TOM ANDERSEN,ARTHUR G.SYLVESTER
s
ARILD ANDRESEN----
NGU-BULL440,20 0 2- PAGE 19Age and petrogenesis of the Tinn granite, Telemark, South Norway, and its geochemical relationship to metarhyolite of the Rjukan Group
TOM ANDERSEN,ARTHUR G.SYLVESTER&ARILDANDRESEN
Andersen,T.,Andresen,A. & Sylvester,A.G. 2002: Age andpetrogenesisofthe Tinn granite,Telem ark,South Norw ay, and itsgeochemicalrelationshipto metarhyoliteof Rjukan Group.NorgesgeologiskeundersokelseBulletin 440,19-26.
TheTinn graniteisa Mid Proterozoic,foliated pluto n, situatedin the centralpartofthe TelemarkSector,South No rway. It isspatiallyassociated wit h met arhyo lit ebelo ng in g totheTudd alFo rm at io n of theRj ukan Group ofthe Telem ark sup racrustalseque nce.SIMS U-Pb dati ngindi catesanage of 14 76±13 Ma,whichisslightlyyo u ng erthan the1500-151 4Ma eru ptio ninte rva l fortheTuddalFormationrhyolite.Rare xenoc ryst iczirco ncoresgiveanage of 1506±10Ma, whic h isindistinguishabl efrom theageof theTuddalFor mat io n.The absenceofolde r inherit edzir- consand evidence fromwho le-rock Nd isotopes sugg estthat noso urce componentolderthan theRjukanGroupis need edin theso urce regionof theTinngranitemagma.Thepreferr edpetrogen eticmod elfortheTinngran it eis a part ial melting-mi xingprocessat mod erat edepth in thecrust, withi n theRju kanGroup volcan ic pile,A mafic magmaacte das a sourceof heat andcont ribute d tothe bulk chemist ryof the grani t ic mag m a.Resettingof thelead isoto pesyst emof the graniteat mineral scale too kplacein Sveconor wegiantime, at 1031±32 Ma.
Tom Andersen and Arild Andresen,Department ofGeology,Universityof Oslo,PO Box 1047Blindern,N-0316 Oslo, Nor way.Arth urG.Sylvestet,Departmentof GeologicalSciences,Universityof Californ ia,Santa Barbara,CA93106-9630, USA.
Introduction
Granitic intr usionsmake up a substanti al comp on ent of the cont ine nta l crust in the southwestern part of the Baltic Shield.Among the Precam b rian granitic rocks of Sout h Norway, a group of granitic orthogneissesin the Telemark sector stands out as especially poorly understood.These rocks are spatially associated with the Mid to Late Proterozoic Telemark Supracrustal sequence(Sigmond et al.
1997),and include the voluminoussouth Telemark Gneisses situatedsouth of the mainoutcrop area of the supracrustal sequence (Ploquin 1972,Dons & Jorde 1978, Kleppe 1980) andthe Tinn granitein the north(Sigmond 1998).
This studypresents new SIMSU-Pb dataforthe emplace- mentage of the Tinn granite,and clarifiesits relationshi p to the Telemark metar hyo lite and to other possib le source rocks.
Geologic setting
The Telemark Supracrustal sequence is a well-preserved sequence of Mid Proterozoic metavolcanic and metasedi- mentary rocks,situated in thecentral part of theTelemark Sector (Dons& Jorde 1978, Sigmond 1998), surrounded by stro ng ly deformed, high er-grade ort ho- and paragneisses (Fig. 1).The supracrusta lsequence consistsof threelith os- trat igrap hic groups separated by angula r unconformities:
theRjukan, Seljordand Bandak Groups, and a fourthgroup, the Heddal Group,conformably overlying the Seljord Group in the eastern part ofthe outcrop area(Dons1960, Sigmond 1998).The RjukanGroupis entire ly metavolcanic,with a thick
sequenceofmetar hyolite(theTuddalFormation)overlainby a thinne r met abasalt ic formatio n (the VemorkFormation). The rhyol it ic rocks were deposited in extensional basins, possibly as part of a continentalrift system(Sigmondet al.
1997),on a migmatiticbasement of unknownage.From the northendoflake Tlnnsjeto the Caledoniannappe front,the Rj ukanGroupis cutby younger mafic to granitic intru sions of the Uvdal plutonic belt (Sigmond et al. 1997,Sigmond 1998).The Seljord Group consistsof quartziteand conglom- erate and the HeddalGroup of quartz arenite with subordi- nate metavolcani c rocks, whereas the Bandak Group is a mixed, volc anic-sedim entary sequence com prisi ng several form at io ns.
Sigmo nd(1998) reported conventionalU-Pb zircon ages for the Tuddal Format ion rhyolite of 1512 +9/-8 Ma and 1499±39Ma, and 1509+ 19/-3for an intrusion of the Uvdal plutonicbelt the bestestimateof the duration ofthe Rjukan Group volcanismwas given as1500-1514 Ma.A volcanicfor- mation of the Bandak Group has been dated atc.1150 Ma (Dahlgrenetal. 1990).From detrital zircon systemat ics, Haas et al.(1999) inferred a maximum depositiona lage of the Seljord Group at 1450 Ma,butnewU-Pb age data from rhyo- lite underlying thetype prof ile of theSeljordGroup suggest thatsome ofthe sed imentary rockspreviou sly assigned to theSeljord Gro up maybe youngerthan 1155 Ma (Laajokiet aI.2000).
The Tinn granite makes up the southern most part of the Uvdal plut onicbelt(Fig.1).It is a fine-grained,palepink, leu- cocratic two-feldsp argranite, wit h minordarkbrownbiot it e
NGU-BULL 440.2002 - PA G E20 TOMANDERSEN, ARTHUR G. SYLVES TER
s
ARILD AND RES ENSeljord.HeddalandBandakGroups
• Quartzites
Intrusiverocks
..x ...x'
~x~x~ TInngranite
IDIIl
Graniticto tonalitic intrusions• Mafic intrusions
RjukanGroup
• VemorkFm metabasalt
• TuddalFm metarhyolite
Gneiss.migmatite
Fig.1. Simp lifiedgeo log ic map of the Tinngrani teandsurro und ingrocks.compiledfro mDo ns & Jorde(1978)and Sigmo nd(1998). Overviewmap:
Simplifiedmapof South Norw ay showi ng regional su bdi visio n used in this st udy:A:0stfol d-Akershus sect o r;K:Kongsbergsector;T:Telema rksecto r;B:
Bambl e sect or;R:Rogaland -Vest Agder secto r.Major shear zon es:MMS:Mjesa-Maqnor shear zone.0MZ:0rj e mylo nite zone;PKS:Por sg run n- Kristia nsandshearzone; KTB:Kongsberg-Telemarkboun da ry;MANUS:Mandal-Ustaosetline;CTF(b ro ken line):Caledonian thrust front.
and magn et it e,and accessory zircon,tit anit e and apatite.
Field observations do not defin eunambigu ou s age relation- shipsbetween the granite and metarhyolite of the Tuddal Formation:The contact between granite and metarhyoliteis gradat io nal,and enclavesor dikes of one inthe otherare now here observed. The grain size of the metarhyolite increases toward the contact with the granite, however, which suggests a local thermal imprint related to the emplacementofthe granite(Sig mond 1998).Both unitsare foliatedparallelto thecontact.Thismaybea result ofdefo r- mati onduri nggraniteemplacement,as has been suggested for other granites in southern Norway (e.g. Elders 1963, Sylvest er 1998),or a result of later tectonic deformation locallycont rolle dby the more competentgranite.
The Tinngranit e is amoderately silica-richgranite(SiO,=
68.3-72.4 wt%,Table 1).Its atomic (Na+K)/AI ratio is well below 1.0.The normative mineralogy is highly leucocratic, wit hadifferent iation index (normat ive qz+ab+qz+ne)well above 80.The samples plot well withi n the 'granite sensu stricto'fieldsin pqand Ab-Or-An classification diagrams (e.g.
Rollinson 1993).Compared to the average of the Tuddal Formation metarhyolite,the Tinn granite is low in SiO,and highin CaOandNa,O (Table1).Furt hermo re,the Tinn gran- ite is meta lumi nous((Ca+ Na+ K)/AI of 1.05-1.08,norm ative co=O),in contrastto the peraluminous composition of the average metarhyolite(normati ve co=1.9).
Analytical methods
The present study is based on two 5-8 kg samples of the Tinn granite(083196-2 and071996-2), for which whole-rock
Sr,Nd, and Pb isotopedata were publishedby Andersen et al.(2001).Thesampleswere crushed to agrain sizeof less than 250 urn using ajaw crusher and a percussion mill.
Zirconswereseparated from the<250urnfraction by acom- bination of Wilfley-tabl e washing,heavy liquid separation (1,1,2,2- tetrabromoethaneand diiodomethan e)and mag- neticseparation.Thefinal,non-mag neti c zirconfraction was then purified by hand picking under a binocu lar micro- scope,andselecte d grains were mount ed on doubly adhe- sive tape,cast inepoxy and polishedfor the ion microprobe study.Electronbackscatterimaging(BSE)inascanning elec- tron microscope(DepartmentofGeology,Oslo)and an elec- tron micropr ob e(Macquarie University,Syd ney)was used both as a prelim inary survey before analysis,and to docu- mentindividual grains afteranalysis.The U-Pb zircondating was perfo rm ed in the NORDSIM laboratory locat ed atthe Swedish Museum ofNatural HistoryinSto ckholm,using a CAMECAIMS1270ionmicroprobe;analyt icalconditionsand data red uct ion procedur esaredescrib edbyWhitehouse et al.(1997, 1999).U-Pbdataare listedwith1<Yerrors in Table1, whereas thederived agesare given with 95% confidence errors.Additional separatesof rock-formingmineralsfor the Pb-Pb isochron st udy were made by a combi nationof heavy liquid and magnetic separation.follo wed by hand picking.
Lead wasseparated andanalysed by methods described by Andersen (1997).Whole-rock and K-felds par lead isotope dataare takenfromAndersen etal.(200 1).
All geochronologiccalculationshavebeenmade using IsoplotlExversion 2.32(Ludw ig 2000).
TOM ANDERSEN,ARTHUR G.SY LVES T ER&ARILD ANDRESEN NGU-BUL L44 0, 2002 - PAGE 21
Table 1. Geo chemicaldataon the Tinngran ite.
Radiogenicisotopes
'47Sm/14'Nd 0.1229 0.1049 '''Nd/14'Nd 0.5120830.511829
±2u 16 10
tDM 1.60 1.69
Major elementanalysisbyXRF(Departmen tofGeology,University ofOslo), trace eleme ntsby ICPMS (Actlabs,Canada).Fe20 3/FeO est im atedaccord ing toRo llinson(1993). Whole-roc k radi- ogenicisoto pe data fromAndersenet al.(2001), and dataonTud dalFm.metarhyolitesfrom Brewer &Men uge(1998).Ndmode l agesare calculate dusingthedeplet edmant lereservoirof De Paolo(198 1).
Fig.2. Backsc atterelec tro nimagesofzirco nsfrom theTinngrani te.
Lo cat io n of SIMSana lytica lspots ind icated(stip ple d ellipse s)wit hnum- bersreferringtoTable 1.lmagesmade aft e r analysis,using aCameca SX- 50 elec tro n microp ro be at GEMOC Natio nal KeyCentre, Macquarie Univer sit y,Australia.a:Mostcommonzirco nin Tinngran ite consistsof ce nt raldoma in wit hoscillato ry,magmatic zonin gand athin, discontin- uou s,BSE-bright overgr ow t h(arrows) .Sample071996-2 .b:Zirco ncrys- tal wit hwell-ro u nd e d,xenoc ryst iccore in oscillat or y zone d ho st.Not e wea k, oscillat or yzoning incore (rig ht part) cut byinterfacebet w e en coreand ho st. Sa m ple083196-2c:Zircon wit ha corrode dxenoc ryst ic core(streng thened by outline).BSE-brightoverg rowthismorest rongly developed inthiscrystal than in thatin b,butit hasbeenpartlybro ke n off(left part)duringcrushing.Sam p le083 196-2.
28.0653 22.9046 1.282289 1.172639
14 11
083196-2 071996-2 Tuddal 083196-2 071996-2 Tuddal
Fm Fm
average average
UTM-reference
4862 4908
66509 66523
Wt%Oxides ClPW-Norm
68.30 72.41 75.07 qz 23.67 29.01 36.9
0.66 0.21 0.27 co 0 0 1.9
14.06 13.10 12.52 or 30.73 30.73 30.6
1.16 0.66 0.73 ab 27.76 29.47 24.2
2.10 1.19 1.47 an 8.27 4.75 0.9
0.05 0.04 0.03 di 1.27 1.31
0.54 0.13 0.40 hs 2.92 1.15 2.7
2.27 1.31 0.25 ilm 1.25 0.40 0.5
3.28 3.48 2.86 mt 1.57 0.89 1.1
5.20 5.20 5.19 ap 0.55 0.10 0.1
0.23 0.04 0.03
1.58 0.83 an% 0.23 0.14 0.04
99.43 98.61
Traceelements,partspermillion
288 310 184
364 319 491
17 17 9
32 41 33
0.46 0.65
Morphology and internal structure of zircons
Zirconsin the Tinn granitearemoderatelyelongated prisms.
BSE images reveal awell-develo ped oscillatory magmatic zoning,in mostgrain sovergro wn by a thin and discontinu- ous, BSE-brightouter zone(Fig.2a).These overgrowthswere toothin to be analysed, but were most likelyformed during met amorphic recrystall ization of the granite. Xenocrystic corespredatingthemain zircon-fo rmingevent,wit h boun d- ariesclearly discordant to the magm atic zoning,are rare.
Amo ng morethan 200 grains mount ed for analysis, only four single crystalscontained coreswhich werevisiblein BSEimages,two ofwhich are show n in Fig.2bandc.
"Rbl"Sr
"Srl"Sr
±2u UTME UTMN
Rb Ba Pb Sr Eu Sa mple
Si0 2 Ti0 2 AI20 3 Fe20 3 FeO MnO MgO CaO Na20 K20 P20S LOI Sum
Geochronology
SIMS U-Pb data and the age of emplacement
Twent y-nine spots on 25 selected zircon grains were analysed by secondary ion mass spect romet ry (Table 2).
Individu alspot analysesareidenti fiedbyNORDSIMlabora- torylog numbers.Core-rim pairs (denoted by
a
and bin Table 2)wereanalysedin grains whereinherite d cores wereI
NGU- BULL 440,2002 - PAGE 22Table 2.SIMS U-Pbdat afortheTinngranite.
TO MAN DERS EN,ARTHURG.SYLVES TER
s
ARILDANDRESEN---
Sam ple Spot
.I01Pb ±u 101Pb ±u lOI>pb ±a p .lOlIpb eoDisc. U Th Pb
"Pb % mU % B~U % Error "Th % %ppm ppm ppm
correlation
Th/U Ol>pblf;'06% X1'Pb ±a
"'Pb "'Pb
Ma Ma
.IO'"Pb eo-
mU Ma Ma
-Pb eo-
,·U
Ma Ma
1O!IPb ±a
uTfh Ma Ma 083196·2
n446-0Ja n446-01b n446-02a n446-02b n446-03a n446-03b
0.09367 0.4 +0.09297 0.3 0.06268 0.5 0.08708 2.7 0.07440 0.9 0.07351 0.8
3.5086 1.2 0.27168 1. 1 3.6029 0.8 0.28106 0.8 0.6768 0.7 0.07832 0.5 1.9590 2.8 0.16315 1.0 1.0245 1.0 0.09986 0.5 0.8 177 1.0 0.08067 0.5
0.95 0.92 0.93 1.00 0.98
1.00
1 259
7 277
-31 3889 -29 903 -43 2430 -53 3444
148 89 149 98 2760 375 4068 19 1 2640 309 4278 365
0.57 186880 0.0 0.54 56240 0.0 0.71 1590 1.2 4.51 1300 lA 1.09 1090 1.7 1.24 980 1.9
1501 7 1487 6 697 10 1362 51 1052 18 1028 17
1529 1550 525 1102 716 607
9 7 3 19 5 4
1549 1597 486 974 614 500
15 11 2 9 3 3
083196-2
n709-01a 0.07750 OA n709-02. +0.09078 0.3 n709-03. +0.09296 0.4
n709~04a-,ore0.09423 0.4 n709-Q4b_rim+0.09093 0.3 n709-05. +0.09298 1.2 n709-06. +0.09334 0.3 n709-07. +0.09385 0.4 n709-08. +0.09303 0.2 n709-09. +0.09215 0.3 n709-10. +0.09240 0.3 n709-11. +0.09088 0.4 n709-12. +0.09419 0.6
1.3477 2.6 0.12612 2.6 2.4729 2.6 0.19756 2.5 2.3622 2.6 0.18430 2.5 3.2895 2.6 0.25318 2.6 2.5881 2.6 0.20642 2.5 1.8845 2.8 0.14700 2.6 2.7464 2.6 0.21339 2.5 2.5787 2.6 0.19928 2.6 3.3108 2.5 0.25811 2.5 2.845 7 2.6 0.22398 2.6 2.9505 2.6 0.23160 2.6 2.7126 2.6 0.21649 2.6 2.6101 2.6 0.20098 2.5
0.99 0.0286 10 -34 0.99 0.02877 7.6 -21 0.99 0.01448 7.6 -29 0.99 0.08109 7.7 -4 0.99 0.02728 7.5 -18 0.91 0.00930 8.1 -43 0.99 0.03531 7.5 -18 0.99 0.01793 7.7 -24 1.00 0.07615 7.5 -1 0.99 0.02870 7.5 -13 0.99 0.03714 8.0 -10 0.99 0.02492 7.5 -14 0.98 0.03421 7.7 -24
3229 878 464 813 632 193 545 891 123 1434 682 453 770 726 193 1205 3332 225 682 588 180 557 805 136 931 507 301 875 1111 246 956 936 277 448 400 116 353 243 86
0.27 0.78 1.64 0.48 0.94 2.76 0.86 1.45 0.54 1.27 0.98 0.89 0.69
4805 OA 12228 0.2 13077 0.1 66534 0.0 8842 0.2 516 3.6 14507 0.1 25674 0.1 93371 0.0 35298 0.1 32852 0.1 13596 0.1 31182 0.1
1134 8 144 2 6 1487 8 1513 8 1445 6 1488 22 1495 7 1505 7 1489 4 1470 5 1476 5 1444 8 1512 10
867 15 1264 19 1231 19 1479 21 1297 19 1076 19 1341 19 1295 19 1484 20 1368 19 1395 20 1332 20 1303 19
766 18 1162 27 1090 26 1455 34 1210 28 884 21 1247 29 1171 27 1480 34 1303 30 1343 31 1263 30 1181 27
570 57 573 43 291 22 1576 117 544 40 187 15 70 1 52 359 27 1483 108 572 42 737 58 498 37 680 51
071796-2 n711-01.
n7l1-02a n711-03a n711-04.
n711-06.
n711-07.
n711-08.
n711-09.
n711-10a n711-12.
+0.09183 0.6 2.6542 2.7 0.20963 2.6 0.08781 lA 2.6250 3.0 0.21681 2.6 0.08476 0.5 1.8583 2.7 0.15901 2.7 +0.09345 0.2 3.0782 2.6 0.23889 2.5 +0.09097 0.4 2.6031 2.6 0.20754 2.6 +0.09268 0.5 2.1991 2.6 0.17208 2.6 +0.09184 0.3 2.7849 2.6 0.21993 2.6 +0.09294 0.5 2.822 2 2.6 0.22023 2.5 0.08541 0.3 1.9206 2.6 0. 16310 2.5 +0.09101 0.5 2.6960 2.6 0.21484 2.5
0.97 0.04561 8.2 -18 974 0.88 0.03204 7.8 -9 821 0.98 0.00845 7.7 ·29 696 1.00 0.06449 7.5 -9 1025 0.99 0.03755 7.6 -17 750 0.98 0.01344 7.7 -33 835 0.99 0.03227 7.6 -14 774 0.98 0.03356 7.7 -15 310 0.99 0.02305 7.6 -29 1760 0.98 0.03322 7.5 -15 1467
724 256 1478 241 986 128 568 304 581 192 1611 179 905 215 321 85 1981 357 2338 421
0.74 1.80 lA 2 0.55 0.77 1.93 1.17 1.03 1.13 1.59
13340 0.1 585 3.2 3619 0.5 31646 0.1 4170 0.5 11330 0.2 5107 0.4 4466 0.4 14102 0.1 1212 1.5
1464 12 1378 27 1310 9 1497 4 1446 8 1481 10 1464 7 1487 10 1325 6 1447 9
1316 20 1308 22 1066 18 1427 20 1302 19 1181 18 1351 19 1361 20 1088 17 1327 19
1227 30 1265 30 951 }4 1381 32 1216 28 1024 24 1282 30 1283 30 974 23 1255 29
901 72 638 49 170 13 1263 92 745 56 270 21 642 48 667 51 461 35 661 49 Pointsn446·01to n446·03 analysedinFebruary 1999,the other pains in january2000. Spotswitha number endingin a havebeen analysed in the centre of a grain,bin the rim.
Analysts:A.AndresenandT.Andersen.
+:indicatesanalysesincludedinthefinal estimateof the emplacement age. Boldtypefacerefersto xenocrysticzirconcores
clearly ob served or suggested by BSE imag es; all ot he r analyses are from thecentralpartof oscillato ryzon ed crys- tals wit hout visib lexeno cryst icco res.One of the analysed grain s (co re-rim pairn446-0 1a/b) isrev ersely discordan t; the ot he rgrai nsran gefro mnear-con cor d an t(1%discordant)to strongly discord ant(>30%disco rd ant).When all po ints are plo t te d in aconco rdia diagram, amajority ofpointscluster alongalead-losslinefrom a MidProte rozo ic upper inte rcep t to alow er interceptwhich ispoorlydef ined, butwithinana- lyt icaluncert ainty of0 Ma (Fig. 3a).Severalpointsfallsignifi- cant ly totheleftof thisline,however, suggest ingthatsome zirco nshave also been affec ted by alead-loss event related to later met am o rph ism (Fig.3b,c).Themet amor p h icover- print may be of Sveconorw eg ian (e.g. Andersen & Munz 1995)or,po ssibl y,Caled oni an age.
When the grainsleast aff ected bySvecono rwegian (or Caled o n ian) lead-lo ss are reg ressed for each sample sepa- rat ely,identicalages of 1476±20(071996-2 , 7single analy- ses) and 1476± 13 Ma (083 196-2,12 po ints) areobta ined (Fig.3b,dj,assumi ng recentlead- lo ss.The less-t han-perfect fit ofthesereg ressionscou ld be caused bythepresenceof unde tected, slight ly old ercoresinthe volu mes sam p led by theion-be am,or bythe effectsof incipient Svecono rw egian lead lo ss.However,further'imp rovement' of theseagesby exclusio nof more pointsisnot justifie d.Theoscilla tory zir- conmusthavegrown duringcrysta lliza t io nof the Tin ngran-
ite magm a,and 7476 ± 73Maisregarded asthe best esti - mate ofits em pla ce mentage.
Of thefou r cores,tw o (n446-02a,n446-03a)come from grains that hav e been thoroughly reset as indi cated by Svecon o rwegian207PbF06Pb ages(Table 2).The two rem ain- ing cores (n446-0 1a and n709-04a) plot marginally tothe right of themainpo pu latio n of zircons(Fig. 3c).Regressio n th roug h aforced lower inte rce pt at 0 Ma yield san age of 1506± 10Ma(Fig.3c),wh ich is slightly,but stillprobabl y sig - nificantly,olde r thanthe age of the main population ofzir- consinsamp le083196-2.
Pb-Pb isotope data and timing of metamorphism
K-fel ds pa r,biotit e,apati te,mag netiteandtitanit eseparated fro m sample 071996-2 give a large rang e of lead isot o p ic compos itions(Table 3),from near-initiallead (K-feld spar) to radiog enic lead with 206Pb/,,)'Pb above90(titanit e).A regres- sio nofall dat a(m ineralsand both whole-rocks)yields a Pb- Pbscatterc hronwit hanage of1031±32 Ma (Isop lot Mod el 2;Ludwi g,2000)and anMSWDof 9.4(Fig. 4).Removalof the whole-rock 083196-2 from the regression incr eases the uncert ainty to ±40 Ma and raises the MSWDslightl y,but does not affec t the age.The 1031 ±32 Ma correlat ion line indi cates partial ho m o genizat io n of the lead isotop es in handspeci me nas wellas intru sionscalein Sveconorwegian
TOM ANDERSEN, ARTHURG.SYLVESTER &ARILD ANDRESEN NGU-BULL 440, 2002 - PAGE 23
0.30
a 083196 -2 b 0831 96-2
0.30 1650
0.26
0.26 0.22
e
:0 0.18 0.22Q. 1000
~
'"
0
N 0.14 800
/
0~
"t;) 0.18 1050 Intercepts at~'1f 0.14 o Ma&1476±13 Ma
0.10 600 t;) (model2)
Allpoints
, '
/
MSWD =10.80.06 400 0.10
0 2 3 4 1.4 1.8 2.2 2.6 3.0 3.4 3.8 4.2
0.29 1600
C 0831 96- 2 0.28 d 071996-2
0.28 0.26
/ /
/ /
0.27 /./ / 0.24
-c-:
=> /~/
~ ///' 0.22
N /'
:0 0.26 /'/'
",Q.
0.20
0 N
0.25 1100
0.18
1000 Interceptsat
0.24 Intercepts at o Ma&1476±20 Ma
0 Ma&1506±10 Ma 0.16 MSWD =8.8
MSWD=1.3
0.23 0.14
3.0 3.2 3.4 3.6 1.4 1.8 2.2 2.6 3.0 3.4 3.8
207p bP 35U 207p bP 35U
Fig. 3. Concord ia diagr am s ofSIMSU-Pb data for theTinn granite (Tab le 1).Dat a-po intsare shown with1s erro rellipses.a: Allpoints, wit hareferen ce recent lead-lo ssline drawnto1500Ma.Not e wide ly discordant grains,plottingbetw eenthe refe renceline andconco rdia. These grainslo st rad ioge nic lead bothinSvecono rwegianand inrecent tim es andareomit te d fro mfurthercon sid erat ion .b:Dat afrom sam p le 083196-2, showinggrainswit hout visib lecores. Reg ressionlinethrou gh a forced low erinterceptat zerois shown.c:Coresinsample083196-2(white)comparedto grainsof mainpopu- lat ionin same sample(g ray).Regressionlines forthe mainpop ulat io n(dotted, seeb) and for cores(forced thr ou gh zero)areshown.d:Data from sam- ple071996-2,showing best-fitrecentlead-lossline.Points indicatedin gray suffe redpartiallead-lo ss in the Svecono rwegianorogeny and havebeen omi tted fro mtheregr ession.
tim e,and suggests that the partia l lead-loss observed in somezircons was indeed dueto a Sveconorwegian meta- morp hicoverprint. Thisage correspondswit hinoverlapping uncertaintieswith a regional lead isotope resetting event detected in metasedimentary rocks and in other felsic int ru-
Table 3.Leadisotopedata onwhole-rocksand rock -form in gmineralsof theTin ngranite.
2O'Pbl'''Pb "'Pb/'''Pb '01
Pb/""Pb 071996-2 Whole-rock 22.843 0.018 15.962 0.019 42.043 0.066
K-feldspar 18.476 0.017 15.640 0.021 37.645 0.067 Magneti te 665 69 0.060 19.110 0.026 71.859 0.127 Biot ite 59.604 0.054 18.692 0.026 74.026 0.131 Tit anite 96.349 0.087 21.389 0.029 72.433 0.128 Apatite 20.681 0.018 15.788 0.021 39.253 0.068
083196-2 Whole-rock 22.274 0.017 15.915 0.019 40.713 0.064
Whole-roc kand K·feldspardata from Andersenetal.(2001)
sions inSout hNorway(Heaman & Smalley,1994;Andersen&
Munz 1995,Simonsen1997).
Discussion
The age of eruption of the Tuddal format ion rhyolite is still not well determined, but the assumptionof ac.14 Ma period of volcanic activity by Sigmond (1998) is reasonable from whatis know nfrommodern andrecentgeologic analogs. In the sout hweste rnUnitedStates, numerou srhyolit ic volcanic centers formed in respon seto Cenozoic crustal extension (e.g.,Lipman1992).One of the largestand beststudied silicic volcanic centers in the world in the Timber Mountain- Oasis Valley caldera complexin southwestern Nevada.It is abou t 100 km long and 50 km wide and has existed for 16 Ma.
Silicicvolcanism predominated between 16 and 6 Ma,the most activity and voluminous magma prod uction was between12 and 10 Ma,sing lecalderaslasted1-2 Ma, some
NGU-BULL44 0,2002 - PA G E24 TOM ANDERSEN,ARTHURG.SYLVESTER&ARILD ANDRES EN
Fig.4.Leadisotopedat afor whol erocksand minerals from the Tinn granite; data from Tab le 2. The age represents an event of Sveconorwegianleadisot opeho mog enizat ion,whose timi ngis compa - rab le to ot her events inSout h Norway (e.g.Andersen & Mun z1995).
Ab breviat io ns:Ttn:tit anite;Mt:magnetite;Bi:bio t ite;Wr: whole rock;Ap:
apati te;Kfsp:potassiumfeldspar(micro cline).
were asshort as 100,000years,andthetimelapsebetw een individual erup tive even ts may have been only severaltens ofyears(Byers etal.1989).Thevolcaniccenter istypified by atleast 35separableim po rt anteruptive events.Basalticvol- canismbegan there at about 9 Ma and continue stopresent (Perryetal. 1998).
Themost preciseU-Pbage reported by Sigm ond (1998) for a Tud dal Format ion rhyol it e(1512+9/-8 Ma)and the 1509 +19/-3 age for a crosscutt ing int rusion com bine to suggest that the rhyoli tic magmat ism had terminated before 1500Ma.Thetwozirconcores dated here(1506±10 Ma)areco evalwith therhyolite and may havebeen inher- ite d from sucha source.An emplacementageof 1476±13
21
J:lQ.
;; 19
:a
"Q.
~
17 083196-2.Wr
• Wr Ap
15 Kfsp
10 30
TIn
Mt Bi
Allpoints:
Age=1031 0:32Mo MSWD =9.4
50 70 90
206Pb/204Pb
Ma thus makes the Tinn granite slightly youngerthanthe Tuddal Formati on rhyolite.Metamorphi sm of the granite post-dated its em placemen t by c.450 Ma,causing only minor lead-lossfrom zircon s.The present geochronologic datathusagree withtheinterpretationthattheTinngranite intr uded theTudd alFormation,and thatthe foliation-con- cordant nature of the contact bet weenthe two unitsis due to defo rmat ion dur ing emplacement or to later Svecon orweg ian(?)deformation.
At 1476 Ma,the Ndisotopi c composit ionof both of the samp les dated inthis studyfallswit hi n the wid e range of variation of theTudd al Formation metarhyolite (Menuge &
Brewer 1996, Brewer & Menuge1998;data for theTinn gran- itefromAndersen et al. 2001,see Tab le1);sample 083196-2 alsooverlaps with themuchmore restric ted range ofvaria- tion of the Vem or k Group metabasalt (Fig.Sa). Depleted mantlemodelages(DePaolo 1981mod el)of1.60 and1.69 Ga are wit hin the range of the Rj ukan Gro up (Brewer &
Menuge 1998).
The Tinn granite hasa very radiogen ic present -day Sr isoto pe comp osit ion,with8'Sr/86Sr wellabove1.0.The reason for this is a very high Rb/Srratio,whichis in turn due to anomalously low Sr contents combined wit h a normal upper-crustal Rbconcent rat ion(Fig 1,see alsoAndersen et al. 2001). In these feat ures,the Tinn granite resembles a group of post-tectonic Sveconorwegian granites from the Telemarksector(Iow- Sr concentration granite s'ofAndersen et al. 2001).and a range of metasedim ent ary rocks and gneissesof uncertain origin from the area westof the Oslo Rift(Andersen & Knudsen 2000).The present-daySrisotope comp osit ion of the Tinn granite fallswithinthe upper part of therangeofthe TuddalFormation(Fig.5b).When recalcu- late d to 1476 Ma,the Tinn granite still overlaps wit h the range of the rhyolite, but at unrealistically low,time -cor-
2 1.4 t,Ga 1.5
o
0.5a o +-~~~-'----'---~~~~---l...~--'-~~---'~~~~-'----'
20
C' (j)
~ 1.0
~
,!!>
co 0.8
o
0.4 0.8t,Ga
1.2 1.6
Fig.5. Sr andNd isotopeevolution diagramsfor theTinn granitecompared totheRju kanGroup.a:Ndisotopes.Gray shadin grepresents tot alrange of Tuddal Fo rm at io n;stripedfield is corresponding range of Vemork Formation(d ata fromBrewer& Menu ge1998). b:Sr isotopes.Gray shadi ng repre- sents tot al range ofTuddal Formation.Data from Kleppe(1980)and Verschureet al.(1990). See discussionintext.
TOM ANDERSEN,ARTHURG.SYLVES TER&ARILD ANDR ES EN NG U- B U L L440,20 0 2 - PAGE 25
rect ed8'Sr/86Sr« 0.70), indicatin gthat theSrisotopesyste m of thegranit e was part lyreset inSveconorweg ian time.The high Rb/Srrat ios in some oftheTuddal Format ion rhyo lite have been attributed to Sveconorwegian Rb- metasomat ism (Verschure et al. 1990),but Andersenet al.(2001) argued that thesimilar Rb/Srratiosobserved in Sveconorwegian low-Sr granites were inherited from the source,because none of these rocks is anoma lously enriched in Rb, yet they have consiste nt lylow Sr concent ratio ns.The same arg ume ntcan be used for theTinngranite.
The presenceof1506 ± 10 Mainheritedzircon coresin theTinn granite,anditspronounced similarityto the Tuddal Formation in Nd isotopesyste mat ics, suggestthat material related to therhyoliteofthe Tuddal Formatio nhas, indeed, been involved in thepetrogenesis of the Tinngranite,as a source for anatec tic melt, as asignificantconta minant,or as a parent magm a. Older,region ally distri buted,possiblepro- toliths in Sout hNorway include pre-1.65Ga TIBequivalents and other, st ill unidentified rocks witha crusta I prehistory back to 1.7-1.9Ga,which haveacted assource terranesfor the Seljord Grou p and other Mid Prot erozoic clastic sedi- mentary rocks (Knudsen et al. 1997,de Haas et al. 1999, Bingen etal. 2001).The Nd isot op e syste mat ics of theTinn granite,andthelack ofpre-1.51 Gainherit ed zircons indicate thatsuch rockswere notsignifica ntas sourcerocks forthe Tinn granite, nor werethey importan tas conta mi nants.
Petrogenesis of the Tinn granite
Thenorm ative mineralog yof theTudd alForm at ion rhyo lite is strongly domin at ed byqz,ab,and or;andnormat iveanis consistently very low,wit hanave rageof 0.97 % (Ta b le 1, datafrom Brew er &Menu g e 1998).The meandifferent iat ion indexis ashigh as 92± 7 (2(T),whichallows differenti at ion and part ialmelti ngofarhyo lit icprecur sorto beadequ ately reproduced by the ab-or-oz syste m,for whic h abu ndan t experi me nta l data areavailable (e.g. Joh annes &Holt z 1996 and referencestherein ).Both fract ion al crysta llizat ion ofa rhyolit ic magm a and parti al melting of Tuddal Formati on rhyolite would producemelt s at thetherm alminimum of thequart z-feld sparcot ect ic boundary in theAb-Or-Oz sys- tem at low to mod erate pressures,and at the albi te -K- feldspar-quar tzeutectic at higher pressures. Atlow to mod- erate pressures,minimum melt s wouldbehigh er in norma- tiveqzandlow er inanandorthan theTinn granite;at high er pressures
c-
5kbar),minim um melts would be less silicic,but would have signif icant ly higher ab/or rat iosthan observed (Johan nes& Holtz1996).The observed range of norm at ive aninthe Tinngranitecould be caused by accumu lationof alkali feldspar and plagioclase in a Tud dal-Iikemagma, but it is highly improbablethat aliquid withlessthan1%norma- tive an could accumu late eno ug h calcic plagioclase to increase theancontent by afact or of 4 to 8.The Tinngranite also has low conce nt rat ions of feld spar-compatibl e trace elements(Ba, Sr,Pb,Eu;Table 1), which doesnotag reewit h the presenceof accumulated feldspar.Thecompositi on al data thus sugges t thattheTinn gran- ite isneither a different iate of aTuddal parent magma, a cum ulate formed fromsucha magma, nor a simpleanatec- tic meltofa Tuddal Format ion protolith .The granite could represent a retarded batch of an undifferenti ated parent magmarelated to the Tuddalrhyolite,but the tim e int erval between the endof rhyoliti c volcanismand emplacementof the granite may betoolon gfor a sing lesilicicmagmatic sys- tem tohave remainedact ive.
The observed compos itions ofthe Tinngranitecanbe adequate lyexplained by mixing between a minimum melt and a lowqz-high an melt, i.e. between an anatect icmelt from a rhyolite-likeprotolith andamafic magma.The thick- ness oftheTudd alFormat ion has beenest imate d tobe min- imum7 km(Sigmond1998).Rockscogenet icwiththe rhyo- lit e, wit h ano ma lo usly low Sr concent ratio n,must also be presentat greater dept h in theTelemark area(And ersen et al. 2001,Andersen & Knudsen 2000). 10-20Maafter eruption of the Tuddal Formation rhyo lite, the volcanic pile and related intrusions atdeeper levels in the crust would proba- bly be hot enough to partially melt when heated up by inj ection of mafic magma.There is abundant evidence of mafic to interm edi atemagm atic act ivity in Telem ark after erupt io n of theTuddalFormation rhyolite (maficmemb ers of the Uvd al pluto nic belt,Vemork Formation basalts, e.g.
Sigm ond 1998),providing a sourcefor the necessary extra thermalenergy and a mafic component. An open-system process, involving maficmag ma and materialderi vedfroma crusta I prot ol ithrelated totheTudd alForm at ion, istherefore thepreferred petrogenetic model fortheTinn granite.The Tinn granite must haveform ed durin g aperiod of crusta l extensio n thatperm ittedmafic magmato ascend to a high crustalleveland,thereby,cause partialmelting of theupper crustas well asto mix and mingle with thederived silicic melts.
Conclusions
Zircon sfrom theTinn granitewere dated at 1476± 13 Ma, whichsug gest s that em p lacement ofthe granitepost -dates thefelsicvolcanismthatgaveriseto the Tudda l Format ion by at least 11 Ma,accepti ng the 1500-1514 Ma age est imat e of theTudd alForm at ionbySig mond(1998). Radiogenic iso- top edata and rare inh erit ed zircon cores indicatethat no material significantl y older than the Rjukan Group was involved in its petrogenesis.Whole-rockmajor eleme ntdata from the Tinn granite suggest that the granit ic mag ma formed by an open syste m processin which part ial melts fromaprot ol ith wit hageand Nd isotope systematics indis- tinguishab lefrom the Tuddal Formation were mixed wit h maficmaterial. Partia lmelti ng may have been induced by emplaceme nt of mafic magm a into the middle to up per crustof theTelem ark secto rafte r term in ation of therhyol it ic volcani sm,but whilethecrust st ill rem ain edhot.
