Laser ablation MC-ICPMS U-Th and U-Th-Pb dating of Quaternary zircons from Jeju Island, Korea

Abstract

Background

Laser ablation inductively coupled plasma mass spectrometry (LA-ICPMS) emerged in the mid-1980s and rapidly became a crucial dating tool. The advent of femtosecond LA systems has substantially reduced volatility-dependent mass fractionation. This study showcases U-Th and U-Th-Pb dating results of Quaternary zircons collected from Jeju Island, Korea, utilizing an advanced femtosecond laser-connected multi-collector (MC)-ICPMS.

Findings

Zircon grains from trachyte samples near the Baeknokdam lake (JJ616-1), Yeongsil (JJ08-1), Chunwangsa (JJ09-1), and Oraidong (JJ09-3) provided weighted mean ²³⁸U-²³⁰Th ages of 28.7 ± 1.6 ka (n = 56/64, MSWD = 3.8), 81.8 ± 10.9 ka (n = 11/12, MSWD = 1.6), 92.6 ± 4.6 ka (n = 49/51, MSWD = 2.2), and 117.6 ± 8.2 ka (n = 48/50, MSWD = 3.2), respectively. The age determination for JJ08-1 zircon aligned well with the recommended value (82 ± 6 ka). Zircons from Sanbangsan (JJ615-1) and Wonmansa (JJ08-2) trachytes yielded common Pb and radioactive disequilibrium-corrected weighted mean ²³⁸U-²⁰⁶Pb ages of 785 ± 5 ka (n = 27/28, MSWD = 0.90) and 743 ± 8 ka (n = 28/30, MSWD = 0.79), respectively. The weighted mean ²³⁸U-²⁰⁶Pb ages of Penglai and 61.308 reference zircons were determined to be 4226 ± 21 ka (n = 22/25, MSWD = 3.8) and 2488 ± 20 ka (n = 19/20, MSWD = 1.8), respectively. These ages are concordant with the recommended values.

Conclusions

Our data provides additional evidence of trachyte magmatism occurring in Jeju Island during the transitional period between the Early and Middle Pleistocene and the Late Pleistocene. The zircon samples analyzed in this study could serve as reference age data for Quaternary geochronology research.

Introduction

In the 1980s, the concept of coupling a laser ablation (LA) system with an inductively coupled plasma mass spectrometer (ICPMS) was introduced (Gray 1985). In LA-ICPMS analysis, a pulsed laser is used to ablate the surface of the sample within a gas-tight chamber, creating a stream of aerosol carried to an ICP source. Within the ICP source, the aerosol is vaporized and converted into ions. These ions are then separated according to their mass-to-charge ratios using a mass analyzer and quantified by a detector. Throughout these processes, elemental and isotopic fractionation is inevitable due to various factors, such as differences in elemental volatility and transportability, and time-dependent changes in particle size distribution (Hirata and Nesbitt 1995; Guillong and Günther 2002; Hirata 2003; Jackson and Günther 2003; Jackson et al. 2004).

Over the past few decades, significant instrumental and strategic advancements have addressed these challenges and enhanced the capabilities of LA-ICPMS (Sylvester and Jackson 2016). The use of helium has improved the transport of ablated aerosols to the ICP, leading to increased signal intensities. The introduction of a two-volume sample chamber design has provided rapid signal response, effectively reducing positional bias and enabling higher spatial resolution. Additionally, the reduction in laser wavelength has resulted in higher absorption and improved ablation of transparent materials, as well as more efficient ionization in the ICP. The initial generation of laser ablation system utilized long wavelength, visible laser beams (Gray 1985), but it was soon recognized that the shorter wavelength, particularly the 213 nm Nd:YAG and 193 nm ArF Excimer, offered more efficient and stable ablation of geologic materials. In the 2000s, the introduction of femtosecond lasers proved highly effective in reducing target heating and minimizing volatility-dependent elemental and isotopic fractionation. This advancement resulted in several benefits, such as higher evaporation efficiency and a shift in the size distribution of sample particles toward smaller sizes, thus enhancing the precision and accuracy of data (Russo et al. 2002; Poitrasson et al. 2003; Diwakar et al. 2013). The application of femtosecond laser technology would be particularly beneficial for absolute element concentration determinations and stable metal and metalloid isotope analyses (Poitrasson and d’Abzac 2017).

The pioneering work of Feng et al. (1993) and Fryer et al. (1993), based on ²⁰⁷Pb/²⁰⁶Pb determinations on Precambrian zircon, showcased the utility of LA-ICPMS as a dating tool. The versatility, ease-of-use, speed, and relatively moderate cost of LA-ICPMS addressed a significant challenge in ion probe dating—the high expense of instrumentation and the limited availability of suitably equipped laboratories. Since the mid-1990s, continuous advancements in the laser system and ICPMS instrumentation have facilitated the dating of zircon and other accessory minerals using the Pb/U decay scheme (Woodhead et al. 2016). The challenges associated with LA-ICPMS U-Pb geochronology, particularly the fractionation of Pb relative to U during ablation, transport, and ionization processes, as well as common Pb correction difficulties, have been extensively discussed (e.g., Jackson et al. 2004). The present-day spatial resolution, and potentially precision, in ²³⁸U-²⁰⁶Pb age determination have become nearly comparable to that achieved by the ion probe technique, provided that well-characterized matrix-matched reference materials are available (Woodhead et al. 2016).

The dating of Quaternary zircon through in situ U-Th analysis began in the 1990s, employing the ion probe technique (e.g., Reid et al. 1997). While the potential of LA-ICPMS in U-Th analysis of zircon was recognized early on (Stirling et al. 2000), its widespread adoption came much later (Bernal et al. 2014). In situ U-Pb dating of Quaternary zircon commenced earnestly in the 2000s using the ion probe (Bacon et al. 2000). Given the analytical challenges in U-Th and U-Pb dating for Quaternary zircon, stemming from extremely small amounts of radiogenic daughter nuclides in the decay chains, the instrumental bias should be corrected with a great care. The use of appropriate mineral standards is critical not only for correcting mass bias and elemental fractionation, but also for ensuring the overall accuracy and reliability of LA-ICPMS age determinations. As part of our effort to develop reference materials, we present the results of U-Th and U-Th-Pb dating for Pleistocene zircons collected from Jeju volcanic island, Korea, using an advanced femtosecond laser-connected multi-collector (MC)-ICPMS.

Principles of zircon U-Th and U-Th-Pb dating

The principles of zircon U-Th and U-Th-Pb dating are well-explained in textbooks (e.g., Faure and Mensing 2005), and only briefly reminded here.

The radioactivity of ²³⁰Th in zircon is attributed to two components. Firstly, there is the excess ²³⁰Th decay, expressed as $\rm{(}{{}^{230}\rm{Th)}}_{\rm{excess}}^{0}$ e^−λt (λ is the decay constant of ²³⁰Th = 9.1706 $\times$ 10^–6 y⁻¹), with a half-life of 75,584 ± 110 years (Cheng et al. 2013). Secondly, there is the activity of supported ²³⁰Th, represented as (²³⁸U)(1 − e^–λt), which increases with time. On the ²³⁰Th/²³²Th (activity ratio, y-axis) and ²³⁸U/²³²Th (activity ratio, x-axis) plot, zircon data conform to a linear relationship (isochron) with a slope of 1 − e^–λt and a y-intercept of ${\rm{(}{}^{230}{\rm{Th}}/{}^{232}{\rm{Th}}\rm{)}}_{\rm{excess}}^{0}$ e^–λt provided that the analyzed parts of the zircon share the same age and initial ²³⁰Th/²³²Th ratio, and have remained closed to U and Th after crystallization. After approximately 5 times the ²³⁰Th half-life (over 350,000 years), a state of secular equilibrium is re-established, resulting in all data points aligning along the 1:1 line (equiline) of the ²³⁰Th/²³²Th-²³⁸U/²³²Th diagram. Consequently, the information needed for zircon U-Th disequilibrium dating is the ²³⁰Th/²³²Th and ²³⁸U/²³²Th ratios.

In the case of old zircon that has attained secular equilibrium among radionuclides in the U-Th-Pb decay chains, the production rates of Pb isotopes at the end of the chains become equal to the decay rates of their respective grandparent nuclides. Consequently, the decay of U and Th isotopes can be treated as if they directly transform into Pb isotopes. However, for young (i.e., Pliocene or younger) zircon, one must consider isotope fractionation effects involving intermediate daughters such as ²³⁰Th and ²³¹Pa. Zircon has the tendency to exclude a significant proportion of ²³⁰Th in the magma. As a result, after decay, there will be a deficiency in ²⁰⁶Pb, leading to the underestimation of the ²³⁸U-²⁰⁶Pb age. This effect was already recognized as early as the 1970s (Mattinson 1973; Ludwig 1977) and quantified by Schärer (1984) with Eq. (1).

$$^{206} {\rm{Pb}} ={\rm{^{238}U}}\left[ {(e^{{\lambda_{238} t}} - 1) + (\lambda_{238} /\lambda_{230} )(f_{{{\rm{Th}}} /{\rm{U}}} - 1)} \right]$$

(1)

where f_Th/U = [Th/U]_{concentration, zircon}/[Th/U]_{concentration, magma}, λ₂₃₈ = decay constants of ²³⁸U = 1.55125 $\times \hspace{0.17em}$10^–10 y⁻¹(Steiger and Jäger 1977), λ₂₃₀ = decay constants of ²³⁰Th = 9.1706 $\times$ 10^–6 y⁻¹ (Cheng et al. 2013) and t = age.

More recently, Sakata (2018) proposed a practical approach for calculating the corrected U-Pb age of Quaternary zircon. This scheme is a kind of ²⁰⁷Pb-correction method based on a modified Tera-Wasserburg concordia. Since zircon typically crystallizes in intermediate to felsic magma, the influence of melt disequilibria can be reasonably neglected. In this case, the modified concordia is constructed using Eqs. (2) and (3) (Sakata et al. 2017). A bisection method between the modified concordia and the common Pb line can be used to calculate the ²³⁸U-²⁰⁶Pb age corrected for the contributions from both initial disequilibria and common Pb, and its uncertainty.

$$\left( {^{206} {\rm{Pb}}/^{238} {\rm{U}}} \right) \, = (e^{{\lambda_{238} t}} - 1) + \left( {\lambda_{238} /\lambda_{230} } \right)\left( {f_{{\rm{Th/U}}} {-} \, 1} \right)(1{-}e^{{{-}\lambda_{230} t}} )e^{{\lambda_{238} t}}$$

(2)

$$\left( {^{207} {\rm{Pb}}/^{235} {\rm{U}}} \right) \, = (e^{{\lambda_{235} t}} - 1) + \left( {\lambda_{235} /\lambda_{231} } \right)\left( {f_{{\rm{Pa/U}}} {-} \, 1} \right)(1 \, {-}e^{{{-}\lambda_{231} t}} )e^{{\lambda_{235} t}}$$

(3)

where f_Pa/U = [Pa/U]_zircon/[Pa/U]_magma, λ₂₃₅ = decay constants of ²³⁵U = 9.8485$\times$10^–10 y⁻¹ (Steiger and Jäger 1977), λ₂₃₁ = decay constants of ²³¹Pa = 2.13$\times$10^–5 y⁻¹ (Lide and Frederikse 1995) and t = age.

For instance, consider a 1 Ma zircon without common Pb and featuring a Th/U ratio five times lower than that of the melt (f_Th/U = 0.2). The difference between disequilibrium-corrected and uncorrected ²³⁸U-²⁰⁶Pb ages for this zircon reaches approximately 9%, significantly surpassing the bounds of analytical uncertainty.

Geology of Jeju Island and sample collection

Jeju Island, the largest island in Korea, is situated on the continental shelf around 80 km off the south coast of the Korean Peninsula (Fig. 1). The volcanic activity on Jeju Island began during the Early Pleistocene, approximately 1.7 million years ago. The island has been shaped by quasi-continuous small-volume volcanism and intermittent large-volume lava effusion. This has resulted in the formation of an elongated symmetrical shield volcano with Mt. Halla at its summit, accompanied by over 300 monogenetic scoria cones, as well as minor tephra rings, maars, and lava domes. Geochronological and geochemical analyses suggest that Jeju volcanic rocks can be divided into three geochemically distinct groups: the Early-Middle Pleistocene high-Al alkalic magma suite (Stage 1), the Middle Pleistocene transitional alkalic magma suite (Stage 2), and the Middle to Late Pleistocene low-Al alkalic magma suite (Stage 3) (Brenna et al. 2015, and references therein). Presently, the island’s surface is predominantly covered with rocks from Stage 3 volcanism. Studies employing optically stimulated luminescence and radiocarbon dating have confirmed that volcanic activity persisted into the Holocene (Yeo et al. 2019, and references therein).

Fig. 1

Digital elevation model of Jeju Island based on ASTER satellite data acquired on 30 November, 2013, with sample locations

In this study, six trachyte samples were collected from inner areas and the outer rim of the island, as shown in Fig. 1. The sampling sites include outcrops near Sanbangsan (sample JJ615-1), the Baeknokdam lake at the summit of Mt. Halla (JJ616-1), Yeongsil (JJ08-1), Wonmansa (JJ08-2), Chunwangsa (JJ09-1), and Oraidong (JJ09-3). Drawing from previously published age data (Koh et al., 2013; Marsden et al. 2021), we presume that the Sanbangsan and Wonmansa trachytes formed during Stage 1, whereas the Baeknokdam, Yeongsil, and Chunwangsa trachytes are associated with Stage 3. The age of the Oraidong trachyte remains undetermined. Additional file 1: Table S1 provides detailed information about these trachytes, encompassing the GPS coordinates, rock types, mineralogy, and texture.

Analytical methods

The major and trace element concentrations of the trachyte whole-rocks were analyzed at Actlabs (Canada) using the combination of X-ray fluorescence spectrometry for fused glass beads and quadrupole ICPMS and ICP optical emission spectrometry for dissolved solutions of the beads.

Zircon grains were extracted through conventional sieving, magnetic, and heavy liquid techniques, and then embedded in epoxy along with reference materials including the 91500, 61.308 (Wiedenbeck et al. 1995, 2004), Plešovice (Sláma et al. 2008), Penglai (Li et al. 2010), FC-1 (Paces and Miller Jr 1993), Temora 2 (Black et al. 2004), and LKZ-1 (Cheong et al. 2019) zircons. The polished surfaces of the zircons were analyzed using a scanning electron microscopy (JEOL JSM-6610LV) at the Korea Basic Science Institute (KBSI) to obtain cathodoluminescence (CL) and backscattered electron images. For isotopic analysis of U, Th, and Pb, a Plasma II MC-ICPMS (Nu Instruments) with a 257 nm femtosecond LA system (Excite Pharos, Teledyne Cetac) was utilized at the KBSI. The instrument’s operational parameters are summarized in Additional file 2: Table S2. The obtained data were processed using the Iolite 2.5 software within the Igor Pro 6.3.5.5 program (Paton et al. 2011). Weighted mean ages were calculated using the Isoplot 3.75 program (Ludwig 2012) and are reported at the 95% confidence level.

For zircon U-Th dating, signal intensities were simultaneously measured using Faraday collectors (for ²³⁸U and ²³²Th) and ion counters (for ²³⁰Th and 228 mass). The measured ²³⁸U/²³²Th ratios were corrected based on the recommended values of the 91500 zircon ([U] = 80.0 μg/g, [Th] = 29.9 μg/g; Wiedenbeck et al. 2004). The ²³⁸U and ²³²Th counts of the 91500 zircon were employed to approximately determine U and Th concentrations in the sample spots. The peak intensity of 228 mass was used for correcting the contribution of molecular zirconium sesquioxide ions to the ²³⁰Th peak (Guillong et al. 2015, 2016; 230 mass = ⁹⁰Zr⁹²ZrO₃⁺ + ⁹¹Zr₂O₃⁺  = 0.7142 × 228 mass (= ⁹⁰Zr₂O₃⁺), while maintaining ThO⁺/Th⁺ below 0.005. The peak tail of ²³²Th, the relative sensitivity between U and Th, and the Faraday-ion counter efficiency were collectively corrected based on the assumption that the Plešovice zircon (²⁰⁶Pb/²³⁸U age = 337 Ma; Sláma et al. 2008) is in ²³⁸U-²³⁰Th secular equilibrium. Radioactivity ratios were calculated using the decay constants proposed by Steiger and Jäger (1977) (for ²³⁸U and ²³²Th) and Cheng et al. (2013) (for ²³⁰Th). After implementing the necessary corrections and calibrations, we confirmed that the radioactivity ratios of (²³⁰Th/²³²Th) and (²³⁸U/²³²Th) for the 91500, FC-1, Temora 2, and LKZ-1 standards plotted on the equiline (Additional file 6: Fig. S1).

For U-Th-Pb dating, signal intensities were measured simultaneously with Faraday collectors (for ²³⁸U and ²³²Th) and ion counters (for ^{208, 207, 206, 204}Pb and ²⁰²Hg). To ensure accurate representation of rapidly transient signals during LA sampling, a peak hopping protocol was implemented. Instrumental mass bias and elemental fractionation, as well as Faraday-ion counter efficiency, were collectively calibrated against the 91500 zircon data. The isotopic ratios of the 91500 zircon reported in Wiedenbeck et al. (1995) were used for this calibration. Each analysis cycle comprised 30 s of measuring the instrumental background by analyzing the carrier gas, followed by a 30-s ablation event, resulting in a total analysis time of approximately 2 min.

Results and discussion

Whole-rock composition

Additional file 3: Table S3 presents the chemical compositions of the trachyte whole-rocks. The SiO₂ contents in these samples range from 59.38 to 65.16 wt. %, accompanied by corresponding total alkali contents (Na₂O + K₂O) ranging from 9.12 to 11.33 weight %. Following the naming scheme of Middlemost (1994), the trachyte samples are designated as either trachyandesite (sample JJ09-3) or trachyte (for the other samples). Based on the K₂O versus SiO₂ relationship (Le Maitre et al. 1989), these samples consistently belong to the shoshonite series. The observed enrichment of large-ion-lithophile elements relative to mid-oceanic ridge basalts and light rare earth element-enriched chondrite-normalized pattern align with the general characteristics of oceanic island basalts (Sun and McDonough 1989). The Th/U ratios span a relatively narrow range between 3.91 and 4.85.

Zircon morphology and CL texture

Figure 2 displays representative CL images of separated zircon grains. The majority of the grains exhibit CL textures that indicate their primary magmatic origin, and they remain mostly non-metamict, unaltered, and lacking xenocrystic cores. Zircon crystals from sample JJ615-1 are transparent or translucent, appearing as euhedral to subhedral prismatic shards with lengths of up to 200 μm. Under CL, they display sector or banded zoning, upon which oscillatory zoning is imposed. While some shards are unzoned and appear homogeneous in CL, most exhibit clear zoning patterns. In contrast, JJ616-1 zircons are transparent and vary in shape from equant to prismatic crystals, ranging in length from approximately 150 to 300 μm. The majority of these grains show evident oscillatory zoning under CL. Zircons from sample JJ08-1 appear transparent or pale brown in color, exhibiting euhedral to subhedral crystal faces with lengths ranging between 100 and 200 μm. Most grains show faint to clear oscillatory zoning, while some display resorption patterns. Needle-shaped inclusions, likely apatite, are frequently observed within the grains. Zircons in sample JJ08-2 are transparent to translucent pale brown crystals, varying in size from < 50 to 300 μm. The grains exhibit euhedral to anhedral faces, and most of them display clear or faint oscillatory zoning under CL. Zircons from sample JJ09-1 are long prismatic euhedral grains, with lengths reaching up to 300 μm. Needle-shaped inclusions are frequently observed within the grains. Under CL, oscillatory or sector zoning is observed in most grains. JJ09-3 zircons are mostly transparent, appearing as euhedral prismatic or subhedral crystals, with lengths between 100 and 200 μm. Sector or oscillatory zoning is observed under CL, and some grains contain needle-shaped inclusions.

Fig. 2

Representative cathodoluminescence images of Jeju zircons. Scale bars represent 100 μm. Certain grains from samples JJ615-1 and JJ616-1 exhibit remnants of analytical spots

U-Th dating

The U-Th concentrations and radioactivity ratios of the zircons are provided in Additional file 4: Table S4. Zircons from sample JJ616-1 near the Baeknokdam lake exhibit moderate to high U (525 ± 252 μg/g, one standard deviation, same hereafter unless otherwise noted) and Th (653 ± 469 μg/g) concentrations, with (²³⁸U/²³²Th) activity ratios ranging from 1.74 to 4.06. These zircons yield a weighted mean ²³⁸U-²³⁰Th age of 28.7 ± 1.6 ka (n = 56/64, mean squared weighted deviation (MSWD) = 3.8) as shown in Fig. 3. Considering the preserved typical igneous texture in these zircons, as well as in those from other samples (Fig. 2), this age is believed to represent the magmatic crystallization. Previous studies near Baeknokdam lake have utilized various dating methods, such as optically stimulated luminescence (Ahn and Hong 2017) and the zircon double dating method (Marsden et al. 2021), which combines U-Th-Pb and (U-Th)/He techniques. The results of these studies ranged from 78 to 2 ka, with a concentrated population between approximately 30 and 20 ka. Our data further support the significance of ~ 30 ka magmatism in the formation of trachyte magma beneath Mt. Halla.

Fig. 3

Uranium–thorium isochron plots for Jeju zircons. The dashed gray lines depict the equiline. The open circles indicate outliers identified through the calculation of weighted mean age. Error bars represent 2 standard errors

Zircons from sample JJ08-1 yield a weighted mean ²³⁸U-²³⁰Th age of 81.8 ± 10.9 ka (n = 11/12, MSWD = 1.6) (Fig. 3), with moderate U (428 ± 343 μg/g) and Th (440 ± 367 μg/g) concentrations. Sample JJ08-1 was taken from the same outcrop containing the SS14-28 zircon standard, which has a recommended ²³⁸U-²³⁰Th age of 82 ± 6 ka (Marsden et al. 2022). Our data provide additional analytical results supporting that age.

JJ09-1 zircons provide a weighted mean ²³⁸U-²³⁰Th age of 92.6 ± 4.6 ka (n = 49/51, MSWD = 2.2) (Fig. 3), with moderate U (307 ± 197 μg/g) and Th (299 ± 366 μg/g) concentrations. The sampling locality is the same as for sample HS100-5 in Marsden et al. (2021), which yielded comparable U-Th dates ranging from 87 ± 23 to 168 ± 117 ka, with a mean of 111 ± 14 ka. Zircons from sample JJ09-3 yield a slightly older weighted mean ²³⁸U-²³⁰Th age of 117.6 ± 8.2 ka (n = 48/50, MSWD = 3.2) (Fig. 3). They also have moderate U (263 ± 188 μg/g) and Th (295 ± 284 μg/g) concentrations.

Our U-Th zircon ages further support the Late Pleistocene magmatic events on Jeju Island as revealed by previous studies (Koh et al. 2013; Marsden et al. 2021).

U-Th-Pb dating

The zircon U-Th-Pb isotope data are listed in Additional file 5: Table S5. The results are discussed below, considering common Pb and radioactive disequilibrium correction.

For the Penglai reference zircon, the f_Th/U (Eq. 1) varies from 0.07 to 0.15, with an average of 0.12 ± 0.03, assuming that the magma Th/U ratio is the same as that reported for the host basalt (Th/U = 4.09; Ho et al. 2000). Applying the scheme proposed by Sakata (2018), the disequilibrium- and common Pb-corrected weighted mean ²³⁸U-²⁰⁶Pb age is determined to be 4,226 ± 21 ka (n = 22/25, MSWD = 3.8), as shown in Fig. 4. In this calculation, 30% error was assigned to f_Th/U. The f_Pa/U (Eq. 3) and common ²⁰⁷Pb/²⁰⁶Pb ratio were assumed to be 3.50 ± 1.05 and 0.84 ± 0.05, respectively. This corrected age aligns well with the recommended ²³⁸U-²⁰⁶Pb age of 4.29 ± 0.05 Ma, as recently reported by further evaluation for the Penglai zircon (Yu et al. 2020). The fraction of common ²⁰⁶Pb (f²⁰⁶Pb_c), based on the ²⁰⁷Pb/²⁰⁶Pb ratio on the normal concordia at the corrected ²³⁸U-²⁰⁶Pb age, ranges from 5 to 15%. Notably, considering only the f²⁰⁶Pb_c, the ²³⁸U-²⁰⁶Pb dates yield a distinctly younger weighted mean of 4,118 ± 22 ka (n = 22/25, MSWD = 5.7), confirming the significance of disequilibrium correction. The ²³²Th-²⁰⁸Pb dating may be less impacted by the disequilibrium effect. However, it should be noted that the measured ²⁰⁸Pb/²⁰⁶Pb ratios of the zircons and assumed common Pb ratio (= 2.0658; Stacey and Kramers 1975) indicate much higher fractions of common ²⁰⁸Pb (f²⁰⁸Pb_c) ranging from 39 to 79%. The ²⁰⁷Pb-corrected ²³²Th-²⁰⁸Pb dates are scattered between 2.8 and 4.2 Ma, appearing much younger than the corrected ²³⁸U-²⁰⁶Pb age. This discrepancy principally resulted from high common Pb fraction, and the uncertainty in the common Pb isotopic composition. To achieve a more precise measurement of the Th-Pb age, it is essential to know the exact ²⁰⁸Pb/²⁰⁶Pb ratio of the common Pb in order to accurately determine the f²⁰⁸Pb_c (= [(²⁰⁸Pb/²⁰⁶Pb)_common/(²⁰⁸Pb/²⁰⁶Pb)_measured] × f²⁰⁶Pb_c).

Fig. 4

Tera-Wasserburg plots for Jeju zircons and standards. Common Pb-uncorrected U-Pb data are plotted with the conventional (in black) and modified (in red) concordia curves. The open circle indicates an outlier identified through the calculation of weighted mean age. Error bars are at the 2σ level. Modified concordia curves are constructed following Sakata (2018), utilizing average f_Th/U (= (Th/U)_zircon/(Th/U)_whole-rock) values. The numbers marked on the concordia curves represent dates in ka. Dashed lines on the plot depict the mixing lines between common and radiogenic Pb

As depicted in Fig. 4, the 61.308 zircon yields a disequilibrium- and common Pb-corrected weighted mean ²³⁸U-²⁰⁶Pb age of 2,488 ± 20 ka (n = 19/20, MSWD = 1.8) when assuming a magma Th/U ratio of 3.89 (the average Th/U of upper continental crust; Rudnick and Gao 2003). The ²⁰⁷Pb-corrected common Pb fraction is determined to be 13–51% for ²⁰⁶Pb and 42–95% for ²⁰⁸Pb. It is worth noting that the 61.308 zircon initially consisted of three crystals (Wiedenbeck et al. 1995). The grains analyzed in this study may correspond to 61.308A, given their U (170 ± 85 μg/g) and Th (179 ± 151 μg/g) concentrations. The thermal ionization mass spectrometric ²³⁸U-²⁰⁶Pb and ²³²Th-²⁰⁸Pb ages of 61.308A zircon were reported as 2,488 ± 4 and 2,538 ± 10 ka, respectively (Wiedenbeck et al. 1995), which align with our measurements. Similar to the Penglai zircon, the ²³²Th-²⁰⁸Pb dates of 61.308 zircon exhibit significant scatter and appear to be younger than the ²³⁸U-²⁰⁶Pb age.

The same correction scheme results in weighted mean common Pb and disequilibrium-corrected ²³⁸U-²⁰⁶Pb ages of 785 ± 5 ka (n = 27/28, MSWD = 0.90) and 743 ± 8 ka (n = 28/30, MSWD = 0.79) for the JJ615-1 and JJ08-2 zircons, respectively (Fig. 4). The f_Th/U values were determined using the whole-rock Th/U ratios, and the f_Pa/U and common Pb isotopic compositions were assumed to be consistent with those used for dating the Penglai and 61.308 zircon. The ²⁰⁷Pb-corrected f²⁰⁶Pb_c is notably higher in the JJ08-2 zircons (0.44 ± 0.13) compared to the JJ615-1 zircons (0.11 ± 0.06). Despite the high common Pb fraction, the JJ08-2 zircons show a good fit of the mixing line between the radiogenic and common Pb (Fig. 4). On the other hand, the ²⁰⁷Pb-corrected ²³²Th-²⁰⁸Pb dates of JJ615-1 zircons (727 ± 11 ka, n = 28, MSWD = 26) are younger and more scattered compared to the corrected ²³⁸U-²⁰⁶Pb dates. The Th-Pb dating method was not applicable for JJ08-2 zircon because ²⁰⁸Pb in this zircon is predominantly composed of common Pb (average > 95%). The zircon U-Pb age of sample JJ615-1 marginally agrees with the whole-rock Ar-Ar age (802 ± 5 ka; Koh et al. 2013) of a trachyte collected from the Sanbangsan area. Interestingly, this age corresponds to the timing of the Matuyama/Brunhes magnetic reversal, defining the Early-to-Middle Pleistocene boundary (~ 780 ka). The zircon U-Pb age of sample JJ08-2 falls within the range of zircon dates (377–988 ka; Marsden et al. 2021) reported for a trachyte lava collected from the same locality.

Concluding remarks

This study employed femtosecond laser-connected MC-ICPMS to determine the U-Th and U-Th-Pb ages of Pleistocene zircons collected from Jeju Island, Korea. The obtained ages for these zircons and reference materials generally agree with the previously recommended ages. Our findings underscore the importance of applying radioactive disequilibrium correction when dating young (< 5 Ma) zircons. Further investigations are required to enhance zircon Th–Pb dating by accurately measuring the common Pb isotopic composition.