EARTH AND PLANETARY SCIENCE LETTERS 17 (1972) 89-94. NORTH-HOLLAND PUBLISHING COMPANY [] ELECTRON MICROPROBE EVALUATION OF TERRESTRIAL BASALTS FOR WHOLE-ROCK K-Ar DATING* Edward A. MANKINEN and G. Brent DALRYMPLE U.S. Geological Survey, Menlo Park, California 94025, USA Received 24 July 1972 Revised version received 8 September 1972 Four basalt samples for whole-rock K-Ar dating were analyzed with an electron microprobe to locate potassium concentrations. Highest concentrations of potassium were found in those mineral phases which were the last to crystal- lize. The two reliable samples had potassium concentrated in fine-grained interstitial feldspar and along grain bound- aries of earlier formed plagioclase crystals. The two unreliable samples had potassium concentrated in the glassy matrix, demonstrating the ineffectiveness of basaltic glass as a retainer of radiogenic argon. In selecting basalt samples for whole-rock K-Ar dating, particular emphasis should be placed on determining the nature and condition of the fine-grained interstitial phases. 1. Introduction The usefulness of the whole-rock K-Ar technique for dating mafic volcanic rocks, particularly basalt, has been well established (see, for example [ 1-4] ). Critical to the success of this technique are the petro- graphic criteria used to select samples that will give meaningful calculated ages. Although these criteria have been discussed briefly by several authors [1, 5-9] and their validity confirmed by successful use, they are largely decuded from petrologic considerations and have not been independently investigated to any sig- nificant degree. We have used an electron microprobe to determine the location of potassium in four basalts whose reli- ability for whole-rock K-Ar dating was not only ap- parent upon petrographic examination but could be established independently. Two of these basalts give reliable whole-rock ages and two do not. Our findings reaffirm the validity of the normal selection criteria and demonstrate that particular emphasis should be placed on determining the nature and condition of the fine-grained interstitial phases when assessing a * Publication authorized by the Director, U.S. Geological Survey. basalt for whole-rock dating. Standard petrographic thin sections, 1" in diame- ter and ground to a high surface polish, were used for this study. Areas of each section representative of the whole rock were selected, photographed, and analyzed with an Applied Research Laboratories electron microprobe operated at 15 kV accelerating voltage. X-ray scanning images of the potassium K s radiation were obtained using an ADP diffracting crys- tal and oscilloscope-mounted polaroid camera. Sam- ple current and calcium K s scanning images were also made and photographed to assist in mineral iden- tification and location. The K-Ar age measurement for sample 509-69-7 was made using flame photo- metry with a lithium internal standard for potassium, isotope dilution for argon, and techniques summa- rized by Dalrymple and Lanphere [9]. 2. Summary of petrographic criteria In principle, a rock that is composed entirely of phases that are good K-Ar geochronometers should itself be a good geochronometer. Practical application of this principle requires that samples for whole-rock dating be sufficiently coarse grained that every com-90 £:A. Mankinen, G.B. Dalrvmple, The electron microprobe and K-Ar dating ponent can be identified. For this reason alone, basalts whose groundmass is clouded and obscured by crypto- crystalline opaque Fe-Ti oxides are usually considered unsuitable for whole-rock dating. Ideally, every phase in a prospective sample should be free of alteration, completely retentive of radiogenic argon, and primary, i.e., originating at the time the rock crystallized. The critical question is, where is the potassium, and is it in components that will have retained the argon that has been generated since the rock crystallized? From petrologic considerations, most of the potassium should be found in the last phases to solidify, and the few observations made with an electron microprobe confirm this [5, 9], except in the case of some Hawai- ian tholeiites with unusually low (~ O. 1%) K 20 con- tents [8]. The fine-grained interstitial phases are the ones most likely to contain the potassium and argon and are therefore the phases whose freshness and ar- gon retention properties are most likely to influence the reliability of the age measurement. In most basalts, these critical phases are fine-grained feldspar and/or glass. Alteration probably is one of the most common bases for rejecting a basalt for dating. The problems with alteration are threefold: (1)alteration may occur after initial crystallization of the rock, and the altering process may release all or part of the radiogenic argon that has accumulated; (2) the argon retention proper- ties of alteration products are largely unknown; and (3) the alteration products often are not primary and, even if retentive, will not record the time of formation of the rock. The degree to which alteration will affect the apparent K-At age of a rock is a direct function of the amount of potassium that that phase contains rela- tive to the rock as a whole. Alteration of a mineral such as olivine and possibly pyroxene, which usually contain a negligible amount of K20 compared with the total rock, will not adversely affect the whole-rock K-Ar age [2, 3, 6]. Studies by Webb and McDougall [10] and McDougall et al. [2] have suggested that some alteration of interstitial material may not result in argon leakage, but with the exception of minerals with negligible K20 content, it is not now possible to determine the amount or type of alteration that is tol- erable. Such samples probably should be avoided when- ever possible, for it is better to reject a suitable sample than to date an unsuitable one. With few exceptions, argon retention properties are known for only the most common minerals. A potas- sium-bearing phase that is widely found in many ma- flc volcanic rocks is glass, and the ability of basaltic glass to retain argon quantitatively is largely unknown. For this reason, basalt which contains glass is usually considered unsuitable for dating. Some basalts that contain small amounts of fresh glass seem to give re- liable ages [2, 10], but there are numerous examples of basaltic glasses
View Full Document
Unlocking...