Simon et al., 2007; Audétat et al., 2008; of intrusive and hydrothermal activity migrate, either sys- Nagaseki and Hayashi, 2008; Wilkinson et al., 2008; Pudack et tematically or randomly, to give rise to the porphyry Cu and al., 2009; Seo et al., 2009). epithermal Au deposit clusters and alignments discussed Transport of Cu and probably also Au was for decades tac- above. itly assumed to be in the form of chloride complexes in the
0361-0128/98/000/000-00 $6.00 24
Excursion métallogénique - Chili 2012 Références page 92 PORPHYRY COPPER SYSTEMS 25 hypersaline liquid phase (e.g., Holland, 1972; Burnham, (Burnham, 1979): a process that may give rise to large in- 1967, 1997; Burnham and Ohmoto, 1980; Candela and Hol- creases in rock volume (Cathles and Shannon, 2007). The sin- land, 1986), but recent experimental work and fluid inclusion gle-phase liquid, the mineralizer in deeply formed porphyry S analysis show that volatile S ligands (H2S ± SO2) in the Cu deposits, may generate the relatively uncommon EDM vapor phase can also act as major Cu- and Au-transporting halo veinlets (Rusk et al., 2008a; Proffett, 2009), whereas the agents (Nagaseki and Hayashi, 2008; Pokrovski et al., 2008, two-phase fluid produces the more common A- and B-type 2009; Seo et al., 2009; Zajacz and Halter, 2009). In contrast, quartz veinlets (e.g., Roedder, 1984, and references therein). Mo may be transported as different, possibly oxochloride The local occurrence of vein dikes (see above), as well as complexes in the hypersaline liquid phase (Ulrich and Mavro- recognition of coexisting melt and aqueous fluid inclusions in genes, 2008). early quartz veinlets (Harris et al., 2003), confirms that Current orthodoxy maintains that the early sodic-calcic al- magma and mineralizing fluid commonly coexist, although teration observed in some porphyry Cu deposits is a product markedly different densities dictate that they typically sepa- of inflowing brine sourced from host-rock sequences (Carten, rate. The stockwork veinlets control and focus continued fluid 1986; Dilles and Einaudi, 1992; Dilles et al., 1995; Seedorff et ascent, with partial dissolution of quartz during cooling al., 2005, 2008), in keeping with theoretical predictions for through its retrograde solubility field (<~550-400°C at pres- fluids following heating paths under silicate-rock−buffered sures <~900 b; Fournier, 1999) enhancing the permeability of conditions (e.g., Giggenbach, 1984, 1997). Light stable iso- the A-type quartz veinlets during at least some of the Cu-Fe tope studies of sodic-calcic alteration in the Yerington district sulfide precipitation (Rusk and Reed, 2002; Redmond et al., support the involvement of externally derived brine from the 2004; Landtwing et al., 2005); synmineral faulting and frac- host sedimentary sequence (Dilles et al., 1992, 1995), al- turing may play a similar role. The quartz-veined cores of though the albite-actinolite alteration there is magnetite de- potassic zones remain barren where temperatures are too structive (Carten, 1986; Dilles et al., 1995). In other cases, high to permit appreciable Cu-Fe sulfide and associated Au however, there is evidence for an origin from hypersaline deposition, potentially giving rise to the bell- and cap-shaped magmatic liquids, with the paucity of contained sulfide min- ore zones described above (e.g., Bingham, Resolution, and eralization being due to excessively high temperatures and Batu Hijau; Babcock et al., 1995; Ballantyne et al., 2003; oxygen fugacities and the consequent deficiency of reduced S Setyandhaka et al., 2008). Fluid pressures may fluctuate from (John, 1989; Clark and Arancibia, 1995; Lang et al., 1995). A lithostatic to hydrostatic during porphyry Cu formation (e.g., magmatic source would certainly be favored where sodic-cal- Ulrich et al., 2001), as a result of both repetitive fracture cic zones are metal bearing (see above). propagation and sealing and reductions in confining pressure As porphyry Cu systems cool through the 700° to 550°C consequent upon surface degradation (see below). These temperature range, the single-phase liquid or, more com- pressure variations may induce changes in the fluid phases monly, coexisting hypersaline liquid and vapor initiate potas- present and consequent remobilization as well as precipita- sic alteration and perhaps the first metal precipitation in and tion of metals (e.g., Klemm et al., 2007; Rusk et al., 2008a). around the early porphyry intrusions (e.g., Eastoe, 1978; Bod- Magmatic-hydrothermal brecciation may be triggered by sud- nar, 1995; Frei, 1995; Ulrich et al., 2001). Nevertheless, in den release of fluid overpressures caused by roof failure many porphyry Cu deposits, it is fluid cooling over the ~550º above large, expanding vapor bubbles (Norton and Cathles, to 350°C range, assisted by fluid-rock interaction, that is 1973; Burnham, 1985), particularly near the ductile-brittle largely responsible for precipitation of the Cu, in low sulfida- transition (Fournier, 1999). tion-state Cu-Fe sulfide assemblages, plus any Au (e.g., Ul- During the protracted potassic alteration event(s) that af- rich et al., 2001; Redmond et al., 2004; Landtwing et al., fect the early and intermineral porphyries and their immedi- 2005; Klemm et al., 2007; Rusk et al., 2008a). In addition, up- ate wall rocks, heated external water, largely meteoric but ward decompression and expansion of the vapor phase causes possibly containing a connate component (e.g., Bingham; rapidly decreasing solubility of the vapor-transported metals Bowman et al., 1987), generates the peripheral propylitic al- (Williams-Jones et al., 2002), as confirmed by their very low teration, mainly by moderate-temperature hydration reac- contents in high-temperature but atmospheric-pressure fu- tions (Meyer and Hemley, 1967). Convective circulation of maroles (Hedenquist, 1995). Such a decrease in solubility the external water takes place where rock permeabilities are leads to wholesale precipitation of the Cu-Fe sulfides to- adequate (Fig. 14): a process that acts as a potent cooling gether with Au, thereby potentially accounting for the typi- mechanism for porphyry Cu systems (Cathles, 1977), particu- cally shallow formation (Cox and Singer, 1992; Sillitoe, 2000) larly after parental intrusions have crystallized and no longer of Au-rich porphyry Cu deposits (Williams-Jones and Hein- exsolve magmatic fluid. rich, 2005). The different Mo complexing (see above), proba- The voluminous vapor readily separates from the coexist- bly assisted by progressive increase of the Mo/Cu ratio in the ing hypersaline liquid and, because of its lower density, as- residual parental melt as crystallization proceeds (Candela cends buoyantly into the 1- to 2-km-thick rock column above and Holland, 1986), results in much of the molybdenite being the porphyry intrusions (e.g., Henley and McNabb, 1978; precipitated not only later than but also spatially separate Hedenquist et al., 1998; Fig. 14). Progressive disproportiona- from the bulk of the Cu ± Au (see above). tion of the contained SO2 (to H2SO4 and H2S) once it and HCl Potassic alteration and associated metal deposition are ini- (plus any HF) condense into ground water (Giggenbach, tiated under near-lithostatic conditions and involve extensive 1992; Rye, 1993) generates the extremely low pH fluid re- hydraulic fracturing of the ductile rock at high strain rates sponsible for the high degrees of base leaching involved in ad- (Fournier, 1999) to generate the pervasive stockwork veining vanced argillic lithocap formation (e.g., Meyer and Hemley,
0361-0128/98/000/000-00 $6.00 25
Excursion métallogénique - Chili 2012 Références page 93 26 RICHARD H. SILLITOE
Depth, High-temperature km fumarole 0 Progressive paleosurface degradation High-temperature lithocap HS epithermal Moderate- temperature lithocap Sericitic ± 2 chlorite-sericite 400°C Potassic IS epithermal 300°C
4 500°C
Rapid Further 300°C
cooling cooling 6 PORPHYRY MAGMA
EARLY LATE Two-phase Single-phase liquid Hypersaline liquid Single-phase liquid Ground (high temperature) (low temperature) water Vapor
FIG. 14. Schematic time slices through the telescoped porphyry Cu system illustrated in Figures 6 and 10 to show the evolution of the main magmatic fluid and alteration-mineralization types in concert with progressive downward magma so- lidification, cooling, and paleosurface degradation. At the early stage (left side), magma is present at the top of the parental chamber, a single-phase, low- to moderate-salinity liquid exits the magma and undergoes phase separation during ascent to generate immiscible hypersaline liquid and vapor, which generate potassic alteration plus contained low sulfidation-state por- phyry Cu ± Au mineralization. The upward-escaping, low-pressure vapor that does not attain the paleosurface as high-tem- perature fumaroles (e.g., Hedenquist, 1995; Hedenquist et al., 1993) forms acidic condensate to produce generally barren advanced argillic alteration. As magma solidification advances downward (middle), the entire system progressively cools, and the rock can fracture in a brittle fashion on cooling below ~400ºC (Fournier, 1999); at this stage, lithostatic gives way to hy- drostatic pressure, and erosion (or some other mechanism) progressively degrades the paleosurface. Under these lower tem- perature conditions, sericitic ± chlorite-sericite alteration zones begin to form from a deeply derived, single-phase aqueous liquid generated by one or both of the methods (see text) postulated by Hedenquist et al. (1998) and Heinrich et al. (2004). Eventually (right side), the sericitic ± chlorite-sericite alteration may cause variable degrees of Cu ± Au removal, but hypo- gene Cu enrichment is also possible in the former. The same liquid continues upward into the lithocap where, upon cooling in an unbuffered environment, it evolves into a high sulfidation-state liquid; if properly focused, it may generate high-sulfi- dation (HS) epithermal deposits. Renewed neutralization of this same liquid on exiting the lithocap and/or aliquots of the deep liquid that bypass the lithocap entirely may give rise to peripheral intermediate-sulfidation (IS) epithermal mineraliza- tion. Based on modeling by Hedenquist et al. (1998), Sillitoe and Hedenquist (2003), and Heinrich (2005).
1967). Focused ascent of the reactive fluid through fault and unlikely to produce much mineralization, thereby possibly ac- other permeable conduits leads to generation of the vuggy, counting for the barren status of many lithocaps (e.g., Heden- residual quartz cores (if pH is <2; Stoffregen, 1987), flanked quist et al., 1998, 2000; Heinrich et al., 2004; Heinrich, 2005). by zoned advanced argillic halos (Table 2; see above) indica- tive of partial outward fluid penetration, neutralization, and Late porphyry Cu system evolution cooling. However, because of the low pressure of the lithocap As the underlying parental magma chambers progressively environment and, hence, low metal-transporting capability of solidify and magma convection ceases, there are marked re- the absorbed vapor (see above), the resultant acidic fluid is ductions in both the heat flux and aqueous fluid supply to the
0361-0128/98/000/000-00 $6.00 26
Excursion métallogénique - Chili 2012 Références page 94 PORPHYRY COPPER SYSTEMS 27 overlying porphyry Cu systems (Dilles, 1987; Shinohara and permeability (Fig. 14). In the skarn environment, the early Hedenquist, 1997), effects that are accompanied by down- two-phase hypersaline liquid plus vapor is likely to be fol- ward propagation of the lithostatic-hydrostatic transition lowed under declining temperature conditions by the single- (Fournier, 1999). Under these lower temperature conditions, phase liquid (e.g., Meinert et al., 1997, 2003; Fig. 14), from the aqueous liquid phase exsolves more slowly from the still which the retrograde skarn Cu ± Au ± Zn, carbonate-re- crystallizing magma and, in turn, advects more slowly and placement Cu or Zn-Pb-Ag-(Au), and sediment-hosted Au- cools, such that it may not intersect its solvus. If this scenario (As-Sb) deposits are formed (e.g., Meinert et al., 1997, 2003; is correct, a single-phase, low- to moderate-salinity (5−20 wt Heinrich, 2005). % NaCl equiv) liquid in the 350° to 250°C temperature range High Zn, Pb, Ag, and Mn contents are recorded in hyper- ascends directly from the parental chambers into overlying saline liquid inclusions from quartz veinlets formed during porphyry Cu systems (Shinohara and Hedenquist, 1997; potassic alteration (Bodnar, 1995; Heinrich et al., 1999; Ul- Hedenquist et al. 1998; Fig. 14). Alternatively, a single-phase rich et al., 1999; Wilkinson et al., 2008), but these chloride- liquid may form, possibly after separation of some brine, by complexed metals (see above) remain in solution because subsequent contraction of vapor of the same composition as it they are not appreciably concentrated in the sulfides present cools at elevated pressures above the critical curve of the fluid in the main porphyry Cu orebodies. Cooling of the hyper- system (Heinrich et al., 2004; Heinrich, 2005). The low-salin- saline liquid in contact with external wall rocks and dilution ity liquid, whose ascent is controlled by the preexisting quartz with meteoric water in the propylitic halos may be the main veinlet stockworks, synmineral faults, and permeability con- causes of Zn, Pb, Ag, and Mn precipitation (Hemley and trasts provided by steep intrusive contacts, appears to be re- Hunt, 1992), giving rise to the geochemical halos of these sponsible for the progressive formation of the chlorite-sericite metals and, in some systems, localized vein concentrations and sericitic alteration, as well as continued advanced argillic (Jerome, 1966; Figs. 6, 10). The largest concentrations of pe- alteration and the principal Cu and Au mineralization in the ripheral Zn, Pb, and Ag are confined to systems hosted by re- overlying lithocaps (Hedenquist et al., 1998; Heinrich et al., ceptive carbonate rocks, where fluid neutralization induces 2004; Rusk et al., 2008b). the precipitation of these metals in skarn and carbonate-re- Admixture of magmatic and meteoric fluids, with the latter placement deposits (Seward and Barnes, 1997). dominant, was long considered necessary to produce sericitic The fluid most likely to lead to appreciable high-sulfidation alteration and the attendant low- to moderate-salinity liquid, Au ± Ag ± Cu mineralization in the relatively barren, early- i.e., 5 to 10× dilution of the hypersaline liquid (e.g., Shep- formed lithocaps is the low- to moderate-salinity, H2S-rich, pard et al., 1971; Taylor, 1974), but recent interpretations of aqueous liquid that produces the underlying sericitic zones stable O and H isotope data reveal that an exclusively mag- (Hedenquist et al., 1998; Heinrich et al., 2004; Heinrich, matic fluid is quite capable of producing the chlorite-sericite 2005; Pudack et al., 2009; Fig. 14). On entering the lithocap and sericitic assemblages (Kusakabe et al., 1990; Hedenquist environment, this intermediate sulfidation-state liquid (form- and Richards, 1998; Hedenquist et al., 1998; Watanabe and ing chalcopyrite and tennantite at depth) becomes unbuffered Hedenquist, 2001; Harris and Golding, 2002; Skewes et al., and easily evolves to a higher sulfidation state on cooling 2003; Rusk et al., 2004; Khashgerel et al., 2006). However, (Einaudi et al., 2003; Sillitoe and Hedenquist, 2003). The meteoric water involvement in late sericitic alteration is by no Cordilleran massive sulfide lodes are localized where the liq- means precluded (e.g., Hedenquist et al., 1998; Harris et al., uid follows pronounced structural permeability spanning the 2005), particularly on the margins of systems where the ad- sericitic to advanced argillic transition (Figs. 6, 10) or, less vecting magmatic liquid may entrain convecting meteoric commonly, encounters reactive carbonate rocks (e.g., Baum- water, although its formerly preeminent role in the porphyry gartner et al., 2008; Bendezú and Fontboté, 2009). However, Cu genetic model (e.g., Beane and Titley, 1981; Hunt, 1991) much of the Au precipitates in the shallower parts of lithocaps is now greatly diminished. Since chlorite-sericite alteration because of the greater likelihood of sharp drops in Au solu- partially or totally reconstitutes potassic assemblages, and bility caused by either intense boiling in upflow conduits or sericitic alteration does the same to potassic and/or chlorite- admixture of the ascendant liquid with cool, inflowing ground sericite assemblages, it is generally impossible to determine if water; in some cases, the latter appears to originate from the the contained metals are inherited from the former sulfide vadose zone (see below) where it was steam heated (Heden- assemblage(s) (e.g., Brimhall, 1979) or newly introduced in quist et al., 1998; Heinrich, 2005, and references therein; the ascendant, still magmatic-sourced aqueous liquid. How- Figs. 6, 14). These shallow Au precipitation processes may be ever, apparent confinement of hypogene Cu enrichment (see particularly effective in permeable phreatic breccias created above) to sericitic alteration overprinting rocks cut by quartz by boiling of the ascendant liquid, vapor buildup beneath sili- veinlet stockworks that formerly contained chalcopyrite ± cified seals, and eventual catastrophic release, perhaps as- bornite may suggest that a large component of the Cu in the sisted by external triggers (faulting, seismic shaking, and/or newly generated high sulfidation-state assemblages is derived deep intrusion contributing gases; e.g., Nairn et al., 2005). by relatively localized remobilization (Sillitoe, 1999b). The low- to moderate-salinity liquids responsible for high- The base and precious metal deposit types in both carbon- sulfidation deposits in lithocaps may, under appropriate struc- ate and noncarbonate wall-rock lithologic units likely form tural and hydrologic conditions, pass into adjoining, less- from the same aqueous magmatic fluids that are involved in altered rocks and undergo sufficient neutralization and porphyry Cu alteration and mineralization, wherever there is reduction during outward flow and wall-rock reaction to pro- provision of lateral fluid access from the porphyry stock or duce liquids appropriate for formation of intermediate-sulfi- dikes via lithologic, structural, and/or hydrothermally induced dation epithermal deposits (Sillitoe, 1999b; Einaudi et al.,
0361-0128/98/000/000-00 $6.00 27
Excursion métallogénique - Chili 2012 Références page 95 28 RICHARD H. SILLITOE
2003; Sillitoe and Hedenquist, 2003; Fig. 14). The above- deposit of the type that is sought, or (3) frontier terranes with cited examples of mineralogic transitions between high- and geologic conditions that are perceived to imply potential. On intermediate-sulfidation mineralization provide support for the basis of recent exploration successes, the first choice has this mechanism. Alternatively, the deeply derived intermedi- been shown to be a wise one, as witnessed by the strings of ate sulfidation-state liquids may bypass the lithocaps entirely high-sulfidation Au and Au-rich porphyry discoveries that and still produce intermediate-sulfidation mineralization at now define the El Indio-Maricunga belt in northern Chile shallow epithermal levels (Sillitoe and Hedenquist, 2003, Fig. and Cajamarca-Huaraz belt in northern Peru (Sillitoe, 2008), 14). as well as discovery of the Resolution porphyry Cu-Mo de- At paleowater tables, near the tops of the lithocaps and posit in the southwestern North American Cu province after nearby areas, the liquid portion of the boiling high- and in- a 100-year exploration history (Manske and Paul, 2002). To termediate-sulfidation fluids follows hydrologic gradients, date, the second choice could be taken to have been less suc- whereas the H2S-bearing vapor (with H2S contributed by the cessful, as shown by the lack of economically significant dis- magma as well as SO2 disproportionation) continues its ascent coveries in the vicinities of the major but isolated Bingham, into the overlying vadose zones. There, it condenses into Butte, Pebble, and Oyu Tolgoi districts, although greenfield ground water to oxidize and produce the low-temperature, exploration is in its infancy in the still poorly defined mag- acidic fluid responsible for the blanketlike, advanced argillic matic arcs that host the last two of these. However, the El alteration zones characteristic of the steam-heated environ- Indio and Yanacocha high-sulfidation Au deposits were ini- ment (Sillitoe, 1993, 1999b; Fig. 10). tially the isolated orebodies that led to eventual definition of As the thermal regimes of porphyry Cu systems decay, shal- the El Indio-Maricunga and Cajamarca-Huaraz belts, respec- lowly generated alteration-mineralization types become tele- tively. The third choice, frontier terranes, obviously involves scoped over more deeply formed ones (e.g., Gustafson, 1978; higher risk but resulted in the recent discoveries of Pebble, Fournier, 1999; Heinrich et al., 2004; Williams-Jones and Oyu Tolgoi, and Reko Diq, for example (Bouley et al., 1995; Heinrich, 2005; Rusk et al., 2008a), thereby causing the se- Perelló et al., 2001, 2008; Kirwin et al., 2003). quence of metal remobilization and reprecipitation events The empirical relationship between well-established mag- emphasized above. Indeed, the tops of porphyry intrusions matic (including postcollisional) arcs containing major, high- may be subjected to at least four distinct alteration-mineral- grade hypogene porphyry Cu and high-sulfidation Au de- ization events, commencing with potassic and ending with ad- posits and contractional tectonic settings characterized by vanced argillic, as temperature fronts retreat downward (Fig. high surface uplift and denudation rates (see above) may 14). The resultant telescoping is potentially more extreme, prove to be a useful criterion for selection of underexplored giving rise to deep penetration of advanced argillic alteration arc segments with incompletely tested potential. Contrac- into porphyry stocks, where porphyry Cu systems undergo ei- tional settings are strongly suggested where entire arc seg- ther rapid, synhydrothermal erosion under high uplift, pluvial ments possess only minor volcanic rock volumes contempora- or glacial conditions (Fig. 14) or, perhaps less commonly, neous with the development of porphyry Cu systems, gravity-induced sector collapse of any overlying volcanic edi- particularly where lithocaps are widely preserved as evidence fices (Sillitoe, 1994; Perelló et al., 1998; Landtwing et al., for shallow erosion. Contractional settings are also likely in 2002; Carman, 2003; Heinrich, 2005; Masterman et al., 2005; belts or districts where porphyry Cu stocks or dike swarms are Rohrlach and Loucks, 2005; Pudack et al., 2009). overprinted on precursor plutons or, in island-arc settings, By the time that the late-mineral porphyry phases are where marine sedimentary rocks only slightly older than the added to porphyry Cu stocks or dike swarms, fluid ascent porphyry Cu systems have been uplifted to ~1 km or more from the parental magma chambers has all but ceased, and K above sea level (Sillitoe, 1998). In arcs where volcanic rocks and metal availability is too limited to generate appreciable are abundant, large-volume ignimbrites, indicative of caldera potassic alteration and mineralization. The only fluid present formation, are taken to seriously downgrade porphyry Cu and is of external origin and produces propylitic alteration similar related epithermal Au potential for the reason given above. to that in the earlier formed propylitic halos. Diatreme brec- The clustering or alignment of both porphyry Cu and cias are preferentially emplaced at this time because external high-sulfidation Au deposits has been shown time and again water access to late-mineral magma bodies, a requirement for to be a highly effective exploration concept. The recent major phreatomagmatic activity, is facilitated. End-stage, ground- porphyry Cu-Mo ± Au discoveries in the productive Col- water incursion into the hot porphyry Cu deposits leads to an- lahuasi (Rosario Oeste), Chuquicamata (Toki cluster; Rivera hydrite veinlet formation, in conformity with the mineral’s and Pardo, 2004; Fig. 3b), Escondida (Pampa Escondida), retrograde solubility (e.g., Rimstidt, 1997). and Los Bronces-Río Blanco (Los Sulfatos; Fig. 5a) districts of Chile are all within <1 to 3 km of the previously known de- Exploration Implications posits, as are the several porphyry Cu-Au discoveries in the Cadia district (Holliday et al., 1999) and high-sulfidation Au Target selection discoveries in the Yanacocha district (Harvey et al., 1999) that When planning exploration programs for porphyry Cu ± were made since mining commenced. In any deposit cluster Mo ± Au, skarn Cu ± Au, or high-sulfidation epithermal Au or alignment, the best deposit may be found first or only after deposits, the preeminent ore types hosted by porphyry Cu several lesser discoveries have already been made (e.g., Hugo systems, the choice is between selection of (1) mature, well- Dummett; Kirwin et al., 2003). Whether or not these and endowed Cu or Au belts, (2) emerging belts with less obvious other deposit clusters and alignments owe their existence to metallogenic credentials but having at least one important fundamental faults or lineaments (see above; Richards, 2000),
0361-0128/98/000/000-00 $6.00 28
Excursion métallogénique - Chili 2012 Références page 96 PORPHYRY COPPER SYSTEMS 29 it is often not obvious if exploration—commonly focused on lithocaps. As a consequence, the most distal and subtle ore areas of pre- or postmineral cover—should target broadly types, sediment-hosted Au on the fringes of carbonate rock- equidimensional deposit arrays or arc-parallel or arc-trans- hosted districts and Zn-Pb-Ag ± Au−bearing intermediate- verse alignments, particularly when only one or two deposits sulfidation epithermal veins and carbonate-replacement bod- have been defined. However, local structural observations, ies on the fringes of lithocaps attract less attention and may be perhaps interpreted from geophysical responses, may assist easily missed. in this regard, although this approach played no part in the It also needs to be emphasized that few porphyry Cu sys- recent discoveries noted above in the Chilean porphyry Cu tems, whatever their exposure level, contain the full spectrum districts. of potential ore types depicted in Figure 6, although the Even cursory inspection of Figure 6 shows clearly that ero- Bingham district with its porphyry Cu-Au-Mo, Cu-Au skarn, sion level is a fundamental control on the mineralization types carbonate-replacement Zn-Pb-Ag-Au, and sediment-hosted that may be anticipated to occur in porphyry Cu systems. If Au deposits (Babcock et al., 1995) and the more shallowly ex- porphyry Cu deposits concealed beneath advanced argillic al- posed Lepanto district with its porphyry Cu-Au, high-sulfida- teration are the principal target, then deeply eroded lithocaps tion Cu-Au-Ag, and intermediate-sulfidation Au-Ag-Cu de- in which quartz-pyrophyllite ± muscovite ± andalusite alter- posits (Hedenquist et al., 2001) are exceptionally well endowed ation is prominent are best selected (e.g., El Salvador). Any in this regard. Nevertheless, many systems contain only one or exposed A-type quartz veinlet stockworks overprinted by two deposit types rather than a full zonal array (Table 3), with sericitic and/or advanced argillic assemblages immediately the presence of the more distal ore types at either porphyry Cu pinpoint the spots for initial scout drilling (Sillitoe, 1995a). In or lithocap levels being independent of the size and grade of contrast, D-type veinlets may be up to 1 km laterally away the porphyry Cu deposits or prospects. Therefore, recognition from the target. Nevertheless, bearing in mind that most ob- of even weakly developed mineralization of a single type may served lithocaps are only erosional remnants, exploration help to direct exploration for potentially higher grade mineral- should focus first around their peripheries in case a porphyry ization of other types elsewhere in the system. Furthermore, Cu deposit has already been exposed. However, should the Mo- as well as Au-rich porphyry Cu deposits may be associated search be for high-sulfidation Au deposits, the shallow parts with Au-endowed lithocaps (e.g., Nevados del Famatina dis- of lithocaps may have the best potential for the discovery of trict; Lozada-Calderón and McPhail, 1996), although lithocaps large, albeit commonly low-grade orebodies. The existence of above any porphyry Cu deposit may lack appreciable high-sul- even minor erosional remnants of steam-heated horizons and fidation mineralization (e.g., Red Mountain; Corn, 1975; Quin- their chalcedonic bases, generated above and at paleowater lan, 1981), at least in their preserved parts. tables, respectively, guarantees that the appropriate near-sur- face level is preserved (Sillitoe, 1999b). Target appraisal Assessment of the likely host-rock lithologic units is also im- Notwithstanding the typical occurrence model depicted in portant during initial appraisals of porphyry Cu belts and dis- Figure 6 and taking into account the critical importance of tricts. Obviously, major skarn, carbonate-replacement, and erosion level, the innumerable variations on the porphyry Cu sediment-hosted Au deposits can only be expected where rel- genetic theme result in a broad spectrum of three-dimen- atively thinly bedded, commonly silty carbonate rocks are sional intrusion, breccia, alteration, and mineralization present. Large, high-grade porphyry Cu deposits seem to be geometries (e.g., Gustafson and Hunt, 1975). At first glance, favored by the “pressure-cooker effect” provided by imper- using representative cross sections of alteration at four high- meable wall rocks, including massive, thickly bedded carbon- grade hypogene porphyry Cu deposits as examples (Fig. 15), ate sequences (e.g., Grasberg), a situation that can also lead these varied geometries are not easy to relate to a standard to the formation of blind high-grade deposits overlain by largely geologic model. Each individual deposit or prospect must be unaltered rocks (e.g., Hugo Dummett, Ridgeway, Pueblo carefully constructed using surface mapping and core log- Viejo). Exceptionally ferrous Fe-rich rocks, relatively uncom- ging, with particular attention paid to the temporal as well as mon in most arc terranes, also appear to assist with develop- spatial relationships of its constituent parts. Only then will the ment of high hypogene Cu grades as well as maximizing the positive and negative geologic features and, hence, its overall wall rock-hosted component of the deposit (e.g., El Teniente, potential become evident. Resolution, Oyu Tolgoi). Highly permeable, noncarbonate In most magmatic arc terranes, it is roughly estimated that host rocks may promote lateral fluid channeling, which may >90 percent of explored porphyry systems lack Cu and Au lead to generation of distal ore types other than structurally concentrations with foreseeable potential, commonly because controlled veins (e.g., Andacollo). Porous and permeable vol- the ore-forming processes, from magma generation through caniclastic and epiclastic sequences also favor large-tonnage to alteration and mineralization, were less than fully opti- orebody development in the lithocap environment, especially mized (e.g., Richards, 2005). Some critical step in the genetic where they happen to be shallowly located with respect to sequence was either poorly developed or entirely missing. paleosurfaces. For example, porphyry Cu prospects containing only weakly The large size of some porphyry Cu systems, with maxi- developed potassic alteration and A-type quartz veinlets in mum radii of ~8 km (e.g., Fig. 9) and maximum areal extents their central parts, indicating a deficiency of early-stage mag- approaching 100 km2 (Singer et al., 2008), complicates their matic fluids, are typically subore grade. Similarly, proximal effective exploration because attention is unavoidably focused skarns lacking hydrous, retrograde overprints are unlikely to on the more prominently altered parts, such as pyrite-bearing host significant Cu-Au deposits. Lithocaps dominated by porphyry Cu mineralization, pyrite halos, and pyrite-rich quartz-alunite or quartz-pyrophyllite alteration but without
0361-0128/98/000/000-00 $6.00 29
Excursion métallogénique - Chili 2012 Références page 97 30 RICHARD H. SILLITOE
TABLE 3. Representative Examples of Various Mineralization-Type Combinations in Porphyry Cu Systems1
District, Porphyry Proximal Distal Carbonate- Sediment- Peripheral High- Intermediate- location deposit skarn skarn replacement Hosted veins/mantos sulfidation sulfidation Reference(s)
Bingham, Bingham Carr Fork and Zn-Pb-Ag- Barneys Babcock et al. Utah Cu-Au-Mo North Shoot (Au) bodies Canyon and (1995), Cunning- Cu-Au Melco Au ham et al. (2004)
Copper Copper East and West Lower Zn-Pb Theodore et al. Canyon, Canyon stock Zone Cu-Au2 Fortitude occurrences (1982), Wotruba Nevada Cu-Mo-Au2 Au3 et al. (1988), Cary et al. (2000)
Superior, Resolution Superior Paul and Knight Arizona Cu-Mo Cu-Ag (1995), Manske and Paul (2002)
Yanacocha, Kupfertal Yanacocha Au Harvey et al. Peru Cu-Au Norte Au occurrences (1999)
Antamina, Cu-Mo Antamina Cu- Zn-Pb-Ag Love et al. (2004), Peru occurrence Zn-Mo-Ag-Au veins Redwood (2004)
Potrerillos, Mina Vieja San Antonio Cu Jerónimo Silica Roja El Hueso Thompson et al. Chile Cu-Mo-Au Au Au Au (2004)
Andacollo, Carmen de Andacollo Reyes (1991) Chile Andacollo Oro Au Cu-Mo-Au
Lepanto, Far Southeast Lepanto Victoria Hedenquist et al. Philippines Cu-Au Cu-Au-Ag and Teresa (2001) Au-Ag-Cu
Wafi-Golpu, Wafi Cu-Au A and Link Link Zone Ryan and Vigar Papua New Zone Au Au (1999) Guinea
Sepon, Padan and Cu-Au Khanong and Discovery, R.H. Sillitoe Laos Thengkham occurrences Thengkham Nalou, etc. (unpub. repts., Mo-Cu South Cu-Au Au 1994–1999), occurrences Smith et al. (2005)
Bau, Cu-Au Cu-Au Zn-Pb Bau Au Percival et al. Malaysia occurrences occurrences occurrences (1990), Sillitoe and Bonham (1990)
Recsk, Recsk Deeps Recsk Deeps Recsk Recsk Lahóca Parád Kisvarsanyi Hungary Cu-Au-Mo Cu-Au Deeps Deeps Cu-Au Au-Ag (1988), Földessy Zn-Cu Zn-Pb and Szebényi (2008)
1 Minor occurrences italicized 2 Porphyry Cu formation likely inhibited by reduced nature of the host porphyry (Meinert, 2000) 3 Proximal to distal skarn transition spanned by Phoenix and Greater Midas pits (Cary et al., 2000) appreciable development of vuggy, residual quartz and asso- be removed and dissipated during the formation of lower tem- ciated silicification, perhaps because fluid pH was too high or perature, pyrite-bearing alteration assemblages (Gammons exposure level is too deep, are much less likely to contain and Williams-Jones, 1997; Sillitoe, 2000; Kesler et al., 2002). major high-sulfidation Au deposits, although Pueblo Viejo Thermal regimes that permit vertically extensive ore zone de- provides a salutary exception (Kesler et al., 1981; Sillitoe et velopment in potassic zones commonly have greater size po- al., 2006). tential than those that were excessively hot internally, thereby Commonly, the highest grade and most coherent porphyry inhibiting sulfide precipitation and giving rise to large, low- Cu deposits are those that retain their early porphyry phases grade or barren cores; the exception is where the resultant and potassic alteration assemblages—with which much of the shell-like orebodies are areally extensive and thick (e.g., Bing- metal content is initially introduced—in essentially unmodi- ham and Resolution; Babcock et al., 1995; Ballantyne et al., fied form. This is particularly the case for Au, which tends to 2003). The enhanced Cu ± Au tenors of many bornite-rich
0361-0128/98/000/000-00 $6.00 30
Excursion métallogénique - Chili 2012 Références page 98 PORPHYRY COPPER SYSTEMS 31
0 800E W E 500m 2600m zone
enrichment Colluvium
Supergene C o
2200m n 0m u ta c t fa u
lt
<0.5% Cu <0.5%
C >0.5% u
1800m
<0.5% C >2.5% Cu a Cu >0.5% 500m -500m
b 500m
3000E 4000E 22400N Postmineral basalt 23200N
Leached + oxidized zone 5800m 2700m RL
>0.2
%C
u
C 5400m u RL
2200m >1.5%
West fault
Cu 5000m RL c >0.5% <0.5% Cu 500m
Late-mineral porphyry
Mineralized porphyry d 500m Alteration Unaltered Propylitic Albite Potassic-propylitic transition Advanced argillic Potassic (alkali feldspar) Sericitic Potassic (biotite) Sericitic over potassic Calcic-potassic Chlorite-sericite over potassic Silica-garnet
FIG. 15. Simplified sections through high-grade hypogene porphyry Cu deposits to illustrate the wide variation in alter- ation zoning patterns and their relationship to porphyry intrusions and Cu tenor. a. El Teniente, Chile (from Cannell et al., 2005). b. Hugo Dummett North at Oyu Tolgoi, Mongolia (from Khashgerel et al., 2008). c. Chuquicamata, Chile (from Os- sandón et al., 2001). d. Ridgeway, New South Wales, Australia (from Wilson et al, 2003). Note that the high-grade (>1.5% Cu) zone at Chuquicamata includes a substantial, but unquantified, contribution from supergene enrichment.
0361-0128/98/000/000-00 $6.00 31
Excursion métallogénique - Chili 2012 Références page 99 32 RICHARD H. SILLITOE porphyry Cu deposits provide the justification for deeper drill In lithocaps, permeable lithologic units are an especially testing of both bornite-bearing and bornite-free potassic al- important control of the largest high-sulfidation epithermal teration zones that are judged to have been only relatively Au deposits, as mentioned previously, in contrast to tight shallowly explored (e.g., Esperanza; Perelló et al., 2004b). rocks, such as little-fractured lava domes and flows, which Abundant hydrothermal magnetite is a good indicator of po- typically host smaller, fault- and fracture-controlled deposits. tentially Au-rich porphyry Cu deposits (Sillitoe, 1979), and Any carbonate rocks affected by the lithocap environment the presence of banded quartz veinlets may be used to iden- may be particularly receptive. However, through-going struc- tify most, but not all, Cu-poor porphyry Au deposits (Vila and tural (fault and fracture network) and hydrothermal (phreatic Sillitoe, 1991). breccia and vuggy, residual quartz) permeability is probably Where lower grade intermineral intrusions and barren late- the most critical requirement for the development of impor- and postmineral intrusions or diatremes are volumetrically tant high-sulfidation deposits; otherwise, inadequate late- important, the original mineralized rock volumes may be stage aqueous liquids from the cooling parental magma physically disrupted and ore-zone geometries radically changed chambers gain access to the lithocaps. Lateral transfer of such and generally rendered less continuous. Where intense chlo- liquids beyond lithocaps to form intermediate-sulfidation epi- rite-sericite or sericitic alteration overprints are developed, thermal deposits is also dependent on the existence of suit- the reconstitution of potassic alteration may result in either able permeability, which in a few cases is the direct continua- reduction or complete stripping of original metal contents. tion of that utilized by the contiguous high-sulfidation Furthermore, even where appreciable Au is retained in chlo- mineralization (e.g., Colquijirca, Tintic). rite-sericite assemblages, flotation recoveries are commonly lower (<60%) than for ore from potassic zones (>80%) be- Concluding Statement cause some of the Au originally in solid solution and encapsu- Porphyry Cu deposits are arguably the most studied and lated in and attached to chalcopyrite ± bornite becomes potentially best known and understood ore deposit type (e.g., linked to introduced pyrite (Sillitoe, 2000). Seedorff et al., 2005), and their relationships with the skarn In the more highly telescoped systems, where sericitic environment have been appreciated for many years (Einaudi and/or advanced argillic assemblages overprint appreciable et al., 1981; Einaudi, 1982). Only in the last decade or so, volumes of potassic and/or chlorite-sericite alteration within however, have the physicochemical connections with the porphyry intrusions, the ensuing effects can be varied. Where high- and intermediate-sulfidation epithermal environment the sericitic alteration is superimposed on quartz veinlet within and around overlying lithocaps been clarified (e.g., stockworks, Cu contents in the form of the high sulfidation- Hedenquist et al., 1998, 2001). The current state of geologic state Cu sulfides may be increased by hypogene enrichment understanding allows explorationists to use a combination of (e.g., Wafi-Golpu). However, if the overprinted high-sulfida- empirical and genetic models with ever-increasing degrees of tion assemblages also contain appreciable arsenical sulfosalts, confidence (Thompson, 1993; Sillitoe and Thompson, 2006). a situation that becomes increasingly likely upward in most Furthermore, the current geologic knowledge base permits systems, the resultant mineralization is less desirable because meaningful deployment of sophisticated geochemical and it is not only refractory if subjected to bacterial heap leaching geophysical techniques in some exploration programs (e.g., but also generates As-rich flotation concentrates that may Kelley et al., 2006; Holliday and Cooke, 2007). prove difficult to market. Nevertheless, there is still a great deal to learn, a fact under- Although many hydrothermal breccias, like the late dia- scored by the relatively recent appreciation of the contrasting tremes mentioned above, are commonly diluents to ore, some metal contents of coexisting hypersaline liquids and vapors magmatic-hydrothermal breccias give rise to anomalously high- (Heinrich et al., 1999; Ulrich et al., 1999) and experimental grade rock volumes despite their intermineral timing. Fur- determination of volatile S complexes as potentially important thermore, magmatic-hydrothermal breccia cemented mainly Cu- and Au-transporting agents throughout porphyry Cu by quartz, tourmaline, and pyrite may be zoned downward systems (Williams-Jones et al., 2002; Nagaseki and Hayashi, over hundreds of meters to chalcopyrite-rich material, which 2008; Pokrovski et al., 2008, 2009). A short, personalized is likely to persist into any underlying biotite-cemented brec- selection of outstanding questions includes the following: cia (e.g., Los Bronces-Río Blanco; Vargas et al., 1999; Fig. 8). (1) what are the fundamental mantle and/or crustal factors Interrelationships between porphyry intrusions and car- that dictate whether youthful arc segments are endowed with bonate host rocks can influence the form and size of skarn de- giant porphyry Cu systems (e.g., central Andes), only incipi- posits, typically with above-average Cu tenors. Where steeply ently developed systems (e.g., Cascades, western United dipping, receptive carbonate rock sequences abut steep por- States), or none at all (e.g., Japan)? (2) what are cross-arc lin- phyry stock contacts, vertically extensive proximal skarn bod- eaments, and can they be demonstrated to play a truly influ- ies may form (e.g., the >1,600-m extent of the Ertsberg East ential role in the localization of porphyry Cu systems? (3) how (Gunung Bijih) Cu-Au deposit, Indonesia; Coutts et al., 1999). important are mafic magmas in the development of the Unusually large, laterally extensive proximal skarn bodies may parental magma chambers beneath porphyry Cu systems, and form preferentially where suitable carbonate host rocks abut what material contributions do they make to the systems the tops of porphyry stocks (e.g., Antamina; Redwood, 2004). themselves? (4) how is the single-phase magmatic liquid Structural permeability linking porphyry stocks to the fringes transferred from the parental magma chambers to porphyry of carbonate rock-dominated districts seems to be a require- Cu stocks or dike swarms, and what distance can be travelled ment for formation of substantial sediment-hosted Au de- by this fluid between exiting the chambers and eventual posits (e.g., Bingham district; Cunningham et al., 2004). phase separation? (5) what are the deep processes that result
0361-0128/98/000/000-00 $6.00 32
Excursion métallogénique - Chili 2012 Références page 100 PORPHYRY COPPER SYSTEMS 33
in some porphyry Cu systems being apparently short lived Arancibia, O.N., and Clark, A.H., 1996, Early magnetite-amphibole-plagio- while others may remain at least intermittently active for up clase alteration-mineralization in the Island Copper porphyry copper-gold- ≥ molybdenum deposit, British Columbia: ECONOMIC GEOLOGY, v. 91, p. to 5 m.y.? (6) why do some porphyry Cu deposits develop 402−438. large and high-grade magmatic-hydrothermal breccias, Arif, J., and Baker, T., 2004, Gold paragenesis and chemistry at Batu Hijau, whereas others have only minor examples or none at all? (7) Indonesia: Implications for gold-rich porphyry copper deposits: Mineral- if externally derived, nonmagmatic brine is responsible for at ium Deposita, v. 39, p. 523−535. least some examples of sodic-calcic alteration, how does it ac- Atkinson, W.W., Jr., Souviron, A., Vehrs, T.I., and Faunes, A., 1996, Geology and mineral zoning of the Los Pelambres porphyry copper deposit, Chile: cess the cores of some porphyry Cu deposits between early Society of Economic Geologists Special Publication 5, p. 131−156. porphyry emplacement and magmatic fluid ascent responsi- Audétat, A., Pettke, T., and Dolejš, D., 2004, Magmatic anhydrite and calcite ble for initiation of potassic alteration (and locally sinuous A- in the ore-forming quartz-monzodiorite magma at Santa Rita, New Mexico type quartz veining)? (8) what controls metal depletion versus (USA): Genetic constraints on porphyry-Cu mineralization: Lithos, v. 72, p. 147−161. enrichment during chlorite-sericite and sericitic overprints? Audétat, A., Pettke, T., Heinrich, C.A., and Bodnar, R.J., 2008, The compo- (9) what are the main mechanisms controlling the bulk sition of magmatic-hydrothermal fluids in barren and mineralized intru- Cu/Au/Mo ratios of porphyry Cu deposits? (10) why is Au sions: ECONOMIC GEOLOGY, v. 103, p. 877−908. transported to the distal limits of only a few porphyry Cu sys- Babcock, R.C., Jr., Ballantyne, G.H., and Phillips, C.H., 1995, Summary of tems for concentration in sediment-hosted deposits, and why the geology of the Bingham district: Arizona Geological Society Digest 20, p. 316−335. are most of these apparently small compared to virtually iden- Ballantyne, G., Marsh, T., Hehnke, C., Andrews, D., Eichenlaub, A., and tical Carlin-type deposits (Cline et al., 2005)? (11) what is the Krahulec, K., 2003, The Resolution copper deposit, a deep, high-grade por- fluid regime responsible for metal zoning in lithocaps, and phyry copper deposit in the Superior district, Arizona: Marco T. Einaudi why are so many lithocaps apparently barren? (12) why do Symposium, Society of Economic Geologists Student Chapter, Colorado School of Mines, Golden, CO, 2003, CD-ROM, 13 p. only a few lithocaps appear to develop intermediate-sulfida- Ballard, J.R., Palin, J.M., Williams, I.S., Campbell, I.H., and Faunes, A., tion epithermal precious metal deposits on their peripheries? 2001, Two ages of porphyry intrusion resolved for the super-giant Effective study of these and other problems will require Chuquicamata copper deposit of northern Chile by ELA-ICP-MS and field-based geochemical and geophysical work and an array of SHRIMP: Geology, v. 29, p. 383−386. evermore sophisticated laboratory equipment for high-preci- Barr, D.A., Fox, P.E., Northcote, K.E., and Preto, V.A., 1976, The alkaline suite porphyry deposits—A summary: Canadian Institute of Mining, Met- sion fluid inclusion and trace element analysis, isotopic deter- allurgy and Petroleum Special Volume 15, p. 359−367. minations, isotopic dating, and experimental work on fluid Barra, F., Ruiz, J., Valencia, V.A., Ochoa-Landín, L., Chesley, J.T., and evolution and metal transport. But more fundamentally, how- Zurcher, L., 2005, Laramide porphyry Cu-Mo mineralization in northern ever, we require better and more detailed documentation of Mexico: Age constraints from Re-Os geochronology in molybdenite: ECO- geologic relationships in porphyry Cu systems worldwide, at NOMIC GEOLOGY, v. 100, p. 1605−1616. Barton, P.B., Jr., and Skinner, B.J., 1967, Sulfide mineral stabilities, in all scales from the thin section to the entire system, and with Barnes, H.L., ed., Geochemistry of hydrothermal ore deposits: New York, greater emphasis on the regional to district context, particu- Holt, Rinehart & Winston, p. 236−333. larly the relationship to igneous evolution. And these geologic Baumgartner, R., Fontboté, L., and Venneman, T., 2008, Mineral zoning and observations must further emphasize the relative timing of in- geochemistry of epithermal polymetallic Zn-Pb-Ag-Cu-Bi mineralization at Cerro de Pasco, Peru: ECONOMIC GEOLOGY, v. 103, p. 493−537. trusion, brecciation, alteration, and mineralization events be- Baumgartner, R., Fontboté, L., Spikings, R., Ovtcharova, M., Schaltegger, U., cause isotopic dating techniques do not and may never have Schneider, J., Page, L., and Gutjahr, M., 2009, Bracketing the age of mag- the required resolution. It is acquisition of this geologic detail matic-hydrothermal activity at the Cerro de Pasco epithermal polymetallic that is going to enable better application of laboratory tech- deposit, central Peru: A U-Pb and 40Ar/39Ar study: ECONOMIC GEOLOGY, v. niques and, hopefully, further clarify the localization and evo- 104, p. 479−504. Beane, R.E., and Titley, S.R., 1981, Porphyry copper deposits. Part II. Hy- lutionary histories of porphyry Cu systems as well as the fun- drothermal alteration and mineralization: ECONOMIC GEOLOGY 75TH AN- damental controls on large size and high hypogene grade. NIVERSARY VOLUME, p. 235−269. Bendezú, R., and Fontboté, L., 2009, Cordilleran epithermal Cu-Zn-Pb-(Au- Acknowledgments Ag) mineralization in the Colquijirca district, central Peru: Deposit-scale mineralogical patterns: ECONOMIC GEOLOGY, v. 104, p. 905–944. Larry Meinert, Economic Geology editor, is thanked for the Bendezú, R., Fontboté, L., and Cosca, M., 2003, Relative age of Cordilleran invitation to contribute this overview and for his vinous base metal lode and replacement deposits, and high sulfidation Au-(Ag) inducements to get it finished. Support over the years from epithermal mineralization in the Colquijirca mining district, central Peru: numerous major and junior companies, particularly Anglo Mineralium Deposita, v. 38, p. 683−694. Bendezú, R., Page, L., Spikings, R., Pecskay, Z., and Fontboté, L., 2008, American, Antofagasta Minerals, Codelco Chile, Billiton, New 40Ar/39Ar alunite ages from the Colquijirca district, Peru: Evidence of CRA Exploration, Minera Escondida, Minorco, RGC Explo- a long period of magmatic SO2 degassing during formation of epithermal ration, and Rio Tinto Mining and Exploration, along with the Au-Ag and Cordilleran polymetallic ores: Mineralium Deposita, v. 43, p. input of innumerable explorationists worldwide deserves hon- 777−789. orable mention. Constructive comments on the manuscript Bissig, T., Lee, J.K.W., Clark, A.H., and Heather, K.B., 2001, The Cenozoic history of volcanism and hydrothermal alteration in the central Andean flat- by Chris Heinrich, Larry Meinert, Pepe Perelló, Jeremy slab region: New 40Ar-39Ar constraints from the El Indio-Pascua Au (-Ag, Richards, Alan Wilson, and the Economic Geology reviewers, Cu) belt, 29°20'-30°30' S: International Geology Review, v. 43, p. 312−340. Regina Baumgartner, John Dilles, and Jeff Hedenquist, led to Bodnar, R.J., 1995, Fluid-inclusion evidence for a magmatic source for met- clarification of ideas and substantial improvements. als in porphyry copper deposits: Mineralogical Association of Canada Short Course Series, v. 23, p. 139−152. Bouley, B.A., St. George, P., and Wetherbee, P.K., 1995, Geology and dis- REFERENCES covery at Pebble Copper, a copper-gold porphyry system in northwest Ambrus, J., 1977, Geology of the El Abra porphyry copper deposit, Chile: Alaska: Canadian Institute of Mining, Metallurgy and Petroleum Special ECONOMIC GEOLOGY, v. 72, p. 1062–1085. Volume 46, p. 422−435.
0361-0128/98/000/000-00 $6.00 33
Excursion métallogénique - Chili 2012 Références page 101 Excursion métallogénique - Chili 2012 Références page 102 JOURNAL OF PETROLOGY VOLUME 52 NUMBERS 7 & 8 PAGES 1591^1617 2011 doi:10.1093/petrology/egq029
Magmatic Evolution of the Giant El Teniente Cu^Mo Deposit, Central Chile
CHARLES R. STERN1*, M. ALEXANDRA SKEWES1 AND ALEJANDRA ARE¤VA LO 2
1DEPARTMENT OF GEOLOGICAL SCIENCES, UNIVERSITY OF COLORADO, BOULDER, CO 80309-0399, USA Downloaded from 2SUPERINTENDENCIA GEOLOGI¤ A, EL TENIENTE, CODELCO-CHILE, RANCAGUA, CHILE
RECEIVED JULY 7, 2009; ACCEPTED MAY 6, 2010 ADVANCE ACCESS PUBLICATION JUNE 28, 2010 http://petrology.oxfordjournals.org/
ElTeniente, the world’slargest Cu deposit, is hosted in Late Miocene volumes of more fractionated, but isotopically equivalent, Late and Pliocene plutons that intrude the older Teniente Volcanic Miocene and Pliocene felsic plutonic rocks that host the deposit Complex (or Farellones Fm; 14·2^6·5 Ma). The Late Miocene were derived from the roof of a large, long-lived, thermally and and Pliocene plutonic host rocks of the deposit include, sequentially, chemically stratified, open-system magma chamber, or magmatic the relatively large (450 km3) Teniente Mafic Complex laccolith plumbing system, recharged from below by mantle-derived magmas. (8 ·9 1·4 Ma),the smaller ( 30 km3) Sewell equigranular tonal- Only when this system fully solidified did post-mineralization ite complex (7·05 0·14 Ma) and associated andesitic sills mafic olivine-hornblende-lamprophyre dikes (3·85 0·18 to
(8 ·2 0·5to6·6 0·4 Ma), small dacitic porphyry stocks 2·9 0·6 Ma) pass through the system from the mantle to the sur- at Universite du Quebec a Chicoutimi on April 21, 2012 (51km3;6·09 0·18 Ma), the unusual Cu- and S-rich ‘Porphyry face. The significant progressive temporal isotopic evolution, to 3 87 86 A’anhydrite-bearing granitoid stock (51km ;5·67 0·19 Ma),the higher Sr/ Sr (from 0·7033 to 0·7049) and lower eNd (from Teniente Dacite Porphyry dike (51km3;5·28 0·10 Ma), minor þ6·2to 1·1), that occurred between the Late Oligocene and latite dikes (4·82 0·09 Ma),and finally a small dacite intrusion Pliocene in the vicinity of El Teniente for mafic mantle-derived (4 ·58 0·10 Ma).These plutonic rocks are all isotopically similar magmas, and by implication their sub-arc mantle-source region, 87 86 to each other ( Sr/ Sr ¼ 0·7039^0·7042; eNd ¼þ2·5toþ3·5) was due in part to increased mantle-source region contamination by and also to the Teniente Volcanic Complex extrusive rocks, but dis- subducted crust tectonically eroded off the continental margin. The tinct from both older Late Oligocene to Early Miocene volcanic post-mineralization olivine-hornblende-lamprophyres also imply ex- 87 86 rocks ( Sr/ Sr ¼ 0·7033^0·7039; eNd ¼þ3·8toþ 6·2) and tensive hydration of the mantle below this portion of the Andean arc younger Pliocene post-mineralization mafic dikes and lavas by the Pliocene, which may have played a role in producing oxidized 87 86 ( Sr/ Sr ¼ 0·7041^0·7049; eNd ¼þ1·1to 1·1). Multiple volatile-rich magmas and mineralization at El Teniente. Cu-mineralized magmatic^hydrothermal breccia pipes were emplaced into these plutonic rocks during the same time period as the felsic porphyry intrusions, between at least 6·31 0·03 and KEY WORDS: Chilean Andes; copper deposits; ElTeniente; magmagen- 4·42 0·02 Ma.These mineralized breccia pipes, which formed by esis; petrochemistry exsolution of magmatic fluids from cooling plutons, have their roots below the deepest level of mining and exploration drilling and were derived from the same magma chamber as the felsic porphyries, INTRODUCTION 44 km below the paleosurface.Toproduce the 100 106 tonnes of The giant El Teniente Cu^Mo deposit, located in central Cu in the deposit requires a batholith-size (4600 km3) amount of Chile (348050S, 70821 0W; Fig. 1), is the world’s largest Cu de- magma with 100 ppm Cu. We suggest that both the mineralized posit, originally containing 100 million metric tonnes magmatic^hydrothermal breccias and the progressively smaller (Mt) of Cu (Skewes et al., 2002, 2005). The deposit is
The Author 2010. Published by Oxford University Press. All *Corresponding author. Telephone: 303-492-7170. Fax: 303-492-2606. rights reserved. For Permissions, please e-mail: journals.permissions@ E-mail: [email protected] oup.com
Excursion métallogénique - Chili 2012 Références page 103 JOURNAL OF PETROLOGY VOLUME 52 NUMBERS 7 & 8 JULY & AUGUST 2011 Downloaded from http://petrology.oxfordjournals.org/
Fig. 1. Location maps of El Teniente and the two other giant Late Miocene and Pliocene Cu deposits in the Andes of central Chile, Los Pelambres and R|¤o Blanco^Los Bronces, modified from Serrano et al. (1996). (a) Map showing tectonic features such as the location of the
Chile Trench, which is the boundary between the Nazca and South American plates, and the depth in kilometers (100 and 150 km dashed at Universite du Quebec a Chicoutimi on April 21, 2012 lines) to the Benioff Zone below South America. The Cu deposits occur just east of the locus of subduction of the Juan Ferna¤ ndez Ridge. This also marks the boundary between the Andean Flat-Slab segment, below which the subduction angle is very low, as indicated by the depths to the upper boundary of the subducted slab, and the Southern Volcanic Zone (SVZ) of active volcanoes (m), below which the subduction angle is steeper. The map also shows the location of the Central Volcanic Zone (CVZ) and some of the Late Eocene and Early Oligocene Cu deposits in northern Chile. (b) Simplified regional geology of central Chile. In this schematic map, both the Coya-Machali (Abanico) and Farellones Formations are included together in the belt of Late Tertiary volcanic rocks.
hosted in Late Miocene and Pliocene igneous rocks (Figs 2 ‘preserves a record of alternating igneous activity and ore and 3). The chronology of the formation of these igneous deposition that affords convincing evidence of the intimate rocks, both in and around the deposit, is reasonably well genetic connection between igneous rocks and ore de- determined (Fig. 4; Cuadra, 1986; Kurtz et al., 1997; posits’. As a contribution to the understanding of this gen- Maksaev et al., 2004), and the published petrochemical etic connection, we present the results of a petrochemical database for understanding the regional magmatic evolu- study of the igneous rocks within the El Teniente deposit. tion at the latitude of El Teniente is among the most de- tailed in the southern Andes (Stern & Skewes, 1995; Nystrom et al., 2003; Kay et al., 2005; Mun‹ oz et al., 2006). GEOLOGICAL BACKGROUND However, the petrochemistry and genesis of the igneous El Teniente, along with Los Pelambres (328S;425 Mt of rocks within the El Teniente deposit have received less at- Cu; Atkinson et al., 1996) and R|¤o Blanco^Los Bronces tention than those surrounding the deposit. It is clear, (338S; 450 Mt of Cu; Warnaars et al., 1985; Serrano et al., from the overlap of the isotopic ages of both the igneous 1996; Frikken et al., 2005), formed during Miocene and rocks (8·9 1·4to4·58 0·10 Ma; Fig. 4) and the multiple Pliocene times in the Andes of central Chile (Fig. 1), and alteration and mineralization events in the deposit are among the youngest and largest Cu^Mo deposits in (6·31 0·03 to 4·42 0·02 Ma), that there is a direct genet- the Andes. They are copper-, sulfur-, iron-, calcium-, mo- ic relation between igneous activity and mineralization. lybdenum- and boron-rich, but gold-poor deposits that This was originally recognized by Lindgren & Bastin share important features. These include their large tonnage (1922), who identified several periods of intrusion and min- and high hypogene copper grade, and the fact that most eralization at El Teniente, and concluded that this deposit of the copper mineralization occurs as primary hypogene
Excursion métallogénique - Chili 2012 Références page 104
1592 STERN et al. EVOLUTION OF EL TENIENTE CU^MO DEPOSIT Downloaded from http://petrology.oxfordjournals.org/ at Universite du Quebec a Chicoutimi on April 21, 2012
Fig. 2. Geological map of the area surrounding the El Teniente copper deposit, modified from Morel & Spro« hnle (1992) and Skewes et al.(2002, 2005).The Braden breccia pipe is located at the intersection of the Teniente Fault Zone, which trends NE^SW between the Teniente River and Agua Amarga fault, and the NW^SE-trending Puquios^Codegua fault.
ore. Another distinctive feature that each of these three (Fig. 1a): the Flat-Slab segment to the north, below which deposits has in common is the presence of large magmat- the angle of subduction has decreased significantly since ic^hydrothermal breccias, both mineralized and unminer- the Miocene and where volcanism is now absent, and the alized (Skewes & Stern, 1994, 1995, 1996). The genesis of Southern Volcanic Zone (SVZ), below which the subduc- these magmatic^hydrothermal breccias has been attribu- tion angle is steeper and volcanism is active. The formation ted to the exsolution of high-temperature magmatic fluids of the three deposits is closely associated in time with the from cooling plutons (Warnaars et al., 1985; Skewes & changing geometry of subduction that has produced this Stern, 1994, 1995, 1996; Vargas et al., 1999; Skewes et al., segmentation of the Andes (Stern, 1989, 2004; Skewes & 2002, 2003). Stern, 1994, 1995; Stern & Skewes, 1995, 2005; Rosenbaum The three deposits in central Chile occur across the et al., 2005). As with the older copper deposits in northern boundary between two major Andean tectonic segments Chile, such as Chuquicamata and El Salvador (Fig. 1a),
Excursion métallogénique - Chili 2012 Références page 105
1593 JOURNAL OF PETROLOGY VOLUME 52 NUMBERS 7 & 8 JULY & AUGUST 2011 Downloaded from http://petrology.oxfordjournals.org/ at Universite du Quebec a Chicoutimi on April 21, 2012
Fig. 3. Geological map of level Teniente 5 (2284 m above sea level) in the mine, modified from Skewes et al. (2002, 2005).The Dacitic Porphyries, north of the Sewell Tonalite, are mapped as a distal portion of this pluton, although they are younger (Maksaev et al., 2004) and have an inde- pendent origin (Guzma¤ n, 1991; Hitschfeld, 2006). The spatial extent of biotite breccias is projected onto this level from where they have been mapped between levels Teniente 4 and 8.
Excursion métallogénique - Chili 2012 Références page 106
1594 STERN et al. EVOLUTION OF EL TENIENTE CU^MO DEPOSIT Downloaded from http://petrology.oxfordjournals.org/
Fig. 4. Schematic chronology of the ages of the Late Miocene and Pliocene volcanic and plutonic host rocks of the El Teniente deposit, includ- ing both K^Ar mineral and whole-rock ages ( ) and U^Pb in zircon ages (g), as well as Re^Os ages of mineralization (¨) and the ages of breccias determined either from Re^Os of molybdenite within the breccia matrix mineral assemblage or, in the case of the Braden Pipe, from K^Ar of sericite formed by alteration of clasts and wall-rock during breccia emplacement.Vertical lines connect igneous rocks with contempor- aneous breccias. Horizontal lines indicate range of ages for regional units and uncertainties for specific rocks. Sources of data are Cuadra (1986), Maksaev et al. (2004), Cannell et al. (2005) and Kay et al. (2005). Skewes & Stern (2007) and Stern et al. (2007) discussed uncertainties in at Universite du Quebec a Chicoutimi on April 21, 2012 the complete published dataset and the cross-cutting field relations that provide the basis for the selection of ages in this figure.
copper mineralization occurred during a relatively re- Teniente mine. He concluded that the Miocene extrusive stricted time interval ( 3 Myr; Fig. 4), at the end of a rocks of the Farellones Formation in the vicinity of El more extended period of magmatic activity (410 My r), Teniente ranged in age from 14 to 8 Ma, and felsic intrusive just prior to the eastward migration of the locus of the rocks within the deposit from 7·4to4·6 Ma. Recently Andean volcanic arc (Maksaev & Zentilli, 1988; Cornejo determined U^Pb in zircon ages for the igneous rocks in et al., 1997). Eastward migration of the magmatic arc the deposit (Maksaev et al., 2004), and Re^Os mineraliza- occurred in central Chile during the Late Miocene and tion ages of molybdenite (Maksaev et al., 2004; Cannell Pliocene, as the angle of subduction decreased (Stern, et al., 2005), are summarized in Fig. 4. 1989, 2004; Stern & Skewes, 1995, 2005; Kay et al., 2005). As the subduction angle decreased, beginning in the Middle Miocene, the crust in central Chile was deformed METHODS and thickened (Godoy et al., 1999; Giambiagi et al., 2001; Samples were collected from both surface localities around Giambiagi & Ramos, 2002), and uplifted and eroded the mine, and from drill core within the hypogene zone (Skewes & Holmgren, 1993; Kurtz et al., 1997; Kay et al., of the mine. Table 1 and Table 1ES in the Electronic 2005). Supplement (available for downloading at http://www.pet- The first comprehensive geological description of El rology.oxfordjournals.org/) list primary igneous minerals Teniente was that of Lindgren & Bastin (1922). They and alteration phases for each sample analyzed, as well as described this deposit as hosted in a sill, formed of andesite the intensity of alteration. This ranges from weak (fresh porphyry and quartz diorite, which was intruded into a plagioclase, some mafic minerals preserved, and igneous thick pile of volcanic rocks. Cuadra (1986) presented a texture intact) to strong (plagioclase partially altered, basic chronology of the development of the deposit based mafic phases replaced, and igneous textures partially to to- on K^Ar dates on extrusive and intrusive igneous rocks, tally obliterated). Petrological descriptions of each rock breccias and alteration events in and surrounding the El unit sampled are presented in the Electronic Supplement.
Excursion métallogénique - Chili 2012 Références page 107
1595 Excursion
métallogénique Table 1: Sample locations, textures and rock names, and primary igneous and secondary alteration mineralogy
Sample or Location* Igneous features Alteration
DDH no. N or S E or W Elevation (m) Texture/rock Igneous minerals Intensity Secondary minerals - ORA FPETROLOGY OF JOURNAL Chili
Mafic intrusive rocks inside the mine (CMET; 8·9 1·4Ma) 2012 DDH540-2082’ 607N 1358E 1766 Gabbro Pl, Cpx, r-Maf Moderate Chl, Act, Bt DDH1411-1630’ 565S 430E 1910 Gabbro Pl(An8–55), r-Px, r-Ol(?) Moderate Bt, Chl
Mafic sills outside the mine (CMET; 8·9 1·4Ma) EX2004-4 34·093S 70·443W Surface Porph basalt Pl(An77–51), Ol(Fo74), Cpx, Ti-Mag Fresh QT-4 6·232·000N 376·000E Surface Diabase Pl Strong Chl, Ep, Cal AS-99-1C 34·097S 70·424W Surface Porph basalt Pl(An51–40), Bt, Mag, r-Px, Fe–Ti Ox Weak Chl
Cerro Montura andesitic sills (8·2 0·5and6·6 0·4Ma) TTc9 6·229·640N 366·730E Surface Andesite Pl(An62–50), Opx, Am, Fe–Ti Ox Fresh
TTc10 6·227·900N 364·600E Surface Andesite Pl(An46–43), Hb, Cpx Fresh 52 VOLUME
Références Sewell Tonalite (7·05 0·14 Ma) 400S/1780E 400S 1780E 2165 Tonalite Pl. Bt, r-Am, Qtz Weak Act, Chl, Bt, Ttn, Ep, Mag 1596 Ttc5 130N 1770E 2284 Tonalite Pl(An34–15), Bt, r-Maf, Qtz Moderate Chl, Ep
Northern Dacite Porphyry (6·11 0·13 Ma) UBR 8 & 7 NUMBERS DDH2370-159 990N 1000E 1920 Dacitic porph Pl(An30–19), r-Am, Qtz Strong Qtz, Chl, Ser, Cal Central Dacite Porphyry (6·08 0·22 Ma) DDH1824-655’ 314N 1340E 2180 Dacitic porph Pl, Bt, r-Maf Strong Qtz, Ser, Cal
Porphyry A (5·67 0·19 Ma) DDH1446-266’ 10N 1650E 2220 ‘Microdiorite’ Pl(An7), K-feld, Qtz, Anh, Bt Weak Chl, Ser DDH1473-970’ 100N 1650E 2000 ‘Andesite’ Pl(An32), Bt, Qtz, Anh Weak UY&AGS 2011 AUGUST & JULY Teniente Dacite Porphyry (PDT; 5·28 0·10 Ma) DDH1134-79’ 1012N 520E 2145 Porph dacite Pl(An24–7), Bt, Qtz, Anh Weak Bt, Ser DDH1134-302’ 1010N 445E 2116 Porph dacite Pl(An27–7), Bt, r-Am, Qtz Weak Ser, Chl, Cal DDH1134-365’ 1006N 437E 2105 Porph dacite Pl(An24–7), Bt, r-Maf, Qtz, K-feld, Anh Weak Chl, Cal, Ser, Anh
Latite dikes (4·82 0·09 Ma) DD1394-92’ 414S 430E 2270 Porph latite Pl(An26–8), r-Maf Moderate Ser, Chl, Cal, Anh
And, andesite; Bas, basalt; Porph, porphyry; Ac, actinolite; Am, amphibole; Anh, anhydrite; Bt, biotite; Cal, calcite; Chl, chlorite; Cpx, clinopyroxene; En, enstatite; Ep, epidote; Hbl, hornblende; Hem, hematite; Idd, iddingsite; K-feld, potassium feldspar; Maf, mafic mineral; Mag, magnetite; Ol, olivine; Opx, orthopyroxene; Or, orthoclase; Ox, oxides; Pl, plagioclase; Px, pyroxene; Qtz, quartz; Ser, sericite; Ttn, titanite; Tur, tourmaline; r, replaced; Weak, plagioclase fresh, some mafic
page minerals preserved, texture intact; Moderate, plagioclase preserved, mafic minerals replaced, texture preserved; Strong, plagioclase can be altered, mafic minerals replaced, partial to total destruction of texture.
108 *Locations inside the mine are given in mine coordinates (Fig. 3). Locations outside the mine are given in UTM or degrees latitude.
Downloaded from from Downloaded http://petrology.oxfordjournals.org/ at Universite du Quebec a Chicoutimi on April 21, 2012 21, April on Chicoutimi a Quebec du Universite at STERN et al. EVOLUTION OF EL TENIENTE CU^MO DEPOSIT
Samples were selected from areas of low vein density (0·7033^0·7039) and eNd (þ3·8toþ 6·2; Figs 5 and 6). and all visible veins were removed from them prior to ana- Although the Coya-Machal|¤ (Abanico) Formation volcan- lysis. In their study of volcanic and plutonic rocks from ic rocks do not crop out either within or in the immediate the Teniente area, Kay et al. (2005) concluded that hydro- vicinity of the El Teniente deposit (Figs 2 and 3), these thermal alteration may have affected loss on ignition rocks occur both to the west and the east of the deposit, (LOI), K2O, Na 2O, Rb, Cs and Ba, but that even strong and they almost certainly also occur at depth below the alteration has not affected the immobile element [rare deposit. earth element (REE), Sr, U,Th, Hf, Nb and Y] chemistry. Extrusive rocks of the Miocene Farellones Formation, lo- Our discussion of the petrogenesis of the rocks focuses on cally referred to as the Teniente Volcanic Complex (Kay constraints provided by these immobile elements and iso- et al., 2005), are the oldest rocks exposed in the immediate topic data that were obtained only for the freshest, least area surrounding the deposit (Fig. 2). The Farellones altered samples. Formation is a sequence of42500 m of lavas, volcanoclastic Major- and trace-element compositions (Tables 2 and 3, rocks, dikes, sills and stocks of basaltic to rhyolitic compos- Downloaded from and Tables 2ES^6ES in the Electronic Supplement) were ition (Vergara et al., 1988; Rivano et al., 1990).The Teniente determined by Actlabs. The Sr and Nd isotopic (Table 4) Volcanic Complex near the deposit has been correlated compositions of 17 selected samples were determined by with the upper part of this formation and dated between solid-source mass-spectrometry techniques (Farmer et al., 14·4 and 6·5 Ma (Cuadra, 1986; Vergara et al., 1988; Kay 1991) at the University of Colorado. Mineral compositions et al., 2005). No extrusive rocks with ages less than 6·5Ma http://petrology.oxfordjournals.org/ were determined using a JEOL electron microprobe at the have been found in the vicinity of the El Teniente deposit. University of Colorado. The extrusive rocks of the Teniente Volcanic Complex were intruded by gabbro, diabase, diorite, tonalite, latite, IGNEOUS ROCKS and dacite porphyry plutons between 12·3 and 4·6Ma Crystalline basement (Cuadra, 1986; Kurtz et al., 1997; Maksaev et al., 2004; Kay et al., 2005). Kurtz et al. (1997) and Kay et al.(2005)divided Paleozoic igneous and metamorphic rocks occur both well these intrusive rocks into the Teniente Plutonic Complex to the west of El Teniente, along the Pacific coast (Fig. 1b), (12 ·3^7·0 Ma) and the Younger Plutonic Complex (6·6to and also to the east, in the High Andean Cordillera along 5 Ma), the latter emplaced after cessation of volcanic ac- the drainage divide between Chile and Argentina. The at Universite du Quebec a Chicoutimi on April 21, 2012 tivity in the region (Fig. 4). As discussed below in more Paleozoic basement west of El Teniente is intruded by detail, the emplacement of the plutonic rocks within the Mesozoic plutonic rocks. These older metamorphic and El Teniente deposit spans the range from 8·9 1·4to igneous rocks may occur in the deep crust below El 4·58 0·10 Ma, and therefore, from an age perspective, Teniente, but they do not crop out within the mine or in they correspond to both the Teniente and the Younger the immediate vicinity surrounding the deposit (Figs 2 Plutonic Complex rocks. and 3). The Teniente Volcanic Complex consists of tholeiitic to Tertiary volcanic rocks calc-alkaline extrusive rocks, which plot in the medium- El Teniente is located in a belt of Middle to Late Miocene to high-K group of convergent plate boundary arc extrusive and intrusive igneous rocks (Fig. 1b), which are magmas (Kay et al., 2005). Mafic rocks of the Teniente part of the Farellones Formation (Fig. 2). Extrusive rocks Volcanic Complex generally have higher La/Yb (4·4^9·2) of the Farellones Formation overlie older continental vol- compared with rocks of the older Coya Machal|¤ canic rocks, up to 3300 m thick, of the Oligocene to Early Formation (Kay et al., 2005), and mafic, intermediate and 87 86 Miocene Coya-Machal|¤ (or Abanico) Formation (415 Ma; silicic rocks also have higher initial Sr/ Sr (0·7039^ Charrier et al., 2002; Mun‹ oz et al., 2006; Montecinos et al., 0·7041) and lower initial eNd (þ2·7toþ 3·6; Figs 5 and 6). 2008), which were initially uplifted and deformed begin- These differences are interpreted to represent a change ning in the Early Miocene (19^16 Ma; Kurtz et al., 1997; from magma genesis below and within relatively thin con- Kay et al., 2005), and again more strongly in the Late tinental crust during the mid-Tertiary, when the Coya Miocene and Pliocene (9^3·5 Ma; Godoy et al., 1999). Machal|¤ Formation was generated, to conditions of thick- Mafic Coya-Machal|¤ (or Abanico) Formation volcanic ened continental crust when the Teniente Volcanic rocks were formed below thin (530 km) crust, or in a Complex formed in the Middle to Late Miocene. transtensional intra-arc basin (Godoy et al., 1999), by rela- Teniente Plutonic Complex rocks have isotopic compos- tively high degrees of partial melting of sub-arc mantle itions similar to Teniente Volcanic Complex extrusive modified to a small degree by the influx from below of rocks, whereas Younger Plutonic Complex felsic granitoids 87 86 slab-derived components (Nystrom et al., 2003; Kay et al., have higher Sr/ Sr (0·70424^0·70441) and lower eNd 2005; Mun‹ oz et al., 2006). This is indicated by their low (þ0·74 to 0·08; Kurtz et al., 1997; Kay et al., 2005). La/Yb ratios (2·3^5·6), as well as their initial 87Sr/86Sr However, as discussed below, the felsic plutonic rocks
Excursion métallogénique - Chili 2012 Références page 109
1597 JOURNAL OF PETROLOGY VOLUME 52 NUMBERS 7 & 8 JULY & AUGUST 2011
Table 2: Major- and trace-element compositions of samples of CMET,Sewell Tonalite and andesite sills
CMET inside mine CMET outside mine Andesite sills Sewell Tonalite
Sample: 540-20820 1411-16300 EX2004-04 QT-4 AS-99-1C TTc9 TTc10 400S/1780E TTc5
SiO2 51·30 51·08 51·94 49·65 59·78 60·90 61·50 62·05 63·70
TiO2 1·07 1·09 0·92 1·00 0·74 0·75 0·77 0·57 0·39
Al2O3 18·68 17·07 18·19 17·96 17·67 16·80 16·90 18·38 17·10
Fe2O3 9·89 11·04 3·50 3·90 3·60 1·71 1·90 FeO 8·78 9·90 3·46 2·20 1·90 1·72 2·40
MnO 0·15 0·08 0·13 0·53 0·20 0·10 0·08 0·07 0·08 Downloaded from MgO 5·23 5·75 5·71 5·29 1·72 2·80 2·60 1·58 1·50 CaO 9·00 6·58 7·87 6·56 4·12 5·60 5·70 4·41 3·90
Na2O3·21 2·15 3·82 4·30 3·97 4·38 4·54 5·98 4·93
K2O1·00 2·56 1·01 1·03 2·05 2·50 2·50 1·47 2·20
P2O5 0·22 0·19 0·20·22 0·20·20 0·22 0·19 0·21 http://petrology.oxfordjournals.org/ LOI 1·07 1·46 0·98 4·53 2·40·94 0·60 1·50 1·80 Total 100·82 99·05 99·55 100·97 99·80 101·07 101·00 99·73 100·17 Cs 5·3187 0·811·72·83·87·46·1 Rb 54 159 26 29 73 82 74 60 96 Sr 535 383 468 595 547 543 700 946 699 Ba 158 100 235 310 517 539 504 501 528 La 9·211·413·411·515·819·116·716·414·4 Ce 20·326·430·326·534·044·938·833·334·7 Pr 2·80 3·40 3·72 3·20 4·22 4·32 3·97 at Universite du Quebec a Chicoutimi on April 21, 2012 Nd 13·315·516·615·818·925·021·916·219·2 Sm 3·30 3·71 4·00 3·90 3·80 5·13 4·26 2·90 3·39 Eu 1·29 1·04 1·22 1·16 1·05 1·18 1·00 0·88 0·88 Gd 3·50 3·11 3·70 3·40 3·40 1·90 2·90 Tb 0·60 0·51 0·60 0·60 0·50 0·60 0·41 0·20 0·29 Dy 3·30 3·00 3·40 3·30 3·10 1·00 1·80 Ho 0·60 0·60 0·70 0·70 0·60 0·20 0·30 Er 1·90 1·60 1·90 2·00 1·80 0·50 0·90 Tm 0·30 0·24 0·28 0·27 0·25 0·07 0·12 Yb 1·70 1·48 1·70 1·60 1·50 2·03 1·27 0·40 0·92 Lu 0·24 0·21 0·25 0·25 0·22 0·26 0·16 0·06 0·11 Y17 151517182115611 Zr 73 56 86 74 164 158 133 104 94 Hf 1·91·72·72·54·94·84·32·73·2 Th 1·31·23·72·911·197·91·74·3 Sc 27 30 21 24 14 15 12 6 6 Cr 263 292 268 41 50 24 Ni 75 198 134 30 520 520 Cu 204 1325 115 20 125 250 480 S 700 11120 400 200 300 400 La/Yb 5·47·77·97·210·59·413·141·015·6 Dy/Yb 1·94 2·00 2·00 2·06 2·07 2·51·96 Sr/Y 33 26 31 35 30 26 36 157 63
Excursion métallogénique - Chili 2012 Références page 110
1598 STERN et al. EVOLUTION OF EL TENIENTE CU^MO DEPOSIT
Table 3: Major- and trace-element compositions of samples of felsic intrusive rocks
Dacitic porphyries Porphyry A Teniente Dacite Porphyry Latite dike
Sample: 2370-159 1824-6550 1473-9700 1446-266 1300-4030 1134-3020 1134-790 1394-920
SiO2 63·96 62·90 48·45 56·34 66·26 64·42 59·89 60·93
TiO2 0·38 0·43 1·01 0·42 0·38 0·53 0·61 0·63
Al2O3 16·30 16·16 18·62 13·86 16·85 17·76 16·73 17·24
Fe2O3 2·57 1·43 7·54 2·74 1·83 0·91·70 2·14 MnO 0·04 0·04 0·03 0·02 0·02 0 0·01 0·02
MgO 1·08 0·89 4·73 1·41 1·01 1·02 2·22 1·86 Downloaded from CaO 2·70 2·74 5·73 6·72 2·82 3·69 4·77 4·61
Na2O5·19 3·84 4·33 3·80 5·53 6·85·55 4·76
K2O3·11 5·98 3·63 5·68 2·64 2·67 2·93 2·23
P2O5 0·14 0·14 0·22 0·17 0·14 0·20·17 0·21
S1·44 3·35 http://petrology.oxfordjournals.org/ LOI 3·44 4·89 4·66 7·99 1·95 3·35 4·83 5·74 Total 98·91 99·45 100·39 102·49 99·43 101·34 99·40 100·37 Cs 7·02·916·35·31·53·93·92·7 Rb 124 142 209 187 66·754·68780·5 Sr 568 415 499 456 699 823 680 742 Ba 564 815 244 601 644 464 762 311 La 12·813·310·624·110·415·15 13·914·6 Ce 24·324·021·453·720·130·74 27·930·9 Pr 2·66 2·47 2·84 6·29 2·42 3·92 3·68 4·00 at Universite du Quebec a Chicoutimi on April 21, 2012 Nd 10·27·40 13·43 26·12 9·015·11 14·115·3 Sm 1·83 1·00 3·44 4·63 1·58 2·51 2·80 2·80 Eu 0·63 0·36 1·00 0·79 0·65 0·69 0·78 0·83 Gd 1·26 0·60 2·96 3·17 1·40 1·90 2·20 2·30 Tb 0·16 0·51 0·38 0·20 0·17 0·30 0·20 Dy 0·82 0·40 2·69 1·89 0·80 0·76 1·50 1·00 Ho 0·15 0·50 0·34 0·18 0·13 0·30 0·20 Er 0·43 0·20 1·39 0·94 0·50 0·35 0·80 0·50 Tm 0·06 0·19 0·13 0·07 0·04 0·12 0·07 Yb 0·42 0·30 1·12 0·79 0·40 0·28 0·80 0·40 Lu 0·07 0·04 0·16 0·11 0·06 0·05 0·11 0·07 Y631395486 Zr 96 95 95 81 94 95 90 108 Hf 2·82·72·83·92·42·62·52·9 Th 3·42·80·85·92·82·72·62·3 Sc 4 3 21 2·64498 Cr 520 520 520 520 11 58 40 42 Ni 520 520 520 520 34 78 520 86 Cu 2210 2520 899 2380 6000 2410 310 1250 S 7300 6500 14400 33500 9200 8600 2800 1420 La/Yb 30·844·39·530·523·754·117·336·5 Dy/Yb 2 2·52·42 2·41·88 2·25 Sr/Y 95 140 37 51 160 206 85 124
Excursion métallogénique - Chili 2012 Références page 111
1599 JOURNAL OF PETROLOGY VOLUME 52 NUMBERS 7 & 8 JULY & AUGUST 2011
Table 4: Sr and Nd isotopic compositions of igneous rocks associated with ElTeniente Cu^Mo deposit
87 86 87 86 143 144 Sample no. SiO2 Rb Sr ( Sr/ Sr)m ( Sr/ Sr)i Nd Sm ( Nd/ Nd)m eNd
Teniente Mafic Complex (CMET; 8·9 1·4Ma) Ex2004-04 51·9 26 468 0·704070 09 0·70405 16·64·00 0·512782 09 þ2·9 1411-1630A 50·1 71 472 0·703992 08 0·70396 14·33·50 0·512818 07 þ3·5 540-2082 51·3 54 535 0·704054 11 0·70404 13·33·30 0·512813 11 þ3·4 QT-4 49·7 29 595 0·704286 14 0·70421 15·83·90 0·512798 09 þ3·1 AS-99-1c 57·8 73 547 0·704032 16 0·70402 18·93·80 0·512803 06 þ3·2
Cerro Montura andesite sills (8·2 0·5and6·6 0·4Ma) Downloaded from TTc9 61·0 82 543 0·70396 08 0·70391 25·05·13 0·51273 16 þ2·7 TTc10 61·5 74 700 0·70390 07 0·70385 21·94·26 0·51279 12 þ3·0
Sewell Tonalite (7·05 0·14 Ma) TTc5 63·7 96 699 0·703930 07 0·70386 19·23·39 0·512770 04 þ2·7
400S/1780E 62·1 60 946 0·704010 12 0·70404 16·22·90 0·512759 36 þ2·5 http://petrology.oxfordjournals.org/
Northern and Central Dacitic Porphyries (6·09 18 Ma) 2370-159 64·0 124 392 0·704095 12 0·70409 8·31·44 0·512802 12 þ3·2 1824-655 62·9 142 415 0·704089 12 0·70408 7·41·00 0·512767 08 þ2·7
Porphyry A (5·67 0·19 Ma) 1473-970 48·5 209 499 0·704105 10 0·70409 13·43·44 0·512799 07 þ3·1 1446-266 56·3 187 456 0·704104 12 0·70406 26·14·63 0·512795 09 þ3·1
Teniente Dacite Porphyry (PDT; 5·28 0·10 Ma) 1134-302 64·4 65 755 0·704072 12 0·70406 14·42·39 0·512757 14 þ2·5 at Universite du Quebec a Chicoutimi on April 21, 2012 1300-403 66·3 67 699 0·704027 12 0·70402 9·01·58 0·512762 27 þ2·6 1134-79 59·9 87 680 0·704047 13 0·70404 14·12·80 0·512785 15 þ2·9
Latite dike (4·82 0·09 Ma) 1394-92 60·9 81 704 0·704051 07 0·70405 15·32·70 0·512757 12 þ2·5
within the El Teniente deposit do not show these same tem- andesitic extrusive rocks of the Farellones Formation poral isotopic variations and all of them are isotopically (Howell & Molloy, 1960; Camus, 1975; Cuadra, 1986). similar to the Teniente Volcanic and Plutonic Complex However, chemical analyses for samples from within the rocks, and distinct from the Young Plutonic Complex mine indicate SiO2 contents that range only from 46·5to granitoids. 52 wt % (Table 2 and Table 2ES in the Electronic Supplement).This indicates that these rocks are essentially basaltic in composition, although some sills on the extreme Teniente Mafic Complex margins of this complex, outside the area of the mine (CMET; 8·9 1·4 Ma) itself, are basaltic andesites and andesites, with SiO2 in The oldest rocks within the El Teniente mine are dark col- the range 56^59·8 wt % (Table 2 and Table 3ES in the ored, with an aphanitic to porphyritic appearance. For Electronic Supplement). many years they were locally known as the ‘Andesites of Geological mapping (Fig. 2; Morel & Spro« hnle,1992) in- the Mine’, but currently they are mapped as the Teniente dicates that the ‘Andesites of the Mine’ constitute part of a Mafic Complex (Fig. 3; Skewes & Are¤ valo, 2000; Burgos, mafic intrusive complex, with the form of a laccolith, that 2002; Skewes et al., 2002, 2005). These rocks are generally intruded rocks of the Teniente Volcanic Complex, as was strongly altered, brecciated and mineralized, and aspects originally suggested by Lindgren & Bastin (1922).The cen- of their original petrology have been obscured. The name tral part of this mafic complex, within which the mine is ‘Andesites of the Mine’ suggests intermediate extrusive located, has a vertical extent of 42000 m. The entire rocks, and they have been correlated in the past with the complex has a volume of at least 50 km3, based on an
Excursion métallogénique - Chili 2012 Références page 112
1600 STERN et al. EVOLUTION OF EL TENIENTE CU^MO DEPOSIT
estimated overall average thickness of 1km. An age of 8·9 2·4 Ma was obtained by fission-track dating on apa- tite from a mafic sill outside the mine (Maksaev et al., 2004), and a similar age of 8·9 1·4Ma(Fig.4)wasdeter- mined by K^Ar whole-rock analysis of a fresh olivine- bearing sample from the western margin of the mafic complex (sample Ex2004-04). There is no chemical distinction between the gabbros (Fig. 7a), diabases and basaltic porphyries (Fig. 7b) that form this complex. The least altered samples from within the mine have SiO2 contents that range between 47 and 52 wt % (Table 2 and Table 2ES in the Electronic Supplement), and chemically they correspond to basalts. Downloaded from They have 6·6^11wt % CaO and 17^21·5wt % Al2O3, consistent with their high calcic plagioclase content. The FeO content (8^11·6 wt %) is high with respect to MgO (55·2wt %),butTiO2 and P2O5 contents are relatively
low, consistent with calc-alkaline affinities for these mafic http://petrology.oxfordjournals.org/
87 86 rocks. Their alkali components and volatile contents vary Fig. 5. Published (open symbols) values of Sr/ Sr vs eNd for igne- according to the type and degree of alteration. The freshest ous rocks of different ages from a transect across the Andes at the lati- tude of El Teniente (348S), compared with values for samples of the rocks, from outside the deposit (samples EX2004-04 and host rocks of the deposit (filled symbols; Table 4), including the AM2; Table 2 and Table 3ES in the Electronic Teniente Mafic Complex (g), Sewell Tonalite (m), felsic porphyries Supplement), have the lowest K O(0·5^1·0 wt %; Fig. 8) (¨), and Porphyry A granitoid ( ). Previously published data are 2 from Stern & Skewes (1995), Kurtz et al. (1997), Nystrom et al.(2003), and H2O (typically 51·0 wt %). The samples with more Kay et al.(2005),Mun‹ oz et al. (2006) and Stern et al. (2010). A lthough intense biotite alteration have higher K2O and H2O con- the youngest felsic plutons in the deposit are the same age tents, the latter as high as 3·5wt%(Fig.8).However,itis (6·09 0·18 to 4·58 0·10 Ma) as rocks from the regionally defined clear that their SiO contents have not been significantly Younger Plutonic Complex (6·6to 5Ma;Kayet al., 2005) they are 2 isotopically similar to the older host-rocks of the deposit as well as to changed by alteration, as the small range of silica contents at Universite du Quebec a Chicoutimi on April 21, 2012 Teniente Volcanic and Plutonic Complex rocks. for samples from within the mine is independent of LOI
Fig. 6. Published (open symbols; data from same sources as for Fig. 5) values of eNd vs SiO2 (wt %) for igneous rocks of different ages from the transect across the Andes at the latitude of El Teniente (348S), compared with values for samples of the host-rocks in the deposit (filled symbols as in Fig. 5).
Excursion métallogénique - Chili 2012 Références page 113
1601 JOURNAL OF PETROLOGY VOLUME 52 NUMBERS 7 & 8 JULY & AUGUST 2011 Downloaded from http://petrology.oxfordjournals.org/
Fig. 7. (a, b) Photomicrographs of two samples from the Teniente Mafic Complex: (a) weakly altered gabbro from within the mine (sample at Universite du Quebec a Chicoutimi on April 21, 2012 DDH1411-1630’;Tables 1 and 2), containing fresh unzoned calcic plagioclase and Fe^Ti oxides, and secondary actinolite and biotite pseudomor- phically replacing olivine (upper left) and clinopyroxene (center); (b) a fresh porphyritic olivine basalt from the margin of the complex, outside the mine (sample EX2004-04; Tables 1 and 2), containing olivine (Fo74; center) and unzoned plagioclase phenocrysts in a finer-grained ground- mass of intergrown clinopyroxene, plagioclase and Fe^Ti oxides. (c). Biotite-rich ‘andesite’ igneous breccia sample DDH1473-940 (Tables 1 and 3), with interstitial anhydrite. It should be noted that the rock is not porphyritic, and in this fresh and unaltered igneous rock neither biotite nor plagioclase are altered to chlorite or sericite. (d) Photomicrograph of PDTsample DDH1134-365’ (Table 1ES) with oscillatory zoned plagio- clase, rounded quartz ‘eyes’, and fresh biotite (biot) phenocrysts in a groundmass containing quartz, plagioclase, potassium feldspar, biotite, an- hydrite and opaque minerals.
and K2O (Table 2 and Table 2ES). Also, plagioclase Sewell Tonalite (7·05 0·14 Ma) and phenocrysts, even in the most strongly altered samples, associated andesitic sills (8·2 0·5to have not been strongly affected by this alteration and pre- 6·6 0·4 Ma) serve high An contents indicative of mafic rocks. The The Sewell Tonalite is one among a number of intermedi- rocks of the Teniente Mafic Complex were clearly not ori- ate to felsic plutons within the Teniente Plutonic Complex ginally andesites converted into more basic compositions that regionally intruded the Teniente Volcanic Complex by secondary hydrothermal alteration processes. rocks between 12·4 and 7·0 Ma (Cuadra, 1986; Kurtz The basaltic rocks of the CMET have La/Yb ranging et al., 1997). The equigranular holocrystalline Sewell from 3·5to8·5 (Figs 9 and 10), Dy/Yb ranging from 1·9to Tonalite complex crops out over an area approximately 2·2 (Fig. 10) and Sr/Y from 19 to 37, and in these respects half as large as the Teniente Mafic Complex (Fig. 2), with they resemble the basalts of the Teniente Volcanic an estimated volume of 30 km3 based on an approximate Complex. Basaltic andesite and andesite sills outside the average thickness of 1km. mine have higher La/Yb (10^13·7) and Sr/Y (30^39) than Cuadra (1986) determined two K^Ar ages of 7·4 1·5 the more mafic samples, but similar Dy/Yb (1·9^2·3) to and 7·1 1·0 Ma for the tonalite. These two ages are in them. Initial 87Sr/86Sr ratios of all the samples range from the range of both a K^Ar age of 7·0 0·4 Ma that he also 0·70396 to 0·70421, and eNd ranges from þ2·9toþ 3·5, obtained for the La Huifa pluton, a porphyritic phase of values that overlap those of Teniente Volcanic Complex the Sewell Tonalite complex which crops out 2 km NE of rocks (Figs 5 and 6). the deposit (Fig. 2; Reich, 2001), and 40Ar/39Ar step heating
Excursion métallogénique - Chili 2012 Références page 114
1602 STERN et al. EVOLUTION OF EL TENIENTE CU^MO DEPOSIT
ages (a plateau age of 6·97 0·1 Ma and a total gas age of 7·05 0·14 Ma) determined by Maksaev et al. (2004) for this same pluton. Maksaev et al. (2004) also determined a younger U^Pb zircon age of 6·15 0·08 Ma for a sample (TT-101) they identified as the Sewell Tonalite, but we do not consider this sample as representative of this pluton (Skewes & Stern, 2007), and this age is younger than a Re^Os age of 6·31 0·03 Ma on molybdenite from a brec- cias complex intruding the tonalite. Therefore, we consider the total gas age they obtained for the La Huifa pluton to be the best age also for the Sewell Tonalite complex (Fig. 4). Cuadra (1986) also dated andesitic sills at Cerro Montura, a few kilometers NW of the mine, at 8·2 0·5 Downloaded from and 6·6 0·4Ma(Fig.4). The Sewell Tonalite and andesitic sills range in compos- ition from 61 to 66 wt % SiO2,with1·5^2·7wt % K2O. The La/Yb ratios of the andesitic sills (9·4^13·1) are lower
and less variable than those of the tonalite (15·6^44·5; http://petrology.oxfordjournals.org/ Fig. 8. K2O vs volatile (LOI, loss on ignition) content (in wt %), Table 2 and Table 4ES). The tonalite has higher La/Yb g of Teniente Mafic Complex samples from within the mine ( ; (Fig. 9) and Sr/Y, lower Yb and Y, and similar to somewhat Table 2 and Table 2ES) compared with the freshest samples from out- side the mine (œ; samples EX2004-04 and AM2; Table 2 and Table higher Dy/Yb (Fig. 10) compared with the mafic rocks of 87 86 3ES), illustrating the correlation of increasing K2O and LOI as the the CMET. Sr/ Sr ratios (0·70385^0·70410) and eNd intensity of alteration increases. The least altered samples have K2O (þ2·2toþ 3·2) are similar for both sills and the tonalite, 51wt %. All these samples have SiO2 between 46·7 and 52 wt %, MgO between 3·5 and 5·8 wt % and FeO 48% (Table 2 and and overlap with the values for the samples from both the Table 2ES), and they all preserve unaltered calcic plagioclase, indicat- Teniente Volcanic and Teniente Mafic Complex (Figs 5 ing that they were originally all basalts, not andesites. and 6). Kurtz et al. (1997) also reported similar La/Yb at Universite du Quebec a Chicoutimi on April 21, 2012
Fig. 9. Rare earth element (REE) concentrations, normalized to chondritic meteorite abundances, for the plutonic host-rocks of the Teniente deposit.
Excursion métallogénique - Chili 2012 Références page 115
1603 JOURNAL OF PETROLOGY VOLUME 52 NUMBERS 7 & 8 JULY & AUGUST 2011 Downloaded from
Fig. 10. La/Yb vs Dy/Yb for the host plutons of theTeniente deposit. Symbols are as in Fig. 9.Trends produced by garnet, amphibole and gabbro (clinopyroxene and plagioclase) crystal fractionation are from Davidson et al.(2007).SiO2 content increases towards higher La/Yb. http://petrology.oxfordjournals.org/ ratios (9·9^20·9), initial 87Sr/86Sr (0·70398^0·70416) and Porphyry A anhydrite-bearing stock eNd (þ1·6toþ 2·8) for regionally contemporaneous felsic (5·67 0·19 Ma) plutons of the Teniente Plutonic Complex (Kurtz et al., The small (51km3) Porphyry A anhydrite-bearing granit- 1997). oid stock, which is 150 m wide at level Teniente-5 (Fig. 3), consists of a number of different lithological units described by Stern et al. (2007). The stock intrudes a Northern and Central dacitic porphyries Cu-rich (41^2 wt % Cu) magmatic^hydrothermal anhyd- (6·09 0·18 Ma) rite breccia emplaced along the contact between the intru- Within the Teniente mine, two small (51km3) dacitic por-
sive rocks of the El Teniente Mafic Complex and the at Universite du Quebec a Chicoutimi on April 21, 2012 phyry bodies, the Northern (1000N, 1000E; Fig 3) and Sewell Tonalite pluton. The margins of the stock are char- Central (700N, 1200E) dacitic porphryies, occur east and acterized by matrix-supported igneous breccias containing NE of the Braden breccia (Fig. 3). Maksaev et al. (2004) ob- abundant angular and rounded clasts of the surrounding tained a U^Pb age of 6·11 0·13 Ma for 11 zircons from mafic and tonalitic rocks. The matrices of these breccias the Northern dacitic porphyry (sample TT-102), and a contain igneous biotite, plagioclase and potassium feld- very similar overall average age of 6·08 0·22 Ma for 20 spar, quartz and both primary magmatic and magmatic^ zircons from the Central dacitic porphyry (sample hydrothermal anhydrite (Fig. 7c). The igneous breccias TT-90). We consider the total average 6·09 0·18 Ma to have been mapped by mine geologists as ‘microdiorite’ be the age of these intrusions (Skewes & Stern, 2007). and ‘andesite’ breccia depending on the color of their Based on their age, they would be considered by Kay matrix, which ranges from light to dark as the relative pro- et al. (2005) as part of the regionally defined Younger portion of biotite increases. The light-colored Porphyry A Plutonic Complex rocks. ‘microdiorite’ forming the central clast-free core of the The Northern and Central dacitic porphyries have SiO2 stock, which is volumetrically the dominant rock type, is contents between 56·8 and 66·2 wt % and K2O between petrologically equivalent to the matrix of the ‘microdiorite’ 1·9 and 7·1wt % (Table 3 and Table 5ES). Their La/Yb igneous breccias found along its margins. (18 ·2^44·3; Fig. 9) and Sr/Y (72^140) ratios are higher As noted by Stern et al. (2007), the ‘andesite’ breccia is than those of the mafic and intermediate rocks of the not extrusive and is not an andesite (Table 3 and Table CMET, but similar to those of the Sewell Tonalite, whereas 6ES), the ‘microdiorite’ breccia is not a diorite, and none their Dy/Yb values (1·9^2·3; Fig. 10) are similar to those of of the rocks that form the Porphyry A stock are porphyrit- both CMET and tonalite samples. Significantly, 87Sr/86Sr ic. These field names are correct only in so far as they sug- ratios (0·70408^0·70409) and eNd (þ2·7toþ 3·2) are simi- gest that these are igneous rocks. However, they are lar for both porphyries, as well as being within the range clearly not typical igneous rocks, and we do not think that of the Teniente Mafic Complex and Sewell Tonalite sam- any common igneous rock names are appropriate for ples, and the Teniente Volcanic and Plutonic Complex these unusual rocks, because of the large amount of rocks (Figs 5 and 6). However, they are isotopically unlike modal primary igneous anhydrite, sulfur and volatiles Younger Plutonic Complex rocks of similar age and com- (Table 3) they contain, but we do not offer alternative 87 86 position, which have higher Sr/ Sr ratios and lower eNd. names for these rocks here.
Excursion métallogénique - Chili 2012 Références page 116
1604 STERN et al. EVOLUTION OF EL TENIENTE CU^MO DEPOSIT
Clast-rich ‘microdiorite’ igneous breccias from along the Braden Pipe, with a maximum width of 300 m (Fig. 3). It margin of the Porphyry A stock have an age determined is truncated to the south by the Braden Pipe, and it is not by U^Pb in zircons of 6·27 0·19 Ma considering all ana- clear if this porphyry also occurs south of the pipe. lyzed spots (sample TT-150; Maksaev et al., 2004). Maksaev et al. (2004) obtained an age of 5·28 0·10 Ma However, a bimodal distribution of ages is apparent, with for zircons from the Teniente Dacite Porphyry dike, which a peak of 6·46 0·11Ma for 13 zircons and 5·67 0·19 Ma has frequently been cited as the causative igneous intrusion for four zircons. We consider the older peak to represent for the mineralization in the deposit (Howell & Molloy, the average of the multi-lithic clasts in the breccia, and 1960; Camus, 1975; Cuadra, 1986), although it clearly cuts the younger to represent the age of the breccia matrix copper mineralized mafic rocks, veins and breccias in the (Fig. 4; Maksaev et al., 2004; Skewes & Stern, 2007; Stern deposit (Skewes et al., 2002, 2005), and its northernmost ex- et al., 2007). Molybdenite in the Cu-rich hydrothermal an- tension, north of the Teniente River valley (Fig. 2), is hydrite breccia intruded by this stock has been dated by unmineralized. Downloaded from Re^Os at 5·89 0·02 Ma (Cannell et al., 2005) and Teniente Dacite Porphry samples have SiO2 contents 5·60 0·02 Ma (Maksaev et al., 2004). These ages are in from 59·9to66·3 wt % and K2Ofrom2·6to2·9wt% agreement with the younger U^Pb age in zircons, and (Table 3 and Table 5ES). Their La/Yb ratios range from they indicate that the formation of this breccia is temporal- 17 ·3to54·1 (Fig. 9). Sr and Nd isotopic ratios are within ly and therefore genetically related to the intrusion of the the range of all the Late Miocene igneous rocks in the de-
Porphyry A stock (Fig. 4). posit (Figs 5 and 6), and similar to the Teniente Volcanic http://petrology.oxfordjournals.org/ The anhydrite-bearing plutonic rocks of the Porphyry A and Plutonic Complex rocks, but distinct from the contem- stock contain 43 wt % S and 40·5 wt % Cu (Table 3 and poraneous felsic granitoids of the Younger Plutonic Table 6ES).The amount of volatiles lost on ignition (LOI) Complex, which have higher 87Sr/86Sr ratios and lower ranges from 3·1^6·6 wt % for ‘andesite’ igneous breccias to eNd. 8·0^10·4 wt % for Porphyry A ‘microdiorite’. However, an indeterminate amount of this LOI is due to the vaporiza- Latite dikes (4·82 0·09 Ma) tion of sulfur during the ignition process. When normal- Latite porphyry ring-dikes, 6^8 m wide, occur concentric- ized for their high LOI and S contents, ‘andesite’ igneous ally surrounding the Braden Pipe (Fig. 3). A small latite breccias have major-element compositions comparable porphyry stock is truncated by the Braden Pipe along its
with those of basalt (normalized SiO2 from 49·5to northeastern edge (Gonza¤ lez, 2006). Maksaev et al. (2004) at Universite du Quebec a Chicoutimi on April 21, 2012 52·3 wt %), whereas the Porphyry A ‘microdiorite’ has obtained a U^Pb age of 4·82 0·09 Ma for zircons from compositions similar to andesite (normalized SiO2 from one latite dike (Fig. 5), which is within the range of the 56·0to61·8 wt %), but with anomalously high CaO that age they also determined for the Braden Pipe (4·81 0·12 reflects the high modal abundance of anhydrite. Despite Ma). The occurrence of blocks of this rock type in the their high CaO, the anorthite content of the primary igne- Braden Pipe suggests that that at least some latite por- ous plagioclase in these rocks, which are of basaltic and an- phyry intruded prior to the formation of this breccia pipe, desitic composition, is relatively low (An532). and may have played a role in the formation of the pipe La/Yb ratios for the ‘andesite’ igneous breccia range (Floody, 2000; Cannell et al., 2005). Maksaev et al. (2004) from 6·6to15·7, overlapping the range of Teniente Mafic also obtained a 40Ar/39Ar plateau age of 4·58 0·10 Ma Complex samples and consistent with the mafic compos- for biotite from a small, weakly mineralized dacite intru- ition of these samples (Fig. 9). La/Yb ratios for the ‘micro- sion deep in the Braden breccia pipe. diorite’ breccia are higher (30·5^79·9), as a result of both Available chemical analyses indicate that the latite is the higher La and lower Yb contents in these samples. chemically similar to the Teniente Dacite Porphyry, with The Sr and Nd isotopic compositions of both the ‘andesite’ 60·9^63·1wt % SiO2 and 2·2^3·5wt%K2O (Table 3 and and ‘microdiorite’ igneous rocks are identical to each Table 5ES), and a high La/Yb of about 35 (Fig. 9). Their other (Figs 5 and 6), and they are isotopically similar to Sr and Nd isotopic ratios are similar to those of the all other Late Miocene igneous rocks in the El Teniente de- Teniente Dacite Porphyry and within the range of all the posit, as well as the Teniente Volcanic and Plutonic Late Miocene igneous rocks in the deposit (Figs 5 and 6), Complex rocks. However, as with the Northern and and similar to both the Teniente Volcanic and Plutonic Central Dacitic Porphyries, they are isotopically distinct Complex rocks, but distinct from contemporaneous felsic from the contemporaneous felsic granitoids of the Younger granitoids of the Younger Plutonic Complex, which have 87 86 87 86 Plutonic Complex, which have higher Sr/ Sr ratios and higher Sr/ Sr ratios and lower eNd. lower eNd (Figs 5 and 6; Kay et al., 2005). Post-mineralization lamprophyre dikes Teniente Dacite Porphyry (5·28 0·10 Ma) (3·85 0·18 to 2·9 0·60 Ma) The Teniente Dacite Porphyry is an elongated subvertical The youngest igneous rocks in the deposit are post- dike-like body, extending 1·5 km to the north of the mineralization olivine-hornblende-lamprophyre dikes,
Excursion métallogénique - Chili 2012 Références page 117
1605 JOURNAL OF PETROLOGY VOLUME 52 NUMBERS 7 & 8 JULY & AUGUST 2011
dated at 3·85 0·18 to 2·9 0·60 Ma (Cuadra, 1986; Kay et al. (2005) concluded that the differences in the iso- Maksaev et al., 2004). These mafic, dark green to black in topic composition of the Younger Plutonic Complex rocks color, fresh or only weakly altered, post-mineralization, (6·6to 5 Ma) from older Teniente Volcanic (46·5Ma) sub-vertical dikes occur both within the El Teniente mine and Plutonic Complex rocks (47·0 Ma), as well as the and in the vicinity of the mine (Cuadra, 1986; Stern et al., higher La/Yb, Sr/Y and other more ‘adakitic’ petrochem- 2010). ical characteristics of the Younger Plutonic Complex felsic The lamprophyre samples with olivine have SiO2 con- rocks, imply differences in magma source regions and tents of 51^53 wt %, Ni of 70^190ppm and Cr 280^ depths of crustal magma generation or fractionation. 390 ppm. Olivine-free lamprophyres range in composition They attributed the generation of the more ‘adakitic’ from 55 to 67 wt % SiO2 and have lower Ni (5100ppm) Younger Plutonic Complex felsic rocks in part to melting and Cr (5200 ppm) than the more mafic olivine-bearing of the base of the crust that thickened during the Late lamprophyres. The lamprophyre dikes have relatively high Miocene. Kay et al. (1999) also suggested, in an earlier Sr (642^916ppm), low Y (12^17ppm) and consequently paper, but in the context of this same model, that fluids Downloaded from high Sr/Yb ratios (43^71), similar to adakitic andesites; released by the breakdown of hydrous phases in the lower however, these rocks, which range from basaltic to dacitic crust as the crust thickens contribute to mineralization in composition, are not adakites (Stern et al., 2010). The processes in the Andes. more mafic lamprophyre samples have low La/Yb ratios Significantly, however, the felsic plutonic rocks within (10^13) that are within the range of mafic rocks in the the El Teniente deposit do not show the same isotopic shift http://petrology.oxfordjournals.org/ youngest units of Late Miocene Teniente Volcanic 87 86 to higher Sr/ Sr ratios and lower eNd (Figs 5 and 6) Complex (9·2^13·2; Kay et al., 2005) and plutonic rocks of that occurs between the regionally defined Teniente the Teniente Mafic Complex. However, the other less Plutonic Complex (12·3^7·0 Ma) and the Younger Plutonic primitive olivine-free lamprophyre dikes have both lower Complex (6·6to 5 Ma). Furthermore, in a recent study Yb (0·80^0·93 ppm) and higher La/Yb ratios (18^22) com- of zircons separated from each of the felsic units in the de- pared with the more mafic lamprophyres (Stern et al., posit, Mun‹ oz et al. (2009) demonstrated that all samples 2010). have high initial 176 Hf/177 Hf ratios and positive eHf with 87 86 · i The lamprophyre dikes have Sr/ Sr from 0 70410 to values ranging from þ6·2toþ 8·5, and there are no dis- 0·70435 and e ¼þ1·1toþ 0·2 (Figs 5 and 6; Stern & Nd tinct differences among the analyzed igneous units. They Skewes, 1995; Kay et al., 2005; Stern et al., 2010). These iso- at Universite du Quebec a Chicoutimi on April 21, 2012 concluded that these characteristics are consistent with a topic values are higher and lower, respectively, than those common magmatic system being the source of the different of other Late Miocene plutonic rocks from the El Teniente intrusive pulses for which the high eHf values record a deposit, but similar to the isotopic compositions of the i relatively juvenile mantle source. We agree, and conclude Younger Plutonic Complex. that the plutonic rocks that host the deposit were not derived from progressively deepening crustal sources. DISCUSSION Instead, we suggest, as elaborated below, that the small The Late Miocene and Pliocene plutonic rocks that host volume of late felsic porphyries with ‘adakitic’ geochemical the El Teniente deposit formed over a relatively brief 4·3 characteristics (high La/Yb and Sr/Y) were formed Myr period, between 8·9 1·4 and 4·58 0·10 Ma (Fig. 4), during the final stages of crystallization of a large, during the waning stages of an extended Andean magmat- long-lived, shallow magma chamber (Fig. 11), and that this ic episode that began in the Oligocene. These plutons, magma chamber was also the source of the aqueous fluids which intruded rocks of theTenienteVolcanic Complex, in- that generated the mineralized breccias and veins in the clude initially the relatively large Teniente Mafic Complex deposit, as well as the metals they carried. laccolith (8·9 1·4 Ma), followed by the intrusion of the Based on a review of information concerning giant Cu somewhat smaller equigranular, holocrystalline Sewell deposits, both Cloos (2001) and Richards (2003, 2005) out- To n a l i t e ( 7 ·05 0·14 Ma). These two plutions overlap in lined models for their generation involving large magma age with the final stages of formation of the Teniente chambers from the cupola of which are derived the felsic Volcanic Complex (14·4^6·5 Ma), and the ages of other porphyries, magmatic^hydrothermal mineralizing fluids plutons in the regionally defined Teniente Plutonic and metals that form such deposits. For El Teniente, the Complex (12·3^7·0 Ma; Fig. 4). After volcanism in the conclusion that the mineralizing aqueous fluids and region ceased, a sequence of much smaller porphyritic metals were derived from the same magmas as formed the felsic stocks and dikes intruded between 6·09 0·18 and host plutons in the deposit is based directly on three lines 4·58 0·10 Ma, the age of the regionally defined Younger of evidence. First is the overlap in the timing of their for- Plutonic Complex (6·6to 5 Ma; Fig. 4), into the very re- mation (Fig. 4). Alteration and mineralization at El stricted 2 km 3 km area within which the deposit is now Teniente occurred over essentially the same time period as mined (Fig. 3). the emplacement of the felsic host plutons, between at
Excursion métallogénique - Chili 2012 Références page 118
1606 STERN et al. EVOLUTION OF EL TENIENTE CU^MO DEPOSIT Downloaded from http://petrology.oxfordjournals.org/
Fig. 11. Model for the multistage development of the El Teniente deposit, modified from Skewes et al. (2002) and Stern & Skewes (2005). The main features of the model include (1) a large, long-lived (42 Myr) open-system magma chamber, crystallizing at 44 km depth, fed from below by mantle-derived mafic magmas and exsolving aqueous fluids through its roof to produce the large breccia pipes that are prominent fea- tures in each deposit; (2) decreasing magma supply in the Late Miocene and Pliocene as subduction angle decreases, leading to crystallization and solidification of this chamber; (3) progressive uplift and erosion that enhances this crystallization and solidification process and results in telescoping of different types of breccia and igneous rocks; (4) progressive igneous differentiation of the magma chamber associated with crystal- lization and volatile loss, generating felsic porphyries that intrude previously mineralized rocks above the chamber. No coeval volcanism is known to have occurred during mineralization, but once the chamber solidified, post-mineralization lamprophyre dikes were emplaced. at Universite du Quebec a Chicoutimi on April 21, 2012 least 47·05 0·14 Ma and 4·42 0·02 Ma (Fig. 4; Skewes homogenization temperatures of boiling fluids, or the an- & Stern, 2007). Although 18 molybdenite Re^Os mineral- hydrite^chalcopyrite sulfur isotope geothermometer ization ages range from only 6·31 0·02 to 4·42 0·02 Ma (Kusakabe et al., 1984, 1990), the d18O of the fluids from (Maksaev et al., 2004; Cannell et al., 2005), mapping and which quartz and anhydrite in these veins precipitated petrological studies at El Teniente have shown that at least was þ5·8toþ 6ø, and the dD of fluids that precipitated one widespread event of biotitization and mineralization sericite in vein haloes is 35ø, similar to aqueous fluids predates emplacement of the Sewell Tonalite, and that this derived from cooling magmas (Skewes et al., 2002, 2005). felsic intrusion cuts biotitized Teniente Mafic Complex Based on homogenization temperatures (45258C) for coex- rocks and mineralized biotite veins (Skewes & Stern, isting vapor-rich and highly saline fluid inclusions in 2007). This implies that the Re^Os ages for molybdenite quartz from both biotite and tourmaline-rich breccias, obtained in the deposit record only the last 1·9Myrofmin- Skewes et al. (2002, 2005) calculated d18Oofþ6·9to eralization and that the temporal extent of mineralization þ 10·6ø and dDof 36ø for the boiling aqueous fluid was actually greater. The lack of older Re^Os ages may re- that precipitated this quartz, consistent with exsolution of flect the fact that the early widespread mineralization the breccia-generating fluids from a cooling magma. associated with biotitization of the Teniente Mafic Finally, for El Teniente, radiogenic isotopes also imply Complex generally does not result in the deposition of con- that the metals in the deposit were derived from the siderable molybdenite, or coarse molybdenite suitable for same magmas as those that formed the host-rocks of dating (Rabbia et al., 2009). the deposit. Lead isotope ratios, measured in galena in Second, fluid inclusions in quartz-bearing veins at vari- the deposit (206Pb/204Pb ¼18·57, 207Pb/204Pb ¼15 ·60, and ous stages of alteration at El Teniente (Skewes et al., 2002, 208Pb/204Pb ¼ 38·49; Puig, 1988; Zentilli et al., 1988), are 2005; Cannell et al., 2005) are consistent with precipitation within the range of the published Pb isotopic compositions from highly saline, high-temperature magmatic fluids, of the Late Miocene Teniente Volcanic and Plutonic both boiling and non-boiling, as well as other fluids with Complex rocks (Kurtz et al.,1997; Kay et al., 2005), consist- variable salinity, including low-salinity vapor-rich inclu- ent with the derivation of this lead, and by implication sions that may have played an important role in transport- the copper in the deposit, from Late Miocene magmas ing and depositing metals (Klemm et al., 2007). (Stern & Skewes, 2005). Osmium is also derived from the Calculated at temperatures derived from either igneous rocks in the deposit, as indicated by the similarity
Excursion métallogénique - Chili 2012 Références page 119
1607 JOURNAL OF PETROLOGY VOLUME 52 NUMBERS 7 & 8 JULY & AUGUST 2011
of 187Os/188Os ratios, measured in chalcopyrite, sphalerite appropriate for calculating the amount of mantle-derived and bornite precipitated at different alteration stages magma from which the Cu in the deposit was ultimately during formation of the deposit (Freydier et al., 1997). If derived. these metals had been derived from the surrounding coun- The full volume of magma from which the Cu was ex- try rocks, greater variability in the 187Os/188Os ratios tracted may have resided in a single large, long-lived would be expected. These osmium isotopic ratios, which magma chamber, or may represent the integrated volume range from 0·171 t o 0 ·223, are more similar to mantle of multiple smaller magma batches forming independent (0·13) than crustal ( 1·0) values, consistent with the deriv- magma chambers and intrusions at different times during ation of Os and other metals from magmas formed in the the 42 Myr period of mineralization. For the five reasons sub-Andean mantle, which was also the source for the listed below, we prefer a model of a single, large, long-lived magmas that formed the host plutons in the deposit. chamber or magmatic plumbing system during the period Finally, d34S values of sulfides (chalcopyrite and pyrite) when the felsic intrusions that host the deposit and the from El Teniente range from 5·6toþ1·6ø, and aver- most prominent dated breccias in the deposit formed. Downloaded from · ø et al age approximately 2 2 (Kusakabe ., 1984, 1990). (1) The lack of occurrence of any mafic rocks in the time This small range is consistent with derivation of sul- period 58·9 1·4to43·85 0·18 Ma suggests that a fur only from igneous rocks associated with the deposit. single magma chamber, or the integrated magmatic A calculated total sulfur isotopic composition of plumbing system, acted as a trap within which mafic þ4·5ø for El Teniente (Kusakabe et al., 1984, 1990) is also http://petrology.oxfordjournals.org/ 34 mantle-derived magmas were mixed and thus unable similar to the d S values of non-mineralized Andean to reach the surface. granitoids, which range from þ3·3toþ 6·1ø (Sasaki (2) The isotopic similarity of all the igneous rocks et al., 1984). emplaced during this time period (Figs 5 and 6) also The size, time period of emplacement, and genesis of the suggests a large well-mixed system, particularly in different Cu-mineralized breccia pipes and associated light of the fact that on a regional scale, smaller iso- veins at El Teniente therefore constrain various aspects of lated felsic plutons of the Younger Plutonic Complex the processes of crystallization and devolatilization in the (Kay et al., 2005) have different isotopic compositions underlying magma chamber from which the igneous rocks from the contemporaneous felsic plutons inside the that host the deposit were derived. Most significantly, the
Teniente deposit. This isotopic change is not observed at Universite du Quebec a Chicoutimi on April 21, 2012 amount of Andean magma with 100ppm Cu required to in the felsic plutons within the Teniente deposit, sug- provide the 100 Mt of Cu in the deposit would be gesting that a single large, long-lived system acted as 3 4600 km if the mechanism of extraction of the Cu is a sponge within which small volumes of isotopically 5100% efficient (Stern & Skewes, 2005). This is consistent distinct melts may have mixed without noticeably 3 with the calculation of Cline & Bodnar (1991) of 60 km changing the isotopic composition of the larger of magma with 62 ppm Cu to produce the 6 Mt of copper system. in the Yerington deposit in Nevada, and with the calcula- (3) The sequence from initially mafic to progressively tions of Cloos (2001) and Richards (2005) that batholith- smaller volumes of isotopically similar felsic differenti- 3 size volumes of magma (41000km ) are required to gener- ates, which can all be explained by fractionation of ate giant Cu deposits. Compilations of Cu contents for arc olivine, plagioclase, pyroxene, amphibole and acces- andesites (median of 60 ppm; Gill, 1981) and arc rocks sory phases, is consistent with a single chamber or with SiO2 452 wt % (average of 70 ppm; Stanton, 1994), integrated system progressively cooling and crystalliz- are lower than 100ppm, which would imply even more ing as subduction angle, and consequently the supply magma required to produce the El Teniente deposit. of mantle-derived magma recharging the system, However, we consider the Cu content of more mafic basalt- decreased from the Late Miocene to the Pliocene. ic magma to be the appropriate value to use, as it was (4) The formation of multiple mineralized breccia pipes sub-arc mantle-derived basalts that recharged the magma and veins contemporaneous with the intrusion of the chamber or magmatic plumbing system above which the felsic plutons in the deposit requires the exsolution of Teniente deposit formed. A sample of a fresh olivine basalt large volumes of magmatic volatiles from the roof or from the Teniente Mafic Complex (sample Ex2004-04; cupola of a magma chamber.The quantity of volatiles, Table 2), which represents the type of basalt initially sulfur and metals in the deposit requires scavenging emplaced in the area of the deposit, has 115ppm of Cu. from a relatively large volume of magma, which Post-mineralization Pliocene olivine lamprophyres, the could be provided by the progressive cooling of a type of mafic magma that was forming in the mantle single large chamber. after the deposit formed, have Cu contents that range (5) Cannell et al. (2005) suggested that vein distributions from 80 to 100ppm and average 92 ppm (Stern et al., in the deposit were controlled by a local stress regime 2010). Therefore we consider a value of 100ppm generated by the intrusion of a large, 4^6 km deep
Excursion métallogénique - Chili 2012 Références page 120
1608 STERN et al. EVOLUTION OF EL TENIENTE CU^MO DEPOSIT
magma chamber that they also interpreted as the Another important implication of the suggestion that source of the felsic porphyries, Braden Pipe and Cu the mineralized breccia pipes, veins and Cu ore in the de- in the deposit. posit were derived from the same magmatic plumbing system as the igneous rocks concerns the depth to the top The development of a large batholith-sized magma of this system. This must have been below the roots of the chamber, containing at least 600 km3 of magma, during breccia pipes, or 44 km based on the fact that these roots a 42Myrperiod,from7·05 0·14 to 4·58 0·01 Ma, occur at 41km below the deepest level of mining and ex- when compression prevented volcanism, is consistent with ploration drilling, which is 43 km below the current sur- known magma extrusion and production rates in the face and deeper below the paleosurface. Cloos (2001) and Andes. Stern et al. (1984) determined that the 105 km3 Richards (2003, 2005) suggested that large magma cham- cone of the Maipo volcano, which formed over only the bers below giant Cu deposits are disk shaped and that last 450 kyr in the High Andes at the same latitude (348S) 3 they focus felsic intrusions and volatile exsolution through
as El Teniente, required extrusion rates of 236 km /Myr. If Downloaded from a narrow cupola. This is consistent with the observation at none of this magma had extruded, it could create an 3 Teniente that the 100 Mt of Cu in the deposit is concen- 600 km chamber in 2 Myr. This rate is at the low end 2 of other estimates of extrusion rates in the Andes (Baker trated into a very small 6 km area (Fig. 3) of brecciated & Francis, 1978; Hildreth et al., 1984), and extrusion and altered host-rocks into which the youngest felsic por- rates are certainly lower than total magma production phyries intruded. Cloos (2001) suggested that disk-shaped http://petrology.oxfordjournals.org/ rates. Magma chambers of this size have occurred at magma chambers of this size may form at the density dis- this latitude of the Andes, as indicated by the eruption of continuity between crystalline basement and the sediment- the 450 km3 tuff of Pudahuel from the Maipo caldera ary cover, at depths of 6^15km. Below Teniente an upper (Stern et al., 1984), and much larger magma chambers are crustal density discontinuity probably occurs at the contact implied by the volumes of single eruptions of silicic between the Coya-Machal|¤ volcanic rocks and the older magmas in the Central Andes of northern Chile (de crystalline basement. Together the Teniente Volcanic Silva, 1989). Complex and Coya-Machali volcanic rocks are 6km Studies of the time frame in which crystal^melt segrega- thick, and we suggest that this is where the initial magma tion takes place from crystal mushes (104^105 years; chamber below the Teniente deposit may have formed (Fig. 11). With time, uplift and erosion brought the roof of
Bachman & Bergantz, 2004), residence times of silicic at Universite du Quebec a Chicoutimi on April 21, 2012 magmas in the crust as indicated by Rb/Sr isochrons the chamber closer to the surface, but never closer than (7 105 years; Halliday et al., 1989), crystal growth rates as 4 km, which is the minimum estimate of the depth of the implied by spot U^Pb analysis of magmatic zircons roots of the larger mineralized breccias in the deposit. (3 105 years; Reid et al., 1997; Brown & Fletcher, 1999), Also, as discussed in detail by both Cloos (2001) and 238U/ 230Th ratios in rhyolites constraining differentiation Richards (2003, 2005), shallower chambers (54 km) may time scales (4105 years; Hawkesworth et al., 2000), and lose volatiles without concentrating metals, as the solubil- time scales estimated to build batholith-size plutons and ity of Cl and metals is lower in aqueous-rich fluids at low allow them to cool (105^106 years; Koyaguchi & Kaneko, pressure. Deeper, mid- to lower crustal chambers may also 1999; Jellinek & DePaolo, 2003) approach the lifetime of underlie and supply magmas to the relatively shallow the processes of felsic pluton intrusion and mineralized chamber below the site of formation of a large Cu deposit breccia emplacement at Teniente. Maksaev et al. (2004) re- (Cloos, 2001; Richards, 2003, 2005; Annnen et al., 2006), ported bimodal zircon age populations (6·46 0·11 and but the thermal gradients in deeper chambers may not be 5·67 0·19 Ma) from different spot U^Pb ages on single as high as in a shallower chamber, inhibiting volatile trans- zircons from the Porphyry A ‘microdiorite’ and suggested fer and concentration of volatiles at the top of the chamber. that the older group of ages may indicate zircons inherited Furthermore, deeper chambers may never reach volatile from a crystallizing magma chamber. Cannell et al.(2005) saturation; a relatively shallow (4^6 km depth) cupola is interpreted changes in vein orientation in the deposit to re- implied for El Teniente by the multiple events of magmatic flect various stages of subsidence and resurgence of the volatile exsolution and mineralized breccia emplacement magma chamber underlying the Teniente deposit. We that generated the deposit (Skewes et al., 2002, 2005). therefore suggest that the episodic recharge of the magma Magmatic differentiation processes taking place in a system below Teniente by hot mantle-derived mafic large magma chamber include a complex combination of magma could either maintain the system as a single cham- recharge of the base of the chamber by mantle-derived ber for a protracted period, or reheat and reactive the mafic magmas, thermal stratification and convection, system through a series of thermal oscillations that allow magma mixing and crystallization, bubble formation, and felsic melts to form, and large quantities of volatiles, sulfur volatile transfer and concentration in the upper cooler and metals to concentrate near the top of the system epi- parts of the chamber, as outlined for such large chambers sodically over an extended period of time. in general by Shinohara et al. (1995), Cloos (2001), Hattori
Excursion métallogénique - Chili 2012 Références page 121
1609 JOURNAL OF PETROLOGY VOLUME 52 NUMBERS 7 & 8 JULY & AUGUST 2011
& Keith (2001), Richards (2003, 2005) and Heinrich et al. moderate-sized chamber prior to the intrusion of the tonal- (2005). The petrochemistry of the sequence of plutons it e ( Fi g. 11). within El Teniente constrains some aspects of the timing of development and evolution of the magma chamber Felsic porphyries (6·09 0·18 to below this deposit. 4·58 0·10 Ma) The Northern and Central Dacitic Porphyries, Teniente Teniente Mafic Complex (8·9 1·4 Ma) Dacite Porphyry and latite dikes all have features in The oldest pluton in the deposit, the Teniente Mafic common. They were intruded after volcanism in the area Complex (CMET), is a laccolith formed by mafic stopped, and their age corresponds to the terminal stages magmas. These magmas petrochemically resemble con- of the regionally defined Young Plutonic Complex. Both temporaneous mafic volcanic rocks of the youngest unit, the felsic porphyries within El Teniente and the Young the Upper Sewell Group (9·3^6·5 Ma), of the Teniente Plutonic Complex rocks have high La/Yb and Sr/Y. Kay
Volcanic Complex. Because the basalts in the core of the et al. (2005) attributed these features to equilibration with Downloaded from Teniente Mafic Complex, inside the mine, resemble the a high-pressure, plagioclase-poor, garnet-bearing residual most mafic members of the Upper Sewell Group, they mineral assemblage formed in the deep crust as the crust may have come directly from the sub-arc mantle, without thickened owing to Late Miocene deformation. This may significant fractionation or residence time in any upper be the case for the Young Plutonic Complex rocks outside
crustal magma chamber. We conclude that the large the deposit, which have different isotopic compositions http://petrology.oxfordjournals.org/ upper crustal magma chamber that subsequently formed from the older Teniente Volcanic and Plutonic Complex below the area of the Teniente deposit (Fig. 11) did not rocks, suggesting an isotopically distinct, possibly deeper exist at the time of the emplacement of the CMET. source. However, the small volume of felsic porphyries inside the Teniente deposit are isotopically similar to both Sewell Tonalite and andesite sills the mafic rocks of the Teniente Mafic Complex and Sewell (8·2 0·5to6·6 0·4 Ma) Tonalite, and we suggest that they were formed by The Sewell Tonalite and sub-contemporaneous andesite low-pressure fractional crystallization processes near the sills are intermediate in composition and both contain roof of the large magma chamber that crystallized and amphibole crystals. Chemically these rocks resemble inter- solidified under the Teniente deposit between 6·3 and
mediate plutons of the Teniente Plutonic Complex. As dis- 4·4 Ma. As outlined above, the presence of this chamber at Universite du Quebec a Chicoutimi on April 21, 2012 cussed by Kay et al. (2005), their chemistry may be during this time period is implied by the contemporaneous explained by derivation from more mafic, but isotopically episode of breccia formation and mineralization (Fig. 4) similar, Teniente Volcanic Complex magmas by crystal resulting from exsolution of magmatic fluids that had sca- fractionation involving amphibole along with pyroxene, venged large quantities of volatiles, metal and sulfur from plagioclase and accessory phases. In the case of the Sewell a large magma body. Magmas formed at greater pressures, Tonalite, this model is consistent with the higher La/Yb below this chamber, such as mantle-derived magmas or and SiO2, but similar Dy/Yb (Fig. 10) of this tonalite com- deep-crustal melts similar to those that formed the Young pared with the isotopically similar basaltic rocks of the Plutonic Complex rocks, did not reach the surface because Teniente Mafic Complex. they mixed into the magma in this chamber. The process of fractionation required to produce this That the felsic porphyries in El Teniente equilibrated relatively large tonalite pluton implies at least a moderate- near the top of a relatively shallow magma chamber is sized upper crustal magma chamber from which both por- also indicated by their porphyritc texture, large proportion phyritic felsic rocks, such as those that occur at La Huifa of phenocrysts implying that they intruded as crystal (Reich, 2001), and the larger equigranular phases of this mushes, and the complex oscillatory zoning exhibited by pluton can form. This pluton was generated when volcan- plagioclase crystals that is interpreted to reflect fluctuation ism was still active, and zoning in plagioclase crystals in of volatile pressure in a shallow crustal environment. the tonalite suggests fluctuating water pressure at relatively Some of the samples of these felsic porphyries contain shallow conditions. We suggests that the Sewell Tonalite what appears to be primary igneous anhydrite, a phase formed during the early developmental and progressive produced by saturation with oxidized sulfur, consistent growth phase of the magma chamber that subsequently with their formation near the volatile-rich roof of a shallow evolved into the larger chamber that generated both the magmatic system, not in the deep crust. dated mineralized breccias and felsic porphyries in the de- Amphibole is a common phenocryst in the porphyries, posit between 6·31 0·03 and 4·42 0·02 Ma (Fig. 4). and we suggest that these magmas formed by fractional However, the Sewell Tonalite cuts early mineralized veins crystallization in the shallow magma chamber involving in the Teniente Mafic Complex (Skewes & Stern, 2007), amphibole, as well as various processes too complicated indicating that the earliest stages of mineralization in the for evaluation by simple models of trace-element evolution, deposit reflect processes of devolatilization of this such as recharge by mafic magmas, magma mixing, and
Excursion métallogénique - Chili 2012 Références page 122
1610 STERN et al. EVOLUTION OF EL TENIENTE CU^MO DEPOSIT
volatile diffusion. The felsic porphyries have higher La/Yb et al. (2007) suggested that this occurred as a result of pre- and SiO2 than the basaltic rocks of the Teniente Mafic cipitation of magmatic^hydrothermal anhydrite from the Complex, but similar Dy/Yb, consistent with fractionation magmatic volatiles escaping the underlying magma cham- of amphibole, clinopyroxene and plagioclase, but not ber along with the Porphyry A magma, and that this pro- garnet (Fig. 10; Davidson et al., 2007; Richards & Kerrich, cess enhanced the CaO, Cu and S contents and resulted 2007). in the high LOI of the ‘microdiorite’ relative to normal ig- neous rocks. Thus we consider the high modal abundance Porphyry A (5·67 0·19 Ma) of magmatic^hydrothermal anhydrite and high CaO, Cu, Despite the abundance of unambiguously hydrothermal S and LOI of the Porphyry A ‘microdiorite’ as a primary anhydrite in this deposit, we argue that anhydrite in the ig- feature of its chemistry, resulting from the crystallization neous rocks forming the Porphyry A stock, as well as in of a supersaturated mixture of cogenetic silicate magma some of the other felsic porphyries, is either a strictly mag- plus magmatic fluids derived together from the same matic or magmatic^hydrothermal phase (Stern et al., underlying magma chamber, and not from a two-stage Downloaded from 2007). It is clear petrographically that anhydrite and process of magma crystallization followed by later second- other major minerals in this stock are primary phases that ary hydrothermal alteration. The volumetrically less sig- crystallized together in the melt-filled space occupied by nificant ‘andesite’ igneous breccias contain lower modal the stock, and are not secondary replacement phases of proportions of interstitial anhydrite that appears texturally any pre-existing rock. The Porphyry A ‘andesites’ and to be strictly magmatic anhydrite, crystallized in a single http://petrology.oxfordjournals.org/ ‘microdiorites’ have REE contents similar to those of stage at the same time as the silicate minerals that form Teniente Mafic Complex rocks and the Sewell Tonalite, re- these rocks. The volatile contents of the ‘andesite’ breccias spectively (Fig. 9), which also suggests that these are igne- are similar to the compositions expected for ous rocks and not the product of extreme alteration of volatile-saturated basalts at 41 kbar pressure (LOI ¼ 2^ pre-exisiting felsic porphyries, all of which have higher 4wt%). La/Yb. The fact that anhydrite crystallized contemporan- The Cu- and S-rich magmas that formed the eously with the other silicate minerals in this stock is sug- anhydrite-bearing Porphyry A pluton and its marginal ig- gested by its planar crystal boundaries with these neous breccias were sourced from the same chamber as minerals, and the inclusion of silicates in anhydrite and the other felsic porphyries in the deposit, as indicated by anhydrite in silicates. Furthermore, we suggest that the at Universite du Quebec a Chicoutimi on April 21, 2012 their isotopic similarity to all the other igneous rocks in relatively sodic content of plagioclase in these rocks, the deposit (Figs 5 and 6). Therefore, these magmas must which are of basaltic and andesitic composition, is due to have formed by magmatic differentiation processes taking the simultaneous crystallization and sequestering of Ca by anhydrite. The similarity of oxygen-isotope temperatures place in this large magma chamber.The more felsic ‘micro- calculated from anhydrite^quartz and anhydrite^magnet- diorites’ have higher La/Yb and lower Yb than the more ite mineral pairs to those calculated from quartz^magnet- mafic ‘andesites’, but similar Dy/Yb to them, consistent ite is consistent with co-crystallization of these three with amphibole fractionation (Fig. 10). However, sequester- phases (Stern et al., 2007). ing of Ca from the melt by the crystallization of anhydrite Cross-cutting relations and chronological (Fig. 4) data may also be responsible for inhibiting the growth of amphi- suggest that the emplacement of the Porphyry A pluton bole in the rocks themselves. Although amphibole is a occurred in conjunction with the generation of the common phase in other igneous rocks containing magmat- Cu-rich anhydrite breccia pipe into which it intrudes. ic anhydrite (Scaillet & Evans, 1999; Barth & Dorais, Generation of this anhydrite breccia by the exsolution of 2000), these generally contain 52 modal % anhydrite, Cu- and S-rich magmatic fluids from the parent magma and those that are volcanic are not fully crystallized and chamber occurred together with, or was followed immedi- equilibrated as are the plutonic rocks that form the ately by, the intrusion of the volatile-saturated and Cu- Porphyry A stock. and S-rich silicate magma from this same chamber that The unusual Cu- and S-rich Porphyry A magmas produced the small Porphyry A pluton. In this sense, we formed during a period of compressive deformation, uplift consider the Porphyry A stock and surrounding contem- and erosion (Skewes & Holmgren, 1993; Skewes & Stern, poraneous anhydrite breccia it intrudes to be the SO2-rich 1994; Kurtz et al., 1997; Garrido et al., 2002), with no evi- equivalent of a carbonatite complex, in which SO2-rich dence of coeval volcanic activity in the area of the deposit silicate melts and their associated magmatic fluids were (Fig. 4; Skewes et al., 2002, 2005; Kay et al., 2005). emplaced together to produce a variety of genetically Compressive deformation and lack of volcanic activity related rocks. during the Late Miocene may have prevented devolatiliza- The poikilitic texture of anhydrite in the Porphyry A tion of the magma in the large, long-lived magma chamber ‘microdiorite’ indicates late crystallization and growth of underlying the El Teniente deposit (Pasteris, 1996; anhydrite relative to the silicate phases in this rock. Stern Oyarzu¤n et al., 2001; Richards et al., 2001).These conditions
Excursion métallogénique - Chili 2012 Références page 123
1611 JOURNAL OF PETROLOGY VOLUME 52 NUMBERS 7 & 8 JULY & AUGUST 2011
allowed the SO2 and H2O contents of magmas near the dikes in the vicinity of El Teniente imply a significant tem- roof of this chamber to increase to the point that they gen- poral increase in the extent of hydration of the erated the multiple breccia pipes in the deposit, one of sub-Andean mantle below this region between the Late which was subsequently intruded by the Cu- and S-rich Miocene and the Pliocene. magmas that crystallized to form the igneous The higher Sr and La/Yb, and the lower Yand Yb of the anhydrite-bearing Porphyry A pluton. We consider volatile Pliocene lamprophyres compared with the Late Miocene transport by bubble formation and resorption (Cloos, basaltic rocks of the Teniente Volcanic and Plutonic 2001) as the physical process responsible for concentration Complex (Stern et al., 2010) suggest both a lower degree of of volatiles within the relatively cool cupola of the magma mantle partial melting and/or a greater proportion of chamber below the deposit (Fig. 11). The relatively high either garnet or hornblende in the mantle source of the pressure (41 kbar) at the roof of the magma chamber lamprophyres. A lower degree of partial melting is consist- underlying the El Teniente deposit at 44 km depth, com- ent with the much smaller volume of Pliocene compared bined with the Cl-rich nature of the volatile-saturated with Late Miocene igneous rocks in the vicinity of Downloaded from magma near the roof of this chamber, as evidenced by the Teniente (Stern & Skewes, 2005), and hornblende in the high salinity of fluid inclusions in hydrothermal breccia mantle source is consistent with both the presence of horn- pipe matrix minerals (Skewes et al., 2002), are other factors blende phenocrysts in the most mafic olivine-bearing that enhance the solubility of Cu and S in the magmas dikes and the extensive hydration of their source as implied near the roof of the chamber (Luhr,1990). by the results of the experimental studies described above. http://petrology.oxfordjournals.org/ The magmas that formed the anhydrite-bearing igneous Carmichael (2002) suggested that the source of Mexican rocks at El Teniente were also highly oxidized, as indicated lamprophyres is metasomatized mantle containing amphi- by both their high ratios of sulfates to sulfides and the pres- bole, similar to the mineralogy of the hornblende- ence of primary magmatic hematite. Recharge of a lherzolite mantle xenoliths observed (Blatter & magma chamber by hydrous, oxidized, mantle-derived Carmichael, 1998). mafic arc magmas (Fig. 11) has been suggested as the Significantly, the less mafic, olivine-free lamprophyre source of S in the case of other S-rich anhydrite-bearing ig- dikes have similar or higher La, but lower Yb, resulting in neous rocks (Hattori, 1993; de Hoog et al., 2004), as well as significantly higher La/Yb compared with the mafic for the large quantity of S in other large Cu deposits lamprophyres (Stern et al., 2010). We suggest that this is (Streck & Dilles, 1998; Hattori & Keith, 2001). at Universite du Quebec a Chicoutimi on April 21, 2012 the product of amphibole fractionation, as heavy REE are Lamprophyre dikes (3·85 0·18 to compatible in amphibole and amphibole is the most abun- 2·9 0·6 Ma) dant observed phenocryst in these lamprophyres. Nearly constant Dy/Yb ratios (2·0^2·3) for lamprophyres that The Pliocene post-mineralization lamprophyre dikes rep- vary between 50·3and61·1wt % SiO are also consistent resent the last phases of igneous activity within the 2 Teniente deposit. Mg-olivine phenocrysts and high MgO, with crystal^liquid fractionation involving amphibole Ni and Cr contents imply a mantle origin for the more rather than garnet (Davidson et al., 2007). Nearly constant mafic lamprophyre dikes (Stern et al., 2010). Petrologically Sr contents in the less mafic compared with the more similar lamprophyres are found in the Mexican volcanic mafic lamprophyres, and the lack of a negative Eu anom- belt, and considered to be the mantle-derived parent for aly, also indicate that plagioclase was not a significant frac- the more typical porphyritic hornblende andesites that tionating crystal phase, consistent with it not being a form the large central-vent stratovolcanoes in this belt phenocryst phase. However, the lack of plagioclase (Carmichael, 2002). Experimental studies suggest that the fractionation was due to the high H2O content of the phenocryst mineral assemblages of these Mexican lampro- lamprophyre magma, which suppresses plagioclase crystal- phyres, which are identical to those from the vicinity of lization and enhances the extent of amphibole crystalliza- Teniente, imply 46 wt % dissolved water in the lampro- tion, and not due to stabilization of garnet at the expense phyre magmas (Moore & Carmichael, 1998; Blatter & of plagioclase caused by high pressure in the magma Carmichael, 2001). This is consistent with a minimum of source or crystal-fractionation region as suggested by Kay 45·2 wt % water in olivine-bearing Mexican magmas as et al. (2005). Garnet may have existed in the mantle-source implied by the water content of melt inclusions in olivine region of the mafic lamprophyres, but we consider the (Cervantes & Wallace, 2003). Carmichael (2002) suggested trend to higher La/Yb and lower Yb observed between that between 6 wt % (the minimum to reproduce the the more and less mafic samples to be due to crystal frac- phenocryst mineral assemblage) and 16 wt % (the max- tionation involving amphibole and not to garnet fraction- imum to saturate the mantle-derived magma at 10 kbar ation. This reflects the increased importance of water in pressure) water, derived from the subducted oceanic litho- the mantle source of these rocks, and not an increase in sphere, was present in the Mexican lamprophyres. Thus the depth of magma generation or subsequent we conclude that the post-mineralization lamprophyre differentiation.
Excursion métallogénique - Chili 2012 Références page 124
1612 STERN et al. EVOLUTION OF EL TENIENTE CU^MO DEPOSIT
Regional temporal isotopic evolution indicate that mafic, intermediate and felsic igneous rocks Previous regional studies of igneous rocks at the latitude of of each of the episodes of igneous activity in the Andes at El Teniente documented significant changes in the isotopic the latitude of El Teniente, are, for each age group, isotop- compositions of magmas formed during the final stages of ically similar to each other (Figs 5 and 6).These data rule the Oligocene to Pliocene magmatic episode in the vicinity out significant intra-crustal contamination or melting of of the Teniente deposit (Figs 5 and 6; Stern & Skewes, isotopically heterogeneous Paleozoic and Mesozoic 1995; Nyst rom et al., 2003; Kay et al., 2005). These reflect Andean continental crust in the generation of the rocks of temporal changes in the isotopic composition of the each age group. source region of these Andean magmas, involving increas- Stern (1991, 2004), Stern & Skewes (1995, 2005) and Kay ing 87Sr/86Sr and decreasing 143Nd/144Nd, that are inter- et al. (2005) concluded, based on both geochemical argu- preted to indicate a progressive increase in the amount of ments and structural grounds, that significant ( 50 km) crustal components incorporated into the Andean fore-arc truncation by subduction erosion has occurred Downloaded from magmas through time. since 9 Ma across the arc^trench gap west of El Teniente, Our data indicate that the isotopic changes to higher and that an important part of the temporal isotopic 87Sr/86Sr and lower 143 Nd/144Nd observed between the changes observed in the igneous rocks generated through Early Miocene and Pliocene occurred in mantle-derived time at this latitude resulted from subduction erosion of olivine-bearing mafic rocks, specifically in the sequence the continental margin and increased contamination by from olivine-basalts of the Early Miocene Coya-Machal|¤ continental components of the sub-arc mantle. Subduction http://petrology.oxfordjournals.org/ Formation, to olivine-basalts of the Late Miocene Teniente of oceanic crust, pelagic and terrigenous sediments, and Mafic Complex, and finally Pliocene post-mineralization continental crust tectonically eroded off the edge of the continent, into the mantle-source region of Andean olivine-lamprophyre dikes (Stern et al., 2010) and magmas, may provide the large amounts of water, sulfur, olivine-bearing basaltic andesite lavas in the Cachapoal copper and boron involved in the generation of the giant river west of the Teniente deposit (Stern & Skewes, 1995). copper deposits of central Chile (Macfarlane, 1999). This trend must therefore result from isotopic variations Significantly, as discussed above, the change from in the sub-arc mantle and may be caused by increased olivine-basalts in the Late Miocene Teniente Mafic mantle-source region contamination by subducted compo- Complex to the isotopically distinct olivine þ amphibole nents (Stern, 1991). Stern et al. (2010) presented a model for lamprophyre post-mineralization dikes emplaced in the at Universite du Quebec a Chicoutimi on April 21, 2012 Sr and Nd isotopic variations caused by different amounts Pliocene implies progressively greater hydration of the of contamination of the sub-arc mantle involving sub- sub-arc mantle over this time period. There is thus a cor- ducted trench sediment containing both marine and terri- relation between the increasing water content of the genous components as calculated by Macfarlane (1999). Andean sub-arc mantle and the increasing degree of con- The model suggests that Coya Machal|¤ magmas could tamination of this mantle by subducted components as form by melting of a mantle contaminated by 1% sub- indicated by the isotopic data. A similar correlation has ducted sediment, the Teniente Volcanic and Plutonic been documented in magmas erupted from different re- Complex rocks by 2% subducted sediment, the Pliocene gions of the subduction-related trans-Mexican volcanic post-mineralization lamprophyre dikes by 4% subducted arc (Cervantes & Wallace, 2003). sediment, and the Late Pliocene basaltic andesite lavas in the Cachopoal river valley by 6% subducted sediment. Kay et al. (2005) presented a model of assimilation of CONCLUSIONS continental crust, consisting of Paleozoic^Triassic granitic Lindgren & Bastin (1922) suggested that El Teniente basement, by a primitive Coya-Machal|¤ olivine basalt that ‘affords convincing evidence of the intimate genetic con- produces the same isotopic variations as does the source nection between igneous rocks and ore deposits’.The petro- region contamination model, but with much higher pro- chemical constraints on the genesis of the igneous rocks in portions of assimilation: 15% to produce the Late the deposit provide important insights into this genetic Miocene Teniente Volcanic and Plutonic Complex rocks, connection. We conclude that both the magmatic^hydro- 30% to produce the Pliocene lamprophyre dikes and thermal breccia pipes and the sequence of felsic porphyry 40% to produce the Cachopoal lavas and northern SVZ plutons in the deposit were derived from a large, long-lived volcanic rocks. However, these amounts of assimilation of ‘productive’ magmatic system, one that involved at least 3 high-SiO2, low-Sr granite are clearly inconsistent with the 4600 km of Andean magma over an 3Myrperiod primitive mafic chemistry of the olivine-bearing basalts in (Fig. 11). This system developed in the upper crust at a the Teniente Mafic Complex and olivine lamprophyre depth of at least 4 km below the surface, as indicated by dikes. Furthermore, the isotopic data, including the recent- the fact that the depth of the roots of the largest breccia ly published data for O and Hf isotopes in zircons from pipes at El Teniente, which are as yet unknown, must all the felsic units in the deposit (Mun‹ oz et al., 2009), extend down to at least this depth.
Excursion métallogénique - Chili 2012 Références page 125
1613 JOURNAL OF PETROLOGY VOLUME 52 NUMBERS 7 & 8 JULY & AUGUST 2011
Decreasing subduction angle, compression and crustal SUPPLEMENTARY DATA thickening also played an important role in the generation Supplementary data for this paper are available at Journal of the El Teniente deposit. They caused the termination of of Petrology online. volcanism, which sealed the magmatic system and allowed volatiles to build up in concentration near the roof of the system, until volatile pressure exceeded confining pressure, generating the large mineralized breccia pipes that charac- REFERENCES terize the deposit. Together they also resulted in a decrease Annen, C., Blundy, J. D. & Sparks, R. S. J. (2006).The genesis of inter- in the volume of the sub-arc mantle and consequently an mediate and silica magmas in deep crustal hot zones. Journal of increase in the relative input of slab-derived fluids into Petrology 47,505^539. this mantle, which was the source region of the magmas Atkinson, W. W., Jr, Souviro¤ n, S., Vehrs, T. & Faunes, A. (1996). Geology and mineral zoning of the Los Pelambres porphyry that fed and maintained the magmatic system from which
copper deposit, Chile. In: Camus, F., Sillitoe, R. H. & Downloaded from the metals, mineralizing fluids and host igneous rocks Petersen, R. (eds) Andean Copper Deposits. Society of Economic were derived. The production of progressively more Geologists Special Publications 5, 131^155. hydrated and oxidized magmas between the Late Bachman, O. & Bergantz, G. W. (2004). On the origin of crystal-poor Miocene and the Pliocene may have facilitated the transfer rhyolites: Extracted from batholithic crystal mushes. Journal of of greater amounts of oxidized sulfur and metals out of Petrology 45, 1565^1582.
the roof of the magmatic system into the shallow crust, be- Baker, M. C. W. & Francis, P.W. (1978). Upper Cenozoic volcanism in http://petrology.oxfordjournals.org/ the Central Andes; ages and volumes. Earth and Planetary Science cause in less oxidized and volatile-poor systems sulfur and Letters 41,175^187. metal are retained near the base of the system as immis- Barth, A. P. & Dorais, M. J. (2000). Magmatic anhydrite in granitic cible sulfur and pyrrhotite (Richards, 2003, 2005). rocks: first occurrence and potential petrologic consequences. In summary, the major factor in producing the giant El American Mineralogist 85, 430^435. Teniente deposit was the development of a large, long-lived Blatter, D. K. & Carmichael, I. S. E. (1998). Hornblende peridotite magma chamber or magmatic plumbing system. The xenoliths from central Mexico reveal the highly oxidized nature of transfer of metals from the magma in this chamber took subarc upper mantle. Geology 26, 1035^1038. Blatter, D. K. & Carmichael, I. S. E. (2001). Hydrous phase equilibria place by the exsolution of volatiles from the roof of this of a Mexican high-silica andesite: a candidate for mantle origin? system as it crystallized, generating large mineralized Geochimica et Cosmochimica Acta 65, 4043^4065. at Universite du Quebec a Chicoutimi on April 21, 2012 breccia pipes and associated veins. Progressively more dif- Brown, S. J. A. & Fletcher, I. R. (1999). SHRIMP U^Pb dating of pre- ferentiated felsic intrusions, of smaller and smaller eruption growth history of zircons from the 340 ka Whakamaru volume, also intruded into the deposit, but these truncated Ignimbrite, New Zealand: evidence for4250 k.y. magma residence already altered, brecciated and mineralized rocks, and times. Geology 27,1035^1038. were not the major process by which Cu was emplaced Burgos, A. (2002). Petrograf|¤a y Geoqu|¤mica de la Diabasas y Diques Basa¤ lticos que constituyen las ‘Andesitas de la Mina’ en el yaci- into the deposit (Skewes et al., 2002, 2005).These felsic por- miento El Teniente, VI Regio¤ n, Chile. Universidad de Concepcio¤ n: phyries have the same isotopic composition as all the Memoria de T|¤tulo, 108 p. more mafic lithologies associated with the deposit, and Camus, F. (1975). Geology of the El Teniente orebody with emphasis they did not form by progressively deeper melting within on wall-rock alteration. Economic Geology 70, 1341^1372. the isotopically heterogeneous Andean crust. They were Cannell, J., Cooke, D., Walshe, J. L. & Stein, H. (2005). Geology, generated instead by a complex combination of differenti- mineralization, alteration, and structural evolution of El Teniente porphyry Cu^Mo deposit. Economic Geology 100, 979^1004. ation processes in the large relatively shallow magma Carmichael, I. S. E. (2002).The andesite aqueduct: perspectives on the chamber below the deposit, including recharge of the base evolution of intermediate magmatism in west^central (105^998W) of the chamber by mantle-derived mafic magmas, thermal Mexico. Contributions to Mineralogy and Petrology 143, 641^663. stratification and convection, magma mixing and crystal- Cervantes, P. & Wallace, P. (2003). The role of water in subduction lization, crystal^liquid fractionation involving amphibole, zone magmatism: New insights from melt inclusions in high-Mg bubble formation, and volatile transfer and concentration basalts from central Mexico. Geology 31, 235^238. in the upper cooler parts of the chamber. Charrier, R., Baeza, O., Elgueta, S., Flynn, J. J., Gans, P., Kay, S. M., Mun‹ oz, N., Wyss, A. R. & Zurita, E. (2002). Evidence for Cenozoic extensional basin development and tectonic inversion south of the flat-slab segment, southern Central Andes, Chile ACKNOWLEDGEMENTS (338^368S.L.). Journal of South American Earth Sciences 15, 117^139. Cline,J. S. & Bodnar, R. J. (1991). Can economic porphyry copper We thank Patricio Zun‹ iga, superintendent of geology at El mineralization be generated by a typical calc-alkaline melt? Teniente, for access to and logistic support within the Journal of Geophysical Research 96, 8113^8126. mine. The manuscript benefited from constructive com- Cloos, M. (2001). Bubbling magma chambers, cupulas and porphyry ments from Jeremy Richards, Victor Maksaev and two copper deposits. International Geology Reviews 43, 285^311. anonymous reviewers. Dan Mitchell assisted with the Cornejo, P., Tosdal, R. M., Mpodozis, C., Tomlinson, A. J., Rivera, O. figures. & Fanning, C. M. (1997). El Salvador, Chile, porphyry copper
Excursion métallogénique - Chili 2012 Références page 126
1614 Excursion métallogénique - Chili 2012 Références page 127 Excursion métallogénique - Chili 2012 Références page 128 Excursion métallogénique - Chili 2012 Références page 129 Excursion métallogénique - Chili 2012 Références page 130 Excursion métallogénique - Chili 2012 Références page 131 Excursion métallogénique - Chili 2012 Références page 132 Excursion métallogénique - Chili 2012 Références page 133 Excursion métallogénique - Chili 2012 Références page 134 Excursion métallogénique - Chili 2012 Références page 135 Excursion métallogénique - Chili 2012 Références page 136 Excursion métallogénique - Chili 2012 Références page 137 Excursion métallogénique - Chili 2012 Références page 138 Excursion métallogénique - Chili 2012 Références page 139 Excursion métallogénique - Chili 2012 Références page 140 Excursion métallogénique - Chili 2012 Références page 141 Excursion métallogénique - Chili 2012 Références page 142 EconomicGeology Vol. 80, 1985, pp. 1544-1565
Porphyry Copper and Tourmaline Brecciasat Los Bronces-Rio Blanco, Chile
FRED W. WARNAARS, EssoPapua New GuineaInc., P.O. Box 2174, Lae, PapuaNew Guinea
CARMEN HOLMGREN D., AND SERGIO BARASSIF. CompadiaDisputada de las CondesS.A, Casilla 16178, Correo9, Providencia,Santiago, Chile
Abstract
The Los Bronces-RioBlanco deposit is locatedon the west side of the Andesin central Chile about69 km fromSantiago. Los Bronces is a brecciacomplex superimposed on the west sideof an earliermajor porphyry copper system. The Rio Blancomine currently exploits the north-centralpart of thisporphyry deposit and has started operating a largecopper-bearing tourmalinebreccia, Sur-Sur, about 2 km southof the presentmine. The Los Bronces-RioBlanco depositwas formed on the east side of the San Francisco batholith.This intrusionis stronglyperaluminous and has a calc-alkalinecomposition with an alkali-calcicaffinity. The batholithtook a minimumof 11.5 m.y.to formfrom the earlyMiocene (20.1 m.y.) to the late Miocene(8.6 m.y.). The porphyrycopper mineralization, alteration, and coppertourmaline breccias were formedover a period of at least2.5 m.y. between7.4 and 4.9 m.y. ago. A postmineralvolcanic neck or diatremeat La Copaerupted within, andremoved a large segmentof, the northernpart of the porphyrycopper system, thus marking the lastevidence of magmaticactivity in the area.K-Ar age determinationsof biotitesindicate the diatreme eruptedin the earlyPliocene between 4.9 and 3.9 m.y. ago. The prebrecciaporphyry system exhibits propylitic, sericitic, silicic, and potassic alteration. A uniquealteration feature of thissystem is the replacementof maficminerals by specularitc and/ortourmaline within the propyliticzone. The porphyrysystem contains disseminated and stockworkcopper-iron-molybdenum sulfide mineralization within an area of about12 km2. LosBronces is composedof at leastseven different copper-bearing tourmaline breccias that form one large contiguouskidney-shaped body about 2 km long and 0.7 km wide, at the presenterosion surface. The brecciabody cropsout at elevationsbetween 4,150 and 3,450 m. The variousbreccias are characterizedby their locations,matrices, clasts, shapes, types, anddegrees of mineralizationand alteration. The brecciasare usuallymonolithic but in some casesare bilithic or heterolithicwith mostclasts consisting of quartz monzoniteor andesitc with locallyminor amountsof quartz latite porphyry,monzodiorite, and vein quartz.The brecciamatrices consist of variableamounts of quartz, tourmaline,specularitc, anhydrite, pyrite, chalcopyrite,bornitc, molybdenite,sericite, chlorite, and rock flour. The sevendifferent breccias types are identifiedfrom oldestto youngestas Ghost, Central, Western,Infiernillo, Anhydrite, Fine Gray,and Donoso. The brecciacomplex has sharp contacts with the surroundingintrusive rocks and andesites. Internally, the brecciacontacts are locally well defined,but elsewherethey coalesce,interfinger, or displaygradational contacts. Breccias at Los Broncesare interpretedas being emplacedexplosively, followed by collapseafter pressurerelease of hydrothermalfluids. The primarymineral distribution is bestknown in the Donosobreccia which hasbeen the centerof miningactivity since its discovery in 1864. In spiteof the coarseand irregular nature of the sulfidesin the matrix,chalcopyrite, pyrite, and specularitcat the 3,670-m, open-pit operatinglevel showa tendencyto be distributedin irregularshells in whichone of the three mineralspredominates in any one shell. The transitionsbetween shellsare rapid. Several semiellipsoidalshells of alternatinghigh and low coppergrades are alsoapparent from the copperdistribution of undergroundlevel 3640 andfrom variouscross sections. The shellsare approximatelyvertical, subparallel to the Donosobreccia contacts, which dip inward. Secondaryenrichment enhanced the primary gradein the southerntwo-thirds of the Los Broncesbreccia complex and in muchof the surroundingporphyry copper system. The degree anddepth of enrichmentare a functionof brecciaand fracture permeability and extend to a depthof morethan 500 m in certainfavorable sectors. The shapeand depth of the enrichment blanketand overlyingleached capping suggest-that the enrichmentprocess is relatedto the presentground-water regime and is still active.
0361-0128/85/442/1544-22$ 2.50 15 4 4 Excursion métallogénique - Chili 2012 Références page 143 LOSBRONCES-RIO BLANCO: Cu & TOURMALINEBRECCIAS 1545
Introduction v v • D a • v THE Los Broncesmine owned by CompafiiaMinera Disputadade Las CondesS.A., a subsidiaryof Exxon vv v t.../ vJ MineralsCompany, is located in the highAndes about 69 km from Santiago(Fig. 1). The presentopen pit is at the northend of a hydrothermalbreccia orebody Bronces•,V-•12• •o •cp _ at an elevation of about 3,600 m above sea level. Pro- • •' • - J,. •. E ' 'A-•c•c•l ductionamounts to about10,000 metrictons per day •/ [ / v R/o B•cb/ • / Bxl at an averagegrade of 1.25 percentCu, with a 0.75 • / • percentCu cut-offgrade. "n/I/oX S•n Enrlque•Bx- •*[. -}1 Rio Blancois an undergroundmine ownedand op- • %> •:.•'•_• h erated by the Andina Division of the state-owned Codelco-Chile. Present production at Rio Blancois 14,000 metric tonsper day, at an averagegrade of about 1.2 percent Cu. The mine is about 800 m east of Los Broncesin the headwaters of the Blanco River, draining northward,whereas Los Broncesis in the headwaters of the San Francisco River draining -•L• Co• 8- S•mpleIocmion• southwestward(Fig. 2). At present,Andina has started ...... • Housing,offices operatinga largecopper tourmaline breccia, Sur-Sur, L t•te • dQcite nearly 2 km southof the presentRio Blancomine, porphyries•.• River with an averagegrade of 1.98 percentCu for the first breccl•s HydrothermolBxOuterBreccio hmlt of the four yearsof production. •tholith Blancoporphyry AugustlnNazario E. wasthe firstto discovercopper at Los Broncesin 1864. High-gradedirect-shipping Farel•nes Inferred ore containing10 to 20 percent copper was mined v•Andesitesformationof•e• Escondidacopperdepositfault and transported on muleback intermittently until FIG. 2. Simplifiedgeologic map of the LosBronces-Rio Blanco 1920. By that time, the grade had dropped to 4.5 area,modified from Irarrazaval et al. (1979) andStambuk (1982). percentand the firstof severalflotation concentrators was installed near the mine site. CompafiiaMinera Disputada de las Condes S.A. was formedin 1916. It acquireda distantmine, E1 Soldado,
?• 71 70 northwestof Santiagoin the early 1950s.Exxon Cor- poration,through a subsidiary,purchased over 90 percentof Disputada'sshares in early 1978. At pres- ent, ExxonMinerals Company operates the company's 32 two minesand a copper smelter.Exxon carried out a mineral resource evaluation at Los Bronces from Scale 0 50 t00Km February1978 to May 1981. A total of 558 holeswas i i i drilled in and around the ore depositwith a total Explanation meterageof over 121,000 m. The presentpaper sum- • River marizesthe geologyas understood from this drilling programand from surfaceand underground mapping /• Recent volcanoes of old workingssince 1977. A uniformnomenclature [-• Tertiaryintrusive and loggingtechnique was introducedduring the evaluationprogram because fifteen geologists partic- •,• Forellonesformation ipatedin it. The terminologyand nomenclature used i• •[--•']Abonico formation :54 is explainedin AppendixI...... The geologicframework of the Los Broncesbrec- 'B• LoVoile volcanics ciaswas presented on a preliminarybasis at the annual o meetingof the AmericanInstitute of Mining, Metal- :•'i• Cretaceousintrusive lurgical,and Petroleum Engineers in 1980. Sincethen
'•]1Candretoceo volconicsus sediments :55 petrological,geochemical, and mineralogicalstudies haveenhanced the understandingof the depositand --•lbasementPoleozoic elucidatedthe mineralzoning. The depth projection MO. 1. Locationmap and simplifiedgeology of central Chile of each of the breccias has been better documented. modifiedafter Drake et al. (1982). Severalage dates near the depositand in the region
Excursion métallogénique - Chili 2012 Références page 144 1546 WARNAARS,HOLMGREN D., AND BARASSIF.
havebeen determined,establishing the localand re- the area, such as that at Colorado-La Parva about 15 gionaltime frame of the deposit. km southof Los Bronces(Drake et al., 1982). Regionaland Local Setting San Francisco Batholith The central Andes between El Teniente and Acon- The LosBronces-Rio Blanco porphyry copper sys- caguapeak are underlainby Jurassicto Cretaceous tem and associated breccias are within the eastern marine stratathat are exposedalong the border with part of the SanFrancisco batholith. This intrusionis Argentina(Fig. 1). These are discordantlyoverlain exposedover an areaof about200 km2, 20 km in a by continentalvolcanic rocks and minor continental north-south direction and 10 km in an east-west di- sedimentsof the Abanicoand/or Coya-Machal{ For- rection.The batholithis largely composedof quartz mations(Aguirre, 1960; Klohn, 1960). These for- monzoniteand quartz monzodiorite(Fig. 3) in the mationsare largely time equivalent (Drake et al., nomenclature of Streckeisen (1976). Aplitic and 1982). The eastern belt of the Abanico Formation syeniticphases are probablylate magmaticdifferen- may be as old as Paleocene(62 m.y.). The western tiatesand form mostlydikes. The texture and com- belt of the AbanicoFormation may be middleto late positionof the batholithare highly variable.Grain Oligocene.After a period of deformation,uplift, and sizesvary from fine to coarseand textures from equi- erosion,renewed volcanismproduced andesitc flows granularto porphyritic.Whole-rock analyses of age- and pyroclasticsinterbedded with continentalsedi- dated rocks are given in Table 1 and their sample ments.These representthe FarellonesFormation of sitesare indicatedin Figure2. The AI•Oa/CaO+ K•O late Oligoceneto Miocene age (Vergaraand Drake, + NagO ratio varies between 1.37 and 1.57, sug- 1978; Drake et al., 1982). gestingthat the rock suite is stronglyperaluminous Widespreadintrusive magmatism along the central (Shand,1927). high Andes formed numerousintrusions within the Five new analysesof unalteredspecimens of the volcanic terrain. One of these, the San Francisco SanFrancisco batholith combined with two analyses batholith, hoststhe Los Broncesand Rio Blancomines. of Oyarzun(1971) and a few of Blondel(1980) and The present undergroundRio Blancomine is in the Lopez and Vergara (1982) from unaltered samples centralpart of a porphyrycopper system (Fig. 2). givean alkali-limeindex (Peacock, 1931) of about56 This systemexhibits propylitic, sericitic, silicic, and which is on the border between calc-alkaline and al- potassichydrothermal alteration assemblageswith kali-calcicmagma affinity. Keith (1978) distinguished disseminatedand stockwork copper-molybdenum typesof magmasuites based on KgO/SiO2diagrams. sulfide mineralization over an area of about 12 km 2. In his classification,a calc-alkalinemagma is defined The limitsof the porphyrycopper as outlined in Fig- as having a KgO percent between 1.2 to 2.5 at 57.5 ure 2 mark the boundarybeyond which no mineral- percentSiOg. Plotting the availablechemical analyses ization nor hydrothermalalteration is visible in the on a K•O/SiO• diagramand interpolatingthe results intrusivehost rock. The porphyrycopper system is irregularlyoval shaped, with its longestaxis trending Q northeast. The Los Broncesmine is within the northern part of a largekidney-shaped hydrothermal copper-tour- malinebreccia complex that was emplacedin, and • Textures superimposedon, the westernpart of the earlierpor- • Equigranular phyry system. XPorphyritlc In addition,several hydrothermal copper-tourma- 6O line brecciasoccur south of the Rio Blancomine (Fig. 2) including the Sur-Sur breccia (Stambuk et al., 1982), and the Monolito, San Enrique, Rio Blanco, 4O Cascada,and Don Luis brecciasshown in Figure 2. Thesebreccias may havebeen part of a secondcon- tiguousbreccia body that wasdisrupted by late and postmineraldacitic to latitic intrusivequartz porphy- ries. Elsewhere,in the vicinity of the Los Bronces- Rio Blanco systemare other brecciassuch as San Manuel,Sur, andAmericana. A postmineralvolcanic neck,or diatreme,at La Copa,north of the Rio Blanco A I0 55 65 90 P mine,probably removed a significantpart of the cen- FIG. 3. Rock compositionof 53 samplesof the San Francisco ter of the initialporphyry copper system. This dacitic batholithcollected near LosBronces and plotted on Streckeisen's neckis time equivalentwith other silicicvolcanics in (1967) classificationdiagram, after Cuadra (1980).
Excursion métallogénique - Chili 2012 Références page 145 LOS BRONCES-RIOBLANCO: Cu & TOURMALINEBRECCIAS 1547
T•,BLE1. Whole-RockAnalyses of IgneousRocks near Los Bronces
Sample no. LB- 1 LB-2 LB-3 LB-7 LB-8 LB- 10 LB- 11 LB- 12
SiO2 59.9 65.3 60.5 63.06 69.61 67.80 69.50 66.13 AI20• 17.1 15.8 16.8 15.40 14.20 15.70 15.50 16.40 Fe•O• 1.40 1.20 2.08 3.65 1.25 0.83 0.63 3.77 FeO 4.28 2.41 3.00 1.41 1.42 0.83 0.91 1.10 CaO 4.24 3.04 5.42 3.43 1.58 2.52 1.75 0.12 MgO 2.44 1.53 3.05 1.74 0.95 0.33 0.41 3.28 Na•O 4.72 4.68 4.60 4.24 4.26 4.38 4.75 0.11 K•O 3.16 3.55 2.22 2.95 3.23 2.57 2.43 4.90 P•O5 0.23 0.15 0.20 0.24 0.10 0.23 0.28 0.34 MnO 0.11 0.08 0.09 0.08 0.05 0.11 0.02 0.011 TiO2 0.78 0.46 0.65 •0.01 •0.09 0.40 H20 0.11 0.16 0.12 0.098 0.21 0.17 0.19 0.61 H•O + 0.11 0.20 0.13 0.26 0.60 1.50 0.66 1.58 C1 0.027 0.01 A/CNK • 1.41 1.40 1.37 1.45 1.57 1.66 1.74 3.20
• A/CNK, molecularratio ofAl•Oa/CaO + Na20 + K•O; if ratiois greaterthan 1.1, the rockis peraluminous(Shand, 1927) Locationsof samplesare given in Figure2; LB-1 = quartzdiorite, LB-2 = granodiorite,LB-3 = granodiorite,LB-7 -- granodiorite, LN-8 = quartzmonzonite, LB-10 -- daciticfragmental breccia, LB-11 -- daciteporphyry dike, LB-12 = mineralizedsericitized quartz monzonite
givesa K57.,5index of about 2.0, indicatinga calc-al- pearance.Fragments of mineralizedbreccias, min- kaline nature for the San Francisco batholith. eralized andesite,and mineralizedquartz monzonite Late and Postmineral Rocks are common. Fragmentsof sedimentaryrocks also occurand were probablyderived from a deepsource. Dacite, latite, and quartz porphyrieswere em- Dacite porphyrydikes cut throughthe diatremebut placed during the waningof or after the porphyry do not crossits perimeter. copper mineralization, south and north of the Rio The chemicalcompositions of sampleLB-10 and Blancomine. Cepeda(1981) referredto someof these LB-11 (Table 1) from La Copa, in additionto three rockssouth of the Rio Blancomine as tonalitepor- samplesby Latorre (1981), suggesta slightlylower phyries. Stambuket al. (1982) introducedthe term alkalinitythan the SanFrancisco batholith. The alkali- Rio BlancoFormation, subdividing the postporphyry lime index of Peacock(1931) is around 59. The in- copperrocks into three members(as if they were terpolated K57.5index is about 1.5 and is lower than stratigraphic units): Don Luis, Rio Blanco, and La the index of the batholith, suggestinga slightlyshal- Copa.The Don Luis intrusivephase consists of dacite lowerorigin of the magma(Keith, 1978). The La Copa porphyriesbut also includessome mineralized tour- rockshave a higher peraluminousindex (1.66-1.74) maline breccias located southeast of the Rio Blanco than those of the San Francisco batholith. mine. The Rio Blancointrusion is largelya quartz la- tite and daciteporphyry dike a few hundredmeters Radiometric Dates wide orientednorth-south and intermittently exposed Initial dating of mineralized and unmineralized over a strikelength of at least3 km. Igneousbreccias rocksnear Los Broncesand Rio Blancowas done by with fragmentaltextures are recordedby Stambuket Vergaraand Drake (1979). They concludedthat the al. (1982) aspart of thisintrusive phase. Minor copper SanFrancisco batholith was emplacedbetween 13.6 stainingin weakly altered porphyriesin the vicinity and 8.4 m.y. ago based on K-Ar determinationsof of the Rio Blancomine where they cutthe porphyry primarybiotite from three differentrock samples. copper depositsuggests that this intrusionwas em- New agedeterminations (Table 2) supporta period placedafter the mainphase of porphyrycopper min- of at least11.5 m.y. for the formationof the SanFran- eralization and alteration and as such is a late mineral ciscobatholith, from 20.1 m.y.to 8.6 m.y. ago.Sample emplacement.The volcanic neck or diatreme of La LB-1 is from fresh unmineralizedquartz monzonite Copais the youngestmagmatic event and forms steep taken from a recent road cut on the south side of the walls,becoming only slightlysmaller at depth. The batholith, about 10 km from the Los Bronces mine neck consistsof fragmentalsand volcanicbreccias of (Fig. 2). The hornblende(20.1 m.y.) yieldsan appar- dacitic compositionwith an aphaniticgroundmass ent age 4.2 m.y. older than the biotite (15.9 m.y.) with sparsesmall phenocrysts of quartz,biotite, pla- from the same sample. Sample LB-2 representsa gioclase,and/or sanidine. Flow texturesand layering quartzmonzonite from a winterrefuge adit alongthe at the southside of the neckgive an ignimbriticap- road betweenthe concentratorand the mine. Only
Excursion métallogénique - Chili 2012 Références page 146 1548 WARNAARS,HOLMGREN D., AND BARASSIF.
TABLE2. Agesof Nine Mineral Samplesfrom LosBronces-Rio the heat sourceof the porphyrycopper or of the vol- Blanco,Chile (for analyticalinformation see Appendix II) canicneck (Hart, 1964; Warnaarset al., 1978). The lower ageof the biotite suggeststhat reheatingtook Sampletype Sampleno. Age (m.y.) place at a lower temperature. Quartz monzonite LB-1 Hb 20.1 q- 2.0 Previouslypublished age datesfrom mineralized Quartz monzonite LB-1 Bi 15.9 q- 0.6 rock at Los Broncesand Rio Blanco (Vergara and Quartz monzonite LB-2 Bi 11.3 q- 0.4 Drake, 1979; Blondel,1980) giveages of 7.4 and5.2 Hornblende diorite LB-3A Hb 18.5 ___1.7 Hornblende diorite LB-3A Bi 12.0 q- 0.5 m.y.Blondel selected a samplefrom the centerof the Granodiorite LB-7 Hb 8.6 _+ 0.9 porphyrysystem and dated secondarybiotite (5.2 Granodiorite LB-7 Bi 7.9 q- 0.4 m.y.). Vergaraand Drake'ssample was taken from Dacite porphyry LB-10 Bi 4.8 q- 0.2 well-mineralizeddrill core below the present open Dacite porphyrydike LB-11 Bi 4.9 ___0.2 pit in theDonoso area in oneof the youngestbreccias Sericitized quartz monzonite LB-12 Se 5.2 q- 0.3 at LosBronces. They dated the biotite, as was verbally Abbreviations: Hb -- hornblende, Bi = biotite, Se = sericite confirmedby Drake, andnot the plagioclase,as was statedin their publication.This biotite is almostcer- tainto be primary,since no secondary biotite is known to occurin that part of the orebody.Because of Ar biotite could be dated (11.3 m.y.) as insufficient degassing,the age musthave been fully reset and hornblendewas available. Sample LB-3 wascollected representsthe minimumage of the startof the hy- 600 m northeastof the open pit at Los Broncesand drothermal activity. Alteration and mineralization 100 m westof the daciteneck just belowthe andesitc couldhave started earlier, before 7.4 m.y., basedon roof pendant.The ageof the hornblendewas deter- the assumptionthat no time periodshould exist be- mined as 18.5 m.y. and the biotite as 12 m.y. The tween the startof the hydrothermalactivity and the youngestdates of the batholithcome from a sample end phaseof the coolingperiod of the batholithas (LB-7) collectedin the DoloresValley north of the representedby the youngestdates (8.6 and7.9 m.y.) San Francisco concentrator 5 km west of Los Bronces. from unmineralizedintrusive rocks (LB-7), collected The hornblendeand biotite agesof thisrock, 8.6 m.y. 5 km away from the deposit. and 7.9 m.y., respectively,are within the limits of Hydrothermalsericite (sample LB-12) wasdated analyticalerror. from a drill core of a well-mineralized,pervasively Hornblendeis the bestmineral to retainradiogenic sericitizedquartz monzonite located between the Los argon(Hart, 1964).The differencebetween the oldest Broncesbreccia body andthe Rio Blancomine. The hornblendedated (20.1 m.y.) and the youngest(8.6 age of 5.2 m.y. representsthe feldspar-destructive m.y.) representsthe minimumtime spanfor the em- phaseof porphyrycopper alteration, which predated placementand coolingperiod of the San Francisco the formation of the Los Bronces breccias. These batholith.A periodof 11.5 m.y. for the emplacement brecciasmust have been emplaced within a shorttime and coolingof the batholith seemslong. However, period,between 5.2 and 4.9 m.y. or lesswhen the intrusions on the west flank of the Andes from E1 Ten- postmineralLa Copadacitic neck was emplaced. We iente throughLos Pelambres(Fig. 1) and possibly conclude that the Los Bronces-RioBlanco hydro- farthernorth and south, may be part of onelarge Ter- thermalsystem, including the brecciaformation, re- tiary batholith that was only recently unroofed. quiredat leasta time spanof 2.5 m.y. to form,be- Batholiths are known to form and consolidate over tween 7.4 and 4.9 m.y. ago. longperiods of timesuch as the SierraNevada batho- Structure lith (about 131 m.y., Albers, 1981) and the Jurassic to CretaceousAndino batholith (about 104 m.y., On a globalplate tectonic scale it maybe significant Zentilli, 1974). that the LosBronces-Rio Blanco system is locatedat SamplesLB-10 and LB-11 (Table 2) are from the the intersectionof the Andesmountain range and the postmineralLa Copadacite neck. SampleLB-11 is a eastwardprojection of the east-northeast-oriented daciteporphyry dike within the neck.The agesof 4.8 JuanFernandez ridge southof the Nazcaplate and and 4.9 m.y. are comparableto the agesof biotites the Challengerfracture zone, off the coastof central from the diatremedetermined by Quirt et al. (1971) Chile (Minsteret al., 1978; Frutos,1981). The struc- andVergara and Drake (1978), whichrange from 4.9 turalintersection may have caused a zoneof weakness to 3.9 m.y. for a fecundmagma to riseinto the upperpart of the The discrepancyof 6.5 m.y. between the horn- crust. blendeand biotite agesin sampleLB-3 (18.5 and 12 The most prominent geologi.c structural trend m.y., respectively)can be best explainedby postu- within and adjacentto the depositis generally latingdegassing of radiogenicAr in biotitesclose to N 60 ø E, typically expressedas joints or fractures
Excursion métallogénique - Chili 2012 Références page 147 LOS BRONCES-RIOBLANCO: Cu & TOURMALINEBRECCIAS 1549
with only sporadicminor fault movements.These The brecciacontacts dip inward on the north, west, structureswere mineralizedbefore, during, and/or and southmargins. The easterncontact is nearlyver- after the emplacementof the variousbreccias. Early tical, suggestinga westwardtilt of the complexof veins that follow this N 60 ø E orientation contain about 15ø after emplacement. mostlyquartz and tourmaline,with only a minor Each breccia at Los Bronces has its own character- amount of s•ulfides. istic matrix, clasts,shape, vugs, type, and degreeof Late, postbrecciamineralized faults and late veins mineralization and alteration. One of the most im- cut the brecciawall and surroundingrocks. Several portantparameters used in distinguishingthe various pebbledikes also follow this east-northeast direction. breccias at Los Bronces is the nature of the breccia They mostlycontain rock flour and occasionallysul- matrix. The matrix consistsof varying amountsof fidesand representa phaseof late-stageventing. quartz, tourmaline,specularite, pyrite, chalcopyrite, Another structural orientation, N 10 ø to 30 ø W, is bornite, molybdenite, chlorite, anhydrite, sericite, more prominentin the andesiteseast and southeast and rock flour. of the brecciacomplex, but it is weaker andless well The brecciasare generallymonolithic but in some developednear Los Bronces.This persistenceaway casesare bilithic. Most of the clastsare composedof from the systemsuggests that the N 10ø to 30 ø W quartz monzonite or andesite, rarely with minor direction is older than the N 60 ø E trend characteristic amountsof quartz latite porphyryand monzodiorite. of the orebody(Barassi et al., 1979). The relativeproportions ofclasts depends on the host A major inferred fault, the Escondida,has an N 40 ø rock within which the breccia was formed. Clasts E orientationand runs parallel to and south of the generallyrange from coarseto very coarse(50-250 upper part of the San FranciscoRiver (Fig. 2). By mm). Some,however, measure tens of metersacross, dextralmovement, it juxtaposedstrongly altered and especiallyalong the outeredge of the brecciacomplex mineralizedquartz monzoniteson the southeastern or alongthe contactbetween two interfingeredbrec- side againstless altered and sparselymineralized cias.The clastsare mostlyangular to subangular,sug- quartz monzoniteson the northwesternside. gestingminor abrasion, rapid emplacement, and quick The Escondidafault must have been active prior coolingof hydrothermalfluids. Smaller clasts, partic- to the emplacementof the coppertourmaline brec- ularly thosein partsof the Centralbreccia, are more cias, because the overall outline of the breccia com- roundedand more hydrothermallyaltered. plex is not offset.The fault was, in fact, floodedand The presenceof fragmentsof one breccia type sealedby abundantquartz and tourmaline during the within anotheris the best criterion for ascertaining early barrentourmaline phase. Because of poor ex- relative age relationshipsbetween breccias.Cross- posure,the fault is difficultto recognizein the field. cuttingrelationships are lessuseful, particularly in drill cores,because it is generally not clear which Breccias at Los Bronces phaseis younger.Rarely, a fragmentof one breccia, Tourmaline breccias form the most favorable host containinga clastof older breccia,forms a clastin a rock for copper and molybdenummineralization third breccia.This breccia, in turn, is cut by apophyses within the majorLos Bronces-RioBlanco porphyry of a fourthbreccia type, thusestablishing a clearrel- copper system. At Los Bronces, seven different in- ative age relationship. terfingeringcoalescing breccias have been recognized An early stageof barrentourmalinization occurred (Cuadra,1980; Warnaars,1980, 1982). At the pres- west, southwest, and east of the Los Bronces breccia ent erosionsurface, they form one elongated kidney- complex.This phase is apparentin clastsin the Central shapedbody that extendsabout 2 km north-southand brecciaand in clastsof the deeper parts of the In- reachesa maximumwidth of 750 m asshown in Figure fiernillo andWestern breccias. The phaseconsists of 4. They are identifiedchronologically as the Ghost, barren quartz-tourmalineveins and nestsor aggre- Central,Western, Infiernillo, Anhydrite, Fine Gray, gatesof tourmalinein quartz monzoniteand was im- andDonoso breccias. The breccia complex is not fully portant in sealingand partly obliteratingthe major exploredat depth but is knownto extendbelow the northeast-southwestEscondida fault zone.The early 3,050-mlevel in its centraleastern part, about1,100 quartz-tourmalinephase developed locally into sul- m belowthe highestbreccia outcrop which is in the fide-poorbreccias where the intrusivefragments were southeasternpart of the complex near Infiernillo peak. rotatedand cemented by quartzand tourmaline. This Geologicprojections of drill-holeinformation on var- type ofbrecciais namedCasino breccia as an informal ious crosssections suggest that somebreccias may field term. It has an erratic distribution and was in- have severalroot zones,much like the rootsof a molar tersected in a few drill holes southwest of the main tooth. brecciacomplex. The outline of the brecciacomplex at about the The LosBronces breccia types are briefly described 3,250-m elevation is smaller than that at the surface. in chronologicalorder in the paragraphsthat follow.
Excursion métallogénique - Chili 2012 Références page 148 1550 WARNAARS, HOLMGREN D., AND BARASS1F.
J••J v v v v v v ¾ v v v v v v v v v v v v v v v v v v v v v EXPLANATION "•RHYOLITE - DACITE NECK v•v v v v v
v v • DaNasoBRECCIA v v •'-• ANHYDRITEBRECCIA v v v v v v v v v '• • FINEGREY BRECClA o v v v v •'•'• INFIE R NILLO BRECCIA •---• WESTERNBRECCIA &•'•'• CENTRALBRECClA • GHOSTBRECCIA '1LaMA BRECCIA • IATITE D•KE INTRUSIVE STOCK (Quartz - ß•' DaNasO ' l [• monzonltetoquartz-dlorlt4 BRECCIA.. ,-• ANDESITEFARELLONESOF FORMATIONTHE MIOCENE 0 0'5 Km , i ! SCALE
ENTRAL ø
WESTERN< o o o o o& o BRECCIA v v/v v v v v v v[v v v v v v v v v. v v • v v v v v v v v v BRECCIA •' v v v ¾ v v v v v •¾ V V V V V V V v • • v v v v v v .... INklERNILLO Inflernillo v v v v • Mountain v v v v S•ECCIA 4,190 ' v v v v v v v v v (SRECCIA,;iI v v v v v v v v v v v v
v v v v v v v
v v v v v v ½/ v v GHOST BRECCIA v v v v v v v v v
of $upergene
FIG. 4. Outlineof the LosBronces breccias simplified from outcropand drill holeinformation.
Table 3 lists some of the characteristics of the indi- quartzmonzonite is difficultto recognize;hence the vidualbreccias and the photographsin Figure 5 il- name.Often, the primary copper and/or molybdenum lustratevarious typical breccia types. gradein the Ghostbreccia is higherthan in the ad- jacentmineralized quartz monzonite rocks. Ghost breccia Ghostbreccias do not formone separate body and Clasts of the Ghost breccia consist of fine to coarse arenot well exposedat the presentsurface. They are fragmentsof quartzmonzonite. These clasts often foundas brecciaremnants peripheral to the whole containdisseminated and sometimesstockwork, por- LosBronces complex, if onedisregards the younger phyry-typemineralization. The Ghostbreccia matrix Donosobreccia (Figs. 4, 6, and7). Remnantsare also is characterizedby rock flour and ground quartz common in the area between the Rio Blanco mine monzonitewith smalldisseminated crystals of quartz, andLos Bronces within the porphyrycopper system. tourmaline,specularitc, and sulfides. Both clasts and The Loma breccia located between Los Bronces and matrixfrequently show homogeneous, moderate to RioBlanco (Fig. 4) outsidethe mainbreccia complex strongquartz-sericite alterations that obscurethe has some of the characteristicsof the Ghost breccia. distinction between them. For this reason,the contact Ghost breccias are considered to be the earliest between the breccia and nonbrecciated mineralized brecciasbecause they are observed as clasts in nearly
Excursion métallogénique - Chili 2012 Références page 149 LOS BRONCES-RIO BLANCO: Cu & TOURMALINE BRECCIAS 1551
o
Excursion métallogénique - Chili 2012 Références page 150 ams. ams. i""1' '1 I , , , , ,.....
"' AD ' '
0 I 2 $ 4 5 6 7 ß 9 I0 CM
E I I I I ' ' ' 0 I 2 $ 4 5 6 cM F I I I I I I I
cpy
Go i 2• 45 6 cms.
FIG. 5. Typical brecciatypes of Los Bronces.A. Ghostbreccia (DDH FF6). B. Central breccia (DDH FF.7). C. Westernbreccia (DDH 4.5). D. Infiernillo breccia(tunnel R5). E. Anhydritebreccia (DDH II9.5). F. Fine Gray breccia(DDH 5). G. Donosobreccia (hand specimen from open pit). For abbreviations, see Table 3, DDH = diamond drill hole.
1552 Excursion métallogénique - Chili 2012 Références page 151 LOS BRONCES-RIO BLANCO: Cu & TOURMALINE BRECCIAS 1553
#- I01,000
OUTLINE LOS BNECCIA COMPLEX EXPLANATION
• SURFACEOUTLINE OUTLINES INTERPRETED FROM DRI•.L HOLE INTERCEPTS • $$S5LEVEL ß.• 3460LEVEL SURFACE ::•:]•1:•$$•5 LEVEL
5 ßN-I00,000 • i'•:/•:.:::•..•3,61 •• •3160 LEVEL
N-99,000 ..
0 0.SKm SUBSURFACE
FIG. 6. Interpreted outline of the Ghost brecciasat different elevations.
all otherbreccias and form large fragments or breccia creasesthe overall grade in severaldrill-hole inter- remnantsthat were brokenup by the emplacement cepts. of youngerbreccia bodies. The Central breccia,developed after the Ghost Central breccia breccia,became clasts in, or wascut by, all the later brecciatypes. A projectionof the Centralbreccia at Clastsof the Centralbreccia are mostlyquartz differentlevels outlines an oval-shaped inverted cone monzonite, except on its south side, where some an- (Fig. 8). The northern part of this cone has been desiteand latite clastsoccur. As a rule, the clastsare erodedor minedaway. Remnants of thesouthern part hydrothermallyaltered and show the effect of quartz- are stillexposed because of the steeptopography of sericitic,silicic, and argillic alterations. The clastsare InfiernilloMountain. The deeperpart of the Central mostlysubangular and morerounded than in any breccialies betweendrilling gridlines B and E and otherbreccia type. The Centralbreccia is character- lines5 and8. However,its extent at depthis unknown izedby a highvolume of blackto darkgray matrix, below the 3,310-m level. composedof fine-grainedtourmaline crystals with minor amountsof quartz, specularite,sericite, sul- Western breccia fides,and rarely, anhydrite. Sulfides are mostly dis- Next, the Western and Infiernillo breccias were seminatedand rarely form the coarse aggregates that formedalong a northwest-southeastaxis. Their em- are so common in the matrix of the Donoso breccia. placementshattered a largepart of the southernand Pyrite is more abundantthan chalcopyrite.Chalco- westernpart of the Central breccia. pyrite in veinlets,and asdisseminations in clasts,in- Mostof the clastsare quartz monzonite fragments
Excursion métallogénique - Chili 2012 Références page 152 1554 WARNAARS,HOLMGREN D., AND BARASSIF.
I 2 $ 4 5 6 7 8 9 10 II 12 I$ 14 15
N-101,000 EXPLA NATION K • • • SURFICEOUTtilE
i
H G • • 34• LEVEL* DRILLHOLE F
E ,
D
C
B
A
BB
CC
DO
EE
G(•
II
dJ
__N99,000 •;:••: 0I • 0-5Kin • • SCALE
OC
I 2 3 4 5 6 7 8 9 J0 II 12 13 14 15
FIG. 7. Interpreted outline of the Donoso,Fine Gray, andAnhydrite brecciasat different elevations.
that display predominantlychloritic alteration and The matrixvolume varies between 2 and 15 percent replacementof the maficminerals by chlorite, spec- and is generallyless than in any of the other breccias. ularitc, and/or tourmaline.Other alterationminerals The matrix consistspredominantly of chlorite and in the clastsare rutile, leucoxene,calcite, and spo- quartzwith lesseramounts of specularitc,tourmaline, radically,epidote. Magnetite is stableand plagioclase epidote, pyrite, chalcopyrite, and magnetite. Open is weakly altered to sericiteand quartz. PrimaryK- vugsare ubiquitous. feldsparis rarely altered.The clastsare mostly angular The Infiernillo brecciais almostentirely confined and rarely mineralized.The Westernbreccia is char- to the andesitesin the southernpart of the breccia acterizedby a green matrix with abundantchlorite complex.The contactwith the nonbrecciatedandes- and rock flour. Tourmaline and specularitcare less ites to the east and south is very sharp and nearly commonthan in other breccias.Open vugsare rare. vertical. Contacts with other breccias to the south, Pyrite andchalcopyrite are mostlyfinely disseminated west, and north are not well exposedon the surface. in the matrix. Underlyingthe andesitcbreccias in the Infiernilloarea are quartz monzonitebreccias similar to the Western Infiernillo breccia breccia(Fig. 9). The similaritiessuggest a geneticre- Angularto subangularandesitc fragments with mi- lationship between the two types. Differences in nor amounts of subangularquartz monzonite and amountand type of matrixmight be explainedby the quartz latite porphyriesform the clastsin the Infier- different behavior and competenceof the various nillo breccia. Mafic minerals in these clasts are clastsduring breccia formation. Lesser amountsof stronglychloritized, and in certainareas, many of the tourmalineand specularitcin the matrix of the West- clastsare silicifiedand have argillizedfeldspars. ern brecciamay reflect mineral zoning.
Excursion métallogénique - Chili 2012 Références page 153 LOS BRONCES-R10BLANCO: Cu & TOURMALINEBRECCIAS 1555
i i i i i i
.EXPLANATION N-IOta,000 • SURFACEOUTLINE -- __ OUTLINESINTERPRETED -- FROM DRILL HOLE
-- INTERCEPTS
-- _ •3 ,3460LEVEL -- • $$10LEVEL SURFACE OUTLINE -- , .:• :[ LOSBRONCE S -- : r: BRECCiACOMPLEX
- N-ioo,ooo\ --
-- __ - -
Z N-99•000
-- 8 ooo -- 0 0-SKin
I 2 $ 4 5 6 7 8 9 I0 II 12 13 14 15 i I I I I I I I I I I I I I
FIG. 8. Interpreted outline of the Central brecciaat differentelevations.
Anhydrite breccia ent.Sulfides in the fragmentsare present in veinlets The Anhydritebreccia is a smallcolumnar breccia andas fine disseminations ofpyrite and chalcopyrite. bodywithin the Infiernillo breccia (Figs. 7 and9). The Fine Graybreccia is namedafter the natureof Fragmentsof Infiernillo and Central breccias are itsmatrix which is typicallygray colored and micro- commonin the Anhydritebreccia and are evidence crystallineconsisting of rock flour, fine-grained tour- of relativelyyoung empiacement. maline,sericite, and quartz. The matrix comprises 10 Weaklyaltered chloritized andesitc comprises most to 40 percent of the breccia volume. In someareas of the clasts,at leastin theupper part. The matrix flowstructures are apparent. Fine-grained chlorite, volumevaries between 5 and60 percentand is char- tourmaline,specularitc, and sulfides are present in acterizedby abundantcoarse-grained anhydrite. minor amounts. Openvugs are normally absent. There is no rock flour TheFine Gray breccia has an irregular outline with in thematrix. The most common opaque minerals are abundantapophyses projecting into the adjacent specularitc,pyrite, chalcopyrite, and molybdenite. breccias. It cuts across and contains fragments ofthe Theseminerals are usually more euhedral and coarse Central, Infiernillo, and Western breccias. grainedthan in the adjacentInfiernillo breccia. In surfaceoutcrop, the anhydrite ishydrated togypsum. Donoso breccia Fine Gray breccia TheDonoso breccia is largely a monolithic breccia. The majorityof the clastsare quartz monzonite and Clastsin thisbreccia are mostly quartz monzonite some are quartz diorite, syenite, and rarely andesitc. and,sporadically, quartz latite porphyries and an- Thefragments are mostly angular to subangular.Not desites.Most clasts show strong quartz-sericite alter- muchabrasive movement took place during breccia- ation.Silicification and chloritization arelocally pres- tion, as rounded clastsare rare. The matrix volume
Excursion métallogénique - Chili 2012 Références page 154 1556 WARNAARS,HOLMGREN D., AND BARASSIF.
,•f///•l'•'r,,,,•., 'e• ;•;/ -- * .• L•I[•.•ILLO
•/• 'Y' ' '•+ + +•• •" '•C•i•' •/•'•• • -• CEntRALeRgccm •//•• + + + • OUARTZ •//•+ + I • WONZONITE
•/•X• + • GHOSTBRECCiA
•-- ZONE • ....'EC•DARY ENRICHMENT•HOLE
FIG. 9. Geologicinterpretation of east-westcross section (line II) through the southpart of the breccia complex.
rangesbetween 5 and 25 percent of the rock mass Donosobreccia and of the northernand western part andconsists of blacktourmaline, quartz, pyrite, chal- of the Westernbreccia are weaklyto moderatelyhy- copyrite,specularitc, and rarely, anhydriteand bor- drothermallyaltered to a propyliticassemblage. They nitc. Sericite, chlorite, and rock flour are typically containchloritized, epidotized,specularitized, and/ sparse.Open vugsare common.The primary copper or tourmalinizedhornblende, primary biotite, and/or content is higher than in any of the other breccias deutericchlorite. Magnetite is mostlystable. Plagio- except, perhaps,for some parts of the Infiernillo claseis slightlysericitized and primary K-feldspar is breccia. largely unaltered. The Donosobreccia is the youngestand northern- The brecciasthat were formedcloser to the (west) mostbreccia of the Los Broncescomplex. It is ellip- centralpart of the porphyrysystem contain fragments soidal in plan and was developed on a northwest- that exhibit strongquartz-sericite alteration. These southeastaxis, with inward-dippingnorthern, west- are the Fine Gray, Ghost,and most of the Central ern, and southern contacts. The eastern contact is breccias.The clastsin the southernand westernpart nearlyvertical (Fig. 7), suggestinga possible westward of the Donoso breccia also contain more sericite and tiltingafter breeeia formation. The rootsof the breecia quartz from the alterationof feldspars,chlorite, and/ pipe lie in the area between east-westdrilling grid or biotite. The clasts of the Infiernillo breccia are lines H and E, and north-south lines 6.5 and 8.5. The mostlystrongly chloritized and/or partly epidotized. breeeiais open at depth. The contactwith the sur- Toward the northernpart of the Infiernillobreccia roundingquartz monzoniteis remarkablysharp. In the clastsare more silicifiedand the feldsparsmore places, the contact with the adjacent Central and sericitizedand/or argillized. Western breeeia also is sharp, but elsewhere the SecondaryK-feldspar and secondarybiotite have breeeia fragmentsare thoroughly mixed and the rarely been observedin intrusiverocks within or pe- boundarycan only be approximated. ripheral to the Los Broncesbreccias. However, sec- ondary biotite is commonly observed in andesites Hydrothermal Alteration closeto the contactwith the quartz monzoniteintru- sion in drill core from holes east of the breccia com- The clastsof the variousbreccias generally were plex. In the Rio Blancomine, secondarybiotite and not alteredduring the breccia-formingstages, except secondaryK-feldspar are common. for somenarrow alterationrims or more pervasively A uniquealteration feature of the LosBronces-Rio in small clasts.Mineral depositiontook place after Blancodeposit is the replacementof maficminerals brecciationand probably formed under relatively low- as well as secondarychlorite by specularitcand/or pressureconditions and rapid cooling of hydrothermal tourmaline within and inside the propylitic zone. fluidsdid not allow adequatetime to alter clasts.The Many aggregatesof specularitc,with or withouttour- degree and type of hydrothermalalteration of the maline, form pseudomorphsafter hornblende,sec- clastsdepended largely on the locationof the breccia ondary chlorite, and/or magnetite. These pseudo- in relationto the earlier developedporphyry system. morphsare preservedwithin the quartz-sericiteal- The clastsof the northern and easternpart of the teration zone.
Excursion métallogénique - Chili 2012 Références page 155 LOS BRONCES-RIO BLANCO: Cu & TOURMALINE BRECCIAS 1557
Mineralization velopedin a propyliticalteration zone, such as in the southand southwest of the brecciacomplex, the Ghost The porphyrycopper mineralization that preceded brecciacontains low copper and molybdenumvalues the brecciaformation is centeredin the vicinity of in the matrix.Ghost breccias that developedbetween the Rio Blancomine. Unusually high tenor of primary drilling grid linesB and C (Figs.4 and 6) are more coppermineralization in the formofchalcopyrite and central to the prebrecciaporphyry copper deposit minor bornite occursin a roof pendantof andesires and often containhigh-grade primary molybdenum and in the underlyingquartz monzonite.The esti- and copper mineralization. matedgrade of the firsteight years of productionafter The Central brecciais to a large extent disrupted 1970 at Rio Blancowas 1.9 percent Cu. The high- by younger breccias, preventing reconstructionof gradeore obtainedin the first years(2.4% Cu) was possiblemineral trends. Also, the Central brecciais partlydue to the miningof a segmentof hydrothermal more complexthan mostbecause it consistsof mul- brecciain the southernpart of the Rio Blancomine. tiple phases.In earlier phases,the mineralizationoc- The northernpart of the high-gradeporphyry cop- cursas coarse-grained aggregates in the matrix. The per center hasbeen removedby a youngerdacite- last phasemay representa fiuidizationstage with rhyolite volcanicneck or diatremenorth of the Rio abundant(up to 80%) matrixmaterial showing flow Blancocave. Strong primary copper and molybdenum structuresand important amountsof disseminated mineralizationwas intersectedby diamonddrilling pyrite and sparseamounts of chalcopyrite. below the western edge of the volcanicneck and The mineralization in the Western breccia consists probablyoccurs southeast of the neckbelow the valley of finely disseminatedpyrite and chalcopyritein the bottom of the Blanco River. matrix.The primary coppercontent is generallylow, The pyrite/chalcopyriteratio increaseswestward rangingfrom 0.3 to 0.6 percent Cu. The brecciais from the Rio Blanco mine toward the Los Bronces consideredwaste except where upgradedby second- breccias. Disseminated mineralization decreases and ary enrichment.Molybdenite occurs irregularly along mineralization in veinlets and on fractures increases. the eastern contact with the Central breccia. This porphyry style of mineralizationis apparentin The gradeof the primarycopper mineralization of the quartz monzoniteclasts in the centraland south- the Fine Gray brecciamust have been low but cannot ern part of the brecciacomplex. Mixing ofclasts within be ascertainedbecause of the strongoverprint of sec- the brecciasprecludes adequately reconstructing a ondary copper enrichment.Pyrite and chalcopyrite zoning pattern of mineralization of the prebreccia with chalcocitecoatings are finely disseminatedwith porphyry system.The copper gradesof clastscon- a low totalvolume percentage mostly varying between tribute to the overall copper contentof mostof the one and three. Ghost,Central, and Fine Gray breccias,and in certain Sulfideminerals in the Infiernillo brecciaare gen- partsof the Infiernillo, Western, and Donosobreccias, erally present as irregular aggregatesin the matrix. in particularin areasof supergeneenrichment. Disseminated mineralization and mineralized veinlets In the intrusiverocks southwest and peripheral to are not significantand most clasts are unmineralized. the brecciacomplex, pyrite is more abundantthan The Infiernillo brecciatypically containsstrong sec- chalcopyriteand both occurmore asstockworks than ondary copper enrichment,consisting of chalcocite as disseminations.The primary gradeshere vary be- coatingson chalcopyriteand pyrite. Covellite is less tween 0.2 and 0.5 percent copper, but supergene frequentlyobserved. Native copper and cuprite are copperenrichment has doubled or tripled the grade. commonthroughout the breccia. Northwest of the Escondida fault, mineralization The amountof chalcopyritepresent below the en- outsidethe brecciacomplex is sparseand is restricted richment zone and the amount of chalcopyritestill mostly to fractures. Mineralization increases, how- visiblewithin the enrichmentblanket suggest a higher ever, toward the volcanicneck at La Copa. gradeof primarymineralization along the easternand Supergenecopper enrichment is economicallysig- southern contacts of the Infiernillo breccia. This nificantin the southerntwo-thirds of the deposit, highergrade rind is possiblydue to inherentbreccia southof the SanFrancisco River (Fig. 4). The copper permeabilitiesbefore mineralization and may indicate contenthas been upgraded two or moretimes its pri- areasof more rapid cooling.Parts of this brecciaare maryvalue in the Ghost,Central, Western, Infiernillo, of economicimportance solely because of goodpri- andFine Gray breccias.Therefore, primary mineral mary grades,which rangebetween 0.25 and 0.9 per- distributionand pyrite chalcopyrite ratios are difficult centCu andwith pyrite to chalcopyriteratios varying to reconstruct in these breccias. from 1:1 to 1:3. Much of the primary copper and molybdenum Molybdenite mineralization is significantwith within the matrix of the Ghost breccia is related to grades over 0.1 percent Mo, particularly near the the areasof the porphyrycopper deposit where the easternand southern margins of the breccia.It iscon- breccia was formed. For instance,where it was de- fined to the matrix and was emplacedafter the de-
Excursion métallogénique - Chili 2012 Références page 156 1558 WARNAARS,HOLMGREN D., AND BARASS1F.
positionof pyrite andchalcopyrite. Ferrimolybdite is Mineral Distribution in the Donoso Breccia ubiquitousin weathered surface outcropsand is abundant in the leached zones. Mineral zoningis difficultto recognizebecause of The Anhydritebreccia has statistically the highest the irregularnature and the coarsegrain size of the total molybdenumcontent (0.051%) comparedwith sulfideminerals in the matrix.Often, the analysisof the otherbreccias (Table 3). Its averagecopper value a split core is not representative,because one side isthe lowest,0.47 percentCu, becauseno supergene mayconsist of matrixwith massivesulfide aggregates enrichmenthas taken place. The degree of copper andthe other side,a largebarren intrusive clast. Min- enrichmentis a functionof permeabilityand porosity eral distributionis important, particularly in grade and both are very low becauseanhydrite sealedall predictionat the variousproduction benches and in the open spaces.This brecciais thusneither leached anticipatingcopper recovery problems (Holmgren nor enrichedbut is surroundedby a circularleached and Marti, 1984). zone over 300m deep (Reyes, 1980; Warnaars, A studyof the distributionof chalcopyrite,pyrite, 1982). The surroundingInfiernillo breccia has a much andspecularite was made of variousproduction levels higher permeability. It is stronglyleached and en- on the north sideof the Donosobreccia by collecting richedas deep as500 m below the surface(Fig. 9). dust samplesfrom bench productiondrill holes on The high primary copper mineralizationin the a 25 X 50-m grid. Representative samples were Donosobreccia attracted the first prospectorsto the mountedinto polishedbriquettes for point counting. area in 1864. This breccia has remained the center The resultsof productionlevel 3670 are illustrated of miningactivity ever since.Mineralization is mainly in Figure 10 and showthat chalcopyriteand pyrite confined to the breccia matrix and occurs as coarse are distributedin an irregularshell-like pattern within aggregatesor irregularpatches of pyrite,chalcopyrite, which one of the two predominates.Note that the andoccasionally bornite. Minor sulfidedisseminations limits of the study area do not necessarilycoincide and stockwork mineralization are found in the clasts with the breccia outline. on the southwestern side of the breccia. The chro- A copper distributionof all availableassay infor- nologicalorder of mineralizationin the matrix is mationof the 3,460-m level (the Pommerantzor main quartz, black tourmaline(dravite), pyrite, chalcopy- haulagelevel) suggestsan irregularellipsoidal distri- rite, and finally specularite.The order of crystalli- bution(Fig. 11). The assaydata are from old hori- zationsuggests a decreasing sulfidation state with time zontaldrill holes,underground channel sampling, and and/or an increasein oxygenfugacity. more recent vertical drill holes. The several semiel- The currently known extent of secondarycopper lipsoidal shellsof alternatinghigh and low copper enrichment is small in the Donoso breccia. One reason gradessuggest multiple phases of mineralization.The isthat between 100 and200 m of the originalDonoso copper distribution in each shell is not uniform, ridgeis assumed to havebeen mined away since 1864. showingstrong local variations, depending on the size Nevertheless,supergene enrichment is, andprobably of clastsand on the amountof chalcopyritein the was,less significant than in other brecciasbecause of matrix. The order of crystallizationin the matrix is the coarse nature of the sulfides in Donoso. The recognized as follows: quartz, tourmaline, pyrite, coarserthe grain size, the smalleris the surfacearea chalcopyrite,and specularite.Mineral zoning and which is available for replacement.A thin film of parageneticrelationships in the matrixtypically are chalcocitecoating chalcopyrite or pyrite may cause not fully developed.Permeability is an importantfac- significantenrichment if the primarysulfide particles tor for mineral deposition.Some open spacesmay are very small. havebeen rapidlyfilled and sealedoff with minerals Late mineral veins,formed after the emplacement early in the sequence,preventing access of mineral- of the variousbreccias, contain chalcopyrite, molyb- izing fluidslate in the sequence.On the other hand denite, luzonite, enargite, tennanite, valleriite, open spacesin certain areas might have become sphalerite,galena, arsenopyrite, barite, calcite,tour- availablefor mineraldeposition only at the endof the maline, alunite, and/or pyrophyllitein a quartz-seri- mineralizingsequence. Since the formationof the cite gangue.These veinsare not of economicsignif- Donosobreccia was probably a multi-phaseevent, icance.They are typicallyfound along the periphery this mineral sequence,with somevariations, was re- of the breccia complex,both inside and outsidethe peatedseveral times. These circumstancesmight ex- actual contact. plain why pyrite, total sulfide,and chalcopyriteand Pebble dikes are probablythe last signsof miner- thus the copperdistribution are erratic on a small alizationrepresenting late-stage or postmineralvent- scaleand also on a largerscale of severalchannel sam- ing. The pebblesare mostlycountry rock andlocally ples or drill hole assayintervals. brecciapebbles. In somecases, they containpyrite The variousshells in the northernpart of the Don- in the rock flour matrixbut rarely coppersulfides. osobreccia are interpreted to be nearly vertical or
Excursion métallogénique - Chili 2012 Références page 157 LOS BRONCES-RIO BLANCO: Cu & TOURMALINE BRECCIAS 1559
FIG.10. Distributionofchalcopyrite andpyrite of operating bench level 3,670 rn in the open pit of the Donoso breccia.
slightlyinclined toward the southeastand perpendic- variousstages of pressurerelease with concomitant ularto the N 30ø W direction.This configuration co- precipitationwhere boiling occurs or differentstages incideswith the generalellipsoidal shape of the entire of decreasein temperature. Donosobreccia with steeply dipping, inward-dipping Interestingly,copper plots of variouslevels above outerlimits (Fig. 7). The shell-likeor onion-ringpat- the Pommerantzlevel (andbelow the currentpro- ternis also apparent in crosssections H-H' andJ (Figs. ductionlevels) show that highergrade shells are not 12 and 13); the crudelyconcave configuration indi- easilyprojectable from one level to the next. Instead, catesthat the shellsare steeplydipping inward. The they are slightlyoffset to the west or to the east,ex- multipleshell-like copper distribution may represent hibitinga tendencyto spiralvertically.
Excursion métallogénique - Chili 2012 Références page 158 1560 WARNAARS,HOLMGREN D., AND BARASS1F.
Summaryand Discussion 2 3 The Los Bronces-RioBlanco deposit was formed on the east side of the San Francisco batholith. This intrusionis stronglyperaluminous and has a calc-al- / -•- kaline compositionwith a alkali-calcicaffinity. The batholithtook a minimumof 11.5 m.y. to form, from 20.1 m.y. (earlyMiocene) to 8.6 m.y. (late Miocene). o ioo The hornblendeage of 20.1 m.y. is the oldestage i I determined so far from a Tertiary intrusion in this METERS part of the Andesof central Chile. A number of in- trusionsalong the central Andean mountainrange were emplacedduring a period of Tertiary magmatic activitythat startedabout 25 m.y. agowith extrusions of large quantitiesof andesitesof the FarellonesFor- D mationand/or part of the AbanicoFormation. This igneousactivity followed a period of magmatic quiescencebetween 62 and 25 m.y. ago as docu- mentedby Drake et al. (1982). The hiatusmay have been a consequenceof flatteningof the subduction plate similarto the model of Jordanet al. (1983) for the present-daysituation of the sectionof the Andean mountainrange between 28 ø and 33 ø15' S. The re- newedmagmatic activity that started25 m.y. agomay
BB have been causedby a steepeningof the subducted plate in this part of the Andes. CC LEGEND: The last magmaticactivity in the area under study ß ) 1.5O%cu gave rise to dacitic (flow) brecciasand dacite por- Ill 1-01 - 1.50 phyriesfrom the volcanicneck or diatremeofLa Copa [] < 0.40 with agesranging from 4.9 to 3.9 m.y. An area of similar acid volcanism lies about 15 km south of Los Broncesat Colorado-LaParva. No youngerages have FIG. 11. Copper distributionat the 3,460-m level (Pommer- been determined in igneousrocks north of 33015' S antz). in this part of centralChile. The youngestage of 3.9
3600
•, ! ::::::' )+ EXPLANATION 3500
:"• Cp 29- 43•/oW c 3400 o
_ ::::: • OM
SCALE 33oo + 0 50 I•Om
i
32o• q.
FIG. 12. Chalcopyritedistribution in an east-westsection (H-H') throughthe Donosobreccia (looking north).
Excursion métallogénique - Chili 2012 Références page 159 LOS BRONCES-RIO BLANCO: Cu & TOURMALINE BRECCIAS 1561
andeach contributed in a differentway to the copper and molybdenummineralization. Los Broncesgives evidence of both explosiveor
SCALE collapsephenomena. The initial breccia-formingpro- cessat LosBronces is interpretedto havebeen a con- I00 200 300m sequenceof hydrothermalpressure buildup that re- suitedin an explosive,disruptive, upward movement. This conceptis consistentwith the presenceof intru- sive clasts and boulders in the andesitc breccia on ...4.õ 5-5 f• 7ø5 8*5 Infiernillo mountainat an elevationof 4,100 m, ap- ß I.' . .fl ' proximately200 m above the contactbetween the andesitesand the intrusivebatholith adjacent to the breccia. Collapsefeatures are alsoapparent and probably • '.• .'. formedafter releaseof the pressurethat initiatedthe brecciation.This would explain the presenceof an- desitc clasts in the Donoso and Central breccias ob- ßa,4• • +• + + EXPLANATIONov..u.o,. servedin the undergroundworkings of the 3,460-m ORIELHOLE • RHYOLITE level at least 300 m below the andesitc-intrusion con- tact outside the breccia. OTALCu• • QUARTZMONZONITE Also, in the south-centralpart of the Infiernillo •,•oo...• • oo•oso•a[cclAbreccia, the contact between the andesitc breccia and • H16H GRA D• O 2 4 •HELLS the quartzmonzonite breccia is gradationaland is 250 m below the regionallyinferred contactbetween the andesitesand intrusions.This changein brecciatype FIC. 13. Geologicinterpretation and copperdistribution of occurs in the center of the Infiernillo breccia at a lower an e•t-west crosssection (line J) through the Donosobreccia. elevationthan in other parts of the Infiernillo and Western breccias,probably because the pressurere- leasewas greatestin the center of the breccia. Dif- ferentialpressure release in the center may alsoex- m.y. may markthe beginningof igneousquiescence plain why the younger, small, columnarAnhydrite that hasprevailed to the presentwhich Jordan et al. brecciapipe was emplacedin this low-pressurearea (1983) postulatedas being due to flatteningof the beforethe Infiernillo-Westernbreccia was completely subductingNazca plate. solidified. TheLos Bronces-Rio Blanco porphyry copper min- Supergeneenrichment and leachingare both sig- eralizationand alterationincluding the brecciafor- nificant in the southern two-thirds of the Los Bronces mationwere generatedover a period of at least 2.5 brecciacomplex. The degreeand extent of both pro- m.y., between7.4 and 4.9 m.y. ago.This makesthe cessesis more a functionof brecciaand fractureper- LosBronces-Rio Blanco deposit one of the youngest meabilitythan of initial primarymineral distribution. in Chile andis comparablein ageto the E1Teniente The depthof the supergenecopper deposition is not porphyrycopper deposit (5 m.y., Clark et al., 1983). known in certain areasof the southernpart of the Hydrothermalbreccias are commonlyassociated brecciacomplex but in placesis more than 500 m with porphyrycopper deposits throughout the world. thick. The supergeneprocesses are related to the The brecciasof the LosBronces-Rio Blanco deposit, present topographyand the present ground-water however,are unique because of theirabundance (both regime, with large seasonalwater table fluctuations to the southof Rio Blancoas well asat LosBronces), in excess of a few hundred meters. The enrichment theirlarge combined size, complexity, and high tour- appearsto be postglacialand geologicallyrecent, malinecontents. The brecciasdominate the system similarto the supergeneenrichment at E1 Teniente from both geometricand economic aspects. (F. Camus,pers. commun.). The Los Bronces breccias were formed after the A postmineralvolcanic neck of La Copawhich re- mainphase of the porphyrycopper system. Intrusive movedpart of a porphyrycopper system is not unique dlastsin the brecciascommonly contain stockwork to the Los Bronces-RioBlanco orebody. Other ex- anddisseminated copper mineralization. The time of amplesare E1 Teniente (Chile), Toquepala (Peru), and emplacementand coolingof the brecciaswas rela- Lepanto (Philippines).The BradenFormation at E1 tively short in comparisonto the time spanfor the Teniente (Camus,1975) is largely a postmineralvol- formationof the entireLos Bronces-Rio Blanco por- canic event that removeda substantialpart of the phyrysystem. Each breccia has its owncharacteristics center of the porphyrycopper orebody and associated
Excursion métallogénique - Chili 2012 Références page 160 1562 WARNAARS,HOLMGREN D., AND BARASSIF.
hydrothermal tourmaline breccias.The tourmaline veins containing specularite, enargite, sulfosalts, brecciasare still preservedas large fragments within chalcopyrite,sphalerite, and galenain a gangueof the diatremeand as a rim alongthe edgeof the Braden quartz,barite, and siderite. Formation.Toquepala also has a barren intrusiveda- 5. Hydrothermalactivity was cut offand disrupted cite thatremoved a substantialpart of the orebodyat by the postmineraldacite porphyries of the La Copa the east side of the current mining operations. volcanic neck or diatreme. MacKibbon,who worked in the Lepanto area, men- tioned (pets. commun.)the existenceof a barren Acknowledgments postmineraldiatreme that containsfragments with We highly appreciatethe encouragementof F. J. porphyry-stylemineralization southeast of Lepanto Sawkins and R. H. Sillitoe to contribute to this breccia and westof the Tirad porphyrycopper deposit. volume.This would have not been possiblewithout As a summary,Figure 14 presentsa schematicdia- the invitation of D. L. Giles, F. H. Bonhamm, and gramof the formationof the LosBronces-Rio Blanco F. J. Sawkins,the organizersof the BrecciationCon- porphyrycopper systemand brecciaemplacements. ferencein ColoradoSprings in September1983. The verticalaxis represents the relative intensityof E. Reyes S. collectedimportant early geological hydrothermalactivity. The horizontalscale is alsoan drill hole and regionalinformation until his fatal ac- approximationbecause more age determinationsare cident in 1980. This paper resulted from valuable requiredto be reallyconfident of the timingof events. contributionsby manygeologists of CompafiiaMinera Figure 14 illustratesthe followinginferred events: Disputada:W. CuadraC., J. CabelloL., V. Irarr•tzaral LI., F. Gonzalez, R. Hein C., R. Le6n B., M. Marti 1. The hydrothermalmineralization and alteration G., R. Mufioz M., J. Urquidi B., C. Walker A., and J. tookplace during the lastphase of the coolingof the Wenke H., who were involved in variousstages of easternpart of the SanFrancisco batholith. the resource evaluation program. To them we are 2. The brecciaswere formed in a shortperiod of mostobliged. time as a result of pressurebuildup during the ret- With great pleasurewe acknowledgethe cooper- rogressionor the waningphase of the disseminated ation and permissionof Compa•ia Minera Disputada stockworkporphyry copper formation. de LosCondes S.A. andExxon Minerals Company in 3. In the porphyry copper system, pyrite was publishingthis paper; in particular J. E. Frost who formedfirst, followedby chalcopyrite,bornitc, and stimulatedthe various studies,critically read the later molybdenite.Subsequent short-lived but sub- manuscript,and gavefinal approvalfor release.Also stantialpulses of brecciationlocalized new deposition Hans Bosshardt,C. L. Dahl, and William Saegartof of quartz, tourmaline,sulfides, and specularitc. Exxonencouraged us and contributedto the under- 4. The youngestphases of mineralizationare rep- standingof the geologyand mineralizationat Los resentedby late, postbrecciamineralized faults and Bronces.We would like to thank E. Klohn H., O. Fer- nandezH., and M. Marti G. of Disputadaand D. Har-
Tourmaline rison of Exxon for their keen interest and continuous
Pyrite • support.We highlyappreciate the cooperationof M. Chalcopyrite -- -- Humphreysof Esso(Australia) and R. D. McNeil of Molybdenite -- -- EssoPapua New Guineafor allocatingtime to prepare Specularite the paper. Many early versionsof the paperwere reviewed by G. Westra,J. F. McKnight,M. J. Mackenzie,and o C. C. Brooks,whose commentariesand suggestions o enhancedthe qualityof the paperconsiderably and to whom we feel most thankful. T We are very gratefulfor the excellentdrafting of
._ .•o o -•c•, ._> M. Ubilla R., E. JorqueraC., D. Baine, and M. Keno and to W. Tep who typed numerousversions of the •]•'• • •Gradua• • • • manuscript.
•o o REFERENCES '•ø•• Abru••• Aguirre,L., 1960, Geologiade LosAndes de Chile central,Prov. Ooohngof / Brecc•a de Aconcagua:Chile Inst. Inv. Geol., Bol. 9, 70 p. bathohth ./' Pulses Albers, J.P., 1981, A lithologic-tectonicframework for the me-
20 8 7 6 5 tallogenicprovinces of California:ECON. GEOL., v. 76, p. 765- Approximate miH•on years before present 790. Barassi F., S., Gonzalez, F., and Warnaars, F., 1979, Traverse FIG. 14. Schematicrepresentation of the sequenceof hydro- mappingin structuraldomains: Santiago, Cia Minera Disputada, thermal eventsduring the formationof the Los Bronces-RioBlanco unpub. rept., p. 1-17. porphyry copper and breccias. Blondel,J. R., 1980, P6rfidode composici6ngranodioritica de la
Excursion métallogénique - Chili 2012 Références page 161 LOS BRONCES-RIO BLANCO: Cu & TOURMALINE BRECCIAS 1563
mina Rio Blanco:Unpub. thesis,Santiago, Univ. Chile, Dept. Lopez,L., andVergara, M., 1982,Geoqulmica y petrog6nesis de Geology, p. 1-88. rocasgranodior•ticas asociadas con el yacimientocuprffero Rio Camus,F., 1975, Geologyof the E1Teniente orebodywith em- Blanco-LosBronces: Rev. Geol. Chile, no. 15, p. 59-70. phasison wall-rock alteration: ECON. GEOL., v. 70, p. 1341- Minster,J. B., andJordan, T. H., 1978, Presentday plate motions: 1372. Jour. Geophys.Research, v. 83, p. 5331-5354. Cepeda, A., 1981, Geologlaecon0mica del distrito R•o Blanco- Ossand0n,G., 1974, Petrograf•ay alteraci6ndel p6rfido dac•tico Disputada:Inst. Inv. Geol. Chile, Bol. 36, p. 1-43. YacimientoEl Teniente: Unpub. thesis,Santiago, Univ. Chile, Clark, A. H., Farrar, E., Camus,F., and Quirt, G. S., 1983, K-Ar Dept. Geology, p. 1-112. age data for the E1Teniente porphyry copper deposit,central Oyarzfin,J., 1971, Contributiona l'etudegeochimique des roches Chile: ECON. GEOL., v. 78, p. 1003-1006. volcaniqueset plutoniquesdu Chile:Unpub. Ph.D. thesis,Univ. Cuadra,W., 1980, Geolog•ay petrog•nesisde la BrechaDonoso, Paris, 195 p. Mina Los Bronces:Unpub. thesis,Santiago, Univ. Chile, Dept. Peacock,M. A., 1931, Classificationof igneousrocks series; Jour. Geology, p. 1-148. Geology, v. 39, p. 54-67. Drake, R., Charrier, R., Thiele, R., Munizaga,F., Padilla,H., and Quirt, S., Clark, A. H., Ferrar, E., and Sillitoe,R. H., 1971, Po- Vergara, M., 1982, Distribuci6n y edadesK/Ar de volcanitas tassium-argonages of porphyry copper depositsin northern post-Neocomianasen la cordillera princip•d entre 32 ø y 36 ø and central Chile labs.I: ECON.GEOL., v. 67, p. 980-981. L.S. Implicacionesestratigr•tficas y tect6nicaspara meso-Cen- Reyes,F., 1980, Actualizati6nde zona de lixiviaci6n:Santiago, ozoicode Chile central:Geol. Cong. Chile, 3rd, Concepci6n, Cia Minera Disputada,unpub. rept., p. 1-9. 1982, Acta, v. 2, p. D41-D78. Shand,S. J., 1927, The eruptive rocks:New York, JohnWiley, Frutos,J., 1981, Andeantectonics as a consequenceof seafloor p. 1-188. spreading:Tectonophysics, v. 72, p. 21-32. Stambuk,V., Blond61,J., and Serrano, L., 1982, Geologla del Hart, S. R., 1964, The petrologyand isotopic-mineral age relations yacimiento R•o Blanco:Cong. Geol. Chile, 3rd, Concepci6n, of a contactzone in the Front Range,Colorado: Jour. Geology, 1982, Acta, v. 2, p. E419-E443. v. 74, p. 493-525. Streckeisen,A., 1976, To each plutonic rock its proper name: Holmgren, C., and Marti, M., 1984, Applied microscopyand Earth-Sci. Rev., v. 12, p. 1-33. metallurgicalforecasting at Los Broncesmine, Chile, in Park, Vergara,M., andDrake, R, 1978, Edadespotasio-arg0n y su im- W., Hausen, D., and Hagni, R. D., eds., Applied mineralogy plicanciaen la geologlaregion•d de Chile: Univ. Chile, Rev. 1985: New York, Am. Inst. Mining Metall. PetroleumEngineers, Comunicaciones, No. 23, p. 1-11. p. 407-417. -- 1979, Eventosmagm•tticos plutonicos en LosAndes de Chile Irarr•zaval, V., LeOn, R., Mufioz, R., and Warnaars, F. W., 1979, centred:Cong. Geol. Chile, 2nd, Arica, 1979, Acta, v. 1, p. Reconnaissancemapping in the area of Los Broncesand Andina: F19-F30. Santiago,Cia Minera Disputada, unpub. rept., p. 1-24. Warnaars, F. W., 1980, Brechas de Cobre y Turmalina en Los Jordan,T. E., Isacks,B. L, Alhnendinger,R. W., Brewer, J. A., Bronces,Chile: Cong. PorphyryCopper Mining, 50th, Santiago, Ramos,V. A., and Ando, C. J., 1983, Andean tectonicsrelated Nov. 1980, Proc., v. 3, p. 175-201. to geometryof subductedNazca plate: Geol. Soc.America Bull., -- 1982, Copper tourmaline breccias at Los Bronces, Chile: v. 94, p. 341-360. Am. Inst. Mining Metall. Petroleum Engineers Trans., v. 272, Keith, S. B., 1978, Paleosubductiongeometries inferred from p. 1902-1911. Cretaceousand Tertiary magmaticpattern in southwestern Warnaars,F. W., Smith,W. H., Bray, R. E., Lanier, G., and Shaf- North America:Geology, v. 6, p. 516-521. iqullah, M., 1978, Geochronologyof igneousintrusions and Klohn, C, 1960, Geologlade la Cordillera de Los Andesde Chile porphyry copper mineralization at Bingham, Utah: ECON. centralProvincias de Santiago,O'Higgins, Colchagua y Curic6, GEOL., v. 73, p. 1242-1249. Chile: Chile Inst. Inv. Geol., Bol. 8, 97 p. Zentilli, M., 1974, Geologicalevolution and metallogeneticre- Latorre, M. J., 1981, La formaci6n Rio Blanco en el area de la lationshipin the Andesof northern Chile between 26 ø and 29 ø Mina Andina:Unpub. thesis,Santiago, Univ. Chile, Dept. Geol- south:Unpub. Ph.D. thesis,Kingston, Ontario, Queen'sUniv., ogy, p. 1-112. p. 1-151.
APPENDIX I
Nomenclature It was essentialto agreeon terminology,nomen- ferent rock types and breccia types were on display clature,abbreviations, and a computercoding system in the loggingfacility. asearly as possible in a resourceevaluation program. A breccia at Los Bronces was defined as a fractured Thiswas particularly critical since 15 geologistswere rock with 5 vol percent or more of the rock material involvedin core loggingand mappingat the same consistingof matrix material, which is mostly tour- time in differentparts of the depositand different maline,quartz, rock flour, sulfides, or specularitc.The geologistswere loggingthe samedrill hole during rock fragmentsmust alsohave been rotated or dis- alternatingshifts using the computerlogs. The time placed.The amountof matrix material is usuallyex- factor was important,because all core drilled each pressedin volume percentage.This is considered daywas described for rockmechanics and geological moreconvenient than describing a ratio of interfrag- purposesthe sameday prior to calculatingcore re- mental(I) versusfragmental (F) rockmaterial as was eovery, core splitting,sampling, and storage.The suggestedat the 1983 BrecciaConference in Colo- programwas very intensewith coregenerated by 16 rado Springs. drill rigsoperating during the lengthof a season.For Brecciasat Los Broncesare usuallymonolithic, oc- thesereasons, reference specimens of mostof the dif- casionallybilithic, and rarely heterolithic. Breccias
Excursion métallogénique - Chili 2012 Références page 162 1564 WARNAARS,HOLMGREN D., AND BARASSIF.
are generallyeasily recognizable. Breccia contacts with the countryrock are, fortunately,mostly sharp and well defined. Contactsbetween variousbreccias, however, are often difficult to define becauseof mix- ing and interfingering.In contactzones, the relative proportionsof the variousbreccias were describedin percentageon the loggingsheet. The loggingsheet illustrated on a reducedscale in Figure A1 is designedfor descriptionsin computer codes.To overcomethe geologist'sreluctance and antipathyto computercoding, ample space is reserved for geologiccommentary, to indicatephotos taken, specimenscollected for polishedand thin sections, X-ray studies,etc. Usuallylogging intervals are se- lectedfor geologicreasons related to changesin rock type, mineralizationand alteration,or fault zones. Otherwise, intervalscoincide with drill core runs.The abbreviationsgiven in TableA1 are usedto quantify the degreeof alterationor mineraloccurrence, in- dicatingtheir approximate percentage of rockvolume. Estimatingsulfide contents and sulfideratios in brecciasis difficultbecause of the irregularnature of sulfide distribution in the breccia matrix; often the visualestimates of percentagesand ratios are verified after receivingthe assayresults of metalsand occa- sionallyof sulfur.The degreeof supergeneenrich- mentis equallyhard to estimatevisually, because in someparts of theorebody the sulfidesare fine grained and disseminated,and elsewhere they are coarse grained.In fine-grainedsulfides, it is mostlyoveres- timated and underestimated in the coarse sulfides, becausethe surfacearea availablefor replacingchal- copyriteand pyrite is greaterwhen the sulfidesare fine grainedthan when the sulfideaggregates are coarsegrained. An agreementwas reached to considersupergene copperenrichment significant if the ratio of chalco- cite/chalcopyriteexceeds 1/lO, which meansthat at least20 percentof the copperis in the formof chal- cocite. The loggingsheet of FigureA1 is largelyself-ex- planatoryand shows all variablesnecessary to char- acterize the various breccias and their alteration and mineralization.
T^BLEA1. Abbreviationsof QuantitativeEstimates of Mineral Occurrencesor Alteration Assemblages
Quantity or Approximate abundance Abbreviation vol percent
Nil NIL Rare RA Trace TR <0.1 Sparse SP 0.1-0.4 Weak WK 0.4-1.0 Moderate MOD 1.0-5.0 Strong STR 5.0-15.0 Intense INT 15.0-40.0 Total TOT >40
Excursion métallogénique - Chili 2012 Références page 163 LOS BRONCES-RIOBLANCO: Cu & TOURMALINEBRECCIAS 1565
APPENDIX II
K-Ar Analytical Data and Agesof Nine Mineral Samplesfrom Los Bronces-RioBlanco, Chile
Rad. 4øAr X10 -•2 Rad. 4øAr 100 X Rad. 4øAr Sample Average Sampletype no. wt % K mole/g Total4øAr 4øK Age (m.y.)
Quartz monzonite Lb- 1 Hb 0.377 0.543 0.077 0.1181 20.1 _+ 2.0 Quartz monzonite Lb- 1 Bi 7.515 8.565 0.421 0.0934 15.9 ___0.6 Quartz monzonite LB-2 Bi 7.608 6.148 0.358 0.0662 11.3 _+ 0.4 Hornblende diorite LB-3A Hb 0.448 0.595 0.087 0.1088 18.5 ___1.7 Hornblende diorite LB-3A Bi 6.442 0.5547 0.304 0.0706 12.0 ___0.5 Granodiorite LB-7 Hb 0.306 0.188 0.0236 0.0505 8.6 _ 0.9 Granodiorite LB-7 Bi 7.362 4.176 0.0206 0.0465 7.9 ___0.4 Dacite porphyry Lb-10 Bi 7.468 2.584 0.205 0.0284 4.8 _ 0.2 Dacite porphyry dike LB-11 Bi 7.446 2.594 0.204 0.0286 4.9 _ 0.2 Sericitizedquartz monzonite LB-12 Se 6.258 2.306 0.242 0.0302 5.2 _+ 0.3
Abbreviations:Hb = hornblende,Bi = biotite, Se = sericite,rad. = radiogenic Constantsemployed: ?,a -- 4.72 X 10-•ø y; ?,e= 0.585 X 10-•ø y; 4øK/K= 1.22 X 10-4 g.g
Excursion métallogénique - Chili 2012 Références page 164 International Journal of Coal Geology 43Ž. 2000 53±82 www.elsevier.nlrlocaterijcoalgeo
Organic petrology, chemical composition, and reflectance of pyrobitumen from the El Soldado Cu deposit, Chile
Nicholas S.F. Wilson ) Department of Earth Sciences, Dalhousie UniÕersity, Halifax, NoÕa Scotia, Canada B3H 3J5 Accepted 14 August 1999
Abstract
Solid pyrobitumenŽ. residual petroleum is found intimately associated with sulfides in several XX manto-typeŽ. strata-bound Cu deposits of central Chile. El SoldadoŽ lat 32838.8 S, long 71806.7 W. is one of the largest deposits of this type Ž.)130 Mt @ 1.5% Cu and is hosted by volcanic and subvolcanic rocks of the Cretaceous Lo Prado Fm. Detailed organic petrology of authigenic sulfide inclusions, degassing vesicles, and anisotropic pyrobitumen and graphitic carbon allowed the elucidation of a complex diagenetic history of a biodegraded reservoir in faulted volcanic and subvolcanic rocks. Due to the overmaturity of the host sequence, biomarker techniquesŽ aliphatic and aromatic. were unsuitable to investigate the composition of the pyrobitumen. However, microprobe and laser ablation inductively coupled plasma mass spectrometryŽ. LA-ICPMS analyses, Rock-Eval pyrolysis, and 13C NMR techniques were used to ascertain the composition of the pyrobitumen. Spatial variations in the maturity of the pyrobitumen were assessed using both random reflectance Ž.Rooand rotational reflectance ŽR maxand Romin.. The data are consistent with the following evolution for El Soldado: As a result of burial in the Cretaceous back-arc basin, petroleum was generated from organic-rich shales in the underlying lower Lo Prado Fm. and migrated into primary and structural porosity in the host rocks, the predominantly volcanic upper Lo Prado Fm. Low-temperature Ž.-908C biodegradation of petroleum developed an early assemblageŽ.Ž. Stage I assemblage of pyrite "sphalerite"chalcopyrite . Geopetal structures related to pressure degassing of semisolid petroleum suggest that migration occurred while the strata were horizontal, at ca. 130 Ma. Continued basinal burial led to thermal degassing, increased maturation and contraction of the petroleum into pyrobitumen. Around 20 Ma later an influx of high-temperatureŽ. ca. 250±3508C Cu-rich fluids induced by regional granitoid intrusionŽ Stage II
) Present address: Department of Geoscience, University of Nevada Las Vegas, 4505 Maryland Parkway, Las Vegas, NV 89154-4010, USA. Tel.: q1-702-895-3250; Fax: q1-702-895-4064; E-mail: [email protected]
0166-5162r00r$ - see front matter q 2000 Elsevier Science B.V. All rights reserved. PII:Excursion S0166-5162 métallogéniqueŽ. 99 00054-3 - Chili 2012 Références page 165 54 N.S.F. WilsonrInternational Journal of Coal Geology 43() 2000 53±82 assemblage. replaced preexisting pyrite and pyrobitumen. The mineralizing fluids altered the pyrobitumen physically and chemically:Ž. a increasing Ro to greater than ca. 5% in the core of the orebodies against a background level of ca. 3%;Ž. b developing anisotropyŽ nongranular, fine to coarse mosaic, and coarse flow mosaic. within the pyrobitumen and locally graphitized the pyrobitumen, developing graphitic carbon; andŽ. c incorporating Cl Ž."Cu"Fe into the pyrobitu- men during the development of anisotropy. q 2000 Elsevier Science B.V. All rights reserved.
Keywords: petroleum; pyrobitumen; manto-type Cu deposits; volcanic-hosted; degassing; reflectance
1. Introduction 1.1. Background In recent years many advances have been made in understanding the association between organic matterŽ. e.g., kerogen, coaly material, petroleum products and the formation of mineral deposits. Many conference and journal volumesŽ e.g., Dean, 1986; Parnell et al., 1993; Giordano, 1996; Hendry et al., 1997; Landais and Gize, 1997; Kettler et al., 1998. have focused on the association of organic matter with both low Ž.-2008C and high temperature Ž.)2008C deposits of many different metal assem- blages. By definition pyrobitumen represents the residual of solidified petroleum and has been described from low temperature carbonate-hosted Pb±ZnŽ e.g., Macqueen, 1986; Marikos et al., 1986.Ž , Hg e.g., Peabody, 1993. , and U depositsŽ e.g., Landais, 1993; Turner et al., 1993. . In Chile pyrobitumen has recently been documented associated within the ores of strata-boundŽ. manto-type Cu deposits Ž Zentilli et al., 1994, 1997; Wilson, 1998; Wilson and Zentilli, 1997, 1999. . One of the largest of these deposits is the El Soldado Cu XX depositŽ Fig. 1; lat 32838.8 S, long 71807.7 W. . Pyrobitumen occurs intimately associ- ated with Cu sulfides dominantly in Õolcanic rocks. The association between organic matter and Cu mineralization and its role in the formation of the deposit have been studied in detail by WilsonŽ. 1998 and Wilson and ZentilliŽ. 1999 . The authors demonstrated that El Soldado formed by two separate stages:Ž. I The first stageŽ. Stage I introduced petroleum during early diagenesis into the host rocks, where abundant pyrite formed within and around bacterially degraded petroleum.Ž. II After significant time and basinal burial, hydrothermal fluidsŽ. Stage II replaced preexisting pyrite by Cu sulfides, pyrobitumen interacted chemically with the mineralizing fluid, and was thermally altered. This paper presents petrographic, compositional, and maturation data for pyrobitumen from the El Soldado deposit and discusses the source, migration, entrapment and destruction of the precursor petroleum. The role of organic matter in the formation of the deposit has been extensively discussed by WilsonŽ. 1998 and Wilson and ZentilliŽ. 1999 and hence not covered in detail in this paper.
2. Geology of the El Soldado deposit The El Soldado Cu deposit Ž.)130 Mt @ 1.5% Cu is located about 120 km NW of Santiago, in the coastal range of central ChileŽ. Fig. 1 . The geology of the deposit is
Excursion métallogénique - Chili 2012 Références page 166 N.S.F. WilsonrInternational Journal of Coal Geology 43() 2000 53±82 55
Fig. 1. Location of the El Soldado Cu deposit in central Chile. compiled from RugeŽ. 1985 , Klohn et al. Ž. 1990 , Zentilli et al. Ž. 1997 and Wilson Ž.1998 . The deposit consists of about a dozen discordant, but strata-bound clusters of orebodies distributed within an area approximately 2 km in length by 1 km in width by 1 km in vertical extentŽ. Figs. 2 and 3 . The Cu sulfide mineralization is primary Ž.hypogene and although oxidized zones do exist near the surface, supergene enrichment was not significant. Orebodies are hosted by the Lower Cretaceous Lo Prado Fm. composed of a lower sedimentary member and an upper volcanic member. The lower member is around 1500 m in thickness under the El Soldado deposit, but varies in thickness along strike. Conformably overlying is the upper member that is around 500 m in thickness and is composed of mafic and felsic flows, and reworked volcanic material with intercalations of dark gray shales and shallow-marine calcareous sediments. The uppermost part of the Lo Prado Fm. is topped by a unit of relatively impermeable reworked volcanic rocks, which acted as cap rock for the petroleum system discussed herein. The Lo Prado Fm. conformably overlies the Upper Jurassic Horqueta Fm. around 4 km in thickness, which is composed of marine carbonates. The Lo Prado Fm. is conformably overlain by
Excursion métallogénique - Chili 2012 Références page 167 56 N.S.F. WilsonrInternational Journal of Coal Geology 43() 2000 53±82
Fig. 2. Simplified geological cross-sectionŽ. N-750 of the El Soldado deposit. between 5 and 7 km of basaltic andesites and redbeds of the Lower Cretaceous Veta Negra FmŽ. Rivano, 1995 . The deposit is strata-boundŽ. i.e., restricted to certain mappable units but discordant with respect to the strata, and mineralization is structurally controlled by faults and fault intersections orientated parallel to the N±S and N±E fault systemsŽ. Bassi, 1988 . The individual orebodies are generally zoned with an outer pyrite-rich halo, followed inwards by abundant chalcopyrite with a bornite±chalcocite core with minor hematite. Sulfides occur as both veinlets and disseminated throughout the host-rocks and the principal gangue mineral is calcite with lesser amounts of quartz, chlorite, epidote, albite, and K-feldspar. Alteration of the host volcanics is not obvious to the naked eye, and the gangue and alteration assemblages are similar. The strata have been tilted into 308 east-dipping north-striking blocks bounded by normal and reverse faultsŽ. Fig. 2 and metamorphosed to prehnite±pumpellyite grade or lowermost greenschist faciesŽ. e.g., chlorite, epidote, and albite . Upper Cretaceous granitoid intrusions occur around 20 km east and northeast of El Soldado and dikes and sills of felsic and intermediate to mafic compositions have been mapped in the deposit. Deposition of this sequenceŽ. Horqueta to Veta Negra Fms. took place in a shallow marine to nonmarine environment during Jurassic to Late Neocomian times, with the depositional environment becoming definitely continental after this periodŽ Ruiz et al.,
Excursion métallogénique - Chili 2012 Références page 168 N.S.F. WilsonrInternational Journal of Coal Geology 43() 2000 53±82 57
Fig. 3. Simplified plan section at the level "0Ž. 830 m.a.s.l and level q100 Ž 930 m.a.s.l outlining the approximate locations of the major orebodies. The major N±S and E±W fault sets are represented by the solid and dashed linesŽ. modified from figure in company report by R. Sillitoe .
1971. . This sequence has been interpreted as a shallow-marine-to-continental transition Ž.Bassi, 1988 , however, the recent recognition of marine fossilsŽ. Rivano, 1995 in the Las Chilcas Fm., which overlies the Veta Negra Fm., is evidence for other marine transgressions in the late Early Cretaceous. The deposit has been interpreted to form in an island arc environment overlying continental crust. Arc magmatism was interrupted in the Early Cretaceous by an intracontinental rifting event which led to the formation of an ensialic trough in central Chile, interpreted by AbergÊ et al.Ž. 1984 to constitute an aborted marginal basin since no ocean crust was created. Flood basalts, alternating with
Excursion métallogénique - Chili 2012 Références page 169 58 N.S.F. WilsonrInternational Journal of Coal Geology 43() 2000 53±82 dacitic flows and pyroclastics and some shallow marine and continental sediments, were deposited in an extensional basin. The genesis of the deposit has been the subject of much controversy, but epigenetic magmaticŽ. Klohn et al., 1990 and metamorphic Ž. Westra, 1988 hypotheses have the most supporters. However, the presence of organic matter, and whether it was significant in the genesis of this deposit, had not previously been considered. Recent work by WilsonŽ. 1998 and Wilson and ZentilliŽ. 1999 suggest that there is no evidence of proximal involvement of magmatic processes in the deposit. The authors suggest that El Soldado formed by the interaction of basinal fluids of metamorphic origin reacting with a biodegraded petroleum and early diagenetic pyrite.
3. Analytical methods 3.1. Samples Rhyolite, andesite, and reworked volcanic rock samples were selected from the upper Member of the Lo Prado Fm. and from siltstones and shales of the lower Member of the Lo Prado Fm. Pyrobitumen occurs in primaryŽ. i.e., amygdules and pores and secondary fracture porosityŽ. e.g., fractures, pores, filling breccia spaces in all lithologies of the upper Lo Prado Fm. and in the basal units of the overlying Veta Negra Fm. In hand specimen pyrobitumen is dark brown or black in color, has a glassy luster, and occurs as angular and globular fragments commonly associated with sparry calcite, and with Fe and Cu sulfides. 3.2. Methods 3.2.1. Organic petrology Study of the petrography of the pyrobitumen placed relative age constrains on the processes operating in the formation of the pyrobitumen. Preparation of polished thin sections of pyrobitumen and metal sulfides was difficult due to the contrasting polishing characteristics of silicates, sulfides and pyrobitumen. Careful preparation using a Durener w polisher with lead laps and an oil-based diamond paste overcame this problem and provided a scratch free surface. Pyrobitumen was studied under reflected light microscopy using a Zeiss Axioplanw microscope and camera attachment in the Fission Track Research Laboratory, Dalhousie University. 3.2.2. Reflectance s Random reflectance in oil Ž.Rooil; h 1.51´ of pyrobitumen was performed using a Zeiss Axioskopq microscope setupŽ. e.g., Mukhopadhyay, 1992 at 546 nm and without any polarizerŽ. American Society for Testing and Materials, 1991 at Global Geoenergy Research, Halifax. Anisotropy within the pyrobitumen grains was measured using rotational reflectance in oil Ž.Romaxand R omin at 546 nm with a polarizer in the light pathŽ. American Society for Testing and Materials, 1991 . 3.2.3. Electron microprobe and LA-ICPMS analyses Sulfur and metal contents of pyrobitumen were analyzed by electron microprobe on polished thin sections. Mineral analyses and X-ray maps were carried out at Dalhousie
Excursion métallogénique - Chili 2012 Références page 170 ARTICLE
L. E. Ramírez . C. Palacios . B. Townley . M. A. Parada . A. N. Sial . J. L. Fernandez-Turiel . D. Gimeno . M. Garcia-Valles . B. Lehmann The Mantos Blancos copper deposit: an upper Jurassic breccia-style hydrothermal system in the Coastal Range of Northern Chile
Abstract The Upper Jurassic Mantos Blancos copper pyrite-chalcopyrite, and followed upwards and laterally by deposit (500 Mt at 1.0% Cu), located in the Coastal Range chalcopyrite-digenite or chalcopyrite-bornite. The assem- of northern Chile, displays two superimposed hydrother- blage digenite–supergene chalcocite characterizes the cen- mal events. An older phyllic alteration probably related to tral portions of high-grade mineralization in the breccia felsic magmatic–hydrothermal brecciation at ∼155 Ma, and bodies. Fluid inclusions show evidence of boiling during younger (141–142 Ma) potassic, propylitic, and sodic the potassic and sodic alteration events, which occurred at alterations, coeval with dioritic and granodioritic stocks temperatures around 450–460°C and 350–410°C, and and sills, and dioritic dikes. Main ore formation is salinities between 3–53 and 13–45 wt% NaCl eq., genetically related to the second hydrothermal event, and respectively. The hydrothermal events occurred during consists of hydrothermal breccias, disseminations and episodic decompression due to fluid overpressuring, stockwork-style mineralization, associated with sodic hydrofracturing, and sharp changes from lithostatic to alteration. Hypogene sulfide assemblages show distinctive hydrostatic conditions. Sulfur isotope results of hypogene vertical and lateral zoning, centered on magmatic and sulfide minerals fall in a narrow range around 0 per mil, hydrothermal breccia bodies, which constitute the feeders suggesting a dominance of magmatic sulfur. Carbon and to mineralization. A barren pyrite root zone is overlain by oxygen isotopic data of calcites from propylitic alteration suggest a mantle-derived carbon and oxygen isotope fractionation due to low-temperature alteration.
L. E. Ramírez (*) . C. Palacios . B. Townley . M. A. Parada Keywords Cu mineralization . Upper jurassic . Departamento de Geología, Universidad de Chile, Coastal range . Northern Chile P.O. Box 13518-21, Santiago, Chile e-mail: [email protected] Tel.: +56-2-9780233 Fax: +56-2-6963050 Introduction
A. N. Sial This paper presents the results of a comprehensive and NEG LABISE Department of Geology, Federal University of Pernambuco, updated study of the Mantos Blancos ore deposit, in the C. P. 7852, Coastal Range of northern Chile (Fig. 1). Pre-mining Recife-PE, 50.732-970, Brazil resources of this deposit are estimated at 500 million metric tons with 1.0% Cu, of which 200 million tons were J. L. Fernandez-Turiel extracted between 1960 and 2002 (Maksaev and Zentilli Institute of Earth Sciences J. Almera, CSIC, Sole i Sabaris, 2002). The remaining ore reserves stand at 142 million tons 08028, Barcelona, Spain with 0.86% Cu, and a resource of 156 million tons with 0.89% Cu (Anglo Base Metals Report, May 2003). . D. Gimeno M. Garcia-Valles The Coastal Range is host to Upper Jurassic to Lower Faculty of Geology, University of Barcelona, Marti i Franques, Cretaceous copper deposits of volcanic-hosted strata- 08028, Barcelona, Spain bound type, and Cretaceous, generally heavily eroded porphyry-type systems, which constitute a NS-trending B. Lehmann metallogenetic province (Camus 2003). The volcanic- Institut für Mineralogie und Mineralische Rohstoffe, Technische Universität Clausthal, hosted strata-bound ore bodies are mainly associated with Adolph Roemer Strasse 2 A, hydrothermal breccia feeder structures, in which the 38678 Clausthal-Zellerfeld, Germany hydrothermal breccias contain at least 50% of the economic
Excursion métallogénique - Chili 2012 Références page 171 sequence by Chávez (1985), but, in general, has been considered as a strata-bound Cu deposit in recent reviews (Espinoza et al. 1996; Maksaev and Zentilli 2002). No detailed studies have been performed since 1985, when Mantos Blancos comprised a series of open pits and underground mines. During the past 20 years, the mine has been transformed into a large open-pit operation, which now provides much better geological exposures and more detailed information. The aim of this paper is to present new data on the geology, hydrothermal alteration and mineralization, fluid inclusions, and stable isotopes, and to discuss the metallogeny and origin of the deposit.
Tectonic and geologic setting
During the Jurassic to Early Cretaceous, a subduction- related magmatic belt was established along the present Coastal Range of northern Chile. It is represented by a 7,000-m thick basaltic to andesitic volcanic pile (La Negra Formation) and granitic to dioritic plutonic rocks. The volcanic sequence evolved with time from an initial stage of tholeiitic affinity to a calc-alkaline composition (Palacios 1984; Rogers and Hawkesworth 1989; Pichowiak et al. 1990; Kramer et al. 2005). Based on radiometric age data and paleontological arguments, the extrusive event oc- curred between the Lower Jurassic to the Oxfordian (Rogers and Hawkesworth 1989; Gelcich et al. 2004; Kramer et al. 2005). The Jurassic volcanic pile was deposited without significant relief building, indicating considerable crustal subsidence, probably related to crustal thinning in an extensional setting (Dallmeyer et al. 1996; Maksaev and Zentilli 2002). Fig. 1 Geological map of the Coastal Cordillera, Northern Chile, and location of the Mantos Blancos ore deposit (star) and the Upper The intrusive rocks, also of calc-alkaline composition, Jurassic volcanic-hosted copper deposits (diamonds). In grey are the include granites, tonalites, granodiorites, and diorites of Middle to Upper Jurassic volcanic rocks of the La Negra Formation, Lower Jurassic to Early Cretaceous age (200–130 Ma; crosses represent Jurassic plutonic rocks. Modified after Maksaev Scheuber and Gonzalez 1999; Oliveros 2005). Tectonic and Zentilli (2002) evolution of the Coastal Range during the Jurassic is interpreted in terms of coupling and decoupling between mineralization and the highest ore grades. The hydro- the subducting oceanic and overriding continental plates thermal breccias are coeval with barren and generally (Scheuber and Gonzalez 1999). From 195 to 155 Ma, an incipiently altered stocks and sills of mainly dioritic intra-magmatic belt was widespread, spatially related to the composition, and are intruded by late mineralization north–south trending, sinistral strike–slip dominant Ataca- dioritic dikes. ma Fault Zone. However, at the end of Jurassic time, due to Sulfide mineralization consists of chalcocite, digenite, foundering of the subducting plate, subduction rollback, bornite, chalcopyrite, and pyrite related to sodic hydro- and decoupling, an east–west-trending extensional regime thermal alteration (Palacios 1990; Wolf et al. 1990). Most developed. At the end of the Jurassic to the Early of these deposits are relatively small, with resources Cretaceous, seismic coupling of the subducted plate is between 10 to 50 million tons grading 1% Cu (Espinoza et suggested by the return of the sinistral strike–slip style of al. 1996). The porphyry-copper-type mineralization is deformation (Scheuber and Gonzalez 1999). associated with granodioritic porphyries and hydrothermal breccias, in which the hypogene mineralization consists of chalcopyrite, pyrite, and minor bornite and molybdenite, Geology of the deposit and occurs coeval with potassic and phyllic alteration (Camus 2003). Rock units recognized within the Mantos Blancos ore The Mantos Blancos ore body, located 30 km NE of deposit consist of a rhyolitic dome and its magmatic– Antofagasta, was described in the past as disseminated hydrothermal breccias, intruded by dioritic and granodi- copper mineralization in a bimodal rhyolite–andesite oritic stocks and sills. The dioritic and granodioritic stocks
Excursion métallogénique - Chili 2012 Références page 172 locally grade upwards into magmatic–hydrothermal brec- laminations are typical, varying between 1 to 4 cm in cias. These rock units are all mineralized to variable thickness. West of the pit, the felsic dome is intercalated degrees. Late mafic dikes crosscut all previously men- with felsic tuffs and andesitic lava flows, and is intruded by tioned rock units and are essentially barren. All the above dioritic and granodioritic sills. The rhyolitic dome consists rock units are informally grouped as the Mantos Blancos of a rhyolite porphyry with fragments of corroded quartz Igneous Complex (MBIC; Fig. 2). The local structural and feldspar phenocrysts (1–5 mm) in an intensively framework at deposit scale is characterized by three groups altered felsic groundmass. of faults: 1) NE- and NW-trending subvertical faults with evidence of sinistral and dextral movements respectively, 2) NS / 50–80° W normal faults, and 3) NS / 50–80° E Rhyolitic magmatic–hydrothermal breccia system normal faults. The MBIC consists of the following major rock units: Several sub-vertical monomictic and matrix-supported rhyolitic magmatic and hydrothermal breccia bodies, have been recognized within the felsic dome intrusion (Figs. 2 Rhyolitic porphyry dome and 3). They consist of irregular bodies, about 100 to 250 m in vertical extent, and semi-oval to circular sections, 50 to The central part of the deposit consists of a rhyolitic dome 100 m in diameter. The matrix is composed of rhyolitic rock (Figs. 2 and 3). The dome structure is partially preserved in flour with intense alteration and disseminated sulfide the open-pit walls, but its geometry has been roughly minerals (Fig. 4a). The fragments are altered, irregular in defined from drill core logs and samples of the early stages shape, poorly sorted, and vary in size between 1 cm and of exploitation of the ore deposit (Chávez 1985), and later several meters. In the centre of the ore deposit, the rhyolitic lithological modeling. Due to pervasive alteration, the magmatic and hydrothermal breccias are intruded by late contacts between different internal flows are very difficult dioritic to granodioritic magmatic–hydrothermal breccias. to observe; however, near-horizontal and vertical flow
Fig. 2 Geological map of the Mantos Blancos ore deposit
Excursion métallogénique - Chili 2012 Références page 173 W E Ore grade > 0.5% Cu 1.000
800 cp-py cp-py
cp-py cp-bor cp-bor cs-dig 600 cs-dig cp-bor 400 cp-py cp-py cp-py cp-py 200 cp-py cp-py py py 0 py py Elevation (m.a.s.l) 0 1 Km py py
Fig. 3 E–W profile of the Mantos Blancos ore deposit. For symbols, and location of profile see Fig. 2
Bimodal stock and sill system they were not observed and described in the earlier study by Chávez (1985). The upper part of the breccia pipes The rhyolite dome is intruded by a subvolcanic complex of exhibit hydrothermal characteristics as evidenced by the porphyritic dioritic and granodioritic stocks and sills. At presence of a matrix mainly composed of hydrothermal least five gently dipping sills of both rock types occur in the gangue and ore minerals. The breccia consists of altered mine, varying in thickness between 10 and 50 m. The angular and subrounded fragments of the rhyolitic dome feeder relationship between the stocks and sills has been and the granodioritic and dioritic porphyries. They are locally observed (Fig. 3). The granodiorite porphyry is poorly sorted and range in size from 1 cm to 15 m. composed of 10 to 30% phenocrysts of hornblende, Downwards in the breccia bodies, magmatic features are plagioclase, quartz, and biotite, in a groundmass of quartz, progressively evident, with granodioritic fragments in an feldspars, biotite, and hematite microlites. The diorite altered and mineralized dioritic matrix, as well as dioritic porphyry has 5 to 10% pyroxene and minor amphibole fragments in a granodioritic matrix (Fig. 4f). phenocrysts in a groundmass of fine-grained pyroxene, plagioclase, and magnetite. In both rock types, the porphyritic texture grades to aphanitic near the intrusive Mafic dyke swarm margins. The diorite porphyry has millimeter-size amyg- dules filled with quartz and quartz-sulfide. Mutual intrusive Intruding all the rock units in Mantos Blancos deposit, relationships between both granodioritic and dioritic rocks partially altered late-ore dioritic dikes were emplaced. They are common, and enclaves of one in the other have been are subvertical and have orientations preferentially NNE, frequently observed. The dioritic enclaves show convolute and subordinate NS–NNW. The dikes are 1 to 12 m wide to flame-like contacts (Fig. 4b) with the host granodiorite, and represent about 15% of the total rock volume in the whereas, the granodioritic enclaves exhibit sharp or deposit. They exhibit porphyritic texture, composed of 10– brecciated contacts with the surrounding diorite. Back- 25% phenocrysts of altered plagioclase, amphibole, and veining between the two lithological types is also observed. minor pyroxene, in a very fine-grained groundmass of Recent 40Ar/39Ar data on amphibole provide ages of feldspar, amphibole, and minor biotite and magnetite. An 142.18±1.01 Ma for the granodiorite, and 141.36±0.52 Ma 40Ar/39Ar date on amphibole from a late-mineral dike in the for the diorite (Oliveros 2005). mine is 142.69±2.08 Ma of age (Oliveros 2005).
Dioritic to granodioritic magmatic–hydrothermal Hydrothermal alteration and mineralization breccia system Two hydrothermal events have been recognized, based on Two polymictic and matrix-supported pipe-like magmatic– the superimposition of alteration minerals and relationship hydrothermal breccias hosted within the rhyolitic dome, at between different stages of veinlets. The first event is the top of some dioritic and granodioritic stocks and represented by the rhyolitic magmatic–hydrothermal spatially related with NS-trending faults, are recognized brecciation hosted by the rhyolitic dome. The second (Figs. 3 and 4c−e). The central and largest breccia body is event, which represents the main stage of mineralization, is crosscut by at least three metric-size sills; two dioritic and hosted mostly within the dioritic to granodioritic mag- one granodioritic in composition. The breccias form near- matic–hydrothermal breccias, dioritic sills, and the rhyo- vertical bodies, with a vertical extent of about 700 m, and litic dome, and may be genetically associated with the diameters between 100 and 500 m. It is likely that these intrusion of dioritic and granodioritic stocks. bodies did not reach the upper levels of the ore deposit, as
Excursion métallogénique - Chili 2012 Références page 174 Fig. 4 Photographs of: a rhyo- litic magmatic-hydrothermal breccia, b dioritic enclave with- in the granodiorite showing convolute contacts, c, d, and e dioritic to granodioritic mag- matic-hydrothermal breccias in which hydrothermal features dominate, f dioritic to grano- dioritic magmatic-hydrothermal breccia with dominating mag- matic features, and g pebble dike
First hydrothermal event rhyolitic dome. In the rhyolitic magmatic–hydrothermal breccias, chalcopyrite and bornite are the most abundant The first hydrothermal event is characterized by the sulfides. Around these bodies the sulfides are chalcopyrite assemblage chalcopyrite, bornite, pyrite, quartz, and ser- and pyrite. The phyllic veinlets contain the sulfide minerals icite. This assemblage occurs: 1) disseminated in the matrix as open space filling within fractures, and often display of irregular and sub-vertical bodies of rhyolitic magmatic– weak alteration halos of sericite and quartz. Due to the hydrothermal breccias, 2) planar veinlets, 3) disseminated intense and widespread superimposition of the main within the rhyolitic dome and in fragments of the (second) hydrothermal event, it was not possible to hydrothermal breccias, and 4) as isolated crystals or as establish the extent and intensity of this first event. It rim assemblages within and on quartz phenocrysts of the probably extended to all rocks of the rhyolitic dome. An
Excursion métallogénique - Chili 2012 Références page 175 40Ar/39Ar age on sericite from this first hydrothermal event feldspar, tourmaline, and biotite are observed in most yields an age of 155.11±0.786 Ma (Oliveros 2005). locations, suggesting that potassic alteration was initially widespread, but was subsequently overprinted and ob- literated by later alteration stages. Dioritic and granodiorit- Second hydrothermal event ic sills, that contain amygdules filled with quartz, chlorite, digenite, chalcopyrite, and traces of K-feldspar and The main hydrothermal alteration and mineralization event tourmaline, intruded the magmatic–hydrothermal breccias. at Mantos Blancos is centered on the dioritic to Propylitic alteration occurs extensively in the whole granodioritic magmatic–hydrothermal breccias and is deposit, affecting all of the rocks (including sills and considered syngenetic with both breccia formation and dikes), and overprinting and obliterating the potassic emplacement of the granodioritic and dioritic stocks and alteration assemblage. It occurs as disseminations and sills. The mineralized zone extends discontinuously for veinlets of quartz, chlorite, epidote, calcite, albite, sericite, 3kminanE–W direction, has a width of up to 1 km and hematite and minor chalcopyrite, galena, and pyrite. These depth of 600 m. The hypogene mineralization occurs minerals also fill amygdules within dioritic sills and dikes. between the elevations of 720 and 450 m asl. (Fig. 3). Laterally, propylitic alteration consists of quartz, chlorite, Primary mineralization developed mainly within and epidote, and pyrite, forming a ring around the orebody at around the magmatic–hydrothermal breccia pipes, yet the least 2 km wide. From elevations of 600 m to the upper part ore deposit exhibits a discontinuous lateral ore grade of the deposit, a swarm of N 25–30° E striking and sub- distribution. The highest Cu grades occur within the vertical pebble-dikes have been observed. These pebble- breccias with lateral zoning to progressively lower dikes are 10- to 20-cm thick and consist of rounded concentrations. This fact suggests that the magmatic– fragments of the rhyolitic dome, dioritic and granodioritic hydrothermal breccia pipes served as the feeder bodies of rocks, set in a matrix of quartz, epidote, calcite, galena, and the main mineralization. pyrite (Fig. 4g). In the second hydrothermal event, the early alteration Both potassic and propylitic alterations were followed stage was potassic and propylitic, followed by sodic by sodic alteration, containing albite (replacing feldspar), alteration. The potassic and propylitic mineral assemblages hematite, pyrite, chalcopyrite, and Ag-rich digenite, with are centered on the dioritic to granodioritic magmatic– minor amounts of quartz. This mineral assemblage is very hydrothermal breccias, affecting all lithologies of the extensive, centered on the magmatic and hydrothermal deposit. These alteration types developed pervasively, breccias, and occurs as disseminations, cavity fillings, and disseminated, filling amygdules within the dioritic sills, sharp veinlets. Sodic alteration and mineralization affected and as weak halos around flame-like veinlets that crosscut all lithological types between elevations of 500 m to the the first generation phyllic veinlets in the rhyolitic dome. surface and spatially coinciding with the current commer- The potassic alteration is characterized by K-feldspar, cial ore zone. Above the elevation of 500 m, the dioritic quartz, tourmaline, biotite–chlorite, magnetite, chalcopy- sills that intruded the magmatic–hydrothermal breccias rite, digenite, and minor pyrite (Fig. 5). Relicts of K- exhibit intense stockwork with a sodic alteration mineral assemblage. As the syn-mineralization granodioritic and HYDROTHERMAL EVENTS dioritic stocks and sills have been dated at 142.18±1.01 and MINERALS First Second 141.36±0.518 Ma (Oliveros 2005), respectively, and a late- Phyllic Potassic Sodic Propylitic ore dike yields an age of 142.69±2.083 (Oliveros 2005), Quartz the age of the main hydrothermal event is constrained Sericite between 141 and 142 Ma. K-feldspar Supergene oxide mineralization has been mined, with Biotite only patches of atacamite, chrysocolla, and malachite Tourmaline remaining. This supergene mineralization was described in Chlorite detail by Chávez (1985). Although he reported primary Albite Epidote chalcocite (late within the hypogene assemblage), our data Calcite indicate the presence of only secondary chalcocite (Fig. 6). Pyrite The secondary sulfides are mainly chalcocite (forming Magnetite zones of high-grade copper mineralization centered over Hematite the magmatic–hydrothermal breccia bodies, with bornite– Chalcopyrite digenite), and weak layers of covellite, together with Bornite cuprite-native copper and tenorite. Digenite Galena Magmatic and Rhyolitic Dioritic and granodioritic Fluid inclusion studies hydrothermal dome and stocks and sills, brecciation events brecciation and dike intrusion. Fluid inclusion studies were carried out on quartz crystals Fig. 5 Hypogene mineral assemblage of the hydrothermal events at of the second hydrothermal event. Samples include quartz the Mantos Blancos ore deposit crystals from potassic, propylitic, and sodic veinlets, and
Excursion métallogénique - Chili 2012 Références page 176 Fig. 6 Microphotographs of a digenite relict in chalcocite, b and c digenite with hematite flakes replaced by chalcocite, d chalcocite with inclusions of hematite flakes, e chalcopyrite replaced by covellite (blue), and f native copper in cuprite (red internal reflections in grey) with replacement rim of tenorite
from potassic and propylitic amygdules of the dioritic sills either liquid CO2 or clathrate formation, freezing point and stocks. A total of 23 samples were taken from the depression measurements rule out the presence of signif- central part of the deposit (Fig. 7), from which 153 icant CO2. Apparent salinities are reported in weight microthermometric measurements of primary inclusions percent NaCl equivalent (wt% eq.), based on the halite were done. Vertical sampling extends to a depth of 850 m. solubility equation for halite-saturated inclusions and on Heating and freezing experiments were conducted on a the final ice-melting temperature for halite-undersaturated Linkam THMS600 stage for homogenization temperatures inclusions (Bodnar and Vityk 1994). The fluid inclusion (Th) up to 450°C and on a Linkam TS1500 stage for Th microthermometric data are presented in Table 1 and Fig. 8. above 450°C. The uncertainty for heating runs is about The highest temperatures were measured in types II and ±2°C at 400°C. IIIb inclusions trapped in quartz from veinlets of the Three fluid inclusion types were recognized, following potassic alteration assemblage within the matrix of the the classification scheme of Nash (1976): I (liquid- magmatic–hydrothermal breccia at elevations between 239 dominant inclusions without halite daughters), II (vapor- and 260 m. The type-II inclusions homogenize between dominant inclusions without halite daughters), and IIIb 550 and 608°C and have salinities of 9.9 to 10.1 wt% NaCl (vapor-dominant inclusions with halite daughters). All eq., whereas, the IIIb-type inclusions have Th values fluid inclusions types have mostly rounded shapes and between 530 and 590°C and salinities ranging from 52 to ranged from 5 to 15 μm. No evidence was observed for 74 wt% NaCl eq. The coexistence of both types of
N S
Ore grade > 0.5% Cu 1.000
900 CP-1-2 CP-1-15 Q-13 Q-103 Q-12 800 Q-8 Q-11 CP-1-22 Q-5 Q-9 Q-102 cp-py 700 Q-4 Q-6 Q-10 Q-7 Q-100 Q-3 Q-2 cp-cs-dig Q-101 600
cp-py 500 cp-dig
400 cp-py cp-py Q-104 Q-1-1 300 Q-105 Q-1 200 cp-py 100
0 Elevation (m) 0 300 600 m Fig. 7 N–S profile of the Mantos Blancos deposit showing the location samples used in the fluid inclusions study. For symbols, and location of profile, see Fig. 2
Excursion métallogénique - Chili 2012 Références page 177 Table 1 Microthermometry data of fluid inclusions from the second hydrothermal event Sample Elevation Size Th (L-v) Th (Halite) %L %V % Tm (ice) Salinity (wt% Remarks N° of (m.a.s.l.) (μm) (°C) (°C) (%in) Halite (°C) NaCl equiv) inclusions
Q-1 239 5–8 601±7 24±9 76±9 −6.5±0.5 9.9±0.7 Veinlets of K-assemblage 5 in MHB 239 5–9 500±20 580±10 10±5 30±4 60±5 71±3.0 5 Q-104 247 8–10 505±15 20±5 20±5 −18.0±2 19.4±3.0 Veinlets of K-assemblage 9 in sill of dioritic porphyry Q-1-1 260 5–10 564±14 550±20 23±8 77±8 −6.7±0.8 10.1±1.0 Veinlets of K−assemblage 5 in MHB 260 5–10 490±10 20±10 20±10 60±10 62±10.0 7 Q-105 260 8–10 465±12 19±6 81±5 −15.0±3.5 18.5±3.0 Veinlets of K-assemblage 5 in sill of dioritic porphyry Q-2 684 5–10 390±12 449±20 11±4 51±6 38±4 52.4±1.6 Veinlets of K-assem- 5 blage in MHB 6–10 462±8 15±10 85±10 −1.5±0.5 2.5±0.8 3 Q-3 684 5–8 404±6 464±6 10±2 50±10 40±8 53.5±0.5 Veinlets of K-assem- 3 blage in MHB 10 455±6 10±5 90±5 −2.0±1 3.3±2.5 2 Q-100 720 5–10 413±13 20±10 80±10 −19.4±1.4 22.2±10 Amygdules filled by K- 5 assemblage in dioritic sill Q-101 720 10–15 380±15 25±10 75±10 −19.4±1.4 22.1±10 Amygdules filled by K- 5 assemblage in dioritic sill Q-4 696 8–10 302±16 349±26 15±6 50±5 35±8 42.2±1.9 Veinlets of Albitic as- 6 semblage in matrix of MHB 8–10 357±23 10±6 90±6 9.9±0.9 13.9±1.1 5 Q-5 696 8 349±20 349±20 6±5 60±10 35±5 42.3±1.6 Veinlets of Albitic as- 2 semblage in MHB 8–15 346±6 9±3 90±7 −9.4±1.2 13.4±1.4 5 Q-6 696 7–10 362±8 10±5 90±5 13.2±1.8 Veinlets of Albitic 5 assemblage in MHB Q-7 708 7–10 356±11 8±2 92±2 −9.7±1.2 14.0±1.4 Veinlets of Albitic 5 assemblage in MHB Q-8 720 8–10 376±25 413±2 10±4 50±2 40±6 47.8±0.3 Veinlets of Albitic 3 assemblage in MHB 5–15 351±23 10±5 90±5 −8.8±1.8 12.6±2.2 3 Q-9 720 8 371 423 8±2 50±4 42±5 48.7 Veinlets of Albitic 1 assemblage in MHB 8–10 313±15 11±7 89±4 −8.5±1.0 12.3±1.3 5 Q-103 768 5–10 358±3 75±10 25±10 −12.5±5.0 15.3±2.5 Veinlets of K-assem- 6 blage in sill of dacitic porphyry Q-10 720 8–10 301±1 90±5 10±5 7.1±0.1 10.6±1.0 Veinlets of Propylitic 2 assemblage in sill of dioritic porphyry CP-1-22 760 8–12 218±25 65±8 35±8 −19±6.8 20±2.4 Amygdules in dioritic 11 porphyry filled by Pro- pylitic assemblage Q-11 780 8–15 269±11 70±10 30±10 −6.6±0.6 9.8±0.9 Veinlets of Propylitic 4 assemblage in RPD Q-12 780 7–12 249±5 68±12 32±12 −7.9±1.3 12.0±2.4 Veinlets of Propylitic 5 assemblage in RPD Q-102 792 8–10 335±5 90±4 10±6 −10.5±0.5 14.5±0.5 Veinlets of Propylitic 2 assemblage in sill of dioritic porphyry
Excursion métallogénique - Chili 2012 Références page 178 Table 1 (continued) Sample Elevation Size Th (L-v) Th (Halite) %L %V % Tm (ice) Salinity (wt% Remarks N° of (m.a.s.l.) (μm) (°C) (°C) (%in) Halite (°C) NaCl equiv) inclusions
Q-13 792 8–10 247±3 70±5 30±5 −6.5±0.5 9.8±0.7 Veinlets of Propylitic 6 assemblage in RPD CP-1–15 816 7–11 187±35 65±10 35±10 −8.8±5.6 12±5.1 Amygdules in dioritic 6 porphyry filled by Pro- pylitic assemblage CP-1–22 816 6–9 318±15 80±10 20±10 −10.1±1.3 14.1±1.2 Veinlets of Propylitic 2 assemblage in dioritic porphyry Th (L+v) Liquid-Vapor homogenization temperature, Th (Halite) halite dissolution temperature, Tm (ice) melting temperature of ice, % L,V, Halite abundance of phases at room conditions, MHB magmatic and hydrothermal breccia, RPD rhyolitic porphyry dome
inclusions within the same growth zone of a quartz crystal, inclusions, in which Th values vary between 340 and 150°C is considered as indicative of deposition from boiling and salinities between 9 and 22 wt% NaCl eq. fluids. In these brines, Th (halite) values are at least 60°C Fluid inclusions in quartz related to the sodic assemblage greater than Th (l−v) values in the same samples (Fig. 9). were difficult to measure due to the limited amounts of Fluid inclusion observations of samples from potassic albite-bearing quartz veinlets. Fluid inclusions in quartz alteration assemblages at an elevation of 684 m also display obtained from these veinlets in the matrix of the magmatic- evidence of boiling: Type-IIIb inclusions have Th values hydrothermal breccia at elevations between 696 and 768 m, between 449 to 464°C and salinities between 52.4 and 53.5 are mainly of types II and IIIb. Evidence of boiling has NaCl eq., and co-exist with vapor-rich type-II inclusions been recognized at elevations of 696 to 720 m asl, in which (with Th between 462 and 415°C, and salinities between both types of inclusions coexist in growth zones of similar 2.5 and 3.3 wt% NaCl eq.). Also in these brines, Th (halite) hydrothermal quartz crystals. The brines have Th values values are at least 65°C greater than Th (l-v) values in the between 349 and 423°C and salinities ranging between 42 same samples. Quartz crystals from potassic alteration and 48 wt% NaCl eq., whereas, the vapor-rich-two phase assemblage in amygdules and veinlets from sills in the inclusions have Th values between 313 and 364°C and diorite contain type I and II inclusions. In these samples, Th salinities between 13 and 14 wt% NaCl eq. Brines in the values decrease systematically with an increase in elevation (from an average of 515°C at 360 m to 365°C at 720 m). In contrast, salinities remain relatively constant (19–22 wt% NaCl eq.). Fluid inclusions associated with propylitic alteration assemblages have been measured in samples from elevations of 720 to 816 m. They correspond to type-I
70 Potassic alteration Sodic alteration 60 Propylitic alteration
50
40
30 Salinity ( wt% NaCl eq.) 20
10
200 250 300 350 400 450 500 550 600 Th (˚C) Fig. 9 Halite dissolution temperature versus liquid-vapor homog- enization temperature of boiled fluid inclusion samples from Fig. 8 Homogenization temperature vs salinity of fluid inclusions potassic and sodic alteration
Excursion métallogénique - Chili 2012 Références page 179 same sample exhibit halite dissolution temperatures greater Carbon and oxygen than the vapor homogenization temperatures. Eighteen calcite samples were analyzed for δ13C and δ18O at the stable isotope laboratory (LABISE) of the Depart- Stable isotope studies ment of Geology, Federal University of Pernambuco, Brazil. CO2 gas was extracted from micro-drilled powder, Sulfur in a high-vacuum line after reaction with 100% orthophos- phoric acid at 25°C for 1 day. CO2 released, after cryogenic Seventeen sulfide samples from the second hydrothermal cleaning, was analyzed in a double inlet, triple collector event were analyzed for δ34S at the Scientific-Technical SIRA II mass spectrometer. Results are reported relative to Services of the University of Barcelona. Sulfide samples PDB, in per mil notation. The uncertainties of the isotope were separated mechanically to obtain splits with 50–80 μg measurements were better than 0.1‰ for carbon and 0.2‰ of sulfur. Between 100 and 300 μg of pure sulfide were for oxygen, based on multiple analyses of an internal 13 18 mixed with V2O5 (1:1), homogenized and packed into laboratory standard (BSC). Values of δ C and δ Oof high-purity tin cups. The sulfur isotopic composition was calcite samples from propylitic alteration stage (of the analyzed using a Continuous Flow-Isotope Ratio Mass second hydrothermal mineralization event) are reported in Spectrometry (CF-EA-IRMS). Samples were combusted in Table 3 and Fig. 11. All samples were taken in the central an elemental analyzer (Carlo Erba EA 1108) connected to a part of the deposit, between elevations of 172 and 900 m Finnigan MAT Delta C gas mass spectrometer via a asl. The carbon isotope values of calcites vary between Finnigan MAT Conflo II interface. Results are expressed in −4.37 and −6.71‰, whereas, the δ18O values fluctuate the per mil notation relative to the international Vienna- between 13.08 to 23.49‰. Canyon Diablo troilite (VCDT) standard. The reproduc- ibility of measurements was ±0.3‰. The δ34S values of 11 samples of pyrite, five samples of chalcopyrite, and one Discussion sample of digenite are reported in Table 2 and Fig. 10. All samples were taken in the central part of the deposit, Based on available radiometric ages and geological between elevations of 450 and 780 m asl. The analyzed observations described in this study, the Mantos Blancos sulfides exhibit δ34S values ranging from −5 to 1.2 per mil, ore deposit was formed by two superimposed Upper Jurassic with a mean value of −1.4‰ and a standard deviation of hydrothermal events. The older event occurred at ∼155 Ma, 1.8‰. Results are similar to those previously reported by coeval with the rhyolitic magmatic–hydrothermal breccia- Sasaki et al. (1984) and Vivallo and Henriquez (1998). tion and phyllic alteration. The younger event represents the Pyrite shows the widest sulfur isotope range in comparison main hydrothermal mineralization (∼141–142 Ma) and is to the Cu-sulfides, and the variation is independent of genetically related to dioritic and granodioritic stocks and alteration types or host rock lithology (Fig. 10). sills and coeval magmatic–hydrothermal brecciation. Prob- ably, both hydrothermal events contributed to extensive but irregularly distributed ore grades of hypogene mineraliza-
Table 2 Sulfur isotope of sul- Sample no. Mineral δ34S (‰) Hydrothermal alterationa Lithologyb fides from the main hydrother- CDT mal event at the Mantos Blancos − ore deposit M-25 Pyrite 2.0 Propylitic Granodiorite CPM-54 Pyrite −1.9 Potassic Diorite CP-122 Pyrite −2.6 Sodic Diorite CPM-53 Pyrite −4.0 Propylitic Rhyolitic dome M-3 Pyrite 1.2 Propylitic MHB M-4-A Pyrite 0.7 Propylitic MHB BC-708 Pyrite −0.1 Potassic MHB P-2-1 Pyrite −0.3 Potassic MHB C-684 Pyrite −1.1 Potassic MHB N-684 Pyrite −1.2 Potassic MHB M-24 Pyrite −5.0 Propylitic MHB M-25 Chalcopyrite −2.1 Propylitic Granodiorite a CPM-54 Chalcopyrite −0.5 Potassic Diorite Hydrothermal alteration stage − associated with the analyzed CPM-54a Chalcopyrite 2.0 Potassic Diorite sulfide CPM-53 Chalcopyrite −4.5 Potassic Rhyolitic dome b Host rock of the sulfide BC-708 Chalcopyrite −1.3 Potassic MHB MHB Magmatic Hydrothermal CPM-54a Digenite −3.2 Potassic Diorite Breccia
Excursion métallogénique - Chili 2012 Références page 180 Table 3 C and O isotope analyses (‰) of calcites from the Mantos Blancos ore deposits 18 18 13 Sample OSMOW(‰) OPDB(‰) CPDB(‰)
56-585 14.98 −15.40 −6.16 56-590 17.42 −13.04 −6.69 VB-1 18.74 −11.71 −5.50 97-230 23.49 −7.14 −6.58 VB-2 17.60 −12.86 −5.36 06-268 13.27 −16.44 −5.13 06-335 15.87 −14.54 −6.27 BC-1 13.91 −16.44 −5.13 33-200 16.72 −13.71 −6.91 33-257 20.81 −9.75 −5.72 33-288 19.87 −10.66 −4.37 33-298 13.08 −17.25 −6.02 DV-1 14.59 −15.78 −5.09 1-14B 16.51 −13.92 −6017 696-41 13.88 −16.47 −6.17 1-14C 16.68 −13.75 −5.42 CPM1-21 16.85 −13.60 −4.75
itic, and a late sodic stage. The potassic and propylitic alteration stages occurred coeval with dioritic and granodioritic porphyry stock intrusions, magmatic–hydro- thermal breccias and late sill and dike emplacements. The late sodic alteration that developed centered around the magmatic–hydrothermal breccias, associated with intense fracturing and brecciation (including in the sills) and the main mineral deposition. The ore grade, alteration, and the copper sulfide mineral zoning indicate that the magmatic– hydrothermal breccia bodies represent the feeders to the hydrothermal system. The hydrothermal activity, was followed by the intrusion of a dioritic dike swarm. An indication of local subsidence is the common occurrence of sills intruded by vertical dikes as part of the same magmatic event. Because the magmatic pressure must exceed the least main horizontal stress and the tensile strength of the rock cover to form discordant intrusions, these intrusive
Fig. 10 δ34S(‰) values of sulfides from the main hydrothermal event at the Mantos Blancos ore deposit (a). Diagrams b and c show the types of alteration and host rock, with which the sulfides are related PDB tion. High-ore-grade mineralization is restricted to the upper part of the magmatic–hydrothermal breccias from the second hydrothermal event. The radiometric ages for the two hydrothermal events reported by Oliveros (2005) agree with previous 40Ar/39Ar (total gas in albite) and whole rock Rb–Sr (errorchrons in strongly altered samples) radiometric ages – (150 146 Ma; Munizaga et al. 1991; Tassinari et al. 1993). Fig. 11 δ13C(‰)vsδ18O(‰) diagram showing the distribution of The younger event is characterized by three types of calcites from the Mantos Blancos ore deposit. Fields and arrows alteration and mineralization: an early potassic, a propyl- after Taylor et al. (1967) and Keller and Hoefs (1995)
Excursion métallogénique - Chili 2012 Références page 181 relationships between sills and dikes are an indication that sequence. These results can be interpreted as boiling events sufficiently thick magmatic overburden was progressively and associated decompression occurring episodically due formed to produce a change of the least principal stress to fluid over-pressuring, hydrofracturing, and sharp chang- from vertical to horizontal (Parada et al. 1997). As this sill– es from lithostatic to hydrostatic conditions. dike relationship has been observed at Mantos Blancos, it is The sulfur isotopic results from hypogene sulfides suggested that the tectonic setting during mineralization suggest a largely magmatic source for sulfide sulfur and corresponded to a local extensional regime, probably indicate a co-genetic relationship for the analyzed sulfide related to a transtensional faulting within the Atacama minerals. C–O isotopes in fresh calcite crystals reported in Fault System. this paper suggest C of magmatic origin, probably of Evidence of boiling associated with potassic alteration mantle provenance (Cartigny et al. 1998), and fractionation has been found in samples up to an elevation of 684 m asl. of O following the trend of low-temperature alteration At this elevation, fluid inclusions Th values exceed 450°C. caused by magmatic–hydrothermal fluids. At such temperatures, rocks in the hydrothermal system behave in a ductile manner: with strain rates smaller than Acknowledgements This study was funded by a FONDEF 10−14/s, rocks of dioritic or granodioritic compositions (CONICYT, Chile), grant DO1-1012, awarded to the authors and the Mantos Blancos division of Anglo American Chile. Permission behave quasiplastically, making brittle fracturing difficult for publication was granted by the University of Chile, the Chilean and allowing fluid pressure to approach lithostatic values Government, and AngloAmerican Chile. We thank the Mantos (Fournier 1991, 1999). As a consequence, the magmatic– Blancos mine geology staff, especially to Jorge Pizarro, with whom hydrothermal breccias most likely did not reach the we had the pleasure of working. Special acknowledgement to Jens Wittenbrink for his constructive comments to the manuscript. Finally, paleosurface, and the hydrothermal system mostly formed this paper was improved through the valuable reviews of Shoji at lithostatic pressure. The hydrothermal fluids within the Kojima, Robert King and Larry Meinert. magmatic–hydrothermal breccias evolved along a cooling trend, as indicated by the fluid inclusion data in quartz of the propylitic assemblage. References The emplacement of dioritic and granodioritic sills crosscutting the magmatic–hydrothermal breccias at dif- Bodnar RJ, Vityk MO (1994) Interpretation of microthermometric – ferent levels, sealed the hydrothermal system, over- data for H2O NaCl fluid inclusions. In: De Vivo B, Frezzotty ML (eds) Fluid inclusion in minerals: methods and applica- pressured the fluids, hydrofractured the rocks, and tions. VPI, Blackburg, Virginia, pp 117–130 produced the sodic boiling. The thermodynamic evolution Camus F (2003) Geología de los sistemas porfíricos en los Andes de of brine into the field of gas+solid salt at 350–400°C Chile. SERNAGEOMIN, Chile, p 267 (conditions under which sodic alteration associated boiling Cartigny P, Harris JW, Javoy M (1998) Eclogitic diamond formation occurred), has important implications regarding the con- at Jwaneng: no room for a recycled component. Science 280:1421–1424 centration of HCl that may be transported when and if Chávez W (1985) Geological setting and the nature and distribution steam escapes into the overlying rocks. Fournier and of disseminated copper mineralization of the Mantos Blancos Thompson (1993) noted an abrupt increase in the concen- district, Antofagasta Province, Chile. Ph.D Thesis, University tration of HCl° in steam when NaCl begins to precipitate at at California, Berkeley, USA, p 142 Deines P (1989) Stable isotope variations in carbonatites. In: Bell K pressures below 300 bars. This increase occurs because (ed) Carbonatites—genesis and evolution. Unwin Hyman, hydrolysis reactions that produce HCl° and NaOH by the London, pp 301–359 reaction of NaCl with H2O become important only at Dallmeyer RD, Brown M, Grocott J, Taylor GK, Treolar PJ (1996) pressures sufficiently low for halite (and probably also Mesozoic magmatic and tectonic events within the Andean plate boundary zone, 26°–27°30′ S, North Chile: Constraints NaOH) to precipitate (Fournier and Thompson 1993). In from 40Ar/39Ar mineral ages. J Geol 104:19–40 addition, an order of magnitude higher than HCl° concen- Espinoza S, Véliz H, Esquivel J, Arias J, Moraga A (1996) The tration is obtained at comparable pressures and tempera- cupriferous province of the Coastal Range, Northern Chile. In: tures when quartz is present. This occurs because quartz Camus F, Sillitoe RH, Petersen R (eds) Andean copper reacts with NaOH to form albite at the expense of K- deposits: new discoveries, mineralization, styles and metallo- geny. Econ Geol, Spec Publ 5:19–32 feldspar or plagioclase (Fournier and Thompson 1993). Fournier RO (1991) The transition from hydrostatic to greater than The limited amounts of quartz-bearing albite veinlets in the hydrostatic fluid pressure in present active continental hydro- deposit support this model. thermal systems in crystalline rock. Geophys Res Lett 18: In addition, as fluids migrated away from the early heat 955–958 – Fournier RO (1999) Hydrothermal processes related to movement of source (the magmatic hydrothermal breccias) and down a fluid from plastic into brittle rock in the magmatic–epithermal thermal gradient, K-feldspar was the stable alteration environment. Econ Geol 94:1193–1212 mineral, as reflected by potassic alteration. The reverse Fournier RO, Thompson JM (1993) Composition of steam in the reaction operated when fluids migrated away from a system NaCl–KCl–H2O–quartz at 600°C. Geochim Cosmochim Acta 57:4365–4375 second heat source (intrusion of sills), conditions under Gelcich S, Davis DW, Spooner ET (2004) Onset of Early Jurassic which the albite stability field expanded at the expense of magmatism in northern Chile: precise U–Pb zircon ages for the K-feldspar (Hezarkhani et al. 1999; Simmons and Browne La Negra Formation and the Flamenco Pluton in the Coastal 2000). Both processes probably occurred at Mantos Cordillera of Chañaral. Proc. IAVCEI General Assembly, Blancos, in which the entire evolution points to a prograde Pucón, Chile (Electronic version) (potassic and propylitic)–retrograde (sodic) hydrothermal
Excursion métallogénique - Chili 2012 Références page 182 Sedimentary Geology 180 (2005) 125–147 www.elsevier.com/locate/sedgeo
The hot spring and geyser sinters of El Tatio, Northern Chile
J.L. Fernandez-Turiel a,*, M. Garcia-Valles b, D. Gimeno-Torrente c, J. Saavedra-Alonso d, S. Martinez-Manent b
a Institute of Earth Sciences J. Almera-CSIC, Sole I Sabaris s/n, 08028 Barcelona, Spain b Faculty of Geology, Dpt. Crystallography, Mineralogy and Mineral Deposits, University of Barcelona, Marti i Franques s/n, 08028 Barcelona, Spain c Faculty of Geology, Dpt. Geochemistry, Petrology and Geological Exploration, University of Barcelona, Marti i Franques s/n, 08028 Barcelona, Spain d IRNASA-CSIC, Cordel de Merinas 40-52, 37008 Salamanca, Spain Received 10 February 2005; received in revised form 13 July 2005; accepted 19 July 2005
Abstract
The siliceous sinter deposits of El Tatio geothermal field in northern Chile have been examined petrographically and mineralogically. These sinters consist of amorphous silica (opal-A) deposited around hot springs and geysers from nearly neutral, silica-saturated, sodium chloride waters. Water cooling and evaporation to dryness are the main processes that control the opal-A deposition in both subaqueous and subaerial settings, in close spatial relation to microbial communities. All fingerprints of organisms observed in the studied sinter samples represent microbes and suggest that the microbial community is moderately diverse (cyanobacteria, green bacteria, and diatoms). The most important ecological parameter is the temperature gradient, which is closely related to the observed depositional settings: 1) Geyser setting: water temperature=70–86 8C (boiling point at El Tatio: 4200 m a.s.l.); coarse laminated sinter macrostructure with rapid local variations; biota comprises non- photosynthetic hyperthermophilic bacteria. 2) Splash areas around geysers: water temperature=60–75 8C; laminated spicule and column macrostructure, locally forming cupolas (b30 cm); predominant Synechococcus-like cyanobacteria. 3) Hot spring setting: water temperature=40–60 8C; laminated spicules and columns and subspherical oncoids characterize the sinter macrostructure; filamentous cyanobacteria Phormidium and diatoms (e.g., Synedra sp.) are the most characteristic microbes. 4) Discharge environments: water temperature=20–40 8C; sinter composed of laminated spicules and oncoids of varied shape; cyanobacterial mats of Phormidium and Calothrix and diatoms (e.g., Synedra sp.) are abundant. El Tatio is a natural laboratory of great interest because the sedimentary macrostructures and microtextures reflect the geological and biological processes involved in the primary deposition and early diagenesis of siliceous sinters. D 2005 Elsevier B.V. All rights reserved.
Keywords: Sinter; Opal; Cyanobacteria; Diatom; Geothermal system; Chile
* Corresponding author. E-mail address: [email protected] (J.L. Fernandez-Turiel).
0037-0738/$ - see front matter D 2005 Elsevier B.V. All rights reserved. doi:10.1016/j.sedgeo.2005.07.005
Excursion métallogénique - Chili 2012 Références page 183 126 J.L. Fernandez-Turiel et al. / Sedimentary Geology 180 (2005) 125–147
1. Introduction vide important constraints in modeling epithermal mineralization (Fournier, 1985), and in scaling pro- Waters emerging with a temperature above the blems in geothermal power plants (Yokoyama et al., human body temperature (36.7 8C) are defined as 1993; Inagaki et al., 1997). Novel extremophiles dis- bhot springsQ by Pentecost et al. (2003). However, covered in hot springs can function optimally at the the mean annual local air temperature has been the relatively extreme levels of temperature, salinity or threshold traditionally used to recognize the limits of acidity found in some industrial manufacturing pro- hot waters in geothermal fields (Galindo et al., 2004). cesses. On the other hand, biotechnical companies are A special type of hot spring is the geyser, where water investigating the physico-chemical activities of cells and steam are ejected episodically through a vent. to carry out tasks at the nanoscale (10À 9 m), espe- Hot springs are the surface manifestation of large cially in silicon applications. hydrothermal systems and are of major importance In addition, modern hot springs are analogues for for geothermal energy and in mineral exploration. understanding the biogeochemical processes implied Epithermal mineral deposits of gold, silver, copper in the formation of some ancient cherts. In this con- and many other elements may originate in these text, the Precambrian cherts are of extraordinary inter- environments (Hedenquist and Lowenstern, 1994). est because their microfossils represent some of the Geothermal systems are present in many geological earliest records of life on Earth and provide a proxy settings, and not necessarily related to volcanism. for the biogeochemical conditions of the early bio- Hot waters usually show high concentrations of sphere (Walter, 1976a,b; Walter et al., 1972, 1992; many elements and can be highly supersaturated Eugster and Chou, 1973; Konhauser and Ferris, 1996; with respect to a variety of minerals. The most Konhauser et al., 2001, 2003). The extreme conditions common deposits precipitated are travertines, made of hot springs allow these settings to serve as natural up of calcium carbonate minerals, and silica sinters, laboratories to explore environments for extraterres- composed of amorphous silica. The process of silica trial life (Treiman, 1998; Farmer, 1999). The silica sinter formation is strongly related to the abrupt deposits associated with active hot springs are very cooling and evaporation of the hot water. In addition small in volume compared to most other types of to these abiogenic processes, a microbial role has rocks at the Earth’s surface. However, they provide also been postulated in sinter buildup because of the an extremely rich potential for research. microbial abundance in these extreme environments. The hot springs and geysers of El Tatio, Chile, are Although fluids reach or even exceed the local boil- one of the least known major geothermal systems. The ing point, these environments are colonized by unique features of the El Tatio geothermal field derive diverse microorganisms. from its geographical location (high altitude, 4200 m Evidence of abundant microbial activity appears in a.s.l.; low latitude, 22 8S) that result in high UV-A and all geothermal fields including, for example, in Yel- UV-B radiation (more than three times the radiation lowstone in the U.S.A. (Weed, 1889a,b,c; Walter et received in central USA), a lower water boiling point al., 1972; Walter, 1976a,b; Brock, 1978; Inagaki et al., (86 8C), and the high contents of potentially toxic 2001; Guidry and Chafetz, 2002, 2003), the Taupo elements in the water (As, B, Sb and others). These Volcanic Zone in New Zealand (Jones and Renaut, features make this system particularly interesting in 1996; Renaut et al., 1996; Jones et al., 1997a,b, 1998, order to interpret the many factors that could influence 1999), the Geysir and Krisuvik hot springs in Iceland the origin and diagenetic evolution of hot-spring silica (Schultze-Lam et al., 1995; Konhauser et al., 2001), deposits. the hot springs in Japan (Inagaki et al., 1997), Lake Geological and geochemical research in El Tatio Bogoria in the Kenya Rift (Jones and Renaut, 1996; has focused on the geothermal resources for power Renaut et al., 1998), Laguna Pastos Grandes in Boli- generation (Ambrust et al., 1974; Davidson and Lah- via (Jones and Renaut, 1994), and El Tatio, Chile sen, 1974; Cusicanqui et al., 1976; Lahsen and Tru- (Jones and Renaut, 1997). jillo, 1976; Giggenbach, 1978; Lahsen, 1988). Hot springs and their silica deposits are of interest Although the geothermal exploration dates back to to many disciplines. They have the potential to pro- 1920s (Tocchi, 1923), systematic geothermal research
Excursion métallogénique - Chili 2012 Références page 184 J.L. Fernandez-Turiel et al. / Sedimentary Geology 180 (2005) 125–147 127 was carried out between 1967 and 1982 by the Cor- 2. Study area poracio´n de Fomento de la Produccio´n together with the United Nations Development Program (CORFO/ El Tatio geothermal field is located in the Andes in UNDP). This geothermal field may be the largest northern Chile (Region II), near the Atacama Desert, undeveloped geothermal resource area in the world 100 km east of the town of Calama and the Chuqui- outside a Natural Park. Near El Tatio, at 30 km to the camata copper mine, and 80 km north of San Pedro de east in Bolivia, lies the Sol de Man˜ana geothermal Atacama (22820S,V 68801W)V (Fig. 1). El Tatio and the field, which is in the Eduardo Avaroa National neighbouring Sol de Man˜ana are the world’s highest Reserve. geyser fields. El Tatio is the largest known geyser Recent studies have described the physical char- field in the southern hemisphere, and the third largest acteristics of the geysers of El Tatio (Glennon and field in the world after Yellowstone, USA, and Dolina Pfaff, 2003), the genesis of siliceous oncoids on the Geizerov, Russia. With more than 80 active geysers, it aprons around the geyser vents and hot spring surpasses the number of active geysers in the geother- pools (Jones and Renaut, 1997), and the extreme mal fields of New Zealand and Iceland (Glennon and conditions of these hot springs as an analogue for Pfaff, 2003). El Tatio contains geysers (~ 8% of the the conditions experienced by early life in Precam- world’s geysers), hot springs (some of which are brian times (Phoenix et al., 2003). The aim of our boiling), broad sinter terraces and aprons, and rare study is to describe and interpret the petrographic mud volcanoes. These thermal features are concen- variability and mineralogical features of the sinter trated in three main zones encompassing an area of deposits of El Tatio geothermal field in order to ~10 km2 (Figs. 1 and 2). Other thermal manifesta- obtain an integrated perspective of the primary tions, such as small hot springs, fumaroles, soffioni, deposition and early diagenesis of silica in this and steaming soils, extend over an area of some 30 unique setting. km2 at elevations from 4200 to 4600 m a.s.l. (Zeil,
Fig. 1. Location of El Tatio geothermal field (APVC: Altiplano–Puna Volcanic Complex, after Zandt et al., 2003). The enlarged map shows the northern (A), central (B), and southern (C) geyser and hot spring areas and the general geological setting (based on Lahsen and Trujillo, 1976).
Excursion métallogénique - Chili 2012 Références page 185 128 J.L. Fernandez-Turiel et al. / Sedimentary Geology 180 (2005) 125–147
Fig. 2. General appearance of the El Tatio geothermal field. Left, northern geyser and hot spring area (zone A in Fig. 1). Right, central hot spring area (zone B in Fig. 1).
1959; Healy and Hochstein, 1973; Cusicanqui et al., eous lava domes and flows that are interpreted as 1976; Lahsen and Trujillo, 1976; Glennon and Pfaff, leaks from a large crustal zone of magma generation 2003). underlying the APVC (De Silva, 1989). Models based The subduction of the Nazca oceanic plate beneath on earthquake seismic observations propose the exis- the South American continental plate produced the tence of a sill-like magma body of regional extent that Andean mountain chain. Along this continental mar- is 1–4 km thick and located at mid-crustal (14–17 km) gin, there are three main volcanically active segments depths (Chmielowski et al., 1999; Leidig and Zandt, (Northern, Central and Southern volcanic zones of the 2003; Zandt et al., 2003). Some volcanoes are very Andes), separated by zones with a low angle of sub- active. For example, Lascar volcano erupted in April duction and without magmatism. The Central Volca- 1993 and produced a Plinian column 25,000 m high nic Zone extends from latitude 14 8Sto288S and has (Gonzalez-Ferran, 1995). over fifty recently active volcanoes. In this zone, since The two major geothermal fields, El Tatio in Chile the Late Miocene, a major ignimbrite eruptive centre, and Sol de Man˜ana in Bolivia, and many other similar in scale to Taupo in New Zealand and Toba in geothermal manifestations indicate that the magmatic Sumatra, resulted in a major volcano–tectonic pro- system of the APVC is currently active (Lahsen, vince at a latitude of 21 8Sto248S, known as the 1988; De Silva, 1989). These two geothermal fields Altiplano–Puna Volcanic Complex (APVC). The are located around the Laguna Colorada ignimbrite geothermal field of El Tatio is located in the APVC caldera complex. El Tatio geothermal field lies in the region (Lahsen, 1982, 1988; De Silva, 1989)(Fig. 1). upper levels of a graben ~4 km wide (WNW–ESE) The APVC developed on the thickest part of the and 6 km long (NNE–SSW), filled by 800–2000 m of central Andean crust (~70 km), between the Atacama volcanic rocks. The western flank of the graben is the basin and the Altiplano (Lahsen and Trujillo, 1976; Serrania de Tucle–Loma Lucero horst, while the east- De Silva, 1989). Although effusive andesitic volcan- ern flank is an alignment of andesitic stratovolcanoes ism dominated throughout the Late Tertiary until the and rhyolitic domes that form the summit of the Late Miocene, voluminous and extensive dacitic Andes at this latitude (~5000 m a.s.l.). ignimbrite eruptions dominated volcanism from the The geological sequence observed at El Tatio area Late Miocene to the earliest Pleistocene (10 to 1 Ma). consists of: a) a basement of Middle Jurassic shallow At least 10,000 km3 of ignimbrites were erupted over marine sediments (sandstones and shales), Upper Jur- 50,000 km2 during that period. The ignimbrites are assic–Early Cretaceous meta-andesites and Upper related to ten major eruptive calderas and edifices, Cretaceous sediments (conglomerates, sandstones with evolutionary histories spanning several million and limolites) that crop out on the western margin years (Leidig and Zandt, 2003; Zandt et al., 2003). of El Tatio graben; b) a Late Tertiary to Late Miocene Some are multiple-eruption resurgent complexes as volcanic sequence made up of ignimbrites, andesites large as Yellowstone or Long Valley caldera in USA. and volcanic agglomerates generated in different vol- No major ignimbrites erupted during the Holocene. canic episodes; and c) a Pliocene–Holocene volcanic The Recent volcanism is characterized by large silic- sequence of dacitic and rhyolitic ignimbrites and other
Excursion métallogénique - Chili 2012 Références page 186 J.L. Fernandez-Turiel et al. / Sedimentary Geology 180 (2005) 125–147 129 pyroclastics, lavas and domes (Lahsen and Trujillo, during this westward migration are steam separation 1976). The volcanic activity ends with the andesitic and dilution. Common features of the surface hot stratovolcanoes and rhyolitic domes located in the waters are: high mineralization (total dissolved solids eastern margin of El Tatio graben, which generated can exceed 9.0 g/l TDS); the prevalent sodium chloride the latter materials. Glacial, alluvial and colluvial type fluids; high concentrations of Ca (up to ~300 mg/ deposits overlay this geological sequence (Fig. 1). In l), SiO2 (up to ~220 mg/l), K (up to ~180 mg/l), B (up addition, there are extensive sinter deposits in the hot to ~150 mg/l), As (up to ~30 mg/l), Li (up to ~30 mg/ spring field that cover the youngest glacial sediments l), Cs (up to ~12 mg/l), Mg (up to ~12 mg/l), Br (up to (Lahsen, personal communication). ~6 mg/l), Sr (up to ~4 mg/l), Rb (up to ~2 mg/l), and The El Tatio hot springs discharge between 250 Sb (up to ~2 mg/l); and the relative low concentrations and 500 l of water per second, depending on seasonal of SO4 (up to ~70 mg/l) (Cusicanqui et al., 1976; changes (Lahsen and Trujillo, 1976). According to Lahsen and Trujillo, 1976; Romero et al., 2003; Fer- geochemical and structural data, the water derives nandez-Turiel et al., 2004). Table 1 shows selected from precipitation in an area located at 12–20 km to chemical analyses of these waters. Boron isotope ratios the E–SE of El Tatio (Giggenbach, 1978). The lateral range between À5.9 and À2.5 (x d11B). The main fluid migration is very likely controlled by the perme- fumarole gases are CO2 (99.5% of total gases) and H2S ability of volcanic rocks along the regional slope, (Cusicanqui et al., 1976). while its ascent is constrained by NW–SE and SE– El Tatio is located in the Central Andean Dry Puna SW fractures and especially by the natural barrier of ecoregion, a montane grassland characterized by low the Serrania de Tucle–Loma Lucero horst. precipitation (b100 mm/year). Rainfall is very seaso- Based on the location (Zandt et al., 2003), it is nal, with rains from December to March, and an eight- likely that the water become heated by the Laguna month-long dry season (Van Damme, 2002; Massud, Colorado ignimbrite caldera complex. The principal 2002). The Dry Puna is oligothermic (mean annual producing aquifer has a temperature of 265 8C and lies temperature range: 8 to 11 8C). Daily temperature at a depth of 800 m (Lahsen and Trujillo, 1976). The variation reaches 35 8C. In winter the temperature surface water discharges at a maximum temperature of can fall to À30 8C. Only flora and fauna highly 86 8C, which is the boiling point for an altitude of 4200 adapted to the extreme temperatures and altitudes m. The two main processes affecting the hot waters live in this ecoregion.
Table 1 Water chemistry of El Tatio hot springs (Fernandez-Turiel et al., 2004). All concentrations are expressed in mg/l M1 M2 M3 M4 M5 M6 M7 M8 pH 6.68 6.43 6.41 6.71 7.10 7.64 6.61 6.55 Ca 82.2 254.4 299.6 176.5 196.1 73.2 292.0 294.4 Mg 2.2 6.2 0.2 11.2 10.0 7.5 2.1 0.8 Na 1135.5 3230.7 3680.4 2399.3 2398.3 1063.8 3654.2 3562.7 K 142.3 163.9 184.1 129.8 124.2 73.6 181.3 177.8 Si 33.4 79.5 102.0 86.7 77.0 57.5 78.1 102.3 Cl 2325.9 6566.4 7684.1 4910.6 4960.2 2091.4 7658.8 7663.8
SO4 21.7 59.6 61.8 44.9 64.4 29.7 62.3 60.6 As 8.49 25.12 30.06 18.47 17.59 7.79 28.76 28.93 B 42.60 126.84 147.50 94.51 95.26 38.28 145.16 143.16 Br 1.59 5.01 5.95 3.64 3.69 1.47 5.63 5.61 Cs 2.96 10.19 12.21 6.98 7.47 1.85 11.35 11.68 Li 5.47 14.44 16.78 11.74 11.60 5.34 16.34 16.52 Rb 1.23 1.72 2.26 1.35 1.42 0.50 1.93 2.01 Sb 0.42 1.37 2.08 1.19 1.19 0.26 1.89 2.03 Sr 1.19 3.67 4.41 2.47 2.74 0.80 4.25 4.31 d11B(x) À5.3 À3.1 À5.9 À2.5 À4.0 À2.8 À3.8 À4.6
Excursion métallogénique - Chili 2012 Références page 187 130 J.L. Fernandez-Turiel et al. / Sedimentary Geology 180 (2005) 125–147
3. Methods matic Siemens D-500 X-ray diffractometer (CuKa radiation, 40 kV, 30 mA, graphite monochromator, This paper is based on field observations and the step size of 0.018 from 108 to 408 2h). Qualitative study of sinter samples collected from various spring chemical determinations of sinters were made by X- deposits at El Tatio during the Chilean winter of ray Photoelectron Spectroscopy (XPS) using a PHI 2003. Petrographic examinations included macro- ESCA-5500 instrument. scopic and microscopic studies. The macroscopic features of the sinter samples were established from hand specimens, cut blocks, and large thin sections. 4. Results Samples were embedded in epoxy resin because of their moderate friability. Thin sections were exam- Water reaches the surface in El Tatio as liquid ined using a petrological light microscope fitted with through hot springs (many are permanent pools), and a digital camera. Small fractured sinter samples were also as steam, by means of spectacular geysers or mounted on aluminum stubs using carbon tape, sput- simply through fractures in the soil. The water flows ter-coated for a maximum of 2 min with graphite (2– across the surface of discharge aprons, in some cases 10 nm coating thickness), and then examined with a forming shallow terraces (e.g., south field), and JEOL J-840 scanning electron microscope, equipped through stream channels until it arrives at the Salado with a Link Systems AN 10000 (Oxford Instru- river. The result is a laterally extensive sinter deposit, a ments) Energy Dispersive X-ray Spectroscopic few dm thick, which in places exceeds 1 m. The deposit (EDS) detector for qualitative determination of and related geothermal activity are relatively recent mineral chemistry (beam current: 1.5 nA; working since the sinter covers some moraines of the last glacial distance: 39 mm; accelerating voltage: 15 kV). The period (A. Lahsen, 2004, com. pers.). The main char- mineralogy was determined by X-ray diffraction acteristics of the sinter deposits at El Tatio are sum- (XRD) analysis of pressed powders using an auto- marised in Table 2. This section describes these
Table 2 Main petrographic characteristics of the silica sinter of El Tatio geothermal field Depositional environment Macrostructure (hand specimens) Microtexture (optical microscope/SEM) Hot-springs Laminated spicules and columns Accretion of cyanobacterial mats (Phormidium sp.) Filled fenestral porosity Heterogeneous fill among spicules made up of fine grained sinter detritus, biota fragments (silicified filaments of Phormidium sp, diatom valves, e.g., Synedra sp.), micro-oncoids and coated grains; tufted Phormidium sp. mats are common. Subspherical oncoids Concentric laminae of Phormidium sp. mats. Discharge environments Laminated spicules Accretion of cyanobacterial mats (Phormidium sp., Calothrix (aprons, terraces, sp., and others); associated growth of diatoms (e.g., Synedra channels) sp.). Common occurrence of halite in apron sinter. Filled fenestral porosity Heterogeneous fill among spicules. The accumulation of diatom valves (e.g., Synedra sp.) is common. Detrital volcanic rocks and minerals are frequent. Oncoids of varied shape Concentric laminae of Phormidium sp. mats. Nucleus: detrital sediment (volcanic rocks and minerals, fine grained sinter, laminated sinter). Geyser cones and Coarse lamination with irregular The outer surface is massive with fenestral cavities covered with splash mounds fenestral porosity (vent mouth) Chloroflexus-like filaments. The sinter interior is highly porous. Coarse lamination with palisade-like External part is very porous; interior is massive. Fenestral fenestral porosity (external wall cone) cavities covered with linear ~10 Am width filaments and finer mucilage filaments. Fine lamination (cone basis) Laminated Succession of 1–200 Am thick massive laminae. Irregular spicules and columns building a cupola lamination; silicified organisms are Synechococcus-like (b30 cm in diameter): the splash mound. microbes.
Excursion métallogénique - Chili 2012 Références page 188 J.L. Fernandez-Turiel et al. / Sedimentary Geology 180 (2005) 125–147 131 deposits in detail according to the depositional envir- laminated microtexture mainly consists of cyanobac- onment where they are formed: terial laminae 10–300 Am thick with varying degrees of silicification (Fig. 4B). The porosity of these 1. hot springs levels varies with the intensity of silicification. In 2. discharge environments (aprons, terraces, channels) the least silicified laminae, a dense disordered asso- 3. geyser cones and splash mounds. ciation of individual silicified filamentous sheaths can be recognized (Fig. 4C). The filaments are 4–7 4.1. Hot-spring sinter Am in external diameter and their length rarely exceeds 100 Am. Organic matter is absent, the A white-light grey laminated sinter, with variable sheaths being completely replaced and/or coated by fenestral porosity, occurs in and around the hot spring opal-A. This aspect is common in hot spring sinters mounds, increasing in thickness towards the water (cf. Jones et al., 2001). These porous laminae com- source. Where the local topographic characteristics monly end at the top with a massive silicified area allow the accumulation of water in small pools, a (Fig. 4D). A similar feature has been observed in rim commonly accretes vertically around the pool Icelandic hot spring sinters (cf. Fig. 4 of Konhauser (Fig. 3). The pool springs are commonly gently sur- et al., 2001). In addition, though less common, thin ging and boiling, and can reach several metres in (b10 Am) silica laminae also occur alone (Fig. 4E diameter and depth, but are usually b2 m wide and and F). 5–20 cm deep. White-light grey subspherical oncoids Although the filaments in the laminae are disor- are common inside the pools. The water temperature dered, their morphology is similar to the filaments of ranges between 40 and 60 8C, with the higher tem- the tufted Phormidium sp. mats found at the top of peratures found in the bubbling pools. The sinter spicules and also in the fenestral cavities. This feature macrostructure is irregularly laminated in hand sam- suggests that they are mainly detrital remains of sili- ple. The laminae build up spicules (Fig. 3A and B). cified Phormidium sp. filaments. Phormidium is a The diameter of these structures ranges between 1–10 filamentous and thermophilic cyanobacterium that mm and the height can exceed 5 cm. The terms lacks heterocysts (nitrogen-fixing cells), and has a bspiculesQ and microstromatolites (where laminae thin, gummy sheath. Phormidium forms slimy mats have microbial templates) are used for mm-scale fea- of filaments arranged in tufts, attached to substrates. tures in most publications on sinter (e.g., Jones and The filaments are long, cylindrical, and may be curved Renaut, 1996, 2003). bColumnsQ has been used for by water flow. Phormidium mats are found worldwide features measured in centimetres, and bmoundQ has on wet soils, and in both standing and running waters been used for macroscale features (e.g., vent in geothermal fields (Walter, 1976b; Cady and Farmer, mounds—measured in metres). Although the size 1996). range of these features has not been used consistently, The amount of fenestral porosity existing among they have become fairly standard terms. the spicules and columns can be significant, with open A laminated microtexture is observed using opti- spaces several millimeters in size (Figs. 3A and 5A). cal microscopy (Figs. 3A and 4A). Fine-grained These fenestrae are partially or completely filled with laminae alternate with palisades composed of the geopetal sediment. When the fenestral macrostructure silicified sheaths of cyanobacteria. Isolated indivi- is observed through optical microscopy, it is clear that dual cyanobacterial filaments can be recognized in the geopetal sediment consists of fine-grained sedi- some fine-grained laminae. Tufts of Phormidium sp. ment of detrital siliceous sinter and micro-oncoids located at the tops of the spicule are the best-pre- (Fig. 5A). The micro-oncoids are commonly several served mats (Fig. 5A). It is important to take into hundred micrometers in diameter, but rarely exceed 1 account that the names Phormidium, Calothrix, etc., mm. A cyanobacterial (Phormidium sp.) lamina com- as used here, are really form genera names because monly envelops the larger micro-oncoids (Fig. 5A). In the identifications of cyanobacteria are based on the addition, ooids and tufted Phormidium sp. mats can morphology of the silicified filaments rather than also be observed (Fig. 4A). Fig. 5B shows the accu- other biological criteria (Jones et al., 2001). The mulation of detrital sediment in the bottom of a fenes-
Excursion métallogénique - Chili 2012 Références page 189 132 J.L. Fernandez-Turiel et al. / Sedimentary Geology 180 (2005) 125–147
Fig. 3. An active hot-spring pool together with a subfossil pool are shown in the lower-left side of figure; white rectangles are sites of the samples enlarged in A and B photographs (A and B hand specimens are embedded in epoxy resin). The vent in the bottom of the subfossil pool implies that this was previously a fountain geyser. A. Macrostructure of hot spring sinter showing laminated spicules at the base. B. Macrostructure of a hot-spring pool rim with oncoids (bottom) and laminated spicules (middle and top). tral cavity. The hydrodynamic conditions in such valve. As a result of these divisions, valve size within microenvironments give rise to the development of a given population gradually decreases. Only rarely coated grains (Fig. 5C). Sediment and coated grains does sexual reproduction occur to produce a larger are made up of sinter detritus that incorporates the valve (Round et al., 1992; SSD, 2005). remains of silicified filaments and diatom valves (Fig. Fragments of laminae are separated from laminated 5D,E, and F). Water also flows through small conduits spicules and columns. These fragments are stirred up in in the sediment (Fig. 5E and F). A smooth silica layer the hot spring pools and give rise to oncoids. Oncoids covers all these elements. are white–light grey in color and variously shaped, SEM examinations also reveal silicified tufted although near spherical shapes predominate in the Phormidium sp. mats (Fig. 6A). Preservation is smallest examples. Generally they do not exceed 1 exceptional; the mat can appear to be flowing in the cm in diameter. During the growth of pool rims, water. Diatoms (e.g., Synedra sp.), possibly in a living oncoids may become cemented to the substrate as position, individually adhere to the filaments (Fig. they become covered by new laminated spicules. Fig. 6B). Diatom frustules can form almost all the bottom 3B shows a section of a pool rim with oncoids overlain sediment (Fig. 6C). The valves appear rectangular by undulating laminae. Rim build-up continued with when viewed from the girdle or side view. Synedra laminated spicules and well developed laminated col- sp. is araphid, i.e., the cells lack a raphe structure on umns. Optical microscope and SEM examinations either valve (Round et al., 1992; SSD, 2005). The reveal that these oncoids present the same microtexture valve size differences (e.g., Fig. 6C) are the result of as the laminated spicules and columns, i.e., a succes- asexual reproduction. The original cell divides and sion of laminae of silicified cyanobacterial filaments separates into one old valve and a smaller new (Fig. 6D).
Excursion métallogénique - Chili 2012 Références page 190 J.L. Fernandez-Turiel et al. / Sedimentary Geology 180 (2005) 125–147 133
Fig. 4. Laminated spicules in a hot spring sinter (A: optical microscope thin section; plane polarized light; pores filled with epoxy resin. B–F: scanning electron microscope (SEM) images). A. Alternation of fine-grained lamina (brownish) and palisades of silicified sheaths of cyanobacteria (yellowish). A fenestral cavity is partially shown in the upper-right part. The cavity bottom is filled of ooids. A tufted Phormidium mat grew on the ooids ( P). B. Microtexture of laminated sinter. Note the intense silicification. C. Detail of palisade lamina showing silicified filamentous sheaths of cyanobacteria. D. Massive top of palisade lamina. E. Thin silica laminae. F. Detail of microtexture of thin laminae from E. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)
The hot spring sinters consist mainly of opal-A. several springs and geysers merge. The sheetflow is This phase coexists with opal-CT in some rim sur- intermittent and variable in intensity; in places, small faces (outer part). This feature has been observed also terraces predominate. The sinter is irregularly lami- in some oncoids. Halite is other common phase in nated in both outcrop and in hand specimens, with a surface sinter samples while sylvite (KCl) and realgar similar appearance to the hot-spring laminated sinter. (AsS) are minor phases. They form spicules. Columns, such as those present in hot-spring pool rims, have not been observed. 4.2. Discharge environments (aprons, terraces, Oncoids are very common on such aprons and ter- channels) sinter races and in stream channels. Evaporation strongly affects the aprons. Water temperatures close to 30–35 The water from hot springs overflows a white–light 8C are recorded in terraces and discharge aprons, grey bumpy sinter surface until it forms stream chan- decreasing to b20 8C a few hundred metres from nels. The sinter aprons are continuous and commonly the hot spring vents. Silica deposition practically several hundred metres wide because the deposits of ceases once the water temperature falls to this value.
Excursion métallogénique - Chili 2012 Références page 191 134 J.L. Fernandez-Turiel et al. / Sedimentary Geology 180 (2005) 125–147
Fig. 5. Occluded fenestral porosity between laminated spicules in a hot spring sinter (A: optical microscope thin section; plane polarized light; pores filled with epoxy resin. B–F: SEM images). A. Alternation of fine-grained laminae (brownish) and palisades of silicified sheaths of cyanobacteria (yellowish). Note the Phormidium mats on the spicules ( P). The fenestral cavity is filled with micro-oncoids and fine-grained sediment. A cyanobacterial (Phormidium sp.) lamina envelopes the larger micro-oncoids. B. Fenestral cavity showing the accumulation of sediment in the bottom. C. Coated grain and accumulation of sediment in a fenestral cavity. D. The coated grain showed in C consists of detrital sinter containing abundant diatom frustules (S, Synedra sp.). E–F. Microtexture of detrital sediment filling a fenestral cavity. Note the conduits through the water flowed. Smooth silica covers and cements cyanobacterial filaments (images E–F) and diatoms (image E; S, Synedra sp.). (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)
The similarity with hot-spring laminated sinter is two other cyanobacteria genera have been observed. also evident in microtextures. The main differences One corresponds to Calothrix sp., and the other genus are the abundance of exceptionally well-preserved has not been identified. Calothrix sp. is a filamentous silicified filamentous mats and diatom valves (i.e., blue-green cyanobacteria with heterocysts (nitrogen the delicate nature of silicification) and the common fixing cells). The filaments taper gradually from presence of evaporite minerals, with halite dominant. base to tip (Fig. 7D and E). It is an epilithic or Phormidium sp. mats were recognized with optical epiphytic organism. The silicified filaments, which microscopy both inside the lamination and on the are separate from each other, are hollow and show spicule tops (Fig. 7A and B, respectively). The silici- longitudinal arrangements of opal-A microspheres on fied Phormidium filaments are coarser than in the hot their surfaces (Figs. 7EF and 8A).The filaments of the spring sinters (approx. 10 Am in diameter) (Fig. 7C). undetermined genus are 1 Am thick and their length The silicified wall is approximately 2 Am thick (Fig. does not exceed 30 Am(Figs. 7E and 8A and B). The 7D). In addition to Phormidium, silicified remains of filaments are separated in the mat and can be slightly
Excursion métallogénique - Chili 2012 Références page 192 J.L. Fernandez-Turiel et al. / Sedimentary Geology 180 (2005) 125–147 135
Fig. 6. Fenestral porosity between laminated spicules in a hot spring sinter (A–C: SEM images. D: optical microscope thin section; plane polarized light; pores filled with epoxy resin are in blue and white). A. Tufted Phormidium sp. mat with curved filaments showing evidence of water flow. B. Diatoms adhering to a Phormidium sp. mat. Note the series of girdle bands fitting the valves. C. Accumulation of diatoms in a fenestral cavity. D. Oncoid incorporated in a laminated spicule; both show the same microtexture. A well-preserved tufted Phormidium mat lies on top of the oncoid ( P). (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.) curved (Fig. 7E and B) or present a typical stick Sinter from the discharge environments is made up appearance (Fig. 8A). Accumulations of diatom of opal-A. Opal-CT has been found in the nuclei of valves (e.g., Synedra sp.) are common in the fenestral some oncoids. The other minerals identified are pro- porosity (Fig. 8C and D). Some have lost their girdle ducts of evaporation. Halite is ubiquitous (Fig. 8E bands, allowing separation of the frustules (Fig. 8D), and F), and sylvite (KCl), sassolite (H3BO3) realgar and making it possible to observe their striae. (AsS), and teruggite (Ca4MgAs2B12O22(OH)12d Detrital volcanic rock fragments and mineral 12(H2O) are also common. The presence of teruggite grains (quartz, plagioclase) are important contributors at El Tatio was reported previously by Rodgers et al. to the detrital sediment in the fenestral porosity of (2002). the discharge environments. Euhedral halite crystals and crusts are common in the sinters that are most 4.3. Geyser cone and splash mound sinter exposed to evaporation (Fig. 8E). The cubes are 10 Am or less in size, but can form aggregates that The geyser cones consist of a subconical mass (the exceed 100 Am(Fig. 8F). White and light grey geyser sensu stricto), perpendicular to the surface, oncoids, which occur in all discharge environments, which is surrounded by a gently sloping discharge have various shapes and their size does not exceed apron that can support splash mounds. The height 10 cm (Fig. 9A). The external concentrically-lami- and basal diameter of the cones vary but both can nated structure wraps around nuclei of varied com- reach 3 m. There are several examples in the northern position, including peloids (Fig. 9B), detrital geyser and hot spring field (Figs. 1 and 2). The unlaminated fine-grained sinter and volcanic rocks temperatures in the vent orifices and overflows around and minerals (Fig. 9C), detrital laminated sinter (Fig. geysers (80–85 8C) are close to the local boiling point 9D), and laminated spicules growing up on a basal (86 8C), but they decrease to ~60 8C in the splash platform. Jones and Renaut (1997) described their mound areas. Hand specimens sampled from the vent patterns of growth. mouth, the external wall, and the base of an extinct
Excursion métallogénique - Chili 2012 Références page 193 136 J.L. Fernandez-Turiel et al. / Sedimentary Geology 180 (2005) 125–147
Fig. 7. Discharge apron sinter (A–B: optical microscope thin sections; plane polarized light; pores filled with epoxy resin (white and blue); C–F: SEM images). A. Palisade of silicified sheaths of cyanobacteria (Phormidium sp.: yellowish) between two fine-grained laminae (dark brown). B. A silicified tufted Phormidium mat in a fenestral cavity. A volcanic rock fragment is in upper right corner of the photomicrograph. C. A silicified tufted Phormidium mat and detrital sediment (lower-right part of image) in a fenestral cavity. D. Detail of the silicified tufted Phormidium mat shown in Fig. 7C. E. Silicified remains of Calothrix sp. Filaments taper gradually from base to tip. Silicified filamentous cyanobacterial mat of undetermined genus (background) and diatom valves are also present. F. Detail of the Calothrix sp. filaments shown in Fig. 7E. Note arrangement of opal-A microspheres on the filament surface. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.) geyser cone are all laminated and white to light grey Optical microscopy and SEM examinations of the (Fig. 10). These sinters are commonly termed geyser- top of the vent mouth show a massive sinter with ite (Jones and Renaut, 1997). The style of lamination, fenestral porosity (Fig. 11A–D). The walls of the however, differs in each case. In the vent mouth, a fenestral cavities are covered with curved filaments coarse lamination with irregular fenestral porosity is 10 Am in diameter and 20–50 Am long. The setting, present (Fig. 10A). The examined sample shows a especially the high temperature, and their morphology geopetal structure. The external surface is very hard indicate that they could be silicified filaments of and has the appearance of being smoothed with a microorganisms like Chloroflexus, a thermophilic fila- coarse abrasive. The external wall of the cone is mentous green bacterium found in waters b75 8C also coarse laminated but has a palisade-like fenestral (e.g., in Yellowstone: DIYNP, 2004)(Fig. 11D). porosity (Fig. 10B). The outer surface, on the other Rare silicified structures 5–8 Am in diameter have hand, is very porous. Samples from the exposed base been seen in this uppermost sinter (Fig. 11B) that of the cone are finely laminated (Fig. 10C). could be related to an undetermined organism. The
Excursion métallogénique - Chili 2012 Références page 194 J.L. Fernandez-Turiel et al. / Sedimentary Geology 180 (2005) 125–147 137
Fig. 8. Discharge apron sinter (A–F: SEM images). A. Silicified filamentous mat of undetermined cyanobacterial genus (note the walking stick shape of filaments), and filamentous remains of Calothrix sp (the right filament shows that is hollow). B. Detail of slightly curved silicified filaments of undetermined cyanobacterial genus. C. Diatom valves (Synedra sp.) adhering to a detrital quartz grain. D. Accumulation of diatom valves (Synedra sp.). Note the striae in the frustule located in the central part of the image, and the girdle bands that adjust the two valves in the central-bottom part of the image. E. Euhedral halite hopper-like crystals and crusts. F. Aggregate of euhedral cubic halite crystals. massive character of sinter disappears in the inner porosity clearly defines the limits of the laminae. The areas. Thus, fenestral porosity clearly dominates the lamina thickness ranges between 1 Am and 200 Am. silica mass in the middle (Fig. 11E) and lower (Fig. Its upper and lower surfaces are irregular and its 11F) parts of the hand specimen shown in Fig. 10A. interior is massive. These microtextures resemble those found in the sin- The splash mounds are hemispherical, 20–30 cm in ters of fountain geysers at Yellowstone, which formed diameter and 15 cm in height (Fig. 13A). The cupola on surfaces that were wetted during eruptions but is ~5 cm thick and is made up of laminated columns were subaerially exposed between eruptions (cf. (Fig. 13B). The splash mound microtexture is also Lowe and Braunstein, 2003). In contrast, the external laminated with spicules resembling those observed in wall of the sinter cone is very porous at the surface hot springs (Fig. 13B–D). The main differences are in (Fig. 12A) and relatively massive in the interior. The the silicified microbial community and the maturation fenestral cavities are covered by silicified linear fila- of the silica. The silicified organisms are Synechococ- ments, ~10 Am wide and up to 150 Am long (Fig. cus-like microbes (Fig. 13E). The silicification in the 12C). A network of silicified mucilage is commonly studied samples is more mature than in the hot spring present on the filaments. The microtexture at the base samples. Thus, opal-CT lepispheres b20 Am in dia- of the cone is laminated (Fig. 12D–F). The fenestral meter commonly cover the spicules (Fig. 13F). Opal-
Excursion métallogénique - Chili 2012 Références page 195 138 J.L. Fernandez-Turiel et al. / Sedimentary Geology 180 (2005) 125–147
Fig. 9. Oncoids from the discharge apron sinter (B–D: optical microscope thin sections; plane polarized light; pores filled with epoxy resin (white and blue). A. Variously shaped oncoids showing the different nature of the nucleus (letters indicate the correspondence with the following thin section images). B. Peloidal nucleus of oncoid daT in image A. C. Detrital unlaminated sinter and volcanic rocks and minerals in the nucleus of oncoid dbT in image A. Detrital laminated sinter elements (lower left) enveloped by an external, concentrically laminated structure (upper right) in the oncoid dcT in image A. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)
A constitutes most of the sinter in the geyser cones The El Tatio sinters reinforce the importance of these and splash mounds, but opal-CT is also a major phase factors. Sedimentary macrostructures and microtex- in their external parts. The external surfaces of the tures of sinters at El Tatio, and their primary and mounds contain halite as a minor phase, and traces of early diagenetic mineralogy, preserve the geological sylvite, sassolite (H3BO3), and quartz. and biological features of this surficial geothermal environment exceptionally well. Although hot waters are enriched in silica and 5. Discussion many other components, the geochemical conditions are usually only favorable for opal-A precipitation at Siliceous sinters that form at hot springs and gey- the surface of the geothermal fields (White et al., sers share common features in all geothermal fields, 1956, 1988; Rimstidt and Cole, 1983). The role of independent of the local geographical and environ- other elements in the abiogenic construction of sinter mental conditions. This trend may extend back in time is usually secondary (e.g., coagulant role of salt to the origin of life on the Earth. Thus, sinters may cations: Iler, 1979) or simply nonexistent. The El provide analogues for understanding early life, and for Tatio sinters consist of amorphous silica (opal-A), developing ideas about possible life on other planets deposited around hot springs and geysers from nearly (Bock and Goode, 1996; Treiman, 1998; Farmer, neutral sodium-chloride, silica-saturated water (Cusi- 1999; Konhauser et al., 2001, 2003). The main com- canqui et al., 1976; Fernandez-Turiel et al., 2004). mon features are the favorable geochemical conditions Only minor concentrations of elements other than Si for silica precipitation and the abundance of microbes. and O have been detected in opal-A samples by X-ray
Excursion métallogénique - Chili 2012 Références page 196 J.L. Fernandez-Turiel et al. / Sedimentary Geology 180 (2005) 125–147 139
Fig. 10. Hand specimens sampled in the vent mouth (A), external wall of the cone (B), and base of the cone (C) of an extinct geyser (samples embedded in epoxy resin). Note macrostructural differences between them. photoelectron spectroscopy (XPS). These include Cl, Opal-A deposition takes place under both subaqu- Na, Ca, S, and B (b5% w/w). eous and subaerial settings that are intermittently wet. When silica supersaturation is high, opal-A nano- Fully subaqueous environments are in the pools, ter- spheres originate from polymeric (silica colloid and races and discharge channels. Intermittently wetted gel) deposition, whereas vitreous opal-A surfaces subaerial zones include geyser cones, splash zones develop from monomeric silica precipitation when around geysers and springs, and discharge aprons. A the degree of supersaturation is low (Iler, 1979; Four- temperature drop is crucial for opal-A precipitation, nier, 1985; Campbell et al., 2001; Ohsawa et al., 2002; but the highest silica deposition rates are recorded in Jones and Renaut, 2004; Lynne and Campbell, 2004). areas where the water may evaporate to dryness. The The latter vitreous deposits have not been observed in macrostructural and microtextural differences between the samples studied, possibly because monomeric Si these depositional environments are mainly the result can polymerize to form silica colloids at pH 7 (Yee et of the available framework where opal-A precipitates, al., 2003). On the other hand, the effects of the poly- which is closely related to the microbial community meric deposition are evident. Opal-A nanospheres living there. Aggradation of sinter occurs principally aggregate to form microspheres then settle on suitable at and above the air–water interface. Water ascends substrates such as the microbial filaments observed in through the previously deposited porous sinter by Figs. 4C 7DF 11D and 12C. The potential growth of meniscoid and capillary action generating thin water microspheres is limited. Thus, the precipitation of films. Spray and splash are significant processes in opal-A continues with the adhesion of microspheres. maintaining moist surfaces. Condensation of steam This process rapidly decreases porosity and may oblit- could be another way to form a thin water film. erate the previous textures. Finally, cooling and evaporation prompt silicification.
Excursion métallogénique - Chili 2012 Références page 197 140 J.L. Fernandez-Turiel et al. / Sedimentary Geology 180 (2005) 125–147
Fig. 11. Sinter microtextures in a geyser vent mouth (A and E–F: optical microscope thin sections; plane polarized light; pores filled with epoxy resin. B–D: SEM images). A. Massive silica sinter and fenestral porosity in the outer surface. B. Silicified undetermined structures in sinter from the external surface. C. Massive silica sinter and fenestral porosity showing silicified filaments; white rectangle is enlarged in D. D. Detail of silicified filaments of Chloroflexus-like microbes. E. Highly porous sinter from middle part of the hand specimen shown in Fig. 10A. F. Highly porous sinter from the lower part of the hand specimen shown in Fig. 10A.
Cations other than silicon are not required to build bell, 2004), and El Tatio is no exception. A complex up the sinter. Some, however, are essential for the community of organisms is adapted to hot water at El microbial life that the sinter substrate supports. In Tatio, including vertebrates (e.g., amphibians; Mendez contrast, it is important to point out the high concen- et al., 2004), invertebrates, plants, and microbes. How- trations of some elements that considered potentially ever, only silicified microbial structures are preserved toxic for many organisms are present in the El Tatio in the studied sinter samples. The presence of such thermal waters. These include B (up to ~150 mg/l), As microbes is evident in the field by the patchy colora- (up to ~30 mg/l), and Sb (up to ~2 mg/l) (Table 1). tion of sinter surfaces and, especially, subaqueous Although it is difficult to ascribe specific sinter substrates. Oranges and browns predominate; red is features to abiogenic or biogenic processes, the scarce. These colours are linked to carotenoids, microbe fingerprints in microstructures and microtex- which are orange, yellow or red pigments associated tures in sinters from geothermal fields around the with vitamin A (Brock, 1994). Carotenoids protect world are evident (Inagaki et al., 2001; Konhauser et the cells from the high UV radiation existing at El al., 2001; Guidry and Chafetz, 2003; Jones and Tatio. However, it seems that high UV radiation is an Renaut, 2003; Smith et al., 2003; Lynne and Camp- incidental factor on microbial development at El
Excursion métallogénique - Chili 2012 Références page 198 J.L. Fernandez-Turiel et al. / Sedimentary Geology 180 (2005) 125–147 141
Fig. 12. Sinter microtextures in the external wall and base of a fossil geyser cone (A–B and D: optical microscope thin sections; plane polarized light; pores filled with epoxy resin. C and E–F: SEM images). A. Highly porous sinter in the outer surface of the external wall of the geyser cone. B. Silica sinter and fenestral porosity in the interior of the external wall cone. C. Silicified filamentous moulds and thin mucilage strands covering the fenestral cavities in the interior of the external wall of the cone. D. Laminated microtexture in the geyser cone base. E. Laminae of variable thickness in the geyser cone base; note the massive nature of the inner part. F. Detail of thinnest laminae.
Tatio because the observed microbial communities has been established for other geothermal fields are similar to those present in other geothermal fields (Walter, 1976a,b; Pentecost, 2003). Furthermore, exposed to lower UV radiation. The bluish colour of the temperature gradient is closely related to the the water is due to the colloidal silica (cf. Ohsawa et depositional setting. The full temperature range com- al., 2002). prises is from 86 8C at the hottest vents to 20 8C, Although some organisms may not have left their where opal-A precipitation ceases. Over this broad fingerprints in the sinter or could have been oblit- range it is possible to distinguish several different erated by both the continuous deposition of silica settings: and the early diagenetic maturation of opal-A, pet- rographic observations provide a strong basis for a) Water temperature between 70–75 and 86 8C. This suggesting that the microbial community is moder- corresponds to the geyser setting, where filaments ately diverse at El Tatio. The microbial community of Chloroflexus-like green bacteria have been is dominated by three major groups of organisms: observed (Fig. 11D). These green bacteria are cyanobacteria, green bacteria, and diatoms. The opti- usually found in Yellowstone waters below 75 8C mal temperature range supporting their growth con- (DIYNP, 2004), although the upper limit of growth stitutes the most important ecological parameter, as was formerly extended to 90 8C(Walter, 1976b).
Excursion métallogénique - Chili 2012 Références page 199 142 J.L. Fernandez-Turiel et al. / Sedimentary Geology 180 (2005) 125–147
Fig. 13. Splash mound sinter (C–F: SEM images). A. Field aspect of splash mounds around an extinct geyser. B. Hand specimen showing laminated spicules and columns. C. Plan view of a group of spicules. D. Detail of a spicule surface. E. Silicified Synechococcus-like microbes in fenestral porosity between spicules. F. Detail of the opal-CT lepispheres on a spicules surface. A microsphere in the lower part (indicated as dhT) shows a hole indicative of the early transformation of opal-A to opal-CT.
Nonphotosynthetic, hyperthermophilic bacteria are bacterium Phormidium and diatoms (e.g., Synedra common where the temperature exceeds 70–73 8C sp.) (Figs. 3A 4A 5A–C). (Brock, 1978, 1994). d) Water temperature from 20 to 40 8C. This setting b) Water temperature from 60 to 70–75 8C. These coincides with the discharge environments (aprons, temperatures are present in the splash areas around terraces, and channels). The microbial assemblages geysers, where silicified Synechococcus-like cya- are complex, consisting mainly of cyanobacterial nobacteria are found (Fig. 13E). Cyanobacteria are mats of Phormidium and Calothrix, but diatoms restricted to b70–73 8C, the upper threshold for (Synedra sp.) are also abundant (Figs. 6A–F and photosynthesis (Brock, 1978, 1994). 7A,C,D). c) Water temperature between 40 and 60 8C. The waters of the hot springs have this temperature. The alternation in dominance by filamentous cya- The abundance of microbial remains, compared to nobacteria or diatoms observed in some hot-spring the higher temperature settings, is notable. The most settings (Brock, 1978; Vinson and Rushforth, 1989; characteristic organisms are the filamentous cyano- Bonny and Jones, 2003) has not been observed at El
Excursion métallogénique - Chili 2012 Références page 200 J.L. Fernandez-Turiel et al. / Sedimentary Geology 180 (2005) 125–147 143
Tatio. Thus, both types of organisms coexist along Most modern sinters consist of opal-A (e.g., those flow paths where the temperature is compatible for forming today throughout Taupo Volcanic Zone: Ora- their growth. kei Korako, Smith et al., 2003; Whakarewarewa, Jones Rapid silicification is necessary to preserve the and Renaut, 2004). In sub-recent sinters opal-C pre- microbe moulds. Thus, for example, the cyanobacter- dominates (e.g., Umukuri in Taupo, New Zealand, ial tissue can be lost to decay in 10–12 days or less Campbell et al., 2001; Lynne and Campbell, 2004). (Bartley, 1996). Afterwards, continued opal-A preci- Finally, in ancient sinters (cherts) quartz is the predo- pitation can help in preservation. However, there is a minant phase, although opal-C and opal-CT, and rarely point, not fixed in time, when the silicification could opal-A, may persist in pre-Cenozoic rocks (e.g., in start to obliterate the microbial fingerprint. Silicified Ordovician rocks at Sarrabus in Sardinia, Italy microbial fingerprints are common in modern and [Gimeno, 1989], and at Palazuelo de las Cuevas in recent sinters. However, the mechanisms that control NW Spain [Ferna´ndez-Turiel et al., 1993]). colloidal silica–microbe interactions are poorly under- Amorphous opal-A predominates in the sinters at El stood. It is controversial whether or not microbes Tatio, whereas opal-CT is present in very specific mediate opal-A precipitation, favoring nucleation on places. More ordered forms such as opal-C and quartz microbial surfaces (Rimstidt and Cole, 1983; have not been recognized. The quartz determined by Schultze-Lam et al., 1995; Ferris et al., 1986, 1988; XRD has been always detrital. Microtextural evidence Jones and Renaut, 1996; Renaut et al., 1996; Jones et of silica maturation is similar to that inferred by Lynne al., 1997a,b), or whether precipitation is due to the and Campbell (2004). The first step consists of the cooling and evaporation of the supersaturated waters formation of holes in opal-A microspheres (Fig. 13F), alone (Lowe and Braunstein, 2003). after which hexagonal platelets emerge. Ultimately, the Recent experiments carried out with cyanobacteria clustering and packing of silica platelets results in the indicate that silica precipitation is largely abiogenic formation of opal-CT lepispheres (Fig. 13F) (Florke et and cyanobacterial surfaces have a negligible effect on al., 1991; Graetsch, 1994). The duration for the opal-A silica nucleation (Yee et al., 2003; Benning et al., to opal-CT transition has been estimated at ~10,000 2004). The microorganisms supply passive surfaces years (Herdianita et al., 2000). However, fumarolic where opal-A precipitation simply occurs as a result activity or weathering can accelerate silica phase tran- of inorganic influences. Thus, although microbial sur- sitions (Lynne and Campbell, 2004). Fumarolic activ- faces do not directly nucleate silica mineral formation, ity may explain the presence of opal-CT in outer areas they may play an important role in the aggregation of of geyser cones and in the surrounding splash mounds, polymeric silica and the deposition of silica colloids. as well as at the surface of some pool rims. Steam Once deposited, the sinter is exposed to a myriad of discharge increases the amount of surface fluids in the factors (fumarolic activity, changing pH, changing deposit, and facilitates the kinetics of localized disso- temperature, water level variations, etc.) that can lution and reprecipitation of silica. alter the primary texture and composition (Lynne and Campbell, 2004). These factors control the early diagenesis of the deposit, i.e., the sinter maturation. 6. Conclusions Sinter maturation introduces textural changes by silica phase transformation (Herdianita et al., 2000; Camp- El Tatio sinters consist of amorphous silica (opal-A) bell et al., 2001; Lynne and Campbell, 2004): opal-A deposited around hot springs and geysers from nearly to opal-CT to opal-C to quartz. The silica transforma- neutral sodium-chloride, silica-saturated water. Only tions are the result of water loss (both water in the minor concentrations of Cl, Na, Ca, S, and B (b5% mineral lattice during maturation and connate pore w/w) have been observed in such opal-A. Water cool- water), repeated solution–precipitation, replacement ing and evaporation to dryness are the main processes and recrystallization (Scurfield and Segnit, 1984; Wil- controlling the opal-A deposition, which takes place in liams et al., 1985; Herdianita et al., 2000; Campbell et both subaqueous and subaerial settings, and is closely al., 2001). This transformation is progressive and related to the complex microbial communities adapted occurs during silica ageing (Herdianita et al., 2000). to the hot water. All fingerprints of organisms observed
Excursion métallogénique - Chili 2012 Références page 201 144 J.L. Fernandez-Turiel et al. / Sedimentary Geology 180 (2005) 125–147 in the studied sinter samples correspond to microbes Chile. IAVCEI, International Symposium on Volcanism. San- and suggest that the microbial community is moder- tiago, Chile, pp. 529–541. Bartley, J.K., 1996. Actualistic taphonomy of cyanobacteria: impli- ately diverse (cyanobacteria, green bacteria, and dia- cations for the Precambrian fossil record. Palaios 11, 571–586. toms). The temperature gradient is the most important Benning, L.G., Phoenix, V.R., Yee, N., Konhauser, K.O., 2004. The ecological parameter, and is closely related to the dynamics of cyanobacterial silicification: an infrared micro-spectro- depositional setting. The temperature range of silica scopic investigation. Geochim. Cosmochim. Acta 68, 743–757. precipitation is from 86 8C (water boiling point at El Bock, G.R., Goode, J.A. (Eds.), 1996. Evolution of Hydrother- mal Ecosystems on Earth (and Mars?), Proceedings of the Tatio, 4200 m.a.s.l.) to 20 8C where opal-A precipita- CIBA Foundation Symposium, vol. 202. Wiley, Chichester, tion ceases. U.K. 334 pp. The main features observed in the sinters of El Tatio Bonny, S., Jones, B., 2003. Microbes and mineral precipitation, are: 1) Geyser setting: water temperature from 70–75 Miette Hot Springs, Jasper National Park, Alberta, Canada. Can. to 86 8C; coarse laminated sinter with short spatial J. Earth Sci. 40, 1483–1500. Brock, T.D., 1978. The habitats. In: Brock, T.D. (Ed.), Thermophilic variations; biota comprises nonphotosynthetic, hy- Microorganisms and Life at High Temperatures. Springer-Verlag, perthermophilic bacteria. 2) Splash areas around gey- New York, USA, pp. 12–38. sers: water temperature from 60 to 70–75 8C; Brock, T.D., 1994. Life at High Temperatures. http://www.bact. laminated spicules and columns building small cupo- wisc.edu/bact303/b1, Search: January, 10th, 2005. las; Synechococcus-like cyanobacteria are predomi- Cady, S.L., Farmer, J.D., 1996. Fossilization processes in siliceous thermal springs: trends in preservation along thermal gradients. nant. 3) Hot spring setting: water temperature In: Bock, G.R., Goode, J.A. (Eds.), Evolution of Hydrothermal between 40 and 60 8C; laminated spicules and col- Ecosystems on Earth (and Mars?), Proceedings of the CIBA umns, and subspherical oncoids; filamentous cyano- Foundation Symposium, vol. 202. Wiley, Chichester, U.K., bacteria (Phormidium) and diatoms (e.g., Synedra) are pp. 150–173. the most characteristic microbes. 4) Discharge envir- Campbell, K.A., Sannazzaro, K., Rodgers, K.A., Herdianita, N.R., Browne, P.R.L., 2001. Sedimentary facies and mineralogy of the onments: water temperature between 20 and 40 8C; Late Pleistocene Umukuri silica sinter, Taupo Volcanic Zone, laminated spicules and oncoids of variable shape; New Zealand. J. Sediment. Res. 71, 727–746. cyanobacterial mats of Phormidium and Calothrix Chmielowski, J., Zandt, G., Haberland, C., 1999. The central and diatoms (e.g., Synedra sp.) are very abundant. Andean Altiplano–Puna magma body. Geophys. Res. Lett. 26, In summary, the hot springs and geysers of El Tatio 783–786. Cusicanqui, H., Mahon, W.A.J., Ellis, A.J., 1976. The Geochemistry are a natural laboratory of extraordinary interest of the El Tatio Geothermal Field, Northern Chile, Proceedings, because their sedimentary macrostructures and micro- Second United Nations Symposium on the Development and Use textures reflect exceptionally the geological and bio- of Geothermal Resources, San Francisco, May 1975, vol. 1. U.S. logical processes involved in the primary deposition Government Printing Office, Washington, D.C., pp. 703–711. and early diagenesis of siliceous sinters. Lawrence Berkeley Laboratory, University of California. Davidson, J., Lahsen, A., 1974. Antofagasta–El Tatio–Laco. Guide book-Excursion A-2. IAVCEI, International Symposium on Volcanism. Santiago, Chile. 61 pp. Acknowledgements de Silva, S.L., 1989. Altiplano–Puna Volcanic Complex of the central Andes. Geology 17, 1102–1106. We express our acknowledgement to J. Illa for DIYNP, 2004. Resources and issues 2004. http://www.nps.gov/yell/ publications/pdfs/handbook/, [Search: January, 10th, 2005]. Di- sample preparation for petrography, and thank the vision of Interpretation, Yellowstone National Park, Yellowstone. technical support of the Scientific–Technical Service Eugster, H.P., Chou, I.M., 1973. Depositional environments of Pre- Unit of the University of Barcelona. We also thank Dr cambrian banded iron-formations. Econ. Geol. 68, 1144–1168. R.W. Renaut for his very helpful comments on the Farmer, J., 1999. Taphonomic modes in microbial fossilization. original manuscript. Commission on Physical Sciences, Mathematics, and Applica- tions (CPSMA) and Space Studies Board (SSB), Size Limits of Very Small Microorganisms. The National Academies Press, pp. 94–102. References Ferna´ndez-Turiel, J.L., Gimeno, D., Lo´pez Soler, A., Querol, X., 1993. Las mineralizaciones fosfa´ticas de los materiales paleo- Ambrust, G., Arias, J., Lahsen, A., Trujillo, P., 1974. Geochemistry zoicos de la provincia de Zamora. Anuario Instituto de Estudios of the hydrothermal alteration at the El Tatio geothermal field, Zamoranos Floria´n de Ocampo, 1992, pp. 463–506 (in Spanish).
Excursion métallogénique - Chili 2012 Références page 202