U-Pb (ID-TIMS) Geochronological Studies of High-Uranium Metamict Zircons: New Opportunities of Familiar Approaches

The possibility of the application of high-uranium self-irradiation-damaged metamict zircon for U-Pb geochronological studies (ID-TIMS) after preliminary high-temperature annealing and acid treatment was demonstrated using zircon from the Li–F granites of the Turga massif (Eastern Transbaikalia). Optimal conditions of high-temperature annealing and acid treatment of zircon with high self-irradiation α-dose can be selected to remove completely a metamict phase and preserve sufficient quantities of the mineral for U-Pb (ID-TIMS) studies.


INTRODUCTION
Self-irradiation caused by the presence of uranium admixture and inclusions of uranium-bearing minerals usually causes the modification of crystal lattice of zircon and local amorphization. As known, the migration ability of radiogenic lead isotopes depends on the state of crystal lattice, and respectively, on the degree of irradiation damages of zircon structure (Holland and Kupl, 1950;Cherniak et al., 1991). The losses of radiogenic lead and uranium cause the disturbance of U/Pb isotope ratios in zircon and, respectively, the discordance of obtained age values. In addition, zircon with damaged structure can contain common lead admixture, which introduces additional errors and uncertainties in the assessment of its U-Pb age.
In some cases, a technique of preliminary hightemperature annealing (Mattinson, 2005) and acid treatment (chemical abrasion) (Makeev, 1981;Mattinson, 1994) are successfully used for U-Pb dating of radiation damaged zircons. The high-temperature annealing partially recovers the zircon crystallinity, while acid treatment provides removal of damaged domains as well as mineral and fluid inclusions. Chemical abrasion is usually applied to zircon with well preserved crystal structure. However, this approach has not yet been used for zircons with heavy self-irradiation damages (with self-irradiation α-dose of D α > 2 × 10 18 α-decay/g, after (Zamyatin et al., 2019)) (Mattinson, 2005;Huyskens et al., 2016;Widmann et al., 2019).
In this paper, the optimal conditions of high-temperature annealing and chemical abrasion of highuranium zircon for its U-Pb dating (ID-TIMS) were determined using zircon from Li-F granites of the Turga massif (Eastern Transbaikalia) as an example. It is shown that the developed methodical approaches are efficient for dating of high-uranium metamict zircons with a high self-irradiation α-dose (D α ).

BRIEF GEOLOGY OF THE TURGA MASSIF
The Turga Li-F granite massif of the Kukulbey complex is located in the Turga River valley (Fig. 1), 350 km southeast of Chita, in the Turga-Kalangui ore zone with rare metal-gold-fluorite mineralization (Ob'yasnitel'naya …, 2001). The massif is restricted to the core of brachyanticlinal fold made up of Lower Jurassic sediments of the Onon and Onon-Borzya groups represented by mudstone, siltstone, sandstone, and conglomerates. In the northwest, the massif is in contact with Lower Jurassic sedimentary rocks, Paleozoic limestone, and dolomite, as well as granitoids of the Undin granite-granodiorite complex (Troshin et al., 1983). The most part of the massif is occupied by the Li-siderophyllite granites of the first phase. The amazonite granites of the second phase compose individual stocks and dike swarms confined to the contact zone with monzonites of the Shakhtama Complex.
Geochronological studies were carried out for representative samples from both rock varieties of the Kukulbey complex: medium-grained equigranular granites with Li-siderophyllite and medium to finegrained lithionite-amazonite-albite granites (Fig. 1). The granites of the Turga massif are subalkaline and ascribed to the peraluminous rare-metal granites, but have the elevated contents of high-field strength elements and contain agpaitic accessory mineralization (Syritso et al., 2021).

METHODS
The composition of zircons from granites of the Turga massif was studied on a Hitachi S-3400N scanning electron microscope equipped with Oxford Instruments X-Max20, AzTec Energy 350 energy dispersive spectrometer at U = 20 kV, I = 1.7 nA, working distance of 10 mm, beam diameter 5 μm, and acquisition time of 30 s (Centre for Geo-Environmental Research and Modelling (GEOMODEL), St Petersburg University Research Park). Raman spectra were recorded on an Horiba Jobin-Yvon LabRam HR 800 spectrometer equipped with a solid-state laser (λ = 532.37 nm, power of 100 mW) and on an Olympus BX 41 microscope equipped with ×10 and ×50 objectives (diameter of analyzed area of 2 μm, acquisition time of 50-100 s, measurement error of 0.5 cm -1 ) at the GEOMODEL Centre. The calibration was made using a 546.07-nm band of mercury lamp, which for used laser corresponds to a Raman shift of 471.26 cm -1 . Cathodoluminescence of the zircon was studied on a TESCAN VEGA3 scanning electron microscope (U = 15 kV, working distance of 12.5-13.0 mm) at the Institute of Precambrian Geology and Geochronology of the Russian Academy of Sciences, in St. Petersburg. Secondary electron (SE) images of zircon crystals were obtained on a HITACHI TM 3000 SEM (Resource Centre for Microscopy and Microanalysis, Research Park of the St. Petersburg State University).
The most transparent zircon crystals (40-300 grains) selected for U-Pb geochronological studies were subjected to high-temperature annealing in a SNOL E5CC muffle furnace in ceramic or quartz crucibles at 850 and 900°С for 48 hours (Mattinson, 2005) with subsequent acid treatment by 35% HF + 15% HNO 3 in proportions 5 : 1 for 2-6 h within a temperature range of 180-230°С. After the preliminary treatment, zircon was analyzed using standard technique (Krogh, 1973). Isotope measurements were carried out using 202 Pb-235 U spike on a multicollector TRITON TI mass spectrometer in static and dynamic (using ion counters) regimes. The accuracy of U/Pb ratios and U and Pb contents was 0.5%. The blanks did not exceed 15 pg Pb and 1 pg U. Experimental data were processed using PbDAT and ISOPLOT softwares (Ludwig, 1991(Ludwig, , 2003. Ages were calculated using uranium decay constants (Steiger and Jager, 1976). Corrections for common lead were introduced using Stacey-Kramers model (Stacey and Kramers, 1975).

L a k e B a i k a l L a k e B a i k a l L a k e B a i k a l T u r g a R .
T u r g a R .
Using formula from (Nasdala et al., 2001), the selfirradiation α-dose (D α ) for the studied zircons was calculated based on U and Th contents. The D α value is 6.0 × 10 18 -1.1 × 10 19 α-decay/g for zircon from the Li-siderophyllite granite and 7.0 × 10 18 -2.3 × 10 19 αdecay/g for zircon from the amazonite granite. Such high D α values indicate the intense development of irradiation-induced metamictization. For instance, at D α > 6 × 10 18 α-decay/g, the content of amorphous phase in zircon is more than 80% (Zhang and Salje, 2001). The U-Pb geochronological studies of these zircons without preliminary treatment usually provide no reliable geochronological data. Therefore, experiments were carried out to determine the optimal conditions of high-temperature annealing and preliminary acid treatment of above described zircons, which would allow one to obtain reliable ages of high-uranium metamict zircons.
High-temperature annealing. The most transparent zircons collected for U-Pb geochronological studies from Li-siderophyllite and amazonite granites of the Turga massif were subjected to high-temperature annealing at 850°C and 900°C for 48 h. In compliance with (Mattinson, 2005), the temperature of 850°C was taken as the standard annealing temperature. According to literature data (Widmann et al., 2019), at temperature higher than 915°С the recrystallization of amorphous domains occurs. In order to improve the crystallinity of weakly damaged domains without recrystallization of highly damaged domains and as a result, to preserve more material after acid treatment, the temperature of some annealing experiments was increased up to 900°С (nos. 4, 7, 10 in Table 1). Visible changes of zircon crystals occured in the insignificant increase in their transparency, change of color (appearance of well expressed red tint), as well as a significant increase of cathodoluminescence (Fig. 2, VI-VIII; Fig. 3, VII-XII), which suggests the partial recovery of their crystallinity. This is confirmed by the Raman spectroscopic studies. The Raman band at 1008 cm -1 (Fig. 6) corresponding to the most structurally sensitive vibrational B 1g mode (ν 3 ) of silicate ion SiO 4 in zircon is considered as indicator (Nasdala et al., 2001(Nasdala et al., , 2004. A shift of mode maximum could be caused by radiation distortion of zircon lattice, formation of solid solutions, and local stresses, while a strong widening of band ν 3 is typical of highly-damaged zircon (Shchapova et al., 2017(Shchapova et al., , 2018. High-temperature annealing of zircons from Lisiderophyllite and amazonite granites caused a well expressed increase of the Raman shift of ν 3 (SiO 4 ) band (toward 1008 cm -1 typical of "ideal" highly crystalline zircon) and a reduction in the band width (FWHM-full width at half maximum) (Fig. 7). This increase for zircon from Li-siderophyllite granite occurs on average from 991 ± 3 to 999 ± 2 cm -1 (the FWHM varies from 27 ± 16 to 15 ± 6 cm -1 ). The annealing temperature does not affect these parameters (Fig. 7a). Zircon from amazonite granite after annealing shows an increase of the Raman shift of ν 3 (SiO 4 ) band on average from 989 ± 3 to 996 ± 2 cm -1 (the FWHM varies from 48 ± 24 to 19 ± 4 cm -1 ). Thereby, after annealing at 900°С, the band of ν 3 PETROLOGY Vol. 29 No. 6 2021 IVANOVA et al. (SiO 4 ) mode in some cases is shifted closer to the "ideal" value of 1008 cm -1 (Fig. 7b), which may indicate that the elevated annealing temperature is more efficient for the recovery of crystallinity of metamict zircon.
Unfortunately, many of the studied zircon crystals lack band of ν 3 (SiO 4 ) mode in the Raman spectra (Fig. 6) due to the high irradiation damages, as well as fluorescence interference caused by the high contents of U, Th, and REE. In addition, all Raman spectra of zircon from the Turga granites display bands in the area of ~810 cm -1 (Fig. 6), which is not typical of this mineral and can be related to the (UO 2 ) 2+ in the lattice or the presence of torbernite microinclusions (Shchapova et al., 2017(Shchapova et al., , 2018. The described changes of zircon in response to the high-temperature annealing in general are consistent with data reported in (Nasdala et al., 1995(Nasdala et al., , 1998(Nasdala et al., , 2004Widmann et al., 2019;Geisler et al., 2001, and others). Annealing at 850°C seems to be optimal for the stud-  Intensity, rel. u.
ied zircons. An increase of annealing temperature up to 900°С likely leads to the more complete recovery of the crystallinity of damaged zones and increase of their stability to leaching, which requires longer dura-tion and likely higher temperature of acid treatment for the efficient removal of a metamict phase. An increase of annealing temperature up to 900°C is correlated with an increase of common lead fraction at decreasing 206 Pb/ 204 Pb ratio in analyzed samples (nos. 4, 7, 10 in Table 1). This tendency can be evidences of the excess recrystallization of zones with losses of radiogenic lead. Acid treatment. To study the efficiency of removal of a metamict phase of zircons during acid treatment, we conducted a series of experiments at different temperature and exposure duration. In most cases, only small fragments (5-20 μm) of zircon crystals have been preserved after acid treatment (Fig. 8).
Zircon from phase I Li-siderophyllite granite of the Turga massif was subjected to acid treatment at 220°С and 230°С for 2 and 4 hours (nos. 1-5, Table 1). A 2-h treatment presumably led to the removal of self-irradiation damaged zones and microinclusions of U-and Th-bearing minerals. Crystalline residues of zircon after 2-h treatment in general have the low content of common Pb (Table 1), while data points of their isotope composition plot on discordia with an upper intercept at 146 ± 4 Ma (MSWD = 0.067, at a zero lower intercept) (Fig. 9а). An increase of acid treatment duration up to 4 h (no. 5, Table 1) resulted in the practically complete dissolution of zircon crystals, and respectively, a sharp decrease of quantities of the min-eral available for study and a significant increase in the U/Pb determination error.
Zircon from phase II amazonite granite of the Turga massif is characterized by the higher ability to the dissolution compared to the zircon from Li-siderophyllite granite, which is consistent with its higher self-irradiation α-dose. A preliminary acid treatment at 180°C for 4 hours is considered to be optimal for this zircon. As seen from Fig. 9b, the data points of isotope composition of zircon residues after treatment under these conditions (nos. 6, 7, 8; Table 1) are approximated by discordia with the lower intercept at 141 ± 1 Ma (MSWD = 0.014, lower intercept at 1575 ± 470 Ma).
The less destructive treatment of zircon from the amazonite granite provides preservation of not only finely crystalline residue, but also significant quantities of seemingly undisturbed or weakly fragmented zircons. An increase of temperature up to 220°С (no. 9, Table 1) or an increase of exposure duration up to 6 h (no. 10, Table 1) lead to a significant reduction of the crystalline residue and increase of measurement error of U/Pb ratios. At the same time, it is probable that an increase of annealing temperature up to 900°С may facilitate the partial recovery of damaged zones and increase of their resistivity to acid treatment (nos. 7 and 10, Table 1). Preliminary annealing at 850°C is inferred to be optimal for this zircon.

CONCLUSIONS
Our studies first demonstrated the principal opportunity of the use of high-uranium highly radiation damaged metamict zircon for U-Pb (ID-TIMS) geochronological studies. It is possible to determine the optimal conditions of high-temperature annealing and acid treatment of zircon with high self-irradiation α-dose to provide a practically complete removal of a metamict phase and to obtain sufficient quantities of zircon for U-Pb (ID-TIMS) dating. Such conditions for zircon with D α @ 2 × 10 18 α-decay/g are high-temperature annealing at 850°C and leaching in a mixture of 35% HF + 15% HNO 3 (5 : 1) at 180°С for 4 h.

ACKNOWLEDGMENTS
We are grateful to A.V. Chugaev (IGEM RAS) for valuable comments, which significantly improved the manuscript.

FUNDING
The studies were supported by the Russian Foundation for Basic Research (project nos. 18-05-00957 and 20-05-00437) and IPGG RAS (theme no. FMNU-2019-0005). 10 μm 10 μm 10 μm 2 μm 2 μm 2 μm 5 μm 5 μm 5 μm 10 μm 10 μm 10 μm 10 μm 10 μm 10 μm 10 μm 10 μm 10 μm This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons license, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons license and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this license, visit http://creativecommons.org/licenses/by/4.0/.