Meso-Cenozoic deformation history of Thailand; insights from calcite U-Pb geochronology

U-Pb dating of calcite veins allows direct dating of brittle deformation events. Here, we apply this method to hydrothermal calcite veins in a fold-and-thrust belt and a large scale strike-slip fault zone in central and western Thailand, in an attempt to shed new light on the regional upper crustal deformation history. Calcite U-Pb dates for the Khao Khwang Fold and Thrust Belt (KKFTB) of 221 ± 7 Ma and 216 ± 3 Ma demonstrate that calcite precipitated during tectonic activity associated with stage II of the Indosinian Orogeny (Late Triassic – Early Jurassic). One additional sample from the KKFTB suggests that the Indosinian calcite has locally been overprinted by a Cenozoic ﬂuid event with a diﬀerent chemistry. For the Three Pagodas Fault Zone (TPFZ), our calcite U-Pb results suggest a complex, protracted history of Cenozoic brittle deformation. Petrographic information combined with contrasting redox-sensitive trace elemental signatures suggest that the vein arrays in the TPFZ precipitated during two distinct events of brittle deformation at 48 and 23 Ma. These dates are interpreted in the context of far-ﬁeld brittle deformation related to the India-Eurasia collision. The presented calcite U-Pb dates are in excellent agreement with published age constraints on the deformation history of Thailand, demonstrating the utility of the method to decipher complex brittle deformation histories. The paper further illustrates some of the complexities in relation to calcite U-Pb dating and provides suggestions for untangling complex datasets that could be applied to future studies on the deformation history of Thailand and other regions.


Introduction
Supplementary data files include an extended method, a series of Tera-Wasserburg Concordia plots showing normalized standards (supplementary figure 1), a table detailing the instrument parameters used during LAICPMS analysis (supplementary table 1). Additional files include an excel spreadsheet with all normalized U-Pb data used in this paper and corresponding trace element concentrations for elementally mapped samples (8b and 12a) Note: analyses have been culled from U-Pb data based on the following: association with cracks/contaminated material, low Pb counts and inconsistent time resolved Pb/Pb or U/Pb signals Delete all unused file types below. Copy/paste for multiples of each file type as needed.

Laboratory Processing
Calcite samples were selected and cut (in~1cm 3 blocks) to reveal internal sections that cross-cut the veins. Subsequently, the calcite pieces were mounted in 1-inch (2.5cm) round epoxy mounts using epoxy cure resin (5g epoxy resin and 1.15g epoxy hardener) and ground (using 800 and 2000 grit sandpaper) and polished (using a 3μm polishing cloth with diamond suspension fluid) to reveal a smooth surface.
Sample imaging was conducted at the British Geological Survey, Nottingham, UK. Cathodoluminescence imaging was conducted with a Technosyn 8200 MKII cold-cathode luminoscope stage attached to a Nikon optical microscope with a Nikon long working distance lens, and equipped with a Zeiss AxioCam MRc5 digital camera; vacuum and electron beam voltage and current were adjusted as required to generate optimum luminescence. Back-scattered electron and charge-contrast imaging were conducted using a FEI QUANTA 600 environmental scanning electron microscope (ESEM) with a working distance of 10 mm. BSE images were recorded using a solid-state (dual-diode) electron detector, with a 20 kV electron beam accelerating voltage, and beam currents between 0.1 and 0.6 nA,. Charge Contrast Images were recorded using a FEI large-field gaseous secondary electron (electron cascade) detector, with 20 kV electron beam accelerating voltage, and beam currents of 1.2 to 4.5 nA.
LA-ICP-MS U-Pb spot-analysis LA-ICP-MS analysis was conducted at the University of Adelaide using an ASI resolution LR Laser Ablation System coupled to an Agilent 7900 mass spectrometer in order to determine U and Pb concentrations. Large spot sizes (110 microns) were selected in order to maximise the signals from elements that were expected to have low concentrations. Only isotopes necessary for U-Pb dating ( 43 Ca, 202 Hg 204 Pb, 206 Pb, 207 Pb, 208 Pb, 232 Th and 238 U) were measured during spot analysis in order to maximise the dwell time on masses expected to have low abundance, such as the isotopes of Pb. Standard-sample bracketing was used, with the NIST614 glass reference material used for fractionation correction of the Pb-Pb ratios, and the WC-1 calcite reference material (Age: 254.4 ± 6.4) for correction of the U-Pb ratios [Li et al. , 2014;Roberts and Walker , 2016;Roberts et al. , 2017]. An in house calcite sample labelled 'Prague' of known stratigraphic age (˜424 Ma) was used as an accuracy check [Farkaš et al. , 2016]. In more detail, a correction factor was calculated based on the offset between the measured age and the known age of WC-1. This factor was then used to correct both the 'Prague' secondary standard and the unknowns.
LA-ICP-MS Elemental Mapping: LAIPCMS elemental mapping was conducted to identify alteration and different growth zones. Before elemental mapping, the surface of samples was gently re-ground (using 2000 grit sandpaper) to just below the laser ablation pits. Following this the surface was re-polished (using a 3μm polishing cloth as before). Maps were created at the University of Adelaide using an ASI resolution LR Laser Ablation System coupled to an Agilent 7900 mass spectrometer (i.e. the same as U-Pb analysis). A square laser beam of 91x91μm was used to create line rasters on selected areas of calcite samples. Data reduction was conducted using Iolite software [Paton et al. , 2011]. Elemental map data was produced using the Monocle plugin for Iolite [Petrus et al. , 2017]. In more detail, polygons, termed regions of interest (ROI) [Petrus et al. , 2017] surrounding to ablation spots were used to query elemental concentrations. Some spot analyses were removed based on anomalous chemistry, particularly high Al and U.
Repeat for any additional Supporting Text Figure S1. T-W Concordia plots showing secondary standards WC-1 and 'Prague' normalised to WC-1's correct age. Both WC-1 and 'Prague' regressions are anchored to the approximate Stacey and Kramers two stage Pb evolution model initial Pb composition (for Phanerozoic samples). Note 'Prague' secondary standard has a high failure rate due to significant zonation in U-Pb and Pb-Pb ratios in certain calcite chips.   The temperature of deformation affecting the Saraburi Group has been estimated in a number of 164 ways (cleavage type, calcite twin morphology, vitrinite reflectance, illite crystallinity), and the 165 approximate estimates based on the latter two techniques range between 160-220° C ± 20° C 166 (Hansberry et al., 2015). Calcite twins range between Type I (thin twins), Type II (tabular thick 167 twins), as well as Type III (bent twins) following the nomenclature of (Burkhard, 1993 Thirteen calcite samples were collected for U-Pb dating from a variety of geographic localities 172 within the KKFTB. Unfortunately, the majority of these samples did not produce useful calcite 173 U-Pb dates (large uncertainties) due to the low concentrations of uranium in the samples (success 174 rate of 23%). The localities that yielded useful results (Table 1)  The quarry for sample 10b has two faces, the northerly face exhibits an exposed, steeply-dipping 198 (70°SSW) bedding surface (Fig. 5B), while the western face is a dip-section (Fig. 5C). The 199 bedding surface exposes bed-perpendicular veins that strike in a N-S direction (Fig. 5B). These  (Table 1), which runs along the TPFZ, near the Srinagarind 247 Dam and reservoir (Fig. 6A). The outcrop section is composed of dark grey to medium grey, 248 fine-grained bedded Ordovician limestone that is strongly boudinaged (Fig. 6). These 'boudins' 249 are tens of meters in length and 30 -40 cm wide and host numerous calcite veins, from which 250 sample 12a was taken (Fig. 6C). Long, sub-horizontal striations mark bounding surfaces of the 251 boudins ( fig. 6B, C), which strongly suggest they are related to strike-slip motion. In addition, 252 the boudinaged layers are folded (Fig. 6), indicating they developed prior to, or accompanied 253 folding. Hence, calcite dating will allow to constrain the timing of regional deformation, 254 associated with fault activity in the TPFZ.

256
Of the sixteen samples that were screened for this work, only four samples provided useful U-Pb 257 dates (Table 1)  Sample 8b reveals a clear primary cleavage that appears to be cross-cut by later veinlets. These 311 later veinlets are enriched in many trace elements such as Al and Mn (Fig. 7). The CL texture of 312 the vein is fairly weak and homogeneous, except for the younger veinlets which are darker. CCI 313 shows a planar fabric that is pervasive throughout the primary calcite at a shallow angle to the 314 cleavage (Fig. 7). This fabric is interpreted as low-temperature deformation twinning of Type 1 315 or 2 due to the apparent narrow width of the twins and lack of recrystallization (Ferrill et al.,   The calcite has a very low CL response, and therefore, calcite crystal outlines and primary 321 growth zoning cannot be ascertained (Fig. 8). In CCI, the calcite exhibits a planar fabric that is 322 patchy in nature (Fig. 8). We interpret this to reflect high-temperature twinning (Type IV; Ferrill   (Figs. 7, 9). The upper intercept 207 Pb/ 206 Pb compositions for populations A 344 and B are 0.765 ± 0.059 and 0.769 ± 0.097, respectively (Fig. 9). Few additional data points were 345 discarded based on significantly elevated Al and/or U concentrations (proxy for detrital input) 346 associated with cracks through the calcite crystals (Fig. 7). where field observations indicate that this fault was most likely active during the Cenozoic. 387 Element mapping of sample 8b revealed the presence of high Al and elevated U (up to 1500% 388 increase) along cracks (Fig. 7). We interpret this as due to the presence of an Al-rich mineral 389 phase (such as a clay mineral) or alteration due to fluid flow along the cracks, and thus such data 390 were discarded. The sample is of particular interest due to the presence of multiple U-Pb data 391 populations (Fig. 9), that are associated with differences in Mn concentrations (Fig. 7). In more 392 detail, a U-Pb age of 197 ± 9 Ma was obtained for analyses in Mn-poor zones of the calcite 393 sample, while the U-Pb analyses in Mn-rich zones produced a much younger U-Pb age of 31 ± 6 394 Ma (Fig. 9). Ablation targets that were set close to boundaries between high and low Mn zones 395 produced mixing ages between both populations (Fig. 7, 9). 396 Chemical zonation within calcite may represent changes in fluid chemistry (and thus potentially 397 different fluid-flow events), or changes in uptake of metals (e.g. Barker and Cox, 2011;Paquette 398 and Reeder, 1995;Reeder et al., 1990). Experimental evidence (Frank et al., 1982) demonstrates 399 that Mn can show oscillatory zoning during calcite growth, related to uptake of Mn 2+ along the 400 calcite crystal surface that inhibits crystal growth. In fact, Mn zonation is the main source of 401 luminescence for calcite in CL imaging (Frank et al., 1982). Mn zonation in sample 8b, however, 402 does not correspond to oscillatory growth patterns (Fig. 7). Given this absence of oscillatory 403 growth patterns and the association between Mn zonation and age populations, we consider it 404 more likely that this zonation reflects changes in fluid chemistry between different 405 precipitation/alteration events. It is, therefore, interpreted that the calcite initially grew during the 406 Indosinian Orogeny (older age population with a poorly defined age of ~197 Ma), and that parts 407 were subsequently recrystallised or altered in a fluid with a higher Mn concentration, associated 408 with a Cenozoic deformation phase (poorly defined ~31 Ma age population). The mixing ages 409 (open ellipses in Figure 9) can then be interpreted as being partially reset Indosinian ages in 410 response to Pb mobility associated with the younger, Cenozoic, Mn-rich fluid infiltration. 411 The ~197 Ma age population (B) in sample 8b corresponds to calcite growth during Indosinian  (Fig. 8). Particularly, the Ce concentration map 436 reveals distinct zonations in the sample (Fig. 8). Therefore, the resulting U-Pb dates for this 437 sample were grouped into two populations associated with Ce-poor (population A) and Ce-rich