Temperatures on Mars from 40Ar/39Ar thermochronology of ALH84001IntroductionThermochronologyImplicationsAcknowledgementsReferencesTemperatures on Mars from40Ar/39Ar thermochronology ofALH84001Benjamin P. Weiss, David L. Shuster, Sarah T. StewartDivision of Geological and Planetary Sciences, 170-25, California Institute of Technology, Pasadena, CA 91125, USA Received 12 February 2002; received in revised form 16 April 2002; accepted 16 May 2002AbstractThe thermal history of Martian meteorite ALH84001 has critical implications for its petrological and deformationalhistory, the age of its trapped atmospheric gases, the timing of the Martian dynamo, and possibly the evolution ofMartian surface temperatures during the last 4 billion years (Gyr). Feldspathic glass in ALH84001 has been datedusing40K/40Ar and40Ar/39Ar chronometry by several laboratories. There is general agreement that thesechronometers were last reset sometime between 3.9 and 4.3 billion years ago (Ga). Using the40Ar/39Ar data fromBogard and Garrison [Meteorit. Planet. Sci. 34 (1999) 451^473] to model several limiting thermal histories of themeteorite, here we show that most of ALH84001 feldspathic glass has probably not been heated to more than V350^500‡C and shocked to peak pressures V s 1 GPa since the glass was last melted. This indicates that most ofALH84001 has been well below these temperatures since 3.9^4.3 Ga. Since these temperatures are below the Curiepoint of magnetite, much of the magnetization recently identified in ALH84001 carbonate [Weiss et al., Earth Planet.Sci. Lett. this issue] must have been acquired by V4 Ga. This also provides an explanation for why ALH84001contains a sample of an apparently ancient Martian atmosphere that is less evolved relative to that on present-dayMars. Our calculations also suggest that for the last 4 Gyr, average surface temperatures on Mars may not have beenmuch higher than the present cold conditions. B 2002 Elsevier Science B.V. All rights reserved.Keywords: Mars; ALH84001; Martian meteorites; argon; thermochronology; temperature; paleomagnetism; climate change1. IntroductionThe meteorite ALH84001 is an orthopyroxenecumulate from Mars with a Sm/Nd age of 4.50Ga [3,4]. This extreme age ^ the oldest known forany planetary rock ^ is re£ected by the multipleepisodes of intense deformation recorded by themeteorite’s petrofabric [5]. During the ¢rst fewhundred million years of its history, ALH84001was partially melted and fractured by severalshock events. In some of these fractures formedzoned carbonate blebs with Rb/Sr and Pb/Pb agesof 3.90 I 0.04 Ga and 4.04 I 0.1 Ga, respectively[6] (although the Rb/Sr dates are controversial[7]). Feldspathic glass, which makes up V1 wt%of the meteorite, is ubiquitously distributedthroughout the rock and very commonly sur-0012-821X / 02 / $ ^ see front matter B 2002 Elsevier Science B.V. All rights reserved.PII: S0012-821X(02)00729-X* Corresponding author. Tel.: +1-626-395-6187;Fax: +1-626-568-0935.E-mail addresses: [email protected] (B.P. Weiss),[email protected] (D.L. Shuster),[email protected] (S.T. Stewart).EPSL 6277 25-7-02 Cyaan Magenta Geel ZwartEarth and Planetary Science Letters 201 (2002) 465^472www.elsevier.com/locate/epslrounds carbonates [5]. The texture, shape, andstructure of this glass indicate that it was shockedand mobilized (and probably melted) at least onceafter the carbonate formed [2,5,8^10]. The lastsuch event was labeled ‘D3’ by Treiman [5], dur-ing which ALH84001 apparently experiencedpeak pressures of 40^60 GPa [5,11] and post-shock temperatures of 400^1000‡C [10^12].Several laboratories have conducted40K/40Arand40Ar/39Ar dating of ALH84001 using steppedheating of bulk grains [1,13^17] and by laserprobe [17]. Essentially all experimenters agreethat the dominant carrier of K and radiogenic40Ar in ALH84001 is feldspathic glass. There isalso near universal agreement that the meteoritelast completely degassed40Ar sometime between3.9 and 4.3 Ga (depending on the compositionassumed for trapped Ar). On the other hand, laserprobe dating [17] has identi¢ed a few locations inthe meteorite having40Ar/39Ar ages of 4.4 Ga.This suggests that while much of the meteoritewas strongly heated sometime during 3.9^4.3Ga, isolated locations may have escaped heatingsince 4.4 Ga, possibly due to the spatially hetero-geneous nature of shock heating [18]. The coinci-dence of the main V4 Ga population of40Ar/39Ar ages with those of impact glasses from thelunar highlands provides the ¢rst direct evidencethat Mars may have experienced a heavy bom-bardment of impactors contemporaneous withthe lunar cataclysm [19].Here we demonstrate that the40Ar/39Ar chro-nometer of ALH84001 was probably last reset bythe D3 shock event that mobilized its feldspathicglass. This was presumably the result of an impacton the Martian surface. We show that ever sincethis time, the meteorite has experienced very mildtemperatures and shock pressures.2. ThermochronologyUsing the39Ar release data of Bogard and Gar-rison [1], we modeled (following the methods of[20]) the temperature dependence of the di¡usioncoe⁄cient of Ar through ALH84001 feldspathicglass using an Arrhenius relationship, D(T)=D0/a2exp(3Ea/RT), for characteristic constants D0/a2, Ea, and the gas constant R (Fig. 1). From thiswe derive best-¢t values of ln(D0/a2)=31.9 I 0.2ln(s31) and Ea= 76.8 I 1.3 kJ mol31(these uncer-tainties are formal errors from the regressionanalysis only), with a linear correlation coe⁄cientr2= 0.998 (number of heating steps n = 12). In thecalculations presented below, we assume that thisArrhenius relationship has held for ALH84001glass since it was last mobilized during the D3deformational event and that the di¡usion do-main size a is constant for all time in our calcu-lations.Because39Ar resides in several distinct phases[1], we used39Ar released only from the high K/Ca phases (i.e., the 200^850‡C steps) in our cal-culation of ln(D/a2)ifor each step i (Fig. 1). Fol-lowing the conclusions of Bogard and Garrison[1] and Turner et al. [17], we assume that39Arreleased during steps s 850‡C (which presentsas anomalously low ages in the age spectrumbetween 80 and 92% cumulative gas release; seeFig. 2) is the product of recoil, and so weexcluded these from the regression (Fig. 1)Wealso excluded the three steps 6 450‡C from theregression (Fig. 1) because some of that gas waspossibly derived from a recently precipitatedweathering product