Introduction

The Coastal Cordillera of northern Chile hosts numerous copper deposits, constituting a NS-trending Late Jurassic to Early Cretaceous metallogenic belt extending 200 km from 22° to 24°S (Fig. 1). This metallogenic belt represents an important source of Chilean copper production (Maksaev and Zentilli 2002). Total annual production of fine copper from these deposits is on the order of 230,000 ton, primarily from the exploitation of six orebodies. The major mines currently operating are located in the Mantos Blancos mining district, which produces 50% of the copper from the belt and has estimated resources of 500 million ton with 1.0% Cu. Two types of ore deposits occur in this copper belt: volcanic-hosted stratabound orebodies (Espinoza et al. 1996; Maksaev and Zentilli 2002) and porphyry copper deposits (Camus 2003). The stratabound orebodies have long been known as “Chilean manto-type,” because initial mining exploited the stratiform section of these deposits. However, they are characterized by central hydrothermal breccias that form feeder structures to the flat-lying stratiform peripheral mineralization. The magmatic-hydrothermal breccias host half of the mineralization and have the highest ore grade in the deposits. These breccias are genetically related to coeval dioritic to granodioritic stocks and sills and are intruded by postmineralization dioritic dikes and stocks. Hypogene sulfide mineralization consists of bornite, digenite, and chalcopyrite related to sodic hydrothermal alteration (Palacios 1990; Ramírez et al. 2006).

Fig. 1
figure 1

Simplified geology map of the Coastal Cordillera between 22°00′ and 24°00′S. Other volcano-hosted, strata-bound type copper deposits are also shown. Inset: geology of the Mantos Blancos district, modified after Ramírez et al. (2006)

Full size image

In spite of their importance, the ages of the copper ore deposits of the Coastal Cordillera in northern Chile are not well known. This applies to the volcanic rocks that host the main ore deposits, the intrusive bodies related to mineralization, and the associated hydrothermal processes. Only scarce K/Ar and Rb–Sr ages between 160 and 130 Ma (Michilla, Buena Esperanza, and Mantos Blancos) have been published (e.g. Chávez 1985; Boric et al. 1990; Tassinari et al. 1993; see discussion by Maksaev and Zentilli 2002). Recently, a Re–Os age of 159 ± 16 Ma (2σ) on chalcocite from the main deposit in the Michilla mining district (22°30′S) has been reported (Tristá-Aguilera et al. 2006); an interval of 163.5 ± 1.9 and 157.4 ± 3.6 Ma has also been proposed for the main mineralization event in this deposit (Oliveros et al. 2007).

The present paper reports a detailed 40Ar/39Ar study on most of the geological formations of the Mantos Blancos district, including intrusive and volcanic rocks and secondary minerals, allowing to determine a chronological evolution of the ore deposits in the district. The ore deposits cannot be easily and precisely dated directly, and this paper shows how a careful interpretation of complex geochronological data obtained on igneous and secondary products can indirectly give precise ages for the ore deposits. This geochronological model is largely based on field observations that are given by Ramírez et al. (2006), who recently proposed a comprehensive two-stage evolution model of the hydrothermal system of the Mantos Blancos ore deposit. In particular, our geochronological data are in agreement with the two mineralization events hypothesized by Ramírez et al. (2006).

Geologic and tectonic setting

Since the Early Jurassic, subduction-related magmatism developed along the Coastal Cordillera. The Jurassic volcanic activity is represented by the La Negra Formation, a 7,000-m thick sequence of mainly basaltic to andesitic lavas with calc–alkaline and minor tholeiitic affinities and sedimentary rocks (García 1967; Palacios 1978, 1984; Rogers and Hawkesworth 1989; Pichowiak 1994; Kramer et al. 2004). Based on radiometric ages and paleontological evidence obtained on marine sedimentary intercalations, the extrusive event has been bracketed between Early Jurassic and Oxfordian (Rogers and Hawkesworth 1989; Basso 2004; Kramer et al. 2004; Oliveros et al. 2006; Cortés et al. 2007), with an apparent maximum of volcanic activity during the Late Jurassic (Callovian to Kimmeridgian age, Oliveros et al. 2006). The volcanic sequence was intruded by Early Jurassic (ca. 173–190 Ma) and Late Jurassic (ca. 142–160 Ma) tonalitic-granodioritic and dioritic plutonic rocks (Pichowiak et al. 1990; Andriessen and Reutter 1994; Pichowiak 1994; Dallmeyer et al. 1996; Scheuber 1994; Scheuber and González 1999; Oliveros et al. 2006). Similar isotopic compositions of both the volcanic and plutonic rocks suggest mantle-derived parental magmas with minor or no crustal contamination (Rogers and Hawkesworth 1989; Lucassen and Franz 1994; Kramer et al. 2004).

The Coastal Cordillera tectonic evolution during the Jurassic is interpreted in terms of a transtensional regime because of oblique subduction (Scheuber and González 1999) and/or to a retreating type boundary system (Grocott et al. 1994). A major arc-linked structure, the Atacama Fault System (AFS), concentrated most of the deformation in the arc during the Jurassic and Early Cretaceous with sinistral strike–slip motion. During the Paleogene Incaic compressive phase, the AFS was reactivated, and the present day Coastal Cordillera of northern Chile was probably formed at this time (Arraigada 2003).

Geology of the Mantos Blancos mining district

The Mantos Blancos ore deposit is hosted within a rhyolitic dome that is supposed to be associated with hydrothermal breccias, which in turn are intruded by bimodal stocks and sills of dioritic and granodioritic composition (Fig. 1), causing a second stage of brecciation and hydrothermal alteration (Ramírez et al. 2006). All these lithological units are mineralized to different degrees (Ramírez et al. 2006). A late mineralization dioritic dyke swarm crosscuts all the exposed rock units. The eastern and western boundaries of the Mantos Blancos ore deposit are marked by NNE to NS-trending faults, which separate the ore deposit from Lower Jurassic tonalitic plutonic rocks (Fig. 1). To the north, the igneous rocks that host the ore deposit intrude a Lower to Middle Jurassic granodiorite (Cortés et al. 2007; Fig. 1). To the south, the ore deposit is limited by an ENE-trending fault, separating it from the Jurassic andesitic to basaltic sequence of the La Negra Formation (Cortés et al. 2007). In much of the district, the fault is covered by Quaternary sediments (Fig. 1); therefore, the stratigraphic relationship between the host rhyolitic dome and the La Negra sequence is unclear.

Dioritic and granodioritic porphyries also crop out 5 km to the S and SW of the main ore deposit (Fig. 1) and appear to be equivalent to the Mantos Blancos bimodal porphyry based on similarities in texture and mineralogical and geochemical composition. Additionally, other small bodies of equigranular granodiorite to diorite also crop out 8 km southwest of the deposit (Fig. 1); however, there is no evidence of a correlation between these rocks and the bimodal porphyry stocks and sills.

The geology of the Mantos Blancos ore deposit is described in detail by Ramírez et al. (2006), who recognized two superimposed hydrothermal events. The first event is coeval with the magmatic-hydrothermal brecciation and its associated phyllic alteration. The hypogene assemblage of this first event consists of chalcopyrite, pyrite, bornite, quartz, and sericite. This assemblage occurs in various modes, including: (a) isolated crystals within quartz phenocrysts of the breccias, (b) disseminated crystals in the felsic dome and in the matrix and fragments of the hydrothermal breccia, and (c) in veinlets. The assemblage is observed in deep drill cores within the rhyolitic dome, at least 300 m away from breccia pipes of the second hydrothermal event (Ramírez et al. 2006). The second event represents the main stage of hydrothermal activity, during which three types of hydrothermal alteration occurred: potassic (K–feldspar, quartz, tourmaline, biotite, chlorite, and magnetite), propylitic (quartz, chlorite, epidote, calcite, albite, hematite, and galena), and sodic (albite, hematite). Mineralization associated with this event consists of chalcopyrite, pyrite, and digenite. These minerals occur as disseminated crystals or as stockworks in the felsic dome, bimodal porphyries, and in the matrix of the dioritic to granodioritic magmatic-hydrothermal breccias. Mineralization in veinlets crosscutting phyllic veins of the first event also occur. Finally, late-ore mafic dikes intruded all of the rock units within the Mantos Blancos copper deposit (Ramírez et al. 2006).

On a regional scale, a widespread low-grade alteration with a typical propylitic assemblage of quartz, chlorite, epidote, calcite, sericite, and titanite affects the volcanic rocks of the La Negra Formation. This alteration is thought to be equivalent to the “alteration on regional scale,” defined by Losert (1974).

Description of the analyzed samples

Three samples of granodioritic and dioritic porphyries located 3 km S and SW from the Mantos Blancos ore deposit, VO16, MB02.4, and MB02.3, as well as one sample (VO8) of an equigranular granodiorite intrusion located 8 km to the SW, were analyzed (Fig. 1). These rocks contain plagioclase, potassic feldspar, quartz, biotite, and hornblende. With the exception of sample VO8, biotites from the intrusive rocks are commonly chloritized, while hornblende crystals remain fresh. Sample VO14 was collected near VO8 (Fig. 1) and belonged to a strongly altered intrusion, probably a granodiorite, with quartz, sericitized plagioclase, and altered mafic minerals. As mentioned previously, whereas VO16, MB02.4, and MB02.3 are thought to be related to the bimodal stocks and dykes inside the Mantos Blancos ore deposit because of their similar petrological and geochemical features, the pluton corresponding to the sample VO8 has no direct correlation with any units known in the deposit.

Sample MB02.2 is an andesite from the La Negra Formation affected by contact metamorphism, probably related to the emplacement of the dioritic and granodioritic porphyries located nearby and described above (Fig. 1). The sample has an hornfelsic texture with plagioclase, biotite, and amphibole in the groundmass and replacing primary plagioclase and pyroxene phenocrysts. Sample VO130B is a basaltic-andesitic lava flow from the La Negra Formation sampled about 7 km south of the Mantos Blancos deposit. The rock is rich in plagioclase phenocrysts, which are strongly sericitized but contain fresher cores; chlorite and epidote are also present as alteration phases in the groundmass and mafic phenocrysts.

Sample CPM1.27 is from a late-mineralization monzonitic dyke, which crosscuts the magmatic-hydrothermal breccias, feeders of the second mineralization event. It contains traces of pyrite and chalcopyrite together with a weak propylitic and albitic alteration; therefore, its age should represent a late stage of the second mineralization event. Plagioclase, potassic feldspar, and amphibole phenocrysts in this rock are strongly altered. Despite this strong alteration, the abundant amphibole phenocrysts (up to 50% of the rock) preserve large fresh internal zones.

Sample S.24 is a dioritic rock collected from drill core, whereas sample P2.8 is a dioritic dyke. In the case of sample S24, the contact relationship with the host rocks is not clear. Both samples might belong to the basic dyke swarm that crosscuts the rhyolitic dome, as their basic chemical composition is very similar and strongly contrasts to that of the acid host rocks. These plagioclase-rich porphyritic rocks contain mafic minerals that are mostly chloritized and small amounts of pyrite and traces of chalcopyrite.

Sample DDH11533-104 is a sample of a drill core from the rhyolitic dome that is weakly mineralized. The sample was taken under the main orebody and contains quartz, plagioclase, biotite, and other mafic minerals as phenocrysts and within the felsitic groundmass. The plagioclase phenocrysts are intensely seritized and albitized, biotites are replaced by iron-rich minerals and other mafic phenocrysts are completely chloritized.

Analytical procedures

Fifteen primary and secondary minerals from 11 rock samples were analyzed using the 40Ar/39Ar step-heating method. Analyzed primary minerals include: single grains of biotite from samples VO8 and VO16, bulk plagioclase from samples VO130B, S24, and P2.8, small clusters of amphibole grains from samples MB02-3, MB02-3, and MB02.4 (heated with a laser system), and a bulk sample of amphibole from sample VO16, which was heated with a high frequency furnace (HF). Analyzed secondary minerals consisted of strongly sericitized plagioclase single grains from samples VO14, VO130B, DDH11533, S24, and P2.8, all of which were heated with a laser system. Mineral separation was carried out using a Frantz magnetic separator followed by careful hand picking under a binocular microscope. Grain sizes are 300–500 μm for sericitized plagioclase and biotite single grains and 125–200 μm for amphibole and fresh plagioclase populations. The weight of the analyzed samples was 0.05–0.8 mg for biotite and sericitized plagioclase single grains, 1.2–1.5 mg for amphibole and plagioclase bulk sample heated with the laser system, and 8.7 mg for the amphibole bulk sample (VO16) heated with the HF furnace. The samples were packed into copper or aluminum foils for the HF furnace or laser experiments, respectively. The samples were then irradiated for 70 h in the Hamilton McMaster University nuclear reactor (Canada) in position 5C along with the Hb3gr hornblende monitor for which an age of 1,072 Ma is adopted (Turner et al. 1971; Renne et al. 1998; Jourdan et al. 2006). The total neutron flux density during irradiation was about 8.8 × 1018 n/cm2. The estimated error bar on the corresponding 40Ar*/39ArK ratio is ±0.4% (2σ) in the volume where the samples were included. For the samples heated with the laser system (Synrad 48–5 CO2 laser), isotopic measurements were performed with a VG3600 mass spectrometer equipped with a Daly system and a photomultiplier. For the samples heated with the HF furnace, the analysis was performed with a mass spectrometer composed of a 120° M.A.S.S.E. tube, a Baür–Signer GS 98 source and a SEV217 Balzers electron multiplier. Total blanks were measured every third step on the laser system and every third sample on the HF furnace system. Argon isotopes were typically on the order of 20–2,000, 100–15,000, and 1–50, for laser experiments, and on the order of 30–430, 10–7,000, and 8–50, for HF furnace experiments, times the blank level for 40Ar, 39Ar and 36Ar, respectively.

The criterion used to define a plateau age includes: (1) at least 70% of the 39Ar released; (2) a minimum of three successive steps in the plateau; and (3) the integrated age of the plateau should agree with each apparent age of the plateau within a two sigma (2σ) confidence level. Plateau and integrated ages are given at the 2σ confidence level. The uncertainties on the 40Ar/39Ar ratios of the monitors are included in the calculation of the integrated and plateau age uncertainties, but the error on the age of the monitor is not included.

The chemical compositions of the primary and secondary minerals were obtained using a CAMECA SX-100 electron microprobe at the Institut des Sciences de la Terre de l’Environnement et de l’Espace of the Université de Montpellier (France) with 20 kV, 10 nA, and 2–5 μm as analytical conditions. A CAMECA SU-30 SEM-probe was also used at the Departamento de Geología, Universidad de Chile (Chile), with 15 kV, 10 nA, and 3–5 μm as analytical conditions.

Results and interpretation of the age data

A summary and detail of the isotopic data are given in Table 1 and Electronic Supplementary Material Table 1 (ESM), respectively. Age and 37ArCa/39ArK ratio spectra are shown in Figs. 2 and 3.

Table 1 Summary of 40Ar/39Ar data on biotite, amphibole, plagioclase, and sericitized plagioclase form lava flows, intrusive rocks, and dykes from Mantos Blancos district
Full size table
Fig. 2
figure 2

40Ar/39Ar age and 37ArCa/39ArK ratio spectra obtained on biotite single grain, amphibole, and plagioclase bulk sample and amphibole small grain clusters sample from the Mantos Blancos district (plutons, lava flows, and dykes). Shaded rectangles show measured 37Ca/39K ratios from microprobe analyses (data on Electronic Supplementary Material Table 2, ESM). Plateau and integrated ages are given at the two sigma (2σ) confidence level; however, apparent ages of each individual step are given at the 1σ (Electronic Supplementary Material Table 1, ESM)

Full size image
Fig. 3
figure 3

40Ar/39Ar age and 37ArCa/39ArK ratio spectra obtained on sericitized plagioclases from the Mantos Blancos district. 40Ar/39Ar age spectra displayed by fresh plagioclase from samples VO130B, P2.8, and S24 are given for comparison. The criteria to define a plateau age are explained in Fig. 2. Attention

Full size image

The granodioritic and dioritic plutons

  1. (a)

    Granodiorite and diorite plutons located 3 km S and SW from the Mantos Blancos deposit

    Both samples VO16 and MB02.3 were sampled from the same isolated hill (Fig. 1). Petrographic observations reveal very similar mineralogy, texture, and alteration, and therefore, they could belong to the same magmatic unit, although this could not be unambiguously demonstrated because of the scarcity of rock exposures. Amphiboles from both samples yield discordant plateau ages of 136.8 ± 1.8 Ma and 142.2 ± 1.0 Ma, respectively (Fig. 2). The amphiboles have variable 37ArCa/39ArK ratio spectra indicating that the analyzed amphiboles were slightly altered. It was noted that the apparent ages of sample MB02.3 are affected by much higher error bars than sample VO16 because of the lower measured argon signals. The inverse correlation diagrams 36Ar/40Ar vs 39Ar/40Ar are not shown, but the isochron ages are listed in Table 1. Sample MB02.3 yields an isochron age of 141.8 ± 3.6 Ma, which is concordant with the plateau age, with an initial 40Ar/36Ar ratio of 258.4 ± 22.0 that is slightly lower than the atmospheric ratio. Sample VO16 yields an isochron age of 140.8 ± 2.0 Ma, also concordant with the plateau age, with an atmospheric initial 40Ar/36Ar ratio of 292.6 ± 13.4. If these two samples belong to the same pluton, its age is more probably given by the VO16 plateau age because of the higher precision on the apparent ages and the better fit on the isochron plot. A plateau age of 145.4 ± 0.6 Ma was obtained on the VO16 biotite (Fig. 2). Nevertheless, a slight bump in the shape at intermediate temperatures, typical of slight chloritization and 39Ar recoil during irradiation, clearly appears on the age spectrum. The slightly older plateau age obtained for the biotite compared with the amphibole may be the result of chloritization and/or some excess argon as frequently observed on plutonic biotite. Therefore, the amphibole plateau age for the VO16 sample is preferred. The amphibole sample MB02.4 displays a well-defined plateau age of 141.6 ± 0.6 Ma, corresponding to pure amphibole, as shown by the constant 37ArCa/39ArK ratio spectrum. The two concordant plateau ages of 142.2 ± 1.0 Ma (VO16), and 141.6 ± 0.6 Ma (MB02.4) may indicate that these samples belong to the same plutonic event, in which case, and because of the higher quality of the data displayed by MB02.4, the latter plateau age is preferred for this pluton.

  2. (b)

    Granodioritic plutons located 8 km SW of the Mantos Blancos deposit

    The plateau age of 148.2 ± 0.6 Ma, obtained for the VO8 biotite sample (Fig. 2), suggests the existence of an older magmatic event. However, some undetected excess argon may affect the biotite from this intrusion.

Volcanic rocks

Plagioclase from a lava flow of the La Negra Formation (sample VO130B) yields a plateau age of 156.3 ± 1.4 Ma on the gas fraction related to the fresh portion of plagioclase, as demonstrated by the nearly flat 37ArCa/39ArK ratio spectrum that is in the range of the Ca/K ratio measured with the electron microprobe (using the relation: 37ArCa/39ArK = 0.54 Ca/K; Fig. 2). Amphibole from a contact metamorphosed volcanic rock (MB02.2) yielded a well defined plateau age of 142.5 ± 0.6 Ma. The 37ArCa/39ArK spectrum shows a slight increase in this ratio vs temperature, demonstrating that the analyzed mineral was either pure or only slightly altered. This rock is probably included in the contact aureole of the pluton(s) corresponding to samples VO16, MB0.3, and MB02.4. The plateau age is very close to that of the pluton (141.6 ± 0.6 Ma), and therefore, it is likely that this age represents the intrusive event.

Dykes

Primary amphibole from the late-mineralization dyke (CPM1.27), which crosscuts the hydrothermal breccias of the second mineralization event, yields a plateau age of 142.7 ± 2.0 Ma (Fig. 2) that corresponds to pure amphibole, as shown by the mostly flat 37ArCa/39ArK ratio spectrum. Plagioclase from samples S24 (assumed to be a dyke) and P2.8 (dyke) located inside the main deposit show complex “stair-shaped” age spectra (Fig. 2). These disturbances are not related to K-rich alteration such as sericitation because of the flat 37ArCa/39ArK ratio spectra that are in the range of the Ca/K composition measured on fresh plagioclase by the electron microprobe (Fig. 2). These complex results are discussed further below.

Alteration minerals

Because of their tiny size, secondary minerals such as sericite could not be separated from the plagioclase and therefore were analyzed by heating single grains of strongly altered plagioclase where sericite represents the dominant alteration mineral. Aguirre et al. (1999) and Fuentes et al. (2005), working with similar material in Cretaceous lavas from the Coastal Cordillera of central Chile, have shown that the obtained plateau ages may represent in some cases the age of the sericite, while the contribution of fresh plagioclase is negligible.

Single grains of strongly altered plagioclase from sample VO14 yield two well-defined plateau ages of 145.6 ± 0.8 Ma and 147.1 ± 0.6 Ma (Fig. 3). The corresponding 37ArCa/39ArK ratios are highly variable and show the dominant degassing of sericite at intermediate temperatures. Both these age values, and the fact that plateau ages are obtained, indicate that the sericite is dominantly degassing on the plateau fraction, and therefore, the plateau ages correspond to the alteration event. Although these ages are not strictly concordant, they are close enough to be considered as representing the same alteration event. Sericitized plagioclase from an andesitic lava flow (sample VO130B) yield widely discordant and much younger apparent ages than the fresh plagioclase of the same rock (156.3 ± 1.4 Ma, Fig. 2). The weighted mean ages are 126.2 ± 0.5 Ma and 147.5 ± 0.7 Ma, considering the 62 and 63% of the total 39Ar released, respectively (Fig. 3). They are obtained when excluding the highest temperature apparent ages because the corresponding high 37ArK/39ArCa ratios suggest that they do not strictly represent sericite degassing. These ages may represent two distinct overprinted alteration events, but this must be confirmed by further work. The oldest age is concordant with the VO14 sericitized plagioclase plateau age.

From within the main orebody, two sericitized plagioclase single grains from sample DDH11533-104 (rhyolitic dome) yielded disturbed age spectra characterized by increasing apparent ages at low temperatures, followed by a flat region, then decreasing ages at high temperatures. These spectra are difficult to interpret because the lowest ages correlate to the highest 37ArK/39ArCa ratios. Nevertheless, one plateau age of 155.4 ± 0.7 Ma is given for one grain, whereas a concordant weighted mean age of 156.6 ± 0.6 Ma was obtained on the other grain (Fig. 3). These two ages correspond to the lowest Ca/K ratios and therefore to the degassing peak of sericite.

The strongly sericitized single grains of plagioclase from the dyke S24 display a similar type of age spectra, with a flat central region yielding a “mini plateau” age of 155.4 ± 0.6 Ma (62% of the 39Ar released) and a plateau age of 154.7 ± 1.2 Ma, both concordant with the DDH11533-104 sericitized plagioclase plateau age (Fig. 3). Curiously, this age is sensibly older than the low intermediate temperature weighted mean age of 139.3 ± 1.0 Ma displayed for the fresh plagioclase of the same sample further discussed later.

The sericitized single grain of plagioclase from the dyke P2.8 displayed apparent ages with large error bars because of the low sericite (K) content of the analyzed grain. An imprecise and young plateau age of 101.6 ± 5.0 Ma was obtained, corresponding to variable 37ArCa/39ArK ratios. Because of the existence of both calcite and albite in this sample and the low quality of the age data, this datum is difficult to interpret.

Discussion

Age of the La Negra Formation

The La Negra Formation, in the Mantos Blancos district, has been precisely dated for the first time; a plateau age of 156.3 ± 1.4 Ma was obtained on plagioclase in an andesitic lava flow (sample VO130B; Fig. 4, Table 2). This age is much younger than previous Rb/Sr age of 186 ± 13 Ma (1σ, MSWD = 4.2; Rogers and Hawkesworth 1989) but roughly concordant with the Ar/Ar ages ranging from ca. 150 to 165 Ma, recently reported (Oliveros et al. 2006; Cortés et al. 2007). Judging from all the existing age data, the latest magmatism of the La Negra arc may have occurred in the Mantos Blancos district.

Fig. 4
figure 4

Synthetic age data of plateau and weighted mean ages obtained on primary and alteration minerals of samples from Mantos Blancos district. bo, bornite; cc, chalcocite; cpy, chalcopyrite; di, digenite; py, pyrite; qtz, quartz; ser, sericite. Mineral assemblages are from Ramírez et al. (2006)

Full size image
Table 2 Rock type, analyzed materials, and geological and chronological interpretation of the accepted 40Ar/39Ar ages for the samples from the Mantos Blancos district
Full size table

Age of the intrusions

One intrusive magmatic event has been robustly dated in the Mantos Blancos mining district. Dioritic and granodiotitic porphyry intrusions located a few kilometers south from the Mantos Blancos ore deposit (Fig. 1) yield concordant plateau ages of 141.6 ± 0.6 (MB02.4 amphibole) and 142.2 ± 1.0 Ma (VO16 amphibole). These ages are confirmed by sample MB02.2 of a metamorphic rock that belongs to the contact aureole of these plutons, yielding an age of 142.5 ± 0.6 Ma on a metamorphic amphibole (Fig. 4, Table 2). A younger age of 136.8 ± 1.8 Ma obtained on amphibole from sample MB02.3 was rejected because of the lower measured argon signals (see above). A biotite K-Ar age of 147 ± 4 Ma (2σ) previously obtained on a sample from the same outcrop as samples VO16 and MB02.3 (Fig. 1; Chávez 1985) is concordant with our amphibole data. A plateau age of 145.4 ± 0.5 Ma was obtained on biotite from sample VO16, but it displays a bump shaped age spectrum typical of chloritized biotites. The chloritization was confirmed by petrographic observations. Therefore, because biotite from plutonic rocks may be affected by undetectable excess 40Ar and chloritization and without extra age data on other mineral to compare with, it is difficult, at the moment, to evaluate the real meaning of the 4 Ma older 40Ar/39Ar age on biotite as well as the K–Ar age from Chávez (1985). As these dated plutons are interpreted to be equivalent to the granodiorite and diorite porphyry stocks and sills of the Mantos Blancos ore deposit, which are supposed to be linked to the second mineralization event inside the copper deposit (Ramírez et al. 2006), an age in the range of 141.6 ± 0.6 Ma and 142.2 ± 1.0 Ma is proposed for this main mineralization event (Fig. 4, Table 2). This age is also consistent with the plateau age of 142.7 ± 2.0 Ma given by the late-mineralization dyke CPM1.27 that crosscuts the hydrothermal breccias of the second mineralization event (Fig. 4, Table 2). This dyke intruded shortly after the intrusion of dioritic and granodioritic porphyries and their associated magmatic-hydrothermal breccias inside the ore deposit.

A granodiorite sample located farther to the SW (VO8, Fig. 1) yielded a plateau age of 148.2 ± 0.6 Ma on biotite, indicating an older magmatic event. Furthermore, a slightly older K–Ar age of 155 ± 4 Ma (Cortés et al. 2007) was obtained on a granodiorite near the VO8 outcrop. Nevertheless, the too-old plateau age obtained on the VO16 biotite compared to amphibole from the same rock leads us to be cautious in the interpretation of the biotite data given by the plutons. Therefore, the various pulses of pluton emplacement based on biotite dating must be verified by analyses of other minerals.

The age spectra displayed by plagioclases of the dykes S24 and P2.8 (Fig. 3) are difficult to interpret. They correspond to flat Ca/K ratios, demonstrating the lack of a K-rich alteration that may have disturbed the isotopic system. The ages obtained on duplicated sericitized plagioclase from sample S24 are higher than the low to intermediate ages of fresh plagioclase (this age is probably valid: see below). Although we cannot dismiss the possibility of excess 40Ar, at least at high temperature, this hypothesis does not explain the younger ages obtained for fresh plagioclase compared to sericite. A possible explanation would be a partial resetting of plagioclase that does not sensibly affect the included secondary sericite. To accept that, we must admit a higher closure temperature for sericite than for plagioclase, and this has never been evidenced to our knowledge. The sericitized plagioclase from sample VO14 that yielded well-defined plateau ages (Fig. 3) consists of a simple mixture system between the two components, plagioclase and sericite (see also examples in Fuentes et al. 2005). These analyzed grains as well as the sericitized plagioclases from samples S24 and P2.8 (Fig. 3) show similar 37ArCa/39ArK ratio spectra. In these diagrams, we clearly observe a first dominant degassing of plagioclase at low temperatures (high 37ArCa/39ArK ratios) followed by a dominant degassing of sericite (low 37ArCa/39ArK ratios) at intermediate temperatures. This means that during the laboratory experiments, the K/Ar system in plagioclase is opened at a temperature where the sericite is still mainly closed. It therefore appears realistic that the low temperature ages (weighted mean = 139.2 ± 1.0 Ma) are the result of a thermal event. This hypothesis is strengthened by the fact that this age is close to the main plutonic and to the second (main) mineralization events. The same conclusions may be deduced for the plagioclase from the dyke P2.8 that displays a low temperature weighted mean age of 142.3 ± 0.7 Ma that may also be geologically significant. Concerning the fresh plagioclase from samples S24 and P2.8, the high temperature ages of 162.6 ± 1.9 Ma and 149.3 ± 0.7 Ma could represent the emplacement ages for these dykes. A plagioclase age of 151.3 ± 1.2 Ma has been obtained for a coarse pyroxene-plagioclase andesite porphyry dyke within the Manto Blancos main deposit (Cornejo et al. 2006); Maksaev (1990) and Scheuber and González (1999) reported emplacement ages between 139 and 155 Ma for the so-called “nonaltered mafic dykes” in the Coastal Cordillera at the latitude of Mantos Blancos. These K–Ar and Ar–Ar radiometric ages, obtained for plagioclase and whole rocks, are in the same range of our data.

Age of the alteration events

Concerning the strongly sericitized plagioclase data, the 40Ar/39Ar age and 37ArCa/39ArK ratio spectra may be complex but are reproducible as demonstrated by the duplicated samples VO14 and DDHH11533.104 (Fig. 3). They are the result of a mixture of mainly plagioclase and sericite but also probably albite and calcite, as observed in thin sections. As previously interpreted by Aguirre et al. (1999) and Fuentes et al. (2005), the apparent ages correlating with the lowest Ca/K ratios are considered as the result of the dominant degassing of sericite. This is strengthened by the existence of corresponding sharp degassing peaks of 39Ar (not shown), characteristic of white micas. These ages, which mostly correspond to intermediate temperature, may be geologically significant.

The strongly sericitized plagioclase sampled in the rhyolitic dome (sample DDH11533.104) yields two concordant ages of 155.4 ± 0.7 Ma (plateau age) and 156.6 ± 0.6 Ma (mini plateau age). Moreover, the duplicated sericitized plagioclase from sample S24, belonging to a basic rock in the rhyolitic dome, displays a plateau age of 154.7 ± 1.2 Ma and a mini plateau age of 155.4 ± 0.6 Ma that are concordant. These ages represent the first mineralization event. They are close to or concordant with the La Negra primary plagioclase plateau age of 156.3 ± 1.4 Ma (Fig. 4).

Based on mineralogical data (sericite, pyrite, and chalcopyrite inclusions within magmatic quartz phenocrysts) and melt inclusion geochemistry, Ramírez et al. (2006) argued that the phyllic alteration of the first hydrothermal event was probably related to the dome emplacement. However, recent U–Pb ages of 180–181 Ma obtained on zircons form the dome and related volcanic rocks from the Mantos Blancos district have been reported by Cornejo et al. (2006). These age data are sensibly older than our present Ar/Ar ages obtained on sericites related to the first mineralization event. This implies that the emplacement of the dome was not coeval with the first mineralization event, as previously suggested by Ramírez et al. (2006).

The convergence of age data at 155–156 Ma apparently indicates a major geological event, which demarks the origin of (1) La Negra andesitic lava series, (2) the first Mantos Blancos mineralization event, and (3) leucogranitic subvolcanic intrusive rocks located 30 km west of the Mantos Blancos district dated at 155.4 ± 2.5 Ma (Rb–Sr isochrone on nonaltered rocks; Pichowiak 1994). The hydrothermal alteration that accompanied the second mineralization event in the main deposit must have occurred within the same time span proposed for that mineralization event, which is between 141.6 ± 0.6 Ma and 142.2 ± 1.0 Ma (see above).

The sericitized plagioclase (sample VO14) from an altered pluton located outside the ore deposit (Fig. 1) displays two plateau ages at 145.6 ± 0.8 and 147.1 ± 0.6 Ma. These ages are close to or concordant with a mini plateau age of 147.5 ± 0.7 Ma for a sericitized plagioclase from the andesitic lava flow (sample VO130B). Furthermore, they are similar to the biotite plateau age of 148.2 ± 0.6 Ma calculated for the intrusion of sample VO8 located in the vicinity of the sample VO14 and 10 km to the east of sample VO130B (Fig. 1). If this last age is valid (i.e., no excess argon), the sericite plateau ages may reflect alteration related to the emplacement of plutonic rocks (sample VO8, ca. 148 Ma). It may also represent a regional scale alteration process, as described for volcanic and plutonic rocks in this region (Losert 1974; Sato 1984). Further work is needed to test this hypothesis.

The significantly lower mini plateau age of 125.4 ± 0.4 Ma obtained on sericitized plagioclase from the andesitic lava flow VO130B could represent an overprinted alteration event affecting the volcanic rocks. Sericitized plagioclase from sample P2.8 displays a poor quality result with a poorly constrained plateau age of 101.6 ± 5.0 Ma.

Summary

Nineteen 40Ar/39Ar plateau ages were obtained from 11 samples of Jurassic andesitic lava series, intrusive bodies, dykes, and alteration minerals from the Mantos Blancos mining district. Although the data concerning rocks from the ore deposit are complex, the Ar/Ar method seems to be capable of detecting complex histories of successive emplacement, alteration, and thermal resetting on a suite of rock samples.

The La Negra Formation series is precisely dated for the first time on one lava flow from the district at 156.3 ± 1.4 Ma. At least one intrusive magmatic event is precisely dated at 141.6 ± 0.6 Ma (MB02.4) and 142.2 ± 1.0 Ma (VO16). Another older magmatic event at 148.5 ± 0.5 Ma (on biotite) may exist but must be confirmed by extra data.

The first mineralization event and associated phyllic alteration affecting the rhyolitic dome are dated on strongly sericitized plagioclases at around 155–156 Ma and are contemporaneous with the La Negra Formation. These data and recent U/Pb ages on the rhyolitic dome (Cornejo et al. 2006) suggest that the first hydrothermal event is not related to the dome emplacement, as previously suggested by Ramírez et al. (2006). The second and main mineralization event occurred around 142 Ma, as demonstrated by plateau ages obtained on both the intrusions genetically linked to this event and one late-mineralization monzonitic dyke crosscutting the orebody. Hydrothermal events associated with the younger magmatic pulse (~142 Ma), and probably with the older one (~148 Ma), have been detected by analyzing strongly sericitized single grains of plagioclase. A still younger sericitization age at 125.4 ± 0.4 Ma must be confirmed by extra age data.

The geochronological data confirm the existence of two distinct hypogene mineralizing events in the genesis of the Mantos Blancos ore deposit, as proposed by Ramírez et al. (2006).