Dehydration-Induced Supralithostatic Fluid Overpressure in Metapelites from Northern Korea

doi:10.21203/rs.3.rs-4677052/v1

Download PDF

Article

Dehydration-Induced Supralithostatic Fluid Overpressure in Metapelites from Northern Korea

https://doi.org/10.21203/rs.3.rs-4677052/v1

This work is licensed under a CC BY 4.0 License

Version 1

posted

You are reading this latest preprint version

Our understanding of orogenic dynamics largely relies on the lithostatic paradigm, which correlates thermodynamic pressure to burial depths. Deviations from lithostatic conditions often involve fluid overpressure, yet the mechanisms in natural rocks remain elusive. In this study, we integrate petrological analysis, radiometric dating, and diffusion chronology to explore the thermal overprinting on metapelites from a Precambrian orogen caused by ~ 110 Ma andesitic porphyry intrusions in Northern Korea. The metapelites’ mineral assemblages, particularly the dissolution-reprecipitation textures of garnet and monazite, reveal rapid heating, dehydration, and reworking within ~ 300 years, reaching 600‒640°C and 7.5‒8.5 kbar at ~ 110 Ma. The barometric results sharply contrast with the shallow emplacement depths of the andesitic porphyries, indicating supralithostatic fluid overpressure. These findings challenge the lithostatic paradigm and suggest that rapid dehydration and deviation from lithostatic pressures may be more prevalent than previously recognized, with significant implications for orogenic reconstructions and seismogenic behaviors at plate boundaries.

Earth and environmental sciences/Solid Earth sciences/Petrology

Earth and environmental sciences/Solid Earth sciences/Mineralogy

Earth and environmental sciences/Planetary science/Seismology

lithostatic paradigm

fluid overpressure

diffusion chronology

rapid dehydration

garnet

Thermodynamic pressure estimated from mineral assemblages is routinely converted into depth by assuming that chemical reactions take place in lithostatic conditions—a fundamental assumption for reconstructing orogenic processes^1,2. In particular, the growth and compositions of garnet, a widespread mineral in crustal rocks, has been widely used as reliable proxies for rocks’ burial or exhumation depths^3,4. However, recent studies have highlighted pressure variations and deviatoric stresses from grain to regional scales^2,4–7, challenging the lithostatic paradigm. Deviations from lithostatic conditions may also arise from rapid dehydration reactions resulting in fluid pressures exceeding those of lithostatic load^8–11. Such fluid overpressure is essential for phenomena like volcanic eruptions and volatile emissions from Earth’s interiors^11,12, and represents a critical factor in models of cyclic slip failure and earthquake nucleation as it can significantly impact fault strength and stress distribution in the crust^13–15.

The extent to which fluid pressure can exceed lithostatic pressure and the duration for which such overpressure can persist remain contentious. Numerical and experimental studies have explored the potential of rapid dehydration in low-permeability rocks to generate supralithostatic fluid overpressure, with estimates ranging from a few bars to kilobars due to uncertainties in kinetic models and boundary conditions^9,10,16–18. Dehydration-induced fluid overpressure leading to slip failure or tectonic tremors could be transient, lasting up to a few hundred years^8,9,19. Despite these insights, there is rare natural rock evidence to validate both the magnitude and timescale of dehydration-induced fluid overpressure. The scarcity of natural examples—partly obscured by its transiency and subsequent deformation or recrystallization^12,20—makes it challenging to link such fluid overpressure to broader tectonic implications.

Prograde growth of garnet is associated with dehydration reactions that release significant amounts of fluids in pulses²¹, which, taking place in low-permeability rocks, could record hints of fluid overpressure in garnet zonation⁸. Besides, the cation diffusivities (Fe-Mn-Mg-Ca) in garnet have been extensively calibrated by a variety of recent experimental and field-based studies^22,23. Garnet diffusion chronology based on these cations can resolve brief events^5,8. In this study, we employ garnet thermobarometry and diffusion chronology, in concert with absolute ages and isotopic analysis, to analyze the pressure variation of transient Cretaceous overprinting on Precambrian metapelites from the Northern Korean Peninsula (NKP; Fig. 1). These metapelites were intruded and heated by subduction-related andesitic porphyries at ca. 110 Ma (Fig. 2a). The resulting rapid dehydration led to new garnet growth with distinct mineral assemblages and chemical zonations^24,25, presenting a unique opportunity to test the magnitude and duration of dehydration-induced fluid overpressure. We reveal a brief (within 0.1‒0.3 kyr) thermal pulse reaching 7.5‒8.5 kbar and 600‒640 ℃ at ca. 110 Ma in the metapelites. The transient pressure disparity observed between this metamorphic overprinting and the andesitic porphyries nearby suggests dehydration-induced fluid overpressure on the order of several kilobars, an explanation that reasonably reconciles field observations, petrological and thermobarometric analysis, and is tested through our numerical simulation. Our results shed lights on the magnitude and timescale of dehydration-induced fluid overpressure, with significant implications for understanding of orogenic processes and seismogenic behaviors at plate boundaries.

Geological background. The NKP is a junction of the Paleoproterozoic Jiao-Liao-Ji, Silurian-Devonian Central Asian, and Cretaceous Western Pacific orogenic belts (Fig. 1a). The Jungsan Group in the northwestern part of the NKP is dominated by meta-sedimentary and meta-volcanic rocks that record Paleoproterozoic deformation and metamorphism during the Jiao-Liao-Ji orogeny (Fig. 1a‒b; ref. ²⁶). In the Cretaceous, extensive biotite and hornblende granites, and granitic to andesitic porphyries peaked at ca. 110 Ma intruded (Figs. 1c and 2a) in association with arc magmatism from the subduction of the Paleo-Pacific plate under the Northeast Asia (Fig. 1a; refs. ^26,27).

We collected the metapelites from the Jungsan Group exposed in a mountain ridge near the Chungsan County, east of Pyongyang (from N39°04′01.78″, E125°24′55.79″ to N39°04′28.78″, E125°25′54.39″; Figs. 1b and 2a‒b). About 100 m away from the locality, crops out a ~ 110 Ma andesitic porphyry (Fig. 2a and e; refs. ^26,27). The metapelite outcrop exhibits dense fractures almost traceable continuously cutting the mountain ridge (Fig. 2b–c). Some of these fractures are filled by veins consisting mainly of quartz and feldspar with minor garnet and staurolite, and cut the regional lineation defined by elongated quartz and kyanite (Fig. 2c‒d). These veins are more abundant close to the andesitic intrusions and appear to brecciate the metapelites (Fig. 2c). Previous studies on the metapelites identified two thermal events: Paleoproterozoic metamorphism (ca. 1850 Ma) of upper-amphibolite facies and a Cretaceous thermal overprint (ca. 110 Ma)^24,25. The first corresponds to the Jiao-Liao-Ji orogeny in association with the assembly of Columbia supercontinent^24–26; the P‒T (pressure and temperature) condition for the latter is estimated at ~ 610°C and ~ 7.9 kbar²⁵, and its nature remains to be resolved.

Petrography and garnet zonation. The metapelites are comprised of garnet and kyanite porphyroblasts in a matrix consisting of garnet in smaller size, biotite, staurolite, plagioclase, rutile, quartz, and ilmenite (Fig. 3a and g; Supplementary Fig. S1). The smaller garnet grains are highly subhedral and brecciated, and in some cases develop within staurolite in the domains where kyanite is absent (Fig. 3g; Supplementary Fig. S1A‒D). Locally, garnet rims are embayed by secondary muscovite and chlorite (Fig. 3a; Supplementary Fig. S1F). Mineral abbreviations throughout text follow the ref. ²⁸.

The garnet porphyroblasts exhibit crosscutting textures in compositional maps (Fig. 3b‒d and h‒j; Supplementary Table T1). The grains appear brecciated characterized by low-Ca bulk garnet domains crosscut by a complex network of high-Ca chemical dykes of 40‒100 µm widths (Fig. 3b and h), which are concordant with the P map (Fig. 3d). The bulk garnet domains have higher X_Mg (Mg/Mg + Fe²⁺) and lower water contents than the garnet chemical dykes, being 0.22‒0.24 vs. 0.16‒0.20 and 200‒240 vs. 260‒500 ppm, respectively (Fig. 3b and h; Supplementary Fig. S2‒S3). The compositional dikes do not correlate with cracks or high-Mn resorption rims (Fig. 3a‒c and g‒i).

The bulk garnet domains contain mineral inclusions of sillimanite, biotite, and quartz, with minor muscovite; the compositional dykes enclose mineral inclusions dominated by muscovite, biotite, staurolite, rutile, and quartz, with minor chlorite and margarite and no aluminosilicate (Fig. 3e‒f and k‒l). Notably, the chemical dyke domains are in broadly equilibrium contact with staurolite in microdomains where kyanite is absent (Fig. 3k‒l; Supplementary Fig. S1D). The high-Mn rims are closely associated with fine-grained secondary muscovite and chlorite (Fig. 3a‒b).

According to the petrographic features and compositional variations of garnet and other minerals (Supplementary Material A1 and Tables T1‒T6), we speculate that the bulk garnet domains record a preexisting background regional metamorphism while the garnet chemical dykes reflect a later thermal overprint. The bulk garnet domains and relict kyanite porphyroblasts may document the Paleoproterozoic high-grade metamorphism. On the other hand, the garnet chemical dykes are in equilibrium with a buffering mineral assemblage of staurolite, muscovite, biotite, rutile, quartz, and plagioclase²⁵, consistent with the dominant inclusions enclosed within garnet chemical dykes (Fig. 3e‒f and k‒l). Given the texture and higher water contents, these chemical dykes likely form via fluid-induced garnet fractures and a subsequent garnet growth that healed the fractures²⁹, similar to the fluid-assisted dissolution and reprecipitation³⁰.

Geochronology of the two thermal events and monazite O isotope. We use in situ zircon‒monazite U‒Pb geochronology, monazite O isotope, and zircon‒monazite‒garnet rare earth element (REE) analyses (Supplementary Tables T7‒T9) to examine the timing of garnet growth and the associated fluid source. Most zircon grains are unzoned to sector zoning in cathodoluminescence (CL) images, with low T/U ratios < 0.1 and flat heavy REE patterns (Supplementary Fig. S4 and Table T7), implying metamorphic origin in the presence of garnet³¹. These metamorphic zircon grains yield a consistent age of ca. 1850 Ma and Ti-in-zircon thermometric³² results of 680‒720°C (Fig. 3v). The REE patterns of these zircon grains are in broad equilibrium with bulk garnet domains (Fig. 3w). Thus, we interpret the bulk garnet recording Paleoproterozoic high-grade metamorphism (ca. 1850 Ma; Event 1).

Monazite grains exhibit irregular domains that are variably depleted in Th and enriched in Y, and thus appear brecciated in chemical maps (Fig. 3m‒u). Similar to garnet chemical dykes, these monazite domains are not related to any physical cracks, implying fluid-associated formation by pre-existing dissolutions/fractures and a subsequent growth³³. They consistently record an age of ca. 110 Ma, despite some high-Th, low-Y monazite relics preserve the age of ca. 1850 Ma (Fig. 3m, s, and x). The REE partitions of the low-Th, high-Y monazite domains suggest broad equilibrium with garnet chemical dykes at temperatures around 600°C (Fig. 3y), so they likely date the growth of garnet chemical dykes (ca. 110 Ma; Event 2). The δ¹⁸O values of ~ 1850 Ma and ~ 110 Ma monazite domains overlap (8.5‒9.5‰, Supplementary Fig. S4C). They are in the broad range of metamorphic fluids from medium to high-grade metapelites³⁴. Thus, the fluids that altered monazite and garnet are likely internally-derived.

Metamorphic P‒T conditions and timescales. We combine phase equilibria modeling and geothermobarometers to determine the P‒T conditions of the two metamorphic events. Given fluids being internally derived, we reasonably consider that the metapelites remain essentially a closed system. In this sense, for the Event 1, we calculate the phase equilibria diagram using measured bulk rock compositions (Supplementary Table T10). The bulk garnet compositions record peak P‒T conditions of ~ 8 kbar and ~ 690°C (Supplementary Material A2 and Fig. S5A‒B), agreeing with independent estimates from geothermobarometers (Supplementary Table T11) and the previous study²⁵. The retrograde P‒T path of the Event 1 was constrained as decompression cooling to < 4 kbar and < 500°C by ca. 1830 Ma (Supplementary Fig. S5A; refs. ^24,25.

For the overprinting event, we account the effect of garnet fractionation³⁵ and use measured bulk rock compositions with the bulk garnet domains removed to calculate the phase equilibria diagram (Fig. 4a; Supplementary Fig. S5C). The observed mineral assemblage is well predicted in the ms + bt + pl + grt + st + rt ± chl fields, and measured compositions of garnet chemical dykes and matrix plagioclase rims collectively define P‒T conditions at 7.5‒8.5 kbar and 600‒640°C (Fig. 4a; Supplementary Fig. S5D). The P‒T conditions agree well with estimates from the garnet-muscovite-plagioclase-quartz (GMPQ)³⁶ and garnet-biotite-plagioclase-quartz (GBPQ)³⁷ geothermobarometers (Fig. 4a; Supplementary Table T11). Minor garnet chemical dykes yielding lower temperatures (580‒600°C; Fig. 4a) reflect the retrograde metamorphism associated with the formation of fine-grained secondary muscovite and chlorite (Fig. 3a‒d). Moreover, we apply the quartz-in-garnet elastic barometry³⁸ that does not require chemical equilibrium to fully-enclosed quartz inclusions within garnet chemical dyke domains, yielding pressures of 8–9 kbar at 600°C (Fig. 4b; Supplementary Table T12), consistent with the phase equilibria modeling.

The sharp compositional discontinuity at the bulk garnet-dyke boundary reflects limited cumulative diffusion effects during the modification and subsequent cooling. Given temperatures above 600 ℃ recorded by the chemical dykes, the diffusion simulation employing the diffusion model²² and the effect of water content on diffusivities²³ yields timescales as short as 0.1‒0.3 kyr (Fig. 4c and Supplementary Fig. S2 and Material A2). The timescales are still considered as upper limit estimates: (1) any more gradual initial profiles than the step function profile used would yield shorter timescales⁵; (2) garnet water contents enhance the diffusivities²³, and thus using the minimum water content (200 ppm) overestimates the timescale; (3) the simulation was conducted at a conservative P‒T condition of 7.5 kbar and 600°C, and higher temperatures would yield shorter timescales. To assess the uncertainty of the best-fit timescales, we incorporate the main uncertainties from the diffusion model, temperature estimation, and garnet water content (see Supplementary Material A2 for details). The result indicates that the best-fit timescales have a relative uncertainty within ± 0.4 orders of magnitude (1σ), with the use of other diffusion models leading to shorter timescales. In sum, the thermal peak and the formation of garnet chemical dykes are brief within 0.1‒0.3 kyr at ca. 110 Ma.

Summary of metamorphic evolution. The bulk garnet domains record a peak metamorphism of ~ 8 kbar and ~ 690°C occurring at ca. 1850 Ma in response to the Paleoproterozoic orogeny, and the terrane subsequently exhumed to upper crustal levels before ca. 1830 Ma^24,25. The kyanite-bearing peak assemblage might be rehydrated during the Paleoproterozoic exhumation^24,25 or over long period residence at the upper crustal levels²⁶ (Fig. 5a‒b). The garnet chemical dykes intersecting the bulk garnet domains document a transient thermal event at ca. 110 Ma, with P‒T conditions of 7.5‒8.5 kbar and 600‒640°C that lasted for less than 0.1‒0.3 kyr (Fig. 4 and Supplementary Fig. S2). The Cretaceous thermal event modified garnet in parallel with the dehydration of muscovite, chlorite, margarite, and biotite, some of which are preserved as inclusions within garnet chemical dykes (Figs. 3e‒f, 3k‒l, and 5). The dehydration of these hydrous minerals at ca. 110 Ma could release significant amounts of fluids that contributed to the high H₂O contents and the complex network of fracture and healing textures observed in garnet chemical dykes (Fig. 5c).

Transient supralithostatic fluid overpressure at ca. 110 Ma. The brief nature of the Cretaceous thermal excursion requires a transient and local heat source^5,39, which is, arguably the most likely, the contemporaneous andesitic porphyritic intrusions nearby (Fig. 2a and 2e; Supplementary Fig. S1G‒H). We note that the converted depth of the thermal anomaly is significantly greater than common emplacement depths of porphyritic to granitic plutons in the region²⁶. The emplacement P‒T conditions of the ca. 110 Ma porphyritic intrusions are determined at 700‒760°C and 1‒3 kbar (Fig. 4a; Supplementary Table T11). Given the uncertainties (± 1‒2 kbar) in phase equilibria modeling and geothermobarometers⁴⁰, a pressure disparity at least of 4.5 ± 3.0 kbar exists between the metapelites and nearby porphyritic intrusions. Such a pressure discrepancy in the juxtaposed units indicates either rapid uplift of the metapelite in response to porphyritic intrusions, or transient deviation from lithostatic conditions.

The first scenario is impossible as the metapelites are not incorporated as xenoliths within the andesitic porphyries; rather, they are intruded by the porphyries, with quartz-feldspar veins prevalent near the intrusions (Fig. 2). Besides, thermochronology and P‒T path reconstructions suggest that the metapelites of the Jungsan Group had already cooled and been exhumed to < 4 kbar and < 500°C by ca. 1830 Ma^24,25. There is no record of tectonic burial reaching depths of 7.5‒8.5 kbar in the region from 1830 to 110 Ma²⁶. Moreover, such fast uplift (4.5 ± 3.0 kbar within hundreds of years) defies any observed or modeled plate convergence and subduction rates^41,42. On the other hand, rapid garnet growth associated with the short-lived pressure variation is consistent with rapid fluctuations in pore fluid pressure due to switches from locally undrained to drained conditions^8,18. Therefore, a non-lithostatic fluid overpressure model is more feasible.

We suggest that the garnet chemical dykes record supralithostatic fluid overpressure due to rapid dehydration in metapelites. Simulations by Auge et al. (1998)⁹ and Aarnes (2010)¹⁹ provide key insights into the process, showing that high-temperature intrusions can rapidly heat and dehydrate country rocks. If a large amount of fluid fails to dissipate, possibly impeded by low permeability of the rocks, localized high pore fluid pressure exceeding lithostatic load can be generated^9,19. The structures of the outcrops and petrographic characteristics align well with the setup in these conceptual simulations. The ca. 110 Ma andesitic porphyries provide enough heat to drive a significant temperature overstepping for dehydration reactions in the metapelites. The metapelites have low permeability due to their deep burial by the Paleoproterozoic orogeny, and contain abundant hydrous minerals likely resulted from the rehydration during the Paleoproterozoic exhumation (Fig. 5a‒b). These metapelites thus could have acted as a source of rapid fluid generation near the intrusions, while those colder rocks further away may act as seals, thus allowing fluid accumulation and eventually overpressure. The brittle fractures or brecciation filled by quart- and feldspar-rich veins (Fig. 2b‒d) are suggestive of supralithostatic fluid overpressure and hydraulic fracturing^14,43. The initial fluids escaped may have fractured garnet, and the further pore fluid overpressure commenced garnet growth and healed the fractures (Fig. 5c). Matrix minerals may have experienced similar fluid-induced fractures, but the evidence is prone to recrystallization²⁰.

Such an ‘autoclave’ scenario^2,5 builds up high fluid pressures in a confined volume. For an independently test, we conduct fluid overpressure simulation using the equations of state and thermodynamic dataset of Holland and Powell (2011) (ds62⁴⁴) and the measured bulk-rock compositions of the metapelites. We assume that the surrounding metapelites are rigid and impermeable so that fluids released by dehydration reactions cannot escape and are trapped in a confined space (isochoric reaction), similar to the ‘autoclave’ geometry^2,5. The resulting elastic compressibility leads to a deviation from the lithostatic pressure and generates fluid overpressure. We model the dehydration of the metapelites through isobaric heating (500 to 600°C) under different pressures (1 − 4 kbar) to reflect the heating from the porphyry intrusions. The calculation methods are detailed in the refs. ^2,5. The degree of fluid overpressure largely depends on the volumes of dehydrated hydrous minerals such as chlorite and muscovite, and changes with the pressures for the isobaric heating (Fig. 6). We find that, at pressures equivalent to the emplacement depths of porphyry intrusions (1 − 2 kbar), high fluid overpressure of 4.5 ± 3 kbar would readily generate even if the confined space has an unrealistically high relative volume expansion (> 80%; Fig. 6). Besides, the heating to 600°C is a conservative estimate and a higher degree of heating would produce greater fluid overpressure⁵. Therefore, to achieve the pressure contrast of 4.5 ± 3 kbar, the metapelites do not need to be perfectly impermeable or rigid, and the dehydration do not require being isochoric. The simulation is a semi-quantitative analysis; nevertheless, the results represent a first-order test and are insensitive to the details of simulation settings^2,5.

In sum, the brief pressure disparity recorded by the metapelites and responsible andesitic porphyries nearby could be reasonably reconciled by transient supralithostatic fluid overpressure generated by rapid dehydration of the metapelites. This brief pressure contrast was followed by rapid brittle hydraulic fracturing and vein formation which released the high pore fluid pressure.

Implications. The growth of garnet and its compositions have been widely used as reliable indicators of burial/exhumation depths to constrain crustal evolution in a wide variety of settings³. The underlying assumption is that the formation of garnet occurs under lithostatic pressures, enabling the direct conversion of thermodynamic pressures derived from garnet compositions into depths^1,3. However, our findings present another example that challenges this paradigm and show that garnet growth could reflect supralithostatic fluid overpressure of several kilobars due to rapid dehydration in low permeability rocks. Previous numerical models have explored the potential of dehydration-induced fluid overpressure^9,10,16–18, yet our study is the first to test this mechanism both on magnitude and timescale in natural metamorphic records, with a specific focus on garnet. Our findings also suggest that while rocks typically do not possess unconfined strength exceeding 100‒200 MPa, the presence of confinement and undrained conditions may have allowed for the development of greater fluid overpressure⁸.

Dehydration reactions that form garnet typically release significant amounts of fluids within narrow P‒T ranges^8,21, and most metamorphic rocks exhibit low permeability^9,10,43. Pulsed dehydration and rapid garnet growth (< 1 Myr) also have been reported from many tectonic zones worldwide, such as the Sifnos accretionary complex⁴⁵, the Alps belt^29,46, the Tianshan subduction belt⁴⁷, the Almagro accretionary complex²⁰, and the Franciscan Complex in California⁸. Some garnets in these regions also exhibit similar fracture and healing textures (e.g., refs. ^20,46), raising the possibility that rapid garnet growth driven by supralithostatic fluid overpressure might be more prevalent than previously recognized. The potential deviation from lithostatic conditions and resulting misinterpreted tectonic reconstruction require more attention or reevaluation.

Furthermore, our research suggests that a favorable situation for dehydration-induced fluid overpressure involves younger thermal pulses superimposed on older metamorphic belts. Given the proximity of many younger orogenies to or originating from ancient orogenic belts^46,48, the associated superimposed metamorphism and rapid garnet growth may record varying levels of fluid overpressures. This effect could potentially lead to an overestimation of burial depths in younger orogens, and might have contributed to the apparent increase in peak pressures observed from Precambrian to present orogens⁴⁹. The presence of dehydration-induced fluid overpressure, in conjunction with stress- and melting-induced pressure variations^2,4–6,8, thus challenges current orogenic reconstructions based on the lithostatic assumption.

The fluid overpressure generated by rapid dehydration of metapelites persisted for several hundred years and was followed by transient brittle failure and hydrofracturing (Figs. 2 and 4c). Many earthquake nucleation models consider dehydration-induced fluid overpressure^13–15, but the mechanism responsible for generating such high pore fluid pressures over timescales relevant to the recurrence of significant earthquakes (100‒1000 years) remains untested in nature. We use garnet diffusion chronology to quantify the duration of dehydration-induced fluid overpressure, and show that dehydration reactions stimulated by rapid heating could produce such transient fluid overpressure. This process may be critical for fault weakness and seismic initiation in thrust or normal fault settings, where rapid temperature increases and fast dehydration of hydrous minerals similar to pelitic rocks in this study could occur^9,16. The brief brittle failure following fluid overpressure may trigger localized low-frequency tectonic tremors^14,15,20. Given the regional distribution of metapelites and subduction-related magmatic intrusions at ca. 110 Ma in the Korea (Figs. 1b–c and 2a), it is conceivable that the localized dehydration-induced seismic events may have accumulated to a larger-scale earthquake and contributed to the development of contemporaneous fault systems and basin evolution in the region²⁶.

CRediT author statement

Yi Zou: Conceptualization, Formal analysis, Visualization, and Writing – original draft, review & editing. Xu Chu: Conceptualization, Methodology, Software, and Writing – original draft, review & editing. Ross N. Mitchell: Conceptualization, Visualization, and Writing – review & editing. Lei Zhao: Data curation, Methodology, and Funding acquisition. Qiuli Li: Data curation, Funding acquisition, and Project administration. Mingguo Zhai: Project administration and Funding acquisition.

ACKNOWLEDGEMENT

This study was supported by the National Natural Science Foundation of China (grants 42220104008, 2023YFF0803802, and 42250202) and the Connaught Research Award administered by the University of Toronto. We appreciate kind help from Dr. Mengdie Chen during sample preparations.

X-ray mapping and mineral compositions. X-ray mapping of garnet and monazite was acquired using microprobe CAMECA SXFiveFE at the Institute of Geology and Geophysics, Chinese Academy of Sciences (IGGCAS). A large LiF spectroscopic crystal was used, and the operating conditions were 20 kV, 100 nA, and 50 ms dwell time. Mineral compositions were analyzed using JEOL JXA-8100 electron microprobe at the IGGCAS and JEOL JXA-6230 at the University of Toronto. The operating conditions were 10 nA or 50 nA (garnet) beam current, 15 kV accelerating voltage, 1‒3 μm spot diameter or focused spot, and 10‒30 s counting time. Mineral standards used were well-defined natural and synthetic minerals. Data were processed with the online ZAF-type correction.

Fourier-transform infrared spectroscopy. The other halves of the representative garnets (Fig. 3a‒d and g‒j) were doubly polished for Fourier transform infrared (FTIR) analysis. Detailed analytical procedures followed the ref.²³. Each section was heated at 200 °C for 36 hours in a dry environment to remove surface absorbed water. The unpolarized FTIR analysis was performed using a Bruker Vertex 70V FTIR spectrometer coupled with a Hyperion 2000 microscope at the IGGCAS. Each spectrum has 128 scans, 4 cm^-1 resolution and an aperture size of 50 × 50 μm. The FTIR analysis was conducted on optically clean garnet domains to avoid any mineral inclusions. We employed the calibration⁵³ to calculate the water contents in garnet:

, where H(x) is the absorption coefficient at given wavenumber x, B is the density factor, and A is an orientation factor set to 1/3. The integration was calculated at a wavenumber range of 2800 to 3750 cm⁻¹, according to the ref.²³. Water contents measured for bulk garnet and chemical dyke domains are 200‒240 and 260‒500 ppm, respectively (Fig. 3b and h), with a relative uncertainty within ~15 % (1σ; ref.²³).

In situ U-Th-Pb-O isotope and trace element analysis. The metapelites were crushed via high voltage of 120 kV pulsing to a grain size of ~300 μm using SelFrag High Voltage Pulse Fragmentor at China University of Geoscience, Wuhan. This helps to produce a high yield of intact zircon and monazite grains. Monazite and zircon grains were then collected by heavy liquid and magnetic separation techniques, and were further handpicked by binocular microscope. The collected monazite grains were mounted in an epoxy resin, along with standard monazites RW-1⁵⁴ and 44069⁵⁵. The same procedures were applied to collected zircon grains, along with standard zircons 91500⁵⁶, Plešovice⁵⁷, and GJ-1⁵⁸. Monazite grains were imaged by back scattered (BSE) images and X-ray mapping, and zircon grains were characterized by cathodoluminescence (CL) images, using the CAMECA SXFiveFE at the IGGCAS.

In situ monazite O isotope was measured using the secondary ion mass spectrometry (Cameca IMS-1280 SIMS) at the IGGCAS. The primary Cs⁺beam was accelerated at 10kV and focused in Gaussian mode, with a beam intensity of ~2.0 nA and a spot size of 30 µm. We used the standard monazites Manangoutry and RW-1⁵⁴ for instrumental mass fractionation corrections, and employed the standard monazite 44069⁵⁵ as unknowns to check instrumental stability. The detailed analytical procedures, conditions, and matrix effect calibration refer to the ref. ⁵⁹.

Measurement of monazite and zircon U-Th-Pb isotopes and trace element concentrations was performed using an Agilent 7500a Q-ICP-MS equipped with a 193 nm laser ablation system (LA-ICP-MS) at the IGGCAS. Detail analytical methods followed the ref. ⁵⁸. For monazite U-Th-Pb isotopes, the analytical conditions were 40 μm spot size, 10 Hz laser repetition, and 25J/cm² energy density. The analyzed spots are the same as or overlap on those domains used for the O isotope analysis. For zircon U-Th-Pb isotopes, the analytical conditions were changed to 40 μm spot size, 5 Hz laser repetition, and 80J/cm² energy density. We used standard monazites RW-1⁵⁴ to correct U-Th-Pb fractionation and instrumental mass discrimination, and employed the standard monazite 44069⁵⁵ as unknowns to check instrumental stability. Similarly, we used the standard zircon 91500⁵⁶ to calibrate the U-Th-Pb isotopes of zircon grains, and used the standard zircons Plešovice⁵⁷ and GJ-1⁵⁸ as unknowns. For trace element determination, we used standard glasses Nist 610 and 612 for calibration, and the internal standard elements were Si for zircon and P for monazite, respectively. In addition, we used the above LA-ICP-MS equipment to determine garnet rare earth elements following the same analytical protocols and standard glasses with those for zircon. Raw data reduction was processed using the Glitter 4.0 program (Macquarie University). Age calculation by the software Isoplot was quoted with 95% confidence.

Bulk-rock composition analysis. Bulk-rock compositions of representative metapelites were analyzed using XRF-1500 X-ray fluorescence spectrometry (XRF) at the IGGCAS. The analytical precision is better than 5% calibrated by international rock standards (USGS) AGV-1. The FeO content was analyzed by a conventional titration method with potassium permanganate as an oxidant. The measured bulk-rock compositions were converted into mol. % and used for phase equilibria modeling.

Phase equilibria modeling and geothermobarometry. We used phase equilibria modeling (pseudosection) and geothermobarometry to determine metamorphic P‒T conditions. Phase equilibria modeling was calculated in chemical system MnO-Na₂O-CaO-K₂O-FeO-MgO-Al₂O₃-SiO₂-H₂O-TiO₂-O₂ (MnNCKFMASHTO) using the computer software Perple_X⁶⁰ (version 6.90) and the thermodynamic dataset ds62⁴⁴. Mineral solution models for garnet, biotite, muscovite, staurolite, chlorite, plagioclase, cordierite, paragonite, margarite, and ilmenite followed the refs. ^61,62. Pure phases are aluminosilicate, rutile, H₂O, and quartz. The H₂O and quartz were set to excess given the widespread presence of hydrous minerals and quartz. The Fe³⁺/Fe²⁺ ratio was determined by the chemical titration. We used the measured bulk-rock compositions by X-ray fluorescence to calculate an initial phase diagram. For the late thermal event recorded by garnet chemical dikes, phase equilibria modeling was calculated using effective bulk compositions³⁵ estimated by subtracting bulk garnet domains from the measured bulk-rock compositions. We used the software ImageJ and thin section scanning and X-ray mapping images (Fig. 3b; Supplementary Fig. S1) to determine the volume percentages of whole garnet grains and the volume proportion of bulk garnet domains assuming no density contrasts between minerals.

For the Cretaceous thermal overprint, we used the garnet-muscovite thermometer and garnet-muscovite-plagioclase-quartz barometer (GMPQ)³⁶ applied to garnet chemical dykes, muscovite inclusions within the chemical dykes, and the rims of plagioclase; and the GBPQ geothermobarometers³⁷ applied to garnet chemical dykes, and rims of matrix biotite and plagioclase. Additionally, we employed the quartz-in-garnet elastic barometry provided in the Excel spreadsheet³⁸ to independently determine the pressure conditions during the growth of garnet chemical dykes. We selected fully-enclosed quartz inclusions in a diameter less than 30 µm within garnet chemical dykes, and avoided quartz inclusions that are exposed to the surfaces of thin sections or surrounded by micro-cracks. Raman spectra of a stress-free quartz reference (464 cm^-1) and quartz inclusions within garnet chemical dykes were acquired using the laser Raman spectrometer HORIBA JOBIN YVON LabRam HR hosted in the IGGCAS. The uncertainty for the Raman band (464 cm^-1) estimated by a least-square fitting is around 0.2 cm^-1 (1σ), which is sufficient to resolve the wavenumber shifts of the quartz inclusions at the level of mostly >1 cm^-1. The detailed procedures and methods followed those described by the previous studies^7,38. Besides, we applied Ti-in-amphibole thermometer⁶³ and Al-in-amphibole barometer⁶⁴ to amphibole grains from the ca. 110 Ma andesitic porphyry intrusions, to determine the crystallization P‒T conditions of the magmas.

Garnet multicomponent diffusion modeling. We employed garnet multicomponent (Ca‒Mn‒Mg‒Fe) diffusion modeling to fit the compositional gradient between garnet chemical dykes and bulk garnet domains. The compositional gradient reflects cumulative diffusion effects during the growth of garnet chemical dykes and cooling below an effective closure temperature for diffusion. Given nearly all garnet chemical dykes record temperature above 600 °C, we ran the simulation at a conservative P‒T condition of 7.5 kbar and 600 °C. The diffusion model²², which considers the effects of temperature, pressure, unit-cell dimension, and oxygen fugacity (fO₂), was used to compute the self-diffusion coefficients of the divalent ions. We incorporated the effect of garnet water content on diffusivities of the divalent ions based on the recent experiment data²³, and used the minimum measured garnet water content (200 ppm). Garnet was considered as an ideal solid solution, and the diffusion coefficient matrix was calculated using the formalism⁶⁵. Garnet was assumed to have center cut with a sphere geometry. Diffusion equations in the sphere geometry were numerically solved using a Nicolson-Crank finite difference scheme (Matlab code from the ref. ²²). We considered convolution effects in profile fitting to account for the spacing averaging of microprobe analyses (convolution factor ε=0.48 μm; ref. ⁶⁶). An initial step-function profile was assumed from which relaxation by intracrystalline diffusion was modeled for a range of timescales. The fitness between modeled and measured profiles was evaluated by calculating the root mean square error (RMSE), and the timescales that yield the minimum RMSE were taken as the best-fit results.

Moulas, E., Schmalholz, S. M., Podladchikov, Y., Tajcmanova, L., Kostopoulos, D. & Baumgartner, L. Relation between mean stress, thermodynamic, and lithostatic pressure. J Metamorph Geol 37, 1-14 (2019).
Luisier, C., Baumgartner, L., Schmalholz, S. M., Siron, G. & Vennemann, T. Metamorphic pressure variation in a coherent Alpine nappe challenges lithostatic pressure paradigm. Nat Commun 10, 4734 (2019).
Caddick, M. J. & Kohn, M. J. Garnet: Witness to the Evolution of Destructive Plate Boundaries. Elements 9, 427-432 (2013).
Tajčmanová, L., Manzotti, P. & Alvaro, M. Under Pressure: High-Pressure Metamorphism in the Alps. Elements 17, 17-22 (2021).
Chu, X., Ague, J. J., Podladchikov, Y. Y. & Tian, M. Ultrafast eclogite formation via melting-induced overpressure. Earth Planet Sc Lett 479, 1-17 (2017).
Plümper, O.. et al. High-magnitude stresses induced by mineral-hydration reactions. Geology 50, 1351-1355 (2022).
Zhong, X., Petley-Ragan, A. J., Incel, S. H. M., Dabrowski, M., Andersen, N. H. & Jamtveit, B. Lower crustal earthquake associated with highly pressurized frictional melts. Nat Geosci 14, 519-525 (2021).
Viete, D. R.. et al. Metamorphic records of multiple seismic cycles during subduction. Sci Adv 4, eaaq0234 (2018).
Ague, J. J., Park, J. & Rye, D. M. Regional metamorphic dehydration and seismic hazard. Geophys Res Lett 25, 4221-4224 (1998).
Wang, W. H. & Wong, T. F. Effects of reaction kinetics and fluid drainage on the development of pore pressure excess in a dehydrating system. Tectonophysics 370, 227-239 (2003).
Peacock, D. C. P., Anderson, M. W., Rotevatn, A., Sanderson, D. J. & Tavarnelli, E. The interdisciplinary use of “overpressure”. J Volcanol Geoth Res 341, 1-5 (2017).
Wang, X. L.. et al. Fluid inclusion evidence for extreme overpressure induced by gas generation in sedimentary basins. Geology 50, 765-770 (2022).
Saffer, D. M. & Tobin, H. J. Hydrogeology and Mechanics of Subduction Zone Forearcs: Fluid Flow and Pore Pressure. Annu Rev Earth Pl Sc 39, 157-186 (2011).
Giuntoli, F.. et al. Methane-hydrogen-rich fluid migration may trigger seismic failure in subduction zones at forearc depths. Nat Commun 15, 480 (2024).
Tarling, M. S., Smith, S. A. F. & Scott, J. M. Fluid overpressure from chemical reactions in serpentinite within the source region of deep episodic tremor. Nat Geosci 12, 1034-1042 (2019).
Fulton, P. M., Saffer, D. M. & Bekins, B. A. A critical evaluation of crustal dehydration as the cause of an overpressured and weak San Andreas Fault. Earth Planet Sc Lett 284, 447-454 (2009).
Connolly, J. A. D. The Mechanics of Metamorphic Fluid Expulsion. Elements 6, 165-172 (2010).
Llana-Funez, S., Wheeler, J. & Faulkner, D. R. Metamorphic reaction rate controlled by fluid pressure not confining pressure: implications of dehydration experiments with gypsum. Contrib Mineral Petr 164, 69-79 (2012).
Aarnes, I. Sill emplacement and contact metamorphism in sedimentary basins. Local processes with global implications. In: Faculty of Mathematics and Natural Sciences). University of Oslo (2010).
Angiboust, S., Ioannidi, P. I. & Muldashev, I. Garnet fracturing reveals ancient unstable slip events hosted in plate interface metasediments. Earth Planet Sc Lett 640, 118794 (2024).
Baxter, E. F. & Caddick, M. J. Garnet growth as a proxy for progressive subduction zone dehydration. Geology 41, 643-646 (2013).
Chu, X. & Ague, J. J. Analysis of experimental data on divalent cation diffusion kinetics in aluminosilicate garnets with application to timescales of peak Barrovian metamorphism, Scotland. Contrib Mineral Petr 170, 1-27 (2015).
Zhang, B. H., Li, B. W., Zhao, C. C. & Yang, X. Z. Large effect of water on Fe-Mg interdiffusion in garnet Earth Planet Sc Lett 505, 20-29 (2019).
Li, Q. L., Zhao, L., Zhang, Y. B., Yang, J. H., Kim, J. N. & Han, R. H. Zircon-titanite-rutile U-Pb system from metamorphic rocks of Jungshan "Group" in Korea: Implications of tectono-thermal events from Paleoproterozoic to Mesozoic. Acta Petrol Sin 32, 3019-3032 (2016).
Chen, M. D.. et al. Zircon and monazite petrochronology studies reveal Mesozoic amphibolite-facies metamorphism overprinting Paleoproterozoic granulite-facies metamorphism: A case study from Rangnim Massif, North Korea. Lithos 398, (2021).
Zhai, M. G.. et al. The geology of North Korea: An overview. Earth-Sci Rev 194, 57-96 (2019).
Wu, F. Y., Han, R. H., Yang, J. H., Wilde, S. A., Zhai, M. G. & Park, S. C. Initial constraints on the timing of granitic magmatism in North Korea using U-Pb zircon geochronology. Chem Geol 238, 232-248 (2007).
Whitney, D. L. & Evans, B. W. Abbreviations for names of rock-forming minerals. American Mineralogist 95, 185-187 (2010).
Broadwell, K. S., Locatelli, M., Verlaguet, A., Agard, P. & Caddick, M. J. Transient and periodic brittle deformation of eclogites during intermediate-depth subduction. Earth Planet Sc Lett 521, 91-102 (2019).
Putnis, A. & John, T. Replacement Processes in the Earth's Crust. Elements 6, 159-164 (2010).
Rubatto, D. Zircon: The Metamorphic Mineral. Reviews in Mineralogy and Geochemistry 83, 261-295 (2017).
Ferry, J. M. & Watson, E. B. New thermodynamic models and revised calibrations for the Ti-in-zircon and Zr-in-rutile thermometers. Contrib Mineral Petr 154, 429-437 (2007).
Harlov, D. E. & Hetherington, C. J. Partial high-grade alteration of monazite using alkali-bearing fluids: Experiment and nature. American Mineralogist 95, 1105-1108 (2010).
Rubatto, D., Putlitz, B., Gauthiez-Putallaz, L., Crépisson, C., Buick, I. S. & Zheng, Y.-F. Measurement of in-situ oxygen isotope ratios in monazite by SHRIMP ion microprobe: Standards, protocols and implications. Chem Geol 380, 84-96 (2014).
Lanari, P. & Engi, M. Local Bulk Composition Effects on Metamorphic Mineral Assemblages. Reviews in Mineralogy and Geochemistry 83, 55-102 (2017).
Wu, C. M. & Zhao, G. C. Recalibration of the garnet-muscovite (GM) geothermometer and the garnet-muscovite-plagioclase-quartz (GMPQ) geobarometer for metapelitic assemblages. J Petrol 47, 2357-2368 (2006).
Wu, C. M., Zhang, J. & Ren, L. D. Empirical garnet-biotite-plagioclase-quartz (GBPQ) geobarometry in medium- to high-grade metapelites. J Petrol 45, 1907-1921 (2004).
Zhong, X.. et al. The geological application of elastic geo-thermobarometry:Example of quartz-in-garnet (QuiG)barometry. Acta Petrol Sin 38, 2933-2948 (2022).
Viete, D. R. & Lister, G. S. On the significance of short-duration regional metamorphism. J Geol Soc London 174, 377-392 (2017).
Palin, R. M., Weller, O. M., Waters, D. J. & Dyck, B. Quantifying geological uncertainty in metamorphic phase equilibria modelling; a Monte Carlo assessment and implications for tectonic interpretations. Geoscience Frontiers 7, 591-607 (2016).
Gerya, T. V., Stockhert, B. & Perchuk, A. L. Exhumation of high-pressure metamorphic rocks in a subduction channel: A numerical simulation. Tectonics 21, (2002).
Agard, P., Yamato, P., Jolivet, L. & Burov, E. Exhumation of oceanic blueschists and eclogites in subduction zones: Timing and mechanisms. Earth-Sci Rev 92, 53-79 (2009).
Norris, R. J. & Henley, R. W. Dewatering of a Metamorphic Pile. Geology 4, 333-336 (1976).
Holland, T. J. B. & Powell, R. An improved and extended internally consistent thermodynamic dataset for phases of petrological interest, involving a new equation of state for solids. J Metamorph Geol 29, 333-383 (2011).
Dragovic, B., Baxter, E. F. & Caddick, M. J. Pulsed dehydration and garnet growth during subduction revealed by zoned garnet geochronology and thermodynamic modeling, Sifnos, Greece. Earth Planet Sc Lett 413, 111-122 (2015).
Herwartz, D., Nagel, T. J., Munker, C., Scherer, E. E. & Froitzheim, N. Tracing two orogenic cycles in one eclogite sample by Lu-Hf garnet chronometry. Nat Geosci 4, 178-183 (2011).
Tan, Z., Agard, P., Gao, J., Hong, T. & Wan, B. Concordant pulse in Mn, Y and HREEs concentrations during UHP eclogitic garnet growth: Transient rock dynamics along a cold subduction plate interface. Earth Planet Sc Lett 530, 115908 (2020).
Zhang, C.. et al. Subduction re-initiation at dying ridge of Neo-Tethys: Insights from mafic and metamafic rocks in Lhaze ophiolitic melange, Yarlung-Tsangbo Suture Zone. Earth Planet Sc Lett 523, (2019).
Brown, M., Johnson, T. E. & Spencer, C. J. Secular changes in metamorphism and metamorphic cooling rates track the evolving plate tectonic regime on Earth. Journal of the Geological Society, London 179, jgs2022-2050 (2022).
Jiao, S. J., Guo, J. H., Harley, S. L. & Peng, P. Geochronology and trace element geochemistry of zircon, monazite and garnet from the garnetite and/or associated other high-grade rocks: Implications for Palaeoproterozoic tectonothermal evolution of the Khondalite Belt, North China Craton. Precambrian Res 237, 78-100 (2013).
Hermann, J. & Rubatto, D. Relating zircon and monazite domains to garnet growth zones: age and duration of granulite facies metamorphism in the Val Malenco lower crust. J Metamorph Geol 21, 833-852 (2003).
Hacker, B., Kylander‐Clark, A. & Holder, R. REE partitioning between monazite and garnet: Implications for petrochronology. J Metamorph Geol 37, 227-237 (2019).
Paterson, M. S. The Determination of Hydroxyl by Infrared-Absorption in Quartz, Silicate-Glasses and Similar Materials. B Mineral 105, 20-29 (1982).
Ling, X. X.. et al. Monazite RW-1: a homogenous natural reference material for SIMS U-Pb and Th-Pb isotopic analysis. Miner Petrol 111, 163-172 (2017).
Aleinikoff, J. N.. et al. Deciphering igneous and metamorphic events in high-grade rocks of the Wilmington Complex, Delaware: Morphology, cathodoluminescence and backscattered electron zoning, and SHRIMP U-Pb geochronology of zircon and monazite. Geol Soc Am Bull 118, 39-64 (2006).
Wiedenbeck, M.. et al. Further characterisation of the 91500 zircon crystal. Geostand Geoanal Res 28, 9-39 (2004).
Slama, J.. et al. Plesovice zircon - A new natural reference material for U-Pb and Hf isotopic microanalysis. Chem Geol 249, 1-35 (2008).
Xie, L. W., Zhang, Y. B., Zhang, H. H., Sun, J. F. & Wu, F. Y. In situ simultaneous determination of trace elements, U-Pb and Lu-Hf isotopes in zircon and baddeleyite. Chinese Sci Bull 53, 1565-1573 (2008).
Wu, L.-G.. et al. Rapid and accurate SIMS microanalysis of monazite oxygen isotopes. J Anal Atom Spectrom 35, 1607-1613 (2020).
Connolly, J. A. D. Computation of phase equilibria by linear programming: A tool for geodynamic modeling and its application to subduction zone decarbonation. Earth and Planetary Science Letters 236, 524-541 (2005).
White, R. W., Powell, R. & Johnson, T. E. The effect of Mn on mineral stability in metapelites revisited: new a-x relations for manganese-bearing minerals. J Metamorph Geol 32, 809-828 (2014).
White, R. W., Powell, R., Holland, T. J. B., Johnson, T. E. & Green, E. C. R. New mineral activity-composition relations for thermodynamic calculations in metapelitic systems. J Metamorph Geol 32, 261-286 (2014).
Liao, Y., Wei, C. & Rehman, H. U. Titanium in calcium amphibole: Behavior and thermometry. American Mineralogist 106, 180-191 (2021).
Schmidt, M. W. Amphibole Composition in Tonalite as a Function of Pressure - an Experimental Calibration of the Al-in-Hornblende Barometer. Contrib Mineral Petr 110, 304-310 (1992).
Lasaga, A. C. MULTICOMPONENT EXCHANGE AND DIFFUSION IN SILICATES. Geochim Cosmochim Ac 43, 455-469 (1979).
Borinski, S. A., Hoppe, U., Chakraborty, S., Ganguly, J. & Bhowmik, S. K. Multicomponent diffusion in garnets I: general theoretical considerations and experimental data for Fe-Mg systems. Contrib Mineral Petr 164, 571-586 (2012).

There is NO Competing Interest.

SupplementaryTablesT1T12.xlsx
Supplementary Tables T1-T12
SupplementaryMaterials.docx

Download PDF

Version 1

posted

You are reading this latest preprint version

Dehydration-Induced Supralithostatic Fluid Overpressure in Metapelites from Northern Korea

Status:

Version 1

Abstract

Figures

Introduction

Results

DISCUSSION

Declarations

CRediT author statement

ACKNOWLEDGEMENT

METHODS

References

Additional Declarations

Supplementary Files

Status:

Version 1