Heat Capacity Behavior of Andradite:

A Multi-Sample and -Methodological Investigation

Charles A. Geiger, Edgar Dachs, Noreen Vielreicher
Department of Chemistry and Physics of Materials
University of Salzburg
Jakob Haringer Strasse 2a
A-5020 Salzburg, Austria

George R. Rossman

Division of Geological and Planetary Sciences
California Institute of Technology,
Pasadena, CA  91125-2500, USA



Andradite, ideal end-member formula Ca3Fe2Si3O12, is one of the common rock-forming silicate garnets found in Earth’s crust. Its P-T stability and thermodynamic properties have been investigated several times using phase-equilibrium results. The heat-capacity behavior, Cp, of a synthetic sample has been studied via calorimetry including a low temperature adiabatic investigation. In addition, various physical properties of andradite have been studied by diffraction and different spectroscopic methods. There are still, however, outstanding questions regarding andradite’s precise thermodynamic behavior. Three issues are: i) Could there be differences in the thermodynamic properties, namely heat capacity, Cp, between synthetic and natural andradite crystals, as observed in the Ca-garnet grossular, Ca3Al2Si3O12? ii) What is the precise thermal nature of the low-temperature magnetic-phase-transition behavior of andradite? and iii) How quantitative are the first and older published calorimetric (i.e., adiabatic and DSC) heat capacity results? In this work, four natural, nearly end-member single crystals and two synthetic polycrystalline andradite samples were carefully characterized, depending on the sample, by optical microscope examination, X-ray powder diffraction, microprobe analysis, and IR and UV/VIS single-crystal spectroscopy. The IR spectra of the different samples often show a main, intense OH- stretching band located at 3563 cm-1, but other OH- bands can sometimes be observed as well. Structural OH- concentrations, calculated from the IR spectra, vary from about 0.006 to 0.240 wt % H2O using a calibration based on grossular. UV/VIS spectra indicate that there can be slight, but not fully understood, differences in the electronic state, probably involving Fe, between synthetic versus natural andradite crystals. Cp behavior, employing the same andradite samples that were used for the other measurements, was determined by relaxation micro-calorimetry between 2 and 300 K and by DSC methods between 150/300 and 700/950 K. The low-temperature Cp results show a magnetic phase transition with a Néel temperature of 11.3 ± 0.2 K, which could be slightly affected by the precise electronic state of the crystals. The published adiabatic calorimetry results on andradite made down to 8 K (Robie et al., 1987) do not provide a full and correct thermal description of this magnetic transition, because it extends to even lower temperatures. The calorimetry Cp measurements for the different samples give a best estimate for the standard third-law entropy at 298.15 K for andradite of So ≈ 324 ± 2 J/mol∙K vs the value of 316.4 ± 2.0 J/mol∙K, as determined from the adiabatic investigation (Robie et al., 1987). A synthetic sample, which may best represent end-member andradite gives So = 324 ± 2.3 J/mol∙K. The published adiabatic data are generally slightly higher in value. Both natural and synthetic crystals give similar So values within experimental uncertainty of about 1.0 %, but one natural andradite, richer in OH, may have a very slightly higher value around So » 326 J/mol∙K. Low temperature DSC measurements made below 298 K agree excellently with those from relaxation calorimetry. The DSC measurements at 298 K show a similarity in Cp behavior among natural and synthetic andradites. A Cp polynomial for use between 300 to about 1000 K was calculated from the data on synthetic andradite (SD23) giving: 

Cp (J/mol∙K) = 636.09(±2) - 3963.1 (±63)·T-0.5 - 7.5883(±0.22)·107·T-2  + 725.67(±023.36)·109·T-3.