High resolution paleothermometry using biogenic silica: A feasibility study

Biogenic silica is produced by numerous organisms that include grasses, protozoa, and Chrysophyceae, as well as marine and freshwater diatoms and sponges (Simpson and Volcani, 1981). It is therefore not uncommon for biogenic silica to be found in environments subject to wildfires. Experimental work on plant-derived opaline phytoliths by Elbaum et al. suggests that the refractive indices (R.I.) of these microfossils increase in response to heating during combustion of their surrounding organic matter (2003). Other work has sought to use phytolith discoloration as an indicator of fire exposure and has had some success (Parr, 2006). However, although innovative, these works have not strongly controlled for temperature, examined other forms of biogenic silica, or probed other parameters. Better controlled work by Sandford (2003) established that siliceous marine demosponge spicules undergo changes to their infrared absorbance spectra upon heating (Sandford, 2003), and dehydration work by him and Jensen et al., offer encouraging indications that comprehensive alterations occur to siliceous sponge spicules heated (Jensen et al., 2009). Based upon these foundations, the goal of this research was to determine whether or not other forms of biogenic silica experience optical, chemical, crystalline or fluorescence changes upon heating. It is hypothesized that if consistent and stable alterations are observed, these results could be used to examine the severity and spatial and temporal variability of wild fires over time and perhaps even in response to changes in climate (Justino et al., 2010). To accomplish this, weighed freshwater peats from two well categorized environments in the Okefenokee Swamp (Cohen et al., 1991), a large freshwater wetland in Southeastern Georgia and Northeastern Florida, were heated in a muffle furnace at predetermined temperatures between 54°C and 1100°C for one hour. Microscopic sponge spicules and diatom frustules within the samples were inspected via refractive index, Scanning Electron Microscope (SEM), Energy-Dispersive X-ray Spectroscopy (EDX/EDS), Ultraviolet-A/blue light excited fluorescence photography, and petrographic microscope, and bulk ash was examined via spectrofluorimetry. Anomolies in the EDX spectra were later modeled using both a Monte Carlo electron transport model and an application of the relativistic equivalent of the binary-encounter Bethe equations to predict x-ray emission in a carbon/silica bi-layered sample, when off beam-axis primary electron scattering is not considered. Sponge spicule refractive indices in both groups vary significantly with treatment temperature. Additionally, upon submersion in double deionized water for approximately 48 hours–5 days, effects in one group begin to reverse. The fluorescence data also reflects a temperature dependency when based in a red/green/blue color space. (R/B), (R/G), and (G/B) ratios increase with heating, while (R-G)/(G-B) increase until 600°C–700°C, above which there is a decline. Investigations of silicon and oxygen K-&agr; characteristic X-ray peak heights with EDX suggest that there is a tendency for the silicon and oxygen emission to increase between room temperature and 450°C for the spicules and 700°C for the diatoms and then to decline significantly thereafter, though the spacing between measurements is coarser for the diatoms. One possible explanation for the aforementioned temperature dependency is that surficial carbon enhances Si and O K-shell X-ray emission until it is oxidized by progressively higher temperatures. The results from a bilayered electron transport model suggest that one variation, electron "moderation" by carbon, is not the mechanism behind the observed temperature dependency. A second more geometrically realistic Monte Carlo simulation of 100,000 electrons and 7 separate carbon thicknesses, meanwhile, failed to demonstrate that carbon thicknesses can enhance characteristic X-ray emissions by silica, even after taking into account electron scattering within the silica and carbon. It is therefore hypothesized that the experimental results might be indirectly related to charge accumulations on the samples stemming from the destruction of surficial organic matter, or actual morphological changes occurring to the silica at length-scales too small to probe with an SEM. Electron trajectory variations due to scattering off trace gas molecules in the microscope chamber, or off of Si, O, or C atoms within the sample, are a possible factor as well