1989 年 37 巻 6 号 p. 343-351
The assumption that the 40Ar/36Ar ratio of Ar trapped in volcanic rocks at eruption is atmospheric often gives a large systematic error in the K-Ar dating of Quaternary volcanic rocks younger than 1 Ma. There are two possible sources of error, the existence of excess Ar and mass-fractionated, initial Ar.
The major source of excess Ar is supposed to be magma. The 40Ar/36Ar ratio for magmatic Ar in the North-East Japan arc is tentatively estimated to be 340±10 from measurements of Ar in large phenocrysts separated from Quaternary volcanic rocks. Separation of phenocrysts is the most effective to decrease the systematic error caused by excess Ar.
The mass-fractionation of initial Ar should give the 38Ar/36Ar ratio which was changed from the atmospheric ratio. This can be checked by determining the 38Ar/36Ar ratio in samples. A peak-height comparison method is preferential to an isotope dilution method in the K-Ar dating of very young volcanic rocks because with the former, both the amount of Ar and the 38Ar/36Ar ratio in a sample can be determined at one measurement, while with the latter, an additional measurement is needed to know the 38Ar/36Ar ratio in the sample. Further, the statistical error in the isotope dilution method overwhelms that of the peakheight comparison method for large atmospheric Ar correction (e.g., 83% correction for 1% relative errors of isotopic ratio measurements). Since the systematic error caused by the mass-fractionated, initial Ar increases greatly with the increasing atmospheric Ar correction, it is important to correct the result for the mass-fractionation. Correction formulae are given.
The error sources other than those mentioned above are uncertainty in the blank correction and instabilities in the sensitivity of spectrometer and the Ar isotopic ratio measurement. Especially, Ar in Al- or other metal foil should be taken into consideration for the blank correction. In order to increase accuracy of the K-Ar age obtained, to check the systematic errors by measuring samples together with standard samples is of great importance in the K-Ar dating of very young volcanic rocks.
A measurable limit of K-Ar age is affected greatly by the atmospheric Ar correction. For andesites containing 5×10-11 ccSTP 36Ar/g and 2 wt% K, the limit is estimated to be 0.004 Ma for 98% atmospheric Ar correction. For most of Japanese Quaternary lavas which contain 2×10-10 ccSTP 36Ar/g and 2 wt% K, the limit is 0.015 Ma for the same atmospheric Ar correction.