Introduction

Terranes are discrete areas of crust with distinct tectonostratigraphic characteristics, separated by shear zones1. The crust enclosed within the bounding shear zones represents disparate continental lithosphere brought together with other distinct crustal elements only later in its evolution. The recognition of terranes is critical in understanding the geological evolution of a region. In particular, Archean terranes may provide fundamental evidence for the switch from dominantly vertical (i.e., stagnant-lid) to horizontal (plate) tectonic processes. One such area where the terrane concept has been highly important in developing models of crustal evolution is in the North Atlantic Craton of West Greenland. Later in the history of an assembled mosaic of terranes, far-field effects of distant plate reorganizations may be important. Such distal forces may reactivate ancient favorably orientated fabrics and also generate new structures. The effects of intra-continental tectonics could be expected in all terranes that went through a mobile-lid stage, regardless of their age. This process is examined in the Archean Akia Terrane.

Mylonites reflect zones of high-strain rate where dominantly ductile deformation at variable temperatures has been localized either along terrane boundaries or along later dislocations within a single terrane. These zones of ductile flow may thus provide key evidence for the kinematics and structure of an orogen and help with the understanding of far-field effects in stable crust. Accurate dating of mylonitization has been a repeated goal in the geosciences, with a wide range of approaches including, among others, thin slab whole rock Rb–Sr dating2, K-Ar microstructurally-constrained and bulk fault gouge dating3, and Ar-Ar dating of neocrystallized fault related phyllosilicates4. However, in some cases results have been difficult to interpret, due to the heterogeneous composition of mylonites containing both deformed relict porphyroclasts, as well as newly crystallized grains. Furthermore, mylonites provide pathways that localize fluids and secondary alteration. As each of the aforementioned geochronological techniques require sample disaggregation, inadvertent physical mixtures of different mineral generations can yield ambiguous mixed ages that fail to accurately capture the timing of geological events, including shearing. Consequently, mylonite dating has been described as one of the most difficult problems in geochronology5.

Southern West Greenland hosts the North Atlantic Craton, retaining a crustal history from at least 3.8 Ga. This region represents one of the largest, best exposed, tracts of Archean crust on Earth6. The craton extends west into Canada7 and east into southeastern Greenland8 and north western Scotland9. In Greenland, the North Atlantic Craton comprises several terranes, including the Akia Terrane, north of Nuuk city, which is a volumetrically large component and cut by various mylonites (Fig. 1). The Akia Terrane consists mainly of igneous rocks which crystallized at c. 3230–3190 Ma and 3070–2970 Ma1,10,11, with the latter being much more voluminous. This c. 3000 Ma Mesoarchean magmatism was coeval with high temperature, low pressure granulite-facies metamorphism12,13. On the northern side of the Akia Terrane 2877–2857 Ma supracrustal rocks are preserved within the Alanngua Complex14. These supracrustal rocks were deposited onto a Mesoarchean basement15, buried to >30 km depth, and partially melted, during regional high-grade metamorphism between 2857 and 2700 Ma16. Apatite Sm–Nd and Lu–Hf isochrons at c. 2700 Ma from the Akia Terrane, indicate pervasive isotopic re-equilibration also occurred during this event17. This Neoarchean metamorphism, which occurred alongside the Akia Terrane margin, is akin to that in adjoining terranes, indicating a shared history of thermal disturbance. Furthermore, such commonality implies these crustal elements were linked at least by this time14,15. Subsequent lower grade (greenschist- to amphibolite facies) c. 2630 Ma and 2540 Ma metamorphism is evidenced by metamorphic zircon overgrowths and neoblastic apatite18 and titanite19, respectively. Foundational Rb–Sr work in the 1970’s and 1980’s in the North Atlantic Craton was interpreted to reflect loss of radiogenic-Sr in biotite from regions south of the Akia Terrane, including the Godthaabsfjord and Isukasia areas at 1700–1600 Ma20,21. It is unclear whether these results reflect any new fabric generation or solely isotopic resetting.

Fig. 1: Schematic geological map of South West Greenland and geographic context of the study region.
figure 1

A Lithological units within the Nuuk-Maniitsoq region14,69. Blue dashed lines are major faults inferred from aerial photography. Yellow star symbol denotes sample locations. B Major lithotectonic units/terranes in the region.

Full size image

Here with the use of thermal history modeling, we seek to link isotopic ages of minerals in the Archean Akia Terrane of South West Greenland to their growth mechanism and ultimately date structures. We present a case study from four adjacent, subparallel, mylonite zones, linked to NE-SW trending sub-vertical terrane boundary parallel faults. We employ in situ biotite Rb–Sr geochronology via laser ablation triple quadrupole inductively coupled plasma mass spectrometry (QQQ-ICPMS) that permits the efficient direct analysis of texturally controlled phyllosilicate minerals linked to the mylonite fabric. We supplement this isotopic data with titanite U–Pb geochronology to evaluate the higher temperature history of these rocks and integrate these results to develop a coherent thermal evolution model to address the timing of mylonite formation. Our findings highlight the effects of the c. 1910–1770 Ma Nagssugtoqidian Orogeny, south of the inferred orogenic front, and caution about the assumption that all structures are Archean in this region.

Results

Four samples along regionally important NE-SW striking shear zones and one sample from an unfoliated granite were selected for geochronology (Fig. 1). All samples were analyzed directly on standard petrographic polished thin sections (Fig. 2).

Fig. 2: False color mineral maps of Akia Terrane mylonites.
figure 2

Tescan TIMA mineral identification images of thin sections from various mylonites in tonalite and felsic gneiss, Akia Terrane, Greenland. The black bar in each thin section image denotes a 10 mm scale.

Full size image

Rb–Sr

Sample 1276; Tonalite: This sample of mylonitic tonalite was collected from a major NE-striking shear zone (Ilivilik Shear Zone), southeast of the Qaqarssuk carbonatite, 2.5 km east of Qoorupiluup Tasia. The shear zone has more than 50 km strike length and a variable width of up to 100 m. The sample is from within the Fiskefjord Complex. A proximal sample of an associated pegmatitic component of the intrusive complex yielded a 3009 ± 4 Ma zircon U–Pb crystallization age (Sample 103622). The tonalite has a heterogeneous texture varying from protomylonite to ultramylonite on the outcrop scale. The sample consists of 50% plagioclase, 23% quartz, 15% biotite, 5% K-feldspar, and accessory augite, calcite, magnetite, titanite, sericite, apatite, amphibole, and zircon (Fig. 2). Plagioclase porphyroclasts are surrounded by a network of quartz. Quartz is medium- to coarse-grained typically displaying undulose extinction, irregular to serrate grain boundaries suggesting recrystallization via subgrain rotation. Plagioclase twinning is rarely bent indicating incipient crystal-plastic deformation. Fine-grained biotite defines high-strain layers (C structures) establishing the mylonitic foliation, whereas medium-grained biotite is commonly associated with S structures, linking it to shearing deformation. Biotite is more prevalent in strain shadows of plagioclase porphyroclasts and also fills rare extensional fissures within the plagioclase, oriented at a high angle to the foliation. Apatite grains are located in the quartz network and are dominated by small euhedral prisms. Titanite is generally medium-grained, subhedral, fractured and oriented parallel to the mylonitic foliation, implying it behaved as a rigid porphyroclast. Forty-nine spot analyses on biotite yield a Rb–Sr isochron age of 1773 ± 12 Ma, with an initial 87Sr/86Sr of 0.7027 ± 0.0042 (MSWD = 0.6; Fig. 3), interpreted as the best estimate of the time of biotite Rb–Sr closure to radiogenic-Sr diffusion. No spatial variation in biotite apparent ages is evident.

Fig. 3: Rb–Sr isochron diagrams.
figure 3

Rb–Sr biotite (muscovite) isochron diagrams for mylonites, Akia Terrane, Greenland. Each panel contains the isotopic results from one sample as indicated by the sample number in the lower right of each plot. Error bars are shown at the two-σ level. Uncertainties around the regressions are at the 95% confidence level.

Full size image

Sample 1226; Felsic gneiss: This mylonitic felsic gneiss was sampled near Puiattoq and is within the Finnefjeld Complex. Also sampled from the Ilivilik shear zone in a location where the mylonitic shear zone is more than 10 meters wide. A proximal dioritic gneiss sample yielded a 2998 ± 6 Ma zircon U–Pb crystallization age23. The sample consists of 44% plagioclase, 18% K-feldspar, 13% quartz, 12% biotite, 7% muscovite, and accessory titanite, augite, apatite, clay, zircon, amphibole, allanite, epidote, and chlorite (Fig. 2). K-feldspar and quartz are dominantly within veins consistent with an early in situ melt genesis. Texturally, the rock is protomylonitic. Quartz is generally elongated with oblique grain-shape preferred orientation, displays undulose extinction, irregular to straight grain boundaries, and chessboard textures suggesting recrystallization via subgrain rotation. Biotite and muscovite are generally oriented parallel to the mylonitic foliation and well-developed S/C structures, linking them to shearing deformation. Epidote (zoisite and clinozoisite) is common throughout the rock, rarely as syn-kinematic megacrysts, likely suggesting fluid percolation during shearing deformation. Forty-eight biotite spot analyses yield an overly dispersed errorchron age of 1523 ± 150 Ma, with geologically meaningless negative initial 87Sr/86Sr of –0.34 ± 0.44 (MSWD = 85). The scatter of these data and the initial 87Sr/86Sr indicate this biotite has undergone secondary alteration that has modified the Rb–Sr systematics. Individual biotite model ages (calculated as two-point isochrons pinned to an average crustal initial 87Sr/86Sr of 0.715 ± 0.015), yield apparent ages of c. 1650 to c. 750 Ma, the oldest of which may best reflect the minimum time through Sr isotope diffusion for the primary biotite. A correlation is apparent between ablation spot petrology and apparent age, where younger apparent ages are associated with a greater area percentage of a Fe- and Mg-poor alteration product of biotite, as visible in BSE images (Fig. 4). Fifty muscovite spot analyses yield a slightly overdispersed errorchron age of 1627 ± 20 Ma, with initial 87Sr/86Sr of 0.7141 ± 0.0063 (MSWD = 5.7; Fig. 3), interpreted as a minimum age estimate on the time of closure to Sr diffusion in muscovite. Muscovite and high Rb/Sr biotite (87Rb/86Sr > 400), inferred as least altered, yield an errorchron age of 1622 ± 18 Ma, with initial 87Sr/86Sr = 0.7147 ± 0.0006 (MSWD = 5.2; Fig. 3), interpreted as the best estimate of the minimum age of closure to Sr diffusion in most grains. The most radiogenic biotite analysis yields a comparable model age of 1648 Ma. If a 87Sr/86Sr of 0.715 is a reasonable estimate for the initial Sr reservoir altered biotite equilibrated with, then alteration occurred at some time after c. 800 Ma.

Fig. 4: Representative backscatter electron images of mica grains.
figure 4

Different mica grains within samples analyzed for Rb–Sr geochronology with various portions of laser ablation sample pits are shown. Each representative image contains the corresponding sample number in the upper left of each panel.

Full size image

Sample 1151; Felsic gneiss: This mylonitic tonalitic gneiss sample, of the Finnefjeld Complex, was recovered from Qooqqut Naerlunnguat. Sampled from a steeply dipping NE-striking mylonitic shear zone with at least 15 km strike length and width (at sample location) of more than 100 m. The outcrop comprises mylonitic tonalitic to dioritic gneisses with compositional layering, isoclinal folding and (locally) dextral shear sense indicators. The outcrop consists of variably sheared and intermingled tonalite, granodiorite, and diorite, with amphibolite boudins. A c. 568 Ma kimberlite dyke intrudes in close proximity to this outcrop and c. 3006 Ma dioritic gneisses have been dated 7 km west of this locality24. The sample comprises 43% plagioclase, 26% quartz, 18% K-feldspar, 8% biotite, 4% muscovite, and accessory minerals include apatite, augite, titanite, allanite, and calcite (Fig. 2). Quartz grains display elongate ribbons with oblique grain-shape preferred orientation, undulose extinction, and serrate grain boundaries, typical of recrystallization via grain boundary migration. Newly formed fine-grained and sub-rounded (equant) quartz grains along the boundaries of large quartz crystals may indicate lower-grade microstructural superposition during retrograde metamorphism/deformation, likely due to quartz recrystallization via subgrain rotation or bulging. Biotite vary in grain size (from fine- to medium-grained) and define discrete S/C structures linking this mineral to shearing deformation. Fifty-three biotite spots yield a slightly over-dispersed errorchron age of 1738 ± 19 Ma, with initial 87Sr/86Sr of 0.683 ± 0.058 (MSWD = 3.1; Fig. 3). Biotite model ages, defined through the initial 87Sr/86Sr from the regression, yield a prominent apparent age peak at c. 1750 Ma. A spatial relationship between the petrographic texture in the ablation locations and their apparent ages is hinted at. Patches of the thin section with a generally smaller grain size and granular texture, appear on average to have younger apparent ages (e.g., potentially a result of minor radiogenic-Sr loss), consistent with the slight over-dispersion on the regression fit.

Sample 1101; Felsic gneiss: This mylonitic felsic gneiss sample from the Fiskejord Complex was recovered from just south of Sillisissannguit Tasiat, and about 2 km east of a major mapped NE-striking fault. Sampled from a steeply dipping NE-striking mylonitic shear zone, subparallel to the Ilivilik shear zone within the Fiskefjord Complex, but to the SE. This shear zone has a strike length of at least 10 km and is several tens of meters wide. The outcrop consists of variably sheared tonalite, granodiorite, and diorite, with pods of norite. The sample comprises 34% quartz, 33% plagioclase, 25% K-feldspar, 3% biotite, and accessory muscovite, apatite, augite, rutile, and calcite (Fig. 2). The sample is compositionally banded with a finer-grained lepidoblastic domain dominated by biotite, whereas the granoblastic domain is quartz-rich with fine-grained biotite commonly pinning quartz microstructures. Plagioclase forms porphyroclasts defining a protomylonitic texture. Quartz grains from both domains displays undulose extinction, irregular to straight grain boundaries, and chessboard textures, suggesting recrystallization via subgrain rotation. In the lepidoblastic domain biotite defines well-developed S/C structures linking it to shearing deformation. Thirty-eight biotite spot analyses yield an isochron age of 1746 ± 18 Ma, with initial 87Sr/86Sr of 0.7102 ± 0.0038 (MSWD = 0.6; Fig. 3).

Sample 912; Granite: This weakly foliated granite sample from the Fiskejord Complex was taken from just south of Annikitsoralak, and adjacent to a NE-SW elongated lake. The outcrop consists of a layer of isoclinally folded granite that is in contact with amphibolite. The sample comprises 45% K-feldspar, 33% quartz, 21% plagioclase, and accessory muscovite, magnetite, chlorite, and zircon (Fig. 5). Muscovite in this sample is very fine-grained (usually <10 μm) and is present as inclusions in K-feldspar (e.g., sericite). Twelve spots on muscovite (plus inevitable K-feldspar mixtures) yield an errorchron of 2369 ± 200 Ma, with an initial 87Sr/86Sr of 0.786 ± 0.0098 (MSWD = 20; Fig. 5). The initial ratio is much greater than in typical Archean crust and implies variable rejuvenation with a radiogenic metamorphic Sr reservoir, consistent with the scatter from an isochron.

Fig. 5: Rb–Sr muscovite isochron diagram.
figure 5

Isochron for unfoliated granite sample 912 Akia Terrane, Greenland. Error bars are shown at the two-σ level. Uncertainties around the regressions are at the 95% confidence level. Inset shows a false color mineral map of the sample thin section, with the black bar denoting a 10 mm scale.

Full size image

U–Pb and trace elements

Titanite in the mylonitic tonalite sample 1276 occurs as individual blocky grains or as grain aggregates. The latter may be displaced with accompanying grain size reduction within the ductile shear fabric (Fig. 6). Sixty-one analyses define a mixing line between radiogenic and common Pb components, with the upper intercept implying a common 207Pb/206Pb of 0.949 ± 0.074 with a lower intercept age of 2941 ± 39 Ma (MSWD = 4.4; Fig. 7). The MSWD indicates that there is scatter greater than analytical uncertainties alone, with individual 207Pb-corrected ages (assuming 2940 Ma common Pb after ref. 25) ranging from 3375 Ma to 2884 Ma with a unimodal peak at 2999 Ma with low excess kurtosis (3.2), implying a distribution deviating slightly from normal. Common Pb (as estimated by f207%) varies from 0 to 58%. Excluding two analyses that may have hit inclusions or epoxy and 18 analyses lying to the right of the regression that have faded internal BSE textures indicative of minor radiogenic-Pb loss, the remaining analyses yield a regression that intersects the concordia at 2971 ± 28 Ma (MSWD = 1.3; n = 40).

Fig. 6: Representative images of Akia Terrane tonalite containing titanite.
figure 6

A False color backscatter electron (BSE) image of sample 1276, Tonalite, showing titanite (green). False color gradient scale bar denotes BSE intensity. The red minerals denote Fe oxide phases. B phase map of the same region.

Full size image
Fig. 7: Tera-Wasserburg Concordia diagram of titanite from sample 1276.
figure 7

Green squares denote main group, gray squares denote potential minor radiogenic-Pb loss, and orange squares denote potential old outliers/analytical mixtures with epoxy. Error bars are shown at the two-σ level. Uncertainty around the regression line is shown at the 95% confidence level.

Full size image

Titanite trace element content may differentiate between magmatic, recrystallized, and metamorphic growth26. A titanite chondrite-normalized rare earth element (REE) plot of analyses from sample 1276 reveals fractionated patterns dominated by light-REE with an average (La/Lu)N of 6.87 and generally a positive Eu anomaly (Fig. 8). Using shape coefficients to mathematically compare chondrite-normalized REE patterns27, we conclude that this titanite is igneous (Fig. 8). Other trace element signatures such as Th/U (average of 1.44) and Th/Pb (average of 2.86) also support an igneous origin for this mineral28,29. Zr-in-titanite temperatures30 imply temperatures of ~770 °C for crystallization, assuming a zircon-quartz buffered system at 0.8 GPa16.

Fig. 8: Titanite trace element geochemistry.
figure 8

A Titanite chondrite-normalized REE patterns from sample 1276 using reference values from ref. 70; B Comparison between sample 1276 titanite (filled orange circles) and typical magmatic (gray unfilled circles) and metamorphic titanite (blue unfilled circles) compiled from the literature27. The reader is referred to ref. 71 for information about the lambda (λn) parameters, which were calculated using BLambdaR72.

Full size image

Discussion

Mica ages

Violin plots of biotite Rb–Sr model ages can be constructed (Fig. 9), assuming an initial 87Sr/86Sr constrained by the intercept of the least squares regression in each sample, provides a reasonable indication of the Sr isotopic reservoir with which the biotite equilibrated. Notably, there is low sensitivity in calculated age to the choice of initial 87Sr/86Sr ratio given the highly radiogenic biotite 87Sr/86Sr ratios31. Biotite model ages for three of the four mylonite samples cluster at c. 1750 Ma, with one sample associated with altered biotite (Fig. 4), showing a greater spread from Paleoproterozoic to Neoproterozoic apparent ages.

Fig. 9: Measures of central tendency and variability in Rb–Sr ages from Akia Terrane mica grains.
figure 9

Violin plots with box and whisker overlay of biotite and muscovite Rb–Sr model ages. Median values are denoted by the central black line in the box, Q1 and Q3 by the bars. Red vertical bar denotes mode at c. 1750 Ma. Muscovite is indicated by the green fill in the violin plot of sample 1226.

Full size image

Empirical field-based studies in Europe have implied Rb–Sr closure temperatures of ~300 °C in biotite and ~500 °C in muscovite32. Muscovite from sample 1226 with altered biotite indicates Rb–Sr apparent dates around c. 1615 Ma that contrast with highly dispersed and seemingly younger biotite apparent model ages. Muscovite in sample 1226 lacks obvious alteration, whereas biotite in places has intergrowths of a lower Mg- and Fe-mica secondary reaction product (Fig. 4). The commonly reported closure temperatures for the Rb–Sr geochronometer in muscovite is in the range of 500 °C to >600 °C33, yet it has been proposed that muscovite can lose Sr due to deformation at temperatures <500 °C34. Furthermore, mineral composition has also been linked to variable Sr retention35. In contrast to the more consistent muscovite ages in sample 1226, the distribution of apparent ages in biotite implies a secondary alteration that has partially rejuvenated its Sr isotopic composition. Nonetheless, most samples show well-defined biotite isochrons with estimates of model ages clustering at c. 1750 Ma (Fig. 9).

Given these data, an important question to address is how closely the biotite Rb–Sr ages of c. 1750 Ma approximate the growth age of the shear zone fabric. In other words, did biotite grow prior to c. 1750 Ma, above its Sr closure temperature, and did the mylonites only cool at this time? To address this question we first need to consider the highest temperature that these rocks experienced.

Primary magmatic temperatures of these TTG are likely around ~750–800 °C16 and, depending on the overprinting history, the titanite U–Pb system could retain its primary magmatic crystallization age if it reflects an igneous mineral. Despite a slight skewness towards younger ages, the titanite U–Pb age distribution is broadly consistent with a single age component with only minor radiogenic-Pb loss or perhaps a slight closure temperature differential in the grain population. The dominant titanite U–Pb age of c. 2970 Ma is similar to the c. 3000 Ma zircon U–Pb magmatic crystallization age for Akia Terrane felsic gneisses24. Trace element patterns of this titanite are also characteristic of a magmatic genesis and the Zr concentration supports a magmatic crystallization temperature of ~770 °C. Both the U–Pb systematics and REE compositional pattern point to magmatic titanite growth during the main regional magmatic emplacement event24, which dominated the isotopic memory of this mineral, with subsequent thermal conditions being unable to wholesale reset its U–Pb system.

Whilst high-temperature granulite-facies conditions have been established for parts of the Akia Terrane in its early history (western Fiskefjord complex and southern Alanngua complex), other regions within the terrane only reached amphibolite facies conditions14,16,36. A subsequent Late Mesoarchaean to Neoarchaean (2857–2700 Ma) granulite-facies metamorphic overprint (~820–850 °C at 8–10 kbar) was reached in the Alanngua complex14 but clearly cannot have pervasively affected the area around sample 1276, as its titanite U–Pb age is not strongly affected by thermally activated diffusive radiogenic-Pb loss post-magmatic crystallization.

With the higher temperature conditions evaluated, we can now better understand the meaning of the biotite Rb–Sr data. We employed inverse thermal history modeling (Supplementary Figure 2) to test whether the measured biotite Rb–Sr dates represent the time of biotite formation/crystallization (i.e., biotite Rb–Sr dates are crystallization ages of newly formed biotite), or record cooling of pre-existing biotite after an earlier metamorphic peak (i.e., biotite Rb–Sr dates are cooling ages; Fig. 10).

Fig. 10: Time-temperature evolution models.
figure 10

Thermal history models constrain the maximum permissible temperature to leave titanite unreset (blue squares) yet entirely thermally reset biotite (green squares). The difference in calculated temperature implies that biotite Rb–Sr reflects a neoblastic age.

Full size image

To test these hypotheses, we first constrained the maximum temperatures the samples could have experienced during the time recorded by the biotite Rb–Sr data (i.e., c. 1750 Ma) by utilizing the titanite U–Pb data. Maximum temperatures were modeled based on available titanite U–Pb ages clustering at 2970 Ma and the experimentally determined diffusion kinetic parameters for the titanite U–Pb system (Ea = 328.7 kJ mol−1 and D0 = 0.001 cm2 s−1)37. The modeling results suggest that the maximum permissible temperatures at 1750 ± 68 Ma recorded by the biotite Rb–Sr data could not exceed ~530 °C, otherwise the titanite would not have been able to retain its U–Pb igneous crystallization signature. Whilst a wide range of Pb diffusion parameters exist for titanite38, with some equivalent to diffusion rates of Sr in titanite39, variation to coefficients within plausible ranges make little difference to the conclusion that biotite temperatures cannot have greatly exceeded ~500–550 °C (Fig. 6).

In the second model, we aimed to constrain the maximum temperature at c. 1750 Ma by utilizing the biotite Rb–Sr model age and grain size data. The biotite grains range in size from ~5 to ~600 μm in diameter (Supplementary Data 6), nevertheless the measured (>64 μm) biotite Rb–Sr dates tends to form tight clusters (c. 1750 Ma) regardless of the grain size. To incorporate this observation into the model, we computed thermal trajectories simultaneously for diffusion domains of different sizes (i.e., 10, 50, 100, 200 μm in diameter), adopting diffusion kinetic parameters for biotite Rb–Sr of Ea = 328.7 kJ mol−1 and D0 = 0.001 cm2 s−1, with the rock cooling from above the Sr blocking temperature40. The maximum temperatures permissible for the observed biotite Rb–Sr age and diffusion domain size are modeled to be in excess of ~610 °C. This temperature estimate on first consideration would appear to be at odds with the former model for Pb diffusion in titanite, as it would imply complete resetting of the titanite U–Pb system and hence complete rejuvenation of its U–Pb ages.

This apparent conflict between the outcomes from the two thermal history models is easily reconciled by the biotite isotopic result dating growth of this mineral at c. 1750 Ma below ~530 °C, rather than reflecting a cooling age. The fact that the biotite Rb–Sr model ages are homogeneous (and the isochrons generally rather well-defined, when not subject to secondary chemical alteration) irrespective of the biotite grain size lends additional support to a growth timing interpretation for the Rb–Sr age of biotite defining the mylonitic fabrics in these samples. A dispersed apparent age spectrum could result from slow cooling, for a biotite population of dissimilar grain sizes (e.g., >5 μm to <1 mm). Yet in the Akia Terrane mylonite data set, disparate Rb–Sr biotite ages correspond only to where petrographic evidence of mineral alteration is detected (Fig. 4). Hence, we see no evidence of slow cooling and conclude the mylonite biotite Rb–Sr apparent age component at c. 1750 Ma is best interpreted as the time of crystallization and, thus, directly constrains the timing of mylonite generation. Furthermore, sericitic muscovite sampled from an unfoliated granite, without mylonite textures, yields a Rb–Sr errorchron of c. 2370 Ma (Fig. 5). While this age may not be directly meaningful, it necessitates regional cooling to this mineral’s radiogenic-Sr retention temperature well before mylonite generation.

Strontium evolution modeling indicates that the initial 87Sr/86Sr of the Akia samples are higher than expected for average crustal Rb/Sr ratios (Supplementary Figure 3). The initial 87Sr/86Sr ratios are consistent with Sr input from a radiogenic source, either a Rb-rich crustal reservoir or metamorphic fluids (e.g., produced through devolatilization reactions41). U–Pb apatite ages of c. 1800–1700 Ma in the Akia Terrane have been interpreted to reflect regrowth from a compositionally distinct reservoir after primary magmatic growth, driven by regional tectonothermal and fluid activity at that time18. Additionally, other works on the terrane have also reported a wide range of apatite U–Pb ages between ≤3000 Ma (magmatic growth) and a dominant c. 1750 Ma (neoblastic or recrystallized) component, consistent with Proterozoic fluids partially to completely mobilizing Sr, which can be dominantly held in this mineral17.

Regional implications

The new biotite crystallization ages at c. 1750 Ma, constraining mylonite genesis, occur in the terminal phase of the Nagssugtoqidian orogeny42,43 and are located ~150 km south of the Nagssugtoqidian orogenic front44. The Nagssugtoqidian Orogen forms a Proterozoic orogenic belt the girdles the northern edge of the North Atlantic Craton and crops out in southwest and southeast Greenland45. The orogen includes a range of Archean and Paleoproterozoic rocks affected by polyphase deformation and high-grade metamorphism42. South-verging structures in its southern extent support an interpretation of subduction under the Ammassalik Intrusive Complex46. The southern boundary of the Nagssugtoqidian Orogen has been defined by the imprint or apparent lack thereof on Paleoproterozoic dolerite dykes47, with the boundary of the orogen located in a near-vertical dextral shear zone48. However, Proterozoic thermal processes are well known from south of the Nagssugtoqidian front. For example, the Rb–Sr isotopic system of biotites in gneisses of the Godthaabsfjord region (Isukasia terrane), south of the Akia Terrane, was interpreted to reflect thermally activated resetting at 1700–1600 Ma21. Additionally, biotite Rb–Sr solution analyses from the Amitsoq gneisses at Isua (Isukasia terrane) yielded isochron ages of 1623 ± 65 Ma20. This isochron was interpreted as tracking low-grade metamorphic re-equilibration of biotite. The timing of this thermal event is also within uncertainty of a Rb–Sr age from an intrusive granite in the same region49. In the Akia Terrane, a c. 1690 Ma Rb–Sr two-point epidote-biotite isochron was considered to record a thermal event at that time50.

The Nagssugtoqidian Orogen’s evolution is characterized by c. 1860–1840 Ma WNW–ESE compression associated high-grade metamorphism and thrusting51. Subsequent c. 1775 Ma sinistral strike-slip shear zones transect the orogen51. Terminal Nagssugtoqidian fabrics dated by high-temperature chronometers (zircon and monazite U–Pb) yield c. 1770 Ma ages whereas amphibole 40Ar/39Ar ages of c. 1700–1740 Ma reflect cooling times during exhumation52. That the major NE-SW shear zone structures dipping to the northwest in the Akia Terrane have identical formation ages (as dated in this paper) to the late-stage uplift processes in the Nagssugtoqidian Orogen supports the interpretation that Proterozoic thrusts propagated southwards deep into the Archean North Atlantic Craton. This important Proterozoic orogenic overprint cautions that, despite being proposed as preserving some of the best evidence for horizontal tectonics in the Archean53, much of the Akia Terrane, if not a larger portion of the North Atlantic Craton, was subject to subsequent structural modification, including growth of new fabrics and modification of some existing isotope systems. Biotite from further south of the Akia Terrane appears to record somewhat younger Rb–Sr isochrons of <1700 Ma20,21,49. These isochrons are commonly interpreted as a function of a late Proterozoic thermal overprint21, linked to minor granitic magmatism49. Hence, it appears possible to distinguish between new fabric growth, in the Akia Terrane at c. 1750 Ma, coeval with epidote alteration50, and thermal overprinting and cooling of biotite with Rb–Sr ages best estimated at c. 1620 Ma20. Apatite U–Pb geochronology also indicates recrystallization of this mineral at c. 1750 Ma, throughout the North Atlantic Craton17,18,54,55.

A distinct timing of granulite-facies metamorphism affected rocks north and south of Maniitsoq, reflecting either distinct terranes, or magmatic emplacement of younger rocks to the north into an older metamorphic basement to the south56. The boundary along Søndre Isortoq has been envisaged as separating distinct crustal blocks (Maniitsoq in the north and Fiskefjord in the south), with different crustal levels exposed in each, which had been formed in distinct continents, and only later amalgamated by continental collision57. Others have proposed that this collision of the Maniitsoq block with the Fiskefjord block occurred by south-directed subduction and convergent tectonism at c. 2560 Ma58. Importantly, the wide commonality of Hf isotopic signatures in zircon crystals from across the North Atlantic Craton, implying a Hadean to Eoarchean crustal nucleus, indicates that crustal blocks within the North Atlantic Craton need not be disparate terranes59. Rather, a crustal nucleus, perhaps formed via early Earth sagduction style processes, may have been extended, fragmented, and recompressed. The imprint of several cycles of deformation, the latter of which has classic thrust geometries and mylonite zones, undoubtedly complicates the disambiguation of Archean from Proterozoic structures and confounds the distinction between horizontal and vertical plate tectonic processes in West Greenland.

Mylonitic zones in the Akia Terrane are also relevant in understanding the nature and timing of a range of later geological processes that modified the crust of this region. For example, Ni-Cu-Co mineralization in norites, which occurs in a curvilinear belt around the Alanngua fjord area, is remobilised and dissected along mylonites60. These norites contain semi-massive pyrrhotite-pentlandite-chalcopyrite-pyrite sulfides, with accessory magnetite and ilmenite, in brecciated units hosting rounded, centimeter-sized and larger, blocks of country rock61. Remobilized sulphides characteristically track into the norites and are locally associated with mylonite zones in the orthogneiss country rocks62. Mylonites, including one dated in this work, transect mineralized sections at Fosillik, displacing norites and remobilising the sulfide assemblage. Of note is that the greatest sulfide accumulation has been observed in proximity to the mylonite zones62, suggesting that Proterozoic tectonism may be important in the Ni-Cu-Co ore evolution.

Additionally, the dated Akia mylonites occur in areas where cataclastic zones have been purported to relate to Earth’s oldest putative bolide impact63 (but see ref. 64). The dating of these mylonites further calls into question the interpretation that the cataclastic zones of the region are related to a Mesoarchean impact, given that those zones have not been dated and Proterozoic deformation fabrics have now been established in the very same area.

Conclusions

Biotite Rb–Sr ages of c. 1750 Ma from mylonite fabrics in the Akia Terrane coincide with the end of the Nagssugtoqidian orogeny and highlight that structures of this age penetrate into the Akia Terrane, in line with the wide thermal halo of this event.

The dated mylonites are from amphibolite facies large shear zones, and the biotite within them grew at c. 1750 Ma rather than being reset by thermally activated volume diffusion at this time. The growth of biotite in mylonite fabrics implies that these major shear zones, which are parallel to other similar shear zones in west Greenland, including in the Nuuk region, were active at this time with fluid mobility and mineral growth.

One of the dated mylonitic shear zones cuts the Fosillik area, where Ni mineralization was remobilized along late mylonites. This suggests that Proterozoic tectonism may be important in the evolution of Ni deposits in the region.

Throughout the North Atlantic Craton Paleoproterozoic Rb–Sr mineral resetting ages are recognized. Here we show that amphibolite facies structures, in the Akia Terrane of South West Greenland, are associated with biotite growth at c. 1750 Ma. These NE-SE shear zones are south of the mapped Nagssugtoqidian deformation front. Thus, in the North Atlantic Craton there appears to be both new mineral growth on major shear zones at c. 1750 Ma, in the Akia Terrane, and also thermally activated biotite Rb–Sr resetting at a slightly later time (1700–1600 Ma), in the Isukasia terrane (e.g., ref. 21). Such young high-grade structures and thermal effects in the North Atlantic Craton highlight that horizontal tectonics have overprinted original Archean relationships across the grain to terrane scale, accommodating unknown amounts of displacement of the stratigraphy and ultimately cautioning that interpretations of ancient geodynamics in the North Atlantic Craton, as a bastion of early Earth modern-style plate tectonics, require detailed evidence.

Methods

Mineral maps

TESCAN Integrated Mineral Analyzer (TIMA) mineral maps of thin sections (Fig. 2) were produced at Curtin University. The TIMA instrument is based around a secondary electron microscope, fusing energy-dispersive X-ray spectroscopy (EDS) and back-scattered electron (BSE) images to create mineral maps. Thin sections were carbon-coated and analysed over their full surface using a grid with 3 μm step size for BSE. A 2-μm step size was used for EDS grids or when BSE contrast changed, that is to say, when a mineral boundary was detected. EDS spectra consist of one thousand counts, are standardized to a pure Mn reference, and are statistically compared to a mineral spectra database for identification purposes. Operational conditions65 consisted of a beam energy of 25 keV, a beam intensity of ~19 keV, a probe current of ~5.3 nA, a spot size of ~80 nm, and a nominal working distance of 15 mm. High-resolution maps of thin sections are provided in Supplementary Data 1.

Isotope analyses

All isotope measurements were carried out at the GeoHistory Facility in the John de Laeter Center, Curtin University using laser ablation techniques on polished thin sections. Rb–Sr data on biotite were collected by LA-ICPMS using a QQQ (Agilent 8900 ICPMS) with a Resonetics RESOlution SE 193 nm excimer UV laser incorporating a dual volume S155 sample cell. The 8900 was run in MS/MS mode using N2O reaction gas. Detailed analytical procedures are provided in Supplementary Methods and all Rb–Sr data are in Supplementary Data 2. Reference biotite CK001 yielded an age of 420 ± 6 Ma (87Sr/86Sri = 0.681 ± 0.066; MSWD = 0.8, n = 53) identical, within uncertainty, to the published value of 413 ± 3 Ma and 87Sr/86Sri of 0.709 ± 0.00165,66 (Supplementary Figure 1).

Titanite U–Pb isotope (Supplementary Data 3) and trace element concentration data (Supplementary Data 4) were also obtained via LA-ICPMS analysis. Individual titanite grains were ablated utilizing a Resonetics RESOlution M-50A-LR system based on a COMPex 102 193 nm excimer UV laser. U–Pb analyses used a 5 Hz laser repetition rate with a fluence of 2.6 J cm−2. The sample cell was flushed with ultrahigh purity He (350 mL min−1) and N2 (3.8 mL min−1). Isotopic abundances were measured using an Agilent 8900 triple quadrupole ICPMS, utilizing high-purity Ar as the plasma gas (1 L min1). The detailed LA-ICPMS method is given in the Supplementary Methods and sample geographic coordinates are provided in Supplementary Data 5.

Inverse thermal history modeling

To explore the meaning of the geochronological data (e.g., crystallization versus cooling versus partial resetting), inverse thermal history modeling using the HeFTy software67 was employed. HeFTy reconstructs time-temperature pathways for multiple thermochronometric systems for which diffusion parameters are known or can be reasonably inferred. The specific objective of the modeling effort was to decipher the meaning of the Rb–Sr dates in biotite. Accordingly, different models were constructed as described in the following sections using grain sizes defined in Supplementary Data 6.