Fluid History of the Athabasca Basin and Its Relation to Uranium Deposits 1

T. Kotzer2 and T.K. Kyser2

Kotzer, T. and Kyser, T.K. (1990): Fluid history of the Athabasca Basin and its relation to uranium deposits; in Summary of Inves­ tigations 1990, Geological Survey, Saskatchewan Energy and Mines, Miscellaneous Report 90-4.

Studies of clays and other minerals in and around un­ mineral deposits (Gustafson and Williams, 1981). conformity-uranium deposits of the Athabasca Basin have resulted in genetic models that involve interaction This report summarizes data from stable and radiogenic between high-temperature basin and basement fluids at isotope and fluid inclusion analyses obtained mainly the unconformity between Aphebian metasedimentary from uranium deposits in the southeastern portion of and overlying Helikian sedimentary rocks (Hoeve and the Athabasca Basin (Figure 1) and places them within Sibbald, 1978; Hoeve and Quirt, 1984 and 1986). These a fluid evolution framework. models have been furthur refined by isotopic and fluid inclusion studies (Pagel et al., 1980; Bray et a/. , 1988; Wallis et al., 1983; Wilson and Kyser, 1987; Kotzer and 1. Fluid Inclusion and Isotopic Evidence Kyser, 1990), which indicate that uranium mineralization for Fluid Movements in the Athabasca has resulted from mixing of a high salinity, metal-bear­ Basin ing basinal brine with a reducing basement fluid at temperatures of 200°C along well-developed fault zones. The results of petrographic work carried out by Zones of fluid mixing are marked by well-developed combined with stable and radiogenic isotopic geochemical haloes containing illite, tourmaline, Mg­ compositions, fluid inclusion and scanning electron data chlorite, euhedral quartz and Ni-Co-As and Cu sulfides. allow formulation of a fluid-mineral-age paragenesis of A regional diagenetic assemblage of illite and kaolinite the Athabasca Basin (Figure 2). occurring within the Manitou Falls Formation throughout the basin attests to the high permeability of the sedi· Correlations between the petrographic and geochemical ments in the basinal aquifers that allowed large-scale data indicate that many of the basin-wide events which fluid flow. late-stage incursion of low-temperature meteoric fluids into the basin along the fault zones which host the uranium deposits has altered the isotopic and chemical composi­ tions of both the clay and uranium minerals. Relatively young K-Ar ages of illite and U-Pb ages of uranium minerals, forma­ tion of kaolinite in the fault zones, B remobilization of uranium into fractures in the re-activated fault zones (Wilson and Kyser, 1987; Katzer and Kyser, 1990) and formation of secondary sulphides with high­ ly variable c534S and Pb isotopic composi­ tions (Kyser et al., this volume) are all in· dicative of these late-stage fluid events.

The various types of fluids and fluid-flow events have characteristic mineralogical and geochemical signatures, so that it is o 50 100km likely that the metallogenic and mineralogic []lAthabasca Basin evolution of the Athabasca Basin has been L Crystalllne Basement controlled by large scale fluid events as­ • Uranium Deposits sociated with prograde and retrograde basin diagenesis. The association between Figu/'9 1 - Map indicating the p/'9sent extent of the Athabasca Basin, Joca· fluid flow events, basin diagenesis, and tions of uranium deposits and major lithostructural domains in the crystaJ. formation in the Athabasca l/ne basement of Saskatchewan (after Hoeve and Sibbald, 1978). MD • Mud· Basin is similar to the mechanism of ore for­ jatik Domain, WO " Wollaston Domain, PLD "' Peter Lake Domain, RD "' mation in other types of sediment-hosted Rotn,nstone Domain.

(1) ProJeC1 funded under NSERC Cooperative Research and DevelOpment PfOje(;t with CAMECO (2) Department GeolOgtcal sciences, University of Slskatchewan. Saskatoon, Saskatchew11n, S7N owo.

Saskatchewan Geological Survey 153 the inclusions reflects heterogeneities in the Hydrothermal late events temp Petrology fluid composition of the fluids during the initial ~••ve Alt'n I (°CI T pCM'e·f1ukb stages of basin diagenesis. 150- quartz overgrowth Ta, Cl5·20 Wt.l - 170 MaCll ~ Primary, three-phase, H20 fluid inclusions, dl-sa. k•ot.-ilite I •n r,KI co• atal containing halite, hematite and phyllosili­ ••t. I b•••ment fluid

154 Summary of Investigations 1990

------·--···-··-···-··· -- - McArthur River a} Eagle Point 20 b)

_] sands ton~ O\iCtgrowth 0 1 g cuhedral QUartz a, 10 I , ow.rgrowths J 0 2

' \00 200 JOO 4 00 500 \00 200 300 400 500 homogcmza!t0n temperature ·c

-::J Co, 30 30 = ?ch.ise (1, v) l'!1!I 2 phase (I+·:) r-::._ 3.,.,aso (l •v•sall) [_:] ~ l)t',asc (• •v+s.111)

cuh1:Clral Q.Jarli: ,. 20 g \l! uhed1al 10 · I Ql,,."\f:IIIIJ 10 quartz s1t.1er 11c sand~lo,,r: (Wt!rgrowlh A. I \0 20 JO •• 10 ,. JO •• sahruty (we 1. NaC l cQ ..v) sahn11y ( wl 'L Na.Cl equiv }

Figure 3 - Fluid inc/us/on histograms indicating the homogenization t&mperatures and salinities of fluid inclusions occurring in parBQenetica//y distinct minerals at a) McArthur River (Bermuda and Phoenix Lake) and b) Eagle Point North. Sim/Jar fluid inclusion temperatures and salinities in euhedral quartz at both McArthur River and Eagle Point North suggest the basinal fluid was quite per­ vasive.

development or restricted fault movements have im­ meteoric waters along re-activated fault structures host­ peded permeability. As most of the uranium deposits ex­ ing the uranium deposits. amined so far in the Athabasca Basin show some evidence of retrograde alteration, it can be concluded T~e occurrence o~ighl~ariable, anomalously high that the later, low-temperature fluid event was 20 Pb/204Pb and Pb/ Pb ratios in most of the sul- widespread. Therefore, the condition of the uranium ore phide minerals analyzed in the Athabasca Basin sug­ deposits today ls dependant upon the amount of per­ gests high lead mobility due to alteration of uranium meability existing around the uranium deposits at the mineralization by the numerous fluid events which have time of incursion of the later, oxidizing fluids. affected the Athabasca Basin (Kotzer et al., in prep).

Sulphur isotopes from sulphides occurring with uranium c) Radiometric Age Determinations ore lend furthur support to the model involving mixing of reducing basement fluids and oxidizing basinal fluids The wide range of ages from the uranium mineralization in fault structures that have focussed fluid flow. Nickel ar­ and sediments (Tremblay, 1982) reflects the complex senide and sulphide at Key Lake and copper sulphide fluid history of the Athabasca Basin. However, the minerals at Bermuda Lake, directly associated with general overlap of U-Pb ages from uranium mineraliza­ uranium minerals, and ~nsidered to be paragenetically tion and Rb-Sr ages of diagenetic clay minerals in the equivalent (Sl), have c5 S values indicative of mixing Athabasca sediments suggests that uranium mineraliza­ between two isotopically distinct sources and sulphide tion and high temperature basin diagenesis are closely formation during relatively closed-system, reducing con­ linked (Katzer and Kyser, 1990). ditions (Katzer et al., in prep.). Later formed iron sul­ phides peripheral to the uranium have a large range of 34 The timing of the high temperature diagenesis event in c5 S values which indicate sulphide formation during the sediments of the Manitou Falls Formation has been highly variable f02 conditions resulting from incursion of determined usinil Rb-Sr systematics on interstitial illites having similar c5 0 and do values. A Rb-Sr isochron

Saskatchewan Geological Survey 155 ca Basin and possibly represents a BASEMENT FLUID (200 ° C) - Mg·GhlOrile in Key major pulse of basement fluids out of Lake gneisses suuounding the fault zones to mix with basinal SMOW uranium deposil (G 1) fluids and furthur the production of 0 + high-grade uranium ore . .iii .,a.0 u c At the Eagle Point North uranium -so • deposit, an Rb-Sr model age of 957 Ei ::, .. c E Ma has been calculated from illite in al­ .. ~ ~ 0 tered pegmatite (Figure 5c). The a: :, BASIN FLUID (200 ° C) - "'.... younger age for the illite at Eagle Point <( -100 - Oiagenetic illite (1 1), kaolinite (Kl ) may indicate that the high temperature c~ dlavite (Tl) and euhedral quartz (Qt) diagenetic fluids in the Athabasca oO - Key Lake,' Midwest Lake Basin persisted for some time or that .iii the nature of the hydrothermal system -1 50 0 was episodic because the illite at Eagle a. 18 modern meteoric • kaol., ill., remobilized uranium ., c u ~ Point has similar cJO and cJ 0 values ' and Fe-chlor. (KJ. 11, U2) E o to the illites having an age of 1477 Ma. Key Lal

18 ca Basin was affected by early, high Figure 4 - Calculated cl 0 and do valutJs for various fluids associated with un­ salinity diagenetic fluids having conformity-type uranium dt1posffs in thtJ Athabasca Basin. Shaded areas repr&­ temperatures near 200°C and by later, StJnt fluids in tJquilibrium with: 1) ~hlorite formed at approKimattJly 200"C in bast1mt1nt rocks at Kt1y LaktJ (Wilson and Kyst1r, 1987) and Bermuda LaktJ in the meteoric fluids having temperatures McArthur River artJa (Bast1ment Fluid), 2) diagentJtic illite at 200"C from Key LaktJ, less than 1OO'C. Coincident with the Midwest LaktJ (Wilson and Kyst1r, 1987), McArthur River and Eagle Point (KotztJr high and low-temperature fluid events and Kyst1r, 1990) (Basin Fluid) and, 3) kaolinite associated with remobil/zed are periods of uranium deposit forma­ uranium in fractures which is similar to thtJ fluids measured in fluid inclusions in tion and destruction, respectively, with siderlte. Also shown are the values for modtJm meteoric waters in thtJ Athabasca the magnitude of uranium formation Basin, the meteoric water /intJ (MWL) and ocean water (SMOW). and destruction directly dependant on the quantities of reactive fluids in­ age of 1477 ± 57 Ma (Figure 5a) for diagenetic illite for­ volved. The late, meteoric fluid event mation in the Athabasca sediments pre-dates the ear­ has remobilized much of the uranium and has had the liest age for uranium emplacement at 1406 Ma (Carl et most pronounced effect on the current state of some of al., 1988) in the basin and suggests that large-scale fluid the uranium deposits in the Athabsca Basin. The mag­ flow occurred before uranium mineralization. The time nitude of uranium deposit destruction is directly related difference of approximately 50 Ma between high to the permeability developed within the sediments and temperature diagenesis and uranium emplacement fault structures hosting the uranium deposits. would allow the fluids to leach sufficient quantities of uranium from heavy minerals in the Athabasca sedi­ 3. References ments_ Armstrong, R.L. and Ramaekers, P. (1985): Sr isotopic study Euhedral quartz-dravite (02-11) assemblages occur in of He!ikian sediment and diabase dikes in the Athabasca the hydrothermally altered sediments associated with Basin, northern, Saskatchewan; Can. J. Earth Sci., v22, the uranium deposits in the Manitou Falls Formation. In p399-407. some areas of the Athabasca Basin, strongly developed Bray, C., Spooner, E.T.C. and Longstaffe, F.J. (1988): Uncon­ zones of euhedral quartz-dravite breccias are evident formity-related uranium mineralization , McClean and appear to be the result of early dissolution of deposits, northern Saskatchewan, : Hydrogen and Athabasca sandstones. An Rb-Sr age of approximately oxygen isotope geochemistry; Can. Mineral., v26, p249- 1270 Ma (Figure 5b) has been determined for this event 268. using the Rb-Sr and 87Sr/ 86Sr ratios from both the tour­ maline and fluids extracted from the fluid inclusions in Carl, C., Hoehndorf, F., Pechmann, E.V., Strnad, J.G., and coexisting euhedral quartz. The age determined from Ruhrmann, G. (1988): Geochronology of the Key Lake this event is similar to the ages of much of the uraninite uranium deposit, Saskatchewan, Canada (abstract); J. Chem, Geol., v70, p133. at Key Lake (Ruhrmann, 1987) and some of the diabase dikes (Armstrong and Ramaekers, 1985) in the Athabas-

156 Summary of Investigations 1990 Gustafson, L.B. and Williams, N. (1981): Sediment-hosted 0.78 ~------, stratiform deposits of copper, lead and zinc; in Skinner, (a) Dlagenelic lllites (11 ) o.n B.J. (ed), Econ. Geo!. (Seventy-fifth Anniversary Volume), p139-179. 0.76 en o.1s Hoeve, J. and Slbbald, T.1.1. (1978): On the genesis of Rabbit Lake and other unconformity-type uranium deposits in 0.74 northern Saskatchewan, Canada; Econ. Geol., v73, p1450- VI o.n 1473. t • l 47"/t57 Ma 0.72 Sr • . 7064 Hoeve, J. and Quirt, D. (1984): Mineralization and host-rock al· Ink, 0.71 teration in relation to clay mineral diagenesis and evolu­ tion of the middle-Proterozoic Athabasca Basin, northern Saskatchewan, Canada; Sask. Ras. Counc., Tech. Rep. #187, 187p. 11 1 e, Rb/ Sr {1986): A common diagenetic-hydrothermal origin -~f-or_u_n-conformity-type uranium and stratiform copper deposits; Sask. Res. Counc., Publ. R-8555-5-A-81 .

Katzer, T.G. and Kyser, T.K., (1990): The use of stable and 0.7090 radiogenic isotopes in the Identification of fluids and VI processes associated with unconformity-type uranium .. 0.7088 deposits; in Beck, LS. and Harper, C.T. (eds.), Modern Ex­ . ploration Techniques, Sask. Geol. Soc., Spec. Publ. 10, .:: 0.7086 C/l p115-131 . 0.70114 ~ t • 1268.9 Ma .. Kotzer, T.G., Kyser, T.K. and Ruhrmann, G., Qn prep.): Sul­ 0.7082 Sr • . 70814 ir:I:. . phur and lead isotopic constraints on the s.ources and ages of the fluids involved with unconformity-type 0.7080 0 .00 0.0 ! 0 .02 0.03 0.04 0.05 0.06 uranium deposits; Can. J. Earth Sci.

7 6 • Ab/ • Sr Pagel, M., Poty, B. and Sheppard, S.M.F. (1980): Contribu­ tions to some Saskatchewan uranium deposits mainly from fluid inclusion and isotopic data; in Ferguson, S. and (c) Eagle Poinl lllile (11?) Goleby, A.B. (eds.), Uranium in the Pine Creek Geosyncline; IAEA, Vienna, p639-654.

2 Ruhrmann, G. (1986): The Gaertner uranium orebody at Key .. en Lake (northern Saskatchewan, Canada) - After three years .. of mining: An update of the geology; in Gilboy, C.F. and .:: Vigrass, L.W. (eds.). Economic Minerals of Saskatchewan, C/l Sask. Gaol. Soc., Spec. Publ. 8, p120-137. ~ . t ';I 95 7 Ma. r :.it!c: l Tremblay, LP. (1982): Geology of the uranium deposits re­ assumed Sr • . 1064 .i n J ~.• lated to the sub-Athabasca unconformity, Saskatchewan, Canada; Geo!. Surv. Can., Pap. 81 ·20, 56p. 0 20 40 ~o 80 100 8 l 8 6 Wallis, R.H., Saracoglu, N., S..ummer, J.J. and Golightly, J.R. Rb/ Sr (1983): Geology of the McClean uranium deposits; Geol. Surv. Can., Pap. 82-11, p71-110.

Figure 5 - Rb-Sr isochrons of: a) Jnterstltlal, dlagenetic itlite in Wilson, M.R. and Kyser, T.K. (1987): Stable isotope the Athabasca sandstones give an age of 1477 ± 57 Ma, b) the geochemistry of alteration associated with the Key Lake euhedral quartz-dravite event gives an age of approximately uranium deposit, Canada; Econ. Geo!., v82, p1540-1557. 1270 Ma and, c) diagenetic itlite at Eagle Point3fves an Rb-Sr model age of 957 Ma assuming an Initial 81Srrsr ratio of 87 Wilson, M.A., Kyser, T.K., Mehnert, H. and Hoeve, J. (1987): 0.7064. Varying thfl assumed initial s,_rs, ratio would have lit· Changes in the H-0-At isotopic composition of clays tie affect on the model age. during retrograde alteration; Geochim. Cosmochim. Acta, v51 , p869-878.

Saskatchewan Geological SuMy 157