We present high-precision chemical abrasion-isotope dilution-thermal ionization mass spectrometry (CA-IDTIMS) U-Pb zircon and isotope dilution-negative-thermal ionization mass spectrometry (ID-N-TIMS) Re-Os molybdenite geochronology of the world-class Tibetan Qulong porphyry Cu-Mo deposit. The data is used to constrain the timing, duration, and to yield implications for the ore-forming processes. The U-Pb data suggest that the preore Rongmucuola pluton crystallized at 17.142 ± 0.014/0.014/0.023 Ma (uncertainties presented as analytical/+ tracer/+ decay constant uncertainties), with emplacements of the synore P porphyry and postore quartz diorite occurring at 16.009 ± 0.016/0.017/0.024 and 15.166 ± 0.010/0.011/0.020 Ma, respectively. The Re-Os analysis of multiple independent molybdenite separations from single molybdenite-bearing quartz veins yields sub-‰ level analytical precision (<1‰), which is comparable with that of modern CA-ID-TIMS U-Pb zircon geochronology. The new Re-Os data indicate that the majority of the metals at Qulong were deposited over a minimum duration of 266 ± 13 k.y., between 16.126 ± 0.008/0.060/0.077 and 15.860 ± 0.010/0.058/0.075 Ma, with the main phase of mineralization being broadly synchronous with the emplacement of the P porphyry. However, our Re-Os data of molybdenite hosted within the Rongmucuola pluton imply that a portion of mineralization also predated the P porphyry and suggest that the P porphyry is an intermineral porphyry stock, although mineralization cut by P porphyry has not been previously documented or observed in this study. Correlating the Re-Os ages with vein types (A-B-D veins) demonstrates that the mineralization process was cyclical, with the presence of at least three short-lived (38 ± 11 to 59 ± 10 k.y.) mineralization pulses between 16.126 ± 0.008 and 16.050 ± 0.005, 16.040 ± 0.007 and 15.981 ± 0.007, and ~15.981 ± 0.007 and 15.860 ± 0.010 Ma. Coupling the Re-Os molybdenite ages and quartz (coprecipitated with the dated molybdenite) fluid inclusion data suggests that the cooling history was also cyclic, and implies a rapid cooling rate during the entire mineralization process (0.55° ± 0.11°C/k.y.), with much faster cooling rates (1.19° ± 0.82° to 1.27° ± 0.53°C/k.y.) for the individual mineralization pulses. The cyclic and rapid cooling process requires an additional cooling mechanism rather than the inefficient conduction, which we attribute to meteoric water circulation.
The presence of mineralization predating the intermineral P porphyry stock and the absence of evidence of an early porphyry stock at Qulong suggest that mineralization potentially can take place without contemporaneous magmatism at mineralization levels. As a result, dating magmatic events may not necessarily bracket the entire mineralization duration of a porphyry system. This highlights the importance of dating ore minerals to reveal a comprehensive picture of the magma-hydrothermal process. In addition, the absence of contemporaneous magmatism during mineralization has broad implications for the classification of porphyry copper deposits and mineral exploration. The timescales of mineralization cycles constrained here via direct dating of ore minerals (tens of k.y.) are comparable with those recently proposed through high-precision U-Pb zircon dating, diffusion modeling, and numerical simulation. We propose that the cyclic mineralization pulses are linked with the periodic release of volatiles from the lower crustal magma chamber, which are common for porphyry copper systems worldwide. The episodic/cyclic metal enrichment process potentially is one of the controlling factors of porphyry copper ore formation and is the key to differentiate the formation of economic and subeconomic porphyry deposits.
Finally, direct comparison of molybdenite Re-Os dates from different laboratories and with the zircon U-Pb system needs to account for the much larger uncertainties from tracer calibration and decay constants, respectively. As a result, we lose the necessary resolution to investigate the ore-forming process at the k.y. level. Therefore, to reduce these uncertainties, calibration between the two chronometers, using shared tracer solutions and a transparent data reduction platform within the community is required.