Congra-2 is a monzogranite from the southern Sierra Madre Occidental. For its cognate sample, collected from the same plutonic body, an emplacement age of 59.4 ± 0.5 (2σ) Ma was proposed on the basis of zircon U-Pb results (Sample Conc-01 in Duque-Trujillo et al., 2014). Hornblende and biotite concentrates from Conc-01 yielded Ar-Ar ages at 55.7 ± 0.5 (2σ) and 56.3 ± 0.3 (2σ) Ma, respectively, both cooling ages that were also obtained by Duque-Trujillo et al. (2014).

This sample is a two mica-bearing deformed granite belonging to the Tres Sabanas pluton, which is located NW of Guatemala City, Guatemala. Torres de León (2016) reported a U-Pb zircon age of 115 ± 4 Ma (2σ) for OV-0421, interpreted as its crystallization age. A cooling age of 102.2 ± 1.2 (2σ) Ma, obtained by the K-Ar method, has also been reported by the same author.

MCH-40 is an andesite of rather hypabyssal origin. This rock was sampled to the east of Jacinto Tirado town, Chiapas (Table 1) from an Early to Middle (?) Jurassic volcanic unit named informally as Andesita Pueblo Viejo (e.g., Castro-Mora et al., 1975; Meneses-Rocha, 2001; Godínez-Urban et al., 2011). Due to its age uncertainty, MCH-40 was also dated by zircon U-Pb technique (see details below).

This rock sample is a metamorphosed granite, collected from the Chiapas Massif near the Cintalapa de Figueroa town. According to several authors (e.g., Damon et al., 1981; Torres et al., 1999; Meneses- Rocha, 2001; Schaaf et al., 2002; Weber et al., 2005), the Chiapas Massif is a Permian-Triassic continental arc and represents the basement of the southern Maya block. This unit is mainly composed of diverse plutonic rocks, most of which are deformed while some are metamorphosed. To better constrain its crystallization age, zircons from MCH-17 were also dated in the current work by employing LA–ICP-MS-based U-Pb method. This sample is the only one from which apatite were analyzed by both LA–ICP-MS and ID–TIMS techniques. This was done mainly to evaluate the U-Pb results obtained by two different analytical methods and to see eventual discrepancies (details in “Analytical Methods” section).

Table 1 Table 1 Lithology, locality, and previously published geochronological data of selected samples.

This sample belongs to the undeformed Rabinal Granite, cropping out north from the Rabinal town in Baja Verapaz, Guatemala. Many zircon aliquots from GT-0340 were previously dated by U-Pb ID–TIMS, yielding an age of 482.6 ± 7 (2σ) Ma (Concordia upper intercept) that was interpreted as its approximate crystallization age (Ortega-Obregón et al., 2008). This age is similar to others obtained in different samples from the same granite (e.g., Solari et al., 2013).

This sample is a granulitic gneiss from the Grenville-aged Oaxacan Complex. The sample is composed by quartz, mesoperthite, rutile, garnet, and by altered mafic minerals. OC-111 was sampled ca. 120 km north of Oaxaca City, to the east of the tectonic contact between the Zapoteco and Mixteco tectonostratigraphic terrains (e.g., see ElíasHerrera et al., 2007). Previous U-Pb zircon dating yielded protolith crystallization ages at ~1220 Ma, although it was also demonstrated that this sample underwent “dry” granulite facies metamorphism at 990 ± 10 Ma (Solari et al., 2014).

Apatite U-Pb analyses were performed by a joint effort between the Laboratorio Universitario de Geoquímica Isotópica (LUGIS, Instituto de Geofísica, CDMX, UNAM) and the Laboratorio de Estudios Isotópicos (LEI, Centro de Geociencias, Campus Juriquilla, UNAM). Apatite U-Pb dating at LUGIS was based on ID–TIMS technique, while in LEI all the six selected samples were analyzed with LA–ICP-MS.

Sample preparation was performed at the Centro de Geociencias (Taller de Molienda and Taller de Laminación), Campus Juriquilla, UNAM. Heavy minerals were concentrated from two narrow grain-size fractions of 63–125 and 125–180 µm, using conventional methods/ instruments like crushing, sieving, Wilfley^TM table, Frantz^TMseparator, and finally by hand-picking under a binocular microscope. For U-Pb dating by ID–TIMS, only ‘sterile’ apatite grains, without visible inclusions and other defects like cracks, were selected to be split into two aliquots with ~300 crystals of different size and morphology (e.g., see Figure 1). Analysis by ID–TIMS was performed only for sample MCH-17. For LA–ICP-MS-based apatite U-Pb dating, ~200 apatite crystals extracted from each sample were mounted with EpoFix^TM Struers in a 2.5 cm diameter plastic ring, and then polished with Buehler SiC MicroCut^TM sandpapers P1500 (~12.5 μm) and P2500 (~8.3 μm), and alumina suspensions over Buehler MicroCloth^TM pads (see an example illustrated on Figure 2). LA–ICP-MS analyses were performed for all the six selected samples. In addition, two samples (MCH-17 and MCH-40) were also prepared for zircon U-Pb dating, from which ca. 100 crystals from each sample were mounted to be analyzed by LA–ICP-MS.

An additional observation of the selected apatite crystals using a binocular microscope, cleaning, chemical treatments, and isotopic measurements were entirely carried out at LUGIS. One of the two aliquots with ~300 apatite crystals was leached in 0.25M HNO. at ambient temperature (24 °C) for 1 min to remove any non-radiogenic Pb from their exterior surface, and then rinsed with de-ionized water. The other one, also made up of ~300 crystals, was cleaned only in Milli-Q water. The aliquots were dried down, transferred into Teflon^TM microcapsules and weighted, then 1 ml of ultrapure 8M HNO. was added. The complete digestion of both aliquots (i.e., leached and non-leached) was performed at 110 °C for 20 hours; then, the solutions were dried down. After evaporation, a drop of 6M HCl was added and dried to convert the produced components to chlorides. The dry residues derived from both aliquots were dissolved in 600 μl of 3M HCl, and heated at 110 °C for 7 hours. Each digested aliquot was split into two fractions; one for isotopic composition (IC) analyses, and a second for isotope dilution (ID) measurements to which a mixed ²⁰⁸Pb–²³⁸U spike was added. U and Pb were separated by chromatographic column chemistry technique employing BioRad^TM AG-1X8 resins within TeflonTM columns.

Figure 1 Figure 1 Photomicrograph of selected apatite crystals for ID–TIMS analysis from sample MCH-17 using (a) transmitted and (b) reflected light. Note that most of crystals are free of inclusions or fractures.

The prepared samples were loaded on Re filaments with the addition of a silica gel–H.PO. mixture and analyzed in static mode. Pb isotopic measurements were performed in a ThermoTM TRITON Plus mass spectrometer; measuring ²⁰⁴Pb with an ion counter, whereas the remaining Pb isotopes (²⁰⁶Pb, ²⁰⁷Pb, and ²⁰⁸Pb) were measured in Faraday cups. U concentrations were determined with a Finnigan^TM MAT-262 mass spectrometer. Finally, 30 and 45 isotope ratios were acquired for U and Pb, respectively. Mass fractionation was taken into account based on PbDat v. 1.24 (Ludwig, 1993) and applying a correction factor of 0.12% for each isotopic pair (²⁰⁶Pb/²⁰⁴Pb, ²⁰⁶Pb/²⁰⁷Pb, and ²⁰⁶Pb/²⁰⁸Pb). The values for initial common Pb were estimated at 250 Ma (Permian–Triassic Chiapas Massif; Damon et al., 1981; Schaaf et al., 2002; Weber et al., 2005), following the well-accepted methodology of Stacey and Kramers (1975). During the experiment at LUGIS, total analytical blanks for Pb and U were obtained as 18 and 16 pg, respectively. However, in further experiments these values may change, and thus, will have to be carefully controlled again.

Spot analyses were performed with a Resonetics RESOlution^TM LPX Pro (193 nm, ArF excimer) laser ablation system, coupled to a Thermo Scientific iCAPTM Qc quadrupole ICP-MS at LEI. Different analytical protocols were employed for zircon and apatite. For zircon, each spot analysis consisted in the acquisition of 15 s of background signal (gas blank), 30 s of ablation, and 15 s to allow the signal to reach the baseline again (wash-out). The spot diameter was of 23 µm, using a fluence of 6 J/cm² with a repetition rate of 5 Hz. For apatite, 35 s of ablation was chosen with the same gas blank and wash-out acquisition times as given above. In this case, a spot of 60 µm was employed, basically to obtain a larger volume of the ablated material to be analyzed with the purpose of compensate for weaker signals of U and Pb. Due to different coupling of laser energy in different materials, the laser pulse energy and repetition rate were reduced to 4 J/cm² and 4 Hz, respectively. For both zircon and apatite, 350 ml/min of He was used as a carrier gas, mixed in the cell with ca. 800 ml/min of Ar. Four and three ml/min of N₂ for zircon and apatite analyses were also added after the cell, respectively, to increase the sensitivity of the plasma. A “squid” signal homogenizer is located right after the ablation cell, to mix properly all the gases added before reaching the plasma. Single-spot analyses in apatite were performed in inclusion-free grains or in those parts in which no visible inclusion or other defects were present (e.g., see in Figure 2).

Figure 2 Figure 2 Photomicrograph of an apatite crystal after ablation in (a) transmitted and (b) reflected light. Spot diameter (center of the crystal) is 60 µm. The selected crystals for LA–ICP-MS analysis are free of inclusions or cracks. Note the concentric deposits of ablated material around the crater in reflected light (b).

Together with those isotopes needed for the U-Pb age calculation (²⁰⁶Pb, ²⁰⁷Pb, ²⁰⁸Pb, ²³²Th, and ²³⁸U), LA–ICP-MS permits to sequentially detect additional elements such as REE (rare earth elements, La to Lu), Sr, Y, Mn, Mg, Cl, etc. These detailed chemical compositions are not shown because they are beyond the scope of this paper. This protocol for apatite analyses was established on the basis of numerous experiments carried out at the LEI Lab during the past years, and may be used for simultaneous U-Pb and fission-track in-situ double dating plus multielemental analysis (e.g. Abdullin et al., 2016a and 2016b).

Tuning of the ICP-MS followed the parameters reported in Solari et al. (2010) and Ortega-Obregón et al. (2014). Corrected isotope ratios, ages and errors were calculated using Iolite software (Paton et al., 2011) and the VizualAge data reduction scheme of Petrus and Kamber (2012). The standard zircon “91500” (Wiedenbeck et al., 1995, 2004) was used as a reference mineral for zircon U-Pb analyses. In the course of this experiment, an age of 337.7 ± 1.5 (2σ) Ma was obtained for the Plešovice zircon used as a secondary standard for zircon U-Pb analyses, yielding a mean square weighted deviation (MSWD) of 1.2 from 20 single-spot U-Pb ages. This value is in line with the standard ²⁰⁶Pb/²³⁸U age of ca. 337 Ma (Sláma et al., 2008).

In the case of apatite U-Pb dating, UcomPbine (Chew et al., 2014) was used to model ²⁰⁷Pb/²⁰⁶Pb initial values and thus force a ²⁰⁷Pb correction that considers the common Pb incorporated by apatite standards at the moment of their crystallization. As is described by Chew et al. (2014), this correction is applied immediately after baseline subtraction and previously to the “raw” standard ratios calculation in order to not compromise down-hole fractionation and mass bias corrections. Once the standard is corrected and down-hole fractionation modeled, the drift in the mass discrimination during a session is subsequently accounted by sample-standard bracketing. Tera and Wasserburg (T–W) Concordia diagrams are commonly used in apatite U-Pb dating, because results are mostly discordant. The lower intercept in T–W, is considered as a “mean” apatite U-Pb age that should have geological significance (e.g., crystallization or cooling age, or the age of a metamorphic event).

For apatite U-Pb dating, the Madagascar apatite was used as a main reference material (with a mean U-Pb age of 485–475 Ma; Thomson et al., 2012; Chew et al., 2014), assuming that matching of the matrix is fundamental during the U-Pb dating by LA–ICP-MS (e.g., Gehrels et al., 2008; Solari et al., 2010, 2015).

Due to limitations in the LEI Lab, no representative and wellstudied secondary U-Pb standard was used for apatite U-Pb dating. However, as was mentioned above, the sample MCH-17 was chosen to be dated by both ID–TIMS and LA–ICP-MS techniques. This sample, thus, was also used to control the final U-Pb ages obtained for all the apatite samples selected (see results below). Age and error calculations, including zircon mean ages and intercept values and uncertainties, were carried out using Isoplot v. 3.7, written by Ludwig (2008).

Zircons from two of the six experimental samples (MCH-17 and MCH-40) were dated by U-Pb employing LA–ICP-MS (LEI; U-Pb results are detailed in Figure 3 and Supplementary Appendix A1). Only results within -5 to 30% of concordance were considered as reliable ages. Weighted average ages were calculated using Isoplot v. 4.15 (Ludwig, 2008) rejecting outliers (less than 95% of confidence).

Sample MCH-17

MCH-17 mainly contains prismatic and bipyramidal zircon crystals, most of which are larger than 100–150 μm length and 50–100 μm width. A detailed observation of their morphology suggests that these crystals are of igneous origin, without evidence from cathodoluminescence of possible metamorphic rims. From a total of 35 zircons dated, 27 passed the concordance filters (those within -5 to 30% of discordance) and were considered meaningful for the interpretation of the igneous crystallization age. The remaining analyses are concordant to slightly discordant (see Supplementary Appendix A1), and define a marked cluster between ~260 and ~250 Ma in the Concordia diagram (Figure 3a). The mean²⁰⁶Pb/²³⁸U age of 253 ± 2.5 (2σ) Ma, displayed in Figure 3b, can be interpreted as the best approximation for the crystallization age of this granitic sample.

Sample MCH-40

Most zircon grains from the andesite sample MCH-40 are small (50–100 μm long, 30–60 μm wide) with poorly developed prism facets and pyramids. 35 zircon crystals were analyzed, 27 of which are considered reliable for age interpretation. The U-Pb ages are concordant and straddle the Concordia curve at ~195 Ma (lower Sinemurian age; Figure 3d). The mean ²⁰⁶Pb/²³⁸U age of 196.1 ± 1.5 (2σ) Ma is, hence, interpreted as the best crystallization age for MCH-40 (Figure 3d).

Apatite U-Pb results obtained from all the six studied rock samples are described below. Detailed information on our LA–ICP-MS U-Pb experiments is given in Supplementary Appendix A2.

Sample Congra-2

Apatite from the youngest sample Congra-2 show a well-defined linear regression in the T–W diagram, which is derived from all the 34 single-spot analyses performed with LA–ICP-MS. A lower intercept age of 60.9 ± 1.5 (2σ) Ma was detected (MSWD = 0.95; Figure 4a). Despite its young age (Paleocene; Duque-Trujillo et al., 2014), a statistically very reliable apatite U-Pb age was obtained for Congra-2, which is also supported by a MSWD that is close to 1. This is due to high U contents that vary broadly from one apatite to another (31.1 to 371.5 ppm with a mean of 108.2 ppm, standard deviation (SD) = 69.5; Supplementary Appendix A2), which was crucial to yield strongly heterogeneous isotopic ratios for the U-Pb system as well as to generate high levels of radiogenic Pb, and thus displaying a perfect order on the Discordia line (Figure 4a) with a very low age error (only 2.5% at 2σ level).

This pluton yielded a crystallization age of 59.4 ± 0.5 (2σ) Ma as it was well documented by zircon U-Pb geochronology, and then cooled to ~56 Ma as suggested by Ar-Ar ages in hornblende and biotite (Sample Conc-01; Duque-Trujillo et al., 2014). In this case, the apatite lower intercept U-Pb age of 60.9 ± 1.5 (2σ) Ma is within the analytical uncertainty of the zircon age, therefore, this apatite U-Pb date also seems to indicate the igneous crystallization age for Congra-2 (Conc01), suggesting that the pluton was likely emplaced at a temperature above the Ar-Ar closure temperature of hornblende and biotite, and lower than the closure temperature of the apatite U-Pb system.

Sample OV-0421

For the Tres Sabanas granite OV-0421, the lower intercept in T–W yielded an apatite U-Pb date of 106.7 ± 8.7 (2σ) Ma, obtained from 42 spots (Figure 4b). These apatites have U contents of 84–177 ppm with a mean of 127 ppm and SD of 21.8 (Supplementary Appendix A2). A high MSWD of 9.89 incorporates a large uncertainty, derived from quite scattered apatite U-Pb results as well as from small age errors (Supplementary Appendix A2; Figure 4b).

Figure 3 Figure 3 U-Pb Concordia diagrams (Wetherill, 1956) and weighted average 206Pb/238U age plots of zircons from samples MCH-17 (a and b, respectively), and MCH-40 (c and d, respectively). Blue bars correspond to rejected data.

The age determined in this work by apatite U-Pb dating, lies between the zircon U-Pb age of 115 ± 4 (2σ) and the biotite K-Ar age of 102.2 ± 1.2 (2σ) Ma, which were previously obtained for the same granite sample by Torres de León (2016). Central values decrease systematically depending on the closure temperatures for each system; from U-Pb in zircon (900 to 800 °C; Lee et al., 1997; Cherniak and Watson, 2001) through U-Pb in apatite (ca. 450–550 °C; Schoene and Bowring, 2007; Chew et al., 2014; Cochrane et al., 2014) to K-Ar in biotite (300 to 350 °C; e.g., Faure and Mensing, 2005). On the basis of these available data, a cooling rate of ~41 °C/Ma can be calculated for the magmatic system from which OV-0421 was formed. This cooling rate could be significant for the whole Tres Sabanas pluton, located in Guatemala.

Sample MCH-40

Apatite crystals extracted from MCH-40 (Sinemurian andesite) fall towards the upper intercept in the T–W Concordia (Figure 4c). A total of 43 dated apatite crystals define a lower intercept of 198 ± 18 (2σ) Ma, yielding an MSWD value of 1.15. The tendency towards the upper intercept in T–W and a large age error of 9.1% at 2SE are apparently linked to very low U concentrations in those apatites, which range between 4.3 and 16.6 ppm with a mean value of 7.2 ppm and SD of 2.4 (see details in Supplementary Appendix A2).

Zircons and apatite from MCH-40 yield similar U-Pb ages of 196.1 ± 1.5 Ma and 198 ± 18 (2σ) Ma, respectively (Figures 3d and 4c). In theory, volcanic or subvolcanic samples like this andesite (or shallow plutons) should yield equivalent crystallization ages independently of the isotopic system applied. Considering that this magma cooled (crystallized) very rapidly, both those U-Pb ages can be interpreted as corresponding to the igneous crystallization age during the Sinemurian Stage. There is a large difference in the age error between the zircon U-Pb and the apatite U-Pb results, which is basically controlled by the strong differences in the abundance of their parent and daughter nuclides (i.e., zircon contains almost no common Pb and is significantly more enriched in U and radiogenic Pb when compared to apatite). It is important to note that Sinemurian igneous ages from Jurassic magmatic rocks and volcaniclastic units (Andesita Pueblo Viejo and La Silla Formation), distributed in the Chiapas Massif area, have also been detected in recent works based on U-Pb, Ar-Ar, and fission-track dating (Godínez-Urban et al., 2011; Abdullin et al., 2018).

Sample MCH-17

Non-leached apatite of sample MCH-17, dated by ID–TIMS, yielded a reversely discordant ²⁰⁶Pb/²³⁸U age of 253.5 ± 2.7 (1σ) Ma (-8.9% of discordance; Table 2; Figure 5). The negative discordance can indicate that these crystals include non-radiogenic Pb in their outermost zones. The leached aliquot, however, yielded a concordant ²⁰⁶Pb/²³⁸U age of 247.1 ± 0.8 Ma with a very small 1σ-error (0.3%; Figure 5; Table 2). The results of this single exercise demonstrate that for some samples, such as MCH-17, a rapid acid leaching may be required to remove possible superficial Pb contamination of non-radiogenic origin (e.g., in apatite from weathered rocks).

Using single-spot analyses performed on 39 apatites by LA– ICP-MS, a lower intercept U-Pb age of 243 ± 12 (2σ; 4.9%) Ma was obtained for the same sample (MSWD = 1.7; Figure 4d), a value that agrees very well with the ID–TIMS-derived concordant U-Pb age of 247.1 ± 0.8 Ma (1σ).In these apatites U vary from 11.5 to 34.7 ppm, giving a mean concentration of 18.7 ppm with a SD of 6.1 (Supplementary Appendix A2).

Figure 4 Figure 4 Tera–Wasserburg Concordia diagrams for U-Pb data of apatites from samples: (a) Congra-2; (b) OV-0421; (c) MCH-40; (d) MCH-17; (e) GT-0340; and, (f) OC-111.

Both apatite U-Pb dates of 247.1 ± 0.8 (1σ) Ma and 243 ± 12 (2σ) Ma obtained for sample MCH-17 by two different analytical techniques are younger than (or, within the analytical uncertainty of) the zircon crystallization age of 253 ± 2.5 (2σ) Ma (Figures 3b, 4d, and 5). In general, our U-Pb results are in close agreement with geochronological data reported for the Chiapas Massif in previous studies (Damon et al., 1981; Schaaf et al., 2002; Weber et al., 2005). For example, K-Ar biotite ages compiled by Torres et al. (1999) mostly lie within the Early Triassic. Triassic cooling ages from the Chiapas Massif were also detected by Rb-Sr in mica–whole rock pairs that range from 244 ± 12 to 214 ± 11 (2σ) Ma (Schaaf et al., 2002). Therefore, our apatite U-Pb ages suggest that the main post-emplacement thermal cooling of the Chiapas Massif started during the earliest Triassic.

Sample GT-0340

A total of 37 apatite crystals analyzed from the Rabinal Granite GT0340 yielded a lower intercept age of 473 ± 31 (2σ; ~6.6%) Ma (Figure 4e), and these grains display U contents between 3.1 and 124.3 ppm with a mean of 19.6 ppm and SD of 19 (see Supplementary Appendix A2). The MSWD of 1.18 suggests that all dated apatites pertain to a single age population. The apatite U-Pb age of 473 ± 31 (2σ) Ma is identical within error to the Rabinal Granite emplacement age of 483 ± 7 (2σ) Ma obtained for the from same rock sample (ID–TIMS-derived zircon U-Pb results, reported by Ortega-Obregón et al., 2008), indicating a shallow crystallization and fast cooling.

Table 2

* Denotes radiogenic lead; (a) Decay constants used ²³⁸U=1.55125 *10-10; ²³⁵U =9.48485*10^-10; ²³⁸U/²³⁵U =137.88; Rho: Error correlation.