Recent Projects

current projects

Numerical model to determine the composition of H2O-NaCl-CaCl2 fluid inclusions based on microthermometric and microanalytical data

Ternary vapor-saturated liquidus phase diagram for the system H2O-NaCl-CaCl2.

Natural fluids approximated by the H2O-NaCl-CaCl2 system are common in a wide range of geologic environments, including sedimentary basins associated with hydrocarbon occurrences and MVT deposits, submarine hydrothermal systems, and other metamorphic, magmatic and hydrothermal environments. Thus, fluid inclusions approximated by the system H2O-NaCl-CaCl2 are common in a diverse range of geologic environments, and a methodology to interpret microthermometric data obtained from these inclusions is necessary to better understand geologic processes such as diagenesis, hydrocarbon migration, evolution of hydrothermal systems, metal transport, metamorphism and crystallization of magmas. The best source of information concerning the compositions of paleo-geologic fluids comes from fluid inclusions. The temperatures at which phase changes occur within a fluid inclusion during heating can be used to estimate the fluid composition, assuming that PTX phase relationships of representative fluid systems are available. In addition, the elemental ratios in saline aqueous inclusions can be determined by microanalysis, for example by laser ablation ICPMS (LA-ICPMS), which provides an additional constraint for determining the fluid composition when combined with microthermometric data.
This study presents a comprehensive set of empirical equations that fully describe the portion of the vapor-saturated H2O-NaCl-CaCl2 system that includes the range of compositions of natural fluid inclusions and the range of available experimental data. The numerical model is a quick and precise way to interpret microthermometric and microanalytical data from H2O-NaCl-CaCl2 fluid inclusions.

Effect of post-entrapment crystallization on the H2O-CO2 systematics of silicate melt inclusions

Concentrations of H2O and CO2 in silicate melt inclusions (MI) are often interpreted to represent magmatic “degassing paths,” in which depressurization of the magma during ascent leads to volatile saturation and release of CO2 and H2O from the melt. This interpretation explicitly assumes that melt inclusions are trapped under conditions of volatile saturation. If the trapped melt is saturated in volatiles, then the melt must exsolve a volatile phase if any post-entrapment crystallization occurs. Using PVTX data for the system NaAlSi3O8-H2O-CO2, we investigate the effect of small fractions of post-entrapment crystallization on the dissolved volatile content in an albitic melt inclusion entrapped under conditions of volatile saturation. As small amounts of albite crystallize on the inclusion walls, the melt becomes supersaturated in volatiles, and so nucleation and growth of a vapor bubble occurs and the pressure inside the MI is estimated from the difference between the molar volume of the volatile phase and the partial molar volume of the volatile in the melt using the EOS for H2O-CO2. Model results show that dissolved volatile contents in the melt vary in a systematic manner during low degrees of post-entrapment crystallization. More importantly,the H2O and CO2 contents of the melt (glass) in the inclusion define trends that are similar to those produced during open-system degassing. The model predicts that the dissolved CO2 in the melt phase may be almost completely lost to the vapor phase during post-entrapment crystallization, even though the vapor bubble occupies less than one volume percent of the melt inclusion. Thus, melt inclusions that all trap a volatile saturated melt with the same volatile concentrations, but experience varying degrees of post-entrapment crystallization on the walls, will define an H2O-CO2 volatile trend similar to an open-system degassing path.

Evolution of internal pressure and H2O dissolved in the melt phase in an H2O-saturated MI during post-entrapment crystallization.

A model for quartz solubility in H2O-NaCl and H2O-CO2 fluids

We build upon the fluid density-based approach (e.g., Fournier, 1983) to develop a predictive model for quartz solubility in aqueous hydrothermal fluids. The model expresses quartz solubility in H2O-NaCl, H2O-CO2 or H2O-NaCl-CO2 fluids as a function of the solubility in pure H2O, the density of the solution, the concentration of the dissolved components, and the density of pure H2O. An advantage of this model is that it appears to be solute independent. For example, the model predicts quartz solubility for H2O-NaCl as well as H2O-CO2 fluids, without any adjustment of the parameters. This result suggests that the model may be used to predict quartz solubility in aqueous fluids of any composition, including non-idealized geologic fluids with several dissolved components.
Another advantage of this model is the ability to predict quartz solubility in immiscible fluid systems, for which experimental quartz solubility measurements in coexisting liquid and vapor are less numerous compared to those for the single-phase region. Fluid phase equilibria control the dissolution and precipitation of quartz in hydrothermal systems, where bulk quartz solubility is a function of quartz solubility in the liquid and vapor phases and the proportions of those phases present. In addition, quartz may exhibit prograde solubility with respect to temperature in the liquid phase and retrograde solubility in the vapor (or vice-versa), within the same P-T range, indicating the possibility of quartz dissolution into one fluid phase concurrent with quartz precipitation from the other. The model can be applied to studies involving vein paragenesis, permeability evolution and reactive transport modeling, and fluid inclusions.

Quartz solubility versus temperature along the critical isobar for an H2O-NaCl fluid of 10 wt% NaCl.

Assessment of the role of boundary layers during melt inclusion trapping

Mathematical modeling predictions of boundary layer effects. t = time, R = crystal growth rate, d = diffusion rate of the chemical species, k = mineral/melt partition coefficient.

Melt inclusions (MI) are used routinely as a source of chemical information on the evolution of magmas. However, some processes occurring either during crystal growth or after MI entrapment can potentially compromise their composition, making the MI composition different from that of the surrounding melt. Arguably one of the most debated processes that can alter MI composition during entrapment is the formation of boundary layers (BL) at the crystal/melt interface. During crystal growth, slow diffusing elements that are incompatible in the host become enriched near the crystal/melt interface, while elements that are compatible become depleted, leading to the formation of a BL. The thickness and chemical composition of a BL are a complex function of crystal growth rate and element diffusion rate in the melt, and MI that trap aliquots of a BL will show stronger enrichment in incompatible elements (or stronger depletion in compatible ones) as an inverse function of their size.

Using the solution to the one-dimensional diffusion equation developed by Smith et al. (1955), and data available from the literature, we modeled BL for compatible and incompatible elements at the crystal/melt interface. MI analyzed from the Bishop Tuff and from Mauna Loa range in size from 8 to 375 mm (Bishop Tuff) and from 9 to 700 mm (Mauna Loa). Plotting ratios of incompatible to compatible elements maximizes any effects that BL would have on MI composition. As seen in the figure, there is no variation in the ratios as function of inclusion size.

Mathematical modeling (top figure) suggests that if BL processes affect MI compositions, enrichment or depletion in trace element concentrations could be high (> 2 times the far-field melt concentration), and would be most highly enriched in the very smallest MI (Fig. 1). However, we analyzed MI as small as 8 mm, and found no clear evidence of enrichment or depletion compared to very large (several 100s mm) MI in the same sample (bottom figure). While we cannot completely rule out that BL effects might be masked by the decrease in analytical precision for very small MI, we suggest that MI as small as 10 mm can be representative of the original melt composition.

Trace element ratios in MI from microanalysis, plotted against inclusion size.

How Much Variation in Homogenization Temperatures of Fluid Inclusions is Acceptable for an FIA?

Fluid Inclusion Assemblage (FIA) describes the most finely discriminated fluid inclusion association that can be identified based on petrography. An FIA thus defines a group of fluid inclusions that were trapped at the same time. This also implies that all the inclusions within an FIA were trapped at the same temperature and pressure, and all trapped a fluid of the same chemical composition. If the inclusions represent the original trapping conditions and have not reequilibrated, all inclusions in an FIA should have the same homogenization temperature. However, fluid inclusions presumed to have been trapped at the same time often show a variation in Th. The goal of this project is to identify FIAs in samples from different major geological environments and determine smallest range in homogenization temperature that might be expected under ideal conditions. Several factors might affect the range in Th within an FIA. The natural temperatyre (and pressure) fluctuations during formation of an FIA will vary depending on the geologic environment, fluid inclusion size may affect the Th (Fig. 1), as well as sample collection and preparation, thermal gradients during Th measurements, etc. Also the factors that influence these variations during and after trapping of fluid inclusions in different environments will be evaluated. In this project we are examining samples of ore and gangue minerals from different types of ore deposit environments. These environments include hydrothermal MVT deposits, deeper magmatic hydrothermal systems, porphyry copper deposits, epithermal deposits, lode-gold deposits in metamorphic environments, and sedimentary (diagenetic) environments.


Relationship between fluid inclusion size and homogenization temperature in fluorite. Fluid inclusions in an FIA have approximately same liquid/vapor ratios, hence using bubble volumes to compare inclusions is appropriate.

Geochemical Evolution of the Campi Flegrei Volcanic System from Studies of Melt Inclusions


Melt inclusion (in the circle) hosted in an olivine crystal (CF-FR-C1 p6).

Silicate-melt inclusions are small droplets (1-several 100 um) of silicate melt entrapped in phenocryst minerals during their growth. Melt inclusions thus provide a sample of the melt that was present in the magma chamber when the phenocryst grew and offer the possibility of reconstructing the chemical composition of the magma (silicate melt + volatiles) during its evolution from formation at mantle depth to its ascent and eruption at the surface. A basic assumption of melt inclusion studies is that the inclusions behave as closed (= isolated) systems after their formation; that is after trapping the silicate melt remains isolated from the evolving melt system in the magma chamber.

In populated areas in which active volcanoes are present, such as Campi Flegrei (Italy), understanding the role of volatiles in magmas provides important information to assess volcanic risks. In particular the volatile content in magmas (e.g. H2O, CO2, Cl, S and F) is of critical importance in determining the eruptive style and magma evolution, because degassing is usually one of the major phenomena before and during an eruption. Campi Flegrei eruptive products were selected based on age, eruptive characteristic, mineralogical and chemical compositions, and structural position of the eruptive center to examine possible relationships between magma chemistry, especially the volatile content, and eruptive style. Melt inclusions were heated inthe Vernadsky Stage in order to produce homogeneous glasses to be analyzed by Electron Microprobe (EMPA), Ion Microprobe (SIMS), Raman Spectroscopy, FTIR and Laser Ablation ICP-MS.

Experimental determination of H2O loss from melt inclusions during laboratory heating: Evidence from Raman spectroscopy

One of the most important assumptions for fluid or melt inclusion studies is that the samples have not gained or lost any material after trapping. However, several workers have suggested that water may be lost from melt inclusions (MI) during laboratory heating. To test this hypothesis, experiments were conducted to quantify H2O loss from MI during laboratory heating - the amount of water in the inclusions was monitored by Raman spectroscopy. Quartz-hosted MI from the early-erupted plinian stage of the Bishop Tuff were heated to 800 C and 1 kbar for 4 to 1512 hours (63 days). Previous studies had shown that unheated MI from this unit of the Bishop Tuff contain 4.8-6.5 wt % H2O. Because many Bishop Tuff MI fluoresce under visible (514 nm) Raman excitation, a method was developed to analyses silicate glasses and MI using an UV (244 nm) excitation source. The Bishop Tuff MI show insignificant H2O loss when heated for less than 12 hours, while up to 75% of the original H2O was lost after 1512 hours of heating. The rate of H2O loss decreases after a few hundred hours, suggesting either a change in H2O speciation or a change in the mechanism of H2O loss. Our results suggest that most silicic MI maintain their original H2O concentration if they are not heated for more than about 12 hours during laboratory studies.


Water content vs. time. Diamond is the average value for each data set, number in parentheses is number of inclusions in that data set, and line represents total range of values.

Determination of REE partitioning behavior between clinopyroxene, orthopyroxene, plagioclase, and silicate melt using melt inclusions from White Island, New Zealand


Clinopyroxene containing multiple "melt inclusion assemblages", containing melt inclusions, plagioclase inclusions, and orthopyroxene inclusions.

One of the most active research areas in igneous and mantle petrology involves the source regions for magmas emplaced into the crust. It is well known that magmas generated in different regions of the mantle and crust show distinct REE patterns. Thus, if the REE concentration of a melt that generated a given igneous rock can be determined, its source region may be inferred. One of the most commonly used methods to determine melt REE concentrations is to measure the REEs in rock minerals and, using "known" partition coefficients, the melt composition can be estimated. Partition coefficients have previously been estimated from both experimental studies and analysis of natural samples. However, many experimental studies involved either unrealistic growth rates or trace element concentrations, or both, or were done on bulk rock samples in which the relationship between phenocrysts and their host is uncertain. This project is examining partitioning behavior of REE and other trace elements between silicate melt and mineral using silicate melt inclusions and mineral inclusions trapped within the same "melt inclusion assemblage" in a host crystal. These contemporaneously trapped inclusions will be analyzed by electron microprobe and laser ablation inductively-coupled mass spectrometry (LA-ICP-MS), and the results will be used to calculate partition coefficients between silicate melt and mineral.

Melt inclusion assemblages: Definition and limits

Melt inclusions have proved to be a successful method of examining magma generation, eruptive styles, and monitors of volatile concentrations in magmas. However, there have been several questions raised about the validity of conclusions based on studies of isolated inclusions in separate crystals. Following the protocol developed for fluid inclusion studies, the melt inclusion assemblage (MIA) concept has been introduced, defined as a group of MI that were trapped contemporaneously. Assuming that the MI in an MIA were trapped at the same time, they should all have the same composition and homogenization temperature, assuming there has been no post-entrapment reequilibration. To test the MIA hypothesis, melt inclusion-bearing pyroxene and plagioclase from White Island, New Zealand are being analyzed using electron microprobe and laser-ablation inductively-coupled mass spectrometry to determine the consistency of data from individual MIAs.


Clinopyroxene containing multiple melt inclusion assemblages. Center of crystal is to the top of the image with growth zones evident by MIA.

Can we determine P-T formation conditions for fluid inclusions that homogenize by halite disappearance?


Fluid inclusions containing liquid, vapor, and a halite crystal that homogenize by halite disappearance are common in magmatic-hydrothermal systems, including porphyry-copper deposits. Previously, only limited experimental data for a composition of 40 weight % NaCl was available to determine the pressure inside such a fluid inclusion at final homogenization (halite dissolution). Using the synthetic fluid inclusion technique, inclusions that homogenize by halite dissapearance were synthesized under halite-saturated conditions at known pressure and temperature. The measured Th L-V (liquid-vapor homogenization) and Tm halite (halite dissappearance) were combined with known pressures to develop a relationship to estimate the minimum trapping pressure and salinity of the fluid. Comparison of experimental results to previous studies reveal an inconsistency between interpreted pressures and known geologic conditions for formation of porphyry systems. These results suggest that most inclusions which homogenize by halite disappearance are the result of either trapping of a solid halite along with the fluid, or reequilibration by necking down, or both.

Temporal and spatial distribution of fluid inclusion characteristics in the porphyry-epithermal environment

Geologic evidence for genetic link between deeper porphyry and shallower epithermal mineralization includes the Lepanto/Far Southeast system in the Phillipines and Red Mountain in Arizona. Geologic evidence and theoretical fluid flow models place reasonable constraints on the temporal and spatial variation of pressure and temperature in these systems. Experimental and theoretical studies of the H2O-NaCl system can be used to predict the fluid properties at the P-T conditions inferred from geological and exeprimental/theoretical studies. The purpose of this study is to combine these data to predict the evolution in fluid properties in time and space in the porphyry-epithermal environment, and to use thse results to predict the temporal and spatial evolution in fluid inclusion characteristics in this environment. These results, in turn, may be used in exploration to predict where in the overall magmatic-hydrothermal system a sample came from, and to provide vectors towards ore.


Oxygen isotope analysis of calcite using laser Raman spectroscopy.


Analysis of the stable oxygen isotopic composition of calcite has many applications in the geosciences. However, traditional techniques to determine d18O require a relatively large amount of sample, which results in an average analysis over a large area, and are destructive. Because 18O is heavier than the more common 16O isotope, 18O-C bonds vibrate at a different frequency then 16O-C bonds within the carbonate group in calcite. Being a vibrational spectroscopic technique, laser Raman analysis is able to resolve the vibrational energy of 18O-C from that of 16O-C bonds. By choosing an arbitrary homogeneous standard with a known d18O value, we can use the ratio of area or height for the 18O and 16O peaks to calculate the d18O value of the calcite at the scale of the size of the laser spot (a few microns), and at a theoretical precision of approximately +/- 1 per mil. The advantage of this technique is the ability to non-destructively analyze d18O values with a spatial resolution of a few microns, which is comparable to the scale of compositional/cathodoluminescent zones observed in many calcites in nature.

Do fluid inclusions in ore-stage gangue minerals record the composition of MVT ore fluids?

Gangue minerals (such as dolomite, calcite, and quartz) associated with Mississippi Valley Type (MVT) base metal deposits are commonly used to study fluid sources, composition and temperature of ore-forming fluids. It is reasonable to assume that primary fluid inclusions hosted in ore-stage gangue minerals which precipitated in the presence of metal-bearing solutions would record the concentrations of metals in this solution. However, previous studies comparing microthermometric data from coeval gangue and ore minerals using visible and infrared microscopy showed distinct and unexpected differences between the two. The purpose of this project is to determine ore-metal contents of fluid inclusions in gangue minerals associated with ore, and to compare these with calculated metal concentrations based on known solubility relationships. Individual fluid inclusions will be analyzed by laser ablation ICP-MS. The results of this study will help us to better undestand which gangue minerals to study to determine sources for ore-forming fluids. These results will also allow us to test the hypothesis that these deposits require at least two separate fluids, one to transport sulfur and another to transport metals. An absence of metals in gangue mineral fluids would not preclude gangue minerals from representing at least one component of the ore-forming solution.



Fluids Home |  Fluids Group |  Projects |  Publications |  Posters |  Facilities |  Software | Photos | Geosciences |  VT Home