El Salvador, Chile Porphyry Copper Deposit Revisited: Geologic and Geochronologic Framework

Home , El Salvador mine, Escondida, Zona Sur

See discussions, stats, and author profiles for this publication at: https://www.researchgate.net/publication/249057232

Article in International Geology Review · January 1997 DOI: 10.1080/00206819709465258

CITATIONS READS 62 763

6 authors, including:

Richard Tosdal Constantino Mpodozis Independent University of Chile

138 PUBLICATIONS 3,988 CITATIONS 144 PUBLICATIONS 8,137 CITATIONS

SEE PROFILE SEE PROFILE

Andrew James Tomlinson Christopher Mark Fanning El Servicio Nacional de Geología y Minería Australian National University

32 PUBLICATIONS 821 CITATIONS 573 PUBLICATIONS 25,836 CITATIONS

SEE PROFILE SEE PROFILE

Some of the authors of this publication are also working on these related projects:

Tectonic evolution of the Andes View project

Cartographie MCF View project

All content following this page was uploaded by Richard Tosdal on 29 May 2014.

The user has requested enhancement of the downloaded file. This article was downloaded by: [Canadian Research Knowledge Network] On: 28 May 2011 Access details: Access Details: [subscription number 932223628] Publisher Taylor & Francis Informa Ltd Registered in England and Wales Registered Number: 1072954 Registered office: Mortimer House, 37- 41 Mortimer Street, London W1T 3JH, UK

International Geology Review Publication details, including instructions for authors and subscription information: http://www.informaworld.com/smpp/title~content=t902953900

El Salvador, Chile Porphyry Copper Deposit Revisited: Geologic and Geochronologic Framework Paula Cornejoa; Richard M. Tosdalb; Constantino Mpodozisc; Andrew J. Tomlinsonc; Orlando Riverac; C. Mark Fanningd a Servicio National de Geologoia y Mineria, Santiago, Chile b U.S. Geological Survey, Menlo Park, California c Servicio National de Geologia y Mineria, Santiago, Chile d Australian National University, Canberra, Australia

Online publication date: 06 July 2010

To cite this Article Cornejo, Paula , Tosdal, Richard M. , Mpodozis, Constantino , Tomlinson, Andrew J. , Rivera, Orlando and Fanning, C. Mark(1997) 'El Salvador, Chile Porphyry Copper Deposit Revisited: Geologic and Geochronologic Framework', International Geology Review, 39: 1, 22 — 54 To link to this Article: DOI: 10.1080/00206819709465258 URL: http://dx.doi.org/10.1080/00206819709465258

PLEASE SCROLL DOWN FOR ARTICLE

Full terms and conditions of use: http://www.informaworld.com/terms-and-conditions-of-access.pdf

This article may be used for research, teaching and private study purposes. Any substantial or systematic reproduction, re-distribution, re-selling, loan or sub-licensing, systematic supply or distribution in any form to anyone is expressly forbidden.

The publisher does not give any warranty express or implied or make any representation that the contents will be complete or accurate or up to date. The accuracy of any instructions, formulae and drug doses should be independently verified with primary sources. The publisher shall not be liable for any loss, actions, claims, proceedings, demand or costs or damages whatsoever or howsoever caused arising directly or indirectly in connection with or arising out of the use of this material. International Geology Review, Vol. 39, 1997, p. 22-54. Copyright © 1997 by V. H. Winston & Son, Inc. All rights reserved.

El Salvador, Chile Porphyry Copper Deposit Revisited: Geologic and Geochronologic Framework1

PAULA CORNEJO, Servicio National de Geologoía y Minería, Avenida Santa María 0104, Santiago, Chile

RICHARD M. TOSDAL, U.S. Geological Survey, 345 Middlefield Road, Menlo Park, California 94025

CONSTANTINO MPODOZIS, ANDREW J. TOMLINSON, ORLANDO RIVERA,2 Servicio National de Geología y Minería, Avenida Santa María 0104, Santiago, Chile

AND C. MARK FANNING Australian National University, Canberra, Australia

Abstract

The Eocene (42 to 41 Ma) El Salvador porphyry copper deposit in the Indio Muerto district, northern Chile (26° 15' S Lat.), formerly thought to have formed at the culmination of a 9-m.y. period of episodic magmatism, is shown by new mapping, U-Pb and K-Ar geochronology, and petrologic data to have formed during the younger of two distinct but superposed magmatic events—a Paleocene (~63 to 58 Ma) and an Eocene (44 to 41 Ma) event. In the district, high-K Paleocene volcano-plutonic activity was characterized by a variety of eruptive styles and magmatic compositions, including a collapse caldera associated with explosive rhyolitic magma tism (El Salvador trap-door caldera), a post-collapse rhyolite dome field (Cerro Indio Muerto), and andesitic-trachyandesitic stratovolcanos (Kilometro Catorce-Los Amarillos sequence). Pre- caldera basement faults were reactivated during Paleocene volcanism as part of the collapse margin of the caldera. Beneath Cerro Indio Muerto, where the porphyry Cu deposit subsequently formed, the intersection of two major basement faults and the NNE-striking rotational axis of tilted ignimbrites of the Paleocene El Salvador caldera localized emplacement of post-collapse rhyolite domes and peripheral dikes and sills. Subsequent Eocene rhyolitic and granodioritic- dacitic porphyries intruded ~14 m.y. after cessation of Paleocene magmatism along the same NNE-striking structural belt through Cerro Indio Muerto as did the post-collapse Paleocene rhyolite domes. Eocene plutonism over a 3-m.y. period was contemporaneous with NW- SE-directed shortening associated with regional sinistral transpression along the Sierra Castillo fault, lying ~10 km to the east. Older Eocene rhyolitic porphyries in the Indio Muerto district were emplaced between 44 and 43 Ma, and have a small uneconomic Cu center associated with a porphyry at Old Camp. The oldest granodioritic-dacitic porphyries also were emplaced at ~44 to Downloaded By: [Canadian Research Knowledge Network] At: 21:38 28 May 2011 43 Ma, but their petrogenetic relation to the rhyolitic porphyries and younger granodioritic- dacitic porphyries in the district is unclear. The main porphyry Cu-Mo-related granodioritic- dacitic stocks in Quebrada Turquesa on Cerro Indio Muerto intruded, cooled, and were mineralized within ~1 m.y. between 42 and 41 Ma. Volumetrically minor late- to post-mineral porphyries are slightly more mafic than earlier granodioritic-dacitic porphyries, a compositional trend possibly repeated on several scales and more than once over the 3-million-year Eocene magmatic history of the Indio Muerto district. This compositional trend requires either addition of basaltic material into an open-system silicic magma chamber or tapping of progres sively deeper levels of a vertically zoned magma chamber. Eocene porphyry magmas were more hydrous and their residual source mineralogy richer in garnet than the relatively anhy drous Paleocene rocks, whose source was rich in pyroxene. The presence of inherited

1This paper is one of a series of contributions (Marcos Zentilli, compiler) to Project No. 342, Ages and Isotopes of South American Ores, of the International Geological Correlation Program. 2Present address: Pasaje Volcán San Francisco 3851, Peñalolén, Santiago, Chile.

0020-6814/97/240/22-33 $10.00 22 EL SALVADOR COPPER DEPOSIT REVISITED 23

zircons in Paleocene and Eocene rocks requires interaction with crustal rocks of Paleozoic and/ or Proterozoic age. Paleocene and Eocene igneous rocks in the Indio Muerto district were emplaced during distinct magmatic-tectonic events that are unrelated, although spatially associated. The district- scale Paleocene and Eocene eruptive styles and geochemical and mineralogic characteristics mimic characteristics of similar-aged igneous rocks throughout northern Chile (20°30' S Lat. to 27° S Lat.), attesting to the regional nature of the Paleocene and Eocene events. Porphyry Cu mineralization in the district furthermore is associated not only with an Eocene granodioritic- dacitic (42 to 41 Ma) complex, but also with one of an older Eocene (44 to 43 Ma) rhyolitic porphyry, implying that a long period of precursor magmatism is not required for generation of the El Salvador porphyry Cu-Mo deposit. Rather, the episodic magmatism preceding porphyry Cu mineralization reflects repeated structural localization through time of superimposed high- level volcano-plutonic complexes in an active magmatic arc.

Introduction PORPHYRY COPPER SYSTEMS are frequently emplaced late in the magmatic history of vol cano-plutonic terranes, and many are believed to have formed in the roots of a coeval volcano (Titley and Beane, 1981; Sillitoe, 1973, 1988; Hedenquist and Lowenstern, 1994). This magmato-hydrothermal association has led to the obvious question of whether protracted precur sor magmatism, including volcanism, might play some role in the formation of porphyry copper systems (Titley, 1982; McCandless and Ruiz, 1993). The El Salvador copper deposit in the Indio Muerto district of northern Chile (Fig. 1), described in the seminal paper by Gustafson and Hunt (1975), is cited as one example where protracted magmatism preceded porphyry copper deposition (e.g., Titley and Beane, 1981; Sillitoe, 1988). Gustafson and Hunt (1975) concluded, on the basis of Rb-Sr and K-Ar geochronology, that the porphyry Cu- Mo deposit formed as the "culmination" of a

Downloaded By: [Canadian Research Knowledge Network] At: 21:38 28 May 2011 ~9-m.y. period of episodic volcanic and subvolcanic activity (Fig. 2), which began in the early Eocene at ~49 to 50 Ma with eruption of rhyolitic domes, was succeeded by the emplace FIG. 1. Distribution of porphyry Cu-Mo deposits and ment at ~45 Ma of subvolcanic quartz-feldspar prospects along the Domeyko fault system in the Andean phyric stocks accompanied by minor porphyry Precordillera, northern Chile. Ages of deposits and pros Cu-Mo mineralization, and peaked with pects are shown parenthetically. emplacement of granodioritic stocks and the main porphyry Cu-Mo deposit between 42 and on a regional scale between 26° and 27° S Lat. 41 Ma (Rb-Sr and K-Ar ages from Gustafson (area of Fig. 3), at least two separate magmatic and Hunt [1975] quoted herein are recalculated episodes having no genetic relation are present: to modern decay constants and quoted to 2a (1) Paleocene (63 to 55 Ma) high-K calcalkaline uncertainties). magmatism represented by extensive volcano- Subsequent geologic and K-Ar geo- genie products including collapse calderas and chronologic studies in the vicinity of El Sal their ignimbrites, rhyolitic dome fields, tra- vador (Cornejo et al., 1993a) demonstrated that chybasalts, and trachyandesites;and (2) middle 24 CORNEJO ET AL.

FIG. 2. Diagram comparing the stratigraphy of Gustafson and Hunt (1975) with that concluded herein. At the time of Gustafson and Hunt's (1975) study, geochronologic data for the rhyolitic ignimbrites and older rocks were unavailable; instead their age assignments were based on correlations with formations in the Copiapó area. The position of the Indio Muerto unconformity in the left column follows that proposed by Clark et al. (1985). See text for discussion. Our quartz rhyolite porphyry unit corresponds to the "quartz porphyries" and Cerro Pelado "quartz rhyolite" stock of Gustafson and Hunt (1975).

Eocene to early Oligocene (49 to 31 Ma) moder The conflicting stratigraphic and chrono ate- to high-K, calcalkaline, intermediate to logic conclusions from the regional studies of silicic magmatism represented by small to mod Cornejo et al. (1993a) and the district study of erately sized, shallow-level stocks and plutons, Gustafson and Hunt (1975) raise questions and only scarce volcanogenic products. This regarding the magmatic history preceding for division was based upon three principal lines of mation of the El Salvador porphyry Cu-Mo evidence: (1) K-Ar geochronology indicating deposit. Was there early Eocene (~45 and ~50 that the two episodes are discrete in time and Ma) volcanic activity in the Indio Muerto dis separated by a 5- to 6-m.y. gap in magmatism; trict? Or could the ~45- and ~50-Ma magmatic (2) petrologic data indicating that they are events be the result of superposition of a different magmatic suites; and (3) different younger (~43- to 41-Ma) magmato-hydrother volcano-magmatogenic and regional structural mal event upon an older early Tertiary (>55 settings of the two suites. The rhyolitic domes Ma) volcanic event? To address this question (and their related volcanic rocks) hosting the El and to provide better constraints on geologic Salvador deposit in the Indio Muerto district conditions conducive for a porphyry Cu-Mo were included in the Paleocene regional vol deposit, the Indio Muerto district was Downloaded By: [Canadian Research Knowledge Network] At: 21:38 28 May 2011 canic unit because of their petrologic similarity remapped and critical units were dated using to Paleocene rhyolitic rocks elsewhere in the the K-Ar and U-Pb methods (Tables 1 and 3, region (Fig. 2). Cornejo et al. (1993a) viewed respectively), including U-Pb spot analyses of the previously reported 49- to 50-Ma Rb-Sr ages individual zircon grains using the sensitive for the rhyolitic rocks (Gustafson and Hunt, high-resolution ion-microprobe (SHRIMP) 1975) as minimum ages resulting from the (Table 4). All K-Ar ages reported herein are thermal effects during emplacement of the listed in Table 1 unless otherwise noted. younger Eocene porphyry complex and associ Recalculated K-Ar ages of Gustafson and Hunt ated hydrothermal alteration system, suggest (1975) are listed in Table 2. Previously pub ing, based on the regional data, that within the lished K-Ar ages (Tobar, 1977; Quirt, 1972; Indio Muerto district a magmatic gap spanning Zentilli, 1974) and unpublished 40Ar/39Ar geo a minimum of 12 m.y. separated the rhyolitic chronology (M. McWilliams, pers. commun. to volcanism (≥ 55 Ma) from the 43- to 41-Ma CODELCO [Corporatión Nacional del porphyry copper complex (Cornejo et al., Cobre-Chile], 1994, 1996) provide additional 1993a; Mpodozis et al., 1994). chronologic constraints. EL SALVADOR COPPER DEPOSIT REVISITED 25 Downloaded By: [Canadian Research Knowledge Network] At: 21:38 28 May 2011

FIG. 3. Regional geologic map of the El Salvador-Potrerillos area, northern Chile. Modified from Comejo et al. (1993a), Tomlinson et al. (1993), and Cornejo and Mpodozis (1996). Abbreviations: SCB = Sierra Castillo Batholith. Downloaded By: [Canadian Research Knowledge Network] At: 21:38 28 May 2011

26 CORNEJO ET AL. 1 TABLE 1. K-Ar Ages from Volcanic and Intrusive Rocks in the El Salvador Area, Northern Chile

Sample UTM Coordinates(N/E) Ar rad, nl/g %Ar, atm. Age, Ma number Location Lithology Material dated %K (2a error)

Llanta Formation (Upper Cretaceous) SB-4722 Cerro El Pingo 7043,425/440,825 Andesitic breccia Hornblende 0.556 1.600 46 73 ±4 SR-3722 Pampa del Carrizo 7123,450/434,650 Dacitic lava Biotite 6.866 19.315 15 71 ±2 EC-3262 Qda. Inés Chica 7111,625/443,800 Pyroxene andesitic lava Whole rock 0.592 1.622 28 69 ±3

Post-Llanta Formation intrus ions (lower Paleocene) SR-4562 Quebrada del Salado 7026,000/486,800 Monzodiorite Biotite 7.504 18.875 23 64±2 SR-2452 Sierra Miranda 7105,900/440,225 Quartz diorite Whole rock 1.641 4.156 38 64±2 10-139 Cuesta San Juan 7093,350/447,530 Pyroxene-hornblende Whole rock 1.282 3.178 75 63 ±4 diorite IP-68 Portal del Inca 7094,460/441,340 Dioritic porphyry Whole rock 0.499 1.178 82 59.7 ± 5.7

El Salvador caldera ignimbrites (Paleocene) Cerros Contreras-La Antena sequence, intracatdera facies ignimbrites 10-142 North base of Cerro 7090,860/446,330 Sanidine-biotite rhyolitic Biotite 5.815 15.257 14 66±2 Contreras tuff IO-81 East of Cerro Contreras 7089,660/447,130 Sanidine-biotite rhyolitic Biotite 6.224 15.416 17 63 ±2 tuff SR-51 NE of Cerro Contreras 7091,900/447,525 Biotite welded tuff Biotite 6.443 15.413 44 61 ±2

Cerro Indio Muerto ignimbrites IP-19 NE flank of Cerro Indio 7096,710/445,560 Rhyolite welded tuff Biotite 6.537 15.351 23 59.4 ± 1.5 Muerto

Cerro El Buitre, Potrerillos area, outflow facies ignimbrites ST-77 East flank of Cerro El 7070,600/466,250 Sanidine-biotite rhyolitic Biotite 7.101 17.016 11 61 ±2 Buitre tuff ST-101 North flank of Cerro El 7073,500/463,050 Sanidine-biotite rhyolitic Biotite 7.055 17.311 11 62 ±2 Buitre tuff

Rhyodacilic domes and lavas of Cerros Contreras-La Antena IO-140 Cerro San Juan 7092,200/448,180 Sanidine-biotite Biotite 6.297 15.154 17 61 ±2 rhyodacitic lava IO-90 Cerro San Juan 7092,350/448,480 Biotite rhyodacitic dome Biotite 5.668 13.030 17 58.2 ± 1.5 IO-89 Cerro San Juan 7092,470/449,000 Sanidine-biotite Biotite 7.181 16.309 25 57.5 ±1.5 rhyodacitic dome SR-59 East of Cerro Contreras 7089,625/448,550 Potassic rhyolitic dome Whole rock 4.007 8.768 16 55.4 ± 1.83 Downloaded By: [Canadian Research Knowledge Network] At: 21:38 28 May 2011

Los Amarillos-Kilómetro Catorce volcanic sequence RKM-123 Kilómetro Catorce area 7188,140/436,560 Altered (alunite) rhyo- Whole rock 1.772 4.330 37 62 ±3 litic tuff related to a dome SR-75-22 Cos. Los Amarillos 7101,375/428,400 Biotite rhyolite welded Biotite 6.475 15.375 14 60±2 tuff IP-912 Cos. Los Amarillos 7106,350/433,925 Sanidine rhyolite welded Whole rock 3.025 6.669 19 55.8 ± 1.83 tuff SR-24-12 W of El Salvador town 7098,150/435,600 Trachyandesitic lava Whole rock 2.060 4.369 16 53.7 ± 1.83 IP-922 W of El Salvador town 7102,450/430,550 Pyroxene microdiorite Whole rock 0.942 2.079 44 55.9 ±2.1

El Salvador porphyry complex (middle Eocene) Quartz -sanidine and quartz-plagioclase rhyolite porphyries EL SALVADOR COPPER DEPOSIT REVISITED 27 IP-11 Cerro Pelado 7099,180/445,900 Quartz-sanidine por• Whole rock 6.712 11.119 41 42.1 ± 1.2 phyry (sericitically altered) ES-7458 Cerro Pelado 7099,339/446,129 Quartz-sanidine rhyolite Sanidine plus 4.503 8.032 56 45.3 ± 2.0 porphyry altered plagioclase IP-10 Saddle between Cerros 7098,580/446,300 Old Camp quartz-plagio• Whole rock 1.892 3.266 30 43.9 ± 1.5 Riolita and Pelado clase porphyry dike

Granodiorite and dacite porphyry IT-9 Level 2445, El Salvador 7096, 410/444, 240 Granodioritic porphyry Biotite 7.718 12.512 21 41.2 ± 1.1 Mine (L type) IT-10 Level 2445, El Salvador 7096,350/444,250 Granodioritic porphyry Biotite 7.212 11.809 25 41.6 ±1.2 Mine (X type) IP-45 Quebrada "M" 7097,470/444,950 Biotite dacitic porphyry Biotite 7.617 13.144 21 43.8 ± 1.2 (A type porphyry) ES-12337 Quebrada "M" 7097,144/444,882 Biotite dacitic porphyry Biotite 5.165 8.352 38 41.1 ± 1.3 (fine grained A- or L-type porphyry) ES-12338 Quebrada "M" 7097,355/444,956 Biotite dacitic porphyry Biotite 5.566 9.029 41 41.2 ± 1.3 (coarse grained A- or L-type porphyry) ES- 12339 Quebrada "M" 7097,464/444,766 Granodioritic porphyry Biotite 5.739 8.629 36 38.3 ±1.23

1Analyses completed in the K-Ar Geochronology Laboratory of SERNAGEOMIN by Carlos Pérez de Arce. Samples with prefixes SR, ST, and SB are from Cornejo et at. (1993a) and EC from Cornejo and Mpodozis (1996). Samples with prefixes IO, IP, and IT represent new ages determined during the course of this study. Samples with ES prefix were collected by Osman Olivares, El Salvador Division, CODELCO. Sample RKM-123 was collected by Orlando Rivera as reported in Rivera (1995). 2Denotes samples located in Figure 3; all others are located in Figure 4. 3Denotes ages interpreted as minimum ages; see text for discussion. Downloaded By: [Canadian Research Knowledge Network] At: 21:38 28 May 2011

28 CORNEJO ET AL.

1 TABLE 2. Recalculated K-Ar Ages Reported by Gustafson and Hunt (1975)

Sample no. Location UTM coordinates Rock unit or Material %K Ar rad, nl/g %Ar, atm. Previous age Recalculated age (N/E) lithology dated and error, Ma2 and error, Ma3

ES-6025B Old Camp 7098,545.7/446,057.8 Quartz-plagioclase Sericite 2.33 3.86 82 45.6 ±1.3 42.1 ±2.6 porphyry ES-6136 Colina de Cobre 7097,448.4/445,056.5 Granodioritic Biotite 7.68 12.56 ±0.10 29.3 40.5 ± 0.9 41.6 ±1.8 (Cu Hill) porphyry ES-3256 Level 2933, El 7096,196.0/444,818.9 K porphyry Sericite 7.30 12.02 ± 0.13 38.4 40.8 ± 0.5 41.9 ± 1.0 Salvador Mine ES-3442A Level 2400, El 7096,213.4/444,908.8 L porphyry Amphibole 0.364 0.618 ±0.010 54.2 42.1 ±2.8 43.2 ± 5.6 Salvador Mine ES-3853 Quebrada Granito 7094,600.6/443,498.2 L porphyry Amphibole 0.503 0.854 ±0.014 58.6 42.1 ± 1.3 43.2 ± 2.6 ES-3853A Quebrada Granito 7094,600.6/443,498.2 L porphyry Biotite 7.29 12.36 ±0.17 26.1 42.0 ± 0.6 43.1 ± 1.2 ES-1910 Level 2400, El 7096,161.3/444,482.7 Biotitized andesite Biotite 40.8 ±1.3 41.8 ±2.64 Salvador Mine ES-4269 Level 2400, El 7095,837.0/443,999.7 "Latite" dike Biotite 6.58 10.86 ± 0.12 13.4 40.9 ± 0.5 42.0 ±1.0 Salvador Mine

1Reported ages represent the average of duplicate analyses for most of the samples; completed by C. Brooks (Univ. de Montréal), except ES-1910, which was completed by Geochron. Sample coordinates converted to UTM by Osman Olivares from mine records. 2Uncertainties reported by Gustafson and Hunt (1975) are 1 sigma (a) uncertainties; recalculated ages are reported to 2σuncertainties. 3Ages recalculated using the decay constants of Steiger and Jäger (1977) and the original analytical data provided courtesy of L.B. Gustafson (pers. commun., 1992). 4Original analytical data were unavailable, so new age determined using the conversion table of Harland et al. (1990). EL SALVADOR COPPER DEPOSIT REVISITED 29

Regional Framework and Pre-Tertiary overlain by Lower Cretaceous continental- Geology of El Salvador Area facies epiclastic rocks. West of the Sierra Castillo-Agua Amarga The El Salvador deposit (26° 15' S) is located fault, the oldest rocks are Jurassic basaltic in the southern segment of the Domeyko fault andesitic and andesitic amygdaloidal lavas and system (Fig. 1), a trench-linked strike-slip fault intercalated fossiliferous limestones of the system having activity contemporaneous with Sierra Fraga Formation. These rocks are over emplacement of numerous middle Eocene to early Oligocene porphyry Cu-Mo deposits in the lain by Lower Cretaceous andesitic lavas, vol Andean Precordillera of northern Chile (Mak- canic breccias, and intercalated continental- saev and Zentilli, 1988; Sillitoe, 1988; Lindsay facies sedimentary rocks of the Mantos Gruesos et al., 1995; Reutter et al., 1996). In the north sequence (Fig. 4). Both units were deposited in ern and central part of the system, fault control a back-arc to intra-arc basin formed during the on the emplacement of porphyry Cu-Mo-related early stages of the Andean orogeny (Mpodozis intrusive complexes is evident where the por and Ramos, 1989) and contrast with the age- phyry Cu complexes are either localized along equivalent sequences deposited east of the master strands of the fault system (e.g., La Sierra Castillo fault. As Tertiary sinistral slip on Escondida and Chuquicamata) or lie within a the Sierra Castillo fault is no more than several short distance of the master fault along tens of km (Tomlinson et al., 1993, unpubl. peripheral subsidiary faults (e.g., Quebrada data), juxtaposition of different volcanic and Blanca and Collaguasi) (Boric et al., 1990; sedimentary facies suggests that the Sierra Cas Tomlinson et al., unpubl. mapping, 1993- tillo fault reactivated a long-lived fault system 1996). The southern Domeyko fault system that separated the carbonate platform margin of consists of the Sierra Castillo fault, which the Jurassic and Cretaceous basin from the passes southward into the Agua Amarga oblique volcanic-filled rift basin to the west. reverse fault (Figs. 1 and 3). The Potrerillos Cu Shortening in the Cretaceous (95 to ~85 Ma) deposit lies east of the Sierra Castillo-Agua deformed large areas of northern Chile and Amarga fault system, within a contempora marked the transition from the dominance of neous fold-and-thrust belt (Olson, 1989; intra-arc extension to that of shortening punc Tomlinson et al., 1993; Tomlinson, 1994). The tuated by periods of extension that has since El Salvador deposit, the oldest deposit of the characterized the Central Andes (Coira et al., Eocene and Oligocene porphyry Cu belt (Sil 1982; Boric et al., 1990; Mpodozis and Ramos, litoe, 1988), lies ~10 km west of the Sierra 1989). During a period of quiescence and Castillo fault. Structural control on the forma extension in the Late Cretaceous, the Llanta tion of this porphyry Cu deposit by the Formation was deposited unconformably over Domeyko system is not obvious (Fig. 3). the Sierra Fraga Formation in an extensional East of the Sierra Castillo-Agua Amarga basin limited on the east by the Mantos Gruesos Downloaded By: [Canadian Research Knowledge Network] At: 21:38 28 May 2011 fault, the oldest rocks consist of a basement of and Kilómetro Catorce faults (Fig. 4). It con upper Paleozoic granitic and volcanic rocks sists of andesitic sandstones, red conglomer unconformably overlain by Upper Triassic to ates, coarse-grained volcaniclastic rocks, sparse Lower Cretaceous sedimentary and volcanic ignimbrites, and high in the formation, ande rocks (Fig. 3, Quebrada del Salitre, Montandón, sitic and basaltic lavas and breccias. Between and Asientos formations; Quebrada Vicuñita 26° and 27° S (Fig. 3) this unit has yielded sequence in Fig. 4). The granitic rocks—the K-Ar mineral and whole-rock ages between 73 Carboniferous and Permian Pedernales and ± 4 Ma and 69 ± 3 Ma (SB-472, SR-372, Sierra Castillo batholiths—represent the mag- EC-326). In the El Salvador area, it is intruded matic arc of western Gondwana (Brook et al., by various Paleocene dioritic stocks with K-Ar 1986; Cornejo et al., 1993a; Mpodozis and biotite and whole-rock ages of 64 + 2 Ma to 59.7 Cornejo, 1994). The Mesozoic sedimentary ± 5.7 Ma (SR-456, SR-245, IO-139, IP-68). The rocks are Lower Jurassic to Neocomian shallow- Mantos Gruesos fault had syn-depositional marine carbonate sequences deposited in a extensional slip prior to post-depositional W-facing platform environment. They contain shortening in the latest Cretaceous to earliest only sparse interbedded volcanic rocks, and are Paleocene, when the basin collapsed and the 30 CORNEJO ET AL. Downloaded By: [Canadian Research Knowledge Network] At: 21:38 28 May 2011

FIG. 4. Geologic map of the El Salvador area showing K-Ar and U-Pb ages determined during this study. Those ages outside the area of the figure are located in Figure 3. Abbreviations: CC = Colina de Cobre. Legend on facing page. EL SALVADOR COPPER DEPOSIT REVISITED 31

Legend, Figure 4 El Salvador Porphyry Complex (middle Eocene) Granodiorite, granodiorite porphyry and dacite Quaternary alluvium porphyry intrusions (43-41 Ma) Atacama Gravels (middle Miocene) Quartz-plagioclase and quartz-sanidine rhyolile porphyry dikes, stocks and laccoliths (44-43 Ma) Paleocene Volcanic Event El Salvador Caldera and Indio Muerto Domes Los Amarillos-Kilometro Catorce Volcanic Sequence Rhyolitic domes, dikes, sills, and associate pyroclastic and epiclastic rocks (58 Ma) Andesitic and trachyandesitic lavas Monzonitic intrusions (59 Ma) Ignimbrites (~60 Ma) Cerros Contreras-La Antena ignimbrites Rhyolitic domes (62 Ma) and basal ignimbrite sequence at Cerro Indio Muerto (63-60 Ma) Dioritic intrusions (64-56? Ma) Paleozoic-Mesozoic Basement west zone East Zone Llanta Formation (Upper Cretaceous) Quebrada del Salitre, Montandon and Asientos a) Sandstone, conglomerates and volcaniclastic Formations, and Quebrada Vicuriita sequence breccias (Upper Triassic to Lower Cretaceous) b) Volcanic breccias and andesitic lavas Marine carbonate sequences, siliciclastic rocks, and scarce interbedded andesitic and basaltic lavas Sierra Fraga Formation and Mantos Gruesos sequence (Jurassic-Lower Cretaceous) Sierra Castillo Batholith Andesitic and basaltic lavas with interbedded (Carboniferous-Permian) marine limestone and continental-facies epiclastic rocks Structures K-Ar ages (in Ma) Normal faults and scissors faults Biotite Whole rock Eocene reverse faults Eocene strike-slip fault (Sierra Castillo System) Drainage Caldera ring fracture (Paleocene) Mine collapse Upper Cretaceous to lower Paleocene reverse fault Upper Cretaceous to lower Paleocene folds Cretaceous normal fault reactivated as reverse fault Paleocene scissors fault reactivated as reverse fault

fault was reactivated as a W-vergent reverse (Cornejo et al., 1993a). In the El Salvador area, fault (Cornejo and Mpodozis, 1996). Shorten silicic eruptive centers are associated with nor ing is particularly evident north of Cerro Indio mal faults that post-date latest Cretaceous to Muerto, where the Llanta Formation is earliest Paleocene shortening (Fig. 4). The deformed into a series of upright folds and cut oldest center is the El Salvador caldera and by the E-vergent Sierra Miranda thrust (Fig. 4). associated ignimbrite sequence (Cornejo et al.,

Downloaded By: [Canadian Research Knowledge Network] At: 21:38 28 May 2011 1993a, 1993b, 1994). On Cerro Indio Muerto, a second eruptive center of coalescing rhyolitic Paleocene Volcanism domes overlies outflow-facies tuffs derived Paleocene high-K calcalkaline volcanic rocks from the adjoining El Salvador caldera. To the include large volumes of clinopyroxene ± west lies a sequence of principally andesitic olivine ± biotite trachybasalt and trachy- rocks intercalated with outflow-facies rhyolitic andesite lava, and of sanidine rhyolitic lava, ignimbrites (Los Amarillos-Kilometro Catorce domes, and tuff (Cornejo et al., 1993a, 1993b, volcanic sequence). The intermediate volcanic 1994). The volcanic rocks unconformably over rocks are thought to have been derived from lie the folded and faulted Cretaceous Llanta coalesced stratovolcanos located west of the Formation and older rocks; the unconformity district, but which now are deeply eroded. corresponds to the Hornitos unconformity of Gustafson and Hunt (1975) (Fig. 2). Silicic El Salvador caldera calderas and rhyolitic dome complexes are rec Cerros Contreras and La Antena, lying south ognized, but eruptive centers for the mafic and of El Salvador, are composed of densely welded intermediate volcanic rocks are not well defined rhyolitic ignimbrites, dipping 10 to 25° east 32 CORNEJO ET AL.

FIG. 5. Generalized cross-sections for Paleocene sequences within the El Salvador area. Section A-A' is through the El Salvador caldera, B-B' through Cerro Indio Muerto, and C-C' through the Los Amarillos volcanic sequence. Downloaded By: [Canadian Research Knowledge Network] At: 21:38 28 May 2011 Sections A-A' and B-B' are located in Figure 4, andC-C ' is in Figure 3. Geochronologic symbols are the same as those used in Figure 4. Abbreviations: min = age considered to be minimum; * = age considered to be anomalously old. Sections are without vertical exaggeration.

and southeast, with a minimum exposed thick from 300 m to >600 m toward the southwest. It ness of 1300 m; the base of the sequence is not is inferred to be a single cooling unit. Pumice exposed (Fig. 5A). The Cerros define a semicir has flattening ratios (length:height) between cular ridge, open to the northwest, which out 10:1 and 20:1 in the thinner sections, ranging up to 80:1 in the thicker sections. This lower lines the form of the El Salvador caldera (Fig. tuff is overlain by brecciated and vesicular 4). Sparse andesitic lavas are interbedded andesitic lavas, including pyroxene trachyan- within the tuff sequence, indicating multiple desites (Fig. 5A). Overlying the andesitic lavas, explosive eruptions. The lower stratigraphic rocks are principally eutaxitic sanidine rhyo levels on the north flank of the ridge consist of litic tuffs with traces of biotite, lithic tuffs with welded rhyolitic tuff (with eutaxitic and rheo- cognate fragments, and sparse andesitic lava morphic textures) that increases in thickness and breccia. Lithic tuffs are lithologically simi- EL SALVADOR COPPER DEPOSIT REVISITED 33

lar pyroclastic dikes that probably represent that the rotational axis (hinge) to the caldera lid eruptive conduits. The highest stratigraphic lies to the northwest and has a northeast-to- levels consist of welded sanidine-biotite rhyo- north strike. Steep NE-striking faults, located litic and dacitic tuffs, with large sanidine crys southeast of Cerro Indio Muerto (Fig. 4), lie tals. Fine-grained porphyritic monzonite to subparallel to the inferred orientation of the quartz monzonite intrudes to high stratigraphic trap-door hinge and may be related to that levels in the tuff sequence near La Cantera structure. They also are subparallel to the (Fig. 4). southernmost segment of the Mantos Gruesos On the southeastern flank of Cerros Con- fault, possibly indicating a control by older pre- treras and La Antena, the ignimbrite sequence caldera faults on the hinge orientation. The is limited by a semicircular normal fault that Mantos Gruesos and Kilometro Catorce faults downdrops the northwestern hanging-wall form the principal structural boundary block of welded, intracaldera-facies tuff against between the Llanta Formation on the west and a footwall of the Sierra Fraga Formation (Fig. the Lower Cretaceous and Jurassic volcanic 4). Near the fault are massive megabreccia, sequences on the east, suggesting that the two representing intracaldera landslide deposits, faults probably are contiguous and continue interbedded with welded tuff. Megabreccia beneath Cerro Indio Muerto and across the clasts are principally Jurassic lavas and Paleo- northwestern margin of the El Salvador caldera cene welded tuffs. The distribution of mega in the approximate location of the inferred trap breccia, coupled with an annular structural door hinge. The El Salvador caldera is smaller margin and intracaldera tuff facies on Cerros than the 15-km-diameter subcircular structure Contreras and La Antena, represent the princi originally proposed by Frances et al. (1983) pal evidence for the El Salvador caldera. from topographic features on LANDSAT imagery. The Cerros Contreras and La Antena intra caldera block also is limited on the northeast by Outflow-facies ignimbrites are recognized 3 a system of subvertical WNW-striking faults km southeast of the caldera margin, where they that place the intracaldera block against the pre- unconformably overlie Jurassic volcanic rocks caldera Mesozoic units, and on the southwest (Fig. 4), and ~20 to 25 km southeast on Cerro by the N-striking Kilómetro Catorce reverse El Buitre near Potrerillos (Fig. 3), where they fault that places Mesozoic rocks and intra overlie Jurassic limestones on the east side of caldera facies against extracaldera Paleocene the Sierra Castillo fault system. They also are andesitic lavas and ignimbrites of the Los present to the west and northwest in the Amarillos-Kilómetro Catorce sequence (Fig. Los Amarillos-Kilómetro Catorce volcanic 4). The WNW-striking faults in the Cuesta San sequence (Fig. 4) and 10 to 15 km southwest in Juan area, although accommodating large dis the Pampa del Inca area (Fig. 3). placements on the caldera fill, fade toward the K-Ar geochronology indicates that explosive WNW into the Cerro Indio Muerto area and volcanism in the El Salvador caldera occurred Downloaded By: [Canadian Research Knowledge Network] At: 21:38 28 May 2011 toward the ESE into the footwall of the caldera- in the Paleocene (Figs. 4 and 5A). Suitable margin normal fault. This suggests that the material to date is unavailable for the lower fault system is a scissors fault system related to tuffs, so inception of explosive volcanism is not collapse during caldera formation. The Kilo- constrained. Tuff above the andesitic member metro Catorce fault is interpreted to be a latest has a biotite K-Ar age of 66 ± 2 Ma (IO-142), Cretaceous to earliest Paleocene reverse fault whereas the stratigraphically highest tuffs have reactivated in the Eocene, but the contrasting K-Ar ages on biotite of 63 ± 2 Ma (10-81) and ignimbritic facies on either side of the fault 61 ± 2 Ma (SR-51). The thin outflow facies tuff suggest that it also was a caldera-related scissors on Cerro El Buitre has identical K-Ar ages on fault in the Paleocene. The structure and vol biotite of 61±2 and 62±2Ma(ST-77-ST-101). canic facies suggest that the caldera complex A diorite intruding into intracaldera tuff on developed as a trap-door caldera. Cuesta San Juan has a whole-rock K-Ar age of The open-to-the-northwest semicircular 63 ± 4 Ma (IO-139). The range of ages suggests form of the Cerros Contreras and La Antena, that explosive volcanism and the last stages of the bounding normal fault, and the east-to- caldera formation may have continued for as southeast dips in the intracaldera block suggest much as 5 million years between ~66 Ma and 61 34 CORNEJO ET AL.

206 238 207 235 Ma. However, the 66 + 2 Ma age (IO-142) is Pb*/ U and Pb*/ U ageg of 59.5 Ma considered to be too old for two reasons: (1) and a 207Pb*/206Pb* age of 58 Ma. The last other Paleocene units regionally correlative to zircon fraction has normal discordance as a the El Salvador area sequence, both east of the result of recent Pb loss, perhaps during deep Sierra Castillo fault (Cerro Valiente sequence) Cenozoic weathering (Mortimer, 1973). The and north of El Salvador in the Exploradora pattern of normal discordance also suggests area, yield K-Ar ages no older than 63 + 2 Ma at that minor quantities of inherited zircon also the base of those sequences (Cornejo et al, may be present. Based upon the overall distribu 1993a; Cornejo and Mpodozis, 1996); and (2) tion of U-Pb ages together with the one concor the low-K content for biotite in IO-142 suggests dant fraction, an age of 59 ± 2 Ma is assigned to some K-loss and potential overestimation of the the quartz monzonite. Evidently, volcanic and age. Therefore, we consider it more likely that plutonic activity in the El Salvador caldera caldera formation occurred over a much shorter ceased by 58 to 59 Ma. period of time, most likely around 63 to 61 Ma in the early Paleocene. Cerro Indio Muerto Post-collapse stocks and domes intrude intra- At Cerro Indio Muerto, the Cretaceous caldera tuff (Figs. 4 and 5A). A dome-derived Llanta Formation is unconformably overlain by sanidine-biotite rhyodacitic lava flow overlying a 100- to 200-m-thick rhyolitic tuff sequence the tuff sequence on Cerro San Juan has a K-Ar correlated by Gustafson and Hunt (1975) to the biotite age of 61 ± 2 Ma (10-140). Along the Hornitos Formation of the Copiapó area (Fig. annular margin of the caldera, the lava flow is 2). These tuffs are unconformably overlain by intruded and overlain by a biotite rhyodacitic alkali-feldspar rhyolite domes and a sequence dome having K-Ar biotite ages of 58.2 ±1.5 Ma of pyroclastic and epiclastic rocks and minor (10-90) and 57.5 ± 1.5 Ma (10-89). East of andesitic lava into which microdiorite sills were Cerro Contreras, a rhyolitic dome also intrud emplaced (Figs. 4 and 5B). This second uncon ing the caldera margin has a K-Ar whole-rock formity is the Indio Muerto unconformity of age of 55.4 ± 1.8 Ma (SR-59). This age is Gustafson and Hunt (1975), and the overlying interpreted to be a minimum because of domes and volcanic rocks constitute their groundmass devitrification, which has been "Indio Muerto rhyolite domes" and "Indio shown in the region to result in K-Ar whole- Muerto series volcanics," respectively (Fig. 2). rock ages that are consistently 4 to 8 m.y. too Rb-Sr whole-rock and mineral isochron ages of young (Cornejo et al., 1993a). 49.3 + 5.6 Ma and 49.2 + 6.4 Ma were reported To the southwest, a biotite-amphibole por- for these rocks, albeit with some uncertainty as phyritic quartz monzonite intruding the tuff to their validity (Gustafson and Hunt, 1975, sequence at La Cantera has a minimum 51.8 + p. 879). 1.7 Ma K-Ar whole-rock age (Cornejo et al., 1993a). Four fractions of zircons from this The base of the rhyolitic tuff sequence con Downloaded By: [Canadian Research Knowledge Network] At: 21:38 28 May 2011 quartz monzonite have discordant U-Pb ages sists of laterally discontinuous, unwelded to ranging between 59 and 61 Ma (Table 3; Fig. slightly welded, air-fall and plinian tuffs con 6A), although their uncertainties overlap containing small pumice lapilli; fragmental quartz, cordia. Two fractions are slightly reversely dis plagioclase, and biotite phenocrysts (with diam cordant, perhaps either because of analytical eters measured in mm); and slightly larger problems or because of internal Pb redistribu alkali-feldspar phenocrysts. These are overlain tion or selective uranium loss during younger by 50 to 200 meters of rhyolitic and dacitic alteration by hydrothermal or metamorphic welded ignimbrites containing sanidine and fluids or during chemical weathering (Mattin- locally biotite. The tuff sequence is cut by NE- son et al., 1996). One of these reversely discor and NW-striking faults and was tilted gently dant fractions also underwent hydrothermal (5° -10°) southwest and south during caldera leaching experiments (Mattinson, 1994), a pro collapse (Fig. 5B). On the southeastern side of cedure that may remove damaged parts of zir Cerro Indio Muerto, zircons were recovered cons resulting in either concordant or normally from a slightly welded rhyolitic tuff (IT-6, Fig. discordant U-Pb ages (Mattinson et al., 1996). 6B) conformably overlying red sandstone One zircon fraction is concordant with immediately above the unconformity with the EL SALVADOR COPPER DEPOSIT REVISITED 35

FIG. 6. A portion of the concordia diagram showing U-Pb geochronologic data for zircons from Paleocene volcanic and plutonic rocks. A. Biotite-amphibole quartz monzonite (IT-17) intruding intracaldera tuff near La Cantera. B. Slightly welded tuff (IT-6) at the base of the Paleocene volcanic sequence, southern flank of Cerro Indio Muerto. All data except for one fraction

Downloaded By: [Canadian Research Knowledge Network] At: 21:38 28 May 2011 are shown. Uncertainty ellipses are shown only for those fractions (filled squares) that were hydrothermally leached. Uncertainties for the other zircon frac tions (filled diamonds) are large because of large common Pb content, and extend over significant parts of the diagram. Line passes only through fractions with negligible common Pb. C. Rhyolite dome (IT-8) on Cerro Indio Muerto. D. Sanidine rhyolite dike (IT-4) on the northeast flank of Cerro Indio Muerto. E. Composite diagram for zircons from Paleocene volcanic and plutonic rocks. Only normally discordant data are shown; reversely discordant data are not plotted. Inheritance discordia for the basal tuff (IT-6) on Cerro Indio Muerto is shown. Stippled symbols are zircons that have undergone step-wise hydrothermal leaching.

underlying Cretaceous Llanta Formation (Figs. ness of the analyses. Three other zircon frac 4 and 5B). The clear, euhedral zircons contain tions underwent hydrothermal leaching experi considerable common Pb (measured 206Pb/204 ments to remove common Pb and to minimize Pb < 54), which limits the chronologic useful effects of Pb loss (Mattinson, 1994). Experi- 36 CORNEJO ET AL.

mental hydrothermal leaching of one fraction northeastern flank of Cerro Indio Muerto, and for three days failed to remove common Pb, but strike northeast, parallel to the alignment of the shifted the 206Pb*/238U and 207Pb*/235U ages to larger domes (Fig. 4), as well as parallel to slightly older ages (60 to 61 Ma) than those in basement faults (such as the Mantos Gruesos conventionally analyzed fractions (59 to 60 fault) and the inferred orientation of the cal Ma). Leaching of the same fraction and a sec dera trap-door hinge. This suggests that these ond size fraction for a longer period of time structures probably controlled magma emplace effectively removed the common Pb, as evi ment. As Cerro Indio Muerto is located on the denced by their high measured 206Pb/204Pb. margin of the El Salvador caldera, collapse- U-Pb ages of leached fractions are discordant related faulting also may explain the gentle because of small amounts of inherited zircons. southwest and south dips on the basement A line through these two fractions intersects unconformity and rhyolitic tuff sequence, concordia at 60 Ma and 783 Ma, with a large and despite an apparent structural conflict between meaningless uncertainty on the upper inter dip directions and inferred NE-strike orienta cept. As the rock can be no older than the tion of the caldera hinge. Within the intra- youngest U-Pb age used to define the inherited caldera block at Cerros Contreras and La discordia array, the base of the tuff sequence on Antena, the caldera lid is dissected by Cerro Indio Muerto is interpreted to be 60 ± 1 numerous faults that bound coherent blocks Ma. Within the same section on the eastern having contrasting dip direction and magni flank of Cerro Indio Muerto, biotite from the tude, and it is likely that a similar structural stratigraphically highest welded rhyolitic tuff architecture characterized the caldera margin. has an essentially identical age of 59.4 ± 1.5 Ma Conversely, Eocene shortening might also have (IP-19). Ages from the tuff sequence agree with produced or accentuated dips on Paleocene those of the adjoining El Salvador caldera, rocks. The facts that Eocene porphyry contacts which is interpreted to have been their source. and sulfide veins in the mine are vertical or dip steeply north (Gustafson and Hunt, 1975; L.B. The core of Cerro Indio Muerto is formed by Gustafson, pers. commun., 1997) and that NW- several coalescing rhyolite domes aligned in a striking faults in the mine cut Paleocene northeast direction (Gustafson and Hunt, 1975; pyroclastic rocks and all units of the Eocene P. Cornejo, unpubl. mapping, 1994), with a porphyry complex (G. Müller, pers. commun., rhyolitic stock located in the headwaters of 1996) suggest that at least some component of Quebrada Aglomerado possibly representing the faulting and tilting is a result of younger, the feeder to a peripheral extrusive center. The probably Eocene, deformation. Thus, the gentle dome rocks consist of massive rhyolite with southwesterly dips in the Cerro Indio Muerto abundant phenocrysts of sanidine, lesser area are inferred to be local phenomena that plagioclase, and trace amounts of biotite in a formed in the hinge area of a fragmented cal flow-banded groundmass with spherulitic dera lid, which later was modified by Eocene or devitrification textures. Quartz phenocrysts are Downloaded By: [Canadian Research Knowledge Network] At: 21:38 28 May 2011 younger deformation. absent or sparse. Flank breccias, block-and-ash flow deposits, small-volume debris flow Geochronologic data for the rhyolitic dome deposits, ash tuff, and fine-grained epiclastic and associated rocks on Cerro Indio Muerto are rocks are preserved on the southeastern flank of limited, but are compatible with their emplace Cerro Indio Muerto, where the rocks filled ment following collapse of the El Salvador cal depressions in an irregular topography result dera. A slightly altered flow-banded aphyric ing from synvolcanic faulting during caldera rhyolite (IT-8) from the dome on the south collapse (Fig. 5B). Epiclastic rocks containing western flank of Cerro Indio Muerto contains fossil plant material and freshwater gastropods euhedral, clear, prismatic zircons. Four zircon were deposited between pyroclastic eruptions in fractions have slightly discordant 206Pb*/238U restricted lakes that filled depressions. Sparse and 207Pb*/235U ages between 55 and 58 Ma, andesitic lava and microdiorite sills also are although, as is the case for most of the Paleo interlayered with the pyroclastic and epiclastic cene rocks, the uncertainties on the individual rocks. analyses intersect concordia. The range of U-Pb Rhyolitic dikes and sills, peripheral to the ages clearly indicates a Paleocene age for the main domes, intrude basement rocks on the dome (Fig. 6C). 207Pb*/206Pb* ages are >58 Ma EL SALVADOR COPPER DEPOSIT REVISITED 37

and range up to 92 Ma, implying a small amount + 3 Ma (RKM-123) (Rivera, 1995). On the of inherited zircons. U-Pb ages of these zircons south, in the Rìo de la Sal valley, discontinuous are similar to those in the welded tuff (see rhyolite ignimbrite lies along the unconformity above), but are slightly younger (Fig. 6E), con at the base of the sequence. In the area of Los sistent with their stratigraphic relations. Dis Amarillos, north and west of the town of El tribution of zircon U-Pb ages in an array Salvador (Figs. 3 and 4), the sequence, which parallel to concordia implies some Pb loss dur also contains a basal ignimbrite, is folded into a ing a younger event, most likely during intense broad syncline with a wavelength of ~11 km; hydrothermal alteration at ~41 Ma (Zentilli, andesitic lavas occupy the core of the fold (Fig. 1974; Gustafson and Hunt, 1975). Eocene Pb 5C). Cornejo et al. (1993a) speculated that this loss is confirmed by SHRIMP 206Pb*/238U ages area might also be a caldera, but this is not of 43.6 + 2.1 Ma and 40.3 ±1.4Ma, determined on the tips of two separate grains (grains 8.1 supported by additional mapping. On the west and 9.1) (Table 4). SHRIMP 206Pb*/238U ages of ern flank of the syncline, a more complex spots on the other 15 zircon tips are between 56 sequence has several sanidine-biotite welded and 62 Ma, similar to 206Pb*/238U ages of multi- rhyolitic tuffs interbedded with andesite and grain zircon fractions (Tables 3 and 4). These trachyandesite lavas. One of the higher welded spot analyses have a mean 206Pb*/238U age of rhyolitic tuffs has a K-Ar biotite age of 60 ± 2 59.1 Ma. Taking the range of ages determined Ma (SR-75-2). Another rhyolite welded tuff, during multigrain and SHRIMP U-Pb analyses, from near the base of the sequence, has a whole- a 58 + 2 Ma age is assigned to the rhyolite dome. rock K-Ar age of 55.8 ± 1.8 Ma (IP.91), and a This age is as much as 8 million years older than trachyandesitic lava from within the core of the previously determined by Rb-Sr (Gustafson and syncline has a whole-rock K-Ar age of 53.7 ±1.8 Ma (SR-24-1). Both of the latter two ages Hunt, 1975), and is similar to the K-Ar ages for are considered to be minimum ages, as the rocks biotite from the post-collapse rhyodacitic dome are devitrified and one tuff (IP-91) lies strati- of Cerro San Juan that was emplaced along the graphically below another tuff (SR-75-2) that El Salvador caldera ring fracture. has an older K-Ar mineral age. The sequence is Zircons from a sanidine rhyolite (IT-4), one intruded by a swarm of N-trending sericitized of the thickest peripheral dikes, have discordant rhyolite dikes and by numerous subvolcanic U-Pb ages (Table 3; Fig. 6D) that are distributed dioritic stocks, one of which has a K-Ar whole- in an array subparallel to concordia. Experi rock age of 55.9 ± 2.1 Ma (IP-92). mentally hydrothermally leached zircons have older U-Pb ages, thereby implying some slight The predominance of intermediate lavas over Pb loss from the zircon. 206Pb*/238U and silicic rocks distinguishes the Los Amarillos- 207Pb*/235U ages are between 56 and 62 Ma. Like zircons from other Paleocene volcanic and Kilometro Catorce volcanic sequence from plutonic rocks, these zircons apparently also equivalent-age volcanic sequences on Cerro Downloaded By: [Canadian Research Knowledge Network] At: 21:38 28 May 2011 contain small amounts of inherited zircons. Indio Muerto and in the El Salvador caldera to From these data alone, an age cannot be the east. However, it is inferred that ignimbrite assigned to the dike. However, in view of the within the Los Amarillos-Kilometro Catorce similarity of the U-Pb data to the other Paleo sequence was derived from the El Salvador cene rocks (Fig. 6E), it is likely that it is about caldera, in view of their lithologic and age the same age, i.e., in the 58- to 60-Ma range. similarities and proximity. Los Amarillos-Kilómetro Catorce volcanic sequence Eocene Cupriferous Porphyry Intrusions West of the Kilómetro Catorce fault are pyroxene trachyandesite and andesite lavas The Eocene porphyritic intrusive complex, intercalated with rhyolitic outflow ignimbrites hosting the El Salvador porphyry Cu-Mo and epiclastic rocks (Fig. 4). These rocks deposit, consists of stocks, dikes, and sills unconformably overlie the Llanta Formation emplaced along a 6-km NNE-trending belt and contain an altered rhyolite dome and flows extending from Quebrada Granito on the SSW that have a whole-rock (alunite) K-Ar age of 62 to Cerro Pelado on the north (Fig. 4). Individual Downloaded By: [Canadian Research Knowledge Network] At: 21:38 28 May 2011

38 CORNEJO ET AL.

1 TABLE 3. U-Pb Geochronologic Data for Zircons from Volcanic and Plutonic Rocks in the Indio Muerto Area

Observed ratios4 Atomic ratios5.6 Ages, Ma

2 3 3 3 3 Fraction Weight, 206Pb, 238U 206Pb 207Pb 208pb 206Pb 207Pb3 207Pb 206Pb3 207Pb3 207Pb 206Pb 238U 235U 206Pb3 238U 206Pb3 mg ppm ppm 204Pb 206Pb 235U

Paleocene volcanic an plutonic rocks Basal tuff: IT-6; UTM 70 95,200N/445,920E N>100(3) 10.1 1.80 220 46.8 0.36074 1.0217 0.00945(0.1) 0.06221 (2.3) 0.04775(2.3) 60.6 61.3 87 N>100(14) 9.9 1.84 226 1742 0.05587 0.3480 0.00945(0.1) 0.06188(0.7) 0.04752(0.7) 60.6 61.0 75 M<63(14) 12.7 3.93 484 7813 0.04925 0.3076 0.00940(0.1) 0.06148(0.4) 0.04743(0.4) 60.3 60.6 71 M>163 10.4 1.52 188 27.6 0.57844 1.5273 0.00932(0.4) 0.06110(7.7) 0.04753(7.4) 59.8 60.2 75 N>163 8.2 1.24 154 26.4 0.60080 1.5611 0.00932(1.5) 0.05341(9.6) 0.04157(9.3) 59.8 33.8 - N<63 10.0 3.98 501 46.7 0.36121 1.0147 0.00918(0.1) 0.05999(1.6) 0.04742(1.6) 58.9 59.2 70 M<63 14.8 3.36 424 53.6 0.31954 0.9016 0.00915(0.1) 0.05744(2.2) 0.04548(2.1) 58.8 56.7 - Rhyolite dome: IT-8; UTM 7094,930N/443,790E M<63 1.8 3.51 455 122 0.16821 0.5945 0.00891(0.1) 0.05848(2.0) 0.04756(1.8) 57.2 57.7 77 N>100 2.4 2.03 268 162 0.13822 0.5804 0.00877(0.2) 0.05785(1.1) 0.04786(1.0) 56.3 57.1 92 M>100 2.7 2.89 385 160 0.13891 0.5798 0.00875(0.1) 0.05690(0.9) 0.04717(0.8) 56.2 56.2 58 N<63 5.2 4.11 552 105 0.18665 0.6356 0.00859(0.1) 0.05653(1.0) 0.04770(0.9) 55.2 55.9 85

Sanidine rhyolite dike: IT-4; UT M 7097,160N/447,510E N>80(R2) 10.2 2.72 333 2141 0.05469 0.4065 0.00945(0.1) 0.06242(1.1) 0.04791(1.0) 60.6 61.5 95 M<100(R3) 7.4 2.87 355 2160 0.05456 0.4438 0.00933(0.1) 0.06153(0.6) 0.04784(0.6) 59.9 60.6 91 M<63 6.6 3.84 506 139 0.15299 0.7156 0.00879(0.1) 0.05794(0.8) 0.04783(0.8) 56.4 57.2 91

Quartz monzonite: IT-17; UTM 7087,610N/442,425E M>163(R2) 5.0 1.34 162 694 0.06813 0.3748 0.00957(0.1) 0.06213(0.9) 0.04707(0.8) 61.4 61.1 53 M>100 6.8 1.70 206 4% 0.07650 0.4018 0.00952(0.1) 0.06166(1.3) 0.04695(1.3) 61.1 60.8 47 N<63 11.2 2.04 255 862 0.06413 0.3546 0.00927(0.1) 0.06031(0.5) 0.04718(0.5) 59.5 59.5 58 M<63 6.7 3.60 277 955 0.06288 0.3617 0.00924(0.1) 0.06063(2.3) 0.04759(2.2) 59.3 59.8 79

Eocene porphyry intusions Cerro Pelado altered quartz-sanidine rhyolite porphyry: IT-3; UTM 7099,190N/445,920E N>100(R2) 5.5 0.74 124 585 0.07278 0.3391 0.00690(0.1) 0.04546(1.1) 0.04777(1.0) 44.3 45.1 88 N>80 5.0 0.95 162 171 0.13308 0.4902 0.00678(0.1) 0.04406(1.4) 0.04714(1.3) 43.5 43.8 56 N>63 6.7 0.98 202 235 0.10967 0.4366 0.00563(0.1) 0.03661(1.5) 0.04719(1.4) 36.2 36.5 59 Downloaded By: [Canadian Research Knowledge Network] At: 21:38 28 May 2011

Cerro Pelade altered quartz-sanidine rhyolite porphyry: ES-7458; UTM 7099,339N/446,129E N>130(R2) 5.1 0.66 117 240 0.10908 0.4244 0.00646(0.3) 0.04274(3.8) 0.04800(3.6) 41.5 42.5 99 N<100(R2) 5.9 1.08 190 613 0.07090 0.3372 0.00654(0.1) 0.04235(0.8) 0.04699(0.8) 42.0 42.1 49

Old Camp quartz-plagioclase porphyry dike: IT-5; UTM 7098,730N/446,730E N>163 13.9 1.63 271 152 0.14403 0.3954 0.00694(0.1) 0.04555(2.3) 0.04762(2.1) 44.6 45.2 81 N<63(14) 7.8 2.76 465 2702 0.05194 0.1872 0.00687(0.2) 0.04409(2.5) 0.04656(2.4) 44.1 43.8 27 M<100(]14) 10.3 2.52 428 160 0.13884 0.3873 0.00681(0.1) 0.04435(0.7) 0.04721(0.7) 43.8 44.1 60 M>163 5.4 2.15 365 101 0.19179 0.5042 0.00680(0.1) 0.04375(2.0) 0.04667(2.0) 43.7 43.5 32 M<63 3.7 3.00 520 193 0.12300 0.3716 0.00669(0.2) 0.04314(3.9) 0.04682(3.7) 43.0 42.9 40 EL SALVADOR COPPER DEPOSIT REVISITED 39 X porphyry: IT-10; UTM 7096,350N/444,250E N<63(R2) 5.7 1.06 177 568 0.07326 0.3020 0.00690(0.1) 0.04519(0.7) 0.04753(0.7) 44.3 44.9 76 M<63(R3> 6.7 1.08 187 568 0.06608 0.3094 0.00668(0.1) 0.04354(0.6) 0.04730(0.6) 42.9 43.3 64

L porphyry: IT-9; UTM 7096,4I0N/444, 240E M»00(R3) 9.0 0.74 131 749 0.06662 0.2438 0.00652(0.1) 0.04231(1.1) 0.04709(1.1) 41.9 42.1 54 N>163(R2) 9.8 0.73 130 569 0.07278 0.2267 0.00652(0.1) 0.04233(0.6) 0.04707(0.6) 41.9 42.1 53 N<100 9.3 1.23 219 807 0.06481 0.2653 0.00651(0.1) 0.04191(1.1) 0.04669(1.1) 41.8 41.7 34 M<100F 8.6 1.61 298 559 0.07335 0.4648 0.00624(0.1) 0.04060(0.6) 0.04718(0.5) 40.1 40.4 58

1Including the El Salvador porphyry copper deposit, north-central Chile. See Figure 4 for location of samples. 2N = nonmagnetic and M = magnetic at 1.8° and 1° side slope on a Franz Isodynamic separator. Zircon sizes 63, 100, and 163 are in microns. Hydrothermal leaching experiments modified from Mattinson (1994); only residues from leaching experiments were analyzed. Initial experiments in Savilex beakers were performed on a hot plate for three days followed by 11 days at 80°C; these experiments are indicated by (3) and (14) to denote total days of leaching. Other experiments were performed in successive steps with a 24-hour leach at 80° C followed by 24 hours at 160°Con hot plates; these experiments are denoted by (R2), which signifies the residue after two leach steps. A third leach step for two hours at 200°C in an oven was performed on several fractions, and these are denoted by (R3), signifying the residue after three leach steps. 3Denotes radiogenic Pb. 4Observed ratios collected on Faraday cups were corrected for 0.125% per unit mass fractionation, based on replicate analyses of NBS 981 and 983. Uncertainties in the 208Pb/206Pb and 207Pb/206Pb ratios are less than 0.1%; the uncertainty in the 206Pb/204Pb ratios are generally less than 5% (2σ). Isotopicdata measured on Finnigan-Mat MAT 262 multiple collector mass spectrometer at the U.S. Geological Survey, Menlo Park, California. 5Atomic ratios were calculated using the following constants: 238U/235U = 137.88; 235U = 0.98485 × 10-9yr-1; 238U = 0.155125 × 10-9yr-1. The ratios were corrected for common Pb ratios based upon Pb-isotopic compositions—208:207:206:204—measured in feldspars. Isotopic compositions have uncertainties of less than 0.1% (2a). ES-7458 = 38.440:15.607:18.524:1; IT-4 = 38.578:15.587:18.521:1; IT-5 = 38.454:15.599,18.492:1; IT-6 = 38.521:15.592:18.521:1; IT-9 = 38.421:15.590:18.519:1; IT-10 - 38.504:15.592:18.548:1; IT-17 = 38.382:15.589:18.439:1. No feldspars were recovered from IT-8; atomic ratios were corrected assuming average common Pb values of IT-4 and IT-6. For IT-3, the atomic ratios were corrected assuming common Pb values of ES-7458 collected from same unit. Corrected for a conservative laboratory procedural blank of 100 picograms of Pb; actual laboratory blanks range from 30 to 90 picograms. 6Uncertainties (2σ) (percent) in atomic ratios are shown in parentheses. 40 CORNEJO ET AL.

TABLE 4. U-Pb-Isotopic Data Determined on the SHRIMP1

206 239 Jrain.spot U, ppm Th,ppm Th/U Pb*, ppm 204Pb/206Pb 206Pb*/238U Uncert. Pb*/ U, Ma

Cerro Indio Muerto rhyolite dome (IT-8) 1.1 176 156 0.89 2 0.000001 0.0093 0.0003 59.4 ±1.8 2.1 282 314 1.11 3 0.002330 0.0094 0.0003 60.5 ± 1.6 3.1 234 249 1.06 3 0.001754 0.0091 0.0002 58.1 ± 1.6 4.1 283 409 1.44 3 0.000113 0.0088 0.0002 56.2 ± 1.5 5.1 117 143 1.22 1 0.002776 0.0088 0.0003 56.4 ± 1.7 6.1 156 227 1.46 2 0.000067 0.0091 0.0003 58.4 ± 2.2 7.1 527 5% 1.13 6 0.000672 0.0092 0.0002 59.2 ± 1.5 8.1* 118 93 0.79 1 0.000001 0.0068 0.0003 43.6 ±2.1 9.1* 124 82 0.66 1 0.000001 0.0063 0.0002 40.3 ± 1.4 10.1 127 124 0.98 1 0.005942 0.0088 0.0003 56.2 ±2.1 11.1 319 584 1.83 4 0.000001 0.0094 0.0003 60.3 ± 1.6 12.1 517 507 0.98 6 0.001066 0.0099 0.0002 63.4 ± 1.6 13.1 829 2354 2.84 13 0.001004 0.0094 0.0003 60.0 ± 1.7 14.1 113 164 1.45 1 0.009329 0.0089 0.0004 57.2 ± 2.6 15.1 307 651 2.12 4 0.000560 0.0095 0.0005 60.7 ± 2.9 16.1 234 421 1.80 3 0.002987 0.0092 0.0003 59.0 ±1.7 17.1 151 248 1.65 2 0.000001 0.0096 0.0004 1.7 ±2.2 Mean of 15 spots 59.1 ± 1.9

Old Camp porphyry (IT-5) 1.1 141 113 0.80 1 0.002584 0.0065 0.0002 41.4 ± 1.4 2.1 504 914 1.81 5 0.000001 0.0069 0.0002 44.6 ±1.1 3.1 832 387 0.47 6 0.000972 0.0064 0.0002 41.3 ±1.1 4.1* 81 29 0.36 1 0.000001 0.0057 0.0003 36.6 ± 1.7 5.1* 226 224 0.99 2 0.000001 0.0062 0.0002 39.5 ± 1.2 6.1 233 81 0.35 2 0.002745 0.0064 0.0002 41.0 ±1.1 7.1 456 460 1.01 4 0.001127 0.0067 0.0002 43.0 ±1.0 8.1 400 198 0.49 3 0.001930 0.0065 0.0004 42.1 ±2.5 9.1 1216 615 0.51 9 0.000405 0.0067 0.0002 43.0 ± 1.1 Mean of 7 spots 42.3 ±1.3

X porphyry (IT-10) 1.1 37 23 0.64 <1 0.0167% 0.0061 0.0003 39.1 ±2.2 2.1 27 11 0.39 <1 0.005398 0.0063 0.0004 40.7 ± 2.4 3.1 47 21 0.45 <1 0.000001 0.0066 0.0003 42.6 ± 1.7 4.1 25 13 0.51 <1 0.000001 0.0065 0.0004 41.7 ±2.2 5.1 37 21 0.56 <1 0.015542 0.0066 0.0005 42.6 ±3.1

Downloaded By: [Canadian Research Knowledge Network] At: 21:38 28 May 2011 6.1* 25 12 0.49 <1 0.005128 0.0055 0.0005 35.2 ± 3.3 7.1* 31 15 0.47 <1 0.000001 0.0059 0.0003 38.0 ± 1.9 8.1 75 42 0.56 1 0.001473 0.0066 0.0003 42.3 ± 2.0 9.1 40 14 0.37 <1 0.003950 0.0068 0.0004 43.9 ± 2.3 Mean of 7 spots 41.8 ±2.3

1SeeTable3 for location of samples. Mean values of spots exclude individual analyses, denoted by *, which are considered to be too young on the basis of geologic and chronologic data presented herein. Uncertainties are given at the one sigma level. Correction for common Pb made on the basis of extrapolation to concordia along a mixing line with common Pb, following Tera and Wasserburg (1972).

intrusions, from northeast to southwest, crop (1975), post-rhyolite dome magmatism in the out at Cerro Pelado, Old Camp, Quebrada "M," Eocene began with porphyritic quartz rhyolite Quebrada Turquesa, and Quebrada Granito. subvolcanic intrusions and possibly extrusive Country rocks are the Llanta Formation and rocks at Cerros Pelado and Riolita and related Paleocene rhyolitic pyroclastic rocks and quartz-plagioclase porphyry intrusions forming domes. According to Gustafson and Hunt an annular dike at Old Camp, an irregular EL SALVADOR COPPER DEPOSIT REVISITED 41

N-trending dike in Quebrada "M," and lac- sericitized plagioclase, quartz, and traces of colithic bodies in Quebrada Turquesa (Fig. 2). biotite in a fine-grained quartzofeldspathic These intrusive rocks, which are characterized groundmass. Feldspar phenocrysts from other by conspicuous quartz phenocrysts, were inter Eocene rocks in the district, in contrast, always preted to have been emplaced at 44 to 45 Ma consist of plagioclase. Sanidine phenocrysts are based upon whole-rock and model mineral Rb- typical of Paleocene rocks in the region, but Sr isochrons of altered rocks from Cerros quartz phenocrysts are not. Pelado and Riolita (ages recalculated by Cor- Altered quartz-sanidine rhyolite from Cerro nejo et al., 1993a), and a sericite K-Ar age on Pelado has a K-Ar alteration age on a phyllically phyllically altered quartz porphyry at Old altered whole-rock sample of 42.1 ± 1.2 Ma Camp. Minor Cu-Mo mineralization is asso (IP-11) and a slightly older sericite 40Ar/39Ar ciated with the quartz porphyry center at Old isochron age of 43.9 ± 1.5 Ma (McWilliams, Camp (Gustafson and Hunt, 1975). 1994, quoting 2a uncertainties). An older K-Ar On Cerro Indio Muerto, several granodiorite age of 45.3 ± 2.0 Ma (ES-7458) was obtained on and granodiorite-dacite porphyry stocks a sanidine concentrate containing partially intruded the quartz porphyries and older coun sericitized plagioclase. Five zircon fractions try rocks between 43 and 41 Ma. Granodioritic from the Cerro Pelado stock also were analyzed intrusions crop out in Quebradas "M," Tur by U-Pb methods; three fractions are from an quesa, and Granito. In Quebrada Turquesa, intensely altered sample (IT-3) and two addi Gustafson and Hunt (1975) described five suc tional fractions are from a less altered part of cessive intrusive phases of decreasing age, on the porphyry (ES-7458) that also was dated the basis of cross-cutting relations, as "X," using K-Ar methods (see above). Three zircon "K," "L," and "A" porphyries, and post-min fractions were experimentally hydrothermally eral latite or dacite porphyry dikes in most leached to remove the abundant common Pb. chemical classifications (Cox et al., 1979; Le All zircon fractions have discordant U-Pb ages Bas et al., 1986). The main porphyry Cu-Mo (Table 3; Fig. 7A), although four analyses over hypogene mineralization event in the Indio lap concordia within the limits of their analyti Muerto district is associated with the X and K cal uncertainty. U-Pb ages for zircons from porphyries (Gustafson and Hunt, 1975), sample IT-3 increase with increasing grain size whereas the younger L and A porphyries are and decreasing uranium content, a pattern typi intra-mineral, with less alteration and contain cal of younger Pb loss and the presence of ing less sulfide. Closely associated with the inherited zircons. It is likely that Pb also has post-mineral dacite dikes are pebble dikes. been lost from zircons from sample ES-7458, Granodiorite porphyry in Quebrada "M" although the evidence is weaker, since these includes rocks lithologically similar to the L fractions were experimentally leached in order and A porphyries of Quebrada Turquesa. Like to minimize the effect of the young Pb loss Downloaded By: [Canadian Research Knowledge Network] At: 21:38 28 May 2011 wise, the granodiorite porphyry in Quebrada (Mattinson, 1994). The >100-micron zircons Granito is similar to the L porphyry. from sample IT-3 have discordant U-Pb ages Quartz-sanidine rhyolite porphyries displaced away from concordia, confirming the presence of small amounts of inherited zircons. Quartz-sanidine rhyolite porphyry—the A line through this fraction and the >80- "quartz rhyolite" of Gustafson and Hunt micron fraction suggests an age of 43 ± 1 Ma for (1975)—underlies Cerros Pelado and Riolita. the Cerro Pelado stock, which agrees with other On Cerro Pelado, it is a subvolcanic stock with ages for the stock within the limits of their sparse concentric and radial dikes and breccias. The Cerro Riolita body has a subhorizontal analytical uncertainties. The inherited compo base, which Gustafson and Hunt (1975) suggest nent is poorly constrained to be Proterozoic or might be the paleosurface over which the rock Paleozoic (1335 ±981 Ma). was extruded, or the base of a sill. The lack of a The chronologic data confirm an Eocene age basal breccia suggests that the sill hypothesis is for the Cerro Pelado stock. No new chronologic more probable. Both porphyries are crystal poor data are available for the rhyolite on Cerro with scarce small phenocrysts of sanidine, Riolita, as no zircons were recovered from the 42 CORNEJO ET AL.

FIG. 7. A portion of the concordia diagram showing U-Pb geochronologic data for zircons from Eocene rhyolite and granodiorite porphyries. A. Phyllically altered quartz-sanidine rhyolite porphyry on Cerro Pelado (IT-3; ES-7458). As Pb toss is evident from these zircons, the chord is drawn only through the zircon fractions with the oldest U-Pb ages. B. Biotite-bearing quartz-plagioclase porphyry (IT-5) that forms the Old Camp dike. Chord is regressed only through the U-Pb data for normally discordant zircons (filled squares). Reversely discordant zircons (stippled diamonds) overlap the chord within their analytical uncertainties. C. X porphyry (IT-10) collected from the 2445-m level in the El Salvador mine. D. L porphyry (IT-9) collected from the 2445-m level in the El Salvador mine. Downloaded By: [Canadian Research Knowledge Network] At: 21:38 28 May 2011 rock, but, based upon petrologic and chemical Quartz-plagioclase rhyolite porphyries criteria (see below), it is assigned a Paleo- Quartz-plagioclase rhyolite porphyries—the cene(?) age. Early crystallization of K-feldspar "quartz porphyry" of Gustafson and Hunt as sanidine indicates high-temperature crystal (1975)—are crystal-rich, characterized by lization of a K2O-rich, H2O-poor magma, whose abundant large phenocrysts of quartz, sodic anhydrous bulk composition impeded the early plagioclase, and conspicuous biotite in a fine crystallization of amphibole and biotite (Naney, grained siliceous groundmass. Fresh rocks are 1983). The dominance of K-feldspar pheno- not present, but the mineralogy and two whole- crysts in the Cerro Pelado porphyry, therefore, rock chemical analyses of altered rock (Gus suggests that some Eocene silicic magmas in the tafson and Hunt, 1975; P. Cornejo, unpubl. district were more anhydrous than the hydrated data, 1994) suggest they are rhyolitic or granodiorite porphyry magmas on Cerro rhyodacitic in composition (67 to 71 wt%

Indio Muerto, where both of the hydrous min SiO2). The quartz-plagioclase porphyry at Old erals, biotite and amphibole (biotitized), are Camp forms a semicircular dike centered on the ubiquitous. Cerro Pelado stock (Gustafson and Hunt, EL SALVADOR COPPER DEPOSIT REVISITED 43

1975), suggesting that the Old Camp body was emplaced first, the water content of the intruded along a ring fracture after emplace parent magma must have increased subse ment of the Cerro Pelado stock had strained and quently, most likely in response to fractional fractured the wall rock. Gustafson and Hunt crystallization of anhydrous phases, thereby (1975) also stressed the irregular dike and sill stabilizing biotite over K-feldspar as the shapes of the quartz-plagioclase porphyry on crystallizing K-rich phase (Naney, 1983; Dilles, Cerro Indio Muerto as linking it more to the 1987). Following fractional crystallization and quartz-sanidine rhyolite porphyry than to the evolution toward a more hydrous magma, younger steep-walled granodioritic intrusions. renewed upward intrusion emplaced the plagio- The Old Camp dike has K-Ar ages of 43.9 ± clase- and biotite-rich Old Camp dike along the 1.5 Ma on whole-rock (IP-10) and 42.1 ± 2.6 concentric fracture system produced by its Ma on sericite (ES-6025B; Table 2), and a predecessor, the Cerro Pelado stock. 40Ar/39Ar isochron age of 43.9 ± 0.5 Ma on Of the other quartz-plagioclase porphyries in biotite (McWilliams, 1994); these ages are con the district, only the porphyry in Quebrada sistent with ages for Cerro Pelado. Five frac "M" has been dated. Here, biotite has a tions of zircons from the Old Camp dike (IT-5) 40Ar/39Ar isochron age of 42.6 ± 0.3 Ma were dated using U-Pb methods; three fractions (McWilliams, 1994). Ages of quartz-plagioclase were analyzed conventionally and two fractions laccolithic bodies in Quebrada Turquesa are underwent hydrothermal leaching experiments unknown, except that they are older than the X (Table 3; Fig. 7B). One fraction is concordant porphyry (Gustafson and Hunt, 1975). with U-Pb ages of 42.9 Ma. Two other fractions are normally discordant, whereas the other two Granodiorite, granodioritic-dacitic porphyry are slightly reversely discordant. Based upon stocks, and dacitic porphyry dikes the concordant fraction, an age of 43 ± 1 Ma Rocks of this suite have a similar mineralogi- age is assigned to the sample. SHRIMP U-Pb cal association dominated by phenocrysts of ages from tips of zircons have a mean plagioclase, hornblende, biotite, and occasion 206 238 Pb*/ U age of 42.3 ± 1.3 Ma (Table 4). ally quartz. The presence, amount, grain size, Discordance in the bulk zircon fractions is the texture, and, to a lesser extent, composition of result of varying amounts of inherited zircon the groundmass vary considerably between and cores, which were avoided during the SHRIMP within individual intrusions (Gustafson and analyses. Regressing the normally discordant Hunt, 1975). The X porphyry is a weakly por- fractions along with the concordant fraction phyritic to equigranular granodiorite, on the indicates that the inherited component has an basis of which Gustafson and Hunt (1975) average age of 819 Ma with an unrealistic uncer distinguished it from the younger, more tainty caused principally by the distribution of strongly porphyritic intrusions, which they data points at the lower end of the chord. The referred to as the "feldspar porphyries" (K, L, two reversely discordant zircon fractions over and A). The K and L porphyries have a ground-

Downloaded By: [Canadian Research Knowledge Network] At: 21:38 28 May 2011 lap the discordia array within their analytical mass dominated by aplitic quartz, alkali-feld uncertainties, and are viewed as having been spar, and biotite, whereas the later A porphyry displaced from the chord as a result of some and post-mineral dacite porphyry dikes have analytical problems. groundmass dominated by sodic plagioclase The quartz-bearing porphyries on Cerro laths and abundant mafic minerals, suggesting a Pelado and at Old Camp are essentially of the more mafic composition. The X, K, and L all same age. Their age similarities, the annular have a similar mineralogy, suggesting similar dike form of Old Camp about Cerro Pelado, and compositions, although the X and K are too their bulk compositional similarities support altered to yield reliable chemical data. The L the interpretation that they may be related and porphyry has a granodioritic-dacitic chemical possibly cogenetic bodies. If so, their miner- composition (Gustafson and Hunt, 1975; P. Cornejo, unpubl. data, 1994), whereas the A alogical differences (sanidine dominant versus porphyry and post-mineral dacite dikes appear plagioclase-biotite dominant) require some to have more mafic dacitic or andesitic chemical chemical evolution within the parent magma. composition (Gustafson and Hunt, 1975), con Since magmato-structural relations suggest sistent with their petrography. that the Cerro Pelado quartz-sanidine rhyolite 44 CORNEJO ET AL.

Because of the intense alteration in the X and discordant 206Pb*/238U, 207Pb*/235U, and K porphyries, primary crystallization ages for 207Pb*/206Pb* ages of 41.9 Ma, 42.1 Ma, and 53 these rocks have not been realized until this to 54 Ma, respectively. One conventional frac study. K-Ar and 40Ar/39Ar ages only date altera tion gave slightly reversely discordant U-Pb tion associated with the porphyry deposit. In ages of 41.8 Ma and 41.7 Ma. The other fraction order to date the onset of granodioritic magma- has younger U-Pb ages of 40.1 Ma and 40.4 Ma, 206 tism in the mine, a sample of X porphyry but the 207Pb*/ Pb* age of 58 Ma is close to collected on the 2445-m level within the mine the age of the leached zircons. There is insuffi was dated using U-Pb methods. Two hydrother- cient spread in the analytical data to define or to mally leached zircon fractions (<63 micron) model a discordia array, but several interpreta (IT-10) have discordant U-Pb ages (Table 3; Fig. tions can be made. One is that at least the 7C). A line through the two points intersects youngest zircon fraction seems to have lost concordia at 41 ± 2 Ma and 389 + 348 Ma. The some Pb, and Pb loss may affect the U-Pb ages upper intercept indicates the presence of an of the other fractions. The consistent inherited zircon component of Paleozoic or 207Pb*/206Pb* ages (between 53 and 58 Ma) of older age. The lower intercept age is inferred to three fractions imply that part of the discordia be the age of the rock. SHRIMP U-Pb isotopic pattern must also result from the presence of data from tips of seven of nine zircon grains inherited zircons. A K-Ar age of 41.2 ± 1.1 Ma have a similar mean 206Pb*/238U age of 41.8 ± has been determined on igneous biotite 2.3 (Table 4). Fine-grained hydrothermal biotite recovered from the same sample as the zircons from the same sample has a K-Ar age of 41.6 ± (IT-9), and Gustafson and Hunt (1975) 1.2 Ma (IT-10). A K-Ar age on sericite of 41.9 + reported an older but imprecise K-Ar horn 1.0 Ma (ES-3256; Table 2) was obtained from blende age of 43.2 ± 5.6 Ma (ES-3442A, Table the younger K porphyry (Gustafson and Hunt, 2) for the L porphyry in the El Salvador mine. 1975). McWilliams (1996) reported 40Ar/39Ar Collectively, the U-Pb and K-Ar data are inter isochron ages from hydrothermal biotite and preted to indicate an age of 41 ±2 Ma for the L sericite in the X and K porphyries ranging from porphyry. 41.4 ± 1.2 Ma to 40.4 ± 1.3 Ma. K-Ar alteration ages from country rocks include a 41.8 ± 2.6 The post-mineral dacite dikes are indis Ma age (ES-1910; Table 2) from biotitized ande- tinguishable in age from the granodiorite por site of the Llanta Formation (Gustafson and phyries. Gustafson and Hunt (1975) reported Hunt, 1975) and a 40.2 ± 1.2 Ma whole-rock an age of 42.0 ± 1.0 Ma (ES-4269, Table 2) from one dike, and biotite from another dike in the age from sericitized Paleocene dome rock 40 39 (recalculated from Quirt [1972] and Zentilli Quebrada Turquesa area has a Ar/ Ar iso [1974]). The U-Pb zircon age, inferred to be the chron age of 41.2 ± 0.5 Ma (McWilliams, crystallization age of the X porphyry, and the 1994). Collectively, the geochronologic data for 40Ar/39Ar and K-Ar ages on alteration minerals the granodioritic complex at Quebrada Tur quesa indicate emplacement and mineralization

Downloaded By: [Canadian Research Knowledge Network] At: 21:38 28 May 2011 are essentially identical, thereby confirming a close temporal connection between the intru between 42 and 41 Ma, probably over a period sion of early granodiorite stocks and the follow <1 m.y. ing hydrothermal alteration associated with the Geochronologic data from Quebradas "M" porphyry Cu-Mo deposit (Gustafson and Hunt, and Granito (Fig. 4) suggest that granodioritic 1975). porphyries have ages similar to both the older 44- to 43-Ma quartz rhyolites and the 42- to 41- The L porphyry, the largest granodiorite- Ma Quebrada Turquesa granodioritic complex. dacite porphyry stock, is younger than most of In Quebrada Granito, a granodiorite porphyry the alteration and mineralization in the deposit, stock, petrologically similar to the L porphyry based on cross-cutting relations (Gustafson and of Quebrada Turquesa, has K-Ar ages of ~43 Ma Hunt, 1975). Two zircon fractions from the L (ES-3853 and 3853A, Table 2) (Gustafson and porphyry (IT-9), where it intrudes the X por Hunt, 1975). 40Ar/39Ar analyses of hornblende phyry (IT-10), were analyzed conventionally, and biotite, also from ES-3853, have isochron whereas two other samples underwent experi ages of 42.3 ± 0.5 Ma and 42.9 ± 0.3 Ma, mental hydrothermal leaching (Fig. 7D). respectively (McWilliams, 1994). At a deep Leached fractions yielded identical but slightly level within the Quebrada "M" open pit, a dark- EL SALVADOR COPPER DEPOSIT REVISITED 45

colored dacite porphyry texturally similar to A main-stage Cu-Mo mineralization is associated porphyries in Quebrada Turquesa, but having with the early intrusions and that petrological large biotite phenocrysts, has a biotite K-Ar age variations and limited chemical data suggest a of 43.8 ± 1.2 Ma (IP-45). Nearby, two different trend toward more intermediate compositions textural varieties of porphyry, also similar to L over time. This trend runs counter to the trend and A porphyries, crop out. The fine-grained in many multiple-intrusion porphyry copper variety has a K-Ar age on biotite of 41.1 ± 1.3 complexes (Titley and Beane, 1981), where a Ma (ES-12337), and a coarse-grained variety mafic or intermediate to silicic compositional has an identical K-Ar age on biotite of 41.2 ± trend is most commonly observed and main- 1.3 Ma (ES-12338). From Colina de Cobre, just stage mineralization is associated with the later to the northeast of Quebrada "M" (Fig. 4), Gus- silicic magmas (Dilles, 1987; Casselman et al., tafson and Hunt (1975) reported a biotite K-Ar 1995). These compositional trends are inter age of 41.6 ± 1.8 Ma (ES-6136; Table 2) from preted to reflect differentiation in a closed another granodiorite porphyry, and between system, with the volatiles necessary for extract Quebradas "M" and Turquesa, McWilliams ing and transporting metals being concentrated (1996) reported a biotite 40Ar/39Ar isochron age in the later silicic magma. Given the available of 40.9 ± 0.2 Ma on a granodiorite porphyry. data, the main El Salvador porphyry system Biotite from still another granodiorite porphyry does not easily fit these models. Instead the El in Quebrada "M" has a K-Ar age of 38.3 ± 1.2 Salvador history suggests derivation either Ma (ES-12339), which is considered too young from a zoned magma chamber or more likely in view of the chronologic data for petro- from an open-system magma chamber, as sug logically similar rocks. gested by mineralogic data cited by Clark (1993), where the silicic magma chamber was Most of the granodioritic porphyries dated at periodically injected by mafic magmas at its Quebrada "M" have ages concordant with their base. As the mafic magma cooled and crystal equivalents in Quebrada Turquesa, suggesting lized, water, sulfur, and chlorine were exsolved that they are comagmatic. However, the biotite- and concentrated in the overlying silicic magma phyric A porphyry in Quebrada "M" and the L (Andres et al., 1991; Matthews et al., 1994a, porphyry in Quebrada Granito have slightly 1994b, 1995). Successive upward intrusion of older, 44- to 43-Ma ages that are similar to ages granodioritic porphyry magmas, perhaps trig of quartz rhyolite intrusions, although the ages gered by mafic magma injection into the magma are within the analytical uncertainties of much chamber (Sparks et al., 1977), provided a focus of the age data from the younger granodioritic- for the flow of exsolved hydrothermal fluids dacitic porphyries. If the porphyries in (Lowenstern, 1994). In this hypothesis, exsolu- Quebradas "M" and Granito are indeed 44 to 43 tion of the ore-forming fluid from the silicic Ma, it suggests that high-level stocks of a variety magma could occur at any stage of the magma of compositions were emplaced approximately chamber history, which in Quebrada Turquesa contemporaneously and were derived either Downloaded By: [Canadian Research Knowledge Network] At: 21:38 28 May 2011 was early. The younger, more mafic porphyry from different parent magmas or as a comag magmas in Quebrada Turquesa reflect either matic suite that tapped different parts of an the tapping of deeper, less evolved parts of the evolving magma chamber. silicic magma chamber or more efficient mixing Potential significance of compositional trends for of the end-member magmas (e.g., Matthews et porphyry copper formation at El Salvador al., 1994a). The close temporal and compositional sim If the A porphyry in Quebrada "M" and the L ilarity of the various granodioritic intrusions in porphyry in Quebrada Granito bridge the age Quebrada Turquesa supports the argument that gap between the older (44 to 43 Ma) quartz they form a cogenetic suite. The quartz rhyo- rhyolite intrusions and the younger (42 to 41 litic intrusive rocks are 1 to 3 m.y. older than Ma) granodiorite porphyries, the entire Eocene the Quebrada Turquesa granodioritic complex, suite could be comagmatic, in which case the a temporal difference suggesting that they main El Salvador copper deposit formed rela could be separate magmatic systems. It is inter tively late in the magmatic history. Considering esting to note that in the history of the the ~3-m.y. time period involved, it seems Quebrada Turquesa granodioritic complex, the unlikely that the entire suite is related via a 46 CORNEJO ET AL.

shallow-level, closed-system magma chamber. If contradicts many models of Andean orogenesis the suite is comagmatic, then the older quartz that argue for short periods of intense shorten rhyolitic and granodioritic-dacite intrusions ing deformation along the arc (e.g., Megard et and younger granodiorite porphyries must be al., 1984; Sebrier et al., 1988; McKee and related by a large, deep magma chamber or by a Noble, 1990). shallow magma chamber that was recharged Eocene shortening also is evident west of the continually by injection of mafic magma, Sierra Castillo fault. The large-amplitude syn- thereby adding the necessary thermal energy cline in the Cerros Los Amarillos area (Fig. 5C) required to sustain an epizonal magma chamber formed at this time, as well as reverse-slip over as much as 3 million years. In this hypothe reactivation of the steeply dipping Kilometro sis, formation of more-mafic A-type porphyry Catorce fault. magmas could have been repeated at various Within the Indio Muerto district, an appar times in the history of the long-lived parent ent structural conflict is present between the magma chamber through mixing of fractionat prominent NNE alignment of the Eocene por ing silicic magma with periodic injections of phyry centers and the dominance of a NW- mafic magma, thus accounting for the —44- to striking structural grain in the El Salvador mine 43-Ma A porphyry in Quebrada "M" and the 42- at Quebrada Turquesa. Inside the mine, the to 41-Ma A porphyry in Quebrada Turquesa. A steep-walled granodiorite porphyry complex return to more-silicic compositions would have has a distinct northwest elongate form with an followed periods of stasis (diminished mafic aspect ratio of ~2.5:1 (Gustafson and Hunt, magma injection and/or inefficient mixing) 1975; CODELCO, unpubl. mapping, 1977). and differentiation in the parent silicic magma Likewise, the ~41-Ma post-mineral dacite por chamber (Matthews et al., 1994a). phyry dikes and pebble dikes at deep levels in Regional and local structural setting of Eocene the mine are predominantly NW striking, with El Salvador magmatism a mean direction of ~310°. The parallelism of these elements to the Eocene shortening direc The principal structural element in the El tion inferred from the Potrerillos geology is Salvador-Potrerillos region is the Sierra Cas tillo fault (Fig. 3). This fault, originally a Juras compatible with their emplacement during sic normal fault, was active as a sinistral strike- NW-SW-directed shortening and indicates a slip fault in the Eocene (Tomlinson et al., 1993; control on the mine-scale magmatism by Mpodozis et al., 1994). East of the fault, a fold- regional transpressive strain. In contrast, the and-thrust belt formed during sinistral trans- NNE alignment of Eocene porphyry centers pression in response to NW-SE (~305° to subparallel to basement faults and the inferred 125°)-directed shortening (Fig. 3) (Tomlinson hinge orientation of the Paleocene trap-door et al., 1993). Deformation had begun by 42 Ma caldera suggests that preexisting structures south of Potrerillos, where hypabyssal intru controlled district-scale distribution of Downloaded By: [Canadian Research Knowledge Network] At: 21:38 28 May 2011 sions and rhyolitic domes were emplaced in part magmatism. along NW-striking (330° to 335°) sinistral strike-slip faults striking at high angles to the NNE-striking (010°) Sierra Castillo fault. These sinistral strike-slip faults represent Petrologic and Geochemical Distinctions Riedel shears formed during regional transpres- Several features distinguish Paleocene from sion (Sanderson and Marchini, 1981; Tomlin Eocene rocks in the El Salvador region. Paleo son et al., 1993). Transpressive deformation cene rocks are high-K, calcalkaline rocks typ continued at least until ~36 Ma when the ified by sanidine and pyroxene phenocrysts, but syntectonic Potrerillos Cobre porphyry was only scarce quartz, biotite, and rarely horn emplaced (Tomlinson, 1994; Marsh et al., blende phenocrysts. They are inferred to have 1995). Deformation ended by 32 Ma when the been derived from relatively anhydrous post-tectonic El Hueso stock was intruded into magmas. In contrast, Eocene rocks are moder the Potrerillos fold-and-thrust belt (Fig. 3; ate to high-K, calcalkaline rocks commonly Tomlinson et al., 1993). Contractile deforma containing significant hornblende and biotite, tion over 6 to 10 m.y. in the Potrerillos area and only rarely containing alkali feldspar as a EL SALVADOR COPPER DEPOSIT REVISITED 47

have equilibrated with a source containing garnet, typical of a higher-pressure residual mineralogy and a thicker crustal column (Mpodozis et al., 1995; Lopez-Escobar, 1982). The intermediate La/Yb of the Paleocene rocks suggest an intermediate crustal thickness at that time. The Cretaceous to Eocene trend is similar to that observed for the late Tertiary volcanic evolution of the Maricunga belt, southeast of El Salvador, which is interpreted to reflect magma generation and eruption through crustal thickening over time (Kay et al., 1994; Mpodozis et al., 1995). Smooth REE patterns for Eocene intrusions at El Salvador lack a Eu anomaly (Gustafson, 1979; Lopez-Escobar, 1982). In contrast, REE patterns for Paleocene rhyolites have a negative Eu anomaly, indicative of plagioclase fractionation during the history FIG. 8. La/Yb versus SiO2 for Upper Cretaceous and of the magma (Gustafson, 1979; Lopez-Escobar, Paleocene volcanic and plutonic sequences (after Cornejo 1982). Gustafson (1979) also reported an REE et al., 1994) and for Eocene granodiorite porphyries at El pattern with a negative Eu anomaly and low La/ Salvador. Paleocene volcanic rocks at El Salvador are part of Yb for a quartz-sanidine rhyolite on Cerro the Cerro Valiente sequence. Chemical data from Gustafson (1979), Lopez-Escobar (1982), and Cornejo et al. (1994). Riolita (L.B. Gustafson, pers. commun., 1997), The Cerro Valiente point with anomalously high La/Yb suggesting that this porphyry is Paleocene. corresponds to the quartz monzonite at La Cantera, which Despite chemical differences, limited radio also has an anomalously low feldspar 206Pb/204Pb value genic isotopic data show no or at best only compared to the feldspar Pb isotopic composition of other Paleocene rocks in the area. suggestive differences. Feldspar Pb-isotopic compositions for Paleocene volcanic rocks on phenocryst phase. These melts were more Cerro Indio Muerto are indistinguishable from hydrous, containing sufficient water to form those of Eocene granodiorite porphyries at El porphyry copper systems. Another difference Salvador (see Table 3 footnotes). An exception between the two suites is the slightly broader is the feldspar Pb-isotopic compositions from compositional range represented by the Paleo the Paleocene quartz monzonite intrusion, which has lower 206Pb/204Pb values, suggesting cene rocks (53 to 78% SiO2) compared to the a much greater range in Pb-isotopic composi Eocene rocks (57 to 75% SiO2) (Cornejo et al., 1993a). tions of Paleocene rocks than of Eocene rocks, Downloaded By: [Canadian Research Knowledge Network] At: 21:38 28 May 2011 Trace elements also vary between suites. For a difference also noted in the porphyry Cu belt example, La/Yb ratios, indicative of the farther north (Williams, 1995). Initial Sr iso residual mineralogy within the magmas' source topic values of Paleocene and Eocene rocks also region (Kay and Kay, 1989; Kay and Abbruzzi, have been inferred to be indistinguishable at 1996), show distinctions with age in the El ~0.7042 (Gustafson and Hunt, 1975; Gus Salvador region. Upper Cretaceous bimodal, tafson, 1979). However, we question the initial high-K, calcalkaline rocks have La/Yb of 6 to Sr values for Paleocene rhyolitic rocks. As Rb- 11, Paleocene rocks have La/Yb of 11 to 19, and Sr ages for these rhyolitic rocks are incorrect, La/Yb in the Eocene porphyries at El Salvador initial Sr-isotopic values derived from iso- range from 20 to 25 (Fig. 8). These trace- chrons must also be suspected until initial Sr element compositions are inferred to indicate values are determined for unaltered rocks. that Late Cretaceous magmas equilibrated with Slightly contaminated or enriched mantle Nd- ε a residual mineralogy characterized by pyrox isotopic values ( Nd = +3 to +1) are known for a ene beneath a thinned continental crust (Corn couple of Eocene stocks at El Salvador (Mak- ejo et al., 1994; Mpodozis et al., 1995). On the saev, 1990; Zentilli et al., 1994), but equivalent other hand, Eocene porphyries are inferred to data is unavailable for any Paleocene rocks. 48 CORNEJO ET AL.

Felsic rocks in both Paleocene and Eocene m.y., from the emplacement of the Paleocene suites are characterized by small quantities of Indio Muerto rhyolitic dome at ~58 Ma to the inherited older zircons. As no xenocrystic zir Eocene quartz rhyolite porphyries at 44 to 43 cons were noted, these older zircons must be Ma. On the regional scale in the El Salvador present as core with new magmatic over area, the magmatic gap is less pronounced, growths. This requires magma interaction with spanning only 5-6 m.y. for units between 26° crustal rocks whose ages are only loosely con and 27°S (Fig. 9). Farther north, the regional strained to be Paleozoic or Proterozoic(?). In temporal gap of magmatism varies from a max this context, Eocene porphyries at El Salvador imum of~11 m.y. between Quebrada Blanca are similar to Eocene and Oligocene porphyries and Chuquicamata (20° 30' to 22° 30' S) (Fig. associated with Cu-Mo deposits at La Escon- 1) to a minimum in the Exploradora area (25° dida, Chuquicamata, and Potrerillos, which to 26° S) where a magmatic gap is not apparent. have inherited Paleozoic zircons (Zentilli et al., Despite the lack of a magmatic gap in the 1994; R.M. Tosdal, unpubl. data, 1995). Crustal Exploradora area, the petrology of pre- and Pb also is evident within rocks associated with post-48-Ma igneous rocks are distinct, and simi these porphyry Cu-Mo deposits, as well as lar to differences observed in El Salvador and within sulfide minerals contained therein (Zen elsewhere between Paleocene and Eocene rocks tilli et al., 1988; Williams, 1995; Tosdal, 1995). (Cornejo and Mpodozis, 1996). The temporal Re-Os-isotopic data for sulfides from the late breaks also are accompanied by changes in the Miocene to Pliocene El Teniente and Andina regional distribution of magmatism, with the (La Disputada) porphyry Cu deposits in central magmatic front jumping to the east in the Chile indicate crustal involvement during gen Eocene, although there is considerable overlap esis of those deposits (Ruiz et al., 1996, 1997), in the magmatic belts as shown by the superim- and it is likely that similar results will be position of Paleocene and Eocene magmatism obtained at El Salvador in view of the remark in the Indio Muerto district. The migration of able isotopic homogeneity of Cenozoic Andean the magmatic front is not gradual but sudden, porphyry Cu deposits (Zentilli et al., 1988). with the Paleocene and Eocene magmatism Crustal involvement in porphyry Cu deposits, being relatively static before and after, a feature indicated by U-Pb-, Pb-, and Re-Os-isotopic that is characteristic of the entire Mesozoic and data, complicates models that derive Eocene Cenozoic magmatic history of the Andes (Coira and Miocene to Pliocene porphyry magmas and et al., 1982; Sillitoe, 1988; Mpodozis and metals in Cu-Mo deposits from mantle sources Ramos, 1989). without any or only limited crustal interaction, as suggested by enriched mantle-like Sr- and Nd-isotopic data (Zentilli et al., 1988, 1994; Clark, 1993; Skewes and Stern, 1995). Resolu Discussion and Conclusions tion of this paradox lies beyond the scope of this Downloaded By: [Canadian Research Knowledge Network] At: 21:38 28 May 2011 Revisions to the stratigraphy of the Indio Muerto paper. district The revised stratigraphic and chronologic framework in the Indio Muerto district docu Regional Temporal Gaps in Paleocene mented herein is shown in Figure 2. Gustafson and Eocene Magmatism and Hunt (1975) presumed that the oldest units Stratigraphic and geochronologic data in the district, outcropping north and east of throughout northern Chile indicate that Paleo Cerro Indio Muerto (Fig. 4), were Upper Creta cene to lower Eocene and middle Eocene to ceous. These rocks now are known to include lower Oligocene magmatic episodes are region Jurassic, Lower Cretaceous, and Upper Creta ally discrete in time and space, and in many ceous rocks (Fig. 4). The supra-adjacent "Hor- places are separated by a significant temporal nitos" unconformity of Gustafson and Hunt gap in magmatism, further attesting to their (1975) is confirmed to be of latest Cretaceous distinction as unrelated events. In the Indio and earliest Paleocene age, but it is not correla Muerto district, the geochronological data indi tive with the Hornitos unconformity proper of cate a temporal gap spanning approximately 14 the Copiapó area, which is of Late Cretaceous EL SALVADOR COPPER DEPOSIT REVISITED 49

age (>78 Ma) (Arévalo et al., 1994; Arévalo, of andesite and trachyandesite flows and 1995). pyroclastic rocks. Within the El Salvador area, For the Tertiary stratigraphy, the most signifi explosive rhyolitic magmatism at the El Sal cant revision involves demonstrating that the vador trap-door caldera and post-collapse high- Indio Muerto rhyolite domes and, by associa K rhyolitic domes were contemporaneous with tion, the "Indio Muerto series" volcanics, are trachyandesitic volcanism from inferred strato ~8 to 10 m.y. older than indicated by the Rb-Sr volcanos. The post-collapse rhyolitic domes of ages of Gustafson and Hunt (1975) and only Cerro Indio Muerto were emplaced near the slightly younger than the rhyolitic ignimbrite intersection of two latest Cretaceous and/or sequence that they unconformably overlie (Fig. earliest Paleocene faults and the trap-door 2). Clark et al. (1985) suggested that the "Indio hinge of the El Salvador caldera. Pyroclastic Muerto" unconformity at the base of the and epiclastic rocks derived from the domes "Indio Muerto series" volcanics and rhyolite were deposited in an irregular topography domes was correlative with the "Cumbre" ero- resulting from synvolcanic faulting during cal sional surface of the Copiapó area, and there dera collapse. Andesitic and trachyandesitic fore was of regional extent related to regional lavas and sills derived from volcanic activity to uplift and erosion. Our data, however, argue the west may once have accounted for the that the unconformity is of local extent and kilometer or so of stratigraphic cover that is related to block faulting and tilting during col inferred to have been eroded from the Indio lapse of the El Salvador caldera. Muerto area since emplacement of the Eocene The chronology of the Eocene "quartz porphyry complex. rhyolite" and porphyry units of Gustafson and After a period of magmatic and tectonic Hunt (1975) is not substantially changed (Fig. quiescence of ~14 m.y., magmatic activity 2), except that the quartz rhyolite unit is con resumed in the Indio Muerto district at ~44 sidered to be a composite unit with the Cerro Ma, approximately contemporaneous with Riolita sill assigned a Paleocene(?) age and the onset of regional transpression along the sin Cerro Pelado stock being of Eocene age. The istral Sierra Castillo fault. Various rhyolitic and new chronologic data reported herein support granodioritic-dacitic porphyry stocks and dikes Gustafson and Hunt's (1975) original inter intruded over a period of ~3 m.y. along a NNE pretation of a close relationship between the trend that intersects the core of the Paleocene Cerro Pelado quartz-sanidine rhyolite stock and rhyolitic dome complex, apparently using the quartz-plagioclase rhyolite porphyries. In addi same channel of magma ascent followed by the tion, the chronologic data confirm the original older dome magmas. Initial magmatism consists K-Ar age determinations of Gustafson and Hunt of quartz-sanidine and quartz-plagioclase rhyo (1975) concerning the age of the younger litic porphyry stocks, dikes, and laccoliths granodioritic and dacitic porphyries. emplaced at 44 to 43 Ma. The first Cu-Mo Downloaded By: [Canadian Research Knowledge Network] At: 21:38 28 May 2011 Geologic framework of the El Salvador area mineralization in the district, at Old Camp, although minor, is associated with this early The Indio Muerto district and El Salvador porphyry Cu-Mo deposit have two parts that are magmatism. The first granodioritic-dacitic por related only spatially. The older magmatic epi phyries seemingly also were emplaced contem sode (~63 to 58 Ma) in the Indio Muerto poraneously in Quebradas Granito and "M," district consists of volcanic and subvolcanic but their petrogenetic relation with the quartz rocks that formed during a period of high-K porphyries and younger granodioritic magma Paleocene magmatism (Cornejo et al., 1993b, tism is unclear. The youngest magmatism in the 1994). Regionally, this period is characterized district, the Quebrada Turquesa granodioritic- by numerous volcanic centers possessing dif dacitic porphyry complex and granodioritic- ferent eruptive styles and magmatic composi dacitic stocks in the Quebrada "M" area, tions, including collapse calderas associated intruded over less than 1 m.y., between 42 and with explosive rhyolitic magmatism, rhyolitic 41 Ma. The main porphyry Cu-Mo mineraliza dome fields with associated outflows and tion in the district is associated with early pyroclastic rocks, and stratovolcanos composed intrusions of the Quebrada Turquesa complex. 50 CORNEJO ET AL.

FIG. 9. Chart showing, by region, the duration of magmatic events for the latest Cretaceous-early Oligocene, and age of individual Eocene-Oligocene porphyry copper deposits. Diagonally ruled area is the period of porphyry copper deposition. Based on original data and compilations of prior work summarized in Cornejo et al. (1993a), Mpodozis et al. (1993),Marinovic et al. (1995,1996), Marsh et al. (1995), Cornejo and Mpodozis (1996),Tomlinson et al. (unpubl. data), and the present study. Time scale after Gradstein and Ogg (1996).

The various stocks of the Quebrada Turquesa Protracted precursor magmatism and porphyry granodioritic complex probably are comagmatic copper formation and were derived successively from a deeper, yet For the El Salvador copper deposit, precursor unexposed, magma chamber, a feature common magmatism lasted no more than 2 to 3 m.y., to many porphyry copper systems. However, which may or may not have been important in unlike many other multiple-intrusion porphyry the formation of the main deposit. As the first copper complexes, at El Salvador there is a Cu-Mo mineralization in the district is associ trend toward more silica-poor units during the ated with one of oldest Eocene intrusions (the waning stages of plutonism. This trend may 43- to 44-Ma Old Camp quartz porphyry), pro reflect the tapping of deeper, more mafic parts tracted magmatism seemingly is not required Downloaded By: [Canadian Research Knowledge Network] At: 21:38 28 May 2011 of a vertically zoned magma chamber or intru for generation of a porphyry Cu-Mo deposit. sion following magma mixing in an open-system Thus a conclusion is consistent with observa silicic magma chamber with mafic magmas tions in districts where porphyry copper sys injected into the base. tems are associated with single, isolated, The petrological distinctions, together with compositionally homogenous intrusions (Titley the temporal gap and spatial reorganization in and Beane, 1981). Another observation is the magmatism and the differences in regional and regional synchronism of mineralization despite volcano-magmatogenic structural setting, dem variations between districts in duration, magni onstrate that the Paleocene and Eocene mag tude, and timing of magmatism. This argues for matic episodes are different and not genetically the greater importance of a regional control on related. These differences are regional in scale the timing of porphyry copper formation, and thus reflect regional fundamental pro rather than on local processes (Fig. 9) cesses, most likely related to different tectonic (McCandless and Ruiz, 1993; this study). conditions and lithospheric architectures (e.g., In the Indio Muerto district, the question of crustal thicknesses) at the time of magma pro whether preceding magmatism was important duction and evolution (Mpodozis et al., 1995). in the formation of the main deposit at 41 to 42 EL SALVADOR COPPER DEPOSIT REVISITED 51

Ma may depend on whether the 43- to 44-Ma report. We also express our thanks to Ricardo quartz porphyries and granodioritic intrusions Rojas, Otilio Chang, Gonzalo Rojas, Osman form a comagmatic suite with the younger Olivares, and Marcelo Mendez (El Salvador granodiorite porphyries or represent separate Division, CODELCO) for their unselfish will systems. In any case, the preceding district- ingness to share their geologic knowledge of the scale magmatism is considerably less than pre Indio Muerto district and the El Salvador viously concluded and less than that docu deposit, and to M. McWilliams, Stanford Uni mented in many porphyry copper districts versity, for allowing the use of his unpublished believed to contain precursor magmatism (Sil- 40Ar/39Ar ages. We also are grateful to Lew litoe, 1988; McCandless and Ruiz, 1993). In the Gustafson for use of the geochronologic analyti latter districts, the question remains as to how cal data of Gustafson and Hunt (1975) and for much of the chronologic proximity of the pre providing the original impetus for this work. We ceding magmatism to the age of the porphyry thank Robert Fleck and L.B. Gustafson for their copper system might result from thermal helpful reviews. This paper is a contribution to effects during emplacement of a younger por IGCP Projects Nos. 342 and 345. phyry complex. Moreover, are precursor igneous rocks truly cogenetic to the porphyry REFERENCES copper system? The El Salvador deposit, pre viously cited as an example of a porphyry cop Andres, R. J., Rose, W. I., Kyle, P. R., deSilva, S., per system emplaced at the culmination of Francis, P., Gardeweg, M., and Moreno Roa, H., protracted district-scale magmatism, clearly 1991, Excessive sulfur dioxide emissions from does not have a long magmatic history. Con Chilean volcanoes: Jour. Volcanol. Geotherm. Res., versely, the geology of the Indio Muerto district v. 46, p. 323-329. suggests that regional and volcanogenic struc Arévalo, C, 1995, Mapa geológico de la Hoja Copiapó: tures focus younger magmas repeatedly Región de Atacama: Documentos de Trabajo No. 8: through geologic time, giving a close spatial Santiago, Servicio Nacional de Geología y Minería, association to otherwise unrelated magmatic (l:100,000-scale map with extended legend). suites—in this case the Paleocene rhyolitic Arévalo, C., Rivera, O., Iriarte, S., and Mpodozis, C., dome and the Eocene porphyries. This fact, 1994, Cuencas extensionales y campos de calderas del Cretácico Superior-Terciario Inferior en la Pre- together with problems of thermal disturbance cordillera de Copiapó (27° -28° S), Chile: VII Con- of isotopic systems during emplacement of the greso Geológico Chileno, Conceptión, Actas, v. 2, p. porphyry system, can lead in some cases to the 1288-1292. appearance of a cogenetic relation between pre Boric, R., Díaz, F., and Maksaev, V, 1990, Geología y cursor district magmatism and the porphyry yacimientos metalíferos de la región de Antofagasta: copper system, given the close spatial and Santiago, Servicio Nacional de Geología y Minería, apparent close temporal relations. This is the Boletín no. 40, 246 p. case for the Indio Muerto district and may be

Downloaded By: [Canadian Research Knowledge Network] At: 21:38 28 May 2011 Brook, M., Pankhurst, R., Sheperd, T., and Shapiro, B., the case for other less well studied districts 1986, Andchron: Andean geochronology and metal - where protracted precursor magmatism is logenesis: London, Overseas Development Admin believed to be present. istration, Open-file Report, 83 p. Casselman, M. J., McMillan, W. J., and Newman, K. M., 1995, Highland Valley porphyry copper deposits Acknowledgments near Kamloops, British Columbia: A review and update with emphasis on the Valley deposit, in Funding for this project was provided by Schroeter, T.G., ed., Porphyry deposits of the north CODELCO and Servicio Nacional de Geología y western Cordillera of North America: Montreal, Minería-Chile (SERNAGEOMIN). We thank Can. Inst. Mining, Metall., Petrol., Spec. vol. 46, p. them for their support. In particular, we express 161-191. our appreciation to Francisco Camus, Clark, A. H., 1993, Are outsize porphyry copper Guillermo Müller, and Pedro Carrasco deposits either anatomically or environmentally dis (CODELCO) for supporting this work, for their tinctive?, in Whiting, B. H., Mason, R., and insights into the geology of the El Salvador Hodgson, C. J., eds., Giant ore deposits: Boulder, CO, region, and for their permission to publish this Soc. Econ. Geol., Spec. Publ. No. 2, p. 213-283. 52 CORNEJO ET AL.

Clark, A. H., Farrar, E., and Zentilli, M., 1985, Discus Hedenquist, J. W., and Lowenstern, J. B., 1994, The role sion of "Relationships between mineralization and of magmas in the formation of hydrothermal ore silicic volcanism in the Central Andes": Jour. Vol deposits: Nature, v. 370, p. 519-527. cano!. Geotherm. Res., v. 24, p. 353-358. Kay, R. W., and Kay, S. M., 1989, Delamination and Coira, B., Davidson, J., Mpodozis, C, and Ramos, V, A., delamination magmatism: Tectonophysics, v. 219, p. 1982, Tectonic and magmatic evolution of the Andes 177-189. of northern Argentina and Chile: Earth Sci. Rev., v. Kay, S. M., and Abbruzzi, J. M., 1996, Magmatic evi 18, p. 303-332. dence for Neogene lithospheric evolution of the Cox, K. G., Bell, J. D., and Pankhurst, R. J., 1979, The central Andean "flat-slab" between 30° S and 32° interpretation of igneous rocks: London, George S: Tectonophysics, v. 259, p. 15-28. Allen and Unwin, 450 p. Kay, S. M., Mpodozis, C, Tittler, A., and Cornejo, P., Cornejo, P., and Mpodozis C., 1996, Geología de la 1994, Tertiary magmatic evolution of the Maricunga región de Sierra Exploradora (Cordillera de mineral belt in Chile: INT. GEOL. REV., v. 36, p. Domeyko 25°-26° S): Santiago, Servicio Nacional 1079-1112. de Geología y Minería-CODELCO, Registered report Le Bas, M. J., Le Maitre, R. W., Streckeisen, A., and IR-96-09, 330 p. Zanettin, B., 1986, A chemical classification of vol Cornejo, P., Mpodozis, C., Kay, S. M., andTomlinson, A. canic rocks based on the total alkali-silica diagram: J., 1994, Volcanismo bimodal potásico en regimen Jour. Petrol., v. 27, p. 745-750. extensional del Cretácico superior-Eoceno en la Lindsay, D. D., Zentilli, M., and Rojas de la Rivera, J., región de El Salvador (26°-27° S), Chile: VII 1995, Evolution of an active ductile to brittle shear Congreso Geológico Chileno, Concepción, Actas, v. system controlling mineralization at the Chuqui- 2, p. 1306-1310. camata porphyry copper deposit, northern Chile: Cornejo, P., Mpodozis, C, Ramírez, C. F., and Tomlin INT. GEOL. REV., V. 37, p. 945-958. son, A. J., 1993a, Estudio geológico de la región Lopez-Escobar, L., 1982. Características geoquímicas de Potrerillos y El Salvador (26°-27° Lat. S): de rocas ígneas asociadas con pórfidos cupríferos Santiago, Servicio Nacional de Geología y chilenos: Revista Geológica de Chile, No. 17, p. 3-19. Lowenstern, J. B., 1994, Dissolved volatile concentra Minería-CODELCO, Registered report IR-93-01, tions in ore-forming magma: Geology, v. 22, p. 258 p. 893-896. Cornejo, P., Mpodozis, C., Kay, S. M., Tomlinson, A. J., Maksaev, V., 1990, Metallogeny, geological evolution, and Ramirez, C. F., 1993b, Upper Cretaceous-Lower and thermochronology of the Chilean Andes Eocene potassic volcanism in an extensional regime between latitudes 21 ° and 26° South, and the origin in the Precordillera of Copiapo, Chile: Second Symp. of major porphyry copper deposits: Unpubl. PhD Int. Géodynamique Andine, Oxford (UK): Paris, thesis, Dalhousie Univ., Halifax, Nova Scotia, Can Éditions de l'ORSTROM, Collection Colloques et ada, 554 p. Séminaires, p. 347-350. Maksaev, V., and Zentilli, M., 1988, Marco metalogénico Dilles, J. H., 1987, Petrology of the Yerington batholith, de los megadepósitos de tipo pórfido cuprífero del Nevada: Evidence for evolution of porphyry copper Norte Grande de Chile: V Congreso Geológico ore fluids: Econ. Geol., v. 82, p. 1750-1789. Chileno, Santiago, Actas, v. 1, p. B181-B218. Downloaded By: [Canadian Research Knowledge Network] At: 21:38 28 May 2011 Frances, P. W., Halls, C, and Baker, M. C. W., 1983, Marinovic, N., Cortés, J., and Garcia, M., 1996, Estudio Relationships between mineralization and silicic vol Geológico Regional de la zona comprendida entre canism in the Central Andes: Jour. Volcanol. Geo Sierra del Buitre y Pampa San Roman: Santiago, therm. Res., v. 18, p. 165-190. Servicio Nacional de Geología y Minería- Gradstein, F. M., and Ogg, J., 1996, A Phanerozoic time CODELCO, Registered Report IR-96-08, 156 p. scale: Episodes, v. 19, p. 3-5. Marinovic, N., Smoje, I., Maksaev, V, Hervé, M., and Gustafson, L. B., 1979, Porphyry copper deposits and Mpodozis, C., 1995, Hoja Aguas Blancas: Región de calc-alkaline volcanism, in McElhinny, M. W., ed., Antofagasta: Santiago, Servicio Nacional de Geo The Earth: Its origin, structure and evolution: logía y Minería, Carta Geológica de Chile No. 70, London, Academic Press, p. 427-468. (l:250,000-scale map with text), 150 p. Gustafson, L. B., and Hunt, J. P., 1975, The porphyry Marsh, T. M., Einaudi, M. T., and McWilliams, M. O., copper deposit at El Salvador, Chile: Econ. Geol., v. 1995, Ar/Ar geochronology of gold and copper min 70, p. 857-912. eralization in the Potrerillos district, Atacama Harland, W. B., Armstrong, R. L., Cox, A. V, Craig, L. region, Chile [abs.]: Geol. Soc. Amer., Abs. Prog., v. E., Smith, A. G., and Smith, D. G., 1990, A geologic 27, no. 6, p. A-329. time scale 1989: Cambridge, UK, Cambridge Univ. Matthews, S. J., Jones, A. P., and Beard, A. D., 1994a, Press, 263 p. Buffering of melt oxygen fugacity by sulphur redox EL SALVADOR COPPER DEPOSIT REVISITED 53

reactions in calc-alkaline magmas: Jour. Geol. Soc. tión del frente volcánico Oligoceno-Mioceno de la London, v. 151, p. 815-823. zona sur de los Andes Centrales: Revista Geológico Matthews, S. J., Jones, A. P., and Gardeweg, M. C., de Chile, v. 22, p. 273-313. 1994b, Lascar Volcano, northern Chile; evidence for Mpodozis, C, Marinovic, N., Smoje, I., and Cuitiño, L., steady-state disequilibrium: Jour. Petrol., v. 35, p. 1993, Estudio Geológico-Estructural de la Cordillera 401-432. de Domeyko entre Sierra Limón Verde y Sierra Matthews, S. J., Sparks, R. S. J., and Gardeweg, M. C. , Mariposas Región de Antofagasta: Santiago, Servicio 1995, The relationships between magma mixing and Nacional de Geologia y Mineria-CODELCO, Regis volatile behaviour at Lascar volcano (23° 22' S, 67° tered report IR-93-04, 282 p. 44' W), northern Chile: Significance for the forma Mpodozis, C, and Ramos, V. A., 1989, The Andes of tion of copper sulphide and magnetite-apatite Chile and Argentina: Circum-Pacific Council for orebodies, in Clark, A. H., ed., Giant ore deposits-II, Energy and Mineral Resources, Earth Sci. Series, v. Controls on the scale of orogenic magmatic-hydro- 11, p. 59-90. thermal mineralization: Kingston, Ontario, QminEx Mpodozis, C., Tomlinson, A., and Cornejo, P., 1994, Associates and Queen's University, p. 146-181. Acerca del control estructural de intrusivos eocenos Mattinson, J. M., 1994, A study of complex discordance y pórfidos cupríferos de la Región de Potrerillos-El in zircons using step-wise dissolution techniques: Salvador: VII Congreso Geológico Chileno, Concep Contrib. Mineral. Petrol., v. 116, p. 117-129. ción, Actas, v. 2, p. 1596-1600. Mattinson, J. M., Graubard, C. M., Parkinson, D. L., and Naney, M., T., 1983, Phase equilibria of rock-forming McClelland, W. C, 1996, U-Pb reverse discordance ferromagnesian silicates in granitic system: Amer. in zircons: The role of fine-scale oscillatory zoning Jour. Sci., v. 276, p. 546-557. and sub-micron transport of Pb, in Basu, A.R., and Olson, S. F., 1989, The stratigraphic and structural Hart, S.R., eds., Earth processes: Reading the iso- setting of the Potrerillos porphyry copper district, topic code: Washington, DC, Amer. Geophys. Union Northern Chile: Revista Geológico de Chile, v. 16, p. Mono. 95, p. 355-370. 3-29. McCandless, T. E., and Ruiz, J., 1993, Rhenium-osmium Quirt, G. S., 1972, A potassium-argon geochronological evidence for regional mineralization in south investigation of the Andean mobile belt of northwestern North America: Science, v. 261, p. central Chile: Unpubl. Ph.D. thesis, Queen's Univ., 1282-1286. Kingston, Ontario, 240 p. McKee, E. H., and Noble, D. C., 1990, Cenozoic tectonic Rivera O., 1995, Proyecto estudio geológico sector Kilo- events, magmatic pulses, and base-and precious- metro Catorce, escala 1:25,000: Unpubl. report, metal mineralization in the Central Andes: Circum- CODELCO, El Salvador, Chile, 50 p. Pacific Council for Energy and Mineral Resources, Reutter, K.-J., Scheuber, E., and Chong, G., 1996, The Earth Sci. Series, v. 11, p. 189-194. Precordilleran fault system of Chuquicamata, north McWilliams, M., 1994, 40Ar/39Ar results from ern Chile: Evidence for reversals along arc-parallel CODELCO El Salvador Samples, Phase 1: Santiago, strike-slip faults: Tectonophysics, v. 259, p. Unpubl. report, CODELCO, 28 p. 213-228. , 1996, 40Ar/39Ar results from CODELCO El Ruiz, J., Freydier, C., McCandless, T., Chesley, J., and

Downloaded By: [Canadian Research Knowledge Network] At: 21:38 28 May 2011 Salvador Samples, Phase 2: Santiago, Unpubl. Munizaga, F., 1996, Re-Os systematics of porphyry report, CODELCO, 11p. copper mineralization in Chile: Anais, XXXIX Megard, F., Noble, D.C., McKee, E. H., and Bellon, H., Cong. Brasil. Geol., v. 7, p. 281-283. 1984, Multiple pulses of Neogene compressive defor Ruiz, J., Freydier, C., McCandless, T., Chesley, J., and mation in the Ayacucho intermontane basin, Andes Munizaga F., 1997, Re-Os isotope systematics of of central Peru: Geol. Soc. Amer. Bull., v. 95, p. sulfides from base metal porphyry and manto-type 1108-1117. mineralization in Chile: INT. GEOL. REV. Mortimer, C, 1973, The Cenozoic history of the south (forthcoming). ern Atacama Desert, Chile: Jour. Geol. Soc. London, Sanderson, D. J., and Marchini, R. D., 1981, Transpres- v. 129, p. 505-526. sion: Jour. Struct. Geol., v. 6, p. 449-458. Mpodozis C., and Cornejo P., 1994, Segmentación And- Sebrier, M., Lavenu, A., Fornari, M., and Soulas, J. P., ina en el Paleozoico superior: Evidencias en los 1988, Tectonics and uplift in Central Andes (Peru, Batolitos Pérmicos Sierra del Castillo y Pedernales: Bolivia, and Northern Chile) from Eocene to pre VII Congreso Géologico Chileno, Conceptión, sent: Geodynamique, v. 3, p. 85-106. Actas, v. 2, p. 1399-1403. Skewes, M. A., and Stern, C. R., 1995, Genesis of the Mpodozis, C., Cornejo, P., Kay, S. M., and Titler, A., giant late Miocene to Pliocene copper deposits of 1995, La Franja de Maricunga: Síntesis de la evolu Central Chile in the context of Andean magmatic 54 CORNEJO ET AL.

and tectonic evolution: INT. GEOL. REV., V. 37, p. Tomlinson, A. J., Mpodozis, C., Cornejo, P., and 893-909. Ramirez, C. F., 1993, Structural geology of the Sierra Sillitoe, R. H., 1973, The tops and bottoms of porphyry Castillo-Agua Amarga fault system, Precordillera of copper deposits: Econ. Geol., v. 68, p. 799-815. Chile, El Salvador-Potrerillos: Second Symp. Int. , 1988, Epochs of intrusion-related copper Géodynamique Andine, Oxford (UK): Paris, Éditio mineralization in the Andes: Jour. South Amer. ns de l'ORSTROM, Collection Colloques et Sémi- Earth Sci., v. 1, p. 89-108. naires, p. 259-262. Sparks, R. S. J., Sigurdsson, H., and Wilson, L., 1977, Tosdal, R. M., 1995, Metal source differences in Ceno- Magma mixing: A mechanism for triggering acid zoic porphyry Cu-Mo-Au deposits in the central explosive eruptions: Nature, v. 267, p. 315-318. Andes between 26° and 28°: An influence on the Steiger, R. H., and Jager, E., 1977, Subcommission on size of porphyry deposits?, in Clark, A. H., ed., Giant geochronology: Convention on the use of decay ore deposits-II, Controls on the scale of orogenic constants in geo- and cosmochronology: Earth magmatic-hydrothermal mineralization: Kingston, Planet. Sci. Lett., v. 36, p. 359-362. Ontario, QminEx Associates and Queen's Univer Tera, F., and Wasserburg, G. J., 1972, U-Th-Pb system- sity, p. 124-138. atics in three Apollo 14 basalts and the problem of Williams, W. C, 1995, The crustal heritage of ore initial Pb in lunar rocks: Earth Planet. Sci. Lett., v. deposits [abs.]: Geol. Soc. Amer., Abs. Prog., v. 27, 14, p. 281-304. no. 6, p. A-410. Titley, S. R., 1982, Geologic setting of porphyry copper Zentilli, M., 1974, Geological evolution and metal- deposits: Southeastern Arizona, in Titley, S. R., ed., logenic relationships in the Andes of northern Chile Advances in geology of the porphyry copper between 26° and 29° South: Unpubl. Ph.D. thesis, deposits: Southwestern North America: Tucson, Queen's Univ., Kingston, Ontario, 446 p. Univ. of Arizona Press, p. 37-58. Zentilli, M., Doe, B., Hedge, C. E., Alvarez, 0., Tidy, E., Titley, S. R., and Beane, R. E., 1981, Porphyry copper and Daroca, J. A., 1988, Isotopos de plomo en yaci- deposits: Part I, geologic settings, petrology, and mientos de tipopórfido cuprífero comparados con tectogenesis, in Skinner, B.J., ed., Econ. Geol., 75th otros depósitos metaliferos en los Andes del norte de Anniv. vol., p. 214-235. Chile y Argentina: V Congreso Geológico Chileno, Tobar, A., 1977, Stratigraphy and structure of the El Santiago, Actas, v. 1, p. B331-B369. Salvador-Potrerillos region, Atacama, Chile: Zentilli, M., Krogh, T. E., Maksaev, V., and Alpers, C. N., Unpubl. Ph.D. thesis, Univ. of California, Berkeley, 1994, Uranium-lead dating of zircons from the Chu- 153 p. quicamata and La Escondida porphyry copper Tomlinson, A. J., 1994, Relaciones entre el porfido deposits, Chile: Inherited zircon cores of Paleozoic cupífero y la falla inversa de la Mina de Potrerillos: age with Tertiary overgrowths: Comunicaciones, no. Un caso de intrusión sintectónica: VII Congreso 45, p. 101-110. Geológico Chileno, Conceptión, Actas, v. 2, p. 1629-1633. Downloaded By: [Canadian Research Knowledge Network] At: 21:38 28 May 2011

View publication stats