| dc.description.transcription | PROCEEDINGS OF THE THIRTEENTH LUNAR AND PLANETARY SCIENCE CONFERENCE, PART 2 JOURNAL OF GEOPHYSICAL RESEARCH, VOL. 88, SUPPLEMENT, PAGES A741-A754, FEBRUARY 15, 1983 NATURE OF THE H CHONDRITE PARENT BODY REGOLITH: EVIDENCE FROM THE DIMMITT BRECCIA Alan E. Rubin, Edward R. D. Scott, G. Jeffrey Taylor, Klaus Keil and Jaclyn S. B. Allen Department of Geology and Institute of Meteoritics University of New Mexico, Albuquerque, New Mexico 87131 T. K. Mayeda and R. N. Clayton Enrico Fermi Institute, University of Chicago, Chicago, Illinois 60637 D. D. Bogard Geochemistry Branch, Johnson Space Center, Houston, Texas 77058 Abstract. The Dimmitt H chondrite regolith breccia consists of (in vol.%) 40% H4 and H5 chondrite clasts, 3% impact melt rock clasts, 0.5% shocked H chondrite clasts, 1.5% exotic clasts (including carbonaceous and LLS chondrites), and 55% gas-rich matrix.. The LLS clast is the best documented example of an ordinary chondrite in a host of a different compositional group. The matrix contains unequilibrated material, which differs from typical H3 material in having little ( 0.2 vol.%) fine-grained opaque silicate matrix, and having 20% of the olivines with compositions in the range Fa 21-24. About 10-15% of this unequilibrated material is probably derived from graphite-magnetite-rich chondrites and 2% from H3.0-3.5 chondrites. The absence of H3 clasts suggests that most of the unequilibrated material was derived from unconsolidated type 3 components. Many exotic clasts may have been derived from planetesimals that accreted to the chondrite parent body prior to regolith develop-ment. One slowly cooled melt rock clast formed beneath a 500-m-thick melt breccia pile on the floor of a large impact crater and was later excavated by additional impacts, incorporated into the regolith and consolidated with other components to form the Dimmitt breccia Introduction Meteorite regolith breccias are clastic rocks that formed by lithification of fragmental regolith material that once resided at the surface of a meteorite parent body. These breccias are characterized by solar wind-im-planted rare gases [Suess et al., 1964], solar flare particle tracks [Pellas et al. 1969], impact-melted clasts of regolith material [e.g., Fodor and Keil, 1973], exotic clasts, Eodor. and oKe hound 9 Unbrecolated meteorites of the same chemical group as the predominant breccia lithology (e.g., carbonaceous chondrite Now at Department of Mineral Sciences, National Museum of Natural History, Smithsonian Institution, Washington, D.C. 20560 Copyright 1983 by the American Geophysical Union. Paper number 2B1421. 0148-0227/83/002B-1421$05.00 clasts in ordinary chondrite breccias; Wasson and Wetheril1, 1979), and a light/dark structure [e.g., Konig et al., 1961]. Because lunar regolith breccias contain the same constituents as the unconsolidated lunar regolith [e.g., Taylor, 1982], meteorite regolith breccias are probably representative samples of the parent body regoliths from which they were derived. Most previous petrologic studies of H chondrite regolith breccias focused on the large prominent clasts [e.g., Fodor and Keil, 1976; Fodor et al., 1976; Keil and Fodor, 1980; Keil et al., 19801. Here we report a study of the matrix and 21 clasts of various sizes (0.2-24 mm) in the Dimmitt H chondrite regolith breccia using petrographic and electron microprobe techniques. In addition, oxygen isotope studies of three clasts (DTI, DT3, and DI4) and instrumental neutron activation analysis (INAA) and 39 Ar/40 Ar age dating of one clast (DT4) are reported. The Dimmitt meteorite (consisting of at least 21 stones with a total weight of 13.5 kg) was found about 1942 near Dimmitt, Texas [Hey, 1966; Hutchison et al., 1977]. Rare gas studies of the dark portions of Dimmitt indicate enrichment in solar-type rare gases [Eberhardt et al., 1966]. Analytical Procedures Polished and unpolished slabs of Dimmitt (Table 1) (total area 300 cm?) were examined visually; polished slabs were also viewed microscopically in reflected light. Polished thin sections of clasts and areas of the clastic matrix were examined microscopically, in transmitted and reflected light. During these studies, many smaller clasts ( < 1 mm) were identified and also studied. Metallic Fe, Ni grains were etched with a dilute solution of nitric acid in alcohol to bring out structural details. Modal analyses were made in transmitted and reflected light on a Zeiss Universal petro-graphic microscope. All clasts (Table 1) and areas of the clastic matrix were analyzed by an electron microprobe; mineral analyses were made using crystal spectrometers, following standard Bence-Albee and ZAF correction procedures [e.g., Fodor and Keil, 1976; Kel, 1967]. Analyses of Fe, Ni for Co were corrected for Co interference from the KB peak of Fe. Major, minor, and trace element contents of DI4 (the largest clast) were determined by M.-S. Ma and R. A. Schmitt (Oregon State University) by TABLE 1. Descriptions and Locations of Class in the Dimmitt H Chondrite Regolith Breccia Clast Area of Dimmitt* Section Clast Type Clast Diameter (mm)?USNM = United States National Museum;?UNM = University of New Mexico +G-M = graphite-magnetite INAA using standard procedures [Wakita et al., 1970; Laul and Schmitt, 1973, 1974]. Broad beam (100 um) electron microprobe analyses of clasts and melt pockets were made to determine bulk comesitigns. Ar- 40 Ar measurements were made on a 0.201 g chip of DI4. This sample and NI-25-2 hornblende fluence monitors were irradiated wigh a nominal fast neutron fluence of 2 x 10 DT4 was subsequently heated for 45 minutes at each of a series of increasing temperature steps, and the isotopic composition of the extracted argon was measured on a mass spectrometer. Temperature was monitored with a thermocouple embedded in the crucible. Argon from the hornblende monitors gave Ar/40 Ar ratios of 0.0312 and 0.0303, which were used to calculate ages from the Ar data for DI4. Details of the irradiation and gas extraction techniques, the hornblende monitor and the age calculations are given by Husain [1974] and Bogard et al. [1976]. Oxygen isotopic analyses of a 10.0 mg sample of clast DT3, two chips of clast DTI (3.6 mg and 5.3 mg), and two chips of clast DI4 (4.7 mg and 2.8 mg) were carried out following previously described procedures (Clayton et al., 1976]. Results Dimmitt consists (in vol.%) of 40+10% light-colored H4 and H5 chondritic clasts, 30.5% impact melt rock clasts, <0.5% shocked H chondrite clasts, 1.5‡0.5% exotic clasts of non-H chondrite parentage, and 55:10% fine-grained ( <2 mm) dark-colored clastic matrix (with 1% interstitial melt; A. Bischoff, personal communication, 1982), in which all clasts are embedded. No agglutinates were found. Clasts and the clastic matrix have been shocked to the same degree: fractured silicates with undulose extinction, melt pockets and "fizzed' troilites are common throughout the meteorite. (Shock melting of troilite produces a fine-grained fizz of irregularly-shaped grains of metallic Fe,Ni in finely crystalline troilite; Scott, 1982. Such shock may also have caused the annealing of solar flare tracks in olivine observed by Martinek [1981]. Melt rock clast DI4 appears to have been degassed together with whole rock Dimmitt (see below). These features suggest that most of the shock occurred during or after lithification of the breccia. Clastic Matrix The dark-colored matrix is clastic (i.e., consists of angular mineral fragments) except for the finest-grained fraction (grain size <15 um). This portion of the matrix was shock-melted, a process which caused lithification of the loose regolith material, forming the tough Dimmitt breccia (A. Bischoff, personal communication, 1982). The matrix contains 40 vol.% chondrules, 0.02-3.0 mm in apparent diameter, of all the textural types found in ordinary chondrites. Many of these chondrules are well defined and contain minor to abundant clear to turbid glass. Approximately 5% of the chondrules have thin rims of fine-grained opaque silicate matrix ('Huss matrix'; Huss et al., 1981) and 0.5% have rims of translucent, glassy-looking, Huss matrix [Scott et al., 1982]. Forty microchondrules (those chondrules <100 um in diameter) of all textural types were identified [Rubin et al., 1982a]. (Dimmitt is the only regolith breccia that we have found to have a micro- Fig. 1. Histograms of the mole% fayalite and ferrosilite concentrations in olivine and low-Ca pyroxene in different areas of the Dimmitt clastic matrix and for the entire clastic matrix. In each of the areas A-D, 30 olivines and 30 low-Ca pyroxenes were measured. The histograms indicate that different areas of the Dimmitt clastic matrix are generally similar in composition. For comparison, compositional ranges in type 4-6 ordinary chondrites are shown (revised from Gomes and Keil, 1980). Area A is from thin section UNM 523, B from UNM 547, C from USM 1595, and D from UNM 489. chondrule-bearing clastic matrix.) Most chon-drules in the matrix are surrounded by angular mineral fragments and none appears to be part of H3 chondrite clasts. Olivine and low-Ca pyroxene compositional patterns show peaks in the ranges characteristic of H4-6 chondrites superimposed on an unequil-ibrated background (Figure 1). However, low-Ca pyroxenes ArO more heterogeneous and more magnesian than olivines, as are those of the more equilibrated H3 chondrites (e.g., Dhajala; Noonan et al., 1976). Analyses of olivines and low-Ca pyroxenes in the matrix of four polished thin sections (areas A-D in Figure 1) indicate that low-Ca pyroxenes have very similar compositional patterns. Although olivines are broadly similar in the four areas, the proportion having compositions of 40% in area ?? 77-24 ??? 1000 FRO eNT SE varies from 17% in area A to variation [(oFa/mean Fa) × 100] and percent mean deviation (IMD; calculated from wt.% Fe) of 120 random olivines are 26 and 15, respectively; the corresponding numbers for 120 random low-Ca pyroxenes are 43 and 32. Olivine and low-Ca pyroxene compositional patterns of glassy chondrules in the matrix are rather similar to random matrix analyses. Olivine in the glassy chondrules shows a prominent peak at Fa 9, and 20% have compositions in the range However, the 47 analyzed glassy chondrite olivines are more heterogeneous, having coefficient of variation of 40 and a %MD of 27. Although the clastic matrix contains compositionally-zoned taenite with normal, M-shaped Ni profiles, taenite analyses scatter on a composition-dimension plot (Figure 2), suggesting that different grains cooled at rates of 0.2-300°C/m.y. (through 500°C). Kamacite is homogeneous with an average of 0.48 wt.% CO. typical of H4-6 chondrites [Afiattalab and Wasson, 1980]. Equilibrated Clasts DTS (H4) and DT6 (H5) chondrite clasts (Table 1) are petrographically indistinguishable from average H4 and H5 chondrites. DT5 contains olivine, low-Ca pyroxene (Table 2) and kamacite (0.44 wt.% Co) with compositions within the ranges characteristic of H4-6 chondrites (Fa 16.9-20.4; Fs 15.7-18.1; 0.33-0.48 wt.% Co; Gomes and Keil, 1980; Afiattalab and Wasson, 1980). Distinct chondrules and obvious chondrule fragments, 0.16-1.6 mm in apparent diameter, constitute 34 vol.% of DT5 and appear somewhat recrystallized. They do not contain glass. DT6 contains olivine, low-Ca (mostly orthorhombic) pyroxene (Table 2) and kamacite (0.45 wt.% Co), also with compositions in the ranges characteristic of H4-6 chondrites. Chondrules (0.25-3.3 mm in apparent diameter) constitute 27 vol.% of DT6. They are readily delineated, but some have extensively recrystallized boundaries and none have glass. Both DIS and DT6 have coherent metallographic cooling rates (Figure 3) of 150°C/m.y. (using the cooling rate curves of Willis and Goldstein, 1981). Evidently, the post-lithification shock experienced by Dimmitt did not affect the central Ni contents of the taenites. Fig. 2. Central Ni concentrations of taenite grains in the Dimmitt clastic matrix plotted against apparent distance from edge of grain. Data do not lie parallel to the cooling rate curves of Willis and Goldstein [1981] and indicate cooling rates of 0.2-300°C/m.y. at different depths before consolidation of the breccala. Table 2. Mineralogy of Dimmitt Clasts and Clastic Matrix M = major phase; m = minor phases; a = accesory phase; hm = homogeneous; ht= heterogeneous; k = kamacite; t = taenite; Tt - tetrataenite; Crm = chromite; Apt = chlorapatite; Wht = whitlockite. * Also contains accessory plessite and schreibersite. +Also contains major graphite-magnetite. Poikilitic Melt-Rock Clast Petrography, , Mineralogy, Composition and Age. DT4, the largest melt-rock clast (Figure 4), is light in color relative to the clastic matrix. It has a prominent polkilitic texture (olivine chadacrysts enclosed by orthopyroxene oiko-crysts), very similar to that of impact-melted clasts in many H, L, and LL ordinary chondrites [e.g., Keil et al., 1980; Rubin et al., 1981; Fodor and Keil, 1975]. DT4 contains olivine, low-Ca pyroxene and kamacite (0.46 wt.% Co) with compositions in the ranges characteristic of H4-6 chondrites (Table 2). DI4 is significantly depleted in metallic Fe, Ni and troilite relative to the clastic matrix; it contains 1.1% normative metallic Fe, Ni (kamacite and taenite) and 0.05% normative troilite. Most of the metallic Fe,Ni occurs in a large (1.4 mm diameter) round assemblage of polycrystalline kamacite (crystals 8-260 jam in size, average, 38 um; 7.0 wt.% Ni) with 5-20 um-wide areas of plessite rimmed by taenite (24- 31 wt.% Ni), and numerous 2-20 um blebs of high-Ni schreibersite (41 wt.% Ni). Martensite and tetrataenite were not encountered. The silicate-normalized bulk composition of clast DI4 (Table 3), determined by broad beam electron probe analysis, is very similar to that of average H chondrites (Keil, 1969], except that the clast is depleted in the volatiles Na,0 and rare earth element (REE) côncenErozions he Diate determined by INAA) are within one standard deviation of those of average chondrites, except for the lighter REE, which appear to be somewhat enriched in the clast (Table 3). The Ir/Au ratio of DT4 (2.3; Table 3) is similar to that of the poikilitic melt-rock clast in the Bovedy I-group chondrite (2.1; Rubin et al., 1981) and lower than that of average H chondrites (3.7; Rambaldi et al., 1979). DI4 has 4deen partiggly, radiogenic Ax , 40 egassed of its ages of the temperature fractions tend to inggase consistently to an age of 4.0 G.y. at an Ar fractional release of 0.95, then to decrease abruptly (Figure 5). The K/Ca ratigg remains nearly constant until the fractional Ar exceeds 0.9, then it also decreases abruptly (Figure 5). The averaged K-Ar 'age' of DI4 Is 2.70 G.y., in agreement with a Dimmitt whole rock age of 2.64 G.y. (Eberhardt et al., 1966], suggesting that DT4 was degassed together with whole rock Dimmitt: 18, isotopic analysis of DT4 yields and 8 values of +4.4°/00 and +3.1° 00, respectively. The clast thus lies close to the range of whole rock values for H4-6 chondrites (Table 4, Figure 6). Thermal history. An approximate cooling rate of DT4 can be calculated by modeling the growth of kamacite from taenite [Smith and Goldstein, 1977]. Assuming that all kamacite growth occurs at a single optimum temperature, an approximate cooling rate can be calculated by dividing the temperature interval over which kamacite growth Fig. 3. Central Ni concentrations of taenite grains in equilibrated clasts DT5 (H4) and DT6 (H5) plotted against apparent distance from edge of grain. Data lie parallel to the cooling rate curves of Willis and Goldstein [1981] and indicate a cooling rate of 150°C/m.y. Fig. 4. Photograph of Dimmitt slab (UNM C10.9) with light-colored poikilitic melt rock clast DI4. The round metallic Fe,Ni-schreibersite assemblage in the center cooled at 30°C/103 y (through 500°C). The clast must have been buried beneath an 500-m-thick melt breccia pile the floor of a several-kilometer-diameter impact crater on the H chondrite parent body. actually occurs by the time taken to grow the average-sized kamacite grain at this temperature. Following Smith and Goldstein [1977] and Taylor et al. [1979], we find that the optimum temperature for kamacite growth in DT4 is 660°C. At that temperature, the time required to grow 38 um-wide kamacite is 2.9×10 sec, based On diffusion coefficient that is most appropriate for the bulk composition (8.9 wt.% Ni; 0.37 wt.% P) of the assemblage [Heyward and Goldstein, 1973]. Kamacite nucleated at 710°C (correspon-ding to the bulk 8.9 wt.% Ni content of the assemblage) and precipitated on taenite borders as DT4 cooled. Using the Fe-Ni-P phase diagram [Romig and Goldstein, 1980], we determine that the approximate temperature at which kamacite growth ceased was 460°C (from the maximum (31 wt.%) Ni content of taenite bordering kamacite]. Thus, kamacite grew over a temperature interval of 250°C. We derive a 9x10 cooling rate of °C/sec or 30°C/103 y for DI4. The thermal event that formed the DT4 parent melt was an earlier more intense event than that which caused the partial loss of Ar in DT4 discussed above. Because the melting event probably totally degassed DI4 (or 158 parent material) of previously accumulated Ar, DT4 must have cooled sufficiently to retain Ar prior to 4.0 G.y. ago, which is the maximum age recorded in Figure 5. Other Melt-Rock Clasts The mineralogy of the other melt-rock clasts in Dimmitt is listed in Table 2. These clasts Include the following: A. Two clasts with skeletal olivine (DT14, DT21). DT14 has large elongated olivines (25 x 100 um) at one end and progressively smaller olivines grading into a microcrystalline mesostasis toward the other end, which appears to be a quenched margin. Olivines in DT21 are very elongated (2 × 50 um), but are too narrow to to obtain quantitative electron probe analyses. B. Four porphyritic clasts with olivine and Table 3. Bulk Composition of Selected Clasts, H, LL, CM Chondrites and CO-CV Matrix low-Ca pyroxene phenocrysts (DT12, DT13, DI15, DT16). These clasts appear somewhat similar to large porphyritic chondrules, but differ from them primarily in having nonspherical shapes. C. Seven clasts which contain relict, apparently unmelted silicate grains with partially resorbed margins and heterogeneous compositions (Table 2 set in a dark-colored microcrystalline groundmass (DT9, DI1O, DT11, DT17, DT18, DT19, DI20). In addition, numerous smaller such clasts, 0.05-0.2 mm in maximum dimension, were also observed. These debris-laden clasts are the most abundant type of melt rock clast in Dimmitt, constituting half of all of the melt rock clasts large enough for microscopic examination. DT11 contains a 200 um-diameter relict porphyritic pyroxene chondrule (Fs Woy) with partially resorbed margins. It also contains an isotropic glass shard (0.1 mm in maximum dimension) similar in composition to the glassy mesostases of some chondrules from type 3 ordinary chondrites (Gooding, 1979]; the shard thus may have been derived from a fragmented chondrule. Clasts similar to these seven occur in Plainview (e.g., PV5b and PV7b of Fodor and Keil, 1976), in the Adams County H5 chondrite [Fodor et al., 1980] and in the Vishnupur LL6 chondrite [Fodor and Keil, 1978]. They formed in the regolith by impact-melting accompanied by loss of metallic Fe, Ni and sulfide as an immiscible liquid and incorporation of unmelted regolith materials before quenching. Shocked H Chondrite Clast DT7 is an angular clast containing a very fine-grained, opaque matrix ( ~ 85 vol.%) consisting of silicates with finely dispersed, very abundant, small (<1 um) grains of metallic Fe, Ni and troilite. Subhedral to anhedral olivine, Fig. 5. Calculated 39 Ar /40 Ar ages and K/Ca jatios of clast DI4 as a function of fraction of released, from stepwise temperature extractions low-Ca pyroxene, and plagioclase grains (Table 2), 2-40 um in maximum dimension, constitute 15 vol.% of the clast. The dispersion of the metallic Fe,Ni and sulfide grains suggests that these phases were mobilized by shock. Olivine is heterogeneous, but has an average composition within the range of H4-6 chondrites, although low-Ca pyroxene (also heterogeneous) has AT average composition in the range characteristic of 14-6 chondrites (Table 2). Bulk atomic ratios of Ca/Si, Fe/Si and Mg/Si are in the ranges characteristic of ordinary chondrites [Wasson, 1974]. The texture of DT7 as well as its mineral and bulk compositions indicate that this clast was probably derived from the heterogeneous Dimmitt clastic matrix by shock. Clasts of Different Chondrite Groups Four chondritic clasts, not of H-group classification, were discovered; such clasts constitute 1% of the Dimmitt breccia. These include an LL5 clast (DT3), two clasts of a new kind of type 3 chondrite with a graphite-magnetite matrix (DTI, DI2), and a carbonaceous chondrite clast (DT8). LLS Chondrite Clast. The bulk chemical composition (determined by broad beam electron Fig. 6. Three isotope plot showing Dimmitt whole rock and clasts DTI (a type 3 chondrite with graphite-magnetite matrix), DT3 (LLS chondrite) and DT4 (poikilitic melt rock of H chondrite composition) compared to the 1 ranges for types 5-6 H and I chondrites. Dimmitt whole rock lies beyond the range of H5-6 chondrites; this may be due to terrestrial weathering or to the presence in Dimmitt of somewhat 0-enriched type material. probe analysis), olivine, low-Ca pyroxene and kamacite compositions, and modal abundance of metallic Fe,Ni all indicate that clast DT3 is an LL-group chondrite (Tables 3 and 4). However, its total He content and oxygen isotopic composition are slightly outside the IL ranges, but in both cases, they are even farther from the H and L chondrite ranges (Table 5; Figure 6). (The LL range is not shown in Figure 6 because only three LL5-6 chondrites have been measured.) DI3 contains barred olivine, porphyritic olivine and cryptocrystalline chondrules (0.4-2.0 mm in diameter) with recrystallized boundaries, set in a recrystallized silicate matrix. The recrystal-lized nature of the chondrules and silicate matrix, presence of augite, fairly homogeneous compositions of olivine (9MD=2) and low-Ca, pyroxene (%MD=5), and abundance of small (3-5 um) grains of plagioclase composition and stoichiometry (Table 2) indicate that the clast is of petrologic type 5. Shock veins, melt pocketS, and undulose extinction and extensive Table 4. Properties of DT3 (LL5 Clast) in Dimmitt Compared to Ordinary Chondrites Table 5. Oxygen Isotopic Compositions of Dimmitt Clasts fracturing of silicates indicate that the clast is of shock facies d [Dodd and Jarosewich, 1979]. We thus classify the clast as an LL5d chondrite. Chondrites Rich in Graphite-Magnetite. Clasts DTI and DT2 are of a new kind of type 3 chondrite with a graphite-magnetite matrix [Scott et al., 1981a,b]. They contain heterogeneous olivines and low-Ca pyroxenes (Table 2), olivines with 0.1 wt.% Ca0, well-defined glassy chondrules (100-500 um in apparent diameter), and bulk Al/Si, Mg/Si, Ca/Si, and Fe/Si atomic ratios in or very near the ordinary chondrite range. Two other clasts of this new kind of type 3 chondrite were found in the Plainview and Weston chondrite regolith breccias [Scott et al., 1981a,b]. The matrix of these clasts is composed of fine-grained graphite and magnetite, rather than the fine-grained opaque and recrystallized silicate matrix 'Huss matrix'; Huss et al., 1981] normally found in type 3 ordinary chondrites. Most of the graphite-magnetite has very little metallic Fe,Ni embedded in it Oxygen isotopic analysis of DT1 gave 18o=+3.40/00, 170= +2.50/00 (Table 5; Figure 6). This point lies on a mass-fractionation line through the data for H4-6 chondrites, but is displaced toward lower ’180 and § 'O values by 0:7 and 0.4 /o, respectively, from the H4-6 chondrite region. This displacement may be result of light Isotope enrichment in magnetite relative to olivine and pyroxene [Anderson et al., 1971]. Carbonaceous Chondrite Clast. DI8 is an angular clast that contains 80 vol.% fine-grained, predominantly opaque silicate matrix, 5 vol.% of 25 11m taenites and 10 um troilites, and 15 vol.% subhedral to anhedral olivine and pyroxene grains (1-140um). Some rounded aggregates of olivine and pyroxene may be porphyritic chondrule fragments. The abundance of large silicate grains distinguishes DT8 from clasts of fine-grained opaque silicate matrix which ace present in some type 3 ordinary chondrites. Analyses of 14 olivines (Fa 3-12; avg. Fa 8 O Fa = 3) and 10 low-Ca pyroxenes (Fs 0-38 avg. Fs 10; O Fs = 12) indicate that these minerals are more magnesian than those of ordinary chondrites and are similar to those of some carbonaceous chondrites, particularly the CV group [McSween, 1977]. However, in this clast, low-Ca pyroxenes are on average slightly more Fe-rich than olivines, unlike most CV chondrites [McSween, 1977]. The bulk composition of DT8 is more Fe-rich than bulk CO-CV chondrites [Keil, 1969], but is more similar to that of CO-CV opaque matrix (Table 3). DT8 has a high modal matrix abundance, like CM chondrites [MeSween, 1979], but is higher in Fe0 and S10, and lower in Cad and S than average CM chondrites. From the high total of the bulk analysis, it is apparent that DI8 is lower in H0 and C relative to CM chondrites, but not CO-CV matrix (Table 3). We conclude that DT8 is a fragment of some kind of carbonaceous chondrite, probably of petrologic type 3. However, the present data are insufficient to render a more specific identification. Graphite-Magnetite Aggregates Numerous 30 um to 1 mm clasts of graphite-magnetite aggregates make up 1 wt.% of Dimmitt [Scott et al., 1981a]. These aggregates are very similar in texture and composition to the graphite-magnetite matrix areas of DT1 and DT2. Such aggregates were probably derived from fragmented class of chondrites like DT1 and DT2 [Scott et al., 1981a]. Isolated graphite- magnetite aggregates have been identified within the clastic matrices of other H and L chondrite regolith breccias as well as in several type 3 ordinary chondrites Scott et al., 1981a; McKinley et al., 1981]. The graphite-magnetite aggregates in type 3 ordinary chondrites generally contain more metallic Fe, Ni than the graphite-magnetite aggregates in the regolith breccias. Discussion Nature of Clastic Matrix Low-Ca pyroxenes in the Dimmitt clastic matrix are very heterogeneous (Figure 1) but do exhibit a peak in their compositional distribution in the range characteristic H4-6 chondrites Can sTalindi Conse and Ron ' 8 sore complicated and comprises three components: a prominent peak in the range characteristic of H4-6 chondrites 48216.9- he Gomes and Keil, 1980), constituting constituting analyses, a minor peak at wag1-24; unequilibrated olivines, ranging from Dazulat These three components constitute different proportions in the four analyzed areas peak is (A-D). For example, the minor Faz-24 beek 15 most prominent in areas C constitutes 30% of the olivines) and least prominent in area A (where it constitutes 7%). Compositional distribution histograms of olivines in the matrices of most H chondrite regolith breccias [e.g., Noonan and Nelen, 1976; Keil and Fodor, 1980; MeSween and Lipschutz, 1980; MeSween et al., 1981] are similar to that of the Dimmitt matrix in having prominent peaks in the range characteristic of H4-6 chondrites superimposed on an unequilibrated background. The proportion of olivines in the H4-6 range in the Dimmitt matrix (40‡5%) is comparable to that of Tysnes Island (50‡5%), Weston (50‡5%), and parts of Leighton (20‡5% in FD3) [MeSween and Lipschutz, 1980; McSween et al. 1981]. [In some regolith breccias, e.g., Pantar II (but not Pantar I), (Fredriksson and Keil, 1963; Lipschutz et al., 1982) and, to lesser extents, Nulles and Cangas de Onis (A. B. Rubin, unpublished data), matrix olivines are almost completely equilibrated.] we found in the Dimmitt Howevound the minor alivine peak at Paz-204 bich matrix, has lot been observed in other H chondrite regolith breccias. The area-to-area matrix heterogeneity in Dimmitt seems to be smaller than that in Leighton [MeSween et al., 1981] and Pantar [Lipschutz et al., 1982]. The simplest explanation for these olivine and low-Ca pyroxene compositional distributions is that they result from the non-uniform admixture of unequilibrated material to comminuted equil-ibrated clasts [e.g., MeSween and Lipschutz, 1980]. About 20% of the glassy chondrules in the Dimmitt matrix (which must be part of the unequilibrated component) also have olivine compositions in the range of H4-6 chondrites, Fa 17-20. This indicates that the prominent peak in this compositional range in the Dimmitt matrix is not due entirely to olivines broken out of equilibrated clasts. It is unlikely that the unequilibrated component in the breccia matrices is primarily normal H3 material. The least equilibrated ordinary chondrites (pétrologic types 3.0-3.5; Sears et al., 1980), which have the most heterogeneous olivines, contain 10-15 vol.% fine-grained opaque silicate matrix [Huss, 1979]. This material is relatively rare in H chondrite regolith breccias; we have only found it as thin rims around 5% of the chondrules in the Dimmitt matrix and around 1% of the chondrules in the matrix of Cangas de Onis. In addition, H3 clasts are very rare in H chondrite regolith breccias; they have been reported only in Weston Noonan and Nelen, 1976]. less abundant in Evidently, H3 clasts were far the H chondrite parent body regolith than were carbonaceous chondrite class. Thus normal consolidated H3 material is probably not a major source of the unequilibrated component in the Dimmitt matrix. Nevertheless, the occurrence of opaque matrix rims around 5% of the chondrules in the Dimmitt matrix suggests that the matrix contains a minor H3.0-3.5 component. [This same component may have contributed the microchondrules to Dimmitt, because Rubin et al. (1982a) determined that, among type 3 chondrites, microchondrules are most abundant in the H and CO groups.] The approximate volume of opaque matrix material in the Dimmitt matrix (0.2‡0.2%) was derived from the proportion of chondrules that have opaque matrix rims (5%), the abundance of chondrules in the matrix (40%), and the maximum (opaque matrix rim volume)/(chondrule volume) ratio (10%). Because type 3.0-3.5 ordinary chondrites contain 10-15 vol.% opaque matrix [Huss, 1979 and the Dimmitt clastic matrix contains 0.2 vol.% opaque matrix, then 2 vol.% of the matrix consists of normal H3.0-3.5 material. The graphite-magnetite aggregates in the clastic matrices of ordinary chondrite regolith breccias were probably derived from fragmented graphite-magnetite-rich type 3 chondrites like DTI and DT2 [Scott et al., 1981a]. Chondrules and silicate grains from such chondritic clasts must also contribute to the unequilibrated component of the clastic matrices. The glassy chondrules in the Dimmitt matrix are smaller than those of most H3 chondrites: 75‡5% of them are 500 um in apparent diameter. These chondrules are similar in size to those in the graphite-magnetite-rich chondrites (100-500 um; Scott et al., 1981b), but are also of similar size to glassy chondrules in the Willaroy H3 chondrite (A. E. Rubin, unpublished data). Olivine and low-Ca pyroxene histograms very similar to those of the Dimmitt clastic matrix (including minor olivine peak at Pa 21-24,) can be obtained by combining the histograms of DT1 and DT2 Scott et al., 1981b], and assuming both clasts contribute equally to the Dimmitt matrix. However, DT2 is small (1 mm) and contains only about 5 chon-drules; thus its olivine and low-Ca pyroxene histograms may not be representative. The Dimmitt matrix contains roughly 2 vol.% graphite-magnetite aggregates. Clast DT1 contains 15 vol.% graphite-magnetite [Scott et al., 1981b]; hence, 10-15% of the clastic matrix is probably derived from graphite-magnetite-rich chondrite clasts. Some small proportion of the Dimmitt matrix may come from H3.6-4.0 chondrites. These contain 1-4 vol.% opaque silicate matrix material and 30% of their olivines have compositions outside the H4-6 range of [Dodd Noonan et al., 1976] However, et al., 1967; compositional distributions of the published olivines in H3.6-4.0 chondrites do not show a minor peak of 21-249 as does the Dimmitt matrix. The source ese olivines is unknown. Thus the unequilibrated material in the Dimmitt matrix probably consists of at least three components: 10 - 15% from graphite- magnetite-rich chondrites, 2% normal H3.0-3.5 material, and the remainder from another source. | |
