Petrochronology of the Dom Feliciano Belt foreland in southernmost Brazil reveals two distinct tectonometamorphic events in the western central Kaoko–Dom Feliciano–Gariep orogen

Abstract

The Dom Feliciano Belt is the South American part of an extensive Neoproterozoic orogenic system that developed during the late Cryogenian–early Cambrian close to the margin of southwest Gondwana. The link of its evolution with the tectonic processes in its African counterpart is still not well understood. P–T estimates, Lu–Hf garnet–whole-rock ages, U–Pb monazite SIMS ages and REE garnet and monazite data from samples of the Porongos and Passo Feio complexes indicate diachronous tectonic evolution of the central Dom Feliciano Belt foreland. Metasedimentary rocks of the eastern Porongos Complex reached previously estimated metamorphic peak conditions of ~ 560–580 °C and 5.8–6.3 kbar at 654 ± 2 Ma, based on Lu–Hf isochron garnet–whole-rock age data. This episode represents an early orogenic thickening in the foreland as a response to the beginning of the transpressive convergent evolution of the belt. The monazite age of 614 ± 6 Ma (U–Pb SIMS) is interpreted as associated with post-exhumation magmatic activity in the foreland and suggests that the eastern Porongos Complex was exhumed sometime between ca. 660 and 615 Ma. The main metamorphic and deformation event in the Porongos Complex’s western region occurred at ~ 545–565 °C and 4.3–5.3 kbar at 563 ± 1 Ma (garnet–whole-rock Lu–Hf isochron age). The exhumation of this part of the foreland is dated using monazite crystallising during garnet breakdown and suggests retrograde metamorphism at 541 ± 7 Ma (U–Pb SIMS). The main metamorphic fabric in the Passo Feio Complex further to the west developed at 571 ± 2 Ma (garnet–whole-rock Lu–Hf isochron age) at 560–580 °C and 4.7–6.4 kbar. The western part of the Porongos Complex and the Passo Feio Complex have deformed at similar P–T conditions and apparent geothermal gradients at ca. 570–565 Ma. These regions record a second crustal thickening event in the Dom Feliciano Belt foreland and the orogenic front migration towards the west as a response to the onset of crustal thickening on the African side of this long-lived transpressive orogenic system.

Graphical abstract

Appendix

Lu–Hf isotope analysis

The Lu–Hf analyses were conducted at the Geological Institute of the Czech Academy of Sciences (chemistry) and Faculty of Science, Charles University (MC–ICP–MS). Garnet concentrates and whole-rock powders were weighted and mixed with the ¹⁷⁶Lu–¹⁸⁰Hf tracer solution. The samples were digested in closed 30 ml Savillex Teflon vials using combined acid attack (HF–HNO₃–HCl). First, 0.5 ml HNO₃ + 2 ml HF (concentrated acids) was added to each sample and left to stand cold in a closed vial for 3 h. Subsequently, the bombs were opened and warmed on the hotplate to 90 °C to evaporate Si and all acids. During this step, the major minerals are attacked by the acids, most of the Si evaporates with the excess HF and the acids used in the following step thus can attack any resistant minerals without being depleted by reaction with major silicate phases. After complete evaporation, the mixture of 1.5 ml HNO₃ + 4.5 ml HF was added to the samples, left on a hotplate for 2 days at 160 °C and then evaporated to dryness. After that, the samples were treated three times with 2 ml of concentrated HNO₃ and evaporated to dryness. Next, 1 ml of 6 M HCl was added and immediately dried down. Finally, 8 ml of 6 M HCl was added and left on a hotplate at 160 °C in a sealed beaker for 24 h. The sample was then evaporated to dryness, and 2 ml of 1 M HCl was added for subsequent column chemistry.

The ion exchange column chemistry follows closely that of Anczkiewicz et al. (2004), which is a down-scale modification of the original setup of Patchett and Tatsumoto (1980). The separation of Hf (+ Ti) and Lu (+ Yb and LREE) fractions is first carried out on a standard cation exchange column using AG50W–X8 resin (200–400 mesh size) and 1 M HCl–0.06 M HF (HFSE elution) and 2.5 M HCl (REE elution). The final purification of Hf from other HFSE and potentially interfering Lu and Yb takes place on a second column with Eichrom LN resin (50–100 µm) using a technique based on Lee (1999) employing the mixture of 2 M HCl–0.1 M. The same column is then used to purify Lu from other REEs and reduce Yb in the Lu cut using 4 M HCl.

All measurements of Lu and Hf fractions were carried out using a THERMO Neptune multi-collector (MC) ICP–MS in the labs of the Faculty of Science, Charles University in Prague. Hafnium isotopic compositions were analysed in a static mode using Faraday cups with the following configuration: L4––¹⁷²Yb, L3––¹⁷⁴Yb, L2––¹⁷⁵Lu, L1––¹⁷⁶Hf, C––¹⁷⁷Hf, H1––¹⁷⁸Hf, H2––¹⁷⁹Hf, H3––¹⁸⁰Hf, H4––¹⁸²W. Samples were aspirated to the instrument in 0.5 M HNO₃–0.25 M HF mixture using CETAC Aridus II desolvating nebuliser. The data acquisition procedure consisted of 40 integration cycles acquired over a period of ~ 6 min, followed by ~ 5 min of washout with a mixture of 1.2 M HNO₃–0.5 M HF composition. The raw data were processed offline using an on-purpose-built calculation EXCEL spreadsheet. Repeated measurements of 50 ppb JMC–475 standards throughout analyses yield ¹⁷⁶Hf/¹⁷⁷Hf = 0.282158 ± 7 (2SE, n = 7), which is in agreement within error with the reference data (Chu et al. 2002). The spike strip** routine employing the ratio of ¹⁷⁹Hf/¹⁷⁷Hf iteratively deconvoluting to the natural value of 0.7325 (Patchett and Tatsumoto 1980) and exponential mass-bias correction were used to obtain Hf isotopic composition and Hf concentration of the spiked samples.

For Lu isotopic analyses, Faraday cup configuration was as follows: L3––¹⁷¹Yb, L2––¹⁷²Yb, L1––¹⁷³Yb, C––¹⁷⁴Yb, H1––¹⁷⁵Lu, H2––¹⁷⁶Lu, H3––¹⁷⁷Hf. The sample aspiration was identical to the Hf measurements, with the difference that HF-free acids were used for sample introduction (0.5 M HNO₃) and washout (1.5 M HNO₃). The data acquisition procedure consisted of 40 integration cycles acquired over a period of ~ 3 min, followed by 4 min of washout. The raw data were processed offline using an on-purpose-built calculation EXCEL spreadsheet. Repeated measurements of natural Lu and Yb standard solutions were carried out to check the accuracy of the isotopic ratio measurements. The mass-bias correction (exponential law) of the spiked ¹⁷⁶Lu/¹⁷⁵Lu ratio was done using the natural Yb present in the sample (reduced in the 3rd step of column chromatography to be ~ 1/10 of the amount of Lu to be suitable for mass bias correction while not causing excessive interference on ¹⁷⁶Lu mass) and the true ratio of ¹⁷⁴Yb/¹⁷²Yb = 1.45198 (Thirlwall and Anczkiewicz 2004). The value of ¹⁷⁶Lu/¹⁷⁵Lu was then used to calculate the concentration of Lu in the samples. The accuracy of the method was checked by measurement of spiked aliquot of BCR–2 reference material, which gave ¹⁷⁶Hf/¹⁷⁷Hf = 0.282859 ± 11, Hf = 5.00 ppm and Lu = 0.514 ppm, which are in good agreement with published values of 0.282866 ± 11 (Jweda et al. 2016), 4.8 ± 0.2 and 0.51 ± 0.02 (U.S. Geological Survey Certificate of Analysis, online source), respectively.

U–Pb monazite dating methodology

The samples were processed through a rock crusher and a hammer mill. Monazite and garnet were separated using a Wilfley table, a Frantz™ isodynamic magnetic separator and heavy liquids in the laboratories of the Department of Geosciences at the UiT, The Arctic University of Norway in Tromsø. Subsequently, monazite grains were handpicked under a binocular microscope, and selected grains were mounted in one-inch epoxy discs. To identify internal microstructures and possible compositional zoning, backscattered electron (BSE) images of monazite grains were made by Zeiss Merlin Scanning Electron Microscope (SEM) housed at the Faculty of Health Sciences of the same university.

Monazite U–Th–Pb analyses by SIMS were performed on a Cameca IMS 1280 ion probe at NordSIMS laboratory hosted by the Swedish Museum of Natural History in Stockholm. Operating parameters concerning primary beam and mass resolution were similar to those used for zircon, broadly following the protocols described (Gasser et al. 2015). The principal difference from zircon is that the monazite analyses employed both a smaller entrance slit (30 µm instead of 75 µm) to limit the secondary beam intensity and a smaller energy slit (30 eV instead of 45 eV) together with a − 30 eV energy offset (applied via sample high voltage) on all the Pb, ThO_x and UO_x (where x = 0, 1 or 2) peaks of interest to minimise matrix differences in potentially chemically diverse monazite and eliminate a small ThNdO₂²⁺ interference on ²⁰⁴Pb identified in earlier monazite studies (e.g. Kirkland et al. 2009). Secondary beam centring and optimisation steps were performed as for zircon but using the CePO₂⁺ matrix peak at nominal mass 203. U–Pb ratios were calibrated against a 425 Ma reference monazite from a metapelite of the Wilmington Complex, Delaware (sample 44,069, Aleinikoff et al. 2006), using a two-dimensional power law calibration approach, i.e. (Pb/U)_true = f (Pb/U)_meas, UO2/U_meas) based on measurement of these ratios in the reference monazite.