KGS Home Geology Home

Kansas Geological Survey, Bulletin 199, Part 2, originally published in 1970


Rb-Sr Geochronologic Investigation of Basic and Ultrabasic Xenoliths from the Stockdale Kimberlite, Riley County, Kansas

by Douglas G. Brookins and Michael J. Woods

Originally published in 1970 as Kansas Geological Survey Bulletin 199, Part 2. This is, in general, the original text as published. The information has not been updated.

Abstract

Five basic to ultrabasic xenoliths from the Stockdale kimberlite have been studied by the Rb-Sr geochronologic method to determine their ages of formation and subsequent thermal history. The data are summarized as follows:

Rock Type Rb-Sr Mineral
Isochron Age (m.y.)
Initial Sr
(87/86)
Diorite (1139i) 1,585 ± 95 0.7016 ± 0.0011
Hypersthene Gabbro (1139h) 82 ± 25 0.7062 ± 0.0002
Hypersthene Gabbro (1139j) 692 ± 108 0.7051 ± 0.0007
Garnet Granulite (1141a) 240 ± 8 0.7042 ± 0.0002
Eclogite (1128g)   0.7044 (maximum)

The diorite age represents the time of formation of igneous material at about 1,600 m.y. ago because (1) the initial Sr (87/86) ratio is extremely low and (2) the constituent minerals have not been recrystallized. The other ages represent post-formational events. One hypersthene gabbro (1139h) has not been closed to its environment (evidenced by secondary K-feldspar) and the age is spurious; the other hypersthene gabbro shows evidence of a thermal event at about 700 m.y. ago. The garnet granulite (1141a) of cognate origin shows evidence of a thermal event at about 240 m.y. ago; and, although an age cannot be calculated due to lack of data, the eclogite (1128g) has been derived from a deep-seated (cognate origin) source as indicated by the low Sr (87/86) ratio. These scattered but internally consistent ages argue for several periods of thermal activity (metamorphism) prior to entrapment in their host kimberlite (the kimberlite itself is thought to have crystallized about 750 m.y. ago in the upper mantle and to have been emplaced about 100 m.y. ago). This further suggests that entrapment occurred under near adiabatic conditions (hot kimberlite mass) or that ascent of the kimberlite as extremely rapid and relatively cool. The data also indicate the extreme complexity of the lower crust and upper mantle in the Midcontinent and the need for many more geochronologic and other investigations of these xenoliths.

Introduction

Five basic to ultrabasic xenoliths from the Stockdale kimberlite, Kansas, have been investigated by the Rb-Sr geochronologic method to learn more about the nature and the history of the lower crust and upper mantle in northeastern Kansas. The Stockdale kimberlite (location: SE SW NE and NE NW SE sec. 23, T. 8 S., R. 6 E.) is one of six kimberlites which intrude Lower Permian sedimentary rocks in northeastern Riley County, Kansas (Brookins, 1970b), although their age of emplacement is thought to be post-Early Cretaceous (Brookins, 1969a). It has been described by Rosa and Brookins (1966) as possessing a typical kimberlitic mineralogy of serpentinized and carbonated olivine and pyroxenes, ilmenite, pyrope, chloritized phlogopite, magnetite, and calcite in a highly serpentinized and carbonated matrix. All of the Riley County kimberlites contain xenoliths, most of which are accidental xenoliths of unaltered to serpentinized sedimentary rocks of probable local origin. Those from Stockdale, however, are of special interest because of their abundance and the diversity of rock types of probable deep-seated origin. These xenoliths have been very briefly described by Brookins (1969b) and in much more detail by Woods (1970) and Woods and Brookins (unpublished data; manuscript submitted). The five xenoliths studied during this investigation will be described later in this note.

The structure and history of the rocks of northeastern Kansas are complex and little detailed information is available. The uppermost rocks in Riley County consist of (except for a small exposure of Dakota sandstone of Lower Cretaceous age) Lower Permian shales, limestones, and cherts. The entire Paleozoic cover over the Precambrian basement rocks is only 2 km or so thick. Wells drilled to the Precambrian rocks in Riley and surrounding Washington, Marshall, Pottawatomie, Geary, and Wabaunsee counties usually reveal the presence of rocks of adamellitic composition although occasional quartzites are noted. This is somewhat surprising in that the uppermost Precambrian rocks in northern Riley County (and surrounding area) are thought to be basalts and arkoses of the Rice Formation (and related strata) which extend into northeastern Kansas from the Lake Superior Province (Lidiak, 1969). It has been previously noted (Brookins, 1969b; 1970b) that xenoliths of the above mentioned rock types (adamellites, quartzites, basalts, arkoses) are extremely rare, suggesting that the kimberlites were injected along structural zones of weakness in a rapid manner in the upper Precambrian crust. Where xenoliths of the above rocks are observed, however, they have been extensively altered (especially the K-feldspar-bearing varieties) to carbonate-rich rocks. This is consistent with the probable existence of a high PCO2 in the kimberlite (Brookins, 1967) and the instability of K-feldspar in such an environment.

The non-sedimentary xenoliths of deep-seated origin are of special interest since they presumably contain information pertinent to the composition and history of the lower crust and upper mantle. These xenoliths consist of a few rare schists and gneisses of probable sedimentary parentage mixed with an abundance of igneous and metaigneous rocks. Some of the xenoliths are apparently truly cognate, i.e., closely associated with the host kimberlite and of probable upper mantle origin. Others have formed under crustal conditions at depths less than 36 km which is the thickness of the crust in northeastern Kansas, and are referred to as accidental. Eclogites, pyroxenites, magnetites, and granulites(?) are representative of the cognate variety; diorites, gabbros, metagabbros, and hypersthene gabbros (and their metamorphic equivalent) are typical of the deep-seated accidental variety. The hypersthene gabbros are not called norites here because some of the hypersthene is of secondary origin. The igneous and metaigneous xenoliths of interest in this investigation are typically rounded to subrounded except where they have been fractured and rimmed by a serpentine- and chlorite-rich armor. These xenoliths range in size from 1 mm to 10 to 15 cm maximum dimension. Many of the xenoliths have been altered internally, especially if the "armor" has been cracked to allow kimberlitic material to infiltrate. When this occurs, the interior of the xenolith is usually completely altered to serpentine and magnetite and often replaced in part or wholly by younger calcite. Samples used in this study (described below) were selected from those showing no signs of alteration on fresh surfaces in hand specimen and in which alteration products identified in thin section could be attributed to reactions in an apparently isochemical system.

Acknowledgments

Financial support was provided by the National Science Foundation (Grant GA-10839), Kansas State University, and the Kansas Geological Survey. M. J. Woods acknowledges partial financial support from the Research Corporation, New York.

Sample Descriptions

The reader is referred to Woods (1970) for a thorough petrographic discussion of the samples used in this study. Only the more salient characteristics of each sample will be presented here except when further description is considered pertinent to the age study.

Sample 1139h

This sample is a hypersthene gabbro (metagabbro in part) consisting of labradorite, 72 percent; hypersthene, 16 percent; augite, 11 percent; magnetite, 1 percent. This sample is comparable to sample 1139j, the main difference being the relative amount of constituent minerals in each. The texture is hypidiomorphic-granular, with the pyroxenes showing partial resorption. Augite has the composition Wo45En33Fs22 and primary hypersthene the composition En73. Secondary hypersthene is more variable in composition, ranging from En76 to En83. Plagioclase has the composition An60 and has been recrystallized (Woods, 1970). Twinning is common, generally on the pericline law. Carlsbad, albite, and combined twins are also noted. Irregular blebs of secondary K-feldspar are ubiquitous within the plagioclase. Extinction is oscillatory, but there is no apparent compositional variation within grains.

Sample 1139i

This sample is a diorite consisting of plagioclase, 49 percent; hornblende, 41 percent; biotite, 8 percent; and sphene (plus magnetite), 2 percent. The plagioclase is near An34 in composition and contains less than 10 percent K-feldspar. The texture is panidiomorphic-granular with hornblende euhedra and laths of biotite occurring between grains of subhedral plagioclase. The rock has a slightly oriented texture with biotite and, to a lesser degree, hornblende exhibiting preferred orientation. Granules of sphene are ubiquitous but are generally concentrated along the outside borders of hornblende grains from which they have exsolved.

This sample has been subjected to some alteration. The outermost 5 mm has been infiltrated with veinlets of hydrothermal calcite and, in this zone, hornblende has been altered in part to ilmenite and leucoxene, plagioclase has been sericitized, and the groundmass seems to consist of serpentine and sericite. Biotite is less common than in the interior of the grain, but the grains do not appear to be highly altered. Proceeding inward, the zone is 3 mm thick and consists of brown hornblende mantled by magnetite surrounded by sericitized plagioclase. The biotite is only slightly altered. The remainder of the xenolith, approximately 6 cm in diameter, is essentially unaltered and consists of green, unaltered hornblende and clear, brown biotite laths surrounded by slightly sericitized plagioclase. Sphene is abundant. Completely unaltered diorite forms an oval 2 x 0.5 cm volume near the center of the xenolith. Feldspar in this volume frequently occurs as euhedra, generally on the order of 0.2 mm maximum dimension. About 50 percent of the plagioclase is twinned on the albite and carlsbad laws, both types being represented equally. Combined twins are rare. The albite twins possess extremely wide lamellae, which is very unusual for oligoclase. X-ray data (Woods, 1970) indicate that the plagioclase in this xenolith alone of all the xenoliths studied has not been recrystallized.

Hornblende in this central zone forms well-terminated euhedra very similar to the feldspar in size. Optically, the hornblende contains (Mg, + Fe2+ - Fe3+ + Mn) = 53 - 57 (Woods, 1970; comparing data from 1139i to that for other dioritic hornblendes tabulated in Deer, Howie, and Zussman, 1962). Biotite appears to be Mg-rich but, because of its dark brown color, is not classified as phlogopite.

The minerals separated for Rb-Sr investigations were taken from the central zone and it is believed that altered material was avoided during the mineral separation and purification.

Sample 1139j

This sample is a hypersthene gabbro containing plagioclase, 55 percent; hypersthene, 25 percent; augite, 19 percent; magnetite, 1 percent. It is very similar to 1139i (described above). The texture is hypidiomorphic-granular with well-developed hypersthene euhedra and larger, but more subhedral, augite grains interspersed with anhedral plagioclase. The plagioclase forms a groundmass of large (up to 2 mm maximum dimension) interlocking grains, within which the pyroxene grains tend to cluster, forming a sub-glomeroporphyritic texture. Pyroxene commonly surrounds magnetite blebs, which often give the appearance of euhedra where they adjoin the straight borders of hypersthene grains. The plagioclase (An59) is generally untwinned and shows undulose extinction. Its structural state is high temperature, indicating recrystallization (Woods, 1970) and, hence, no compositional variation across individual grains. The grains are generally on the order of 1 to 2 mm maximum dimension, but there are abundant smaller (about 0.5 mm maximum dimension) unresorbed fragments of an earlier generation plagioclase. These fragments are generally twinned on the pericline law. Some of the plagioclase contains irregular fractures and a poorly developed cleavage. Cryptocrystalline feldspar (K-feldspar?) and glass fill some of the fractures and cleavage traces. Some grains contain oriented rutile needles while a few contain vacuoles.

Hypersthene occurs in two habits, as a primary crystallate and as an alteration product on augite. In both occurrences the composition is identical, near En73, suggesting that hypersthene phenocrysts were precipitating at the same time that the augite was reacting with the magma. Pleochroism is more pronounced in the secondary variety but it exhibits poorer crystallinity. Inclusions are rare and consist of ore minerals and biotite and an unidentified acicular mineral.

Augite generally forms subhedral phenocrysts which are conspicuously larger than hypersthene crystals. The composition from optical properties is approximately Ca45Mg37Fe18, although a number of grains display concentric zoning. Inclusions of biotite are always present, in some cases being confined to the center of an augite grain.

Sample 1141a

This sample is a garnet granulite which could also be described as a uralized gabbro or plagioclase-bearing pyroxene granulite containing clinopyroxene, 55 percent; plagioclase, 21 percent; hypersthene, 16 percent; garnet, 8 percent. The texture is hypidiomorphic granular with large (to 10 mm maximum dimension) interlocking crystals of pyroxene surrounding irregular pods of plagioclase (An58). Hypersthene forms irregular patches mantling clinopyroxene. Garnet is plentiful as an alteration product and generally occurs as patches along pyroxene-plagioclase interfaces, although it also forms as subhedral and euhedral crystals approximately 0.2 mm in maximum dimension. This latter occurrence may be indicative of a primary origin. A number of isometric euhedra have an apparent spinel morphology but are otherwise identical to the garnet; they always occur in association with garnet and proof of their identification as spinel (while probable) is lacking.

The clinopyroxene has the composition, from optical data, of Wo44En30Fs26, but this value is in error due to aluminum and other cations which affect the optical properties as well as the stoichiometry. Some of the grains show zoning, with variations in optic angle to ±4° noted. Exsolution lamellae of bronzite averaging 0.2 mm in width are developed parallel to the host (100). These are not generally terminated within the crystal and in some cases have been further exsolved as alteration garnet. The bronzite has the approximate composition En82. Patches of bronzite which are optically continuous with the lamellae are also noted.

Orthopyroxene occurs as either a late magmatic phase or an alteration product of clinopyroxene. It generally occurs as long, disconnected patches where clinopyroxene adjoins plagioclase. Sinuous but optically continuous patches over 4 mm in maximum dimension are also observed. Magnetite fills fracture zones, and platy inclusions of probable hematite are often noted. Occasionally, where orthopyroxene is in contact with plagioclase, a fibrous mineral believed to be kyanite (Woods, 1970) is developed.

Plagioclase generally forms completely enclosed pods up to 6 mm maximum dimension. Each pod consists of aggregates of anhedral grains, each with rounded irregular borders. Zoning is present in most, but not all, grains. The range in composition is from An55 to An60. Sinuous exsolution blebs of lower refractive index than the host indicate that the feldspar may be considered antiperthite. Inclusions are ubiquitous but not abundant, consisting mainly of needles (rutile?) and short stringers of bubbles. Twinning is generally on the pericline law and occurs in phenocrysts showing a low degree of zoning. Albite twins are much less common and manebach and carlsbad twins extremely rare. In many cases recrystallization has completely removed all twinning, and a high temperature structural state is indicated by X-ray data.

Garnet euhedra occur frequently along the contacts between ortho- or clinopyroxene and plagioclase, as anhedral blebs within orthopyroxene grains, and as thin shells partially mantling orthopyroxene. Some euhedra are completely surrounded by plagioclase grains, isolated from pyroxene grains.

Sample 1128g

This sample is an eclogite containing garnet, 87 percent, and clinopyroxene, 13 percent. It consists essentially of a continuous groundmass of non-crystalline garnet in which are contained a few rounded fragments of omphacitic clinopyroxene. The pyroxene has been partly resorbed but the process was interrupted before completion.

The pyroxene is an omphacite exhibiting cleavage parallel to (110), and rare parting parallel to (100). A few grains contain black granular material which also occurs along the borders of grains and is believed to be incipient garnet. One pyroxene grain contains apparent exsolved orthopyroxene.

Garnet is massive and no crystal faces are discernible. Based on optical determinations, it appears to be a grossular-bearing almandine-pyrope. Some grains contain a microcrystalline mineral which occurs as sinuous patches and may be quartz (suggested by its low birefringence and the fact that quartz is commonly found in some omphacite-rich eclogites). Some of the sinuous material shows a faint pleochroism, however, suggesting that it may be vestigial pyroxene.

A very few short stringers of a dark reddish-brown mineral occur within the garnet. High birefringence suggests rutile. This possible rutile and the microcrystalline mineral mentioned above comprise less than 1 percent of the total specimen.

Analytical Methods and Results

Minerals were separated from whole-rock samples by hand picking (using a binocular microscope), a Frantz isodynamic separator, and by heavy liquid techniques. The details are given by Woods (1970). Unless otherwise specified, purity of the separates is greater than 99 percent. It should be pointed out here that, for Rb-Sr mineral isochron work, 100 percent purity is not an absolute necessity as is the case for other methods (i.e., K-Ar method) since the impurity that may be present has been derived from the same isochemical system. If, for example, a biotite separate is contaminated by 0.01 percent plagioclase, the effect will be reflected in a lower Rb/Sr ratio and lower Sr (87/86) ratio in the mixture, but the analysis will still yield a point on the mineral isochron. The advantage, of course, in working with very pure separates is to provide a greater spread in Rb/Sr ratios which allows the investigator to evaluate possible cases of open-system conditions.

Rb and Sr concentrations were determined by replicate X-ray spectrography using the U.S. Geological Survey basalt BCR-1 as a reference standard. The Rb/Sr ratios for the unknowns were determined in two ways: (1) by assuming Rb/Sr for BCR-1 = 0.145 and correcting the measured ratios to this value by peak height comparison, and (2) by calculating approximate Rb and Sr concentrations for the unknowns from peak heights (intensities) relative to BCR-1 assuming the latter to contain 48.4 ppm Rb and 333.8 ppm Sr (average of data presented in Hurley, 1969). Both methods yield results reproducible to ±3 percent of the mean, and this value has been ascribed as the error about the Rb87/Sr86 data reported in Table 1.

Table 1--Rb-Sr Data.

Sample Rb (ppm) Sr (ppm) Rb87/Sr86 Sr87/Sr86
R1139h 27 462 0.17 0.7063
P1139h 49 1072 0.13 0.7065
X1139h 14 54 0.75 0.7071
R1141a 14 131 0.41 0.7058
P1141a 23 1193 0.05 0.7041
X1141a (c) 12 30 1.16 0.7080
X1141a (o) 20 18 3.26 0.7150
R1139i 86 358 0.70 0.7146
P1139i 80 1430 0.16 0.7056
H1139i 39 104 1.10 0.7256
B1139i 84 171 1.43 0.7345
R1139j 9 634 0.04 0.7058
P1139j 23 1416 0.04 0.7046
X1139j (c) 10 31 0.93 0.7137
X1139j (o) 10 78 0.34 0.7094
Notes to Table 1: R = whole rock, P = plagioclase, H = hornblende,
B = biotite, X (c) = clinopyroxene, X (o) = orthopyroxene.

Strontium for isotopic study was separated by standard methods (see Chaudhuri and Brookins, 1969) and analyzed for its isotopic composition on a surface ionization, 6-inch, 60-degree, single filament, direction focusing mass spectrometer. The error about a single Sr (87/86) determination, normalized to Sr (86/88) = 0.1194, is ± 0.0005. Three runs on the Eimer and Amend interlaboratory standard SrCO3 (lot no. 492327) yielded Sr (87/86) = 0.7082, 0.7082, and 0.7083, values well within the limits of error for Sr (87/86) = 0.7081 ±0.0003 for twenty other runs on the standard from the same instrument at Kansas State University. The Sr (87/86) data are also presented in Table 1.

The data for all xenoliths except 1128g are presented graphically in Figures 1 through 4. The errors in age and initial Sr (87/86) have been calculated from the cubic least-squares computerized program described by York (1966). We use 5,000 m.y. as the half life of Rb87, resulting in use of λ = 1.39 X 10-11y-1 (decay constant) in the approximate age equation:

t = [(Sr87 /Sr86) measured - (Sr87 /Sr86) initial] / (Rb87 /Sr86) (λ)

Some laboratories use λ = 1.47 X 10-11y-1 as the decay constant for Rb87 (half life = 4,700 m.y.) which would result in a 6 percent lowering of the ages reported here but, since the ages are so different, it is irrelevant in interpreting the results.

Figure 1--Rb-Sr mineral isochron for diorite (1139i).

T=1,585 ± 95 m.y.

Figure 2--Rb-Sr possible mineral isochron for norite (1139h).

T=82 ± 25 m.y.

Figure 3--Rb-Sr mineral isochron for norite (1139j).

T=692 ± 108 m.y.

Figure 34--Rb-Sr mineral isochron for granulite (1141a).

T=240 ± 8 m.y.

Discussion

Four distinct ages (or, more properly, dates) are noted for xenoliths 1139h, 1139i, 1139j, and 1141a. No "true" age can be calculated for 1128g although the data (Table 1) do allow some limits on its genetic history to be placed (discussed below).

The diorite (1139i) yields a mineral isochron age of 1,585 ± 95 m.y. (Fig. 1) with initial intercept of Sr (87/86) = 0.7016 ± 0.0011. The initial intercept is extremely low as many rocks of dioritic or gabbroic composition of upper crustal origin yield initial ratios in the range 0.704 to 0.707. The value for 1139i is within the postulated range of upper mantle Sr (Hurley, 1967) of 0.702 ± 0.001, although its mineralogy and chemistry (Woods, 1970) argue for a deep crustal origin. The fact that the initial intercept is so low, coupled with the lack of recrystallization of the plagioclase, indicates that the rock has not been recrystallized (nor in any other way affected) since it became an isochemical system at about 1,600 m.y. ago. This date is significant in that it is the first reliable date for basement material in northeastern Kansas, as most of the ages reported by Muehlberger, Hedge, Denison, and Marvin (1966) are based on samples taken from cuttings of wells drilled into the upper Precambrian rocks and range from 1,320 to 1,950 m.y. Their work is difficult to evaluate since the ages reported were determined by the K-Ar method and the Rb-Sr mineral method. In the former the possibility of Ar loss (or, rarely, K loss or Ar gain) exists, and, in the latter, initial Sr (87/86) is assumed. This assumption is very serious for if the rock has been metamorphosed the initial ratio can only be accurately determined by the mineral isochron method. Use of an arbitrary value (especially if too low which is usually the case) will result in too high an age, hence many of their mineral dates by the Rb-Sr mineral method may be in error. The same is true for the whole-rock dates reported. Finally, the approximate 1,600 m.y. age reported for 1139i precludes its origin from the basalts of the Rice Formation of presumed Grenville (1,000 m.y. ±) age discussed by Lidiak (1969).

The two hypersthene gabbros (metagabbros in part), 1139h and 1139j, yield very different mineral isochron ages of 82 ± 25 m.y. [with initial Sr (87/86) = 0.7062 ± 0.0002] and 692 ± 108 m.y. [with initial Sr (87/86) = 0.7051 ± 0.0007] respectively. The former is thought to be spurious because of (1) its extremely low value and (2) the presence of exsolved or added amounts of K-feldspar indicating open-system conditions subsequent to the last pronounced thermal event affecting the xenolith. The 82 ± 25 m.y. date is within the error limits of the 100 ± 20 m.y. age of emplacement for the kimberlites proposed by Brookins (1970b), but no strong thermal effects were associated with the latter event. This is evidenced by (1) preservation of pre-emplacement K-Ar ages in chloritized phlogopites from which a temperature of injection of less than 150°C is inferred (Brookins, 1969a), (2) the lack of pyrometamorphic contact effects in the contact country rocks and in some unaltered accidental xenoliths of limestone and shale (Franks, 1966), (3) fission track dates of approximately 115 m.y. on apatites from altered adamellite xenoliths again indicating low temperatures accompanying emplacement (C. W. Naeser, written communication; Brookins, 1970c), and (4) the fact that the other xenoliths studied in this investigation do not yield such a low date. It is inconceivable that a thermal event strong enough to reset radiogenic Sr87 within 1139h would not have similarly affected the other xenoliths.

Both 1139h and 11391 have been recrystallized, however, as evidenced by X-ray study of their constituent plagioclases (Woods, 1970) and the fact that both mineral isochrons (Figs. 2 and 3) yield initial Sr (87/86) ratios greater than 0.705 (and significantly higher than the 0.7016 -± 0.0011 value reported for the diorite, 1139i).

The 692 ± 108 m.y. mineral isochron age reported for 1139j most probably represents a time of metamorphism, perhaps related to a Grenville-type event(?). In any event it is clearly separate from the other dates reported here. Since the pyrope crystallized in the upper mantle and the hypersthene gabbro in the crust, it is coincidental that this date falls within the limits of error of probable age of crystallization of pyrope from the Stockdale kimberlite of 745 ± 100 m.y. reported by Brookins (1969c). Further, following the arguments given above concerning the 82 ± 25 m.y. date for 1139h, if a thermal event related to entrapment of the xenoliths by the kimberlite at about 700 m.y. was strong enough to reset radiogenic Sr87 in 1139j, why were not the minerals of the other xenoliths reset at this date? Since hydrothermal (or metasomatic) alteration is not evidenced in 1139j and because it has been recrystallized, we interpret this age (approximately 700 m.y.) as a real age of metamorphism within the crust in northeastern Kansas.

The age of 240 ± 8 m.y. [with initial Sr (87/86) = 0.7042 ± 0.0002] for the garnet granulite shown in Figure 4 is also clearly different from the ages reported for the other xenoliths. The plagioclase has been recrystallized and it shows other signs of internal reequilibration but no evidence for externally derived chemical alteration. These factors coupled with the relatively high initial ratio argue for a strong thermal event about 240 m.y. ago. This date is well documented elsewhere in the United States, especially in New England where it is thought to represent a precursor to separation of the continents by sea-floor spreading. The mineralogy of 1141a argues for its derivation in the upper mantle at depths corresponding to a load pressure of approximately 13.5 kilobars. It is entirely plausible, considering current thoughts on plate tectonics, that metamorphisms occurred in the upper mantle long after metamorphisms ended in the overlying crustal rocks. Hence the difference in ages for 1139j and 1141a are not at all inconsistent. The data merely suggest a crustal metamorphism about 700 m.y. ago and a later metamorphism 240 m.y. ago occurring at greater depths, probably in the upper mantle. If the latter event had occurred after entrapment of the xenolith into the rising kimberlite then 1139i and 1139j should also yield this age and they do not. This 240 ± 8 m.y. age is also interesting in that it places some further limits on the history of the host kimberlite. Although captured in the upper mantle by the kimberlite, the metamorphism probably occurred prior to its capture, for otherwise the pyrope (745 ± 100 m.y.; Brookins, 1969c) would yield an age close to that time. It can be argued that the initial ratio (0.702) used for the pyrope age calculation is too low, but carbonatitic material from the kimberlites (Brookins, 1967) indicates that the initial Sr (87/86) ratio for the kimberlite mass must have been lower than 0.704 (and probably lower than 0.703). This information tends to indicate rapid piercement of the upper mantle and lower crust by an essentially solid kimberlite mass long after initial crystallization about 745 m.y. ago and also after 240 m.y. ago but before 100 m.y ago (probable age of emplacement). This matter is discussed in Brookins (1970c) and Brookins and Woods (1970).

The data for the eclogite, 1128g, do not allow an exact age to be calculated. However, using the data of Table 1 for 1128g and assuming an initial ratio of 0.7016 (diorite, 1139i, value), an age of 415 m.y. results. The minimum age (100 m.y.) is obtained if the initial ratio is increased to 0.7037. It is not unlikely that 1128g was also affected by the 240 m.y. event that is reflected in 1141a since both are upper mantle in origin. Future mineral analyses will help resolve this possibility. It is also noteworthy that the initial Sr (87/86) ratio for 1128g must be lower than 0.7044 (Table 1) and, since 1128g has been recrystallized and there is evidence for interruption of internal processes within the xenolith (Woods, 1970), the Sr (87/86) ratio is probably close to 0.7027 (which corresponds to the value required to yield a 240 m.y. date). It is further probable that the clinopyroxene from 1128g will possess a Rb/Sr ratio significantly higher than 0.18 and that the garnet will yield a Rb./Sr ratio much less than 0.1 because appreciable grossular is present in the garnet portion of 1128g (and Sr is greatly enriched in Ca sites, whereas Rb would be expected to be enriched in the alkali sites in the clinopyroxene since it is omphacitic in nature). There is no evidence for infiltration of the xenolith by fluids derived from an external source, so the Rb contained within sample must have been trapped there during initial crystallization.

Contamination of the xenoliths by the host kimberlite in such an incipient manner as not to be noted in even thin-section analysis is unlikely. Brookins (1967) has demonstrated the presence of two types of carbonate material in the Riley County kimberlites: a carbonatitic phase containing nearly 3000 ppm Sr and possessing Sr (87/86) values averaging 0.7038, and a sedimentary-derived portion containing less Sr (75 to 300 ± ppm) and a significantly higher Sr (87/86) range (0.7070 to 0.7093). Carbonate minerals comprise nearly 20 percent of the entire kimberlite plus inclusions; hence, these two varieties of carbonates are suspect. The very low Sr concentrations (Table 1) of the xenoliths and their constituent minerals coupled with the preservation of very different Sr (87/86) ratios and the fact that internally consistent isochrons (Figs. 1 through 4) have been defined preclude contamination from the carbonates. It is, however, likely that much of the Rb and possibly some of the radiogenic Sr87 is contained in micro- or cryptocrystalline veins or in inclusions (Erlank, 1969) in the constituent minerals of the xenoliths; but it is just as likely that both Rb and Sr87 so contained were derived from within the xenolith and not from an external source.

Conclusions

The conclusions reached from this investigation are few but significant. These are:

  1. The first reliable age of crustal basement rocks in northeastern Kansas is given as 1,585 ± 95 m.y. by the mineral isochron (Fig. 1) for a diorite xenolith (sample 1139i). This date is more reliable than those reported by Muehlberger, et al. (1966) because it has been obtained from a well-defined isochron.
  2. A lower(?) crustal metamorphic event at about 700 ± 100 m.y. is indicated by the mineral isochron for a hypersthene gabbro (sample 1139j) xenolith. This date may represent a reflection of a Grenville (1,000 ± 200 m.y.) event.
  3. Metamorphism, probably in the upper mantle but certainly in a different T, P range than that affecting 1139j, at 240 m.y. is indicated by a garnet granulite (sample 1141a). This date may be a reflection of upper mantle unrest known to have occurred, and generally attributed to plate tectonic activity just prior to the start of rapid sea-floor spreading. An eclogite xenolith (sample 1128g) may have been affected by this same metamorphism but the data are inconclusive except to state that its Sr (87/86) value (0.7044; Table 1) is significantly lower than the initial Sr (87/86) ratios for the hypersthene gabbros (1139h and 1139j) thus precluding a similar history.
  4. One hypersthene gabbro (sample 1139h) yields a spurious date of 82 ± 25 m.y. which is not attributed to a thermal event at that time (close to the time of emplacement of the kimberlites; Brookins, 1970b), because the other zenoliths are unaffected (i.e., do not show this date) but rather were exposed to open-system conditions within the xenolith. This may be evidenced by the presence of small blebs of K-feldspar within the plagioclase (P1139h) which are probably not due to exsolution phenomena (Woods, 1970).
  5. The kimberlite may have crystallized 745 ± 100 m.y. ago in the upper mantle at depths greater than those in which the garnet granulite (1141a) and eclogite (1128g) were formed, but injection of the kimberlite and concomitant entrapment of the xenoliths must have occurred after 240 m.y. and before 100 m.y. ago. This implies very rapid injection under near adiabatic conditions at depth, and an essentially completely crystallized kimberlitic mass. Emplacement in the lower crust was halted several times (evidenced by several sets of kink bands in micas according to Brookins, 1970b) but was extremely rapid and in at least one case explosive (Brookins, 1970a, 1970c) near the surface.
  6. Ascent to the surface through the upper Precambrian basement rocks probably occurred along fault or joint planes associated with the Abilene anticline. This accounts for the paucity of xenoliths of adamellites, arkoses, and basalts known to underlie the Paleozoic rocks of northern Riley County (Lidiak, 1969).
  7. Although most of the xenoliths of a deep-seated origin are altered, some are protected by a serpentine- and chlorite-rich (± carbonate minerals) armor which probably resulted from rapid reaction with the rising kimberlite. The serpentinization (and carbonation) probably did not take place when the xenoliths were initially trapped, although laboratory evidence indicates the probable existence of a H2O- and CO2- charged fluid in hypothetical kimberlite fluids stable to the T, P range of the upper mantle (Brookins, 1970b). It is more likely that the kimberlite-xenolith reaction produced a series of hydrothermal-like alterations in the outer portions of the xenoliths which were more prone to serpentinization at shallower depths within the crust.
  8. Although many more data are needed to substantiate the conclusions and inferences stated above, it is clear that a detailed Rb-Sr geochronologic study of the unaltered, deep-seated xenoliths of the Riley County kimberlites will help unravel the complex lower crust and upper mantle history of the Midcontinent. No other means of studying the history of these deep-seated rocks can offer so much potential, especially when coupled with a careful mineralogical and chemical study. Similar conclusions concerning xenoliths from the Delegate kimberlite pipe, Australia, have been reached by Compston and Lovering (1969).
  9. Finally, contamination from the kimberlitic host material or from later sedimentary-derived carbonate material is precluded by the significantly different Sr contents and isotopic compositions of the xenoliths relative to the possible sources for contamination.

References

Brookins, D. G., 1967, The strontium geochemistry of carbonates in kimberlites and limestones from Riley County, Kansas: Earth Plan. Sci. Letters, v. 2, p. 235-240.

Brookins, D. G., 1969a, The significance of K-Ar dates on altered kimberlitic phlogopite from Riley County, Kansas: Jour. Geol., v. 77, p. 102-107.

Brookins, D. G., 1969b, Riley County, Kansas, kimberlites and their inclusions (abs.): Geol. Soc. America, Abstracts for 1969, Pt. 2, South-Central Section, p. 4.

Brookins, D. G., 1969c, Possible age of crystallization of pyrope from the Stockdale kimberlite, Kansas: Geochem. Jour., v. 3, p. 135-140.

Brookins, D. G., 1970a, Kimberlite at Winkler Crater, Kansas: Geol. Soc. America, Bull., v. 81, p. 541-546.

Brookins, D. G., 1970b, The kimberlites of Riley County, Kansas: Kansas Geol. Survey, Bull. 200, 32 p. [available online]

Brookins, D. G., 1970c, Mechanism(s) of emplacement of Riley County, Kansas kimberlites (abs.): Geol. Soc. America, Abstracts with Programs, v. 2, p. 271.

Brookins, D. G., and Woods, M.J., 1970, High pressure mineral reactions in a pyroxenite granulite nodule from the Stockdale kimberlite, Riley County, Kansas: Kansas Geol. Survey, Bull. 199, pt. 3, 6 p. [available online]

Chaudhuri, S., and Brookins, D. G., 1969, The Rb-Sr whole-rock age of the Stearns Shale, eastern Kansas, before and after acid leaching experiments: Geol. Soc. America Bull., v. 80, p. 2605-2610.

Compston, W., and Lovering, J. F., 1969, The strontium isotopic geochemistry of granulitic and eclogitic inclusions from the basic pipes at Delegate, eastern Australia: Geochim. Cosmochim. Acta, v. 33, p. 691-700.

Deer, W. A., Howie, R. A., and Zussman, J., 1962, Rock-forming Minerals, Vol. 2: John Wiley and Sons, Inc., New York, 374 p.

Erlank, A. J., 1969, Microprobe investigation of potassium distribution in mafic and ultramafic nodules (abs.): Trans., Am. Geophys. Union, v. 50, p. 343.

Franks, P. C., 1966, Ozark Precambrian-Paleozoic relations: Discussion of igneous rocks exposed in eastern Kansas: Am. Assoc. Petroleum Geologists, Bull., v. 50, p. 1035-1042.

Hurley, P. M., 1967, 11, IV. Rb87-Sr87 Relationships in the Differentiation of the Mantle; in Ultramafic and Related Rocks, P. J. Wyllie, ed.: John Wiley and Sons, Inc., New York, p. 372-375.

Hurley, P. M., ed., 1969, Variations in isotopic abundances of strontium, calcium, and argon and related topics: M. I. T.-1382-17. Seventeenth Ann. Prog. Rpt. for 1969, U.S. Atomic Energy Comm., Contract AT (30-1)-1382, p. 99-101.

Lidiak, E. G., 1969, Burled Precambrian rocks of eastern Kansas (abs.): Geol. Soc. America, Abstracts for 1969, Pt. 2, South-Central Section, p. 17-18.

Muehlberger, W. R., Hedge, C. E., Denison, R. E., and Marvin, R. F., 1966, Geochronology of the Midcontinent Region, United States, 3. Southern area: Jour. Geophys. Res., v. 71, p. 5409-5426.

Rosa, F., and Brookins, D. G., 1966, The mineralogy of the Stockdale kimberlite pipe, Riley County, Kansas: Kansas Acad. Sci., Trans., v. 69, p. 335-344.

Woods, M. J., 1970, Petrography and geochronology of basic and ultrabasic inclusions from kimberlites of Riley County, Kansas: Unpub. M.S. thesis, Dept. Geol., Kansas State University, 95 p.

York, D., 1966, Least-squares fitting of a straight line: Canadian Jour. Physics, v. 44, p. 1079-1086.


Kansas Geological Survey, Geology
Placed on web Sept. 24, 2008; originally published in Dec. 1970.
Comments to webadmin@kgs.ku.edu
The URL for this page is http://www.kgs.ku.edu/Publications/Bulletins/199_2/index.html