In situ LA-ICP-MS U–Pb dating of metavolcanics of Norrbotten, Sweden: Records of extended...

(2007) 163–181www.elsevier.com/locate/chemgeo

Chemical Geology 240

In situ LA-ICP-MS U–Pb dating of metavolcanics of Norrbotten,Sweden: Records of extended geological histories in

complex titanite grains

C.D. Storey a,c,⁎, M.P. Smith b,c, T.E. Jeffries c

a Department of Earth Sciences, University of Bristol, Wills Memorial Building, Queen's Road, Bristol, BS8 1RJ, UKb School of the Environment, University of Brighton, Lewes Road, Brighton, BN1 4GJ, UK

c Department of Mineralogy, The Natural History Museum, Cromwell Road, South Kensington, London, SW7 5BD, UK

Received 23 October 2006; received in revised form 19 January 2007; accepted 1 February 2007

Editor: P. Deines

Abstract

In situ U–Pb dating of a variety of mineral phases is an important goal in petrology. This study reports data chiefly from titanite,but also from rutile and apatite, obtained using the laser ablation (LA)-ICP-MS methodology on polished thick sections in order toretain as much petrologic information as possible, and allowing trace element analyses from adjacent areas to the U–Pb analyses.The samples analysed come from Svecofennian intermediate to acid volcanic rocks of the Porphyry Group within the major ironore province of Norrbotten, northern Sweden, where titanite is a common phase associated with the mineralisation. Using a laserbeam of 30–45 μm in diameter, accurate and relatively precise (2–3% 2σ on 206Pb/238U ratios) data can be obtained on titanite.Similar precision on Pb/U ratios can be obtained in rutile, but accuracy cannot be assessed because of the lack of well characterisedstandards. However, the data are a priori accurate as they are geologically reasonable. Apatite shows reverse discordance, whichcould be explained by a number of scenarios. However, the 207Pb/206Pb ages are fairly precise (1–3% 2σ) and the calculated age isgeologically reasonable, suggesting that the data may be accurate.

The titanite grains studied show complex internal structures. Rare earth element (REE) analysis by LA-ICP-MS demonstratesthat the core and rim zones are distinct and supports a model for two-stage evolution in two out of three samples, with the rims inthese samples being enriched in total REE, and particularly in LREE in one sample, and displaying a positive Y-anomaly incontrast to the cores. In situ U–Pb analysis reveals that core zones from all samples retain distinct older ages of c.2050 Ma, whereasthe reworked rims have either distinct younger or strongly reset ages. The older cores are interpreted as representing the firstalteration/metamorphism of the volcanic pile. Titanite rims record a U–Pb age of 1870±24 Ma (2σ) at Luossavaara, and 1826±15 Ma (2σ) at the smaller Fe oxide–apatite body at Gruvberget. In both cases, the ages are consistent with previous agedeterminations from the Fe oxide–apatite deposits. A sample from Malmberget records a slightly different scenario, where U–Pbages spread along concordia from c.1920 to 1708 Ma. We attribute this pattern to metamorphic reworking of primary metasomatictitanite. Rutile ages from inclusions within the core of Gruvberget titanite essentially mimic the ages obtained from the surroundingtitanite and suggest that they have undergone the same history. Apatite inclusions within titanite at Malmberget record a youngerage of 1584±12 Ma (2σ), which we attribute to later events known to have affected the area.

⁎ Corresponding author. Department of Earth Sciences, University of Bristol, Wills Memorial Building, Queen's Road, Bristol, BS8 1RJ, UK.Tel.: +44 117 9545377; fax: +44 117 9253385.

E-mail address: [email protected] (C.D. Storey).

0009-2541/$ - see front matter © 2007 Elsevier B.V. All rights reserved.doi:10.1016/j.chemgeo.2007.02.004

mailto:[email protected]

http://dx.doi.org/10.1016/j.chemgeo.2007.02.004

164 C.D. Storey et al. / Chemical Geology 240 (2007) 163–181

These data indicate that the Porphyry Group volcanics must be at least c.2050 Ma in age, in contrast to previous agedeterminations from zircon which suggested a deposition age of c.1960 to 1880 Ma. The large grain size of the titanites studied,coupled with the ability for the first time to obtain in situ U–Pb ages from small areas, has allowed probing of this hithertounrecognised event. Smaller zircon grain size, coupled with extremely saline hot, pressurised (high P–T) metasomatic fluids,relating to regional magnetite–apatite mineralisation in the period 1900 to 1800 Ma, may have resulted in resetting of zircon U–Pbsystematics.© 2007 Elsevier B.V. All rights reserved.

Keywords: Laser ablation; Titanite; U–Pb; In situ; ICP-MS; Norrbotten

1. Introduction

Since the advent of laser ablation (LA)-ICP-MS, anumber of investigations have studied the potential ofU–Pb geochronology by this method (e.g. Fryer et al.,1993; Hirata and Nesbitt, 1995; Horn et al., 2000; Liet al., 2001; Fernandez-Suarez et al., 2002; Horstwoodet al., 2003; Jeffries et al., 2003; Tiepolo, 2003; Smithet al., 2004b; Hirata et al., 2005; Simonetti et al., 2005).These studies have been directed mainly towardsanalysing zircon, with the exception of Horstwoodet al. (2003), who also presented some data frommonazite and xenotime and Smith et al. (2004b), whoanalysed columbite–tantalite. One of the major issuesrestricting the advancement towards routine analysis ofother minerals for U–Pb LA-ICP-MS geochronology isthe problem of common Pb correction. This issue hasbeen recently addressed (Horstwood et al., 2003; Hirataet al., 2005; Storey et al., 2006) and now offers theprospect of analysing phases with elevated common Pb.Another issue is the necessity, or otherwise, for matrixmatching external standards for U–Pb LA-ICP-MSgeochronology. This issue has also been addressedrecently (Horstwood et al., 2003; Smith et al., 2004b;Storey et al., 2006) and it seems possible that strictmatrix matching may not be necessary. This isencouraging, since identifying and characterising suit-able, homogeneous standards for U–Pb LA-ICP-MSgeochronology is not simple, and opens up the attractiveproposition of analysing phases other than zircon, forwhich well characterised standards are available. Storeyet al. (2006) addressed in some detail the possibility ofU–Pb dating of titanite by LA-ICP-MS and showed thatthis was an achievable goal. In the study of Storey et al.(2006), zircon was used to externally normalise the Pb/U ratios in titanite and a common Pb correction wasperformed by measurement of the stable 204Pb isotope,corrected for isobaric interference from Hg. Titanite is acommon accessory phase and one that is sensitive toreactions taking place throughout the P–T–t–x evolu-

tion of the rock. For this reason, in situ U–Pb analysiswould be ideal, linked to mineral chemistry data andenhanced imaging. It is also of great utility if analysescan be carried out in such a way as to retain as muchpetrographic information as possible so that texturalinformation and mineral chemistry data can be readilyrelated to the petrologic evolution.

In this study we present data from petrographicpolished thick sections containing chiefly titanite, butalso rutile and apatite. The titanite grains analysed arecomplex but we have developed a technique with whichwe can collect back-scattered electron (BSE) imagesfrom the grains, and perform multi in situ U–Pb isotopicand trace element analyses from contrasting areas withinindividual grains. Importantly, we demonstrate that coreand rim relationships within titanite have distinct traceelement signatures and U–Pb age. We also present datafrom rutile and apatite inclusions within titanite anddiscuss how these data complement the titanite data andrepresent probably accurate U–Pb analyses of thesephases by LA-ICP-MS. The overall aim of the study wasto investigate in detail the timing of petrogenesis andmineralisation-related hydrothermal fluid flow duringthe tectonic and geological evolution of the Fennoscan-dian Shield.

2. Geological background

The major iron ore province of northern Sweden ishosted within Palaeoproterozoic rocks, mainly Karelian(2.5–2.0 Ga) and Svecofennian (1.95–1.85 Ga) in age,which extend from northern Sweden into Finland andparts of northern Norway (Fig. 1). The geology andmetallogeny of Norrbotten have recently been reviewed(Carlon, 2000; Bergman et al., 2001). The Palaeopro-terozoic rocks comprise the Karelian Greenstone Groupand the stratigraphically overlying intermediate to acidrocks of the Svecofennian Porphyry and PorphyriteGroups. The Haparanda and calc–alkaline and alkali–calcic Perthite Monzonite granite suites intrude all these

Fig. 1. Simplified geological map of study area in Norrbotten, northern Sweden (see inset). Modified from (Bergman et al., 2001).

165C.D. Storey et al. / Chemical Geology 240 (2007) 163–181

rocks. These granitiods are related to the development ofthe Svecokarelian Orogeny from 1.9 to 1.8 Ga (Skiöld,1987). Note that we use Svecofennian here to refer to thesupracrustal rock sequence and Svecokarelian to refer tothe orogenic period (Weihed, 2004). Both the Green-stone Group and intermediate to acid volcanics hostmajor Fe oxide–apatite ore bodies, the most notable ofwhich are at Kirunavaara and Malmberget (Geijer,1931). Copper deposits occur either spatially associatedwith the iron ores (e.g. Pahtohavare (Lindblom et al.,1996)) or hosted within regional-scale shear zones (e.g.Aitik, Nautanen (Martinsson and Wanhainen, 2004)).

The samples used in this study come from theintermediate to acid volcanic rocks of the PorphyryGroup at points where they host Fe oxide–apatitemineralisation at Luossavaara, Malmberget and Gruvber-get (Fig. 1). Sample 03 Luoss 01 was taken from thefootwall to the Luossavaara magnetite body, which is anorthern extension of the Kirunavaara magnetite–apatitebody. The footwall lithology is hydrothermally altered,amygdaloidal trachyandesite lava. The trachyandesite hasbeen extensively albitised, and actinolite and magnetite

also occur as alteration phases. Titanite occurs withinamygdales accompanied by magnetite and actinolite.

The Malmberget Fe oxide–apatite deposit occursnear to Gällivare in the south of the area (Fig. 1). Itconsists of several major and many minor magnetite–apatite and hematite–apatite bodies developed in abroad structure involving at least two phases of folding(Martinsson and Virkkunen, 2004). The bodies are alsometamorphosed, with the extent of recrystallisationincreasing to the north towards the contact with a Linastage (c.1.79 Ga (Skiöld, 1988)) granite intrusion. Thehost rocks to the deposits are basic to acid Svecofennianmetavolcanics, with some similarities to those in theKiruna area that have been affected by multi-stagealteration including K-feldspar, albite, scapolite, biotite,iron oxides and tourmaline. The sample 03 Valk 1 wastaken from albitised and scapolitised felsic metavolca-nics in the hanging wall to the Välkomman magnetite–apatite body. Titanite occurs as coarse (up to 5 mm indiameter) grains intergrown with apatite and magnetite.

Gruvberget is a much smaller Fe oxide–apatite bodylocated in the Svappavaara area to the south of Kiruna

Table 1Machine conditions and protocols for LA-ICP-MS U–Pb and traceelement analyses

Laser parametersLaser NewWave Research UP213Wavelength 213 nmPulse width 3 nsEnergydistribution

Homogenised, flat beam, aperture

Pulse energy 0.01–0.1 mJ per pulseEnergy density 4 J cm−2

Focus Fixed at surfaceRepetition rate 20 HzRaster scan speed 10 μm s−1

Nominal spot sizediameter

30–60 μm (unknowns), 60 μm (standard)

ICP-MS parametersICP-MS Thermo Elemental PlasmaQuad 3 with ‘S-

option’Forward power 1350 W

Gas flows:Coolant (plasma) Ar: 13 l min−1

Auxiliary Ar: 0.8 l min−1

Sample transport He: c. 1.1 l min−1, Ar: c. 0.9 l min−1

Analysis protocolScanning mode Peak hopping, 1 point per peakAcquisition mode Time resolved analysisAnalysis duration 180 s, (c. 60 s background, 120 s signal)Dwell times:201Hg, 204Hg/Pb,206Pb,208Pb, 232Th, 238U

10 ms

207Pb, 235U 30 ms


(Fig. 1 (Martinsson and Virkkunen, 2004)). It isoverprinted by copper mineralisation in the footwall,and to a lesser extent the hanging wall to the mainhematite–magnetite–apatite vein. Both the ore body andits wall rocks are cross-cut by a series of (meta-)doleritedykes. The wall rocks are affected by albite–scapolite–actinolite–magnetite alteration, and their protoliths werebasic to intermediate metavolcanic rocks, probably ofsimilar age to the Porphyry Group around Kirunavaara.Sample 03 Gruv 22 was taken from the hanging wall tothe west of the Fe oxide–apatite vein, and consists of anintensely scapolitised metavolcanic. The sample alsocontains significant actinolite, and minor magnetite andalbite. Titanite in this sample is developed as large (up to2 cm) porphyroblasts nucleated on pre-existing rutilegrains.

3. Methods

Samples were cut into polished thick sections ofc.150 μm thickness revealing large crystals of titaniteexposed at the polished surface. Samples were studiedoptically under microscope with transmitted andreflected light, and also as uncoated specimens in aJEOL 5900LV SEM at the Natural History Museum,London (NHM) operating at low vacuum conditions tominimise charging effects. This method was chosen inorder to eliminate the need for coating the surface with aconductive metal, which could potentially contribute alarge, undesirable common Pb signal to in situ U–Pbisotope analyses; Backscattered Electron Images (BEI)were collected from grains of interest. U–Pb isotopeanalyses were undertaken first, in order to minimise thechances of common Pb contamination. Samples werethen very lightly polished and cleaned to removeexcavated material from the vicinity of laser ablation(LA) pits, then carbon coated for electron microprobeanalysis (EPMA), using a Cameca SX50 electronmicroprobe housed at NHM operating with a 15 kVaccelerating voltage and a beam current of 20 nA. Thisanalysis was required in order to gain a chemicalreference analysis against which to normalise subse-quent LA-ICP-MS trace element analyses and convertinto concentration; for this we used Ca. EPMA spotswere taken as close as possible to the U–Pb LA pits.After this the coating was carefully removed for traceelement analyses, which were performed by LA-ICP-MS at the site of each EPMA analysis. Samples werefinally gently polished and cleaned before being re-coated for further high quality electron imaging.Detailed description of methodology for LA-ICP-MSanalysis follows separately.

3.1. U–Pb isotope analysis

Sample surfaces were cleaned by carefully rinsingwith dilute (c.5%) HNO3 and deionised water. Oncefully dry, the samples were mounted in a speciallyadapted laser cell for thick sections, and loaded into aUP213 Nd:YAG laser (λ = 213 nm) port, linked to aPlasmaQuad 3 quadrupole ICP-MS housed at NHM.Operating conditions for Laser Ablation ICP-MS (LA-ICP-MS) analyses are listed in Table 1. Laser beamdiameter and line scan length (for U–Pb isotopeanalyses) are listed individually for each analysis inTable 2. Line raster scans were performed within areasof distinct BEI contrast that appeared likely to representdifferent generations of crystal growth. Also, crystalswithin different samples contained inclusions of apatiteor rutile, and some of these inclusions were chosen foranalysis. Normalisation and age calculation was per-formed off-line using the established macro spreadsheet-based procedure LAMTRACE, developed by SimonJackson of Macquarie University, Australia. Plotting of

Table 2LA-ICP-MS U–Pb isotope ratio and calculated age data

Sample/anal #

Min(1)

Description(2)

Beam diameter(μm)

Scan length(μm)

Isotopic ratios and errors (2σ) Calculated ages and errors (Ma, 2σ)

207Pb/235U Error 206Pb/238U Error Rho 207Pb/206Pb Error 207Pb/235U Error 206Pb/238U Error 207Pb/206Pb Error

03 Luoss 01oc16a05 t p 45 60 5.744 0.170 0.343 0.009 0.60 0.121 0.003 1938 57 1901 49 1978 49oc16a06 t p 30 70 5.337 0.126 0.332 0.006 0.63 0.117 0.002 1875 44 1848 32 1904 35oc16a07 t p 45 60 5.676 0.167 0.342 0.009 0.87 0.120 0.002 1928 57 1897 47 1960 29oc16a08 t d 45 60 5.584 0.175 0.333 0.007 0.70 0.121 0.003 1914 60 1854 37 1978 45oc16a09 t b 45 65 5.625 0.226 0.349 0.011 0.68 0.117 0.003 1920 77 1928 62 1911 57oc16a10 t b 45 60 5.264 0.149 0.329 0.006 0.65 0.116 0.003 1863 53 1831 35 1899 41oc16a11 t d 45 60 6.708 0.305 0.382 0.011 0.90 0.127 0.003 2074 94 2084 59 2063 48oc16a12 t p 45 60 5.345 0.197 0.331 0.008 0.32 0.117 0.004 1876 69 1842 44 1913 71oc16a13 t b 45 65 5.540 0.131 0.337 0.007 0.75 0.119 0.002 1907 45 1873 41 1944 31oc16a14 t b 45 60 5.435 0.164 0.340 0.009 0.66 0.116 0.003 1890 57 1886 50 1895 44oc16a16 t b 45 60 5.606 0.151 0.345 0.007 0.32 0.118 0.003 1917 52 1912 37 1923 53mr24b05 t b 45 70 5.815 0.181 0.322 0.008 0.76 0.131 0.003 1949 61 1801 46 2109 43mr24b06 t d 45 55 5.656 0.115 0.346 0.008 0.71 0.118 0.002 1925 39 1917 42 1932 31mr24b07 t d 45 60 6.260 0.312 0.354 0.018 0.91 0.128 0.003 2013 100 1955 98 2072 43mr24b08 t d 30 50 6.494 0.280 0.373 0.014 0.86 0.126 0.003 2045 88 2045 77 2045 45mr24b10 t b 30 50 5.776 0.123 0.342 0.007 0.45 0.123 0.003 1943 41 1895 38 1994 43mr24b11 t d 45 60 6.183 0.149 0.362 0.008 0.69 0.124 0.002 2002 48 1992 42 2012 36mr24b12 t d 45 67 5.516 0.113 0.337 0.009 0.78 0.119 0.002 1903 39 1872 49 1936 32mr24b13 t d 45 67 5.565 0.177 0.340 0.010 0.78 0.119 0.002 1911 61 1884 53 1939 39mr24b14 t d 45 60 5.086 0.117 0.316 0.006 0.63 0.117 0.002 1834 42 1770 32 1906 35mr24b15 t d 45 67 5.763 0.157 0.347 0.010 0.82 0.120 0.002 1941 53 1921 54 1962 33mr24b16 t b 45 60 5.935 0.180 0.362 0.011 0.85 0.119 0.002 1966 60 1993 59 1939 32

03 gruv 22mr25a05 t b 45 60 5.044 0.107 0.332 0.009 0.76 0.110 0.002 1827 39 1847 48 1804 30mr25a06 t b 45 50 5.114 0.126 0.334 0.008 0.86 0.111 0.001 1838 45 1856 47 1818 24mr25a07 t d 45 60 4.981 0.145 0.320 0.009 0.88 0.113 0.002 1816 53 1791 52 1845 26mr25a09 t d 30 50 4.783 0.106 0.319 0.006 0.30 0.109 0.003 1782 39 1783 31 1780 42mr25a10 t b 45 80 4.832 0.114 0.317 0.009 0.90 0.110 0.001 1791 42 1777 48 1806 21mr25a11 t b 45 80 4.994 0.125 0.322 0.008 0.86 0.112 0.001 1818 45 1801 46 1838 24mr25a12 t d 45 50 6.399 0.292 0.365 0.017 0.91 0.127 0.002 2032 93 2003 95 2061 40mr25a14 t b 45 50 5.069 0.110 0.327 0.007 0.75 0.113 0.002 1831 40 1823 42 1840 29mr25b12 t b 45 50 4.938 0.175 0.320 0.012 0.81 0.112 0.002 1809 64 1787 65 1833 40mr25b13 t b 45 45 5.139 0.210 0.329 0.013 0.84 0.113 0.003 1843 75 1834 72 1853 42mr25b14 t b 45 56 4.893 0.307 0.321 0.019 0.93 0.111 0.002 1801 113 1793 108 1810 41mr25b15 t d 45 70 5.448 0.250 0.335 0.011 0.61 0.118 0.004 1892 87 1864 59 1923 70mr25b16 t b 45 60 5.292 0.164 0.340 0.008 0.66 0.113 0.003 1868 58 1889 46 1844 43mr25a13 r 45 50 5.041 0.175 0.329 0.012 0.72 0.111 0.003 1826 63 1835 64 1816 47

(continued on next page)

167C.D.Storey

etal.

/Chem

icalGeology

240(2007)

163–181

Table 2 (continued )

Sample/anal #

Min(1)

Description(2)

Beam diameter(μm)

Scan length(μm)

Isotopic ratios and errors (2σ) Calculated ages and errors (Ma, 2σ)

207Pb/235U Error 206Pb/238U Error Rho 207Pb/206Pb Error 207Pb/235U Error 206Pb/238U Error 207Pb/206Pb Error

03 gruv 22mr25a16 r 45 57 5.204 0.164 0.333 0.011 0.61 0.113 0.003 1853 58 1853 60 1853 52mr25b07 r 45 60 6.270 0.431 0.354 0.014 0.69 0.128 0.006 2014 139 1956 75 2074 105mr25b08 r 45 60 5.391 0.180 0.342 0.017 0.20 0.114 0.006 1883 63 1898 94 1868 100mr25b10 r 45 50 5.600 0.268 0.346 0.016 0.69 0.117 0.004 1916 92 1914 87 1918 71

03 valk 1mr29a05 t b 45 70 4.365 0.110 0.301 0.008 0.77 0.105 0.002 1706 43 1696 43 1718 30mr29a06 t b 45 60 4.474 0.098 0.306 0.011 0.74 0.106 0.003 1726 38 1723 62 1730 43mr29a07 t b 45 60 4.574 0.142 0.306 0.009 0.87 0.108 0.002 1744 54 1721 52 1773 28mr29a08 t b 45 85 4.491 0.122 0.303 0.009 0.85 0.107 0.002 1729 47 1707 52 1756 28mr29a09 t b 45 70 4.762 0.112 0.314 0.007 0.72 0.110 0.002 1778 42 1762 42 1797 32mr29a10 t b 45 60 5.236 0.161 0.337 0.010 0.85 0.113 0.002 1858 57 1873 54 1841 30mr29a11 t d 45 65 4.990 0.112 0.326 0.008 0.61 0.111 0.002 1818 41 1819 42 1816 37mr29a13 t b 45 55 4.847 0.206 0.320 0.012 0.86 0.110 0.002 1793 76 1790 66 1796 39mr29a14 t b 45 60 5.023 0.157 0.327 0.007 0.75 0.112 0.002 1823 57 1821 39 1825 38mr29a16 t d 45 65 5.487 0.165 0.349 0.013 0.84 0.114 0.002 1899 57 1929 69 1865 36mr30a05 t b 45 65 4.754 0.097 0.314 0.006 0.53 0.110 0.002 1777 36 1760 32 1796 34mr30a06 t p 45 55 5.267 0.134 0.341 0.010 0.79 0.112 0.002 1864 47 1892 56 1831 33mr30a07 t d 45 35 5.606 0.197 0.345 0.013 0.87 0.118 0.002 1917 67 1910 74 1924 37mr30a09 t b 45 60 4.988 0.126 0.324 0.008 0.84 0.112 0.002 1817 46 1807 42 1828 25mr30a10 t p 45 70 5.137 0.262 0.325 0.012 0.88 0.115 0.003 1842 94 1813 68 1875 48mr30a14 t d 45 75 6.780 0.290 0.388 0.014 0.85 0.127 0.003 2083 89 2111 79 2055 47mr30a16 t b 45 60 4.862 0.157 0.323 0.009 0.88 0.109 0.002 1796 58 1802 51 1787 27mr30b12 t d 45 60 4.526 0.086 0.304 0.005 0.73 0.108 0.001 1736 33 1711 26 1766 23mr30b13 t b 45 55 5.349 0.152 0.335 0.010 0.91 0.116 0.001 1877 53 1863 54 1892 23mr30b15 t d 45 70 6.705 0.177 0.374 0.009 0.63 0.130 0.003 2073 55 2048 47 2099 45mr30a08 a 45 70 3.932 0.091 0.294 0.008 0.37 0.097 0.003 1620 38 1661 45 1568 44mr30a11 a 60 90 3.894 0.138 0.292 0.011 0.78 0.097 0.002 1612 57 1650 59 1563 37mr30b05 a 60 100 4.145 0.136 0.306 0.008 0.83 0.098 0.002 1663 54 1722 45 1589 29mr30b06 a 60 95 4.255 0.089 0.318 0.008 0.79 0.097 0.001 1685 35 1779 44 1569 24mr30b07 a 60 100 4.170 0.122 0.309 0.010 0.93 0.098 0.001 1668 49 1735 57 1584 20mr30b08 a 45 100 4.114 0.104 0.302 0.006 0.70 0.099 0.002 1657 42 1699 35 1604 29mr30b09 a 60 70 4.132 0.085 0.305 0.007 0.69 0.098 0.002 1661 34 1714 38 1594 27

(1) Mineral description: t — titanite, r — rutile, a — apatite.(2) Analysis description (titanite only): b — BEI bright, d — BEI dark, p — BEI patchy.

168C.D.Storey

etal.

/Chem

icalGeology

240(2007)

163–181

Table 3LA-ICP-MS trace element analyses

Typicalprecision

BCR-2G BCR-2Gn=14

2σ Luossavaara Gruvberget Valkommen

(2σ onNIST612)

Accepted⁎ Mean Core Rim Core Rim Rutile

89Y 0.7 32.5 32.7 2.7 1436.7 139.5 1664.0 4985.2 19.0 53.9 4128.3 1193.1 0.9 9935.4 14671.8139La 0.5 25.3 24.9 1.8 76.5 54.0 5816.9 7408.5 76.0 218.9 382.2 274.7 8.1 1361.5 4222.7140Ce 1.5 53.6 41.6 2.4 381.5 111.5 13870.8 17918.3 96.0 1144.0 3231.4 1608.2 4.8 6995.0 12705.5141Pr 0.9 6.8 6.1 0.4 105.0 20.5 739.7 1314.5 10.4 83.7 394.8 175.5 0.7 806.6 1401.0145Nd 0.3 28.6 27.8 1.7 643.3 100.8 2705.3 4155.0 36.9 300.9 2240.1 949.6 2.9 4049.6 7069.3147Sm 1.3 6.7 6.4 0.5 273.0 24.8 413.2 645.3 4.1 32.9 553.4 212.8 b0.595 1014.7 1752.5151Eu 0.5 2.0 1.9 0.1 220.2 8.1 81.7 120.4 1.9 12.6 79.8 36.4 b0.162 122.9 207.5157Gd 0.3 6.8 6.5 0.6 306.7 22.3 330.8 528.2 3.1 18.9 497.9 181.5 b0.787 1000.6 1651.2159Tb 0.7 1.0 1.0 0.1 50.9 3.4 46.3 75.0 0.5 2.1 69.4 25.0 b0.106 167.4 273.4163Dy 1.5 6.4 6.5 0.7 330.1 22.2 303.8 469.6 3.2 11.8 424.3 145.5 b0.523 1098.2 1763.9165Ho 1.2 1.3 1.3 0.2 68.4 4.7 62.9 94.8 0.6 2.3 86.5 29.8 b0.071 224.3 347.3167Er 1.8 3.7 3.6 0.4 199.2 16.3 190.0 271.7 2.3 7.8 249.2 81.7 b0.515 640.6 975.2169Tm 0.8 0.5 0.5 0.1 29.5 2.5 27.8 41.1 0.4 1.3 36.7 12.3 b0.122 97.9 153.3173Yb 0.1 3.3 3.6 0.4 213.1 20.6 197.9 275.9 2.0 10.8 240.9 78.0 b0.751 680.1 1025.4175Lu 0.6 0.5 0.5 0.0 31.9 4.9 26.8 39.0 0.5 2.0 30.2 10.5 b0.116 73.3 108.5208Pb 2.4 11.8 0.9 0.3 0.3 15.6 27.4 0.29 2.1 22.9 20.1 2.3 18.2 38232Th 0.2 6.0 6.2 0.5 1.1 1.4 295.2 531.2 0.8 6.7 131.0 56.6 1.9 115.7 260238U 1.6 1.7 1.1 0.1 142.3 2.0 32.3 49.4 16.0 30.1 30.2 69.1 67.1 79.9 97

The unknowns are representative analyses from core and rim areas. Concentrations shown are in ppm.


U–Pb data was performed by ISOPLOT v.3 (Ludwig,2003), a Microsoft Excel plug-in.

Using our analytical set-up at NHM we are able todetect and correct the occurrence of common Pb in LA-ICP-MS analyses of titanite via measurement of thestable 204Pb isotope, corrected for interference from204Hg aided by the use of in-line Au traps on carrier gaslines (Storey et al., 2006). In the case of all of theanalyses presented here, it was found that after carefulscrutiny of each datum common Pb could not bedetected, apart from in one case (Section 4.1.2), andanalyses contained very strong radiogenic Pb signals.Effectively, we attempted to correct the Pb/U ratios bydeducting common Pb with model terrestrial values(Stacey and Kramers, 1975) at the age of crystal-lisation, based on the measured 204Pb, and in each casethis resulted in negative values, implying that too muchcommon Pb had been deducted. In this case it is evidentthat 204Pb in the analyses was not present above thedetection limit for our analytical set-up. Once normal-ised to zircon geostandard 91500 (Wiedenbeck et al.,1995), nearly all analyses overlapped concordia. In thecase of these particular samples, it was found that thedata did not require a common Pb correction, based onthe lack of common 204Pb and our experience ofdetecting common Pb in LA-ICP-MS U–Pb analyses atNHM (Fernandez-Suarez et al., 2002; Jeffries et al.,

2003; Storey et al., 2006). Any discordant points werefound to be best explained by Pb-loss rather than highcommon Pb content (see Section 4.1). It is suggestedthat any common Pb present for all analyses was belowdetection and unlikely to result in large U–Pb agedifference, at least within the quoted precision of ouranalyses. The time resolved plot of Pb/Pb and U/Pbratios from each analysis was studied and onlyanalyses, or coherent parts of analyses, where wecould be confident that no mixed age data, secondaryphase or potentially contaminated fracture was encoun-tered during analysis, were chosen for final agecalculation.

3.2. Trace element analysis

In situ trace element analysis was performed by LA-ICP-MS. Following EPMA, the carbon coating wascarefully removed and sample surfaces cleaned beforebeing placed back into the LA chamber. Samples wereanalysed using the same analytical equipment as that forU–Pb analysis, except that the in-line Au traps wereremoved. Analytical protocol for LA-ICP-MS analysis islisted in Table 1 for U–Pb analysis. The main differencesfor trace element analysiswere thatwe used a fixed spot of85 μm on the standard and 45 μm on the unknown, thepulse energy was 0.18 mJ on the standard and 0.05 mJ on


the unknown, and we used a repetition rate of 11 Hz; allisotopes collected (Table 3) were measured for 10 ns. Allanalyses used NIST 612 glass as a standard, and used Cadetermined by electronmicroprobe as an internal standardto convert element ratios to absolute concentrations.

4. Results

4.1. U–Pb

Altogether 84 unknown analyses of all minerals wereperformed, and of these 17 were rejected due to criteriadescribed below; mostly these were poor analyses. For

Fig. 2. Analytical results from Luossavaara (03 Luoss 01). A) U–Pb concordiand ages calculated at 2σ. Inset is a BEI of one typical grain analysed andB) Cumulative probability diagram shows bimodal distribution of 207Pb/206

C) Chondrite normalised rare earth element plot from titanite, with Yttrium

titanite specifically, of 65 analyses 10 were finallyrejected.

4.1.1. Luossavaara (03 Luoss 01)Over two sessions a total of 24 unknown titanite

analyses, of which 2 were rejected following processingand analysing of the time resolved data; both were notconsidered to be reliable analyses and may havecontained either mixtures of different age zones,inclusions of different phases or areas of Pb-loss. Twodistinct groups of ages emerge (Fig. 2A): a youngercluster of ages in the range 1950–1850 Ma representingthe majority of analyses and a distinct older cluster at

a diagram showing titanite analyses. All error ellipses are plotted at 2σ,sites within the grain that have dark BEI and retained old U–Pb age.Pb ages within titanite, demonstrating two discrete growth episodes.is plotted as a pseudo-lanthanide between Dy and Ho.


2100–2000 Ma. Some of the analyses are slightlydiscordant on chords towards modern-day Pb-loss(Fig. 2A); hence, 207Pb/206Pb ages will best representtheir true age. A cumulative probability density plot ofall 207Pb/206Pb ages (Fig. 2B) also demonstrates thebimodal nature of ages, but it also suggests a mixingeffect between the youngest ages and the older ages.Taking only the youngest ages, with 207Pb/206Pb ages ofc.1900 Ma or younger, yields a Concordia age of 1870±24 Ma (2σ). Taking all older analyses, with 207Pb/206Pbages in excess of 2000 Ma, and performing a regressionthrough zero age Pb-loss, yields an upper intercept ageof 2058±47 Ma. Considering the remainder of analyses,those lying between these two end-member groups, andregressing through zero age Pb-loss to these analyses,yields an upper intercept age of 1950± 9 Ma.

The older group of ages were all from analyses ofareas of titanite with dark BEI signal and, coupled withthe distinct peak of ages, it seems fairly apparent that theupper intercept age of c.2058 Ma is a reasonableestimate for initial crystallisation of the titanite. Theyoungest group came from BEI bright areas andgenerally close to the rims of grains and/or fractures.This young group at c.1870 Ma is likely to represent theage of titanite overgrowth and/or recrystallisation(dissolution/reprecipitation) of titanite at that time. Theintermediate group of analyses came from both BEI darkand bright areas within titanite with no clear relation-ship. The age calculated is probably meaningless andrepresents either a mix of c.2058 Ma titanite andc.1870 Ma titanite, or else represents c.2058 Ma titanitethat has undergone Pb-loss at c.1870 Ma and has beenpartially reset. We favour the latter interpretation sincedistinct age variations across sharp boundaries duringlinear ablation would be clearly visible in the timeresolved isotope ratios, and this is not the case.

4.1.2. Gruvberget (03 Gruv 22)A total of 16 titanite analyses were performed, of

which 3 were rejected due to poor analyses. The majorityof analyses form a tight cluster at c.1800 to 1900Ma, oneanalysis is distinctly older at c.2050Ma (Fig. 3A). The oldanalysis comes from a BEI dark area within titanitebetween two inclusions of rutile. A cumulative probabilitydensity plot (Fig. 3B) displays a largely bimodaldistribution, although one of the younger ages is slightlyolder than the main array at c.1900 Ma. This analysis isfrom a BEI dark area close to a rutile inclusion. Thisanalysis could represent a distinct growth event or elsecould represent amixture of c.2050Ma titanite and c.1850to 1800 Ma titanite, or old titanite with partial Pb-loss atc.1850 to 1800 Ma. A weighted average of 207Pb/206Pb

ages for the younger group of analyses results in an age of1826±15 Ma (2σ). The old analysis has a Concordia ageof 2049±28 Ma (2σ). It seems likely that initialc.2050 Ma titanite underwent reworking at c.1826 Ma.

A total of 8 rutile analyses on inclusions withintitanite were made, of which 3 were rejected as beingpoor analyses. The rutiles spread along concordia fromc.2000 Ma to c.1850 Ma (Fig. 3A). There is no clearrelationship between size of inclusion or position ofanalysis within inclusion and U–Pb age. The results arebest visualised on a cumulative probability density plotof 207Pb/206Pb ages (Fig. 3B), where they can bedirectly compared with the ages of the surroundingtitanite. It can be clearly seen that there is a strongcorrelation between the rutile and titanite ages. Ittherefore follows that the older age of rutile, similar tothe oldest age of titanite, represents the minimum age ofthe initial crystallisation of rutile, which could havebeen reset at the time of c.2050 Ma titanite crystal-lisation. The spread of rutile ages down to close to thec.1826 Ma age of titanite recrystallisation/Pb-loss and/ornew growth, suggests that the originally older rutile hasalso undergone partial Pb-loss.

4.1.3. Välkomman (03 Valk 1)A total of 25 analyses on titanite were performed, of

which 5 were rejected following processing. Of these, 4were due to poor analyses as for sample 03 Luoss 01, andone analysis contained large amounts of common Pb,potentially from a fracture or a second phase. Nearly allanalyses overlap concordia and only one is slightlydiscordant (Fig. 4A). A clear bimodal distribution of207Pb/206Pb ages can be seen in Fig. 4B, with an olderpeak (n = 2) at 2050 to 2100 Ma and an array of youngerages (n = 18) between 1900 and 1700 Ma. The distinctolder group come from BEI dark areas; one analysis wasin the centre of a large grain shielded between apatiteinclusions and one analysis was at the rim of a smallergrain in contact with matrix quartz. The two concordantolder analyses yield a Concordia age of 2073±19 Ma(2σ). The younger group come from analyses fromvarious locations within the grains and no clearrelationship emerges between BEI response, distancefrom grain boundary or inclusions and U–Pb age. Thespread of analyses along concordia could be explained byrecrystallisation and/or new growth of titanite atc.1900Ma followed by a number of scenarios: 1) isotopicclosure to Pb close to c.1900 Ma, followed by a singlethermal event at c.1700Ma above the closure temperaturefor Pb in titanite. In this scenario all analyses lie on ashallow chord between c.1900 and 1700 Ma. 2) retentionof temperatures above the closure temperature for Pb in

Fig. 3. Analytical results from Gruvberget (03 Gruv 22). A) U–Pb concordia diagram showing titanite and rutile analyses. All error ellipses are plottedat 2σ, and ages calculated at 2σ. Titanite analyses shown as fine line ellipses and rutile as solid line ellipses. Inset is a BEI showing BEI dark areawithin the grain retaining old U–Pb systematics and bright areas are rutile. B) Cumulative probability diagram shows bimodal age distribution oftitanite and rutile that are essentially identical. C) Chondrite normalised rare earth element plot from titanite, with partial data from rutile; Yttrium isplotted as a pseudo-lanthanide between Dy and Ho.


titanite following c.1900 Ma until final cooling atc.1700 Ma. In this scenario any of the apparent agessimply reflect the amount of diffusion of Pb within thetitanite at the site of ablation. 3) a series of punctuatedthermal events above the closure temperature for Pb intitanite between c.1900 and 1700 Ma. In this scenario,there may be discrete pulses of ages within the array, butthe precision of analyses does not allow their deconvolu-tion. Any of these scenarios may result in partial Pb-losswithin the grain, with Pb diffusing towards the grainmargin and/or fractures and escaping the system.However, the lack of correlation between age and distance

from grain boundary, argues for multiple diffusiondomains within the crystal, which is not best explainedby a simple, thermally activated volume diffusionmechanism. Alternatively, within the same time con-straints, the data could be explained by dissolution/reprecipitation of titanite. In this scenario, dissolution/reprecipitation event(s) could better explain the lack of asimple crystallographic control on apparent Pb diffusion.It would also better explain the lack of a recognised high-grade metamorphic event at this time to account forthermally activated Pb diffusion in titanite above theclosure temperature of c.650–700 °C (Scott and St-Onge,

Fig. 4. Analytical results from Malmberget (Välkomman; 03 Valk 1). A) U–Pb concordia diagram showing titanite and apatite analyses. All errorellipses are plotted at 2σ, and ages calculated at 2σ. Inset is a BEI showing areas within the grain containing old U–Pb and volumetrically superiorreset regions. Also shown are apatite inclusions analysed. B) Cumulative probability diagram shows bimodal distribution of titanite 207Pb/206Pb ages.C) Chondrite normalised rare earth element plot from titanite, with Yttrium is plotted as a pseudo-lanthanide between Dy and Ho.


1995; Pidgeon et al., 1996; Frost et al., 2000), particularlyin these large grains c.5 mm in diameter. In terms of thetiming of dissolution/reprecipitation events, within thelimits of precision it is impossible to discern whethersingle or multiple are responsible for the apparent spreadalong concordia. What can be stated, however, is that theonset of this event(s) is represented by the oldestconcordant (Concordia) age of 1920±23 Ma (2σ) andthe younger end of this array is represented by theyoungest concordant (Concordia) age of 1708±20 Ma(2σ). However, these ages should be considered as aminimum and maximum age, respectively, for the event.

A total of 9 analyses of apatite were made oninclusions of apatite within titanite. Of these, 2 were

rejected due to poor analyses. The apatites are allslightly reversely discordant (Fig. 4A) but clearly lie ona regression line towards modern-day Pb-loss. Theupper intercept of the discordia line, anchored through0 Ma, is 1584±12 Ma and the similar approach to takethe weighted average of 207Pb/206Pb ages yields an ageof 1583±10 Ma (2σ).

4.2. Trace element analysis

The results of trace element analyses of the titanitegrains are summarised in Table 3, and shown inFigs. 2C–4C. At both Luossavaara (Fig. 2C) andGruvberget (Fig. 3C) the trace element analyses clearly


differentiate the core and rim zones identified from BEI(Fig. 5A and B) and which define different agepopulations in U–Pb analyses. At Luossavaara the Fe-depleted core zones show low rare earth element (REE)concentrations (102 to 103 times chondritic abundance)relative to the Fe-rich, zoned rim (104 to 105 timeschondritic abundance). The REE patterns in grain coresat Luossavaara are typically relatively flat, without amarked difference in abundance between the light(LREE) and heavy (HREE) rare earths. There is a slightabundance maximum at Nd, and typically a positive Eu-anomaly (Eu/(0.5×(Sm+Gd))). In contrast the rimsshow a strong LREE enrichment with an abundance

Fig. 5. BEI's of titanite grains showing clear core and rime features andsites of LA-ICP-MS analyses. Note sites of LA-ICP-MS analyses. Theelongate black pits are line rasters for U–Pb analysis, the black spotsare for trace element analyses, which were superposed on electronprobe analyses. A) Sample from Luossavaara (03 Luoss 01). All of thecrystal displayed is titanite. Note dark core clearly altered and replacedby brighter titanite and distinct, zoned bright rim overgrowth. B)Sample from Gruvberget (03 Gruv 22) showing BEI darker titaniteimpinged and overgrown by BEI brighter titanite. The BEI very brightareas within titanite are rutile grains.

peak at La. There is a slight negative Eu-anomaly.Yttrium is plotted as a pseudo-lanthanide between Dyand Ho, and shows a positive anomaly in the rims whichis absent in the grain cores.

The core zones at Gruvberget show a similar overallrelative depletion in the REE to that seen at Luossa-vaara, but in this instance are LREE enriched, whilst stillshowing a positive Eu-anomaly. The rim REE patternsare similar to those at Luossavaara, but show a lessmarked distinction between the LREE and HREE, andLa is depleted relative to the other LREE (Ce–Sm).

Inmarked contrast to the other two samples studied theVälkomman titanites show very homogeneous REEpatterns. These have a slight increased abundance of theLREE relative to the HREE, a negative Eu-anomaly and apositive Y-anomaly. They also show depletion in Larelative to the other LREE. They share an overall highabundance of the REE with the rim zones of othermetasomatic titanites.

5. Discussion

5.1. Origin of titanite

Titanite does not occur as a primary igneous phasewithin the lavas in any of the studied samples. AtLuossavaara it occurs as a metasomatic/metamorphicphase within amygdales, at Gruvberget as a metasomatic/metamorphic overgrowth on rutile, and at Välkomman asa vein and cavity fill alongside apatite and magnetite.In all these cases, therefore, the earliest age that titaniteU–Pb systematics could be expected to record is the firstalteration and metamorphism of the volcanic pile. Rutileis the precursor phase at Gruvberget, and it also occurswithin early hydrothermal phases (rutilated quartz) atLuossavaara. Early metasomatic assemblages in amyg-dales involving rutile, quartz and calc-silicate minerals(zoisite) can react to form titanite bearing assemblages attemperatures between 300–500 °C depending on pressureand XCO2 in the fluid phase (Hunt and Kerrick, 1977).Simple reactions involving calcite, quartz and anatase orrutile can lower the conditions of formation of titaniteduring metamorphism to below 300 °C (Schuiling andVink, 1967). Titanite formation is typically limited to lowpressure (below 2 kbar) and low XCO2 in the fluid phase(Frost et al., 2000). The upper limit of titanite stability incalc-silicate bearing assemblages is given by reactionsinvolving the formation of rutile and anorthite at ∼ 450–500 °C. Although the P–T–x conditions of titaniteformation cannot be accurately established at this time,it seems likely (particularly at Gruvberget) that titaniteinitially formed during prograde metamorphism of the


Porphyry Group volcanics, possibly during burial, as aresult of autometamorphism within the accumulatingvolcanic pile. Retrograde breakdown of titanite toquartz+ rutile bearing assemblages is also observed atLuossavaara.

The trace element compositions of the core zones atLuossavaara and Gruvberget are consistent with thisinterpretation, particularly as they are associated withdistinctive older concordant (c.2050 Ma) U–Pbdomains. Recent studies of trace element partitioningbetween titanite and melts (Tiepolo et al., 2002;Prowatke and Klemme, 2005) indicate a preference forthe MREE (Sm–Gd) over the LREE and HREE withinthe titanite lattice. An additional control on titanite meltREE-partitioning is exerted by melt composition andstructure, and in a similar way fluid chemistry and henceionic speciation will have an additional control inhydrothermal fluid–titanite REE-partitioning. In theabsence of a strong fluid (melt or hydrothermal) controlon the REE distribution, equilibrium trace elementdistribution patterns in titanite should show a normal-ised abundance maximum between Nd and Gd. Positiveeuropium anomalies reflect a relatively low oxygenfugacity during crystallisation (Sverjensky, 1984; Bau,1991). We therefore interpret the core zones of thesegrains as indicative of equilibrium fractionation of theREE during metamorphism of the Porphyry Grouplavas. The core REE patterns at Gruvberget sharefeatures such as a negative Y-anomaly and variationbetween slight negative and positive Eu-anomalies withLuossavaara, but are significantly richer in the LREE.This may reflect a slightly more evolved initial sourcefor the REE than at Luossavaara, closed systemfractionation of the REE during metamorphic growth(Brugger et al., 2000; Smith et al., 2004a) or thepresence of an LREE-rich fluid during initial titanitegrowth. We reemphasise that the core zones at bothLuossavaara and Gruvberget formed at a distinctivelyearlier time (c.2050 Ma) than the rims that werereworked at c.1870 Ma and c.1826 Ma, respectively.

The rim zones of the titanite grains studied atLuossavaara and Gruvberget are clearly distinct fromthe cores. The REE content of these zones reflects thegeneral trend of trace element enrichment associatedwith Fe oxide–apatite mineralisation across the Norr-botten district, and in Fe oxide–apatite ores elsewhere inthe world, with weak to moderate LREE/HREEfractionation and negative Eu-anomaly (Frietsch andPerdahl, 1995). These zones can therefore be directlyrelated to metasomatic reworking (dissolution/repreci-pitation) on metamorphic titanite during the formationof the Fe oxide–apatite bodies.

Despite the large grain size and evident chemicalzoning of the Välkomman titanites they show remark-ably homogeneous REE patterns. Given the similarity ofthe REE pattern with the metasomatic rims of the otherdeposits, we interpret the bulk of the titanite to havebeen reworked during mineralisation. In this case, thepreservation of old concordant Pb/U ratios (2050 to2100 Ma) rather than any potentially different, earlierREE pattern similar to the cores in Luossavaara andGruvberget titanites, we can only assume is related tosampling; our adjacent trace element analysis in theselimited cases potentially are not representative of the U–Pb site of analysis. An alternative, however, is that theinitial REE pattern was fortuitously identical to the latermetasomatic overprint. The REE appear not to bedisturbed during the c.1920–1708 Ma event, which weassume to be a metamorphic overprint, based on thedeformed and metamorphosed nature of the MalmbergetFe oxide–apatite bodies.

5.2. U–Pb systematics and the utility of titanite

Titanite is a useful mineral for U–Pb geochronologysince it can incorporate 10's to 100's ppmU into its latticeduring crystallisation (Frost et al., 2000), and it has arelatively high closure temperature for the U–Pb systemof c.650–700 °C (Scott and St-Onge, 1995; Pidgeon et al.,1996; Frost et al., 2000). Moreover, it can be clearlyrelated to the P–T–t–x history of the rock (Carswell et al.,1996; Frost et al., 2000; Piccoli et al., 2000).

In this study we have demonstrated how in situmeasurement of Pb/U isotope ratios can reveal andresolve age variation within single grains of titanite.Furthermore, we have demonstrated that the agevariation correlates with trace element variation withinthese grains. This gives a high degree of confidence thatwe are measuring discrete events in the evolution of thetitanite. Moreover, the old ages preserved in titanitecores, though volumetrically inferior, are consistentacross three different samples from different deposits inthe province. It is, therefore, unlikely that the old ages inthe cores represent older labile Pb produced frombreakdown of pre-existing older phases. This affords ahigh degree of confidence that we are measuring ageologically meaningful age from these titanite cores(the details of which are discussed in Section 5.4).

Previous U–Pb geochronology from these rocks andore deposits has not suggested an old component withineither titanite or zircon; all ages fall in the range 1800–1900 Ma. However, all previous age determinationswere by ID-TIMS on whole grain or multi-grainaliquots. In the case of the titanites, the very small


volume of old titanite within the grains, which requiredcareful identification and sampling, would be massivelydiluted within a multi-grain ID-TIMS analysis. There-fore, the “average” age of the analysis would bedominated by the volumetrically far superior rims andareas of core that had undergone preferential Pb-loss. Itis considered that only such a careful in situ study wouldbe likely to reveal this enigmatic multi-age preservationof titanite and is analogous to the advent of ionmicroprobe U–Pb geochronology on zircon, wherepreviously unrecognised “events” have been deconvo-luted from complex grains (e.g. Friend and Kinny,1995). In terms of why zircon does not apparentlyrecord any older ages than c.1900 Ma, despite having ahigher closure temperature than titanite (N900 °C (Leeet al., 1997)) the answer may lie in the grain size and thefluid regime during 1900–1800 Ma reworking. Grainsizes of zircon within the felsic rocks of the PorphyryGroup are typically of the order 200×100×100 μm,whereas the titanite grains analysed here are up to 5 mmin diameter. Hence, volume diffusion of Pb withinzircon leading to resetting of U–Pb may be moreeffective simply as a result of smaller grain size. It is alsoimportant to consider that the most kinetically favour-able mechanism of isotopic re-equilibration in zircon isnot volume diffusion. Whilst there is no evidence thatmetamorphic conditions during later over-printingattained more than greenschist to amphibolite facieswithin the studied areas, there is evidence of Zrmobilisation during mineralisation within comparablerocks c.50 km WSW of Kiruna at Tjårrojåkka (Edfeltet al., 2005). Particularly pertinent is the recentdiscovery that zircon can grow during very low-gradeprehnite–pumpellyite facies metamorphism at c.250 °C,providing compelling evidence of zirconium mobilityeven at low temperatures, possibly transported asfluorine complexes in aqueous fluids (Rasmussen,2005). Recent studies of pseudomorphic replacementof sparingly soluble salts in solid solution–aqueoussolution systems have shown that direct dissolution–reprecipitation is one of the main mechanisms occurring(Putnis, 2002; Pollok et al., 2004; Putnis and Putnis,2004) and has been observed directly in zircon-meltsystems (Tomaschek et al., 2003; Tomaschek, 2004)yielding homogeneous, non-cathodoluminescent grains(Tichomirowa et al., 2005). Although zircon solubilitiesin aqueous fluids are limited (Ayers and Watson, 1991)the fluid conditions during mineralisation in magnetite–apatite deposits are extreme enough to actively promotezircon solubility even in the absence of significantfluorine concentrations in the fluid phase (c.f. Rubinet al., 1993). Tomaschek et al. (2003) suggested that

solution–reprecipitation recrystallisation of zirconwould be promoted by the presence of a high-pressureaqueous fluid phase. Typical fluids in late stage veinsassociated with the Kirunavaara–Luossavaara system,and at Gruvberget have in excess of 30 wt.% salts andwere probably trapped at temperatures and pressures inexcess of 2–3 kbars and 300–400 °C (Broman andMartinsson, 2000; Harlov et al., 2002). Further studiesinto the textures, trace element composition and oxygenisotope systematics of zircons may shed more light onthis problem.

5.3. Rutile and apatite U–Pb geochronology byLA-ICP-MS

In situ U–Pb geochronology of rutile and apatite ispresently limited. Pb–Pb dating of carbonado (poly-crystalline diamonds) by ion microprobe using quartz,rutile and clay mineral inclusions has been attempted(Sano et al., 2002) with very poor precision. Pb–Pbdating of apatite and titanite has been conducted withgood precision (207Pb/206Pb b0.3% 2SE) and accuracyby multicollector LA-ICP-MS, although titanite Pb/Pbages were consistently too young by c.1% (Willigerset al., 2002). However, even with this level of precisionit cannot be demonstrated by Pb isotopes alone thatanalyses are concordant with respect to U–Pb system-atics. One attempt at ion microprobe dating of Archaeanapatite has been made, which yielded U–Pb isochronages with poor precision (5–15% 1σ), but only for the238U decay system (Sano et al., 1999); again, noguarantee of U–Pb concordance can be assured fromthese data alone. One attempt at ion microprobe datingof rutile is reported (Steven and Armstrong, 2003), butthe data are discordant and imprecise, although theyquote a final age error of c.4% (1σ). In general, U–Pbgeochronology of these phases is difficult due to lowconcentrations of U (average 33 ppm in apatite, 39 ppmin rutile (Heaman and Parrish, 1991)) and high contentsof common Pb. This situation is exacerbated by in situgeochronology where a very small analyte volume issampled and where a common Pb correction is moredifficult. In ion microprobe geochronology the situationis made more difficult by the necessity of matrixmatching, and well characterised apatite and rutile U–Pb standards do not exist at present. As a result of thestrong possibility that matrix matching for U–Pbgeochronology in LA-ICP-MS may not be a strictrequirement, we felt justified in attempting to analysePb/U ratios in apatite and rutile using zircon as anexternal standard, as for titanite. The rutile analyses arenon-controversial, since they essentially mimic the


surrounding titanite within the sample analysed. Thedata are concordant and suggest they share the samethermal history following formation of the surroundingtitanite at c.2050 Ma. Rutile is generally believed tohave a lower closure temperature for Pb of around400 °C (Mezger et al., 1989), but more recent studieshave suggested temperatures of around 600 °C (Cher-niak, 2000; Vry and Baker, 2006), which may accountfor the preservation of similar U–Pb ages to titanite inthis study. Also, the rutiles occur as inclusions withintitanite, which would also affect their ability to undergoPb-loss and would alter the closure temperature tohigher values within a Pb-retentive host mineral, such astitanite. In the absence of a separate ‘known’ rutileanalysed by the same protocol, we suggest that the datacomplement the titanite data and are a priori accurate.With respect to the apatite data, the situation is morecomplicated since the data are mostly moderatelyreversely discordant and much younger (N100 Ma)than the titanite data. Reverse discordance may beattributed to elemental fractionation of Pb and Uisotopes relative to the external standard during laserablation (Wiedenbeck et al., 1995; Horn et al., 2000;Ketchum et al., 2001; Smith et al., 2004b), due toredistribution of Pb on a micron scale (Compston, 1999;Carson et al., 2002), or to the presence of labilecomponents (Wiedenbeck et al., 1995). However, theeffect on 207Pb/206Pb age should not be hindered sincethis trend shifts analyses along a chord through zero ageto the concordia intercept (Jeffries et al., 2003; Smithet al., 2004b). The extremely reproducible 207Pb/206Pbages from the apatites measured here suggest that theage may be accurate, regardless of the precisemechanism responsible for producing reverse discor-dance. The younger age could be explained by the lowerclosure temperature of apatite for the U–Pb system of425–500 °C (Watson et al., 1985; Cherniak et al., 1991)compared to 650–700 °C for titanite and c.600 °C forrutile (op cit). One explanation is that the area remainedin a mid-crustal position following cooling through theclosure temperature for titanite at c.1708 Ma andcontinued to cool slowly until it passed through theclosure temperature for apatite at c.1583 Ma. Alterna-tively, a later thermal event could have disturbed onlythe apatites, with conditions only rising above theapatite closure temperature at this time. Romer (1996)records U–Pb ages of 1613–1620 Ma from titanitesfilling fractures from this group of deposits, so it islikely that later events affected this deposit and mayhave resulted in disturbance of the U–Pb system inapatite. Therefore, this implies that the apatite data aregeologically reasonable and may be accurate.

The prognosis for U–Pb geochronology of rutile andapatite by LA-ICP-MS is very good and encourages furthereffort. In the case of the samples here, the U contents(c.67 ppm rutile, Table 3; not measured in apatite but likelycomparable to titanite due to similar ion yields and Pb/Uprecision) were sufficient to allow reasonably precisemeasurements and the general low common Pb content(below detection) negated the requirement for a commonPb correction in this particular sample set. However, not allsamples are likely to be so amenable to analysis and shouldbe treated on their own merit; certainly, the issue ofcommon Pb correction should not be overlooked. As toprecisely why the phases studied from the iron ore provinceof northern Sweden should display such elevated radio-genic to common Pb concentrations is slightly enigmatic,but presumably relates to the fluid present duringhydrothermal alteration and metamorphism.

5.4. Implications for regional geological evolution ofthe Fennoscandian shield

Current models for the tectonic evolution of theFennoscandian shield in Norrbotten suggest that thePorphyry Group lavas and associated sediments wereemplaced in the period c.1.96–1.88 Ga (Skiöld andCliff, 1984; Welin, 1987; Bergman et al., 2001). TheLuossavaara and Kirunavaara deposits have previouslybeen considered to have formed synchronously with theenclosing trachyandesitic rocks at c.1.88–1.89 Ga(Welin, 1987; Cliff et al., 1990; Romer et al., 1994).Our data do not support either the depositional ages ofthe supracrustal successions or the view that the ore wassynchronous with enclosing volcanics, but suggest thatthe formation of the Porphyry Group was a pre-Svecokarelian event. It therefore seems more likelythat the previously recorded ages represent majorresetting at c.1.9–1.8 Ga during hydrothermal activityand ore emplacement; in that case, we currently do nothave a reliable age for portions of the stratigraphy innorthern Sweden.

On the basis of the titanite core ages determined in thisstudy we propose that the Porphyry Group in the Kirunaarea is the result of volcanism and initial burialmetamorphism in the period 2.1–1.9 Ga. This links thegenesis of the Porphyry and Porphyrite Groups to theclosing stages of the Greenstone Group basic volcanism,rather than to magmatism related to accretion during theSvecokarelian orogeny. Martinsson (1997) inferred theLinkaluoppal Formation (the upper formation of theGreenstone Group) to be formed during basin shoaling,uplift and erosion leading to a shift to within-plate-basaltand low K tholeiite volcanics. Current models for the


development of the Fennoscandian Shield in Sweden andFinland suggest rifting of theArchaean craton, continentalbreak up and the formation of a passive margin in theperiod 2.45–2.1 Ga followed by the inception of ajuvenile arc system around 1.94 Ga. The Svecokarelianorogeny then consisted of the accretion of this arcsystem to the margin of the Archaean craton during theperiod 1.9–1.8 Ga (Nironen, 1997). However, Lahti-nen et al. (2002) detected evidence for the existence of2.1–2.0 Ga Proterozoic crust in detrital zircons withalkaline affinities from the Central and SouthernSvecofennian sedimentary domains in Finland, andnoted that potential source regions were sparse in theFennoscandian Shield. The Kiruna Porphyries maytherefore actually represent volcanics generated duringthe closing stages of basin inversion or the products ofearly arc processes in this period, which are nowlargely represented by detrital zircon populations.

Subsequent metamorphism and intrusion-related heatand fluid flow events during the Svecokarelian Orogenybetween 1.9 to 1.8 Ga resulted in the regional-scalerecrystallisation of zircon within the volcanic succession,promoted by the circulation of extremely saline, hightemperature and pressure, metasomatic fluids related tothe formation of the magnetite–apatite deposits. The agesdetermined from titanite rims at Luossavaara andGruvberget in this study are in agreement with previousage determinations for Fe oxide–apatite mineralisation inthe district (Cliff et al., 1990; Romer et al., 1994). Our newdata now firmly tie thismineralisation in time to the periodof granitoid magmatism (Haparanda Suite and PerthiteMonzonite Suite) and deformation which post dated arc-related volcanism, but is related to the accretion of arcsystems onto the Archaean craton. The Fe oxide–apatitesystems unequivocally post-date volcanism and must berelated to deeper crustal processes. The range of agespreserved by titanite from the Malmberget area isconsistent with the known age range of metamorphismand late Svecokarelian magmatism in the area (Bergmanet al., 2001), and is consistent with the extended fluid flowhistory inferred from TIMS U–Pb titanite geochronologyat the nearby Aitik Cu–(Au) deposit (Wanhainen et al.,2005).

6. Conclusions

LA-ICP-MS is a powerful technique for in situ U–Pb dating of titanite and probably rutile and apatite aswell. The issue of common Pb within these phasesshould be tackled by measurement of the 204Pbisotope, but in the case of these particular samplesthe common Pb content was sufficiently low not to

necessitate a common Pb correction within the stateduncertainties. The ability to perform several differentin situ chemical, isotopic and imaging methods on thesame petrographic thick section is ideally suited foraccurate correlation between the different data thatthese methods provide.

The large titanite grains in this study contained smallareas in the core of the grains that variably retained oldconcordant U–Pb ages and REE and trace elementinformation. This information has never been extractedbefore due to previous analyses using whole grain ID-TIMS methods. The retention of old c.2050 Ma cores intitanite from across the region does not tally with zirconU–Pb ages, which do not record ages older thanc.1900 Ma. It is suggested that extremely saline, hightemperature and pressure metosomatic fluids, related toregional magnetite–apatite mineralisation, may haveresulted in recrystallisation and isotopic resetting ofzircon in this period.

Ages derived from titanite rim analyses are consis-tent with previous age determinations from the Feoxide–apatite deposits of the Norrbotten region, andprovide evidence of a temporal relationship betweenmineralisation and the period of Sveconkarelian mag-matism and accompanying metamorphism represen-ted by the Haparanda Suite and Perthite MonzoniteSuite granitoids. The trace element chemistry of the rimzones in each case studied here are consistent withtitanite recrystallisation (dissolution/reprecipitation)during mineralisation. The ages derived from corezones, however, clearly indicate that mineralisation wasnot coeval with the host volcanic sequence. The coreages provide minimum age estimates for the PorphyryGroup volcanics, as initial titanite formation probablytook place during early metamorphism of the volcanicpile. The ages of c.2050 Ma are significantly older thanprevious thought for these volcanics, and suggest anorigin for the volcanics via intermediate and alkalinemagmatism either during basin uplift, or in an early arc-related extensional setting, prior to accretion during theSvecokarelian orogeny.

Acknowledgements

This work was funded by European Union —Regional Development Fund Georange Program grant89121. We would like to acknowledge Conor Ryan andRaquel Garcia-Sanchez for aid with trace elementanalysis, and Tony Wighton for sample preparation.Anton Kearsley and John Spratt provided assistancewith SEM imaging and electron microprobe analyses.Javier Fernandez-Suarez is gratefully acknowledged for


discussions during the development of this technique.Reviewers David Chew and Emma Rehnström arethanked for constructive and honest comments thathelped to improve the manuscript.

References

Ayers, J.C., Watson, E.B., 1991. Solubility of apatite, zircon, and rutilein supercritical aqueous fluids with implications for subductionzone geochemistry. Philosophical Transactions of the RoyalSociety of London, Series A A-335, 365–375.

Bau, M., 1991. Rare earth element mobility during hydrothermal andmetamorphic fluid-rock interaction and the significance of theoxidation state of europium. Chemical Geology 93, 219–230.

Bergman, S., Kubler, L., Martinsson, O., 2001. Description of regionalgeological and geophysical maps of northern Norrbotten county(east of the Caledonian orogen). Sveriges geologiska under-soÉkning. Ser. Ba. OÉversiktskartor med beskrivningar 56.

Broman, C., Martinsson, O., 2000. Fluid inclusions associated withCu–Au mineralization in Norrbotten, Sweden. L.U.T. researchreport 2000:06 2nd GEODE Fennoscandian Shield field workshopon Palaeoproterozoic and Archean greenstone belts and VMSdistricts in the Fennoscandian Shield, p. 7.

Brugger, J., Lahaye, Y., Costa, S., Lambert, D., Bateman, R., 2000.Inhomogeneous distribution of REE in scheelite and dynamics ofArchaean hydrothermal systems (Mt. Charlotte and Drysdale golddeposits, Western Australia). Contributions to Mineralogy andPetrology 139 (3), 251–264.

Carlon, C.J., 2000. Iron oxide systems and base metal mineralisation innorthern Sweden. In: Porter, T.M. (Ed.), Hydrothermal iron oxidecopper-gold and related deposits: a global perspective. AustralianMineral Foundation, Glenside, Australia, pp. 283–296.

Carson, C.J., Ague, J.J., Grove, M., Coath, C.D., Harrison, T.M., 2002.U–Pb isotopic behaviour of zircon during upper-amphibolitefacies fluid infiltration in the Napier Complex, east Antarctica.Earth and Planetary Science Letters 199, 287–310.

Carswell, D.A., Wilson, R.N., Zhai, M., 1996. Ultra-high-pres-sure aluminous titanites in carbonate-bearing eclogites at Shuanghein Dabieshan, central China.Mineralogical Magazine 60, 461–471.

Cherniak, D.J., 2000. Pb diffusion in rutile. Contributions to Mineralogyand Petrology 139, 198–207.

Cherniak, D.J., Lanford, W.A., Ryerson, F.J., 1991. Lead diffusion inapatite and zircon using ion implantation andRutherford backscatteringtechniques. Geochimica et Cosmochimica Acta 55 (6), 1663–1673.

Cliff, R.A., Rickard, D., Blake, K., 1990. Isotope systematics of theKiruna magnetite ores, Sweden. 1. Age of the ore. EconomicGeology 85, 1770–1776.

Compston, W., 1999. Geological age by instrumental analysis: the29th Hallimond Lecture. Mineralogical Magazine 63, 297–311.

Edfelt, A., Armstrong, R.N., Smith, M., Martinsson, O., 2005.Alteration paragenesis and mineral chemistry of the Tjårrojåkkaapatite-iron and Cu(–Au) occurrences, Kiruna area, northernSweden. Mineralium Deposita 40 (4), 409–434.

Fernandez-Suarez, J., Alonso, G.G., Jeffries, T.E., 2002. Theimportance of along-margin terrane transport in northern Gond-wana: insights from detrital zircon parentage in Neoproterozoicrocks from Iberia and Brittany. Earth and Planetary Science Letters204 (1–2), 75–88.

Friend, C.R.L., Kinny, P.D., 1995. New evidence for protolith ages ofLewisian granulites, Northwest Scotland. Geology (Boulder) 23(11), 1027–1030.

Frietsch, R., Perdahl, J.-A., 1995. Rare earth elements in apatite andmagnetite in Kiruna-type iron ores and some other iron ore types.Ore Geology Reviews 9, 489–510.

Frost, B.R., Chamberlain, K.R., Schumacher, J.C., Scott, D.J., Moser,D.E., 2000. Sphene (titanite); phase relations and role as ageochronometer. Chemical Geology 172, 131–148.

Fryer, B.J., Jackson, S.E., Longerich, H.P., 1993. The application oflaser ablation microprobe-inductively coupled plasma-mass spec-trometry (LAM-ICP-MS) to in situ (U)–Pb geochronology.Chemical Geology 109 (1–4), 1–8.

Geijer, P., 1931. Berggrunden inommalmtrakten Kiruna-Gällivare-Pajala.Sveriges Geologiska Undersokning (SGU) Series C 366 255 pp.

Harlov, D.E., Andersson, U.B., Förster, H.-J., Nyström, J.O., Dulski,P., Broman, C., 2002. Apatite-monazite relations in the Kiiruna-vaara magnetite–apatite ore, northern Sweden. Chemical Geology191, 47–72.

Heaman, L., Parrish, R., 1991. U–Pb geochronology of accessoryminerals. Short Course Handbook on Applications of RadiogenicIsotope Systems to Problems in Geology. Mineralogical Associ-ation of Canada, Toronto. 59–102 pp.

Hirata, T., Nesbitt, R.W., 1995. U–Pb isotope geochronology ofzircon; evaluation of the laser probe-inductively coupled plasma-mass spectrometry technique. Geochimica et Cosmochimica Acta59 (12), 2491–2500.

Hirata, T., Iizuka, T., Orihashi, Y., 2005. Reduction of mercurybackground on ICP-mass spectrometry for in situ U–Pb agedeterminations of zircon samples. Journal of Analytical AtomicSpectrometry 20, 696–701.

Horn, I., Rudnick, R.L., McDonough, W.F., 2000. Precise elementaland isotope ratio determination by simultaneous solution nebuli-zation and laser ablation-ICP-MS; application to U–Pb geochro-nology. Chemical Geology 164 (3–4), 281–301.

Horstwood, M.S.A., Foster, G.L., Parrish, R.R., Noble, S.R., Nowell,G.M., 2003. Common-Pb corrected in situ U–Pb accessorymineral geochronology by LA-MC-ICP-MS. Journal of AnalyticalAtomic Spectrometry 18 (8), 837–846.

Hunt, J.A., Kerrick, D.M., 1977. The stability of sphene; experimentalredetermination and geologic implications. Geochimica et Cos-mochimica Acta 41, 279–288.

Jeffries, T.E., Fernandez-Suarez, J.,Corfu, F.,Alonso,G.G., 2003.Advancesin U–Pb geochronology using a frequency quintupled Nd: YAG basedlaser ablation system (lambda = 213 nm) and quadrupole based ICP-MS. Journal of Analytical Atomic Spectrometry 18 (8), 847–855.

Ketchum, J.W.F., Jackson, S.E., Culshaw, N.G., Barr, S.M., 2001.Depositional and tectonic setting of Palaeoproterozoic LowerAillik group, Makkovik Province, Canada: evolution of a passivemargin-foredeep sequence based on petrochemistry and U–Pb(TIMS and LAM-ICP-MS) geochronology. Precambrian Research105, 331–356.

Lahtinen, R., Huhma, H., Kousa, J., 2002. Contrasting sourcecomponents of the Paleoproterozoic Svecofennian metasediments:detrital zircon U–Pb, Sm–Nd and geochemical data. PrecambrianResearch 116, 81–109.

Lee, J.K.W., Williams, I.S., Ellis, D.J., 1997. Pb, U and Th diffusion innatural zircon. Nature (London) 390 (6656), 159–162.

Li, X.-H., Liang, X., Sun, M., Guan, H., Malpas, J.G., 2001. Precise206Pb/238U age determination on zircons by laser ablationmicroprobe-inductively coupled plasma-mass spectrometry usingcontinuous linear ablation. Chemical Geology 175, 209–219.

Lindblom, S., Broman, C., Martinsson, O., 1996. Magmatic-hydrothermal fluids in the Pahtohavare Cu–Au deposit ingreenstone at Kiruna, Sweden. Mineralium Deposita 31, 307–318.


Ludwig, K.R., 2003. User's manual for Isoplot 3.00 a geochronolog-ical toolkit for Microsoft Excel.

Martinsson, O., 1997. Tectonic Setting and Metallogeny of theKiruna Greenstones, PhD Thesis, Luleå University of Technology,19 pp.

Martinsson, O., Virkkunen, R., 2004. Apatite Iron Ores in theGällivare, Svappavaara, and Jukkasjärvi Areas. Society ofEconomic Geologists Guidebook Series, 33. 167–172 pp.

Martinsson, O., Wanhainen, C., 2004. Character of Cu–Au miner-alisation and related hydrothermal alteration along the NautanenDeformation Zone, Gällivare Area, Northern Sweden. Society ofEconomic Geologists Guidebook Series, 33. 149–160 pp.

Mezger, K., Hanson,G.N., Bohlen, S.R., 1989.High-precisionU–Pb agesof metamorphic rutile — application to the cooling history of high-grade terranes. Earth and Planetary Science Letters 96, 106–118.

Nironen, M., 1997. The Svecofennian Orogen: a tectonic model.Precambrian Research 86, 21–44.

Piccoli, P., Candela, P., Rivers, M., 2000. Interpreting magmaticprocesses from accessory phases: titanite — a small-recorder oflarge-scale processes. Transactions of the Royal Society ofEdinburgh. Earth Sciences 91, 257–267.

Pidgeon, R.T., Bosch, D., Bruguier, O., 1996. Inherited zircon andtitanite U–Pb systems in an archaean syenite from southwesternAustralia: Implications for U–Pb stability of titanite. Earth andPlanetary Science Letters 141 (1–4), 187–198.

Pollok, K., Geisler, T., Putnis, A., 2004. How does a replacement frontproceed? Observations on chlorapatite–hydroxylapatite replace-ments. Geochimica et Cosmochimica Acta 68 (11), A184.

Prowatke, S., Klemme, S., 2005. Effect of melt composition on thepartitioning of trace elements between titanite and silicate melt.Geochimica et Cosmochimica Acta 69, 695–709.

Putnis, A., 2002. Mineral replacement reactions: from macroscopicobservations to microscopic mechanisms. Mineralogical Magazine66, 689–708.

Putnis, A., Putnis, C.V., 2004. Reactive fronts at the nanoscale: thenature and origin of a coupled dissolution–reprecipitationinterface. Geochimica et Cosmochimica Acta 68 (11), A180.

Rasmussen, B., 2005. Zircon growth in very low-grade metasedimen-tary rocks: evidence for zirconium mobility at ∼ 250 °C. Contri-butions to Mineralogy and Petrology 150 (2), 146–155.

Romer, R.L., 1996. U–Pb systematics of stilbite-bearing low-temperature mineral assemblages from the Malmberget iron ore,northern Sweden. Geochimica et Cosmochimica Acta 60 (11),1951–1961.

Romer, R.L., Martinsson, O., Perdahl, J.-A., 1994. Geochronology ofthe Kiruna iron ores and hydrothermal alterations. EconomicGeology 89, 1249–1261.

Rubin, J.N., Henry, C.D., Price, J.G., 1993. The mobility of zirconiumand other “immobile” elements during hydrothermal alteration.Chemical Geology 110, 29–47.

Sano, Y., Oyama, T., Terada, K., Hidaka, H., 1999. Ion microprobe U–Pb dating of apatite. Chemical Geology 153, 249–258.

Sano, Y., Yokochi, R., Terada, K., Chaves, M.L., Ozima, M., 2002. Ionmicroprobe Pb–Pb dating of carbonado, polycrystalline diamond.Precambrian Research 113, 155–168.

Schuiling, R.D., Vink, B.W., 1967. Stability relations of sometitanium-minerals (sphene, perovskite, rutile, anatase). Geochimicaet Cosmochimica Acta 31, 2399–2411.

Scott, D.J., St-Onge, M.R., 1995. Constraints on Pb closuretemperature in titanite based on rocks from the Ungava orogen,Canada: implications for U–Pb geochronology and P–T–t pathdeterminations. Geology 23 (12), 1123–1126.

Simonetti, A., Heaman, L.M., Hartlaub, R.P., Creaser, R.A.,MacHattie, T.G., Böhm, C., 2005. U–Pb zircon dating by laserablation-MC-ICP-MS using a new multiple ion counting Faradycollector array. Journal of Analytical Atomic Spectrometry 20,677–686.

Skiöld, T., 1987. Aspects of the Proterozoic Geochronology ofnorthern Sweden. Precambrian Research 35, 161–167.

Skiöld, T., 1988. Implications of new U–Pb zircon chronology to earlyProterozoic crustal accretion in northern Sweden. PrecambrianResearch 32, 35–44.

Skiöld, T., Cliff, R.A., 1984. Sm–Nd and U–Pb dating of earlyProterozoic mafic–felsic volcanism in northernmost Sweden.Precambrian Research 26, 1–13.

Smith, M.P., Henderson, P., Jeffries, T.E.R., Long, J., Williams, C.T.,2004a. The rare earth elements and uranium in garnets from theBeinn an Dubhaich aureole, Skye, Scotland, UK: constraints onprocesses in a dynamic hydrothermal system. Journal of Petrology45 (3), 457–484.

Smith, S.R., Foster, G.L., Romer, R.L., Tindle, A.G., Kelley, S.P.,Noble, S.R., Horstwood, M., Breaks, F.W., 2004b. U–Pbcolumbite–tantalite chronology of rare-element pegmatites usingTIMS and laser ablation-multi collector-ICP-MS. Contributions toMineralogy and Petrology 147 (5), 549–564.

Stacey, J.S., Kramers, J.D., 1975. Approximation of terrestrial leadisotope evolution by a two-stage model. Earth and PlanetaryScience Letters 26 (2), 207–221.

Steven, N., Armstrong, R., 2003. A metamorphosed Proterozoiccarbonaceous shale-hosted Co–Ni–Cu deposit at Kalumbila,Kabompo Dome; the Copperbelt ore shale in northwesternZambia. Economic Geology and the Bulletin of the Society ofEconomic Geologists 98 (5), 893–909.

Storey, C.D., Jeffries, T.E., Smith, M., 2006. Common lead correctedlaser ablation ICP-MS systematics and geochronology of titanite.Chemical Geology 227, 37–52.

Sverjensky, D.A., 1984. Europium redox equilibria in aqueoussolution. Earth and Planetary Science Letters 67, 70–78.

Tichomirowa, M., Whitehouse, M.J., Nasdala, L., 2005. Resorption,growth, solid state recrystallisation and annealing of granulitefacies zircon — a case study from the Central Erzegerbrige,Bohemian Massif. Lithos 82, 25–50.

Tiepolo, M., 2003. In situ Pb geochronology of zircon with laserablation-inductively coupled plasma-sector field mass spectrom-etry. Chemical Geology 199 (1–2), 159–177.

Tiepolo, M., Oberti, R., Vannucci, R., 2002. Trace element incorpo-ration in titanite: constraints from experimentally determined solid/liquid partition coefficients. Chemical Geology 191, 105–119.

Tomaschek, F., 2004. Transient porosity at reaction fronts: replacementtextures from flux-grown zircon. Geochimica et CosmochimicaActa 68 (11), A184.

Tomaschek, F., Kennedy, A.K., Villa, I.M., Lagos, M., Ballhaus, C.,2003. Zircons from Syros, Cyclades, Greece — recrystallisationand mobilization of zircon during high-pressure metamorphism.Journal of Petrology 44, 1977–2002.

Vry, J.K., Baker, J.A., 2006. LA-MC-ICPMS Pb–Pb dating of rutilefrom slowly cooled granulites: confirmation of the high closuretemperature for Pb diffusion in rutile. Geochimica et Cosmochi-mica Acta 70, 1807–1820.

Wanhainen, C., Billström, K., Martinsson, O., Stein, H., Nordin, R.,2005. 160Ma of magmatic/hydrothermal and metamorphic activityin the Gällivare area: Re–Os dating of molybdenite and U–Pbdating of titanite from the Aitik Cu–Au–Ag deposit, northernSweden. Mineralium Deposita 40, 435–447.


Watson, E.B., Harrison, T.M., Ryerson, F.J., 1985. Diffusion of Sm, Sr, andPb in fluorapatite. Geochimica et Cosmochimica Acta 49, 1813–1823.

Weihed, P., 2004. Overview of the Geology and Tectonic Setting ofNorthern Sweden. Society of Economic Geologists, GuidebookSeries, vol. 33. 1–15 pp.

Welin, E., 1987. The depositional evolution of the Svecofenniansupracrustal sequence in Finland and Sweden. PrecambrianResearch 35, 95–113.

Wiedenbeck, M., Alle, P., Corfu, F., Griffin, W.L., Meier, M., Oberli,F., Vonquadt, A., Roddick, J.C., Speigel, W., 1995. 3 NaturalZircon Standards for U–Th–Pb, Lu–Hf, Trace-Element and ReeAnalyses. Geostandards Newsletter 19 (1), 1–23.

Willigers, B.J.A., Baker, J.A., Krogstad, E.J., Peate, D.W., 2002.Precise and accurate in situ Pb–Pb dating of apatite, monazite, andsphene by laser ablation multiple-collector ICP-MS. GeochimicaEt Cosmochimica Acta 66 (6), 1051–1066.

Date post:	15-Nov-2023
Category:	Documents
Upload:	nhm
View:	0 times
Download:	0 times

In situ LA-ICP-MS U–Pb dating of metavolcanics of Norrbotten, Sweden: Records of extended...

Documents