This paper focuses on reasons for the appearance of tetrad effects in chondrite-normalized REE distribution patterns of granitoids (Li-F granites, peralklaine granites, ongonites, fluorine-rich rhyolites, and granitic pegmatites). The analysis of published data showed that the alteration of such rocks by high- and/or low-temperature metasomatic processes does not result in most cases in the appearance or enhancement of M-type tetrad effects in REE patterns. These processes are accompanied by the removal or addition of lanthanides, a W-type sag appears between Gd and Ho, and negative or positive Ce anomalies develop sometimes in REE patterns. The formation conditions of peculiar rocks enriched in Ca and F from the Ary Bulak ongonite massif (eastern Transbaikalia) and the character of REE distribution in these rocks and melt inclusion glasses were discussed. Based on the obtained data and the analysis of numerous publications, it was concluded that REE tetrad effects in rare-metal granitoids are caused by fluoride-silicate liquid immiscibility and extensive melt differentiation in the accumulation chambers of fluorine-rich magmas. A considerable increase in fluorine content in a homogeneous granitoid melt can cause its heterogenization (liquation) and formation of fluoride melts of various compositions. The redistribution of lanthanides between the immiscible liquid phases of granitoid magma will result in the formation of M-type tetrad effects in the silicate melts, because the REE patterns of fluoride melts exhibit pronounced W-type tetrad effects. The maximum M-type tetrad effect between La and Nd, which is observed in many rare-metal granitoids, is related to the character of REE partitioning between fluoride and silicate melts and F- and Cl-rich magmatic fluids. The low non-chondritic Y/Ho ratio (<15) of many rare-metal granitoids may be indicative of a contribution of fluoride melts to the differentiation of F-rich silicic magmas, from which these rocks were formed. The influence of high-temperature F-Cl-bearing fluids on melts and/or granitoid rocks results in an increase in Y/Ho ratio owing to the elevated solubility of Ho in such fluids.
Analysis the development of large fluid segregations in a flux of small fluid bubbles during the degassing of granitic (pegmatitic) melts indicates that the velocity of the buoyant ascent of fluid bubbles depends on their sizes, the viscosity and density of the melts, and the duration of melt flow. Possible variants of the primary and secondary boiling of magma are discussed depending on the P-T conditions and concentrations of H2O, F, B, and other components dissolved in the magma. The possible density ranges of the fluid phases are considered, along with the viscosity and density of granitic (pegmatitic) melts, velocities of the buoyant ascent of fluid bubbles in them, and the processes of their coalescence and accumulation in the temperature range of 650–850°C. Provisional evaluates are obtained for the duration of melt crystallization and the development of intrusive massifs and dikes of granites and syngenetic intragranite and epigenetic (intruded into the host rocks) granite pegmatites. Simulation data and geological observations suggest that large fluid segregations were formed already in the magma chambers in which the heterogeneous granite (pegmatitic) magma was derived, before its emplacement into the host rocks. These generation regions could be magma chamber areas within granite intrusions, in which melts enriched in volatiles were accumulated and then degassed with the release of fluid phases of various composition and density. The crystallization of fluid-rich melts under favorable conditions gives rise to granites with miarolitic structures. The emplacement of heterogeneous pegmatitic magma (which consists of immiscible silicate melts and large fluid segregations) into the host rocks results in that these segregations (would-be miaroles) occur in any part of the pegmatite-hosting chamber. This explains why miaroles of significantly different composition and with broadly varying proportions of their filling minerals may occur in various parts of pegmatite veins or their swells, as well as near contacts with the host rocks.
The P-V-T-X properties of H2O-salt systems are compared depending on the solubility coefficient of compounds contained in these systems and the presence or absence of critical phenomena in the saturated solutions. Data on synthetic and natural inclusions captured in minerals at elevated temperatures and pressures and employed to discuss the principal features of phase diagrams of the H2O-NaCl system (type I) and H2O-NaF system (type II or P-Q type). It is demonstrated how characteristics of magmatic fluids of various types are manifested during the development of miarolitic pegmatites (Malkhan field in Transbaikalia) and during the crystallization of F-rich ongonitic melts (Ary-Bulak Massif in eastern Transbaikalia). Characteristics of solutions and gas-rich (gaseous) fluid inclusions in quartz phenocrysts from porphyritic ongonites (disappearance of the liquid regardless of its density and the overall salinity near the critical point of water, distinctive features of the dissolution of the crystalline phase, and the ability of the inclusions to withstand heating to 1400°C without decrepitation), and the richness of the fluid-magmatic system as a whole in F suggest that the ongonite melt crystallized in the presence of low-density NaF-bearing fluids of the P-Q type with a minor admixture of chlorides. It is important to identify the type of solutions in the fluid inclusions, because without knowing this type, it is impossible to accurately enough calculate the pressure at the temperatures of inclusion capture. For example, the unwarranted classification of solutions of type II (P-Q) in inclusions with the chloride system results in a significant overestimation of the calculated fluid pressures. A technique is proposed for studying the high-temperature immiscibility region in P-Q systems based on data obtained on gaseous fluid inclusions.