U–Pb geochronology reveals evidence of a Late Devonian hydrothermal event, and protracted hydrothermal–epithermal system, within the Mount Painter Inlier, northern Flinders Ranges, South Australia

Abstract

The Mount Gee Sinter and the Radium Ridge Breccia within the Mount Painter Inlier, South Australia, preserve evidence of a hydrothermal event peaking during the Late Devonian (ca 365 Ma). Prior to this study, limited data relating to this event were available, but our results of 846 LA-ICPMS U–Pb monazite analyses indicate the timing of this hydrothermal event. The dominant monazite population age of ca 363 Ma for the Mount Gee Sinter represents earlier phases of a protracted hydrothermal–epithermal system, whereas a later epithermal phase cross-cuts rocks hosting ca 220 Ma age zircon. Parts of the Radium Ridge Breccia have been recently interpreted as a series of Early Cretaceous glacial events. Zircon ages (604 zircon analyses) from the Radium Ridge Breccia define a detrital population dominated by ca 1585 Ma ages, consistent with derivation of clasts within the breccia from local granitic and metasedimentary basement. The Radium Ridge Breccia is, however, dominated by a ca 367 Ma aged monazite population, probably reflecting overprinting of the local Mesoproterozoic granitic basement rock during the same hydrothermal event as formed the earlier phase of the Mount Gee Sinter. It is interpreted that the monazite ca 365 Ma age reflects a significant Late Devonian hydrothermal event, evidenced locally in the southern Mount Painter Inlier, resulting from localised heat flow but with thermal implications for the regional geology of the northwestern Curnamona Province.

Keywords:

• Late Devonian hydrothermal event
• Radium Ridge Breccia
• Mount Gee Sinter
• Mount Painter Inlier
• LA-ICPMS
• geochronology

Introduction

One of South Australia’s eminent geologists, Reginald Sprigg AO (Keeling & Hore, 2007; McGowran & Hill, 2015; Weidenbach, 2008), described the spectacular accumulations of crystals and vughular quartz of Mt Gee at Arkaroola as ‘the Crystal Mountain’ representing the tip of an intriguing geological mystery, and the formation of the Mt Gee area as posing ‘one of the great geological conundrums of our day’ (Sprigg, 1984). As a student, Sprigg studied at Arkaroola, and later, as a government geologist (Geological Survey of South Australia), he worked at Arkaroola. In 1968, he purchased Arkaroola, converting the property to a wilderness sanctuary and tourist destination. He was well travelled locally and internationally in his geological career; he was not one to make flippant geological comments.

The Arkaroola property includes the southern area of the Mesoproterozoic Mount Painter Inlier (MPI) within the Mount Painter region (Hore, 2015) (Figure 1). The MPI and the adjacent Mount Babbage Inlier are a relatively small but exposed western section of the Moolawatana Domain (Conor & Preiss, 2008) of the Curnamona Province. Within the MPI, the ridgeline of Mt Gee and the nearby summit of Mount Painter represent the best exposures of the multi-phased Mount Gee Sinter (MGS) (Drexel & Major, 1987), formed by hydrothermal/epithermal siliceous fluids. The earliest phase is SiO₂ ± Fe ± P ± rare earth element (REE)-rich, comprising monazite-rich layered quartz–hematite masses, exposed primarily on Mt Gee (Figure 1). This earliest phase may be associated with localised hydrothermal brecciation and uranium mineralisation of the primarily granitic rock. The latest phase displays many typical epithermal textures. This epithermal quartz phase overprints and/or intrudes many of the earlier phases on Mt Gee. Mt Gee is an expression of a fossil epithermal environment, which is typically associated with geysers and hot pools (Brugger et al., 2011; Drexel & Major, 1987; Sprigg, 1945; Stillwell & Edwards, 1945). The Mt Painter outcrops display a breccia-style melange, consisting primarily of MGS clasts, cemented by the late epithermal quartz.

Figure 1. Location and general geology of the northeastern Flinders Ranges, focusing on the southern area of the Mesoproterozoic Mount Painter Inlier that hosts the Mount Gee Sinter, Mt Gee and the Radium Ridge Breccia.

Also in the Mt Gee–Mt Painter region is the enigmatic Radium Ridge Breccia (RRB) (Figure 1; Drexel & Major, 1987; Hore et al., 2020), which is associated with the MGS. The RRB rocks range from quartz–feldspar-breccia and granitic breccia to sedimentary arenaceous/arkosic rocks. Some of these are affected by major brecciation, whereas others are considerably less affected and preserve primary sedimentary structures. Clasts are predominantly locally derived and angular to sub-rounded, and range from granules to boulders. Hematite appears to have been introduced into much of the RRB, as well as chloritic alteration and silicification (Collins, 1976). The RRB contains a tillite horizon, the Sprigg Tillite Member (STM) (Hore et al., 2015, 2020); both the RRB and STM are intruded or cross-cut by the late-stage epithermal phase of the MGS.

Youles (1975, 1986) and Major (1976) interpreted the RRB as sedimentary rocks, whereas others suggest a diatreme-related, hydrothermal origin or lateritisation resulting in formation of the breccias (Coats & Blissett, 1971; Drexel, 1980; Drexel & Major, 1987, 1990; Lambert et al., 1982).

Prior to this study, and the recent study of Hore et al. (2020), the MGS and the RRB were generally thought to have formed during the Permo-Carboniferous to Cenozoic (Brugger et al., 2011; Coats & Blissett, 1971; Idnurm & Heinrich, 1993). Youles (1975), however, suggested the most probable age of the RRB to be Neoproterozoic. From field interpretation, Drexel and Major (1990) suggested Late Ordovician to perhaps Silurian and, owing to the high degree of preservation of the MGS, suggested an age perhaps as young as Cenozoic. Results of a limited number of geochronological analyses of monazite and molybdenite hosted in the RRB were interpreted to represent the age of uranium mineralisation from Ordovician to Devonian (Elburg et al., 2003; Elburg et al., 2013; Pidgeon, 1979; Skirrow et al., 2011). A comprehensive set of geochronological zircon analyses of the STM (Hore et al., 2015, 2020) supports an Early Cretaceous age for the tillite.

The absolute ages of the RRB and MGS, however, are uncertain. This uncertainty has held back the understanding of the significance of these units in terms of regional geological evolution, and exploration models, particularly for unconformity-style uranium mineralisation, which we introduce in this paper.

The objective of this contribution is to present the first comprehensive laser ablation inductively coupled plasma mass spectrometry (LA-ICPMS) U–Pb isotope age data set of monazite analyses from the MGS, and monazite and zircon from the RRB. The new U–Pb results, in combination with field relationships and textural evidence, indicate a complex, multi-stage history recorded by the MGS and RRB, and, in particular, provide, from multiple samples, strong evidence of a significant Late Devonian hydrothermal event in the MPI.

This manuscript shows that the Mount Painter region continues to provide new evidence of geological processes that would intrigue and fascinate Reginald Sprigg.

Geological setting

The MGS (Drexel & Major, 1987) and RRB (Drexel & Major, 1987; Hore et al., 2020) are within the Mesoproterozoic rocks of the southern MPI (Figures 1 and 2). In addition to a spatial relationship, there is also a possible genetic relationship between the MGS and RRB (Stillwell & Edwards, 1945). The Mount Painter region (Hore, 2015) of the northern Flinders Ranges is centred on the ?Paleoproterozoic–Mesoproterozoic Mount Painter and Mount Babbage inliers (Cowley et al., 2011; Fricke & Hore, 2011; Hore et al., 2020; Neumann et al., 2010; Teale, 1993).

Figure 2. MGS and RRB geochronology sample locations. Satellite imagery overlain by generalised geology with MGS samples locations in blue, and RRB in red. Numerous uranium prospects within RRB are highlighted.

Mount Painter Inlier

Mesoproterozoic rocks of the MPI include the Radium Creek Group (Cowley et al., 2011), which consists of strongly foliated quartz–feldspar–biotite gneisses, schists and metasedimentary rocks, intruded by 1585–1550 Ma U- and Th-rich granites and granodiorites of the Coulthard and Moolawatana suites (Cowley et al., 2011; Fricke & Hore, 2011; Neumann et al., 2010), and mafic dykes and pegmatites.

In this study area of the southern central portion of the MPI, these metamorphics and granites are, in part, unconformably overlain by the unmetamorphosed unit of the Early Cretaceous RRB (Drexel & Major, 1987; Hore et al., 2020). The MGS is hosted in and intrudes the Mesoproterozoic basement rocks and the RRB.

The history of the Mesoproterozoic MPI includes uplift, erosion and granitic exposures, with reburial commencing about 850 Ma by sediments of the Adelaide Rift Complex (Preiss, 2000; Veevers et al., 1997). These Neoproterozoic sediments progressively buried the Mesoproterozoic MPI rocks to depths of 12–15 km (McLaren et al., 2002; McLaren et al., 2006).

These Adelaidean (Preiss, 1987, 2000) and underlying Mesoproterozoic rocks were variably deformed, particularly during the ca 510–480 Ma Delamerian Orogeny (Preiss, 1995). There is a general increase in metamorphic grade in the Adelaidean rocks towards the MPI, attributed to the high radiogenic heat production of the Mesoproterozoic granites and gneisses. Their heat production rate averages ∼16 μWm⁻³, more than four times the heat production for average granites and three times that of the global upper crust average (McLaren et al., 2002; Nuemann, 2001; Neumann et al., 2000; Sandiford et al., 1998). At ca 450–440 Ma, the Mesoproterozoic rocks were intruded at a depth of ∼14 km by the I- and S-type British Empire Granite, formation of which is also attributed radiogenic heat owing to the burial of these high heat-producing Mesoproterozoic rocks (McLaren et al., 2002, 2006).

The region was later subjected to episodic and abrupt uplift, during which, exhumation resulted in a combined minimum of 6–7 km of denudation, recorded in three cooling episodes at 430 Ma, 400 Ma and 330 Ma (McLaren et al., 2000, 2006). Post 330 Ma, the Mount Painter region has been exposed from depths of at least 8 km, with a minimum of 3 km of this denudation occurring in the interval 330–320 Ma (McLaren et al., 2000, 2002; Mitchell et al., 2002). Continued uplift of the Mesoproterozoic basement during the late Paleozoic and Mesozoic is supported by low-temperature thermochronology (Foster et al., 1994; Mitchell et al., 1996, 2002; Weisheit et al., 2014), exposing the basement rocks at the surface prior to the Cretaceous.

The basement rocks of the MPI were thought to be reburied by Cretaceous sediments, which contributed to the reheating prior to a final phase of basement cooling in the Late Cretaceous (Foster et al., 1994; Gordon, 2007; Mitchell et al., 1996, 2002). Evidence from apatite fission track analysis indicates that uplift and regional cooling of up to 60°C occurred in the Flinders Ranges during the Eocene, accompanied by an increase in sedimentation rates into surrounding basins (Mitchell et al., 2002). Foster et al. (1994) found evidence for the flow of large volumes of fluids during this time in the Paralana Fault. Zircon (U–Th)/He apatite, and sphene fission track ages indicate reheating during the mid–late Cretaceous times, which cannot be explained by burial alone (Weisheit et al., 2014). From ca 40 to 10 Ma, much of the cover was eroded during neotectonic uplift of the basement, leading to the present surface expression that includes remnants of the MGS and Early Cretaceous RRB.

Today, the active Paralana Hot Springs (Figure 1), located on the Paralana Fault, discharge meteoric water at a temperate of ∼60 °C, a result of radiogenically heated sub-terranean waters that percolated down to hot rocks and then upwelling to the surface via faults within the MPI (Anitori et al., 2002; Brugger et al., 2005; Mawson, 1927).

Mount Gee Sinter

The rocks of the MGS (Drexel & Major, 1987) are interpreted to have been formed via multiple episodes of primarily silica- and iron-rich fluids. The two fluid phases we suggest are the earliest, and the latest are defined by differences in mineralogy as well as by field their relationships. The earliest SiO₂ ± Fe ± P ± REE-rich phase of the MGS contains magnetite–hematite and monazite. The later phase of the MGS lacks the monazite and is generally Fe-poor. It displays many epithermal textures (Sprigg, 1945; Stillwell & Edwards, 1954) with euhedral quartz forming crusts and vugh linings with radiating quartz crystals in cylindrical masses. A common and characteristic variety of this late-stage silica was deposited as thick encrustations around prismatic laumontite crystals up to 0.1 m long. In all but one locality, the laumontite has dissolved to leave the ‘nail-holes’, locally referred to as ‘nail-hole’ quartz (Figure 3; Coats & Blissett, 1971).

Figure 3. Representative Mount Gee Sinter textures. (a) Mt Gee in foreground (left of centre), capped by MGS dipping ∼30° east. Image looking north. Freeling Heights in background (scale: single-vehicle track on right of image). (b) MGS texture; radiating quartz crystals surrounding casts of what was originally acicular laumontite crystals—‘nail-hole’ quartz (scale bar 1 cm graduations). (c) MGS texture; quartz replacing ?calcite/barite/fluorite rhomboidal crystals (scale bar 1 cm graduations). (d) MGS texture; euhedral quartz in-filling voids (scale bar 1 cm graduations).

The MGS outcrops (and drill-hole intersections) are limited to the south-central part of the MPI (Figures 1 and 2), covering an area of less than a few square kilometres. The main outcrops of the MGS include both the topographic highs of Mt Gee (20–50 m thick) (Figure 3a), which has the highest abundance of SiO₂ ± Fe ± P ± REE-rich rocks mixed with Si-rich geodic quartz, and Mt Painter (up to 200 m thick), which is primarily a breccia dominated by MGS clasts cemented by Si-rich geodic quartz. Smaller outcrops of MGS occur along Radium Ridge (Figure 2). The geodic and ‘nail-hole’ quartz of the MGS occurs mostly around Mt Gee, but is also hosted within the brecciated granite of the RRB (and the underlying fracture veins within the Mesoproterozoic metasedimentary rocks and granites) in the area extending from Mt Gee NE towards Paralana Hot Springs on the Paralana Fault (Figure 1).

The layered to partly clastic quartz–hematite sequence of the early stages of the MGS includes the presence of specular hematite, aggregates of hematite including martite and minor monazite associated primarily with the hematite, some of which is locally enclosed by aggregates of hematite or martite (Drexel & Major, 1987; Freytag, 1995; Stillwell & Edwards, 1945; Whitehead, 1976).

Monazite grains in the earliest MGS are euhedral to subhedral, prismatic, tabular crystals with sizes up to 0.5 mm, and in some exceptional cases up to 1.5 mm. They occur either as singular, large fractured crystals or as polycrystalline aggregates. Monazite is ubiquitously associated with euhedral hematite, which is likely a replacement of magnetite (Stillwell & Edwards, 1945).

The later MGS quartz ranges from dark ‘liver’ red, hematite-rich quartz to white, hematite-free quartz. In places, the earliest Fe ± P ± REE-rich silica phase has been brecciated with gaps (voids can be metres in diameter) recemented or lined with the later euhedral epithermal quartz. Various textural features of the siliceous exposure of Mt Gee and Mt Painter are diagnostic of an epithermal origin (Figure 3b–d) (Brugger et al., 2011; Drexel & Hore, 2004; Drexel & Major, 1987; Hilyard, 1998; Idnurm & Heinrich, 1993; Sprigg, 1945, 1989; Stillwell & Edwards, 1945, 1954; Youles, 1978). Evidence suggesting that late-stage activity is epithermal in character is provided by the fluid inclusion homogenisation temperatures of 110–200 °C (Collins, 1977; Simpson & Topp, 1997). The ‘nail-hole’ quartz latter phase of the MGS is the only known lithology to cross-cut the Early Cretaceous STM (Hore et al., 2015, 2020) of the RRB. No clasts of the earlier phase are known as clasts within the STM, or any STM clasts within the MGS.

Interpretation of the age of the MGS has been based previously on field observations with previous authors typically providing multiple options, highlighting the complexity of the field relationships (Coats & Blissett, 1971; Drexel & Major, 1990; Sprigg, 1945; Sullivan, 1945).

Radium Ridge Breccia

The RRB consists mostly of granitic rocks, breccia and tillite, interpreted by Hore et al. (2020) as periglacial/glacial/proglacial sedimentary facies associated with an Early Cretaceous glacial event. Also, the RRB probably includes brecciated rocks of an earlier subterranean hydrothermal brecciation event, which is subsequently overprinted by deep cryogenic weathering associated with the Early Cretaceous glacial event. At some localities, we have identified within the exposures the unconformity between the hydrothermal breccia and the glacial components of the RRB (Hore et al., 2020).

The lithologies of the RRB are detailed in Drexel and Major (1987) and Hore et al. (2020), with petrological descriptions by Whitehead (1976). In brief, they include angular to rounded fragments varying from microscopic size to at least 10 m in diameter, set in a red, brown, green or grey matrix. Rock types recorded as coherent rocks and fragments within the breccias include granite, granitic gneiss, gneiss, schist, quartzite, carbonate rocks and unmetamorphosed siltstone and sandstone; the large majority of fragments are of granitic composition. The metamorphic rock types match local basement geology. Unmetamorphosed clasts are found in the brecciated portion of the RRB reported by Youles (1969) and supported by field observations by Bob Major and are also reported in the distinctive tillite portion (Hore et al., 2020). These clasts are not locally derived.

The degree of brecciation varies from incipient and local comminution of basement rocks to incoherent, matrix-supported breccia of granitic rock. Clast abrasion is evident in thin-section (Whitehead, 1976). Some outcrops of granitic breccia consist of extensively fractured and crushed granitic rock in which much of the feldspar has been granulated to a finer grainsize. This rock flour matrix is dominantly quartz and feldspar of granitic origin. Fine-grained silty to sandy sediments can be within the breccia, in some cases with cross-bedded or graded layering. Rounded clasts of hematitic siltstone and breccia are enclosed within the breccia and sediments, suggesting repetition of the process that formed them.

Accessory minerals in the breccia include hematite, apatite, zircon, rutile/ilmenite and varying amounts of monazite. These grains can be either euhedral, rounded, fractured or fragmented, with granulation and shatter planes (Davy & Whitehead, 1968; Whitehead, 1976). The breccia ranges from a few metres to 300+ m thick.

In part, the RRB is intruded by Fe ± Mn ± U-rich fluids associated with uranium mineralisation hosted in hematitic breccia, positioned at or near the unconformity with the underlying Mesoproterozoic basement. Some of the RRB is intruded by the later epithermal ‘nail-hole’ quartz phase of the MGS. Hematitic and chloritic alteration is abundant in the breccia along with silicification, and significant potassium alteration (Drexel & Major, 1987, 1990; Weisheit et al., 2013) that may have an authigenic component formed by the precipitation of potassium feldspar during weathering (Steveson, 1978).

The STM (Hore et al., 2020) differs lithologically from the majority of the RRB in that it contains exotic and unmetamorphosed clasts. To date, recognition of the STM is limited to the immediate Mt Gee area where its competent lithology is highlighted within the RRB. U–Pb zircon geochronology of the STM has provided a maximum depositional age of ca 220 Ma (Hore et al., 2015, 2020) and U–Pb monazite geochronology recognised age populations of ca 370 Ma, ca 430 Ma and ca 487 Ma (Hore et al., 2020). The STM is correlated with the lower Aptian Sheehan Tillite Member (Alley et al., 2020). As such, the dominant glacial facies of the RRB is interpreted to be Early Cretaceous.

Previous geochronology of MGS, RRB and STM

Published geochronology of the MGS, RRB and STM is presented in Table 1 and summarised below.

Table 1. Summary of previously published age data for hydrothermal–epithermal rocks and breccias associated with the MGS, RRB and uranium mineralisation, and for the Sprigg Tillite Member.

Rock type	Age (Ma)	Type	Analyses (n)	Reference
Mount Gee Sinter (Sunshine Pound)	ca 426	U–Pb monazite	1	Elburg et al. (2003)
Mount Gee Sinter	ca 220	Isochron (2 point)	1	Elburg et al. (2013)
Radium Ridge Breccias (U-ore)	400 ± 50	U–Pb U-ore	1	Kleeman (1946)
	ca 360	U–Pb U-ore		Cooper et al. (2008) recalculation of Kleeman (1946)
Radium Ridge Breccias (U-ore)	390 ± 50	U–Pb monazite	2	Pidgeon (1979)
Radium Ridge Breccias (U-ore)	440 ± 50	U–Pb monazite	1	Pidgeon (1979)
	650 ± 100	U–Pb monazite		Youles (1981) recalculation of Pidgeon (1979)
U-ore	ca 462	U–Pb monazite	1	Elburg et al. (2003)
U-ore (Mount Gee)	ca 290	Uraninite (chemical age)	5	Bogacz et al. (2007)
Radium Ridge Breccias (U-ore)	362 ± 2	Re–Os molybdenite	3	Skirrow et al. (2011)
Radium Ridge Breccias (U-ore)	355 ± 5	U–Pb monazite	8	Elburg et al. (2013)
Davidite (Mt Gee West)	286 ± 6	Pb–Pb thorite	1	Brugger et al. (2011)
Brannerite (Hidden Valley)	367 ± 13	U–Pb	8	Wülser (2009)
Sprigg Tillite Member	315 ± 9	U–Pb zircon (max age)	175	Brugger et al. (2011)
Sprigg Tillite Member	264 ± 3	U–Pb zircon (max age)	222	Hore et al. (2015)
Sprigg Tillite Member	221 ± 3	U–Pb zircon (max age)	428	Hore et al. (2020)
Sprigg Tillite Member	344 ± 5	U–Pb monazite (max age)	60	Hore et al. (2020)

Uraniferous hematitic breccias samples of RRB were dated by Pidgeon (1979), and monazite concentrates returned a bimodal U–Pb age of 390 ± 50 Ma and 440 ± 50 Ma. Youles (1981) later reinterpreted Pigeon’s isotopic age to 650 ± 100 Ma to support his interpretation that the RRB was a Sturtian tillite, with an unexplained possible reset age of 280 ± 40 Ma.

A monazite separate from a quartz–hematite sample of MGS collected 4.5 km south of Mt Gee returned a ²⁰⁶Pb/²³⁸U age of ca 426 Ma (Elburg et al., 2003). The suggested interpretation of this age by Elburg et al. (2003) is that the monazite may have been derived from the host granites recrystallised during a magmatic–hydrothermal event.

Elburg et al. (2003) also analysed monazite from an Fe–Si sample collected from the Mt Painter No. 2 Working, Radium Ridge, and reported a ²⁰⁷Pb/²⁰⁶Pb age of ca 462 Ma interpreted to represent a phase of uranium mineralisation.

Subsequently, Elburg et al. (2013) reported LA-ICPMS analyses of monazite (n = 8) hosted in hematitic breccia sample, collected ∼1 km west of the Mt Painter No. 2 Working. These gave a U–Pb age of 355 ± 5 Ma and interpreted to represent a phase of iron–U–REE mineralisation. At Mt Gee, a poorly constrained age of a single 219 ± 25 Ma 2-point isochron of quartz–fluorite is interpreted to represent the age of the MGS hydrothermal system overprinting the 355 ± 5 Ma phase of iron–U–REE mineralisation (Elburg et al., 2013).

Skirrow et al. (2011) calculated an Re–Os age of 362 ± 2 Ma for molybdenite collected from an Armchair Prospect drill-core breccia sample and interpreted this to represent a phase of uranium mineralisation. Kleeman (1946) produced a U–Pb age of 400 ± 50 Ma for samarskite associated with monazite, hematite and brown feldspar in a quartz vein located near Mt Gee, which was recalculated by Cooper et al. (2008) to ca 360 Ma using current decay constants. In the same area, Brugger et al. (2011) produced an age of 284 ± 25 Ma from a davidite crystal hosted in a Delamerian pegmatite. Similarly, Bogacz et al. (2007) determined a chemical age of ca 290 Ma for uraninite hosted in the RRB. Wülser (2009) analysed pegmatite-hosted brannerite at ‘Hidden Valley’, for which a discordia line gave an upper intersect at 367 ± 13 Ma and a lower intercept at 21 ± 39 Ma.

Based on paleomagnetic data, Idnurm and Heinrich (1993) interpreted a Permo-Carboniferous age for the RRB, MGS and STM. The poles that plot close to the Permo-Carboniferous segment of the Australian polar wander path, however, also intersect the Cretaceous pole path (Idnurm & Heinrich, 1989). Analysis of zircon separated from the STM has returned ages as young as ca 220 Ma (Hore et al., 2020).

Sample descriptions

The main purpose of this study is to provide constraints on the timing of the MGS hydrothermal/epithermal process and the brecciation, sedimentation and fluid flow of the RRB.

Accordingly, we have chosen five float samples of MGS quartz collected from locations near the type section of Mt Gee (Figure 2) (NB: Mt Gee is a registered Geological Monument and therefore was not directly sampled from outcrop). These samples can be correlated with outcrops on Mt Gee. Eight samples of the RRB collected from various outcrops (Figure 2) have also been analysed. Table 2 provides a summary of the geochronological samples from this study and their locations.

Table 2. Summary of MGS and RRB geochronological samples, their location and LA-ICPMS analysis format (thin-section or mount).

Sample no.	Rock type	GDA94 mE	GDA94 mN	Location	Thin-section or mount
2065343	MGS	340709	6654945	Mt Gee area	Thin-section
2065345	MGS	340405	6655337	Mt Gee area	Thin-section
2138303	MGS	340260	6653920	Mt Gee area	Thin-section
2138311	MGS	340260	6653920	Mt Gee area	Thin-section
2413921	MGS	342081	6654513	Mt Painter	Mount
2381596	RRB	343577	6657648	Mt Ward	Mount
2381600	RRB	344225	6657682	Turquoise Prospect	Mount
2413916	RRB	344199	6655359	East Painter Prospect	Mount
2413917	RRB	342704	6656612	Smiler Prospect	Mount
2430629	RRB	340078	6653949	Mt Painter No. 6 Working	Mount
2381597	RRB	340175	6655424	Mt Gee West	Mount
2413919	RRB	340355	6654998	Mt Gee West (Boomerang)	Mount
2381598	RRB	340100	6655395	Mt Gee West	Mount

MGS sample descriptions

MGS sample 2065343

The MGS rock sample 2065343 (Figure 4a) is composed of moderately abundant small to larger (∼2–5 mm) aggregates of lustrous metallic hematite and closely associated very fine-grained red patches of an undetermined mineral, enclosed by translucent grey to white quartz.

Figure 4. Images of the Mount Gee Sinter geochronological samples (10c or $2 coin for scale or 10 cm scale bar): (a) R2065343; (b) R2065345; (c) R2138303; (d) R2138311; (e) R2413921.

The sample represents quartz–hematite–monazite hydrothermal–epithermal rock, with paragenetic assemblages: (1) of minor, early subhedral quartz crystals, (2) overgrown by abundant radiating quartz grains and associated hematite and monazite, and (3) with late radiating quartz projecting into minor vughs. The radiating quartz contains abundant fluid inclusions loosely concentrated along crystal growth faces.

Monazite is present in trace amounts as euhedral prisms that are both blocky and prismatic. Most of the monazite is within the aggregates of finer-grained quartz and variably oriented hematite crystals that range from ∼20 μm to ∼2 mm as discrete crystals or concentrated in aggregates.

The initial stages of this rock are considered to have crystallised from a single SiO₂–Fe–P–REE-bearing hydrothermal fluid. The last vugh-filling stage has a different mineral assemblage and texture, and most likely formed by circulation of a small amount of a different fluid lacking the P and REE components of the principal hydrothermal fluid.

MGS sample 2065345

The MGS rock sample 2065345 (Figure 4b) represents a quartz–hematite–monazite rock, inferred to have crystallised from SiO₂–Fe–P–REE-bearing hydrothermal fluid as open space-filling deposit. It has a paragenetic sequence: early finer-grained quartz + hematite + monazite is overgrown by later-formed quartz. The quartz displays sub-radiating internal microstructure and minute fluid inclusions.

Hematite is moderately abundant. Most occurs as variably oriented bladed crystals that range widely from ∼0.2 mm up to ∼2 mm diameter. They occur in patches within the finer-grained quartz. Monazite occurs within finer-grained quartz and hematite.

MGS sample 2138303

The MGS rock sample 2138303 (Figure 4c) represents a dominantly black and patchy white rock, composed of small patches of metallic dark grey and red hematite, in translucent pale grey to white hard siliceous matrix. The capping red ‘sediment’ appears to have had blades of metallic dark grey and reddish hematite forcefully injected into the layers. There is a final capping of Fe-poor quartz following on from the red ‘sediment’. The ‘sediment’ is suggested by Drexel and Major (1987) to be a finely laminated reddish-brown chemical sedimentary strata of silt-sized quartz with minor fine-grained hematite, possibly deposited originally as silica-rich gel. Determination of the origin of this ‘sediment’ is beyond the scope of this paper.

Hematite is moderately abundant, mostly in the finer-grained quartz, where it forms bladed crystals ranging from ∼1.5 mm down to ∼20 μm long. Crystals form within clusters and have a specific orientation resembling a mass fluid flow. Monazite is in minor amounts in the quartz–hematite-rich patches. The monazite crystals are up to 1 mm long and, like the hematite blades, have minor fracturing pre- or syn- relative to the silica matrix. Some hematite also occurs as submicron-sized specks that are concentrated in loose clouds; this hematite is responsible for the red appearance in patches through the hand sample. The ‘sediment’ is capped by Fe–P–REE-poor quartz. This quartz also cross-cuts the Fe–P–REE-rich phase.

MGS sample 2138311

The MGS rock sample 2138311 (Figure 4d) represents a dominantly red ‘sediment’ rock with blades of red hematite forcefully injected into some layers. The metallic dark grey patches include minor red hematite, in translucent pale grey to white hard siliceous matrix. There is an injection of Fe-poor and monazite-free quartz cross-cutting the red ‘sediment’. This red ‘sediment’ rock is associated with both the metallic hematite and Fe-poor quartz.

The red ‘sediment’ layers consist of microscopic hematite, quartz and monazite in ‘graded bedding’ with coarser and variably fractured grains up to 0.5 mm of hematite and monazite. The ‘sediment’ is ‘intruded’ by areas of larger bladed hematite to ∼1.0 mm, which are clumped together with a specific orientation like a mass fluid flow. Euhedral monazite (1.5 mm) are also present in minor amounts, mostly in the quartz–hematite-rich patches. Mineral abundances and textures differ: the early formed patches contain finer-grained quartz and more abundant hematite and monazite, whereas the later formed are dominated by coarser-grained quartz and contain only trace hematite and monazite.

MGS sample 2413921—Mt Painter

The MGS rock sample 2413921 (Figure 4e) is a massive dark red hematite-rich ochreous rock with rounded aggregates or pods of specular hematite. It contains angular clasts of granitic rock and Mt Gee type quartz–hematite, and vughy Mt Gee type quartz. The red hematite has some laminations and in parts siliceous silts with graded bedding and cross-bedding.

This is a silicified hematite-bearing siltstone, and the distribution and general appearance of the hematite flakes suggests a clastic sedimentary origin. One coarser-grained layer contains a higher concentration of this hematite and small angular cleavage fragments of monazite.

It is suggested that specular hematite and monazite were present in local rocks that, when eroded, supplied much of the detrital material now forming this sediment (Whitehead, 1976).

RRB sample descriptions

RRB sample 2381596—Mt Ward

Sample 2381596 (Figure 5a) was collected from an outcrop of cross-bedded glacial sediments (Figure 5b) on an approximately horizontal ridge on the flank of Mt Ward. Petrological thin-sections reveal some relatively large angular fragments 0.5 mm to several millimetres in diameter, mainly of quartz and microcline, in a matrix of arkosic sandstone composed of moderately well-sorted sub-rounded to angular grains of quartz and microcline, with minor muscovite flakes and heavy mineral grains generally within the size range 0.05–0.15 mm. The sedimentary rock contains some larger fragments of locally derived immature sediment in a matrix of moderately well-sorted more mature detrital material. The rock is cemented mainly by hematite and chlorite, and it does not appear to have been metamorphosed (Whitehead, 1976).

Figure 5. Images of the Radium Ridge Breccia geochronological samples and their outcrop location: (a) R2381596 sample Mt Ward; (b) R2381596 Mt Ward outcrop; (c) R2381600 sample Turquoise Prospect; (d) R2381600 Turquoise Prospect outcrop; (e) R2413916 sample East Painter Prospect; (f) R2413916 East Painter Prospect outcrop; (g) R2413917 sample Smiler Prospect; (h) R2413917 Smiler Prospect outcrop.

RRB sample 2381600—Turquoise Prospect

Sample 2381600 (Figure 5c) was collected from an outcrop (Figure 5d) of RRB at the Turquoise Prospect. The sampled outcrop of hematic breccia is poorly bedded but contains clasts of subangular to rounded granite, some of which have been kaolinised.

RRB sample 2413916—East Painter Prospect

Sample 2413916 (Figure 5e) was collected from the base of a 50 m-high outcrop (Figure 5f) of hematitic breccia (mass flow glacial sediments) from along Heighty Creek.

This is a sedimentary breccia in which the larger clasts were derived primarily from granitic or gneissic rocks. Similarly, the matrix is composed mainly of quartz and microcline but differs in composition in that it contains an abundance of hematite mostly of clastic origin. This coarse breccia contains large clasts up to 1 m in size. Some of these larger clasts are angular, and others partly rounded. The clasts also contain relatively large crystals of monazite over 0.5 mm in diameter. The rock has been silicified by coarse-grained quartz (Whitehead, 1976). The matrix is siltstone with an abundance of specular hematite crystals and fragments. It contains scattered quartz grains and/or fragments 0.05–0.15 mm in size, a few angular fragments of microcline and monazite. Some of the quartz grains in the matrix are rounded, but very few of the microcline fragments are rounded or sub-rounded. Hematite and monazite fragments appear angular.

RRB sample 2413917—Smiler Prospect

Sample 2413917 (Figure 5g) was collected from an outcrop of hematitic breccia (glacial sediments) (Figure 5h) at the northeastern area of the Smiler Prospect (Coats & Blissett, 1971). The specular hematitic breccia, with manganese coating, contains abundant basement clasts, such as gneisses and biotite schists, and rare angular pebble clasts of white quartz.

RRB sample 2430629—Mt Painter No. 6 Working

Hematitic breccia (glacial sediments) outcrops occur along the high ridge, trending approximately north and south, for about 100 m north from the open cut of the Mt Painter No. 6 Working. Sample 2430629 (Figure 6a) was collected from an outcrop (Figure 6b) of this hematitic breccia, which consists of subangular to sub-rounded quartzite and gneiss, ranging from boulders to coarse sand, in a matrix largely of hematite and feldspar.

Figure 6. Images of the Radium Ridge Breccia geochronological samples: (a) R2430629 sample Mt Painter No. 6 Working; (b) R2430629 Mt Painter No. 6 Working outcrop; (c) R2381597 sample Mt Gee West; (d) R2381597 Mt Gee West outcrop; (e) R2413919 sample Mt Gee West; (f) R2413919 Mt Gee West outcrop; (g) R2381598 sample Mt Gee West; (h) R2381598 Mt Gee West sandstone outcrop.

In thin-section, the hematitic breccia contains specular hematite, minor monazite, some euhedral quartz and rounded quartz grains. There are a few crystals, also aggregates of crystals and angular fragments of monazite up to about 0.8 mm in size. Specular hematite and monazite crystals are fractured, and many of the fragments have been at least slightly displaced relative to one another; a few appear partly rounded or abraded. The sample contains clasts of hematitic breccia and a finer-grained ‘siltstone’ that contains smaller fragments of specular hematite and a few small fragments of monazite (Hore et al., 2020).

The finer-grained bands in this specimen have all the characteristics of a fine-grained sediment. The coarser-grained hematitic breccia is composed predominantly of separate fragments (clasts) of hematite, quartz, biotite, chlorite, monazite and zircon that are immature and proximal to source (Hore et al., 2020; Whitehead, 1976).

RRB sample 2381597—Mt Gee West

Sample 2381597 (Figure 6c) was collected from an outcrop exposed in a historic drill pad area on the otherwise steep slopes near Mt Gee. The weathered and poorly layered sandstone outcrop (Figure 6d) contains occasional rounded polished lonestones, which may be glacial dropstones. This exposure is similar to other outcrops of the RRB with crude layering and an attitude following the general topography of the area. After deposition and burial, low-grade recrystallisation of the rock produced fine-grained non-foliated sericite + minor hematite matrix replacement; this occurred at P–T conditions equivalent to the lower greenschist facies and may be due to the intruding late-stage MGS noted in outcrop.

The outcropping rocks represent a range of clastic sediments, including siltstone, sandstone and granule breccia. Some parts of the heterogenous rock have an ultra-fine-grained texture, whereas other lithologies are coarse-grained and gritty tillites, which contain fragments ranging up to several centimetres in diameter. Uncommon smaller angular lithic fragments of micrographic K-feldspar (orthoclase) and quartz are derived from a massive felsic unmetamorphosed granitoid. The crystal fragments of the finer-grained siltstones and sandstone, and the matrix of the coarser sediments, are derived mainly from a sericite–hematite altered felsic crystalline rock source. The sandy material has ‘flow structures’. This suggests that the sandy material was emplaced within the granitic breccia, picking up tillite and clasts from within the tillite during this process. Whitehead (1976) suggests that the rock is a tillite that had been slightly reworked by water after deposition by ice.

RRB sample 2413919—Mt Gee West

Sample 2413919 (Figure 6e) from Mt Gee West area was collected from the informally named ‘boomerang’, a mass of hematite breccia with a curved outcrop shape resembling a boomerang (Figure 6f). The ∼10 m-thick outcrop, on the western slope of Mt Gee, extends north–south for ∼300 m just north of the Mt Gee summit (Drexel & Major, 1987, 1990).

In general, the ‘boomerang’ ranges from fine-grained clast-free near the base to clast-rich but is not clast-supported towards the upper portion. Petrological descriptions show abundant brecciated hematite blades and fractured monazite <0.1 mm grains in a siliceous matrix. These mineral clasts have been fragmented and combined with altered clasts of the quartzo-feldspathic local rock (Marathon Resources Limited, 2008). Late-stage MGS cross-cuts the fine-grained siliceous hematitic rock. The ‘boomerang’ is interpreted to be silicified glacial sediments (Hore et al., 2020).

RRB sample 2381598—Mt Gee West

Sample 2381598 (Figure 6g) was collected from an outcrop of brown to purple sandstone–siltstone with a massive to slightly tabular morphology (Figure 6h), approximately 100 m west of ‘the bench’. Locally, these sandstones and siltstones are feldspathic and form rare lenses and beds in the granitic breccia, as noted by Youles et al. (1971) and Major (1976). The sampled lens of sandstone is about 1 m thick and wide, and more than 75 m long, and is the best-known exposure of its type.

The contact between the coarse feldspathic lithology and the fine-grained sands is sharp and well defined, although Steveson’s (1975) petrological description is that it is irregular in thin-section. The sandstone is fine-grained and homogenous, although there are a few crystals that can be seen with the naked eye. The siltstone consists of detrital grains of quartz, feldspar and muscovite that have been compressed together leaving little or no intergranular matrix material. The average grainsize of the rock is 0.05–0.1 mm. The grains have well-defined irregular outlines, and there has clearly been sufficient compression of the rock to cause deformation and possibly some recrystallisation of the quartz and feldspars such that original rounded detrital shapes no longer exist, and the grains are closely packed. The sample contains small quantities of accessory heavy minerals, of which rutile is the most abundant. The sample is a siltstone that shows the effects of compression, and the rock has a compact and granular texture. Monazite grains (>50 µm) suitable for geochronological analysis were not recovered from the separation process, with only zircon available for geochronological analysis.

Analytical procedures

Geochronology, SEM, petrology

For the five MGS samples, five polished thin-sections were prepared by Adelaide Petrographic Laboratories for in situ monazite sample analyses. A further nine samples were processed for monazite (one MGS and seven RRB) and zircon (eight RRB) separation and mounting by commercial mineral separating laboratory (Geotrack International, Melbourne, Victoria). These monazite and zircon samples were mounted in epoxy resin and polished to expose grain centres.

All samples were imaged under transmitted light, back-scattered electron (BSE) imaging and cathodoluminescence (CL) using both a Phillips XL-30 and Quanta600 scanning electron microscope located at Adelaide Microscopy, University of Adelaide.

Monazite and zircon were analysed via LA-ICPMS methods at Adelaide Microscopy, University of Adelaide, as described in detail by Payne et al. (2008, monazite) and Wade et al. (2007, zircon). Monazite data were collected during nine different analytical sessions on LA-ICPMS at Adelaide Microscopy, and zircon data were collected from two different analytical sessions. Refer to online Appendix 1 and 2 for details of the monazite and zircon LA-ICPMS analytical procedures. Raw data were processed using Griffin et al. (2008), and weighted mean ages and probability plots were generated using IsoplotR (Vermeesch, 2018). Weighted mean ages for standards are given in online data files (Table S1 [monazite] and S2 [zircon]). For individual spot analyses, the 1 s uncertainty is quoted, and all other ages are given with uncertainty shown at the 95% confidence level. Details of monazite and zircon standards and results are given in online data files (Tables S3 [monazite standards], S4 [zircon standards], S5 [summary of sessions and relative monazite standard results], S6 [summary of sessions and relative zircon standard results], S7–S8 [monazite spot data] and S9 [zircon spot data]). In calculating ages for monazite data, a concordance filter of ±5% was first applied to eliminate analyses severely affected by Pb loss. Non-radiogenic Pb was also widespread in the samples resulting in spuriously old ages for some analytical spots. Careful filtering of the data has been undertaken to eliminate data points that are affected by non-radiogenic Pb and or Pb loss. Zircon U–Pb data are filtered for discordance using a threshold of ±10% discordance.

Table 3. Summary of MGS new monazite U–Pb geochronological data for six samples of the Mount Gee Sinter.

Sample no.	Mineral	General location	No. of grains	No. of analyses	Conc. analyses	Age (Ma)
2065343A	Monazite	Mt Gee area	22	60	28	352.6 ± 4.9 (n = 24)
2065343B	Monazite	Mt Gee area	39	105	29	371.3 ± 2.3 (n = 29)
2065345	Monazite	Mt Gee area	45	117	14	351.7 ± 6.3 (n = 11)
2138303	Monazite	Mt Gee area	18	32	10	322.9 ± 4.1 (n = 8)
2138311	Monazite	Mt Gee area	40	54	47	356.6 ± 3.1 (n = 47)
2413921	Monazite	Mt Painter	59	75	65	362.4 ± 1.5 (n = 65)

In addition to U–Th–Pb, REE and trace element concentrations were also determined for monazite in samples 2381596, 2381597 and 2381600 (online data file Table S10). The trace-element distribution of monazites for sample 2065343B was imaged using a Cameca SXFive Electron Microprobe at Adelaide Microscopy, University of Adelaide.

Results

A summary of new monazite and zircon geochronology data is presented in Tables 3 and 4.

Table 4. Summary of RRB new monazite and zircon U–Pb geochronological data.

Sample no.	Mineral	General location	No. of grains	No. of analyses	Conc. analyses	Age (Ma)
2381596	Monazite	Mt Ward	44	50	43	368.7 ± 1.7 (n = 41)
2381600	Monazite	Turquoise Prospect	38	50	34	364.2 ± 2.0 (n = 32)
2413916	Monazite	East Painter Prospect	33	60	46	371.2 ± 1.8 (n = 46)
2413917	Monazite	Smiler Prospect	34	60	49	363.7 ± 1.8 (n = 39)
2430629	Monazite	Mt Painter No. 6 Working	66	73	71	367.5 ± 1.4 (n = 71)
2381597	Monazite	Mt Gee West	41	50	38	367.6 ± 1.4 (n = 38)
2413919	Monazite	Mt Gee West	54	60	54	364.6 ± 1.5 (n = 54)
2381596	Zircon	Mt Ward	83	87	75	1586 ± 6 (n = 75)
2381600	Zircon	Turquoise Prospect	48	50	47	1584 ± 9 (n = 43)
2413916	Zircon	East Painter Prospect	60	60	47	1567 ± 8 (n = 45)
2413917	Zircon	Smiler Prospect	58	60	58	1576 ± 7 (n = 58)
2430629	Zircon	Mt Painter No. 6 Working	70	70	62	1596 ± 13 (n = 23)
						1700 ± 12 (n = 26)
						1854 ± 21 (n = 9)
2381597	Zircon	Mt Gee West	99	102	84	1584 ± 6 (n = 82)
2413919	Zircon	Mt Gee West	59	60	58	1573 ± 7 (n = 58)
2381598	Zircon	Mt Gee West	114	115	81	1576 ± 7 (n = 81)

MGS samples

Monazite was analysed in six samples of the Fe ± P ± REE-rich siliceous rocks of the MGS using both the in situ (thin-section) method and grain mounts from mineral separation. Electron microprobe mapping of a typical monazites from the MGS, sample 2065343B, reveals compositional homogeneity in P and also highlights the presence of both quartz (Si) and hematite (Fe) inclusions within the monazite (Figure 7). REE (e.g. La) composition appears relatively homogeneous, as do U, Th and Pb across the analysed monazite (Figure 7). However, there are zones of elevated S and Ca both within the monazites and also as obvious inclusions. These elements likely reflect the presence of both nano-scale and larger inclusions of anhydrite (or gypsum) and apatite within the monazite.

Figure 7. Electron probe micro-analysis chemical maps of two representative monazites from the Mount Gee Sinter, sample R206343B. Count scale range from blue (low) to red (high). Note the first monazite is also shown in Figure 8b, with LA-ICPMS U–Pb analytical sites indicated.

MGS sample 2065343 A

In polished thin-section 2065343 A monazite forms include small euhedral prisms, some with blocky shapes, and others with more prismatic shapes (Figure 8a). Most are very small (20–100 μm in size), with some larger crystals (∼400–800 μm) in size. Fracturing of monazite (and hematite) grains is widespread. Sixty analyses were made on 22 monazite grains from sample 2065343 A. Of the 60 analyses, 28 are ±5% discordant and yield ages that range from 754 Ma (analysis 343 A-38) to 324 Ma (analysis 343 A-12); (Figure 9a). Analysis 343 A-38 has the highest U content of the concordant analyses and very high ²⁰⁶Pb, likely indicating non-radiogenic Pb and a spurious age. The remaining analyses cluster at ca 350 Ma, although there is a spread of analyses along concordia towards older ²⁰⁶Pb/²³⁸U ages, suggesting there is an influence of non-radiogenic Pb on some of the analyses. Omitting the four of these analyses, which are statistical outliers, permits calculation of a weighted mean ²⁰⁶Pb/²³⁸U age of 352.6 ± 4.9 Ma (Figure 9b; n = 24, MSWD = 3.2, prob. = 0.00). The statistics of this weighted mean age suggest scatter in excess of a single population; however, there is not a clear break in the ages that justify a different calculation, and 352.6 ± 4.9 Ma is taken as a conservative estimate of the timing of monazite crystallisation in this sample.

Figure 8. BSE images of representative monazite from Mount Gee Sinter samples along with location of LA-ICPMS analyses and their corresponding ²⁰⁶Pb/²³⁸U ages: (a) sample 2065343 A; (b) sample 2065343B; (c) sample 2065345; (d) sample 2138303; (e) sample 2138311; (f) sample 2413921.

Figure 9. Summary of LA-ICPMS geochronology of monazite from the Radium Ridge Breccia. Wertherill Concordia and weighted mean age plots are shown for each sample. (a, b) 2065343 A; (c, d) 2065343B; (e, f) 2065345; (g, h) 2138303; (I, j) 2138311; (k, l) 2413921.

MGS sample 2065343B

Sample 2065343B is a different thin-section of the same rock type presented for sample 2065343 A. Monazite in polished section from sample 2065343B forms small euhedral prisms, some with blocky shapes and others with more prismatic shapes. Most are very small (20–100 μm) in size, with some larger crystals (400–800 μm) in size. Fracturing of monazite (and hematite) grains is widespread. Back-scatter electron images suggest the monazite is compositionally homogeneous (Figure 8b). One hundred and five analyses were completed on 39 monazite grains, of which 36 are ±5% discordant (Figure 9c). The ²⁰⁶Pb/²³⁸U ages predominantly cluster at ca 370 Ma, although some of the analyses are older, up to 565 Ma (analysis 343B_49). These older ages generally, although not universally, correlate with analyses with higher total Pb counts, for example analyses 343B_37, which has the highest total Pb content. We consider the older ages to be a result of the presence of a component of non-radiogenic Pb. Two analyses are younger than the ca 370 Ma population, 343B_67 and 343B_65, which have ages of 339.5 and 337.4 Ma, respectively. These two ages do not fall within the ca 370 Ma population. Omitting five of the oldest and two of the youngest of the concordant data as statistical outliers, a weighted mean ²⁰⁶Pb/²³⁸U age of 371.3 ± 2.3 Ma (Figure 9d; n = 29, MSWD = 3.3, prob. = 0.00) can be calculated. The MSWD suggests more than one population within the data; however, the ages occur as a continuum centred at ca 370 Ma, with no obvious break in ages to suggest where those populations might lie. Therefore, 371.3 ± 2.3 Ma represents a conservative estimate of the timing of monazite crystallisation in this sample. The two younger analyses suggest there may be some component of Pb loss that has affected this sample leading to slightly younger ages.

MGS sample 2065345

In polished section 2065345, monazite occurs in a minor amount as small euhedral crystals ∼100–200 μm, and up to ∼400 μm in size. Most occur within a single larger centimetre-sized patch of finer-grained quartz and hematite. No compositional variation is evident in the back-scatter electron image (Figure 8c). One hundred and seventeen analyses were completed on 45 monazite grains for sample 2065345. Only 14 analyses are ±5% discordant; the remainder have geologically unreasonable ages and extend towards non-radiogenic Pb compositions on the concordia diagram. Even the near-concordant analyses have a spread of ages, ranging from 537 Ma (analysis 345-11-11) to 295 Ma (analysis 345-16-01; Figure 9e) making the timing of crystallisation of monazite in this sample difficult to determine with confidence. By omitting the two oldest analyses and the youngest analysis from the 14 analyses as statistical outliers, a weighted mean ²⁰⁶Pb/²³⁸U age of 351.7 ± 6.3 Ma (Figure 9f; n = 11, MSWD = 3.43, prob. = 0.00) can be calculated. As for the previous samples, the MSWD suggests more than one population within the data. As the bulk of the analyses show evidence for non-radiogenic Pb, and only 14 analyses are near-concordant, 351.7 ± 6.3 Ma must be considered as a broad estimate of the timing of monazite crystallisation in this sample.

MGS sample 2138303

In polished section 2138303, monazite forms small euhedral prisms, with some sub-rounded and fragments noted. Monazite crystals are up to 1 mm long, and they display minor fracturing pre- or syn-intrusion of the hosting silica matrix and appear to be homogenous in back-scatter electron images (Figure 8d). Thirty-two analyses were completed on 18 monazite grains, with the majority of the data being strongly discordant and trending towards non-radiogenic composition (Figure 9g). Only 10 analyses are ±5% discordant, and these range from 366.9 Ma (analysis 303 A_19) to ca 307.9 Ma (analysis 303 A_04) and cluster at ca 330 Ma. By omitting the two analyses at ca 360 Ma the remaining eight analyses can be combined to yield a weighted mean ²⁰⁶Pb/²³⁸U age 322.9 ± 4.1 Ma (Figure 9h; MSWD = 4.4, prob. = 0.00). The high MSWD suggests this is not a statistically single population, however, and the age is taken as an estimate only.

MGS sample 2138311

Most monazite grains in polished section R2138311 are ∼0.5 mm in length with euhedral monazite up to 3 mm in size also present, all variably fractured. The BSE images suggest compositionally homogenous monazite (Figure 8e). Fifty-four analyses were completed on 40 monazite grains, the majority of which are near-concordant and cluster at ca 360 Ma (Figure 9i). Omitting the seven analyses not within the ±5% discordance filter, the remaining 47 analyses can be pooled to yield a weighted mean ²⁰⁶Pb/²³⁸U age of 356.6 ± 3.1 Ma (Figure 9j; MSWD = 2.50, prob. = 0.00).

MGS sample 2413921

Monazites from sample 2413921 were analysed in an epoxy mount from a mineral separate. The monazites are angular to sub-rounded, with those more tabular grains being up to 3 mm in length. Monazite cores exhibit a variety of internal textures, including oscillatory and chemical zoning under BSE (Figure 8f). Seventy-five analyses were completed on 59 monazite grains, with 65 analyses within the ±5% discordance filter and clustering at ca 360 Ma (Figure 9k). Several of the discordant analyses are reversely discordant, suggesting possible non-radiogenic Th is present in the monazite (Parrish, 1990). Pooling these analyses yields a weighted mean ²⁰⁶Pb/²³⁸U age of 362.4 ± 1.5 Ma (Figure 9l; n = 65, MSWD = 1.57, prob. = 0.00).

RRB samples

Eight samples representative of the RRB were collected. Monazite was analysed from seven samples and zircon from all eight samples, and a summary of the results is given in Table 4.

RRB sample 2381596—Mt Ward Prospect

Monazites from sample 2381596 display a wide range of sizes, from 50 to 300 µm in length. They are euhedral to sub-rounded, and variably fractured to fragmental. BSE images of representative monazite crystals show either internal oscillatory zoning or patchy compositional zoning (Figure 10a). Fifty analyses were completed on 44 monazite grains, of which 43 are ±5% discordant (Figure 11a). Rejecting the oldest and youngest analyses, which are statistical outliers, a weighted mean ²⁰⁶Pb/²³⁸U age of 368.7 ± 1.7 Ma (Figure 11b; MSWD = 1.50, prob. = 0.02) can be calculated from 41 analyses.

Figure 10. Back-scatter electron microscope images of representative monazites from samples of the Radium Ridge Breccia: (a) 2381596; (b) 2381600; (c) 2413916; (d) 2413917; (e) 2430629; (f) 2381597; (g) 2413919.

Figure 11. Summary of LA-ICPMS geochronology of monazite from the Radium Ridge Breccia. Wertherill Concordia and weighted mean age plots are shown for each sample. (a, b) 2381596; (c, d) 2381600; (e, f) 2413916; (g, h) 2413917; (I, j) 2430629; (k, l) 2381597; (m, n) 2413919.

Zircons from sample 2381596 are euhedral to sub-rounded with most grains displaying oscillatory zoning (Figure 12a). Eighty-seven analyses were completed on 83 zircon grains. The data form a discordia trend from ca 1585 Ma towards zero (Figure 13a). A ²⁰⁷Pb/²⁰⁶Pb weighted mean age of the 75 analyses within ±10% discordance gives an age of 1586 ± 6 Ma (n = 75, MSWD = 0.57, prob. = 1.0).

Figure 12. Representative cathodoluminesence images for zircons from samples of the Radium Ridge Breccia: (a) 2381596; (b) 2381600; (c) 2413916; (d) 2413917; (e) 2430629; (f) 2381597; (g) 2381598; (h) 2413919.

Figure 13. Wetherill concordia plots and kernel density distribution of zircon LA-ICPMS U–Pb data from samples of the Radium Ridge Breccia: (a) 2381596; (b) 2381600; (c) 2413916; (d) 2413917; (e) 2430629; (f) 2381597; (g) 2413919; (h) 2381598.

RRB sample 2381600—Turquoise Prospect

Monazites from sample 2381600 range in size from 50 to 350 µm in length, range from euhedral, equant, to sub-rounded and rounded, and are variably fractured. BSE images of representative monazite crystals show many with patchy compositional zoning, and a few with oscillatory zoning (Figure 10b). Fifty analyses were completed on 38 monazite grains, of which 34 are less than ±5% discordant and predominantly cluster at around 365 Ma (Figure 11c). Analyses M600-36 and M600-26 are not within this cluster, with ages ca 580 Ma and ca 525 Ma, respectively. These analyses are interpreted to be affected by non-radiogenic Pb, as they have Pb values of 10 540 and 570 ppm, respectively, values that are significantly greater than other analyses in this sample, with all other near-concordant analyses having <135 ppm Pb, with an average Pb value of 63 ppm (see Appendix S8). Excluding these two outliers, the remaining 32 analyses can be pooled to yield a weighed mean ²⁰⁶Pb/²³⁸U age of 364.2 ± 2.0 Ma (Figure 11d; MSWD = 0.86, prob. = 0.68).

Zircon grains from sample 2381600 range in size from 50 to 150 µm in length and are euhedral, subhedral, sub-rounded and fragmental, with most displaying oscillatory zoning (Figure 12b). Fifty analyses were completed on 48 zircon grains from sample 2381600. A discordia line trends from ca 1590 Ma, and several analyses have older ²⁰⁷Pb/²⁰⁶Pb ages, including analyses Z600-18 and Z600-7 which have ages of 1754 and 1731 Ma, respectively. These older apparent ages correspond to zircons that have generally higher ²⁰⁴Pb counts than the other more concordant analyses, suggesting that these ages may be an artefact of non-radiogenic Pb. Omitting the oldest four analyses and pooling the most concordant analyses yields a weighted mean ²⁰⁷Pb/²⁰⁶Pb age of 1584 ± 9 Ma (Figure 13b; n = 43, MSWD = 0.32, prob. = 1.00).

RRB sample 2413916—East Painter Prospect

Monazite grains from sample 2413916 range from euhedral, subangular, to sub-rounded, with some fragmental and many fractured. Euhedral grains are up to 300 by 150 µm in size, and most have either internal oscillatory zoning or patchy compositional zoning, with some grains having compositional zoning superimposed on oscillatory zoning (Figure 10c). Sixty analyses were completed on 33 grains. Of the 60 analyses, 46 are less than ±5% discordant (Figure 11e). The ages cluster at ca 370 Ma, with one analysis giving an age of 327 Ma (analysis 3916 M-52). Pooling the most concordant ages yields a weighted mean ²⁰⁶Pb/²³⁸U age of 371.2 ± 1.7 Ma (Figure 11f; n = 45, MSWD = 1.39, prob. = 0.05).

Zircons from sample 2413916 range in size from 50 to 200 µm in length and are euhedral with rounded grains and grain fragments. Approximately half of the grains display oscillatory zoning with the remainder displaying featureless uniform light grey or dark CL response (Figure 12c). Sixty analyses were completed on 60 zircon grains. Analyses range from concordant to heavily discordant and mostly cluster at ca 1570 Ma (Figure 13c). A single analysis yields a ²⁰⁷Pb/²⁰⁶Pb age of ca 2630 Ma (analysis 3916-59). Excluding the Archean grain and a single older statistical outlier (analysis 3916-60), a ²⁰⁷Pb/²⁰⁶Pb weighted mean age of 1567 ± 8 Ma (n = 45, MSWD = 0.70, prob. = 0.93) is obtained from the concordant and near-concordant analyses.

RRB sample 2413917—Smiler Prospect

Monazite grains from sample 2413917 range from euhedral to subangular and sub-rounded, with many fragments and fractured grains. Euhedral grains are up to 400 µm in length. BSE images of representative monazite crystals show that most have either internal oscillatory zoning or patchy compositional zoning (Figure 10d). Sixty analyses were completed on 34 monazite grains, of which 49 are less than ±5% discordant and have ²⁰⁶Pb/²³⁸U ages that cluster at ca 360 Ma (Figure 11g). A weighted mean ²⁰⁶Pb/²³⁸U age from these data is 361.4 ± 1.6 Ma (n = 49); however, the MSWD of 5.42 suggests this is not a single population. The four oldest analyses have four of the highest total Pb contents (analyses 3917 M-55, 3917 M-24, 3917 M-27 and 3917 M-56), suggesting that these older ages could be an artefact of non-radiogenic Pb. In addition, there appears to be a spread of ages from the cluster at ca 360 Ma to the lower limit of ca 320 Ma. This could reflect non-recent Pb loss post ca 360 Ma. Omitting the four oldest and six youngest analyses as statistical outliers results in a weighted mean ²⁰⁶Pb/²³⁸U age of 363.7 ± 1.8 Ma (Figure 11h; n = 39, MSWD = 1.27, prob. = 0.12), which represents a conservative estimate for the timing of monazite crystallisation in this sample.

Zircons from sample 2413917 range in size from 50 to 200 µm in length and range from euhedral with rounding (sub-rounded) and some grain fragments. Most grains display oscillatory zoning (Figure 12d). Sixty analyses were completed on 58 zircon grains, of which 58 are less than ±10% discordant cluster at ca 1580 Ma (Figure 13d). A ²⁰⁷Pb/²⁰⁶Pb weighted mean age of 1576 ± 7 Ma (MSWD = 0.55, prob. = 1.00) can be calculated from the data.

RRB sample 2430629—Mt Painter No. 6 workings

Monazite grains from sample 2430629 range from 50 to 300 µm in length and are euhedral, subangular, to sub-rounded, with many fragmental and fractured (Figure 17g–i). BSE images of show that most grains have internal oscillatory zoning with some displaying patchy zoning (Figure 10e). Seventy-three analyses were completed on 66 monazite grains, of which 71 are less than ±5% discordant and cluster at ca 365 Ma (Figure 11i). The two discordant analyses have older ²⁰⁶Pb/²³⁸U ages and a relatively high Pb content suggesting that non-radiogenic Pb has resulted in spurious ages. A weighted mean ²⁰⁶Pb/²³⁸U age calculation of 367.5 ± 1.4 Ma (Figure 11; MSWD = 0.69, prob. = 0.98) can be obtained from 71 analyses.

Most zircon grains from sample 2430629 are sub-rounded to rounded with some euhedral. They are up to 120 µm in diameter with a few up to 300 µm. The majority of the grains display oscillatory zoning (Figure 12e). Seventy analyses were completed on 70 zircon grains, of which 62 are less than ±10% discordant with ²⁰⁷Pb/²⁰⁶Pb ages ranging from ca 2500 Ma to ca 1547 Ma. This indicates a detrital age spectrum consistent with a sedimentary origin for the sample (Figure 13e). The youngest 23 of the concordant analyses yield a weighted mean ²⁰⁷Pb/²⁰⁶Pb age of 1596 ± 13 Ma (MSWD = 0.54, prob. = 0.96). A second population of 26 analyses gave a ²⁰⁷Pb/²⁰⁶Pb weighted mean age of 1700 ± 12 Ma (MSWD = 1.38, prob. = 0.10). A third grouping of nine concordant zircon analyses gave a ²⁰⁷Pb/²⁰⁶Pb weighted mean age of 1854 ± 21 Ma (MSWD = 0.27, prob. = 0.97). The remaining four analyses yield ages between ca 2360 and ca 2510 Ma.

RRB sample 2381597—Mt Gee West

Monazites from sample 2381597 are generally sub-rounded and subangular, with minor rounded, fragmental and euhedral grains. Sizes reach up to 400 µm in length. BSE images of representative monazite crystals show that most grains have either internal oscillatory zoning or patchy compositional zoning (Figure 10f). Fifty analyses were completed on 41 monazite grains, of which 38 are less than ±5% discordant with ²⁰⁶Pb/²³⁸U ages clustering at ca 365 Ma (Figure 11k). Several of the analyses are reversely discordant. A weighted mean ²⁰⁶Pb/²³⁸U age calculated from 38 concordant and near-concordant analyses is 367.6 ± 1.4 Ma (Figure 11l; MSWD = 1.73, prob. = 0.004).

Most zircons from sample 2381597 are generally euhedral and blocky with some displaying strong evidence of rounding, ranging in size up to 200–300 µm in length. Most of the grains display oscillatory zoning with the remainder having no zoning (Figure 12f). One hundred and two analyses were completed on 99 zircon grains. The ²⁰⁷Pb/²⁰⁶Pb ages cluster at ca 1590 Ma, and the analyses form a discordia trend with an upper intercept age of 1595 ± 6 Ma and a lower intercept of 198.3 ± 12.6 Ma (Figure 13f; MSWD = 1.4). Two analyses have ²⁰⁷Pb/²⁰⁶Pb ages ca 1685 Ma (analyses Z597-8, Z597-81) and do not fall into the main age population. Omitting the 17 analyses with discordance greater than ±10% and the two ca 1685 Ma ages, a weighted mean ²⁰⁷Pb/²⁰⁶Pb age of 1584 ± 6 Ma can be calculated (n = 82, MSWD = 0.6, prob. = 1.00).

RRB sample 2413919—Mt Gee West

Monazite from sample 2413919 ranges from 50 to 300 µm in length, with generally subangular, sub-rounded and fragmental shapes. BSE images of representative monazite crystals show that most grains have patchy compositional zoning with a minor amount displaying oscillatory zoning (Figure 10g). Sixty analyses were completed on 54 monazite grains with all except four analyses being less than ±5% discordant (Figure 11m). ²⁰⁶Pb/²³⁸U ages cluster at ca 365 Ma, and a weighted mean ²⁰⁶Pb/²³⁸U age of 364.6 ± 1.52 Ma (Figure 11n; MSWD = 0.68, prob. = 0.97) can be calculated from 54 analyses.

Most zircons from sample 2413919 are blocky, subangular to sub-rounded, up to 200 µm in length. Most of the grains display oscillatory zoning (Figure 12g). Sixty analyses were completed on 59 zircon grains, of which 58 are less than ±10% discordant with ²⁰⁷Pb/²⁰⁶Pb ages ranging between ca 1626 and ca 1526 Ma (Figure 13g). A ²⁰⁷Pb/²⁰⁶Pb weighted mean age of 1573 ± 7 Ma (n = 58, MSWD = 0.46, prob. = 1.00) can be calculated from 58 analyses.

RRB sample 2381598—Mt Gee West

No monazite was obtained from sample 2381598; zircon grains are generally subangular to sub-rounded, and up to 100 µm in length, with some well rounded. Most of the grains display oscillatory zoning (Figure 12h). One hundred and fifteen analyses were completed on 114 zircon grains, of which 81 are less than ±10% discordant and have ²⁰⁷Pb/²⁰⁶Pb ages that cluster at ca 1580 Ma. A discordia line calculated from all the data gives an upper intercept of 1582 ± 7 Ma and a lower intercept of 238.0 ± 31.1 Ma (Figure 13h; n = 114, MSWD = 1.5, prob. 0.00). A ²⁰⁷Pb/²⁰⁶Pb weighted mean age calculated from the most concordant analyses is 1576 ± 7 Ma (n = 81, MSWD = 0.85, prob. = 0.82).

Trace element chemistry of monazite from the RRB

Monazite grains in three samples of RRB (2381596, 2381597 and 2381600) were analysed by LA-ICPMS for trace element composition (27 elements in total; Table S10). Monazite from the RRB show very steep REE trends, with enrichment in light REE, a modest negative Eu anomaly and low heavy REE (Figure 14a). The REE pattern for the RRB is steeper than other examples of monazite, such as metamorphic monazite from Mount Woods Inlier (Gawler Craton; Forbes et al., 2011) or a magmatic/hydrothermal monazite within carbonatite (Eureka carbonatite, Namibia; Broom-Fendley et al., 2020). The light REE content of the RRB monazites is more similar to an example of hydrothermal monazite from an iron oxide–coper–gold deposit (Prominent Hill Mine, Gawler Craton; Forbes et al., 2015) compared with examples of metamorphic or magmatic/hydrothermal monazite. The RRB monazites have a slightly lower La abundance than the hydrothermal monazite example, but generally more than metamorphic or the magmatic/hydrothermal monazites (Figure 14b).

Figure 14. Summary of LA-ICPMS chemistry of monazite from samples 2381596, 2381597 and 2381600 of the Radium Ridge Breccia. Comparison monazite data from Forbes et al. (2011, 2015) and Broom-Fendley et al. (2020). (a) Chondrite-normalised rare earth element diagram. Chondrite normalisation factors after McDonough and Sun (1995). (b) La vs La + Ce plot. Values in ppm.

Discussion

Geochronology

The total numbers of concordant age analyses presented in this study are 193 monazite ages for the MGS, 335 monazite ages for the RRB and 512 zircon ages for the RRB. Three significant results emerge from these new data: (1) monazites within the Fe-rich phase of the MGS have a mean age of ca 363 Ma; (2) monazites within the RRB display a maxima at ca 367 Ma; and (3) zircons in the RRB are dominated by ca 1585 Ma ages.

Timing of monazite crystallisation within the MGS and RRB

Monazite data from the MGS are notable for the degree of discordance and reverse discordance, and evidence for non-radiogenic Pb incorporated within the monazite grains. As the LA-ICPMS system utilised in this study cannot correct for non-radiogenic Pb owing to isobaric interference from trace Hg in the carrier gas and the low concentrations of the non-radiogenic isotope ²⁰⁴Pb, this has been a limitation on the accuracy of the ages produced. Nevertheless, an attempt has been made to filter out the least reliable age estimates and to pool data with minimum ²⁰⁴Pb counts and/or total Pb content. Despite the filtering, the population statistics for many of the weighted mean ages calculated in this study are not consistent with simple populations, as the scatter observed from the data is greater than expected if it was solely a function of analytical uncertainty. Rather, there are geological reasons why the data do not fall into statistically acceptable age populations.

First, it is important to emphasise that the textures of the MGS are consistent with banded space-filling deposits. They are mostly fine-grained and show evidence of formation in an epithermal environment. The MGS and RRB samples have abundant monazite that is universally single phase, with no evidence for overgrowth textures or resorption features such as might be expected from multi-phase hydrothermal systems or recrystallisation during subsequent events. Therefore, texturally the monazites from both these rock types should be expected to produce a single age or very limited range of ages. While most of the analyses cluster between ca 320 and ca 380 Ma, many samples contain apparently concordant analyses with ages ca 520 and older. We interpret the majority of these older ages to reflect the incorporation of non-radiogenic Pb into the monazite either during crystallisation or potentially during subsequent fluid migration through the rocks.

A key piece of evidence to support this is the presence of anhydrite as both inclusions and potentially micro- or nano-particles within the monazite (Figure 7). The presence of S in the monazite lattice creates significant lattice distortion that can enable the incorporation of ions with large radii, including rare earth elements, Mo, Zr and, critically for the present study, Pb (Broom-Fendley et al., 2020). For this reason, monazites with CaSO₄ have been noted to contain elevated non-radiogenic Pb (Krenn et al., 2011). An example of this in the present data set is from sample 2065343B, where analysis 26a is located near a higher S region of the monazite than analysis 26 b, located on the same monazite crystal (compare Figures 7 and 8b). The ages of these two analyses are not within analytical uncertainty; rather, analysis 26a has an age of 386 ± 6 Ma, and 26 b an age of 364 ± 7 Ma. We interpret the older age to reflect the incorporation of some degree of non-radiogenic Pb, and the most likely crystallisation age of the monazite is therefore ca 364 Ma.

The majority of weighted mean ²⁰⁶Pb/²³⁸U ages for the MGS and the RRB are between ca 351 and ca 371 Ma (Figure 15; Tables 3 and 4). Only one sample deviates from this range. MGS sample 2138303 yields a weighted mean age of 322.9 ± 4.1 Ma from only 10 analyses and, like many of the calculations in the present study, does not form a statistically significant cluster. Furthermore, two near-concordant analyses 11.1 and 11.2 from this sample have ages 366 ± 5 Ma and 367 ± 6 Ma, respectively, and are therefore within the range of the pooled weighted mean ages for all other samples. In detail, analyses 11.1 and 11.2 lie on the same monazite as analysis 11.3, the latter of which has a ²⁰⁶Pb/²³⁸U age of 324 ± 6 Ma. Interestingly, these three analyses have subtly different degrees of concordance, with 11.1 being 102% concordant, 11.2 being 100% concordant and analysis 11.3 being the least concordant at 105%. Therefore, it is possible to interpret the slightly lower age of analysis 11.3 as being the result of minor disturbance to the U–Pb system and therefore not to reflect the true age of monazite crystallisation, at least for this single grain. The presence of 10 analyses that are ±5% discordant and yet have ages of ca 320 Ma suggests that there is a secondary process that has affected the monazites in this sample. We note that samples 2065343B and 2138303 of the MGS and samples 2413916 and 2413917 of the RRB also have individual spot ages at ca 320 Ma, although the dominant age for these samples is ca 360 Ma. The significance of the ca 320 Ma ages is uncertain; as more than one sample preserves ca 320 Ma analyses, it is possible that this reflects a geological process such as Pb mobility subsequent to initial crystallisation. We note that in metamorphic environments, monazite has been shown to undergo recrystallisation at temperatures around 700–800 °C, depending on the composition of the host rock (e.g. Rubatto et al., 2001). However, monazite may undergo some modification at lower temperatures (∼200 to 340 °C) in an oxidising, saline fluid-rich environment (e.g. Hawkins & Bowring, 1997; Poitrasson et al., 2000).

Figure 15. Summary of weighted mean ²⁰⁶Pb/²³⁸U ages derived from monazites in samples from the Mount Gee Sinter and the Radium Ridge Breccia.

In addition, we note that some monazite analyses are reversely discordant, particularly within the samples of the RRB. Reverse discordance in monazite may reflect the presence of non-radiogenic Th in the mineral leading to accumulation of excess ²⁰⁶Pb (Parrish, 1990). Back-scatter electron images of RRB monazites show some degree of zonation, with some analyses of lighter zones having a higher Th relative to the darker zones, which could indicate the potential for non-radiogenic Th (similar results are documented in Catlos, 2013). Mobility of U could also account for some degree of reverse discordance, although the degree to which Th or U mobility has affected the current data is beyond the scope of this study.

We consider the statistical variability in each sample analysed in this study to reflect the cumulative effect of variable non-radiogenic components, minor Pb loss and possibly variable non-radiogenic Th or U mobility. Hawkins and Bowring (1997) documented a similar variability in monazite ages from Paleoproteozoic granitic dykes in the western US and noted that concordant dates from single monazite crystals do not necessarily reflect a rock’s crystallisation age but must be closely examined for possible open system behaviour, especially where there is a possibility of fluid–mineral interaction under low-temperature conditions.

With these caveats to the filtering and interpretation of the geochronology data in mind, we interpret the monazite data from the MGS to reflect a period of hydrothermal activity that occurred between ca 370 and ca 350 Ma. An average of the weighted mean ages for the MGS is 362.9 ± 7.7 Ma, which can be considered the most likely time of formation of the hydrothermal system. Likewise, the hydrothermal monazite preserved in samples from the RRB has slightly less scatter in the calculated weighted mean ages, ranging from 363.7 ± 1.8 (sample 2413917) to 371.2 ± 1.7 Ma (sample 2413916). Averaging these results gives an age of 366.9 ± 2.4 Ma for the hydrothermal formation of monazite hosted in the RRB and, within the limits of uncertainty, in the MGS.

MGS and RRB monazite geochronology

The average of the weighted mean ages for the RRB monazite at ca 367 Ma (Figure 15) is similar to the average of the weighted mean ages for the Mount Gee Sinter monazite at ca 363 Ma (Figure 15), which records an early phase of the MGS. These ages potentially correlate with the limited dating results of Elburg et al. (2013) on monazite from an uraniferous Fe-rich outcrop at Radium Ridge, which gave a U–Pb age of 355 ± 5 Ma. The ages also match the molybdenite Re–Os age of 362 ± 2 Ma representing the timing of the U mineralisation at the Armchair Prospect (Skirrow et al., 2011).

Petrologically, the monazite grains in the MGS samples are predominantly euhedral and are interpreted to be primary and formed in situ. However, we note that many are also sub-rounded, rounded, fractured or fragmental, or a combination of these (e.g. Figure 8d). These morphologies of the monazite indicate both a primary and secondary event. Monazite is associated with bladed hematite crystals within the siliceous matrix.

The secondary geological process we highlight, potentially affecting the MGS monazites at ca 320 Ma, is possibly reflected in earlier work by Wülser (2009) on pegmatitic davidite from Mt Gee West, which was interpreted to represent either (1) a major loss of its uranium content or (2) a gain of radiogenic lead mobilised from uranium minerals occurring around 290 Ma. Similarly, a chemical age of ca 290 Ma was determined Bogacz et al. (2007) for uraninite hosted in the RRB. Although not well constrained, these two analyses may be signs of remobilisation or a later hydrothermal event at ca 320 Ma or an additional event during the Permian.

RRB zircon geochronology

Zircons within the analysed rocks of the RRB display a dominant age peak at ca 1585 Ma (Figure 13), similar in age to the local Coulthard Suite granites (Fricke & Hore, 2011). Euhedral zircon grains are common, but some grains are fractured, fragmental or rounded, indicating secondary physical processes. For each sample, zircon ages tend to define a discordia line with an upper intercept at ca 1585 Ma and a lower intercept that varies from either the present day (approximately half the samples) to approximately 380–360 Ma. Resetting at this time fits with the monazite hydrothermal growth event noted in the monazites of the MGS and RRB, and therefore it is possible that some of the zircon was also affected by the Late Devonian event.

RRB sample 2430629 from the Mt Painter No. 6 Workings, contains older inherited zircon populations at ca 1700 Ma and ca 1850 Ma, plus four grains aged between ca 2360 and ca 2510 Ma. These older inherited ages are absent from all other samples analysed (except sample 2413916 with a single analysis aged ca 2630 Ma). The ca 1700 and 1850 Ma ages are not as common as inheritance at Mt Painter, but are very common further west, in magmatic rocks of the Gawler Craton. However, Fanning et al. (2003), Ogilvie (2006) and Shafton (2006) do report these, and Archean ages, from the local Radium Creek Group metasedimentary rocks. The detrital zircon ages obtained on these RRB rocks are probably the result of local basement rock input in the formation of these breccias.

Development of the MGS

We provide evidence that the MGS hydrothermal–epithermal system continued either episodically or continuously over many millions of years in the southern MPI, with a number of pulses or events of silica-rich (±Fe ± P ± REE) fluids identified. The earlier phase(s), centred on ca 365 Ma, are more akin to a hydrothermal environment with anhedral silica masses associated with Fe ± P ± REE (indicated by the presence of monazite) deposited as infill of open space created by sub-surface hydrothermal brecciation. The latter Si-rich Fe-poor phase, also known as the ‘nail-hole’ quartz phase, which cross-cuts the STM that hosts zircons as young as ca 220 Ma, is more likely linked to an epithermal environment with open space infill and euhedral quartz crystal growth. The early phase of the MGS has been positioned near-surface in relatively recent times, after the Early Cretaceous, owing to the tectonic uplift and coincident erosion. This early phase has been overprinted by a later near-surface or venting siliceous phases of the MGS that includes chemical sedimentary rock deposition and other ‘sinter-like’ textures, observed by the authors, and similarly by Brugger et al. (2011) and Drexel and Major (1990), who associated them with geysers and hot pools. Whitehead (1976) suggested that the red ‘sediment’ layers were predominantly chemical precipitates and that the fragments of finer-grained hematite and monazite at the base of some layers probably represent locally reworked material.

An epithermal period, possibly extending from the Early Cretaceous to the present day, is represented in the rock history by the ‘epithermal’ and ‘sinter-like’ phase of the MGS, and currently by the active Paralana Hot Springs, which is the only remnant of the longstanding hydrothermal activity in the Mt Painter Inlier (Anitori et al., 2002; Brugger et al., 2005; Mawson, 1927).

The naming of the MGS by Drexel and Major (1987) is misleading, with the unit having an obvious early hydrothermal quartz–iron (P ± REE) component and a latter phase with ‘sinter-like’ textures. The latter phase could also be termed ‘epithermal’; however, this term pre-supposes ascending hot water of direct magmatic origin; there is no evidence that there has been igneous activity in the Mount Painter region since Paleozoic times.

Indications are that the MGS is the result of a mixing of a siliceous hydrothermal system later overprinted by an atypical epithermal environment. Coupled with the limited ‘sinter-like’ textures, the lack of mineralisation commonly associated with epithermal systems and the possible extended duration of the system that formed the MGS, revision of Mount Gee Sinter as a name for the unit on whole may be required to dismiss any ambiguity. This could be considered in future research work.

MGS monazite

Textural evidence of monazite within the MGS suggests that monazite was present very early in the history of these rocks. The general texture suggests that the hematite and monazite crystallised in situ, but there has been some later physical rearrangement before quartz filled interstices and replaced earlier matrix. The MGS shows a few instances of primary fluid inclusion homogenisation temperatures of approximately 300 °C in monazite, indicating the temperature of monazite formation (Collins, 1977; Drexel & Major, 1990; Simpson & Topp, 1997). Primary inclusions in the early quartz formed at about 300 °C and in the vughy epithermal quartz phase in the range of approximately 110–200 °C (Collins, 1977).

Development of the RRB

The RRB is a diverse group of rocks, mostly matrix-supported and containing a variety of clast types, where the matrix can vary widely and include sedimentary facies. Part of the RRB is interpreted by Hore et al. (2020) as preserved remnants of periglacial–glacial–proglacial facies of an Early Cretaceous event. The STM (Hore et al., 2020) of the RRB correlates with the early Aptian Sheehan Glaciation (Alley et al., 2020). Also, the RRB probably includes brecciated rocks of an earlier subterranean hydrothermal brecciation event during the Late Devonian.

The RRB displays either sharp or gradational contacts with the underlying basement granites and metasedimentary rocks (at times highly fractured). The RRB unit is polygenetic, preserving many distinct overprinting phases, including intrusion of a uranium-bearing Fe–Mn fluid and by the later phase of the MGS.

A number of these unmetamorphosed sedimentary outcrops of the RRB are highly radioactive. Some contain uranium-bearing minerals such as uraninite, and secondary minerals such as torbernite. Primary mineralisation is associated with the intrusion of Fe ± Mn ± U-rich fluids above the fractured or in-situ brecciated Meso-proterozoic basement and deposited within the glacial sediments.

The later uranium mineralisation is associated with a hematitic–manganese oxide hydrothermal event, probably coeval with the late MGS event. Silicification, chloritisation and kaolinisation of the RRB are well developed. This latter uranium mineralisation in the RRB is equivalent to unconformity-style mineralisation, described from many other regions, such as within the Athabasca Basin of Canada (Alexandre et al., 2009; Cuney et al., 2003; Fayek & Kyser, 1997; Hecht & Cuney, 2000; Hoeve & Sibbald, 1978; Lorilleux et al., 2003; Reid et al., 2014).

Future additional research is required on the unconformity-style uranium model per se, as well as separating out the RRB into two disparate parts—the hydrothermally brecciated basement and the unconformably overlying glacial component—with separate nomenclature to give clarity to the recent interpretation of the RRB.

RRB monazite

The textures of RRB monazite analysed in this study are representative of what is normally seen in hydrothermal systems. However, the voluminous unpublished petrological work described by Sylvia Whitehead (e.g. 1976), describes varied textures ranging from euhedral to subhedral, fractured to fragmental and sub-rounded to rounded. Many textures indicate physical processes of fracturing and physical abrasion that are interpreted to have occurred after monazite crystallisation but have not disturbed the U–Pb signature of the original hydrothermal monazite.

Nearly all the RRB monazite crystals (and specular hematite) are fractured, and many of the fragments (mostly angular) have been at least slightly displaced relative to one another. Fractured monazite crystals may also aggregate with textures that suggest they have been locally transported and accumulated in a fluid indicating a sedimentary origin. A few crystals are abraded or sub-rounded and are found associated with sedimentary facies, which are suggested to be the result of physical weathering, transport and deposition. Where monazite crystals are observed to have formed in situ within transported granitic clasts, they retain their euhedral structure. Fluid inclusion homogenisation temperatures of primary monazite in the RRB range from 300 °C to slightly in excess of 400 °C (Collins, 1977), interpreted as the temperature of monazite formation.

Paleozoic evolution of the Southern MPI

Figure 16 illustrates the tectonic uplift history of the MPI based on thermochronology, geochronology and fluid inclusion analyses of various rocks and minerals currently exposed basement rocks of the Mount Painter Inlier as a temperature (depth) vs age diagram.

Figure 16. (a) Simplified schematic temperature (depth) vs age diagram of published thermochronology, geochronology and fluid inclusion analyses of currently exposed basement rocks of the Mount Painter Inlier. The depth is calculated assuming a geothermal gradient of 40 °C/km and 20 °C at the surface. Currently exposed basement rock was at ∼15 km depth at 500 Ma and reached the surface during the late Jurassic (ca 150 Ma). Here they remained, affected by the Early Cretaceous glacial period. A subsequent rise in basement temperature may indicate possible burial by Cretaceous and Cenozoic sediments or may reflect the epithermal Mount Gee Sinter event (modified from Weisheit et al., 2014; Wülser, 2009). (b) Diagrammatic summary at ca 440 Ma when the British Empire Granite formed in basement rocks owing to radiogenic heat while covered by ∼14 km of Adelaidean rocks. (c) Diagrammatic summary at ca 365 Ma. Tectonic uplift combined with erosion had removed >7 km of Adelaidean rocks in the preceding 80 Ma. At this time, Si–Fe–P–REE–U-rich hydrothermal fluids at 6–9 km depth formed monazite at temperatures of ∼300 °C, possibly in hydraulically fractured rocks. The driving heat source may be a continuation of radiogenic heat or from the mantle (magmatic). (d) Diagrammatic summary at ca 130 Ma. Owing to prolonged uplift and erosion, the ca 365 Ma aged monazite-rich rocks have been exposed to surficial glacial conditions and contribute to the formation of the Radium Ridge Breccia (see Figure 17).

Owing to the high concentrations of heat-producing elements in the Mount Painter basement rocks, the region is characterised by elevated thermal gradients (Sandiford et al., 1998). The elevated thermal gradients resulted in the Mount Painter basement staying above ∼300 °C for an extended period of time from the Delamerian Orogeny (ca 500 Ma) to ca 330 Ma. This is despite it residing at depths significantly shallower than 10 km, and possibly even shallower at 9 to 6 km around 360 Ma, based on an elevated thermal gradient of 40 °C/km (McLaren et al., 2002, 2006; Sandiford et al., 1998).

The formation of the ca 450–440 Ma British Empire Granite, at temperatures of at least 720–750 °C, is attributed to radiogenic heat, with the melting occurring at around 15–16 km depth, at the level of the Proterozoic granites and metasedimentary rocks (McLaren et al., 2006) (Figure 16a). Hartley (2000) obtained muscovite ⁴⁰Ar/³⁹Ar dates from granite and pegmatite that record a range of ages between 445 and 394 Ma. The interpretation is that at ca 445 Ma, the temperatures of the presently exposed rocks began cooling rapidly below ∼350 °C. The accelerated cooling phase at ca 445–394 Ma suggests that the basement rocks were denuded as a result of tectonism that caused the reactivation of basement fault zones and rapid unroofing of the MPI. McLaren et al. (2002) proposed three periods of post-Delamerian cooling; at 430 Ma, 400 Ma and ca 330–325 Ma. These cooling events of the Mt Painter region are interpreted as basement uplift resulting in a combined minimum of 6–7 km of denudation. The third period at ca 330–325 Ma is interpreted to have been moderately fast (∼4–8 °C Ma⁻¹), which would have brought the MPI below 200 °C (cooling by at least 150 °C), and is interpreted to have likely been associated with 3 km of denudation.

The ca 440 Ma thermal event and attendant hydrothermal circulation are attributed to in-situ heating via radioactive decay of K, Th and U (McLaren et al., 2006), which we suggest possibly persisted at least until 365 Ma, producing a period of hydrothermal activity and the circulation of the SiO₂–Fe–REE–P ± U-fluids producing the earliest phase of the MGS and formation of hydrothermal monazite in basement rocks (Figure 16b). This is despite relative cooling of the MPI during two intervening periods: at approximately 430 Ma and 400 Ma (McLaren et al., 2006). There is no known magmatic activity recorded in the Mt Painter region around 365 Ma. The thermochronology suggests that the currently exposed rocks in the southern MPI were at a depth of ∼9 to 6 km at ca 365 Ma (Figure 16; Hartley, 2000; McLaren et al., 2002; Pointon, 2010).

In addition to the previous geochronology (Brugger et al., 2011; Elburg et al., 2013; Skirrow et al., 2011) that correlates with our ca 365 Ma hydrothermal event and the thermochronology presented in Figure 16, there are a number of other thermal geological events documented to have occurred around 365 Ma in the immediate Mt Painter region (Hore, 2015): Rb–Sr bulk rock isochrons on a granodiorite collected from Paralana Plateau yielded an age of 375 ± 10 Ma (Compston et al., 1966); Rb–Sr bulk rock isochrons on the pegmatoidal garnet–muscovite British Empire Granite yielded an age of 375 ± 5 Ma interpreted to represent isotopic resetting (Neumann, 1996); soda leucogranite (the Needles) yielded a muscovite–whole rock Rb–Sr age of 372 ± 2 Ma (Preiss, 1995); and in granite from drill-core Skeleton #2, a regression line has an upper intercept at 1564 ± 4 Ma and a lower intercept at 348 ± 55 Ma. From this discordia trend, it is inferred that the zircons lost radiogenic Pb in the early Carboniferous (Fanning, 1991), and U–Pb isotopic measurements of zircon leachate and residual from a Yerila Granite sample from the Mount Babbage Inlier gives intercepts of 1556 ± 10 Ma and 344 ± 23 Ma (Cooper et al., 1982).

Regionally in South Australia and adjacent states, localised magmatic events have been recognised during the mid-Paleozoic, including:

• Rb–Sr date of about 370 Ma for basement granite in Moomba No. 1 (Webb, 1974).

• From a drill hole sited west of Woodgate Bore near the Strzelecki Track, one whole-rock K–Ar date on the upper basaltic lava of the Ooloo Hill Formation yielded an age of 367 ± 9 Ma (Sheard, 2012).

• Carboniferous granodiorites, the Big Lake Suite in the Warburton Basin, 250 km NE of the MPI, gave U–Pb zircon ages of 342 ± 28 and 310 ± 17 (Boucher, 2001; Gatehouse et al., 1995; Krieg et al., 1995).

• This ca 365 Ma age has been correlated with the Alice Springs Orogeny to the north and events in the Lachlan Orogen to the east (Buick et al., 2001; Haines et al., 2001). These events raise the possibility of localised regional magmatism, which may be an additional driver of Devonian hydrothermal activity in the Mount Painter area, although this is not conclusive.

Deposition and overprinting of the RRB

The average of the weighted mean age for the hydrothermal monazite preserved in samples from the RRB is ca 367 Ma (Figure 15), representing a hydrothermal event in basement rocks at a depth of ∼9 to 6 km (Figure 16c).

The RRB, including the glacial facies, is dominated by ca 367 Ma aged monazites, which are believed to have originated at considerable depth during hydrothermal fracturing in the MPI. During the Early Cretaceous, as a result of tectonic uplift coincident with erosion (Figures 16d and 17a), the RRB lay at the surface. Reworking of the RRB monazite is evident and is defined by their sedimentary and abraded textures, indicating exposure to surficial physical processes associated with flowing water. This would suggest that the sedimentary phase of the RRB is much younger than the crystallisation of the monazite. Hore et al. (2020) present the evidence for an Early Cretaceous age for the sedimentary phase of the RRB.

Figure 17. Diagrammatic summary of monazite transport derived from basement samples in Early Cretaceous glacial facies of the Radium Ridge Breccia. (a) General overview of the topography at an early phase of the Early Cretaceous glaciation. Red stars indicate the ca 365 Ma monazites that formed in conditions of ∼300 °C and in rocks at depths of 6–9 km. Tectonic uplift combined with erosion has brought those rocks to near-surface during the late Jurassic to Early Cretaceous. During the Early Cretaceous, glacial environments have produced a variety of glacial facies to form the STM and the sedimentary portion of the RRB. Local rocks, hosting the ca 365 Ma monazites, have contributed to these facies during (b) pre-glaciation, (c) glaciation and (d) deglaciation. (e) The current breccia exposures of the RRB preserve a representative Early Cretaceous landscape, which has evolved from rocks exhumed from depths of 6–9 km, at which monazites formed at ca 365 Ma within these rocks (modified from Hore et al., 2020).

Figure 17 schematically highlights the glacial facies and processes that would have incorporated Late Devonian monazites into Early Cretaceous RRB. This includes periglacial geomorphological processes (Figure 17b) including cryogenic weathering (permafrost), glacial geomorphological processes (Figure 17c) including moraines and tillite, and geomorphological processes associated with deglaciation (Figure 17d), all of which can account for the rudimentary sedimentary features found in the breccias that have incorporated the ca 367 Ma aged monazites.

We suggest that portions of the RRB are preserved, owing to induration by Fe ± Mn ± SiO₂ ± U, possibly correlating with fluid flow during the late Mt Gee siliceous event.

Conclusions

This study presents the first comprehensive U–Pb monazite and zircon geochronological data set from the MGS and RRB within the MPI. Our U–Pb isotope results indicate a localised but significant and protracted hydrothermal event during the Late Devonian.

The initial phase of the MGS occurred at ca 365 Ma with the crystallisation at a depth of ∼6–9 km of monazite in a silica–Fe-rich fluid during a hydrothermal event, which probably included hydrothermal/hydraulic brecciation. This hydrothermal event also included a U-rich phase. The latter phases of the MGS were near-surface to venting, evident by ‘epithermal’ and ‘sinter-like’ textures.

The RRB has a component that possibly reflects the sub-surface hydrothermal/hydraulic brecciation phase, but current exposures mainly reflect the Early Cretaceous glacial facies presented by Hore et al. (2020) that contain basement-derived clasts-matrix hosting ca 365 Ma aged monazite. However, monazite hosted within local granites has not been analysed, and so the presence of ca 365 Ma monazite is conjectural. The RRB is intruded by the latter ‘epithermal’ phase of the MGS, a phase that also cross-cuts the STM that contains ca 220 Ma aged zircons. The timing of the ‘sinter-like’ phase of the MGS, relative to formation of the glacial facies of the RRB, is not well constrained, but hot siliceous fluids venting into ice or a cold-water pooled environment cannot be discounted. This may account for many of the unusual textures of the MGS still to be reconciled.

Also, a late phase of the MGS may have been synchronous with the Fe–Mn ± U-rich fluid phase and the associated silicification, which intruded into the RRB forming above the unconformity with the basement (which may be fractured). This Fe–Mn ± U-rich phase likely represents a late Early Cretaceous unconformity-style uranium-mineralising event. The Fe–Mn ± U-rich fluids, silicification and late MGS silica-rich phases noted in the RRB have contributed to its preservation by rendering the breccia resistant to weathering and erosion.

The age of the monazite from this study, and inferred temperature of their formation based on fluid inclusions (Collins, 1977), correlates well with the regional basement exhumation/depth model (Figure 16). The MGS and RRB monazites analysed from present-day surface samples formed during a Late Devonian hydrothermal event at a depth of ∼9 to 6 km.

This presentation of the MGS being a protracted system supports historic claims of it being a long-lived atypical hydrothermal–epithermal system generated by radiogenic heat. The hydrothermal monazites in the RRB, being the same age as those in the MGS, in combination indicate a relatively localised but significant Late Devonian hydrothermal event within the southern MPI.

Acknowledgements

Marg and Doug Sprigg of Arkaroola Pastoral Lease and Arkaroola Tourist Complex are thanked for their long-lasting and continued support to researchers of varying fields. Published with permission of Director Geological Survey of South Australia. The first author thanks Professor Ian Plimer for many detailed discussions and enjoyable times in the field scrambling over the ‘sinter-like’ rocks. Zoë French of Department for Energy and Mining is thanked for her generous assistance in drafting of images.

David Kelsey and Sarah Gilbert of Adelaide Microscopy are thanked for their assistance and patience helping the first author with the analytical process in the SEM and LA-ICPMS with dating and mineral mapping. Justin Payne is thanked for providing the standards used for the comparative study. Geoff Fraser and Justin Payne are thanked for constructive reviews that greatly improved this contribution.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Data availability statement

The data that support the findings of this study are openly available from Figshare, Hore et al. (2020), https://doi.org/10.6084/m9.figshare.12611600