DiadFit: An Open-Source Python3 Tool for Peak fitting of Raman Data from silicate melts and CO2 fluids

Abstract

We present a new Open-Source Python3 tool, DiadFit, for efficient processing of Raman spectroscopydata associated with silicate melts and CO2 fluids. DiadFit can fit Fermi diads, hot bands, 13C peaks,and Ne lines using various background and peak shapes. Thus, it is highly suited for workflows involving melt inclusion vapour bubbles and fluid inclusions (FI). We include generic peak fitting functions(e.g., for fitting carbonate and S-rich phases), and functions to quantify the area ratio of the silicate vs.H2O region of spectra collected on silicate glasses to determine H2O contents. DiadFit can convert between temperature, pressure and density using the CO2 equation of state (EOS), allowing calculation ofFI entrapment pressures (and depths in the crust) DiadFit can also calculate CO2 density from microthermometry measurements of Thomog . DiadFit can propagate errors in these EOS calculations using MonteCarlo methods. Documentation and worked examples are available (https://bit.ly/DiadFitRTD,https://bit.ly/DiadFitYouTube)

Determining the Pressure – Temperature – Composition (P-T-X) conditions of magma storage

“…In stark contrast to uncertainties associated with model choice when performing mineral thermobarometry or melt inclusion barometry, fluid inclusion barometry (Section 4) relies on the CO 2 equation of state to convert the measured density of a CO 2-rich fluid into a P for a specified entrapment temperature. Other than an ideal gas law (which does a poor job at high P), different published CO 2 equation of states predict very similar P for a given CO 2 density and T (~ 1-5% difference, Böttcher et al, 2012;Lamadrid et al, 2017;Span and Wagner, 1996;Sterner and Pitzer, 1994;Wieser and DeVitre, 2023). For example, at 1150°C for ρ 𝐶𝑂 2 =0.8 g/cm 3 , the relatively simple empirical expression of Sterner and Pitzer (1994) gives 5.008 kbar, while the more complex thermodynamic model of Span and Wanger (1996) gives 4.956 kbar (~3% difference).…”

“…The elevated crustal thickness in OIBs, corresponding to pressures of 3-4 kbar or greater, means that mineral-based barometers with uncertainties of 2-3 kbar can begin to distinguish storage at the Moho vs. the shallow (e.g., . Mineral-based barometers are also aided by the fact that OIB lithosphere can be extremely thick relative to MORB as a result of conductive cooling as the oceanic crust moves away from the ridge (e.g., 45-60 km thick in the Galápagos, Gibson and Geist, 2010;50-110 km thick in the Hawaiian Islands, Li et al, 2004), and many OIB magmas are stored at sub Moho depths (e.g., Barker et al, 2021;DeVitre et al, 2023;, Again, this makes uncertainties of 2-3 kbar less problematic. It is only in volcanic arcs, and particularly continental arcs with thicker crust (Profeta et al, 2016) that mineral-melt barometers can reliably distinguish between storage in the upper, middle and lower crust.…”

“…Relative to melt inclusion analyses and mineral barometry, estimates of magma storage depths from FIs are significantly more precise than both melt inclusions and mineral barometry, and have fewer sources of systematic uncertainty relating to the conversion of the measured quantity (density) to pressure (Dayton et al, 2023). Unlike the complex relationships relating pressure to mineral components (e.g., Jd in Cpx) or dissolved H 2 O and CO 2 concentrations in silicate melts, the CO 2 equation of state is extremely well constrained, with very little offset between different parameterizations (~3% at 7 kbar, Span and Wagner, 1996;Sterner and Pitzer, 1994;Wieser and DeVitre, 2023). Additionally, unlike many mineral-melt barometers which are very temperature sensitive, the P calculated from the EOS is not all that sensitive to T (Fig.…”

“…24a). To demonstrate the high precision of this method, we propagate a ±30 K uncertainty in trapping temperature and a Raman analytical error of ±0.02 g/cm 3 using Monte-Carlo methods (Wieser and DeVitre, 2023), which yields a 12% 1σ error at 0.6 kbar, a 7% error at 2.2 kbar, and a 5% error at 5.2 kbar. These errors are significantly smaller than other petrological barometers (Fig.…”

Reliability of Raman analyses of CO2-rich fluid inclusions as a geobarometer at Kīlauea

“…3; Wieser et al, 2021;Tucker et al, 2019). The uncertainties on CO 2 -dominated fluid inclusion pressures are much smaller as they are only a result of peak fitting, drift corrections, and uncertainty in the temperature of fluid trapping/re-equilibration (Wieser and DeVitre, 2023). These sources of uncertainty were propagated in DiadFit (Wieser and DeVitre, 2013) using Monte Carlo simulations considering 50 K uncertainty on the temperature (see Supplementary Information for details on temperature) and a 1σ uncertainty on density based on peak fit uncertainties of CO 2 spectra as well as the uncertainty in the Ne correction model.…”