Our careful analysis (Akagi et al., 1993) disclosed a W type tetrad effect in kimuraite, an REE carbonate mineral. To put constraint on the genetic stage responsible for the appearance of the effect, rare earth, element (REE) abundances were compared between kimuraite and its paragenetic mineral (lanthanite) and between kimuraite and its host rock. The Nd isotopic compositions of kimuraite and the host rock were also compared. Neodymium isotopic analyses conclude that the REE in kimuraite has a different source to its host rock. The similar magnitude of the tetrad effect among the paragenetic minerals of kimuraite rules out the process of mineralization as the stage responsible for the tetrad effect. It is inferred that kimuraite and lanthanite formed in situ from aqueous solutions enriched in REE which had infiltered the fissures of the host rock. This solution is considered to be responsible primarily for the tetrad effect of kimuraite.
REE abundances of the average shales of “North American shale composite” (NASC) and “Post-Archean Australian average shale” (PAAS) contain small but significant convex tetrad effect variations of the M-type relative to the chondritic REE. This is revealed in their REE patterns normalized by a USGS reference basalt of BCR-1. PAAS has a larger convex tetrad effect variation than NASC. The average shales are favorably used to normalize REE analyses of various geochemical samples. The tetrad effects of the average shales, however, call into question conventional interpretations of their shale-normalized REE patterns. Smooth shale-normalized REE patterns do not mean the absence of tetrad effect variations in their REE abundances relative to the chondritic REE abundances.
The Yatsugatake volcano, located on the volcanic front in central Japan, formed above an unusually deep Wadati-Benioff zone (ca. 160 km) as a trench-side volcano. In northeast Japan, incompatible element compositions and 87Sr/86Sr of basalts in the area above the Wadati-Benioff zone where it is ca. 110–170 km deep are clearly different from those where the Wadati-Benioff zone is more than ca. 180 km deep. These differences are inferred to have resulted from the differences in the dehydration reaction at the base of the mantle wedge dragged by the subduction and the contribution from a high 87Sr/86Sr reservoir (probably subcontinental upper mantle): the former zone; amphibole and chlorite dehydration and higher contribution from the high 87Sr/86Sr reservoir, the latter zone; phlogopite dehydration and lower contribution from that. The depth to the Wadati-Benioff zone beneath the Yatsugatake volcano is somewhat shallower than that of phlogopite dehydration, and the basalts show similar ratios between incompatible elements and 87Sr/86Sr to those from the amphibole and chlorite dehydration zone of northeast Japan. These observations show that the Yatsugatake magmas are produced by addition of amphibole and chlorite fluids, and the edifice is situated on the furthest position from the trench above those dehydration zone at the base of mantle wedge. It means that in central Japan there are no Quaternary volcanoes corresponding to the volcanoes in northeast Japan overlying the Wadati-Benioff zone less than ca. 160 km depth, though the amphibole and chlorite at the base of the mantle wedge exceed decomposition pressure (ca. 110 km) beneath the trenchside of the Yatsugatake volcano. This can be explained by the effects of subduction of the Philippine Sea plate, over the previously subducting Pacific plate beneath this area. Beneath the forearc-side of the Yatsugatake volcano, through the subduction of the Philippine Sea plate, the temperature of the mantle wedge has been lowered to suppress the generation of mantle diapirs. However, beneath the Yatsugatake volcano, such thermal disturbance has not taken place due to a possible splitting in the subducting Philippine Sea plate, sub-parallel to the subduction direction (NNW) in Miocene to Pliocene. The splitting might have been caused by tearing the Philippine Sea plate into the eastern and western parts in subducting around the Izu-Peninsula. Around the Yatsugatake volcano, the migration of the volcanic front toward the backarc-side is observed during mid-Pliocene (before the changing of the subduction direction of the Philippine Sea plate from NNW to NW), which might have been caused by the process of formation of this double-overlapping subduction zone.
For the accurate determination of trace gold in rocks and iron minerals by graphite furnace atomic absorption spectrometry, a two-stage solvent extraction method (diethyl ether and MIBK) was designed to prevent strong interference from iron and to effectively concentrate gold. The interference was perfectly condoled by adjusting some operating conditions in the solvent extraction. In the analytical method developed, the recovery of gold was >90%, and the precision (C.V.) was 11.7% (1 μg/l). When 0.5 ml of MIBK was used in the second extraction, the detection limit for gold was 0.13 μg/kg (concentration in rock) at S/N = 2. The gold concentration of geological standard rocks (JB-2 and JB-3) measured by the method proposed in this work was in close agreement with the recommended values. The gold concentration of sandstone and shale collected from the Shimanto belt, southwestern Japan, was in the range 1.0–2.5 and 1.1–1.8 μg/kg, respectively, while the gold concentration of pyrite contained in the above rocks was about ten times higher than that of the corresponding rocks.
A reliable determination method of gold in geothermal water at the ng/l level was developed by a combination of selective concentration method of gold using anion exchange resin and inductively coupled plasma quadrupole mass spectrometry (ICP-MS). When geothermal water samples were concentrated twenty-five fold, the detection limit was 0.4 ng/1. From reproducible determinations (N = 5) using the calibration curve method, the gold concentration in a Hachobaru geothermal water, Japan, was found to be in the range of 66 to 76 ng/l, while the standard error was in the range of 9.7 to 21.4%.
Carbon isotopic ratios and rare-earth element (REE) contents were measured for six Central African carbonados and two Brazilian ones. The δ13C values of all the Central African samples and one of the Brazilian samples are from –29.7‰ to –24.4‰, which are within the range of organic matters such as petroleum and coal, and are much lower than the typical values for ordinary diamonds of around -5‰. Another Brazilian sample gives diamond-like δ13C value of –8.8‰, suggesting that this is actually not a carbonado but an aggregate of ordinary diamond such as bort and ballas. Samples with a larger amount of impurities show higher REE contents and higher light-REE/heavy-REE ratios, implying the existence of a light-REE-enriched mineral, such as florencite, along grain boundaries between diamond crystals. Chondrite-normalized REE abundance patterns of the samples are similar to crustal materials such as shales rather than to kimberlites and ordinary diamonds, which are much more light-REE enriched than most of the studied samples. The Brazilian sample with a higher δ13C value, however, shows a kimberlite-like REE pattern which is clearly different from that of the other Brazilian sample. From our data, the crustal origin of carbonado is preferable to its genesis in the mantle.
A flow-through system, GC/GC/C/IRMS (gas chromatograph/gas chromatograph/combustion/isotope ratio mass spectrometer) was developed for carbon isotope analysis of methane in a sample of low level concentration. In this system, methane was roughly separated from a large amount of N2 and O2 in the first GC with packed column, then introduced into GC/C/IRMS and carbon isotopic composition was measured. The δ13C of methane in a sample of 92 ml in volume with concentration of 11 ppm was successfully obtained with a standard deviation of ±0.4‰.