Keywords

1 Introduction and Geological Setting

La Horqueta Formation (Dessanti 1956; González Díaz 1981) crops out on a 12 km-wide strip developed from the Seco de las Peñas River to Agua de la Piedra creek (Fig. 1; Cuerda and Cingolani 1998; Cingolani et al. 2003a), within the San Rafael block . It is in tectonic contact with Carboniferous units, either by reverse faults or by an angular unconformity (Tickyj et al. this volume).

Fig. 1
figure 1

a Geological sketch map of the studied area within the San Rafael block , where outcrops and sampling zone of La Horqueta Formation are located northward. b Regional view toward the West near La Horqueta type area. It is shown the deformed outcrops of La Horqueta Fm superposed by the Upper Paleozoic El Imperial Fm. The Quaternary Diamante Volcano is also shown

La Horqueta Formation was deposited in a marine environment and comprises dominantly metasandstones, although metasiltstones, metapelites, and rare metaconglomerates are also present. The matrix of the metasandstones was recrystallized into chlorite, illite, quartz, albite, and minor smectite. Foliation is penetrative in some layers; ductile deformed clasts are present as well as pseudomatrix. In less-deformed metawackes, the relictic clasts are mainly composed of monocrystalline and polycrystalline quartz, sedimentary and metasedimentary lithoclasts, with scarce volcanic and limestone lithoclasts, and rare feldspars. The fine-grained levels are metamorphosed to phyllites and they comprise oriented illite and chlorite with scarce quartz and feldspar grains (Tickyj et al. this volume). In several outcrops quartz veins cutting the La Horqueta unit are conspicuous (Fig. 2c, d, e). Toward north, (Los Gateados river; Fig. 1) the unit consists of muscovite–biotite schists interlayered with quartzitic schists showing granolepidoblastic textures.

Fig. 2
figure 2

a Google satellite image around Diamante river type section, within the area of the La Picaza old mine and Agua de la Chilena. b Outcrops of La Horqueta Formation at Agua de la Piedra section. c, d, e Some details of abundant quartz veins crosscutting folded layers of La Horqueta Formation at Agua de la Chilena and Agua de la Piedra outcrops

The base of the succession is not exposed and it is overlaid through unconformity by the Upper Carboniferous marine-glacial-continental unit (El Imperial Formation). La Horqueta Formation was affected by deformational events; it is folded and develops cleavage. The regional metamorphic conditions slightly increase from south to north (Criado Roqué 1972; Criado Roqué and Ibáñez 1979), ranging from very low (anchizone) to low grade (epizone). Maximum Silurian–Devonian depositional age was determined using U-Pb detrital zircon dating (Cingolani et al. 2008; Tickyj et al. this volume).

La Horqueta Formation is intruded by a granitic stock known as Agua de la Chilena, which extends over 5 km2 of the northwestern part of the San Rafael Block and it is covered by Quaternary volcanic rocks (Cingolani et al. 2005a). The stock is composed of diorites, tonalites and biotitic-horblendiferous, and leucocratic granodiorites; grain size is medium to fine. Xenoliths and enclaves are frequent. Their mineralogical constituents are subhedral to anhedral quartz (27–33%), subhedral to euhedral altered alkaline feldspars (10–20%) and plagioclases (51–59%) as well as biotite, amphiboles, and epidotes. Accessory minerals are apatite and zircon, and scarce titanite. The texture is granular hypidiomorphic and locally pegmatitic.

The stock was dated using the Rb-Sr method on whole rock and biotite on one sample, giving an age of 256 ± 2 Ma, with a Ri = 0.7073 ± 0.0001. However, a more accurate age is obtained combining whole-rock Rb-Sr of five samples with data from feldspars and biotite, which assigned a Guadalupian–Lopingian age of 257 ± 3 Ma, with a Ri = 0.7069 ± 0.0003 and MSWD of 8.6 following ISOPLOT model 3. The stock would have been emplaced after the Orogenic San Rafael Phase (Asselian—Sakmarian), and during the latest stages of volcanic activity linked to the Cochicó Group (Cingolani et al. 2005a). Based on the presence of amphibole together with biotite a metaluminous series with calcoalcaline characteristics can be assumed, which are typical of magmatic arcs related to the subduction of the paleopacific plate within the southwestern margin of Gondwana. This Permian magmatism could have originated the mineralization of El Rodeo and Las Picazas sulfides mines (arsenopyrite, pyrite and sphalerite), as well as the hydrothermal hematite of the Alto Molle mine (within the La Horqueta Formation ; (Cingolani et al. 2005a).

The present work focus on provenance analyses of La Horqueta Formation based on whole-rock geochemistry and Sm–Nd data, which altogether with the information presented in Tickyj et al. (this volume), particularly regarding detrital zircon dating and Rb-Sr whole-rock data, give insights into source composition and the comparison with Río Seco de los Castaños Formation.

2 Sampling and Analytical Techniques

Sampling was done (see Tickyj et al. this volume) at Los Gateados and La Horqueta type areas (Fig. 1 and Table 1). A total of eighteen samples were selected for chemical analyses done at ACME Labs, Canada. Major elements were obtained by inductively coupled plasma element spectroscopy (ICP-ES) on fusion beads and the loss on ignition (LOI) was calculated by weight after ignition at 1000 °C. Mo, Cu, Pb, Zn, Ni, As, Cd, Sb, Bi, Ag, Au, Hg, Tl, and Se were analyzed by inductively coupled plasma mass spectroscopy (ICP-MS) after leaching each sample with 3 ml 2:2:2 HCl–HNO3–H2O at 95 °C for 1 hour and later diluted to 10 ml. Rare earth elements (REE) and certain trace elements (Ba, Be, Co, Cs, Ga, Hf, Nb, Rb, Sn, Sr, Ta, Sc, Th, U, V, W, Zr, Y, La, Ce, Pr, Nd, Sm, Eu, Gd, Tb, Dy, Ho, Er, Tm, Yb, Lu) were analyzed by ICP-MS following lithium metaborate/tetraborate fusion and nitric acid digestion. Detection limits are: 0.01% for major elements, except for Fe2O3 which is 0.04%; 0.1 ppm for Mo, Cu, Pb, Cd, Sb, Bi, Ag, Tl, Cs, Hf, Nb, Rb, Ta, U, Zr, Y, La, and Ce; 1 ppm for Zn, Ba, Be, Sn, and Sc; 0.5 ppm for As, Au, Ga, Sr, and W; 0.01 ppm for Hg, Tm, Lu, and Tb; 0.2 ppm for Co and Th; 8 ppm for V; 20 ppm for Ni; 0.002 ppm for Cr; 0.02 ppm for Pr, Eu, and Ho; 0.3 ppm for Nd; 0.05 ppm for Sm, Gd, Dy, and Yb and 0.03 ppm for Er. Data are presented in Tables 2, 3, 4 and 5.

Table 1 GPS location of studied samples (after Tickyj et al,. this volume)
Table 2 Major elements (expressed in %) of La Horqueta Formation
Table 3 Trace elements (expressed in ppm) of La Horqueta Formation
Table 4 Rare earth elements (expressed in ppm) of La Horqueta Formation
Table 5 Selected ratios of La Horqueta Formation

Seven whole-rock samples were used for Sm–Nd determinations; they were spiked with mixed 149Sm–150Nd tracer and dissolved in Teflon vial using an HF–HNO3 mixture and 6 N HCl until complete material dissolution. The cationic resin AG-50 W-X8 (200–400 mesh) were used for column separation of the REE, followed by Sm and Nd separation using anionic politeflon HDEHP LN-B50-A (100–200 μm) resin according to Patchett and Ruiz (1987). Each sample was dried to a solid and then loaded with 0.25 N H3PO4 on appropriated filament (single Ta for Sm and triple Ta–Re–Ta for Nd). Isotopic ratios were measured in static mode with a VG Sector 54 multicollector mass spectrometer at the Laboratorio de Geología Isotópica, Universidade Federal do Rio Grande do Sul (LGI-UFRGS, Porto Alegre, Brazil). 100–120 ratios with a 0.5–1 V 144Nd beam were normally collected. Nd ratios were normalized to 146Nd/144Nd = 0.72190. All analyses were adjusted for variations instrumental bias due to periodic adjustment of collector positions as monitored by measurements of our internal standards. Measurements for the Spex 143Nd/144Nd are 0.511130 ± 0.000010. Correction for blank was insignificant for Nd isotopic compositions and generally insignificant for Sm/Nd ratios. ƒSm/Nd is the fractional deviation of the sample 147Sm/144Nd from achondritic reference and is calculated as (147Sm/144Nd)sample/(147Sm/144Nd)CHUR − 1. The εNd indicates the deviation of the 143Nd/144Nd value of the sample from that of CHUR (DePaolo and Wasserburg 1976) and it is calculated as εNd(0) = {[(143Nd/144Nd)sample(t=0)/0.512638] − 1} * 10,000, whereas εNd(t=420 Ma) = {[(143Nd/144Nd)sample (t)/(143Nd/144Nd)CHUR (t)] − 1} * 10,000. Parameters used are: (147Sm/144Nd)CHUR = 0.1967. (143Nd/144Nd)CHUR = 0.512638. T DM (model ages) were calculated based on the depleted mantle model (DePaolo 1981) and on the three-stage model (DePaolo et al. 1991), as indicated in Table 6.

Table 6 Sm–Nd data of La Horqueta Formation

3 Whole-Rock Geochemistry

Data are presented in Tables 2, 3, 4, and 5. The use of whole-rock geochemistry to described provenance composition has been proven to be useful in the context of the pre-Carboniferous clastic units of the San Rafael block of the Cuyania terrane , as demonstrated by Cingolani et al. (2003b), Manassero et al. (2009), and Abre et al. (2011). Therefore, despite the remobilization that could have occurred due to low-grade metamorphism, it is expected that geochemical proxies using trace and REE of La Horqueta Formation would still reflect source compositions.

In the description of the geochemical proxies that follows, the sample HOR27 is treated separately due to their unique characteristics with respect to the whole dataset. A comparison to Río Seco de los Castaños Formation is introduced, since both units of the San Rafael block show similarities.

La Horqueta Formation shows SiO2 concentrations ranging from 44.69 to 83.17%, Al2O3 is between 7.32 and 24.48%, Fe2O3 ranges from 3.33 to 10.89%, CaO is present in low concentrations (0.59% on average), Na2O contents ranges from 0.08 to 2.16%, whereas K2O is between 1.23 and 6.84% (Table 2). Sample HOR27 has SiO2, Al2O3, and K2O in the range of variation of the unit, but has lower Fe2O3 (2.43%), and higher CaO and Na2O contents (2.63 and 2.68%, respectively). Some of the quartz veins cutting La Horqueta Formation were also analyzed for Ag and Au with negative results.

Weathering: The Chemical Index of Alteration (CIA; Nesbitt and Young 1982) is used to evaluate the extent of primary material transformation caused by weathering. The index is calculated using mole fractions as follows: CIA = {Al2O3/(Al2O3 + CaO* + Na2O + K2O)} × 100, where CaO* refers to the calcium associated with silicate minerals.

For La Horqueta Formation , values range from 61 to 77, indicating intermediate weathering conditions. The exception is sample HOR27 that has a CIA value of 55, typical of unweathered crystalline rocks of granodioritic composition (Fig. 3a; Table 2; Nesbitt and Young 1989). In the ACNK diagram the samples display a general weathering trend, that starts parallel to the A-CN boundary and to the weathering path of the upper continental crust (UCC), but shows deviation toward the K apex for samples with the highest CIA, indicating K2O enrichment comparing to UCC value (McLennan et al. 2006; Table 2). Such behavior is in accordance to XRD mineralogical and petrographical data. Comparing to data from Río Seco de los Castaños (Manassero et al. 2009) it is evident that both units have the same range of CIA variation, and similar weathering trends, although La Horqueta Formation show higher K2O enrichment (Fig. 3a).

Fig. 3
figure 3

a In the A-CN-K display La Horqueta Formation plot along a vertical array parallel to the expected weathering trend (field of vertical lines) for average upper crustal rocks; UCC values according to Taylor and McLennan (1985). b Th/U versus Th based on McLennan et al. (1993). c Th/Sc versus Zr/Sc display (McLennan et al. 2003): the ratios are typical of sedimentary rocks derived from unrecycled upper continental crust with a minor input of more depleted source composition. d The Y/Ni and Cr/V ratios are used to discriminate the input of a mafic source (McLennan et al. 1993). UCC values according to McLennan et al. (2006), while PAAS is following Taylor and McLennan (1985). e Chondrite normalized REE patterns; PAAS post-Archean Australian shales pattern (Nance and Taylor 1976) is draw for comparison. EuN/Eu* = EuN/(0.67SmN + 0.33TbN). f La/Th versus Hf after McLennan et al. (1980). For comparison range of variation of geochemical proxies of Río Seco de los Castaños Formation are shown as gray areas

Weathering effects could also be detected analyzing Th/U ratios and their variation regarding Th concentrations (McLennan et al. 1993), although attempts performed on Ordovician clastic units of the San Rafael block have led to uncertain results (Abre 2007; Abre et al. this volume). Compared to UCC averages of Th (10.7 ppm) and U (2.8 ppm) according to McLennan et al. (2006), most of the samples from La Horqueta Formation (including HOR27) are enriched in Th concentrations (14 ppm on average), and show similar to enriched U concentrations (4 ppm is the average of the unit as well as the U content of HOR27). Nonetheless, the Th/U ratios are in general around 3.5–4, which is typical for unrecycled samples derived from the UCC, although a few samples have higher Th/U ratios (maximum value 6.02) indicating weathering (Fig. 3b). Sample HOR27 has a Th/U ratio of 3.99, therefore clustering along with all samples. The Río Seco de los Castaños Formation shows a narrower spread of data, since values lower than the UCC are not present (Fig. 3b).

Recycling: Resistant heavy minerals tend to be concentrated during reworking, and this effect could be deciphered by analyzing the content of elements typically carried on such heavy minerals. The Zr/Sc ratios of La Horqueta Formation range between 5.2 and 36.6, indicating that the detrital components were not recycled (Fig. 3c). Sample HOR27, with a Zr/Sc ratio of 29.31 shows the same tendency. Noteworthy are those samples with Zr/Sc ratios lower than the UCC average (14; McLennan et al. 2006), which is a response of Sc concentrations above and Zr lower than UCC average (13.6 ppm and 190 ppm, respectively; see Table 3), indicating a derivation from a source less evolved than the average UCC. The same was deduced for Río Seco de los Castaños Formation, since its narrower range of values indicate a depleted source composition and recycling was even less important comparing to La Horqueta Formation (Fig. 3c; Manassero et al. 2009).

Source composition: The average composition of the source rocks (s) could be determined through REE patterns, the character of the Eu anomaly and the content of certain trace elements which tends to be either concentrated in silicic (such as La and Th) or mafic (Sc, Cr, Co) rocks (Taylor and McLennan 1985).

The Th/Sc ratios of La Horqueta Formation range from 0.50 to 1.36 (average 0.90). Those samples with Th/Sc ratios around average UCC (0.79; McLennan et al. 2006) are explained as derived from a felsic source with a composition similar to average UCC, with low Th/Sc ratios could have been derived from a depleted source (Fig. 3c, where it is clear the similarity to Río Seco de los Castaños Formation). La/Th ratios between 2.78 and 4.93 support felsic source rocks composition (Fig. 3f), as it was deduced for Río Seco de los Castaños Formation, although the latter display a narrower range of values ruling out any recycling (Manassero et al. 2009). Cr/V and Y/Ni ratios are between average values for UCC and Post-Archean Australian Shales (PAAS), as it is shown in Fig. 3d (see again similarities to Río Seco de los Castaños Formation; Manassero et al. 2009). The exception is sample HOR15 that display a Cr/V ratio of 2.0 due to Cr enrichment compared to UCC average of 83 ppm (McLennan et al. 2006), which along with Zr enrichment and scarce effects of recycling could indicate the presence of Cr-rich resistant heavy minerals such as first-cycle spinel that had been found in several sedimentary sequences of the San Rafael block (e.g., Abre et al. 2009, 2011). Although an ophiolitic source can be neglected, most of the samples show contents of Sc (up to 31 ppm), Cr (up to 231 ppm), and V (up to 219 ppm) above average UCC, indicating the influence of a less-evolved source.

The REE contents of La Horqueta Formation are enriched compared with PAAS, although the chondrite normalized REE patterns are parallel (Fig. 3e). The negative Eu anomaly (EuN/Eu* of 0.66 on average) typical for detrital rocks derived from UCC is present. Noteworthy is the REE pattern of sample HOR27, which is parallel to PAAS regarding Light-REE, but shows extremely low concentrations of Heavy-REE; additionally, the Eu anomaly is the less negative of all samples analyzed (0.82; Table 5). Río Seco de los Castaños Formation also shows REE patterns parallel to PAAS with certain enrichment particularly regarding Heavy-REE and a negative Eu anomaly of 0.68 on average (Manassero et al. 2009), resulting therefore very similar to La Horqueta Formation , suggesting similar source composition.

The geochemical composition of sample HOR27 is not easily explained, since in summary, it shows CIA values typical for unweathered granodioritic rocks but a REE pattern that does not match such igneous compositions neither any other; therefore, laboratory errors cannot be rejected.

4 Isotope Geochemistry

Sm–Nd: Seven samples from La Horqueta Formation were analyzed using the Sm–Nd system and data are presented on Table 6. εNd (t = 420 Ma) values range from −1.49 to −6.53, the ƒSm/Nd are between −0.35 and −0.43, while the T 1DM (average crustal residence age calculated following DePaolo 1981) range from 1.17 to 1.50 Ga and T 2DM (calculated following DePaolo et al. 1991) ranges from 1.28 to 1.66 Ga. These data indicate that the average Nd isotopic signatures of the source rocks are rather a mix of both, an old upper crust and an arc component, and opposite to Río Seco de los Castaños Formation, fractionation is absent (Fig. 4a).

Fig. 4
figure 4

a ƒSm/Nd versus εNd(t) and b εNd versus age. The range of Cerro La Ventana Formation Nd data (Cingolani et al. 2005b) and of Río Seco de los Castaños Formation are drawn for comparison. CHUR Chondritic Uniform Reservoir

The εNd values are within the range of variation of data from the Cerro La Ventana Formation (Fig. 3b), which is part of the basement of the Cuyania terrane outcropping within the San Rafael block (data from Cingolani et al. 2005b; Cingolani et al. this volume); The T DM ages are comparable to those from Mesoproterozoic basement rocks of the Cuyania terrane studied by Kay et al. (1996) and summarized in Cingolani et al. (this volume) consistent with derivation from the nearest Grenvillian-age crustal source such as Cerro La Ventana Formation, exposed in the Ponón Trehué area. Furthermore, the Nd signature is similar to that of Río Seco de los Castaños Formation (Manassero et al. 2009), as well as to the Ordovician Pavón and Ponón Trehué Formations (Cingolani et al. 2003b; Abre et al. 2011). Ordovician to Silurian clastic sequences studied from the Precordillera s.st. (as part of the Cuyania terrane ) also display the same range of εNd and T DM values (Gleason et al. 2007; Abre et al. 2012).

Rb-Sr: Seven metapelites and six samples of micaschists were analyzed by Tickyj et al. (2001) and Tickyj et al. (this volume). The recalculated age obtained using an Isoplot/Ex Model 3 (Ludwig 2008) is 372.8 ± 8.1 Ma, initial 87Sr/86Sr: 0.7164 ± 0.0012 and MSWD 8.4. The Rb-Sr data indicate that the low-grade metamorphism and folding events of La Horqueta Formation are Devonian in age. Furthermore, the age obtained agree with previous K-Ar ages reported by Linares and González (1990). The same methodology applied to Río Seco de los Castaños Formation indicate a very low-grade metamorphic age of 336 ± 23 Ma (Lower Carboniferous, Cingolani and Varela 2008). This is younger than the Rb-Sr metamorphic age of La Horqueta Formation , although both are probably linked to the final Chanic tectonic phase that occurred as a response to the accretion of Chilenia terrane at western proto-Andean Gondwana margin (Ramos et al. 1984).

5 Provenance Discussion

Geochemical analyses and particularly the Th/Sc and La/Th ratios, REE patterns and Eu anomalies indicate a derivation from a felsic source with a composition similar to average UCC, although Th/Sc and Zr/Sc ratios lower than the UCC average, along with Sc, Cr, and V concentrations suggest a provenance from source rocks slightly less evolved than the average upper continental crust. Similar conclusions were found for Río Seco de los Castaños Formation (Manassero et al. 2009) as well as for the Ordovician sequences of the San Rafael Block (Cingolani et al. 2003b; Abre et al. 2011). The agreement observed when comparing with the Sm–Nd signature of the Mesoproterozoic basement (Cerro La Ventana Formation) give further provenance constraints.

Zircon age patterns for La Horqueta Formation indicate four main populations, which in order of abundance correspond to the Mesoproterozoic (Grenvillian cycle), Neoproterozoic (Pampean–Brasiliano cycle), Paleoproterozoic and Upper Cambrian–Devonian (Famatinian cycle). A main derivation from the Mesoproterozoic basement of the San Rafael Block and Pampia terrane is supported, as well as a detrital input from the Río de la Plata craton and the Famatinian belt. Sample HOR27 shows however a different pattern, with a dominance of Famatinian grains, followed in abundance by the Mesoproterozoic population, the Neoproterozoic, the Paleoproterozoic and showing a few Neoarchean detrital zircon grains; it also comprises the younger detrital zircons found within the unit (ca. 0.4 Ga; Cingolani et al. 2008; Tickyj et al. this volume).

These age patterns are rather different comparing with Río Seco de los Castaños Formation which shows a dominance of Famatinian and Pampean–Brasiliano detrital zircons and lower amounts of Mesoproterozoic grains (Fig. 5). Such differences in age patterns indicate that the source rocks providing detritus to both basins were not the same.

Fig. 5
figure 5

Comparative U-Pb ages of the detrital zircons represented in percentage “pie diagrams” for La Horqueta and Río Seco de los Castaños Formations. Different colors are from recognized South American orogenic cycles: Archean to Paleoproterozoic; Mesoproterozoic, Brasiliano (Neoproterozoic–Early Cambrian), and Famatinian (Middle Cambrian–Devonian)

6 Conclusions

  1. (a)

    CIA values of La Horqueta Formation indicate intermediate weathering conditions, and samples with the highest CIA are enriched in K2O comparing to UCC; some Th/U ratios support this. Zr/Sc ratios point to mainly unrecycled detritus.

  2. (b)

    Th/Sc, La/Th, and Th/U ratios, REE patterns, and negative Eu anomalies are typical for detrital rocks derived from unrecycled UCC. However, Sc, Cr, and V concentrations along with low Th/Sc ratios suggest a provenance from source rocks slightly less evolved than the average upper continental crust. Sources compositions are similar to that of Río Seco de los Castaños Formation.

  3. (c)

    The εNd values are within the range of variation of data from the Mesoproterozoic Cerro La Ventana Formation, which is part of the basement of the Cuyania terrane outcropping within the San Rafael Block . These isotopic data are also similar to that of the Río Seco de los Castaños Formation.

  4. (d)

    Detrital zircon age patterns indicate a provenance from Mesoproterozoic (Grenvillian), Pampean–Brasiliano, and Famatinian cycles, in order of abundance.

  5. (e)

    Comparison with Río Seco de los Castaños Formation indicate similar source composition based on geochemical proxies but the age of such rocks are different, according to detrital zircon age patterns. The main difference is that the Río Seco de los Castaños Formation contains larger proportion of Ordovician zircon grains while the La Horqueta Formation contains few Ordovician zircon ones and much more Grenville-aged zircons.