Propylitic alteration, as a distal alteration zone of porphyry-centered hydrothermal systems, has been traditionally considered poor of useful information for mineral exploration. Therefore, the controls on the formation of this alteration facies have not been studied in depth since a few years ago. However, several recent papers demonstrated that the trace element composition of epidote and chlorite has significant potential to provide spatial compositional trends useful for mineral exploration (i.e., vectoring toward mineralized/hydrothermal centers within propylitically altered terrain, differentiate propylitic alteration around ore deposits from the similar assemblages developed during regional metamorphism), because they are able to record the dispersion of many elements as a consequence of changing physicochemical conditions across the propylitic facies (Cooke et al., 2014; Wilkinson et al., 2015; Cooke et al., 2020a; Ahmed et al., 2020; Baker et al., 2020; Cooke et al., 2020b; Pacey et al., 2020; Wilkinson et al., 2020; Li et al., 2022). In this scenario, data on the mineral chemistry of propylitic alteration minerals related to ore deposits worldwide are needed, to improve understanding of this widespread and still incompletely studied alteration style.

These features are clearly exposed in southern Ecuador where the metallogenic belts, dating to Jurassic and to Tertiary times, host notable Au, Cu, Mo, and Ag resources (PRODEMINCA, 2000; Schutte et al., 2010; Schutte et al., 2012). In the Ponce Enriquez Mining District (PEMD hereafter), the Cretaceous tholeiitic basalts of Pallatanga Formation (Fm.), covered by Miocene ignimbrites, are intruded by Miocene Au-porphyries and host an important gold mineralization. These mafic rocks are deeply altered by a mineralogical assemblage dominated by the occurrence of chlorite and epidote. The Ponce Enriquez gold mineralization is not well known. The environmental implications linked to the mining activity in the area have been tackled in several works (PRODEMINCA, 2000; Appleton et al., 2001; Carling et al., 2013; Pesantes et al., 2019; Jiménez-Oyola et al., 2021a; Jiménez-Oyola et al., 2021b), but the fluid evolution, the mechanisms that control the formation of this huge gold mineralization hosted within “green rocks,” the relationship between the mineralization and the propylitic alteration, and the possible relationships with the nearby porphyry systems are not still fully reconstructed.

In this paper, we present a case study concerning the propylitic alteration of PEMD, focusing on the trace element chemical composition of the secondary minerals (particularly epidote and chlorite) and on fluid inclusion data. The aim is to shed light on the fluid evolution and conditions that led the development of propylitic alteration, and on the possible relationships with the occurrence of porphyry systems only few hundreds of meters far. The results help improving the knowledge of the “green rocks” in the southern Ecuador, opening the possibility to employ the epidote and chlorite chemistry as a tool in mineral exploration to assess the relationships to causative porphyry systems. Moreover, we add new data for understanding the formation of propylitic alteration style and for using trace element chemistry of epidote and chlorite in mineral exploration elsewhere.

The PEMD (Azuay Province) is sited in southern Ecuador on the western flank of the Western Cordillera and is comprised between two hydrographic basins: the Río Siete basin to the south and the Río Guanache basin to the north (Figure 1). The Pallatanga Fm., the Yunguilla Fm., and the Saraguro Group (Las Trancas Fm.) outcrop in this area (Figure 1) (PRODEMINCA, 2000). The Pallatanga Fm. (Cretaceous) corresponds to an ophiolitic association, and it is composed of a sequence of oceanic basalts, pillow lavas, hyaloclastites, and massive dolerites with aphanitic and variolitic textures (Kerr et al., 2002; Srivastava et al., 2008). The material presents geochemical features of accreted igneous oceanic terranes that include calc-alkaline material of an island arc sequence, and oceanic plateau material (Kerr et al., 2002). Most of the rocks of the Pallatanga Fm. are aphanitic, but locally these are enriched in phenocrysts of pyroxene and plagioclase and locally small gabbroid bodies occur inside the main body. From the geochemical point of view, the main rocks of the Pallatanga Fm. are basalt with a SiO₂ content varying from 47% to 53% and MgO from 7% to 9% and tholeiitic affinity (Kerr et al., 2002). With regard to the trace elements, they have high concentrations in Y, Nb and V. The La/Yb ratio is low (<2.5), the Zr/Th ratio is >200 while the chondrite-normalized REE show flat patterns (Kerr et al., 2002).

The Yunguilla Fm. (Upper Cretaceous, Maastrichtian) corresponds to a turbidite fan sequence made up of siltstones, feldspathic quartz-type sandstones, and quartz arenites (Duque, 2000). According to Thalmann (1946), this formation is more than 1,000 m thick.

The Saraguro Group (Late Eocene—Early Miocene) is mainly composed of welded ignimbrites and lavas interbedded with sedimentary rocks (Siravo et al., 2021). It is divided into two main sequences: the lower sequence (Portovelo Unit and Las Trancas Fm.), composed of quartz-diorites and granodiorites/tonalites (Egüez et al., 2017), and the upper sequence (Plancharumi, La Fortuna, Jubones, and La Paz Fm.), made up of rhyolitic ignimbrites (PRODEMINCA, 2000). The Portovelo Unit is composed of andesitic to basaltic lavas and andesitic ignimbrites that outcrop north of the Piñas-Portovelo Fault. The Las Trancas Fm. (Oligocene) sited at southwest of PEMD, lies on the Pallatanga Fm. and it is composed of andesitic to dacitic ignimbrites, conglomerates, argillites, and feldspars, muscovite, and quartz-rich sandstones (PRODEMINCA, 2000). The Plancharumi Fm. (Oligocene) is made up of stratified rhyolitic ignimbrites, rhyolitic breccias, and lavas interbedded with ash, sandstone, and claystone (Dunkley and Gaibor, 1997). The La Fortuna Fm. (Early Miocene) is located northward of the Jubones Fault and is composed by aphyric ignimbrites. It reaches a maximum thickness of 600 m (Pratt et al., 1997). The Jubones Fm. (Early Miocene) represents the top of the Saraguro Group, and it can reach 500 m of thickness (Hungerbühler, 1997). It is a rhyolitic welded ignimbrite rich in quartz, biotite, and plagioclase crystals. The Saraguro Group presents multiple Tertiary intrusions, and it hosts several epithermal deposits of Miocene age that make up the Azuay and El Oro districts (PRODEMINCA, 2000; Schutte et al., 2012).

In the PEMD, porphyry Cu (Chaucha) and porphyry Au-Cu deposits (Gaby-Papa Grande) are related with the large Oligocene to Miocene intrusive centers. The Gaby-Papa Grande hornblende Au-Cu porphyry (20.54 ± 0.08 million years, Schutte et al., 2012) is located to the north of Bella Rica Mining Cooperative (Figure 1). It is intruded in Cretaceous mafic volcanic rocks and shows euhedral hornblende and plagioclase phenocrysts. The rocks outcropping around these porphyry intrusions belong to the Pallatanga Fm. and mostly consist of dark greenish-gray basaltic andesites with aphanitic texture and andesites with a porphyritic texture with millimeter-sized or close to a centimeter plagioclase phenocrysts with breccias, veins, and mineralization in stockwork (PRODEMINCA, 2000).

In the PEMD sector, different fault systems with a N-S, NW-SE, or WNW-ESE orientation (Figure 1) of Cenozoic age (PRODEMINCA, 2000) are present. In the northeast of Ponce Enríquez, the Tenguel River Fault is present while in the southwest the Margarita Fault divides the Bella Rica basalts (Pallatanga Fm.) from the turbiditic sediments (Yunguilla Fm.) The Guanache Fault is north dipping (50°–80°), parallel to the Guanache River (Figure 1) and the Eloy Calderón vein is located in correspondence of this fault (Dunkley and Gaibor, 1997). The other main fault system presents an E-W and NE-SW directions. The Los Ratones Fault has dips between 60° and 90° to the NW; in the NE sector, a mineralized fault is observed in the Muyuyacu sector. The Río 9 de Octubre Fault tilts to the SE and it extends to the Papa Grande porphyritic complex (PRODEMINCA, 2000). Here the mineralized veins show E-W strike direction. The mineralized Tres de Mayo Fault presents a N-S direction, and it dips to the east displaced by the regional transversal faults. In general, the hydrothermal mineralization in the Bella Rica area is N-S oriented on transverse fault structures and east-dipping veins due to an existing paleosuture. Minerals such as pyrrhotite, pyrite, galena, hematite, sphalerite, chalcopyrite, magnetite, molybdenite, calcite, chlorite, among others, have been identified (AMBIENCONSUL, 2017).

Twenty-one rock samples were collected in the area of the Bella Rica Mining Cooperative and in the northern side of the area (Figure 1). Thirty-one thin sections were prepared in the Petrography Laboratory of the Engineering in Earth Sciences (FICT) of the Polytechnic University of Litoral (ESPOL) of Guayaquil, Ecuador. Polished thin sections for reflected light optical microscopy, scanning electron microscopy - backscattered electron imaging (SEM-BSEI), electron probe micro-analyzer (EPMA), and Laser Ablation Inductively Coupled Plasma Mass Spectrometry (LA-ICPMS) analysis were prepared in the laboratories of the Earth Science Department of Pisa, Italy.

Bulk rock geochemistry was carried out on 10 samples on pulps obtained by crushing in a jaw mill and pulverized in agate jars rock chips picked out of macroscopically visible quartz and sulfide veins, to gain information on the propylitically altered host rock. Major element composition and Loss On Ignition (LOI) were determined at Actlabs (Canada) by the Inductively Coupled Plasma Optical Emission Spectrometry (ICP-OES) technique using the Varian 720-ES. Bulk rock trace elements were determined after HNO₃-HF acid-digestion by means of a Perkin-Elmer NexION 300x ICP-MS at Earth Science Department of Pisa.

FE-SEM-EDS analyses were performed using the FEI Quanta FEG 450 apparatus housed at the Centre for Instrument Sharing of University of Pisa (CISUP). The EPMA analyses of chlorite, epidote, feldspar, and amphibole were performed on carbon-coated and polished thin sections, using a JEOL JXA-8200 Super-Probe at the Earth Sciences Department Ardito Desio, University of Milan, Italy. Operating conditions were 15 kV and a beam current of 5 nA. The instrument is equipped with five wavelength-dispersive spectrometers (WDS) with a range of LiF, PET and TAP crystals, and one additional EDS detector, and was internally calibrated using mineral standards and metals of natural and synthetic origin, notably: grossular (Si Kα, Ca Kα, Al Kα), omphacite (Na Kα), forsterite (Mg Kα), fayalite (Fe Kα), ilmenite (Ti Kα), orthoclase (K Kα), rhodonite (Mn Kα), pure metal Cr (Cr Kα). Raw element data were ZAF-corrected using a phi-rho-Z analysis program, and corrected element contents were converted to oxide contents in weight percent (wt%) assuming stoichiometry. FeO_tot represents total iron (oxide) content.

In-situ trace element analyses on epidote and chlorite, previously analyzed by means of EPMA, were performed by laser ablation inductively coupled plasma mass spectrometry (LA-ICPMS). The LA-ICPMS analyses were performed using a PerkinElmer NexION 2000 ICP-MS coupled with a New Wave Research-193 Ar-F 193 nm excimer laser, at the Earth Science Department of Pisa, in the laboratories of the Centre for Instrument Sharing of University of Pisa (CISUP). The laser was operated at a repetition rate of 10 Hz using spot sizes of 35 and 40 µm and a 4.6 J/cm² energy density. Synthetic standard glass NIST 612 was used for external calibration. The accuracy of the analyses was monitored using the synthetic standard glass NIST 610 and CaO and SiO₂ was used as internal standards for epidote and chlorite respectively. Data reduction was performed with the software package IOLITE (Paton et al., 2011). Edges of grains were carefully avoided to ensure no contamination of the analyses occurred due to accidental ablation of adjacent minerals.

Doubly polished thin sections (100–300 µm-thick) were prepared for petrography and microthermometric determinations of fluid inclusions. Fluid inclusion analyses were carried out using a Linkam THMS 600 heating-freezing stage at the Earth Sciences Department of the University of Pisa. The accuracy of measurements is estimated at ±2°C at 398°C controlled by the melting point of K₂Cr₂O₇, ±0.1°C at 0°C and ±0.2°C at −56.6°C controlled by using certified pure water and CO₂-bearing synthetic fluid inclusions (Synthetic Fluid Inclusion Reference Set, Bubbles Inc., United States). Salinities of fluid inclusions, expressed in weight percent equivalent NaCl, were calculated from final ice-melting temperature, using the equation of Bodnar (1993) for the H₂O-NaCl system. Eutectic temperature (when observable) was used to constrain overall composition of trapped fluids by comparison with published data for various salt-water systems (Crawford, 1981).

Despite the economic importance of the gold mineralization occurring in Southern Ecuador very little is known about the physical-chemical conditions of fluids that led to the formation of the hydrothermally altered rocks hosting the mineralization. This work aims to fill this gap, by using the PEMD as a case study for improving the general knowledge of the propylitically altered rocks in Ecuador and add new data for the use of trace element chemistry of epidote and chlorite in mineral exploration elsewhere.

In the PEMD, the pervasive hydrothermal alteration, mainly affected the mafic volcanic rocks of the Pallatanga Fm. and is composed of quartz, epidote, and chlorite, with subordinate amounts of illite, titanite and prehnite. This alteration assemblage may be similar to that of typical lower greenschist facies metamorphism (e.g., Turner, 1981). However, propylitic alteration, normally developed as an outer halo of many porphyries Cu-Au-Mo deposits (Lowell and Guilbert, 1970; Seedorf et al., 2005; Sillitoe, 2010), is similarly expressed by the same mineralogical assemblage (quartz, chlorite, epidote, prehnite). Recent papers demonstrated that the trace element content of epidote (particularly As and Sb content) can be used to discriminate epidote of metamorphic origin from epidote owing to propylitic alteration halo of porphyry copper deposits, and also between epidote from giant porphyry deposits and epidote from small porphyry deposits (Cooke et al., 2014; 2020a; b; Wilkinson et al., 2020). In particular, Cooke et al. (2020b) showed that epidote from the giant porphyry deposits is characterized by As and Sb contents that are one or more orders of magnitude greater than epidote from small porphyry deposits, whereas metamorphic epidote is characterized by very low As and Sb (commonly below detection limits; Baker et al., 2017). The high trace element and metal content of epidote from the propylitic halo of porphyry deposits has been attributed to the capacity of porphyry systems (large porphyry systems in particular) to flux metals, and this can be recorded in the trace element geochemistry of the distal propylitic alteration minerals (e.g., epidote), with greater distal pathfinder metal contents detected in epidote peripheral to larger deposits (Cooke et al., 2014). The As and Sb content of epidote analyzed in the altered mafic volcanic rocks of the Pallatanga Fm. fall in the field of porphyry deposits (Figure 11). This indicates that the pervasive alteration characteristic of mafic volcanic rocks of the Pallatanga Fm. can be related to the propylitic alteration halo of porphyry copper deposits. Most of the As and Sb data on epidotes plots in the lower left part of the porphyry data field (Figure 11) and therefore could be interpreted as epidote collected from either close to a giant porphyry deposit or distal to a small porphyry deposit. The coexistence of pyrite with epidote must also be considered, since As and Sb are more likely to be incorporated into pyrite than epidote when these minerals co-precipitate (Cooke et al., 2014; 2020a; b). In our case there is no evidence of the constant coexistence of epidote and pyrite, as a consequence we consider more realistic the second interpretation, i.e., distal to a small porphyry deposit.

Fluid inclusion data support the interpretation of propylitic hydrothermal alteration linked to porphyry systems for the “green rocks” associated with PEMD. Indeed, fluid inclusions in quartz veins associated with epidote and chlorite show that the fluid circulated in the mafic volcanic rocks of Pallatanga Fm. had low salinity (around 1.5 wt% NaCl_equiv.) suggesting a meteoric origin and temperatures between 300°C and 350°C. These data are basically consistent with the conditions of propylitic alteration formation of porphyry-type deposits (Seedorf et al., 2005; Sillitoe, 2010; Ni et al., 2015; Li et al., 2017) and are also typical of the propylitic alteration found in several active hydrothermal systems (Bird et al., 1984; Reyes, 1990; Berger and Velde, 1992). At least two pulses of hydrothermal fluids have been recognized thanks to fluid inclusions: the first one (probably responsible for the main development of propylitic alteration) characterized by temperature around 345°C, the second one with temperature around 305°C, indicating the cooling of the hydrothermal fluids. The lower temperature event corresponds quite well to the modal value of the temperature calculated by chlorite geothermometers (around 315°C, Figure 9). This suggests that chlorite composition may have re-equilibrated with the temperature conditions characteristics of the second and relatively lower temperature fluid. However, assuming a possible error of at least ±30°C for chlorite geothermometry (Fulignati, 2020 and reference therein), it is difficult to discriminate between the two hydrothermal pulses. The temperatures estimated with chlorite geothermometers are basically in the range of the homogenization temperatures measured in fluid inclusions in quartz associated with the PEMD propylitic alteration. This indicates that these homogenization temperatures can be probably considered as representing the true trapping temperature of the fluids, and pressure correction is not required. This also means that propylitic alteration in PEMD developed at relatively low-pressure conditions (below 10–20 MPa) probably under hydrostatic regime, suggesting a shallow depth environment. Fluid inclusion data did not reveal any evidence of boiling processes.

Taking into account the above considerations and the geological context of the area, we hypothesize that PEMD hydrothermally altered rocks could be envisaged as the propylitic halo of the Miocene hornblende-bearing Au-Cu porphyry deposits of Gaby-Papa Grande (Schutte et al., 2012). The application of chlorite proximitor (Wilkinson et al., 2015) gave a calculated distance from the collected samples to the causative porphyry systems ranging between 1,300 and 1,600 m, which may be plausible for Gaby-Papa Grande (Figure 1).

The improved knowledge of the propylitic alteration affecting the Pallatanga mafic volcanics can be useful for interpreting rock alteration and discriminate porphyry-related propylitic style during mineral exploration in this region, as well as elsewhere. An open question concerns the relationships existing between the vein gold mineralization and the propylitic alteration event. Are they coeval or not? To shed light on these topics, pivotal for the comprehension of the Ponce Enriquez gold mineralization genesis, further research has to be addressed to carry out dating data on the mineralization and propylitic alteration mineralogical assemblages.

This work, for the first time, tackles the study of propylitic alteration in the Ponce Enriquez mining district (Southern Ecuador) as a case study to improve understanding of the formation of propylitically altered rocks and their possible use as a mineral exploration tool. The investigation of these “green rocks” is particularly relevant because they host a significant gold mineralization, exploited in the last decades. The main conclusions of the study are as follow:

- The mafic volcanic rocks of the Pallatanga Fm. are affected by a pervasive alteration characterized by a mineralogical assemblage formed by chlorite + epidote + quartz + titanite ± illite ± prehnite ± calcite.

The results of this paper represent the first pieces of the puzzle concerning the PEMD mineralization. These, integrated and corroborated with further analytical studies, will help in defining a conceptual working model for the Ecuadorian gold endowment. This will be useful in the mineral exploration of the Azuay and El Oro Provinces of Ecuador, where several gold mineralizations occur in a general context similar to PEMD, and also in other regions worldwide where gold mineralization occur associated to propylitic alteration facies.

This work furthermore confirms that trace element analysis of epidote and chlorite, coupled with other methods of investigation (i.e., fluid inclusion data), is a powerful tool in mineral exploration, being able to discriminate between propylitic hydrothermal alteration halos formed by porphyry systems from similar mineral assemblages formed through regional metamorphic processes.