V31D-0679
Vanadium Partitioning and Mantle Oxidation State: New Experimental Data
Vanadium exists in multiple valences in natural basaltic melts, namely V2+, V3+, V4+ and V5+. Because most crystalline phases prefer to incorporate V3+ rather than V4+ and V5+, the crystal/silicate-melt partitioning of vanadium (DVcry/melt) tends to decrease with increasing oxygen fugacity (fO2). Such dependence has been experimentally demonstrated and used to estimate the fO2 of mantle and mantle-derived rocks. Recent modelling of V and V/Sc systematics in basalts has lead to the view that the relative fO2 of the upper mantle is constant, both through time and among the sources of different types of basaltic magmas (i.e. MORB, OIB and IAB). This is contrary to the notion given by other oxygen barometric methods on peridotites and basalts, which indicate an upper mantle heterogeneous in relative fO2. To explore further the potential of V abundances and V/Sc ratios to estimate the relative fO2 of mantle peridotites and basalts, and in particular to understand variations in mantle oxidation state better, we carried out an experimental campaign aimed at measuring DVcry/melt for all the relevant phases of the upper mantle (i.e. olivine, orthopyroxene, clinopyroxene, garnet and spinel) over a range of fO2 conditions large enough to pin down not only the behaviour of V3+ and V4+ but also V2+ and V5+. Experiments were done in 1-atm vertical tube furnaces (1300°C) and piston-cylinder apparatus (1275-1450°C and 1.5-3.2 GPa). For the high-pressure experiments, fO2 was controlled by the Re-ReOx/2 equilibrium (10-9 to 10-0.7 bar), whereas for the 1-atm experiments, fO2 was controlled by Ar-CO-CO2- O2 gas mixes (10-18 to 10-0.7 bar). Five starting compositions were used to ensure the presence of all the desired phases. Experimental products were analysed for major elements by electron microprobe and for trace elements by laser-ablation ICP-MS, which enables V to be measured precisely even at very low concentrations. Partition coefficients for all phases plot as approximately sigmoid-shaped curves in log D-log fO2 space. Details of the shape of the curve are controlled by the relative preference of each crystalline phase for a specific valence of V. For instance, orthopyroxene appears to particularly like V4+, so that the log DVopx/melt-log fO2 and log DVcpx/melt-log fO2 curves converge in the region of the diagram dominated by V4+, diverging in the regions dominated by V3+ and V5+. Contrary to previous studies, our results do not suggest a systematic increase in DVcpx/opx with decreasing fO2. Olivine and spinel, on the other hand, strongly prefer V3+ relative to V4+ and V5+ and hence for olivine and spinel the difference in partition coefficients between reducing and oxidizing conditions are more pronounced than that for pyroxenes. At high-pressure, DVgrt/melt and DVcpx/melt are very similar to each other, but the values of DVcpx/melt are about one order of magnitude higher than those obtained at 1 atm at comparable fO2. The cause of this discrepancy is being investigated.
V31D-0680
Diffusion experiments with natural alkaline melts: estimation of diffusion coefficients for a multicomponent system
Magma mixing is known to be interplay between convection and diffusion. Widespread evidence for magma mixing in the Canary Islands, Spain, motivated this work on the diffusion component of this process. We performed a time series of diffusion experiments at 1350°C using alkaline melts from volcanic products from this province. The two end-members are: 1) an alkali basalt (43% SiO2; η=4.412 Pa·s) and 2) a phonolite (59% SiO2; η=1000 Pa·s). For every experiment, a 3 mm thick disk of alkali basaltic glass and a 6 mm thick disk of phonolitic glass were loaded in a 5mm diameter Pt open capsule. They were arranged in a buoyantly stable geometry, where the denser material is placed at the bottom (alkali basalt, ρ=2.73g/cm3) and the lighter material at the top (phonolite, ρ=2.36 g/cm3). We run experiments during 4, 25, 49 and 125 hours at 1350°C at atmospheric conditions. This temperature, well above the liquidus of the system, has been kept constant during the whole experimental runs, with a negligible thermal gradient (< 2°C). No forced convection was applied so that the diffusion process takes over and the compositional gradient becomes the only parameter enhancing the mixing process. Microprobe analyses were performed along longitudinal lines from sections of the resulting products. In order to fit a curve to the experimental data, the diffusion equation is solved using the finite difference method. The conspicuous asymmetry of the diffusion profiles suggests that the D-value is strongly compositional dependent. A better fit can be obtained when the curve is splitted in two and different D-values are calculated for each end- member. Our first estimation of the diffusion coefficients for the major elements will be shown in this work.
V31D-0681
Chaotic Mixing in Magmatic Systems: a new experiment
Previous studies on magma mixing systems have evidenced that mixing processes could be controlled by chaotic dynamics. These processes are thought to be the source of fractal structures propagating within natural magmatic systems, from meter to the micrometer length scale (Perugini et al., 2006. EPSL, 234: 669-680 and references therein). We have developed a device for experimental studies of chaotic mixing dynamics in silicate melts at high temperatures (up to 1700°C). This device has been inspired by the journal bearing or eccentric cylinder geometry for viscous fluids for the study of chaotic mixing in slow flows (Swanson and Ottino, 1990. J. Fluid Mech., 213:227-249). This geometry is thought to be an ideal system for chaotic studies because a) it is experimentally accessible/feasible for silicate rheologies and b) it is subject to an analytical solution for the stream function. In the journal bearing system the flow region, is confined in the torus between the centers of the two cylinders. Their central axes are parallel but not coincident, i. e. the cylinders are eccentric. In order to generate chaos in a flow, the streamlines must be time dependent, resulting in alternating movements between the two cylinders. This means that at least one of the cylinders has alternating rotation directions. The dimension of this new experimental device follows the required main dimensionless numbers for a chaotic flow. Our first experimental goal is to characterize the mixing process in a prototypical system (haplogranite-haplobasalt)under variable mixing protocols. http://www.min.geo.uni- muenchen.de/
V31D-0682
Crystal Chemistry of Carbonate Apatites from High-Pressure Synthesis
Relatively large (50-200 μm) crystals of carbonate-bearing hydroxylapatite (CHAP) and fluorapatite (CFAP) have been grown from carbonate-rich melts at 1-3 GPa and used to determine structural details beyond the resolution of Rietveld powder diffraction methods, using X-ray single-crystal structure and FTIR spectroscopy. The new information includes the structural location of the channel (type A) and phosphate group (type B) carbonate ions in various composition series, as well as the location of the excess fluoride anion in francolite, substitution mechanisms, and identification of the hydrogen carbonate (bicarbonate) ion as a new apatite channel species. For equivalent conditions of synthesis, the uptake of A-B carbonate is greater for Na-bearing CHAP (up to 2 carbonate ions pfu) than Na-bearing CFAP (about 0.4 pfu). The Na cation and A and B carbonate ions are locally coupled in ratios of 1:1:1 in CHAP and 1:1:2 in CFAP, to minimize the effects of charge compensation and spatial accommodation. An extensive data base of type A and B site occupancies reveals that the amount of A carbonate in type A-B CHAP and CFAP is considerably under represented by the relative band areas for asymmetric stretching (ν3) and out-of-plane bending (ν2) of carbonate in FTIR spectra. The weaker absorption intensity and shift to higher wavenumbers of type A bands indicates that the carbonate ion is bound more weakly in the apatite channel than in the interior of the crystal structure. Thus literature spectra for apatites, and especially for apatite biomineralisation, showing dominant amounts of B carbonate should be re-evaluated.
V31D-0683
Determination of Partial Molar Volume of Ferrous Iron in Silicate Melts
The density of silicate liquids and partial molar volumes of oxide components have been a topic of intense study over several decades. Nevertheless, only a handful of density measurements have been performed in ferrous- iron dominated melts. Ferrous iron is among the most important elements in the crust and mantle. Knowledge of the partial molar volume of ferrous iron is critical to calculating phase saturation, exchange equilibrium, density etc. in major rock-forming phases as a function of pressure. We have developed a protocol for measuring density of silicate melts under reducing conditions using a single-bob Archimedean device. We are currently performing density measurements in the CaO-FeO-Al2O3-SiO2 system with the goal of accurately constraining the partial molar volume of ferrous iron as a function of temperature and composition. The device as currently configured uses a Mo crucible and Mo bob assembly. Preliminary work in mildly reducing melts suggest that up to 3 mol% Mo oxide in the silicate melt can coexist with Mo metal. On quench, Mo comes out of solution forming globular "micro-nuggets" and/or dendritic crystals. Under more reducing conditions in Fe-rich compositions, Fe in the silicate reacts with Mo metal to form "micro-nuggets" of equilibrium immiscible Fe-Mo liquid alloy. We believe that under these conditions the Fe content of the melt and the formation of "micro-nuggets" are governed by the reactions: \begin{eqnarray} {Mo(s) + \frac{y}{2} O2 \ (\!g)} &↔& {MoO y \ (\! silicate \ melt) ↔ Mo(\! alloy) + \frac{y}{2} O2 \ (\!g)}\nonumber {FeO(\! silicate \ melt)} &↔& {\rm Fe(\! alloy) + \frac{1}{2}O2 \ (\!g)}\nonumber \end{eqnarray} A review of the literature suggests that redox chemistry might also be controlling the formation of "micro-nuggets" of alloys of Fe and other transition metals. We will perform drop quench experiments into water, air, and air slowly to investigate this hypothesis. Since the crucible acts as an infinite reservoir of Mo, the maximum Fe content of the silicate melt is defined by saturation with the Fe-Mo alloy. Changing the oxygen fugacity also changes the Fe and Mo content of the alloy. This will permit resolution of both Fe and Mo partial molar volume independently.
V31D-0684
Experimental Constraints on Pt-Pd-Au Partitioning in an H2O-Haplogranite-Sulfide-Oxide phase assemblages
We performed experiments to to elucidate the role of sulfide and magnetite crystallization on the Pt, Pd and Au metal budgets of a water-saturated haplogranite magma. Experiments were carried out by equilibrating a NaCl- KCl +/-HCl aqueous solution, haplogranite melt, magnetite and two sulfide assemblages with elemental Pt, Pd, and Au. Oxygen fugacity was buffered at approximately Ni-NiO by the intrinsic buffering capacity of the experimental apparatus, and fS2 was fixed at Ni-NiO by sulfide phase assemblages of pyrrhotite (po) plus intermediate solid solution (Iss), and at a second fS2 by an Iss plus bornite assemblage. The experiments were carried out at 800C and 150 MPa, slightly higher than the critical pressure in the NaCl-KCl-H2O system for varying durations. Experimental run products were analyzed for major elements by using wavelength dispersive spectrometry electron probe microanalysis (EPMA) at UNLV and for trace elements via laser ablation inductively coupled plasma mass spectrometry (LA-ICPMS) at ETH in Zurich Switzerland. Resultant data were used to calculate Nernst-type partition coefficients for Pt, Pd and Au between the silicate melt and all solid phases. Calculated partition coefficients for Pt and Pd between all sulfides and melt suggest that po more efficiently sequesters Pd relative to Pt from the melt. Sulfide saturation of a melt and subsequent crystallization of po, Iss and bn will decrease significantly the concentration of Pd, Pt and Au in the melt. Po and Iss effectively fractionates Pd from Pt by at least one of magnitude resulting in a strong decrease in the Pt/Pd ratio of the melt. Additionally, moderate to high concentrations of Pd (1-2 wt%) in Iss at both fS2 values decreases Au solubility in Iss by an order of magnitude from the Au solubilities reported by Jugo et. al. (1999, Lithos). The presence of HCl in the assemblage has no statistically meaningful effect on the solubilities and partitioning behavior of the investigated elements. The results will be used to model the Pt, Pd, Au budget of an evolving magmatic system.
V31D-0685
Iron Diffusivity in Water Saturated Rhyolite Melt
We have quantified experimentally the bulk chemical diffusivity of iron and the solubility of magnetite in peraluminous, water-saturated rhyolite melt at 100 MPa and 800°C by performing experiments in which we equilibrated a single crystal of magnetite with water-saturated rhyolite melt. The oxygen fugacity of each run was buffered at nickel-nickel oxide (NNO) and the assemblage was saturated with a 1.8 wt. % NaCl eq. NaCl-KCl- FeCl2-HCl-H2O volatile phase. The experimental charge contained a cylinder of magnetite (activity Fe3O4=1), cored from a single crystal of magnetite and placed at the base of a gold capsule, synthetic rhyolite glass placed above the magnetite cylinder and aqueous vapor which occupied the remaining capsule volume. The concentration profiles of FeO (and Na2O, K2O, Al2O3, SiO2 and Cl) in the quenched melt (i.e., glass) were measured over a distance of 400 microns beginning at the magnetite-rhyolite interface and moving orthogonally away from this interface into the glass until the concentration of iron fell below the limit of detection. Diffusion profiles were fit by inverting the measured concentrations of iron in the melt through the error function and solving for the diffusion coefficient assuming a stationary planar boundary; the near-intersection of the error function regression with the origin justifies this assumption. The calculated bulk chemical diffusivity for iron in H2O- saturated rhyolite is 4 E-10 cm2 sec-1; this measured diffusivity is consistent, albeit one-half to one order of magnitude lower than data for other divalent elements (Ca, Mg, Sn) in rhyolite. The Co value used to fit the diffusion profiles is consistent with published data for the equilibrium concentration of iron in rhyolite melt and, thus, the data yield the solubility of iron in water-saturated rhyolite melt. The aluminum saturation index (ASI) of the melt, hence concentrations of Na2O, K2O and Al2O3, remains essentially constant in the melt across the entire measured diffusion length indicating that the bulk diffusivity of iron is not affected by coupled diffusion with these major elements. The chlorine concentration in the melt, however, increases markedly toward the magnetite-glass interface. This finding suggests that iron and chlorine are associated strongly in the melt and that the presence of iron in the melt, owing to magnetite dissolution increases significantly the chlorine "solubility" in the melt. The new results constrain the growth and dissolution rates of iron-bearing minerals during the evolution of hydrous felsic melt, including magma mixing, and the apparent association of iron and chlorine in the melt provides important constraints on the mass transfer of iron, chlorine and other metals, to an exsolved volatile phase and how this impacts the acidity, hence metal-scavenging potential, of the volatile phase.
V31D-0686
New experimental approach toward the determination of Os partitioning between metal and silicate
Highly siderophile elements (HSE; including Re, Os, Ir, Ru, Rh, Pt, Pd, Au), generally defined by their extremely high partition coefficients between Fe-rich metal and silicate (Dmet/sil >104), are of great interests for understanding core formation processes. Of these, Os is a key element due to applications of the 187Re- 187Os isotope system. Precise experimental determination of the metal-silicate partition coefficients for HSE have been hindered by analytical difficulties such as: 1) extremely low HSE concentrations in quenched glass that are close to or less than the detection limits of applicable measurement methods, and 2) the common development of HSE micro-nuggets scattered in the silicate glass. The presence of such nuggets can potentially lead to incorrect interpretations regarding Dmet/sil values. Here we employed a new approach to determine the DOsmet/sil at natural Os abundances by applying graphite capsule experiments in a piston cylinder apparatus. A paired sample of well-characterized iron meteorite and a komatiitic basalt, which have very different 187Os/188Os ratios, was put in a graphite capsule and processed at 10 kbar, 1450oC for 24-92 hours. The 187Os/188Os and Os concentration in the quenched glass containing a dispersion of metal nuggets were simultaneously measured by isotope dilution (NTIMS) by processing very small (~1 mg) samples liberated from the capsule. When the results are plotted on a 187Os/188Os-1/Os diagram, we observed a linear correlation which converges at the composition of the starting metal. A projection of the trend to the Os isotopic composition of the starting silicate indicates a pure silicate concentration that is several-times lower Os than the starting basalt. This method allows estimation of a D value for silicate free of nuggets without the need to separate nuggets from the silicate. Our new result implies extremely low solubility of Os in the silicate melt, evidently resulting from Os removal from the silicate. The Os concentration in the endmember silicate glass free of nuggets gives DOsmet/sil = 6×105.
V31D-0687
Quantifying Element Mass Transfer at Subduction Zone Conditions by using the Hydrothermal Diamond Anvil Cell and in-situ X-Ray Fluorescence
Our understanding of the physicochemical processes attending prograde metamorphism of subducting oceanic plates derives largely from inferences made from the chemistry of arc volcanics and traditional quench experiments. Using the chemistry of arc volcanics introduces complications owing to the history of melt + crystals during their protracted ascent through the magmatic plumbing system. Quench phenomena may affect measured element abundances in recovered phases, thus, in this project we aim to circumvent these problems by refining a technique using the hydrothermal diamond anvil cell (HDAC) to perform in situ experiments and utilize white synchrotron radiation and X-ray fluorescence (SR-XRF) to quantify element mobility of REE-monazite in H2O ± HCl ± NaCl during simulated prograde heating and compression at relevant P-T conditions of subduction zones. Our first target has been the dissolution of monazite at pressures from 2 to 5 GPa and temperatures to 800 °C. Schmidt et al. (2007; Lithos) report data at 2 GPa which serve as a benchmark for us to evaluate our results. Monazite is a common accessory mineral in a variety of igneous and metamorphic rocks associated with subduction zones and the REE chemistry of monazite has provided insight into processes attending magma evolution in arc systems. The interaction of REE-monazite with an aqueous fluid, generated by dehydration of the subducting oceanic plate, may significantly affect REE ratios in monazite which ultimately complicate geochemical inferences based on monazite chemistry. Solutions of known REE concentration are made and used as standards. The peak areas determined during experiments at slab conditions are compared to the peak areas of the known solutions to resolve the concentration of the REE in the sample. The partitioning of HFSE and LFSE are also of interest, and are being examined in the same experimental procedure. This technique allows us to quantify the fluid mobility of REE, HFSE, and LFSE at the conditions approximating subduction zones.
V31D-0688
Experimental Determination of Trace Element Partition Coefficients Between Zircon, Garnet and Melt
The problem of relating ages, as calculated by zircon U-Pb geochronology, to processes and hence geoological events is central to understanding mountain building and crustal evolution. Accurate P-T-t paths can only be produced if zircon growth can be linked to specific rock and mineral processes used to establish pressure and temperature values for metamorphic episodes. As a major metamorphic mineral in crustal events, garnet is widely used as a thermobarometric tool, and linking garnet growth to zircon formation could be used to refine the interpretation of U-Pb ages. Attempts to resolve this issue have focussed on REE partitioning between zircon and garnet, both of which strongly incorporate the HREE into their structure, and so it is possible there is a distinct REE partitioning signature which will highlight whether the two minerals have grown in equilibrium. There are two complementary methods to obtaining this information, empirical and experimental. Empirical methods of determining this signature using carefully selected rocks have proved troublesome, with a wide range of partitioning signatures found. This work has used experimental techniques to produce zircon-melt, garnet-melt and zircon-garnet-melt partition coefficients at a range of P-T conditions using synthetic materials. Zircon and garnet are grown in trace element equilibrium with a water-undersaturated granitic melt, which represents partial melts formed in the lower crust during anatexis. Temperature ranges from 850°C to 1000°C at a pressure of 5Kbar were produced using internally heated gas apparatus. Trace element concentrations were measured using SIMS analysis at the Ion Microprobe Facility at the University of Edinburgh. The experimental data produced will be applied to interpret chemical signatures in zircon in garnet-bearing metamorphic rocks, and will provide an objective basis for interpretation of the timing of growth or recrystallisation of zircon in many high-grade terrains.
V31D-0689
Site occupancy and distribution of Ti in zircon
Application of the Ti-in-zircon thermometer requires an understanding Ti site occupancy in zircon, the influence of chemical variables on Ti-saturation and other parameters that may influence Ti-distribution within zircon. Zircon was synthesized in phase assemblages in which Ti and Si activity was varied systematically. Ti K-XANES spectra were used to directly determine Ti site occupancy in synthetic zircons. The energy and intensity of the pre-edge feature clearly shows that Ti predominately resides on the Si site in zircon. Thus Si activity directly controls Ti solubility in zircon. The study of zircons crystallized under different TiO2-ZrO2-SiO2 phase assemblages can also be used to critically evaluate the difficulties and concerns surrounding the nucleation and crystallization of buffering-phases within experimental systems. In addition to site occupancy, this study has focused on the distribution of Ti within natural and experimentally produced zircon populations. Some synthetic zircons show notable Ti sector zoning in the absence of any correlative SEM-CL zoning. In addition to experimental systems, natural zircons studied in this program have shown that Ti-distribution can show antithetic or non-correlative distribution with regard to SEM-CL signal. This is not unexpected, so long as factors controlling Ti-saturation and saturation of CL-activating elements are decoupled during the growth of zircon, and if Ti occupies a different site to elements responsible for CL-activators. This research does not invalidate the use of zircon thermometry, but addresses a number of concerns regarding how, when and where such thermometry may be applied. Because other factors also influence Ti-saturation in zircon, attempts should be made to estimate Si and Ti activity during zircon growth. Furthermore, this research brings into question the use of common zircon mapping techniques - such as BSE and CL imaging - for the purpose of targeting and mapping Ti-distribution.
V31D-0690
Olivine Dissolution into MORB Melt: an Experimental and Theoretical Study
Olivine dissolution in a MORB melt experiments were conducted using a piston-cylinder apparatus at nominally 1250, 1350 and 1450°C temperatures and 0.5 and 1 GPa pressures. The experimental procedures follow those of Zhang et al. (1989). The goals of this work are to quantify diffusive crystal dissolution rates of olivine in basalt and to model convective crystal dissolution rates. From experiments, we (1) extract interface melt composition and examine the time scale for it to reach steady state concentration (Shaw, 2004); and (2) evaluate effective binary diffusion coefficient (EBDC) of the key dissolving component (MgO for olivine dissolution). At 0.5 GPa and 1250°C, interface melt composition changes in the first few minutes, and the e-folding time to reach a "steady state" MgO concentration is 171±71 s (1σ error). Within this time period, dissolution is controlled by both interface reaction and diffusion. Much beyond this time period, the dissolution is controlled by diffusion only. At 0.5 GPa but higher temperatures, the "steady state" MgO concentration is reached rapidly and the time scale (< 1 minute) cannot be resolved. It is hence inferred that the interface reaction rate increases more rapidly with temperature than MgO diffusivity, and olivine dissolution tends to be diffusion- controlled as temperature increases. Based on experimental data at 1250-1450°C and 0.5-1.0 GPa, both MgO diffusivity and the "steady-state" interface MgO concentration increase with increasing temperature and decrease with increasing pressure. Hence olivine dissolution rate in this melt also increases with temperature and decreases with pressure. From the results, we will model diffusive and convective dissolution rates in nature.
V31D-0691
Experimental Peridotite – melt reaction at one atmosphere: A textural and chemical study.
Sieve-textured clinopyroxene and spinel are common in mantle xenoliths and have been interpreted to be the result of partial melting, mantle metasomatism and host magma – xenolith reaction during transport. In this paper we test the latter hypothesis with a series of reduced and oxidized experiments at 1200 and 1156 C at one atmosphere using a synthetic leucitite melt and discs of natural peridotite. The experiments were performed at atmospheric pressure so that large sample volumes could be used; this allowed us to use natural peridotite with a grain size of several millimeters which facilitates sample preparation and analysis. The peridotite used in the reaction couples was cored from a 30 cm diameter anhydrous lherzolite xenolith from Meerfelder Maar in the West Eifel volcanic field, Germany (see Witt-Eickschen et al., 1998) for a detailed description of the xenoliths from this locality). The melt used was a synthetic version of a leucitite from the Rockeskyller Kopf volcano, also in the West Eifel. Our results show that sieve texture development on clinopyroxene and spinel in mantle xenoliths is the result of a multistage reaction process. In the first step, orthopyroxene undergoes incongruent dissolution to produce a silica- and alkali-rich melt together with olivine. As this melt migrates along grain boundaries it causes incongruent dissolution of clinopyroxene and spinel. The incongruent dissolution mechanism involves complete dissolution of the clinopyroxene or spinel followed by nucleation and growth of a secondary clinopyroxene or spinel once the reacting melt is saturated. The reaction of orthopyroxene, clinopyroxene and spinel with infiltrated host magma results in a range of melt compositions that are very similar to those observed in nature that have been interpreted to be due to very small degrees of partial melting.
V31D-0692
Fluorescences Of Inclusion Oils With Respect To Their Maturities And Sources:Simulation In Diamond Anvil Cell
Evolution of fluorescence color of inclusion oils has been simulated by measuring in-situ the fluorescence of `live' oils generated from thirteen oil-prone kerogens from different depositional environments during a closed system pyrolysis in a diamond anvil cell at three heating rates (3, 8, and 25°C /min) up to 600°C. The measured fluorescence intensity of samples increases considerably within maturation intervals close to oil windows, while the fluorescence spectra of oils generated from all studied kerogens exhibit exclusively a progressive blue-shift of peak wavelengths (λmax) and red/green quotient (Q) upon increasing maturity. The observation is consistent with the maturity dependence of spectral shift trend widely recognized in natural hydrocarbon inclusions or crude oils. This study furthermore reveals that the acclaimed direction of spectral shift for inclusion oils is mostly independent of sources of their parental kerogens, implying that some reverse or anomalous trends reported in inclusion oils may be attributed to other processes subsequent to their generation, which significantly altered the fluorescence properties of oils. However, the experimental maturity corresponding to each color (λmax or Q) of oils can vary significantly (± 0.2 %Ro) among their sourced kerogens, suggesting that single fluorescence color of crude or inclusion oil is both maturity- and source-dependent and therefore may not be a good indication of its maturity. In addition, the blue-shift of cumulative oils generated from all kerogens approaches similar minima λmax around 564 nm or Q- value around 0.6 at maturity close to the middle or late stage of oil generation, implying that most late cumulative oils may exhibit similar colors. The oils generated in a maturity interval in late stage, however, can exhibit color of shorter wavelength less than the minimum.