Fragments of juvenile Siderian continental crust in the Rio Piranhas-Seridó Domain , Borborema Province , Northeastern Brazil , as deduced by zircon

1CPRM-Serviço Geológico do Brasil (NANA/SUREG-RE), Rua Professor Antônio Henrique de Melo, 2010, Natal, RN, Brazil, CEP: 59078-580 2CPRM-Serviço Geológico do Brasil (SEDE), Setor Bancário Norte SBN Quadra 02, Bloco H Asa Norte, Brasília, DF, Brazil, CEP: 70040-904 3CPRM-Serviço Geológico do Brasil (SUREG-RE), Av. Sul, 2291, Recife, PE, Brazil, CEP:50770-011. 4CPRM-Serviço Geológico do Brasil (GEREMI/SUREG-SP), Rua Costa 55, Consolação, São Paulo, SP, Brazil, CEP 01304-010. 5CPRM-Serviço Geológico do Brasil (NUBA/SUREG-SP) Rua Voluntários da Pátria, 475, 1o andar cj. 10, Curitiba, PR, Brazil, CEP: 80020-926.


Introduction
The Siderian Period (2.5 to 2.3 Ga) of the Paleoproterozoic era has been the target of intense debate in the world's geological literature because there is no consensus about its geotectonic evolution. Lines of research advocate the existence of an apparent tectonic calm-stagnation (shutdown by Condie et al. 2009 or tectono-magmatic Lull by Spencer et al. 2018), while other lines of research argue that this period, in fact, is not so calm in comparison to the processes that generate juvenile crusts (Pehrsson et al. 2014). The use of new tools into regional geological mapping, e.g., aerogeophysical data from gamma spectrometry combined with the popularization and refinement of geochronological (U-Pb) and isopotonic (Sm-Nd) methods, has promoted, in recent years, significant advances in the cartographic individualization of geological units other than those already known in the Borborema Province.
In the Borborema Province, in northeastern Brazil ( Figure  1), rocks from the Siderian age, mostly composed of TTG-type gneisses, are reported in the Middle-Coreaú Domain (Santos et al. 2009); they compose part of the Granjeiro Complex (Hollanda et al. 2015;Pitarello et al. 2019) and locally in the Alto Moxotó Domain (Melo et al. 2002;Santos et al. 2013, Santos et al. 2015Brito Neves et al. 2020). Rocks from the Siderian age have not yet been well documented from a cartographic point of view so far in the central portion of the Rio Piranhas-Seridó Domain (PSD). So far, the existing geological descriptions in the literature have treated its basement as formed mainly by gneisses and migmatites of diverse compositions, with ages ranging from the Riacian to the Orosirian (Hackspacher et al. 1990;Legrand et al. 1991;Dantas et al. 2007;Hollanda et al. 2011;Costa and Dantas 2014;. Neoarchaean and Paleoarchaean ages have also been described for rocks that compose the PSD basement (Oliveira et al. 2013;Costa et al. 2018;Santos et al. 2020;Cavalcante et al. 2019;Ferreira et al. 2020a;2020b).
The present study presents geological, geophysical, isotopic and geochronological integration data, which allowed the individualization of geological units from the Siderian age (one of which was proven to be juvenile) in two areas located in the central part of the PSD. This documentation of juvenile Siderian rocks in the PSD contributes to a better understanding of this period in Borborema, as well as as in other areas of Brazil, such as the Bacajá Domain in the Amazon Craton (Vasquez et al. 2008) and the Mineiro Belt in the São Francisco Craton (Seixas et al. 2012).

Regional Geological Context
This study was carried out in the northern portion of the Borborema Province (Ebert 1970;Almeida et al. 1981), northeast of Brazil (Fig. 1). This province is composed of a complex mosaic of rocks that underwent a tectonic evolution related to a collage of lithospheric fragments from Archean and Paleoproterozoic ages surrounded by sequences of metasedimentary rocks. Such fragments belong mainly to the São Francisco/Congo and São Luis/West African cratons (Trompette 1994;Brito Neves et al. 2002;Cordani et al. 2003). In the final stage of evolution, the Brasiliano orogeny outlined the current structure of the Borborema Province, clearly marked by the establishment of transient shear zones associated with high-volume Ediacaran plutonic activity (Brito Neves 1975;Almeida et al. 1981;Santos and Brito Neves 1984;Jardim de Sá 1994;Vauchez et al. 1995;Neves et al. 1996;Archanjo et al. 1999;Brito Neves et al. 2000). According to Caxito (2013) the geological evolution of the Borborema Province is particularly based on two main hypotheses: (i) in the first, its crust would have grown in complete tectonic cycles with crustal rifting, opening and closing of oceans, establishment of subduction and continental collision zones (Santos 1996;Brito Neves et al. 2000;Medeiros 2004); (ii) in the second, it would have been formed by a single continental block with a stable Archean-Paleoproterozoic basement since 2.0 Ga, with establishment followed by an inversion of the ensialic basins on the basement (Neves 2003;Neves et al. 2006).
In relation to this subdivision, the areas investigated in this study are located in the central portion of the Piranhas-Seridó Domain River (PSD), which would be delimited to the south, east and west, respectively,by the Patos, Picuí-João Câmara and Portalegre shear zones, while the northern limit of this domain is covered by phanerozoic covers of the Potiguar Basin (Fig. 1).
The PSD is described as a domain constituted by a crystalline basement predominantly from the Riacian/Orosian age, containing Archean remnants (Oliveira et al. 2013;Costa et al. 2019;Santos et al. 2020;Cavalcante et al. 2018Cavalcante et al. , 2019Medeiros et al. 2021). The set of these units occurs is partially capped by Neoproterozoic metasupracrustal rocks from the Seridó Group (Jucurutu, Ecuador and Seridó formations, Van Schmus et al. 2003), all of which are intruded by Brazilian igneous bodies, in addition to dikes and Mesozoic and Cenozoic volcanic plugs (Figure 1).
The main geological and geochronological studies carried out on the PSD basement rocks between 1980 and 2000 indicated that the generation of rocks with U-Pb ages ranged between 2.2 and 2.1 Ga (Table 1). These rocks were gathered in the Caicó Complex (Hackspacher et al. 1990;Dantas 1992;Jardim de Sá 1994), whose major constitution is given by orthogneisses of different compositions (sometimes with augen textures), migmatitic gneisses, banded gneisses, metamafic and metaultramafic rocks, paragneisses, paraderivative shales and marbles.
The first pieces of evidence of older rocks for the basement of the PSD were proposed by Negrão et al. (2005) based on isotopic studies by the Sm-Nd method, with ages found of model T agesDM ranging from 2.97 Ga to 3.88 Ga. Later, Dantas et al. (2008) based on U-Pb zircon determinations, identified Siderian ages of 2336 ± 12 Ma in a calcisilicate rock intercalated with amphibolites and 2331 ± 37 Ma in a tonalitic biotite-hornblende gneiss, including these rocks in the Santa Luzia Sequence. In the same region addressed by Negrão et al. (2005) and Dantas et al. (2008), Costa and Dantas (2014) individualized and mapped a Siderian age unit, naming it the Arabia Complex instead of the Santa Luzia Sequence proposed by Dantas et al (2008), owing to issues relative to the stratigraphic nomenclature code, since the term sequence does not apply to the designation of formal units.

Analytical procedures
The samples used in the U-Pb and Sm-Nd petrographic and isotopic studies were collected during the geological mapping work carried out in two areas of the PSD, Area I, north of the domain, and Area II, center of the domain (Fig. 1), with the aid of aerogeophysical data from gamma spectrometry collected by CPRM in the Paraíba-Rio Grande do Norte Aerogeophysical Project (LASA SA and Prospectors 2010). This project surveyed high-resolution aeromagnetic and aerogammaspectrometric profiles, with flight and control lines spaced 500 m and 5,000 m apart, oriented in the NS and EW directions, respectively, with the flight height set at 100 meters over the ground.
The U-Pb geochronological zircon data were collected from three samples from the two study areas: AP-242B (orthogneiss -Area I); AP-01C (amphibolite -Area II); and AP-02 (metamorphic clinopyroxene-hornblendite -Area II). The samples were prepared in the petrography laboratory of the CPRM (Regional Superintendence of Recife) according to the following the steps: (i) crushing of approximately 10 kg of each sample with a reduction to the fraction <2 cm; (ii) milling (Mineral Technology Laboratory of the Department of Mining Engineering, UFPE); (iii) drying the ground sample in an oven at a temperature of 80°C; (iv) sieving to collect the grains that pass through the 100 mesh sieve and are retained in the 170 mesh sieve; (v) gravimetric separation with heavy liquids; (vi) magnetic separation in a FRANTZ isodynamic separator; (vii) separation of the zircon crystals with the aid of a magnifying glass.
The U-Pb zircon isotope analysis of the AP-242B sample was performed in an ionic microprobe at the High Resolution Geochronology Laboratory, located at the University of São Paulo (USP), using the SHRIMP IIe/MC model. The zircon concentrates were mounted in epoxy resin at the Research Center in Geochronology and Isotopic Geochemistry (CPGeo/USP), where the assemblies were imaged by cathodoluminescence (CL) in a Scanning Electron Microscope (SEM).
The acquisition of isotopic data was performed with a 24 μm primary electron beam. The concentration of U-Pb-Th was normalized using the international standard of zircon SL13 (238 ppm) and the 206Pb/238U ratio was bracketed using the Temora 2 natural standard, aged 416.78 Ma (Black et al. 2004). Secondary pattern analyses (Z6266 - 566±4 Ma, Stern & Amelin, 2003) were used to measure accuracy. The analysis conditions were: 6 scans, dead time = 25ηs and source slit = 80μm. For data reduction, the Squid 1.06 program (Ludwig 2002) was used.
Samples AP-02 and AP-01C were analyzed via LA-ICPMS at the Geochronological Research Center (University of São Paulo ) and at the Geochronology Laboratory (Federal University of Ouro Preto, UFOP), respectively. The internal structure of the zircon crystals of these samples was investigated using electron backscattered images (BSE) from a SEM (FEI, Quanta 450) at CPRM in Brasília,, under conditions of 20 kV, about 100 µA. In both laboratories the ratios were determined between isotopes 202 Hg, 204 (Pb+Hg), 206 Pb, 207 Pb, 208 Pb, 232 Th and 238 U.
At UFOP, the analyzss were performed using the ICP-MS Element 2 spectrometer, a thermo-Finnigan magnetic sector monocollector model, coupled to an Excimer M50 193 nm laser system. Grain ablation was performed in 30 µm spots. In all analyses, the BB international standard ) was used to correct the equipment's drift, as well as the fractionation between the U and Pb isotopes. Accuracy was checked through analyses performed using the GJ-1 standard, 608.5±1.5 Ma (Jackson et al. 2004). Data were acquired in peak jumping mode for 20s of background, followed by 20s of sample ablation. Raw data reduction, which includes fixes for background, derived from the equipment and common lead, was performed by the Glitter® software.
The analyses carried out at USP used a LA-MC-ICP-MS Neptune device (Thermo-Finnigan) coupled to the ArF Excimer laser (λ=193nm, Photon Machines). The ablation conditions were: 32-µm spot, frequency of 6 Hz and fluence of about 6 mJ of intensity. Acquisition took place in 50 cycles of 1 second each. The analytical sequence was 2 blanks, 3 GJ-1 standards, 12 samples, 2 blanks and 3 GJ-1 standards. Accuracy was controlled through various analyses in the international Mud Tank standard (Black and Gulson 1978). Data reduction was performed in a spreadsheet developed in the laboratory and includes blank correction, equipment drift and common lead correction.
The calculations for all ages and the generation of graphs were performed in Excel® with the resources of the ISOPLOT 4.15 supplement from the Berkerley Geochronology Center (Ludwig 2012).
The sample AP-242B was analyzed by the Sm-Nd method in whole rock at the Geochronology Laboratory of the University of Brasília (UnB), following the methodology described by Gioia and Pimentel (2000). In this procedure, about 50 mg of powdered sample was mixed with a tracer solution of 149Sm and 150Nd. The sample was dissolved in Savillex® capsules through successive acid attacks on HF, HNO3 and HCl. The Sm and Nd contents were extracted through cation exchange columns, made in Teflon and filled with LN-Spec resin. The Sm and Nd salts were deposited on rhenium filaments with nitric acid and evaporated. The readings of the ratios were carried out in a Finnigan MAT 262 mass spectrometer in static mode. The 143 No/ 144 Nd ratio was normalized on the basis of the 146 No/ 144 Nd ratio of 0.7219. The value of TDM was calculated using the model of De Paolo (1981), using the ISOPLOT 4.15 software (Ludwig, 2012). The uncertainties for the 147 Sm/ 144 Nd e 143 Nd/ 144 Nd ratios are less than ±0.5% (2σ) and ±0.005% (2σ), respectively, based on repeated analyses in international standards BHVO-1 and BCR-1.

Petrography of the Arabia Complex
The rocks that compose the Arabia Complex were identified in two work areas, I and II (Figures 1, 2). Area I is located in the northern portion of the Piranhas-Seridó River Domain between the municipalities of Lajes and Pedro Avelino, while Area II is located in the central portion of this same domain, around the Ferro de Saquinho deposit, in the municipality of Cruzeta. In Area I, the rocks that form the Arabia Complex are predominantly composed of orthogneisses and migmatitic, leucocratic, medium to coarse inequigranular gneisses with granodioritic to tonalitic composition, sometimes monzogranitic (Figures 3 A, B). There are also lenses and metric bodies of fine to medium textured amphibolites, embedded in orthoderived rocks ( Figure 3C), or in the form of centimeter to metric xenoliths in orthoderived rocks of this complex ( Figure 3D).
In order to perform the geological interpretation and correlation of the gamma spectrometric data, the ternary composition (RGB) of the three radioelements potassium (K), thorium equivalent (eTh) and uranium equivalent (eU) was used; there was differentiation radiometric between the rocks from Area I. The composition allowed the individualization of the rocks that compose the Arabia Complex from the other adjacent orthogneissic rocks, which are correlated to the Caicó Complex. Ternary composition data (RGB) ( Figure  5A) show that, in relation to the Caicó Complex, the Arabia Complex in its southern and central portions is enriched in K, while eU and eTh are predominant in the northern portion (Figure 5 B, C, D).
In Area II, around the Saquinho iron deposit (Figures  2, 6), there are bodies of metamorphic clinopyroxenehornblendite. These bodies are elongated, sometimes intercalated with felsic gneisses (Figure 7) and spatially associated with lenses and amphibolite bodies from the Caicó Complex ( Figure 6).
Micropetrographic data show a rock with a granonematoblastic, fine to coarse inequigranular, serial and non-oriented texture. The rock was originally pyroxenite, submitted to retrometamorphism, i.e., pyroxenes were hydrated and replaced by hornblende. Retrometamorphism also caused the instability of the hornblende, which, in turn, was replaced with actinolite, and it was classified as actinolytic hornblende.

U-Pb in zircon and Sm-Nd in whole rock
A sample (AP-242B) of biotite-amphibole orthogneiss (Figure 9) from Area I was analyzed (SHRIMP). Zircon crystals are small to large (30 to 300 µm), colorless, translucent to transparent bipyramidal prisms. In a cathodoluminescence image, one can see a complex internal structure (Figure 10a), with the presence of nuclei and edges. Oscillatory zoning is the most frequent feature, although edges, recesses and chaotic texture can be identified. U-Pb isotopic composition was determined in fifteen points in eleven crystals (Table 2). Eleven out of fifteen points were located in crystals without borders or in nuclei (main group), while the other four points were located in more homogeneous areas, suspected of being regrowth borders. The main group had Th/U ratios between 0.52 and 0.83 and a coherent set of 207 Pb/ 235 Huh 206 Pb/ 238 U ratios. The most concordant data (9 points) allowed the calculation of the concordant age of 2456 ± 4 Ma, with an MSWD value of 2.0 and a probability of concordance of 0.15 (Figure 10b). The isotopic data obtained at the edges were very scattered and age could not be calculated. The Th/U ratios of these points were in the order of 0.30, a little smaller than those found in the nuclei. It can be interpreted that the age of 2456 ± 4 Ma represents the crystallization age of the gneiss igneous protolith and that the superposed metamorphic event disturbed the system, but was not able to generate the isotopic rebalance of the mineral.
The same orthogneiss sample was analyzed using the Sm-Nd technique. The contents of Sm and Nd were relatively low (6.063 and 32.574 ppm, respectively) and showed a 147 Sm/ 144 Nd of ratio 0.1125, within the expected range for felsic rocks. This result, in association with the 143 Nd/ 144 Nd ratio of 0.511339+/-0.000017, indicates the age model TDM of 2.56 Ga and εNd value(t=2.46 Ga) of +1.20, which allows inferring a juvenile source for the magma generating the igneous protolith or gneiss.
For Area II, two samples were dated for U-Pb via LA-ICP-MS. Sample AP-01C is an amphibolite from the Caicó complex, dark green in color, with no apparent structure ( Figure 11). This rock has large zircon crystals (150-350 µm),     which are prisms, sometimes bipyramidal, and fragments. Most crystals are clear, but some have metamitic surfaces while others are heavily fractured. The BSE image shows crystals with a discrete oscillatory zoning, homogeneous crystals, crystals with a spongy texture and mixed crystals (Figure 12a). Owing to morphological diversity, 30 points in 27 crystals were analyzed (Table 3). The data are relatively dispersed along the concord (Figure 12b). A crystal showed 207 Pb/ 206 Pb apparent age of 2453±20 Ma, much higher and different from the others and therefore it is considered the result of crustal contamination. The main group of data, despite the dispersion, has similar behavior. Thus, their most concordant data were used to calculate the age of 2129 ± 52 Ma ( Figure  12b), with MSWD of 0.31 and probability of agreement of 0.58. This result is interpreted as the crystallization age of the amphibolite protolith. The alignment of the other points in this group indicates loss of lead in the Neoproterozoic, but age could not be specified.
The second sample from Area II (AP-02) consists of a metamorphic clinopyroxene-hornblendite ( Figure 13). The zircon crystals recovered from this sample are large prisms (150-250 µm) with little expressive bipyramid. They are clear crystals with few fractures and inclusions. The predominant internal structure in a BSE image (Figure 14a) is oscillatory zoning, although sectored zoning and homogeneous edges could be seen some of them Despite the morphological diversity, the isotopic data obtained at 24 points (from 20 crystals) are similar (Table 3). The concord diagram ( Figure  14b) shows agglomeration of data at about 2.4 Ga and some scattered points with reverse and normal disagreement. The selection of the most concordant points allowed the calculation of the age of 2381 ± 16 Ma, with MSWD of 1.09 and probability of concordance of 0.30 (Figure 14b). This result is interpreted as the crystallization age of hornblendite.

Discussion and conclusions
The integration of geological, geophysical and isotopic data carried out in this study will allow a better understanding of the geological record during the Siderian period in two distinct areas of the Piranhas-Seridó Domain. In Area I (north of the domain), using geological cartography with the support of aerogeophysical data, an expressive portion of a fragment of juvenile continental crust of Siderian age could be individualized. It is found to be amalgamated between the Rhyacian rocks from the Caicó Complex. In Area II (center of the domain), the first evidence of the existence of older rocks for the crystalline basement came through U-Pb geochronological data carried out on subsurface rock samples (stratigraphic hole FD-SE-002 carried out by SGB -CPRM in 2014), which identified banded gneisses and Archean metamafic rocks (2512 Ma and 2501 Ma -Cavalcante et al. 2019). In an attempt to identify the outcropping presence of these rocks on the surface, this study identified a siderian rock (metamorphic clinopyroxene-hornblendite) with an age of 2381 ± 16 Ma, which is here attributed to the Arabian Complex. This hornblentite is spatially associated with amphibolite bodies whose protolith has a crystallization age of 2129 ± 52 Ma; thus, it can be correlated with the amphibolites from the Caicó Complex.
The ages determined in this study corroborate data reported by other authors , which indicate the presence of geological units of Siderian ages for the FIGURE 12. a) Electron backscattered image of zircon crystals from the AP-01C sample, with the identification of the crystal number and the 207Pb/206bp apparent ages in italics The colors of the circles correspond to those used in the concord diagram. b) Concordia diagram for the LA-ICP-MS data of the AP-01C sample, belonging to the Caicó Complex. Dashed lines indicate points that were not used for age calculation. The green dots represent data interpreted as crustal contamination. The data with a high degree of disagreement are shown in red.   crystalline basement of the PSD, which are about 250 Ma older than those that compose the Caicó Complex, which, according to data from the regional literature, show an age ranging between 2.2 and 2.1 Ga, thus corroborating the dating of the amphibolitic body of Area II. Our U-Pb geochronological results, together with the Sm and Nd isotopic data, clearly indicate the existence of a Siderian juvenile event in the PSD, which needs to be looked into in more detail in further studies. The recognition of this juvenile Siderian continental crust for this portion of the Borborema Province corroborates data from the world literature that increasingly finds evidence of the action of geotectonic activity in this period in different cratons.
The results found in this reinforce the importance of carrying out geological mapping work combined with aerogeophysical data supported by geochronological (U-Pb) and isotopic (Sm-Nd, Lu-Hf) methods, even in areas with considerable prior geological knowledge. Similar studies should be carried out in other areas of the PSD to improve the geotectonic knowledge of the units that compose the crystalline basement in this domain.  FIGURE 14. a) Electron backscattered image of zircon crystals from the AP-02 sample, with the identification of the crystal number and the 207Pb/206bp apparent ages in italics The colors of the circles correspond to those used in the concord diagram. b) Concordia diagram for the LA-ICP-MS data from the AP-02 sample. Dashed lines indicate points that were not used for age calculation.