TECTONICS, VOL. 8, NO. 4, PAGES 865-880, AUGUST 1989
40Ar/39Ar AGE CONSTRAINTSON DEFORMATION AND METANDRPHISM IN THE MAIN CENTRAL THRUST ZONE AND TIBETAN SLAB, EASTERN NEPAL HIMALAYA
Mary S. Hubbard1
Department of Earth, Atmospheric and Planetary Sciences, Massachusetts Institute of Technology, Cambridge
T. Mark Harrison
Department of Geological Sciences, State University of New York at Albany
Abstract. In eastern Nepal, structural 6.4+0.8 Ma (Tc=225ñ20ø C). An adjacent and petrologic observations suggest that gneiss yields a 9.1ñ0.2 Ma biotite movement on the Main Central Thrust (MCT), isochron and a K-feldspar minimum age of the development of an inverted metamorphic 3.6_+0.2 Ma (Tc=210-+50øC). In the upper sequence, and leucogranite plutonism were Tibetan Slab, samples collected >200 m genetically related. Samples from across from visible intrusives yield a biotite the MCT zone and its hanging wall, the isochron age of 20.2!0.2 Ma and a complex Tibetan Slab, were collected from four hornblende age spectrum suggestive of locations in the Everest region for excess argon. At the fourth location an 40Ar/39Ar studies: (1) the MCT zone, (2) augen gneiss, a pegmatite, and a the lower Tibetan Slab, (3) the upper tourmaline-bearing leucogranite, all in Tibetan Slab away from intrusive rocks, mutual contact, yield indistinguishable and (4) the upper Tibetan Slab adjacent to mineral ages. The average biotite and and including leucogranitic material. muscovite isochron ages for these samples Within the MCT zone, muscovite and are 17.0_+1.4 and 16.6ñ0.2 Ma, hornblende yield cooling ages of 12.0-+0.2 respectively. The minimum age for K- Ma and 20.9ñ0.2 Ma, respectively. K- feldspar from the leucogranite is 15.5ñ1.8 feldspar gives a minimum age of 8.0-+0.2 Ma Ma. As thermobarometric data from the MCT (closure t•rature (Tc) =220ñ15ø C) . In zone in this area suggest synmetamorphic the lower Tibetan Slab a sample from a deformation at t•ratures of ~500ø- sheared pegmatite yields muscovite and 550ø C, the MCT hornblende (Tc=~500ø C) biotite isochron ages of 7.7_+0.4 Ma and dates this event at ~21 Ma. Diffusion 7.5_+0.6 Ma and a K-feldspar minimum age of experiments on hornblendes from the MCT zone provide data which support a maximum duration of peak metamorphic t•ratures of <_2 Ma. Biotite ages as old as ~20 Ma from the upper Tibetan Slab near 1Nowat Geologisches Institute, ETH leucogranites indicate that leucogranite Zentrum, Z•rich, Switzerland. intrusion was essentially coeval with deformation and metamorphism in the MCT Copyright 1989 zone. Very young ages in a ductile shear by the American Geophysical Union. zone in the lower Tibetan Slab suggest that there has been deformation within the Paper number 89TC00708. Tibetan Slab that postdates major movement 0278-7407/89/89TC-00708510.00 on the MCT. 866 Hubbardand Harrison: 40Ar/39ArAges, Nepal Himalaya
INTRODUCTION the Himalaya' (1) the southernmost, and possibly seismically active, Main Frontal Collision between the Indian and Thrust, (2) the Main Boundary Thrust, and Eurasian continents occurred between ~40 (3) to the north the Main Central Thrust. and 50 Ma [Molnar, 1984]. Convergence The Main Frontal thrust separates terraced between these two continents has continued alluvial deposits from the overriding to the present, creating the spectacular Early Miocene - Late Pleistocene age Himalayan range and the elevated Tibetan molassic rocks, locally known as the Plateau. The original zone of collision, Siwaliks [Gansser, 1981]. To the north the Indus-Tsangpo suture, is marked by the Siwaliks are overridden along the Main ophiolitic rocks in southern Tibet. The Boundary Thrust (MBT) by the Lesser high peaks and rugged topography of the Himalayan sequence, which consists of a Himalaya coincide with metamorphic and predominantly clastic sedimentary sequence intrusive rocks within the Indian plate. of Precambrian to late Paleozoic age, These crystalline rocks are part of the although fossil control is scarce upper plate to a major, north- dipping [St6cklin, 1980]. This sequence generally thrust known as the Main Central Thrust has been metamorphosed to a low grade, but (MCT). local areas remain unmetamorphosed . The The MCT is a major intracontinental Main Central Thrust (MCT) forms the fault zone that accomodated portions of northern boundary of the Lesser Himalaya the postcollisional convergence between (Figure 1). The upper plate to this India and Eurasia. This structure may structure is the Tibetan Slab with high- have accomodated more than 100 km of grade sillimanite-bearing gneissic rocks displacement [Gansser, 1981]. It has been which have been subjected to widespread proposed that movement on the MCT may have anatexis and intruded by leucogranitic influenced metamorphism and the material. Rocks adjacent to the MCT are development of leucogranites in the characterized by an inverted metamorphism Himalaya [LeFort, 1975, 1981]. such that metamorphic grade increases Unfortunately, these hypotheses have toward structurally higher levels remained virtually untestable because [Gansser, 1964; P•cher, 1977; Hodges et relative and absolute ages of al., 1988]. These rocks generally dip metamorphism, thrusting, and leucogranitic homoclinally to the north, thus the •mplacement are lacking. Many of the metamorphic grade increases northward. A leucogranites have been dated (see LeFort variety of ideas have been proposed that et al. [1987] for summary and Copeland et link the inverted metamorphism with al. [1988b] for a discussion of possible deformation along the MCT [LeFort, 1975; ambiguities in age assessment), but the Maruo and Kizaki, 1983; Brunel and Kienast, timing of metamorphism and deformation 1986]. In central Nepal, LeFort [1981] along the MCT has not been studied in proposed that the formation and detail. emplacement of the Manaslu leucogranite in In the Everest area of eastern Nepal the Tibetan Slab is also related to the MCT is a thrust zone in which movement on the MCT. Radiometric ages, temperatures reached 500ø-700øC during from a variety of techniques, range from thrusting [Hubbard, 1989]. Amphibolite 14.3+0.6 to 25.0!_0.5 Ma (2-sigma occurs in the lower part of the zone where uncertainties are reported herein for all calculated temperatures are close to the radiometric ages) for intrusives within closure temperature for the K-Ar system in the Tibetan Slab [SchRrer et al., 1986; hornblende (~500 ø C) . This coincidence Deniel et al., 1987], although Copeland et provides an opportunity to date thrust al. [1988a] have criticized the basis for movement and to compare this age with the assessment of many of the older ages. ages of metamorphism within the hanging Movement on the MCT is generally thought wall and with the ages of leucogranites. to predate movement on the MBT, although In this paper we report the results of a no unambiguous age constraints exist for regional 40Ar/39Arstudy of hornblende, the temtooral development of either biotite, muscovite, and potassium feldspar structure. from the MCT zone and its upper plate. Study Area _ GEOIf•IC SETTING The structural setting and tectonic Three laterally continuous, north- stratigraphy of the Everest region are dipping thrusts have been identified in described in detail by Hubbard [1988]. In Hubbardand Harrison: 40Ar/39ArAges, Nepal Himalaya 867
86o40 ' MCT zone and across the Tibetan Slab. Metamorphic isograds and lithologic contacts are roughly parallel to the dominant schistosity. The lower boundary of the MCT zone is a lithologic contact between an augen gneiss in the MCT zone and a chlorite-muscovite schist in the uppermost Lesser Himalaya. The Lesser Himalayan section is a thick sequence of fine-grained psammite, schist, and quartzite. Locally, biotite and garnet occur in the uppermost unit of chlorite-muscovite schist. The upper boundary of the MCT zone is within a biotite gneiss unit at a level that separates highly sheared rocks of the thrust zone from overlying, less-deformed but migmatizedgneissic rocks of the Tibetan Slab. Shear zones do exist north of the MCT zone, but deformation does not appear to be as widespread. Lithologies within the Tibetan Slab include biotite gneiss, calc-silicate rock, quartzite, augen gneiss, granitic gneiss and pelitic T. schist. Garnet and sillimanite are common in pelitic andgneissic rocks throughout this section. Anatectic granites and small-scale granitic intrusions occur in the lower Tibetan Slab, but the frequency and volume of graniticmaterial increases toward the north. There are several large 5km' 27030 ' leucogranitic bodies in this region, such :.'•'• GRANITE • MCTZONE as the south face of Nuptse or the Makalu massif, but the majority of the granitic I•1 TIBETANSLAB ::1• LESSERHIMALAYA material occurs in networks of dikes and sills. Some of the larger intrusive Fig. 1. Simplified geologic map of the bodies have retained xenoliths of the study area in eastern Nepal. Sample country rock. Schistosity within these locations for 40Ar/39Arstudy are xenoliths has often remained parallel to indicated by the circled place names. the country rock schistosity in the area. Samples 87H13F, 87H13E, 87H12C, and Crosscutting relationships in the dikes 87H12B are from the Main Central Thrust and sills suggest multiple generations o• (MCT) zone. Samples 87H21C, and 87H21D intrusion. are from Ghat. Samples 87H19A, and Metamormhism 87H19Bare from Na. Samples 87H18A, _ 87H18B, and 87H18C are from Ngozumba. Metamorphic assemblages across the MCT zone and into the Tibetan Slab suggest an general, the MCT in this area is a 3- to increase in metamorphic grade toward 5- km-thick, north-dipping shear zone higher structural levels. Mineral rim consisting of a variety of lithologies thermobarometric data [Hubbard, 1989] from including augen gneiss, slate, phyllite, samples across the MCT zone and lower quartzite, talc-schist, amphibolite, Tibetan Slab in eastern Nepal show an pelitic schist, calc-silicate rock, and increase up-structure in tentoeratures of banded biotite gneiss. Metamorphic grade metamorphic equilibration from 505ø_+50øC increases up-structure, or northward, from (2-sigma uncertainty) in the lower MCT chlorite-bearing schist in the Lesser zone to 725ø_+96øC at about 1.5 kilometers Himalaya, which contains garnets in the structurally above the upper boundary of upper ~100 m, through the staurolite and the MCT zone (Figure 2). Above that level kyanite isograds in the lower MCT zone, to t•ratures decrease to 593ø_+50øCat ~8 sillimanite-bearing gneiss in the upper km structurally above the MCT zone. 868 Hubbardand Harrison' 40Ar/39ArAges, Nepal Himalaya
900 MCT zone (a) i
800
700
600
500
400 i ß i ! i ß i ß 0 5 10 15 20 STRUCTURAL DISTANCE FROM KY ISOGRAD (krn) y = 680.2246 - 28.3211x R = 0.54 1200 (b) MCT zone
1000
800
600
400
200 . I i I .. -5 0 5 10 15 20 STRUCTURAL DISTANCE FROM KY ISOGRAD (kin) Fig. 2. Summaryof thermobarometric data plotted against structural distance from the kyanite isograd [from Hubbard, 1989] (a) temperature, (b) pressure.
Pressures across the same section decrease variety of techniques. Because of from 685+160 MPa in the lower MCT zone to limitations of the individual techniques 328_+102 MPa at ~8 km above the MCT zone there has been a mixed degree of success. (Figure 1). These data support the model Table 1 sun•narizes the majority of of LeFort [1975] that metamorphism was existing age data for this region. synchronous with movement on the MCT. Ferrara et al. [1983] used the Rb-Sr In the upper Tibetan Slab, metamorphic method to obtain both whole- rock temperatures and pressures from samples isochrons and mineral isochron ages for across a folded section are nearly samples from the south side of the peaks uniform, with average values of 712øC and Nuptse and Lhotse. Two suites of samples 533 MPa. It has been proposed [Hubbard, gave whole- rock ages of 52!1 Ma and 1989] that these samples either came from 13.7_+7 Ma. One of the samples from the 52 one structural horizon, thus experiencing Ma suite yielded a 17.3_+0.3 Ma internal identical uplift histories, or they may isochron from K-feldspar, biotite, have been heated during the period of muscovite, and whole rock. Another leucogranitic intrusion in this area. In sample, not included in either whole- rock the latter case, regional metamorphism in isochron, had an internal four-mineral the upper Tibetan Slab would be the result isochron age of 16.5+0.5 Ma. Ferrara et of advection of heat during intrusion. a!. suggested that the middle Miocene ages could represent a period of regional PREVIOUS GEOCHRONOLOGY uplift, but they also recognize that Sr heterogeneity may be a problem in Tertiary_ Granites of the Everest Area Himalayan leucogranites. SchRrer et al. [1986] determined a A small number of workers have monazite U-Pb age of 14.4!0.6 Ma from an attealoted to obtain ages of the Tertiary undeformed granite sample from the north leucogranites in eastern Nepal using a side of Mount Everest. This age is Hubbardand Harrison' 40Ar/39ArAges, Nepal Himalaya 869
TABLE 1. Previous Geochronologic Results, Mount Everest Region
S_y_•tem Locat ion Rock Type Mineral Age* Source Rb/Sr Lhotse glacier leucogranite whole rock 52+1 Ma 1 Rb/Sr Lhotse glacier leucogranite whole rock/biotite 17.3+0.3 Ma 1 muscovit e/K- feldspar Rb/Sr Nuptse glacier leucogranite whole rock 14+7 Ma 1 Rb/Sr Lhotse glacier leucogranite whole rock/biotite 16.5+0.5 Ma 1 tourmaline/K-feldspar muscovite Rb/Sr near Na orthogneiss whole rock/biotite 15.6+0.6 Ma 1 Rb/Sr near Na orthogneiss whole rock/biotite 15.3+0.6 Ma 1 Rb/Sr Lhotse glacier paragneiss whole rock/biotite 16.4+0.3 Ma 1 U/Pb Mount Everest leucogranite monazite 14.4+0.6 Ma 2 U/Pb Makalu leucogranite monazite 21.9+0.2 Ma 3 U/Pb Makalu leucogranite monazite 24.0+0.4 Ma 3 U/Pb north side Everest leucogranite zircon/monazite 20+1 Ma 4 40Ar/39Ar north side Everest leucogranite biotite 17.1+0.2 Ma 4 40Ar/39Ar north side Everest leucogranite muscovite 16.7+0.4 Ma 4 40Ar/39Ar north side Everest leucogranite K-feldspar 16.2+0.2 Ma 4 K-Ar Lho La Pass leucogranite biotite 19•_2 Ma 5 K-Ar Pumori leucogranite biotite 47+5 Ma 5 K-Ar Everest basecamp leucogranite biotite 19!_2 Ma 5 K-Ar north side Nuptse mica schist biotite 17.0+_1.6 Ma 5 K-Ar north side Nuptse amphibolite hornblende 24+2 Ma 5 K-Ar north side Nuptse biotite gneiss biotite 18.0+1.6 Ma 5 K-Ar north side Nuptse amphibolite amphibole 56+4 Ma 5 K-Ar Pangboche amphibolite biotite 40+4 Ma 5 K-Ar Pangboche amphibolite hornblende 350+40 Ma 5 K-Ar Tyangpoche biotite gneiss biotite 16.0+1.8 Ma 5 K-Ar south of Namche augen gneiss biotite 13.4+1.2 Ma 5 K-Ar south of Namche migmatite biotite 10.5+2.0 Ma 5 K-Ar south of Namche migmatitic diorite biotite 10+2 Ma 5 K-Ar south of Ghat biotite gneiss biotite 14.0+1.0 Ma 5 K-Ar Puiyan (MCT) zone augen gneiss biotite 9.0+1.0 Ma 5 K-Ar Puiyan (MCT) zone biotite gneiss biotite 5.5+0.6 Ma 5 K-Ar north of KhariLa (MCT) biotite gneiss biotite 3.4+0.2 Ma 5 K-Ar south of KhariLa (MCT) mica schist biotite 8.5+0.8 Ma 5 * With two sigma uncertainty. Sources' 1, Ferrara et al. [1983]' 2, Sch•rer et al. [1986]; 3, Sch•rer [1984]' 4, Copeland et al. [1987]; 5, Krummenacher et al. [1978] considerably younger than zircon and Metamorphic Rocks of the Tibetan Slab monazite crystallization ages determined by Copeland et al. [1988b] from Ferrara et al. [1983] attempted to neighboring Everest granites at 20+_1Ma determine the age of gneissic samples from and ages of 21.9!_0.2 Ma and 24.0+0.4 Ma the south face of Lhotse using Rb-Sr on determined by SchRrer [1984 ] on monazites whole- rock samples. Significant scatter from granites of the Makalu massif, in the data did not permit calculation of roughly 20 km to the southeast. These age a meaningful age but rather was differences may result from multiple interpreted as indicating isotopic intrusive events. heterogeneity in the sedimentary Copeland et al. [ 1987 ] used the protolith. They did obtain an age of 40Ar/39Artechnique on biotite, muscovite, 16.4!0.3 Ma from a biotite-whole- rock and K-feldspar from a granite just north pair that is consistent with the ages of of Everest and obtained cooling ages of leucogranites from the same area. 17.1+0.2, 16.7+0.4, and 16.2!0.2 Ma, Krunmenacher et al. [ 1978 ] obtained K- respectively. Similar K-Ar ages were Ar biotite ages on schistose and gneissic determined by Krummenacher et al. [1978 ] rocks from across the Tibetan Slab. In on two biotite separates from granites general, their ages range from 10.5+2.0 to near the Everest base csalo in Nepal at 18.0+_1.6 Ma. With a few exceptions the 19!_2. Krummenacher et al. [1978] also ages tend to decrease structurally measured a 47+5 Ma biotite from a granite downsection. One anomalously old age of in the Everest region. This latter age 40+4 Ma may be due to excess Ar. could either represent an older intrusion Copeland et al. [1988b] found or reflect excess Ar. I! inherited monazite in a leucogranite 870 Hubbardand Harrison: 40Ar/39ArAges, Nepal Himalaya
Fig. 3. Protomylonitic texture in pegmatite from Ghat. Field of view is 13 mm. sample from northwest of Everest that from the lowest unit in the MCT zone. In revealed an upper intercept of 471!10 Ma, this sample, K-feldspar porphyroclasts suggesting a protolith in the Tibetan Slab (
Fig. 4. Outcrop that was sampled at Ngozumba. The augen gneiss (ag), pegn•tite (p), and leucogranite (lg) are labeled.
Upper Tibetan Slab' Na Correction factors to account for interfering nuclear reaction products are Two samples were collected south of Na given, together with the analytical in the Gokyo valley at a site > 200 m from results for each sample, in Appendix visible leucogranitic intrusives. The 1,available on microfiche 1. intent was to analyze samples from this Closure temperatures for hornblende, structural level whose thermal history biotite, and muscovite were estimated from reflects uplift and not intrusion. Sample hydrothermal diffusion studies (see 87H19A is an amphibolite and 87H19B is a McDougall and Harrison [ 1988 ] for a tourmaline-bearing biotite gneiss. Both review). Closure temperatures for K- samples are foliated but neither shows feldspars were calculated using the evidence of intense deformation. diffusion parameters estimated from measured39Ar release during laboratory Upper Tibetan Slab' Ngozumba heating. Three of the K-feldspars yielded reasonable values for closure temperature. The outcrop at Ngozumba (Figure 4), in Unfortunately, the K-feldspars from the upper Gokyo valley, consists of an gneissic and granitic rocks of the upper augen gneiss (87H18A) that is intruded by Tibetan Slab were partly sericitized and pegmatite sills (87H18B) and cut by a gave unrealistically low closure tourmaline-bearing leucogranite (87H18C). temperatures. For these samples a closure Myrmekites are present in the augen gneiss temperature of 200øC was assumed, as this and the pegmatite. The pegmatite and the temperature is within the uncertainty of leucogranite contain chlorite, and closure temperatures calculated for other muscovite is sericitized in the augen K-feldspars of similar composition from gneiss, the pegmatite, and the the study area. leucogranite.
ANALYTICAL TECHNIQUES
Mineral separates were obtained using 1Supplementaldata are available with standard magnetic and heavy liquid entire article on microfiche. Order separation techniques. Sample from American Geophysical Union, 2000 preparation, irradiation, and analytical Florida Avenue, N.W., Washington, DC procedures are the same as those described 20009. Document T89-003; $2.50. Payment by Harrison and Fitz Gerald [1986]. must accompany order. 872 Hubbardand Harrison' 40Ar/39ArAges, Nepal Himalaya
RESULTS (Figure 5) reveals clear age discordance and does not define a plateau. A 90-100 MCT zone Ma total gas age predates continental collision and is therefore unlikely. This Muscovite, biotite, and K-feldspar were age is also inconsistent with the younger analyzed from sample 87H13F, and biotite ages from sample 87H13F, which was was analyzed from the adjacent sample collected just 5 m away. Excess argon 87H13E. Release spectra from both samples must, again, be the cause of this are shown in Figure 5. Biotite and unrealistically old age. The relatively muscovite from 87H13F yield plateaux at flat but discordant age spectrum -40 Ma and-12 Ma, respectively. As underscores the cautions recommended by biotite typically has a lower closure Gaber et al. [1988] regarding age spectra temperature than muscovite, this reversal interpretation from biotite. suggests the presence of excess argon. The biotite from sample 87H12C yields a Excess argon is a relatively common plateau age of -60 Ma, which is problem in biotites and, in certain cases, considerably older than the hornblende age is not detectable with the isochron (Figure 6). The hornblende data are approach because of the high Ar solubility complicated by the presence of two, in biotite and high 40Ar/36Arratio of the distinct, trapped argon components trapped gas. Muscovite, however, [Heizler and Harrison, 1988], but the generally appears less susceptible to the sample yields an isochron age of 20.9!_0.3 effects of excess argon, so we interpret Ma (Figure 6). This age is consistent the muscovite isochron age of 12.0+0.2 Ma with the muscovite and K-feldspar ages to be a meaningful cooling age. The K- from sample 87H13F collected at a slightly feldspar release spectrum is "saddle- lower structural level within the MCT shaped," which is also a sign of excess zone. The biotite does not yield any argon [Dalrymple et al., 1975; Harrison simple correlations. and McDougall, 1980 ] . The minimum age on A hornblende separate from sanple this spectrum at •8 Ma appears sensible 87H12B gives a complex release spectrum given the consistency with the muscovite (Figure 7) which could be interpreted as age. The closure temperature for this K- an excess Ar uptake profile superimposed feldspar is calculated to be Tc=220+15øC. on an Ar loss profile from a Paleozoic age Biotite was the only mineral analyzed [e.g., Harrison and McDougall, 1980] or as in sample 87H13E. The release spectrum "saddle- shaped" due to excess Ar [e.g., Harrison and McDougall, 1981], in which case no chronological meaning could be
'•10 I ,i i inferred from the spectrum. We suggest that the former explanation is correct and 100 '------have performed laboratory heating 87H13E BIOTITE 90 experiments to test this conclusion. Preirradiated, coarse-grained, 80 aliquants of 87H12B and 87H12C hornblendes were hydrothermally treated for 416 hours at 800øC and 1 kbar. Following the heat [ 60 treatment, the 39Ar concentrations of both c 50 samples were determined relative to their 87H13F BIOTITE starting materials, allowing calculation = 40 ß .• of D/a 2, which is the ratio of Ar 5o diffusivity (D) to the square of the diffusion radius (a). We chose to use 39Ar 2O MUSCOVITE• rather than 40Ar* (*radiogenic) for this 10 , purpose, as the nonuniform 40At* 87H13F K-FELDSPAR 0 distribution implicit in Figure 7 would O0 seriously compromise assumptions required to calculate model diffusion coefficients Cumulctive% 39Ar releosed [Crank, 1975] . The measured losses for Fig. 5. Release spectra for mineral 87H12B and 87H12C of 6 and 18%, separates from samples 87H13E and 87H13F respectively, correspond to values of D/a2 from the lower MCT zone. of 2x10-10 and 2x10-9 cm2/s, respectively. Hubbardand Harrison' 40Ar/39ArAges, Nepal Himalaya 873
0.01
100 0.004 90 87H12C j 87H12C
80 Biotite 70 o. .- 1,
o 40
30 o.oo• 5.8 Ma 20 • i '
10t Hornblende[ 0 I •%. , ø0 ,00 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8
Cumulative• 39Ar released 39Ar/4øAr
Fig. 6. Release spectra for biotite and hornblende from sample 87H12C (MCT zone), shown together with the K/Ca plot and isochron plot for the hornblende. Two trapped components yield isochrons that indicate roughly the same age.
Similar values for 37Ar loss (7 and 15%, and 87H12C were collected from the same respectively) strongly suggest that the Ar outcrop, they must have experienced loss is from the hornblende rather than similar thermal histories but responded to some other phase. We note that the Ar loss quite differently. From Figure 7 measured 40Ar* loss for 87H12C was similar we can infer that the rising age gradient to the 39Ar loss due to the uniform in the latter portion of gas release for distribution of both (Figure 6), but that 87H12B reflects 50% Ar loss from an 40Ar* loss in 87H12Bwas appreciably original age of -450 Ma (suggested to us higher, possibly due to the preferential by the data of Copeland et al. [1988b]), degassing of excess 40Ar held in near whereas sample 87H12C was virtually boundary sites. completely degassed (say >95%). Figure 8 The contrast in D/a 2 of an order of shows the relationship between fractional magnitude is likely a consequence of Ar loss and KDt/a2 (wheret is time) for a variety of geometries. Viewing the differing diffusion radii between the two samples. Baldwin and Harrison [ 1988 ] have hornblende crystals as spherical in shape (see Harrison [1981] and Baldwin and shown that metamorphic hornblende with variable Fe/Mg contents, when subjected to Harrison [1988] for an explanation), we hydrothermal treatment similar to that can see from Figure 8 that a combined utilized for 87H12B and 87H12C, yield an K2Dt/a2 parameterof 0.3 wouldeffect the activation energy for Ar transport of ~60 observed 50% Ar loss in 87H12B hornblende. kcal/mol but contain broadly varying Since we know that for an equivalent diffusion dimensions (controlled by temperature and heating duration 87H12C exsolution, defect structure, cleavage, hornblende would have a K2Dt/a2 value ten fractures, etc.) that strongly effect times higher (i.e., 3), from Figure 8 we overall Ar loss. Since our samples 87H12B infer an Ar loss in excess of 95%, similar 874 Hubbardand Harrison' 40Ar/39ArAges, Nepal Himalaya
1.0
0.9 Spher 0.8 Cube Cylinder • 0.7
•o 0.6
0.5
0.01 • 0.4 600 " •.5 !/ a = radiusofa sphere,and 0.2 / cylinder, half thickness 87H12B hornblende 500 i i I i I 0.1 ofslob. half edge ofcube 0 0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 •00 •2 Dr/a2 Fig. 8. Plot of fraction Ar loss versus 300 •2Dt/a2 for a variety of geometries.
200 Lower Tibetan Slab' Ghat
lOO K-feldspar from san/ole 87H21C gives a minimumage from the release spectrum (Figure 9) of 3.6+0.2 Ma (Tc = 210!_50), whereas biotite yields an isochron age of 9.1+0.2 Ma. Sample 87H21D is from the adjacent sheared pegmatite and yields a K- Cumulative:• 3eAr released feldspar minimum age of 6.4+0.8 Ma (Figure 10). Biotite and muscovite isochrons from Fig. 7. Release spectra and K/Ca plot 87H21D give ages of 7.5+0.6 Ma and 7.7+0.4 for hornblende from sample 87H12B (MCT Ma, respectively. The behavior of the zone). spectra for the K-feldspars from these two
2O to that observed. Thus our hypothesis that the latter portion of the 87H12B hornblende age spectrum represents incomplete degassing rather than excess Ar incorporation appears justified on the basis that (1) the form of the spectrum differs from our commonexpectation of excess Ar (see McDougall and Harrison o lO [1988] for review), (2) the degree of 40Ar loss during laboratory heating suggests excess40Ar in a nearboundary siting, (3)
the flat K/Ca plot (Figure 7) is not 5 indicative of the exsolved structures
needed to yield "saddle-shaped" spectra, -- but if, however, there are fine-grained intergrowths containing excess Ar, then o recoil homogenization may tend to damp 0 10 20 30 40 50 60 70 80 90 100 this effect, and (4) the relative 39Ar losses from adjacent san/oles (87H12B and CumulativeX; 39 Ar released 87H12C) suggest a higher Ar retentivity Fig. 9. Release spectra for biotite and for 87H12B. K-feldspar from sample 87H21C (Ghat). Hubbardand Harrison- 40Ar/39ArAges, Nepal Himalaya 875
20 ! ß , , I i i ,
50
K-FSIDSPA• 15 40
'-' 30 10 BI OTI TE
!
MUSC¸VITE 87H19B BIOTITE
lO
, 0 16 2•3 3•) 4•3 5•) 6•3 7•) 8•) 9•)100
Cumulative:• 3eAr released 11a Cumulative• 3gAr released Fig. 10. Release spectra for muscovite, biotite, and K-feldspar from sample 87H21D (Ghat). 1
san•les are markedly different. Almost o.• 50%of the 39Ar released is youngerthan 5 , Ma for sample 87H21C, with the remaining gas fraction younger than 15 Ma. For sample 87H21D, less than 50% of the 39At released from the K-feldspar is younger 0.01 , than 15 Ma, with the remaining gas 60 - i I i I I i i , i i fractions older than 20 Ma. San•le 87H21D appears to have been affected by excess 55 argon. õ0 .•------•r•• Upper Tibetan Slab' Na
Release spectra for hornblende from san•le 87H19A and biotite from san•le • 35 87H19B are shown together in Figure 11a. o 30 The age spectum for the hornblende • 25 suggests marginal excess argon uptake with • 20 87H19a Hornblende a minimum age on the spectrum of 22.7+4.4 1 st run Ma. A biotite isochron age for the second 15 2nd run through ninth increments gives an age of 10 20.2+0.2 Ma (mean square of weighted deviates (MSWD)=16, trapped 5 40Ar/36Ar=-299!_10). A secondanalysis of 0 the hornblende (Figure llb) reveals a o similar age spectrum to the first run but 11b Cumulotive% 39 Ar releosed in detail is more diagnostic of excess argon in the high-temperature steps. Fig. 11. (a) Release spectra for biotite Upper Tibetan Slab' Ngozumba from sample 87H19B and hornblende first run from sample 87H19A (Na), (b) Release Biotite and K-feldspar were analyzed spectra for hornblende second run, from sample 87H18A, an augen gneiss in the 87H19A. upper Tibetan Slab (Figure 12). The 876 Hubbardand Harrison- 40Ar/39ArAges, Nepal Himalaya
feldspar age spectrum is irregular and yields a minimum age (21.7ñ0.8 Ma) which is older than the mica ages from this sample. The biotite and muscovite isochron ages are nearly indistinguishable at 16.8ñ1.4 Ma (steps 1-11, MSWD=33.7, trapped 40Ar/36Ar=-308_+3)and 17.0ñ0.4 Ma (steps 2-10, MSWD=4.2, trapped o 20 40Ar/36Ar=-285•18),respectively. The biotite release spectrum is somewhat irregular, and there is scatter about the isochron as indicated by the high MSWD.
10 K-feldspar and muscovite release spectra for the leucogranitic sample 87H18C are shown in Figure 14. Again the K-feldspar spectrum is irregular, but in this sample the minimum age of 15.5ñ1.8 Ma is younger 0 0 10 20 30 40 50 60 70 80 90 100 than the muscovite isochron age of 16.6ñ0.4 (steps 1-11, MSWD=2.1, Curnulotive•. 3gAr releosed 40Ar/36Ar=-304_+4).
Fig. 12. Release spectra for biotite and TECTONIC IMPLICATIONS K-feldspar from sample 87H18A (Ngozumba). A summary of the mineral ages from this study are given in Table 2. As discussed release spectrum for K-feldspar is in the previous section, some of these irregular and, again, suggestive of excess argon. The minimum age on this spectrum ages may reflect excess argon and are is 18.7+0.8 Ma, which is older than the therefore less likely to be actual cooling biotite isochron age of 17.2+0.8 Ma (steps ages. From the subset of data considered 2-10, MSWD=5,trapped 40Ar/36Ar=-301ñ11). meaningful a representative age for each mineral, from each site, was plotted Cooling ages were obtained for versus mineral closure temperature (Figure muscovite, biotite, and feldspar from the 15). Taken together, these ages should pegmatñre sample 87H18B from Ngozumba reveal a consistent pattern from which (Figure 13). As in sample 87H18A, the tectonic rates and timing can be inferred.
4o i i i I i i i - $7H18B 87H18C
3o 3o _ •]• K-FELDSPAR•' PAR •1 -- BIOTITE • •0 ':' 20
MUSCOVITE U a. MUSCOVITE
10
, , , , , , , , 0 10 20 30 40 50 60 70 80 90 100 0 16 2•) 3• 4•) 5•) 6•) 7• 8•) 9•)100
Cumut•lve %39 Ar reteased Curnulotive:• 3ilAr releosed Fig. 13. Release spectra for muscovite, Fig. 14. Release spectra for muscovite biotite, and K-feldspar for ssmple 87H18B and K-feldspar for sample 87H18C (Ngozumba) . (Ngozumba) . Hubbard and Harrison: 40Ar/39ArAges, Nepal Himalaya 877
TABLE2. Summaryof 40Ar/39ArAges
Sample Rock Typ• Mineral Age Closure Temperature MCT zone 87H13F* augen gneiss K-feldspar 8.0+0.2 Ma (M) 220+15 øC 87H13F augen gneiss biotite 36.3+0.4 Ma (M) 300+50 øC 87H13F* augen gneiss muscovite 12.0+0.2 Ha (I) 350+50 øC 87H13E biotite gneiss biotite 88.8+0.6 Ha (M) 300+50 øC 87H12C amphibolite biotite 58.3+3.4 Ma (I) 300+50 øC 87H12C* amphibolite hornblende 20.9ñ0.2 Ma (I) 500+50 øC 87H12B amphibolite hornblende 204.0+5.2 Ma (M) 500+50 ø C Ghat 87H21C* biotite gneiss K-feldspar 3.6+0.2 Ha (M) 210+50 øC 87H21C biotite gneiss biotite 9.1+0.2 Ha (I) 300+50 øC 87H21D pegmatite K-feldspar 6.4+0.8 Ha (H) 225+20 øC 87H21D* pegmat it e b iot it e 7.5+0.6 Ha (I) 300+50 øC 87H21D* pegmat it e muscovite 7.7+0.4 Ha (I) 350+50 øC Na 87H19B* biotite gneiss biotite 20.2+0.2 Ha (I) 300+50 øC 87H19A* amph•lite hornblende 22.7+4.4 Ha (M) 510+50 øC Ngozumbe 87H18A augen gneiss K-feldspar 18.7+0.8 Ma (M) 250(?) 87H18A augen gneiss biotite 17.2+0.8 Ma (I) 300ñ50øC 87H18B pegmatite K-feldspar 21.7ñ0.8 Ma (M) 250(?) 87H18B* pegmatite biotite 16.8+1.4 Ma (I) 300ñ50øC 87H18B pegmatite muscovite 17.0ñ0.4 Ma (I) 350+50ø C 87H18C* leucogranite K-feldspar 15.5+1.8 Ma (M) 250 (?) 87H180' lcucegr•nite muscovite 16.6ñ0.4 Ma {I} 350ñ50ø C (I) is age determined from isochron and (M) is age from minimum on release spectra. * Data plotted in Figure 15.
If our interpretation of the age age of our less retentive hornblende spectrum for 87H12B hornblende is correct (87H12C) of 20.9+0.2 Mm, with a closure (see results section), we can use our temperature that falls within the range of knowledge of the diffusion properties and metamorphic temperatures, likely reflects metamorphic temperatures to constrain the the age of thrust movement at the duration of heating. Our Ar loss culmination of this metamorphism. experiments imply a frequency- In the upper Tibetan Slab, similar ages factor/diffusion-dimension parameter (15-20 Ma) to that of the MCT were (Do/a2) of ~1500, assumingan activation recorded at Na and Ngozumba in the upper energy of 65 kcal/mol [Harrison, 1981]. Tibetan Slab by biotite, muscovite, and K- Metamorphic temperatures for this portion feldspar separates. These ages are of the MCT (Figure la) are in the range consistent with 17-19 Ma K-Ar biotite ages ~5000-550 øc. An estimate of the maximum [Krummenacher et al., 1978] and 16-17 Ma duration of heating can be made from Rb/Sr whole rock mineral isochrons Figure 8 which shows that -50% Ar loss [Ferrara et al., 1983] for granite samples requires a K2Dt/a2 value of ~0.3. Our from the Everest area in Nepal but are diffusion parameters lead us to calculate somewhat younger than the U-Pb monazite a (D/a2) 500øCof 5x10-15 cm2/s, which and zircon ages of 20ñ1 Ma obtained by translates to a time at peak temperature Copeland etal. [1988b] for a granite of -0.25 Ma, although a parabolic sample from just north of Mount Everest. temperature history could yield a duration Although field evidence suggests that the of ~2.50 Ma. Shorter durations result for leucogranites in the Everest region higher temperatures. In the lower MCT represent multiple generations, the zone, peak metamorphic temperatures are collective geochronologic data suggest interpreted as synchrous with movement on that MCT metamorphism and deformation was the thrust [Hubbard, 1989]. The isochron generally coeval with leucogranitic 878 Hubbardand Harrison- 40Ar/39ArAges, Nepal Himalaya
600
500
LU 400 rr
•' 200
IO0
0 5 10 15 2 2 30
TIME (Ma)
Fig. 15. Plot of age versus closure temperature for mineral separates from the MCT zone, Ghat, Na, and Ngozumba. Cross section provides structural reference for sample locations. Sample number and source of age (isochron or minimum from release spectra) are given in Table 2. intrusion in the upper Tibetan Slab. mica ages at Ghat of -8 Ma are from an The samples from Na were collected some intermediate position between the MCT zone distance (>200 m) from any observable to the south, with a muscovite age (deemed intrusive rocks and permit us to examine the most reliable) of 12.0+0.2 Ma, and the the thermal history of the upper Tibetan Na/Ngozumba region in the north, with-17 Slab away from the direct influence of the Ma ages. This age progression reversal leucogranites. Unfortunately, the suggests either (1) eraplacement of this hornblende age spectrum (Figure 11a) is slab from depth along unrecognized complex and duplication (Figure 11b) structures since 8 Ma or (2) localized strongly suggests contamination by excess deformation between -12 and -8 Ma Ar. No age significance should be placed accompanied by hot fluid circulation on the -40-55 Ma ages. The 20.2+0.2 Ma sufficient to reset the mica ages. biotite from Na (Figure 11a) is slightly Whichever the case, this deformation older than the -17 Ma mica ages from indicates that there has been localized Ngozumba, but the significance of this age shear deformation within the Tibetan Slab difference is difficult to assess in view and that this deformation significantly of the problematic nature of our other postdates movement of the MCT. biotite age spectra from this study. Biotite and muscovite from a CONCLUSIONS mylonitized pegmatite at Ghat record -8 Ma ages, which constrains the age of In the Everest region of eastern Nepal deformation to be not younger than -8 Ma. a 40Ar/39Ar hornblende age from the MCT Cooling by uplift of the inverted zone suggests thrust movement and metamorphic sequence would produce mineral metamorphism at -21 Ma. Mica ages from a ages which would become younger leucogranite and adjacent gneiss within monotonically from north to south. The the Tibetan Slab suggest that granitic four mica ages from Ngozumba yield an age intrusion was contemporaneous with MCT of 16.9+0.3 Ma, similar to mica ages from thrusting and metamorphism. These data the leucogranite belt farther west are consistent with the LeFort [1981] [Copeland et al., 1988a]. However, the model of granite generation associated Hubbardand Harrison' 40Ar/39ArAges, Nepal Himalaya 879 with movement on the MCT. Peak inherited radiogenic Pb in monazite and temperatures were probably maintained for its implications for U-Pb systematics, less than-2 Ma. Nature, 333, 760-763, 1988b. A biotite from the upper Tibetan Slab Crank, J., The Mathematics of Diffusion, away from intrusives is -20 Ma, only 414 pp., Oxford University Press, New slightly older than biotite ages from the York, 1975. leucogranites and adjacent gneisses and Dalrymple, G.B., C.S. Gromme, and R.W. reflects regional heating associated with White, Potassium-Argon age and leucogranitic intrusion. paleomagnetism of diabase dikes in The -8 Ma mica ages from a shear zone Liberia' Initiation of central Atlantic in the lower Tibetan Slab indicate rifting, Geol. Soc. Am. Bull., 86, 399- relatively recent deformation in the 411, 1975. Higher Himalaya, which may have also been Deniel, C., P. Vidal, A. Fernandez, P. Le important in shaping the tectonic history Fort, and J. Peucat, Isotopic study of of this area. the Manaslu granite (Himalaya, Nepal)- Inferences on the age and source of Acknowledgments. We wish to thank Kip Himalayan leucogranites, Contrib. Hodges, Peter Molnar, Leigh Royden, Clark Mineral. Petrol.: 96, 78-92, 1987. Burchfiel, and Gary Johnson for useful Ferrara, G., B. Lombardo, and S. Tonarini, discussions and comments on an early Rb/Sr geochronology of granites and version of this manuscript. R.J. Fleck, gneisses from the Mount Everest region, T.C. Onstott, B.M. Page, and an anonymous Nepal Himalaya, Geol. Rundsch., 72(1), reviewer are gratefully acknowledged for 119-136, 1983. helpful reviews. Thanks go to Sange Dorje Gaber, L.J., K.A. Foland, and C.E. Sherpa and Mountain Travel Nepal for Corbato, On the significance of argon logistical support during sample release from biotite and amphibole collection and Matt Heizler for technical assistance. Matt, Lynn, and Ashley during 40Ar/39Arvacuum heating, Heizler are thanked for their endless Geochim. Cosmochim. Acta, 32, 2457-2465, hospitality during data collection. This 1988. research was supported by the American Gansser, A., Geology_ of the Himalayas, 289 Alpine Club, The Explorer' s Club, The pp., Wiley-Interscience, New York, 1964. Geological Society of America, and NSF Gansser, A., The geodynamic history of the grants EAR-8414417 and EAR-8707569. Himalaya, in Zagros: Hindu Kush. Himalaya Geodynamic Evolution, edited by REFERENCES H.K. Gupta and F.M. Delany, Geodyn. Ser., vol. 3, pp. 111-121, AGU, Baldwin, S.L., and T. M. Harrison, Washington D.C., 1981. Diffusion of 40Ar in metamorphic Harrison, T.M., Diffusion of 40Ar in hornblende, Eos Trans. AGU, 69, 522, hornblende, Contrib. Mineral. Petrol., 1988. 7__8,324-331, 1981. Brunel, M., and J.R. Kienast, •.tude p6tro- structurale des chevauchements ductile Harrison, T.M., and J.D. Fitz Gerald, Exsolution in hornblende and its himalayens sur la transversale de L'Everest-Makalu (N6pal oriental), Can. consequencesfor 40Ar/39Arage spectra J. Earth Sci.; 23, 1117-1137, 1986. and closure temperature, Geochim. Copeland, P., T.M. Harrison, R. Parrish, Cosmochim. Acta. 50: 247-253, 1986. B.C. Burchfiel, and K.V. Hodges, Harrison, T.M., and I. McDougall, Constraints on the age of normal Investigations of an intrusive contact, faulting, north face of Mt. Everest' northwest Nelson, N.Z., II, Diffusion of Implications for Oligo-Miocene uplift radiogenicand excess40Ar in hornblende. (abstract), Eos Trans. AGU 68(44), revealedby 40Ar/39Arage spectrum 1444, 1987. analysis, Geochim. Cosmochim. Acta, 44, Copeland, P., T.M Harrison, and P. LeFort, 2005-2020, 1980. Cooling history of the Manaslu granite, Harrison, T.M., and I. McDougall, Excess north-central Nepal (abstract), Geol. argon in metamorphic rocks from the Soc. Am. Abstr. Programs, 20, 1988a. Broken Hill region, New South Wales- Copeland, P., R. Parrish, and Implications for 40Ar/39Arage spectra T.M. Harrison, Identification of and the thermal history of the region, 88O Hubbardand Harrison- 40Ar/39ArAges, Nepal Himalaya
Earth Planet. Sci. Lett., 55, 123-149, J. Geoph_vs. Res,, 86, Bll, 10,545- 1981. 10,568, 1981. Heizler, M., and T.M. Harrison, Multiple Le Fort, P., M. Ctuney, C. Deniel, C. trapped argon isotope components France-Lanord, S.M.F. Sheppard, B.N. revealedby 40Ar/39Arisochron analysis, Upreti, and P. Vidal, Crystal generation Geochim. Cosmochim. Acta, 52, 1295-1303, of the Himalayan leucogranites, 1988. Tectonophysics, 134, 39-57, 1987. Hodges, K.V., M.S. Hubbard, and D.S. Maruo, Y., and K. Kizaki, Thermal Silverberg, Metamorphic constraints on structure in the nappes of the eastern the thermal evolution of the central Nepal Himalayas, in Granites of Himalayan orogen, Philos. Trans. R. Soc. .Himalaya, Karakorum and Hindu-Kush, London, Ser A,326, 257-280, 1988. edited by F.A. Shams, pp. 271-286, Hubbard,M., Thermobarometry,40Ar/39Ar Punjab University, Lahore, Pakistan, geochronology, and structure of the Main 1983. Central Thrust zone and Tibetan Slab, McDougall, I., and T.M. Harrison, Eastern Nepal Himalaya, Ph.D. thesis, Geochronology and thermochronology by 169 pp., Mass. Inst. of Technol., Mass., the 40-QAr/9•ArMethod, 212 pp., Oxford 1988. University Press, New York, 1988. Hubbard, M., Thermobarometric constraints Molnar, P., Structure and tectonics of the on the thermal history of the Main Himalaya: Constraints and implications Central Thrust Zone and Tibetan Slab, of geophysical data, Annu. Rev. Earth Eastern Nepal, J. Metamorph. Geol.,. 7, Planet. Sci,, 12, 489-518, 1984. 19-30, 1989. Sch'•rer, U., The effect of initial 230Th Krummenacher, D., A.M. Bassett, F.A. disequilibrium an young U-Pb ages- The Kingery, and H.F. Layne, Metamorphism Makalu case, Himalaya, Earth Planet,. and K-Ar age determinations in eastern Sci. Lett.;_ 67, 191-204, 1984. Nepal, in Tectonic Geology_ of the Sch'Rrer, U., R.H. Xu, and C.J. All•gre, Himalaya, edited by P.S. Saklani, pp. U-(Th)-Pb systematics and ages of 151-166, Today and Tomorrow's Printing. Himalayan leucogranites, South Tibet, and publication, New Delhi, India, 1978. Earth Planet. Sci. Lett., 77, 35-48, La Tour, T.E., Geochemical model for the 1986. symplectic formation of myrmekite during amphibolite-grade progressive mylonitization of granite, Geol.. Soc. T. M. Harrison, Department of Am Abstr. Programst 19(7), 741, 1987. Geological Sciences, State University of Le Fort, P., Himalayas- The collided New York at ALbany, Albany, NY 12222. range, Present knowledge of the M. S. Hubbard, Geologisches Institut, continental arc, Am. J. Sci. 275-A• 1- ETH Zentrum, CH-8092, Z•rich, Switzerland. 44, 1975. Le Fort, P., Manaslu leucogranite' A (Received October 28, 1988; collision signature of the Himalaya, A revised March 9, 1989; model for its genesis and eraplacement, accepted April 5, 1989. )