Assessment of purification and contamination correction methods for analysing the oxygen isotope composition from biogenic silica

作者：Chapligin B^*; Meyer H; Bryan A; Snyder J; Kemnitz H

来源：Chemical Geology, 2012, 300: 185-199.

DOI：10.1016/j.chemgeo.2012.01.004

• ICP-OES
• XRF
• EDS
• Diatoms
• Species-effect
• El%26apos
• gygytgyn

Mineral particles and other contaminants cause a serious and undesirable shift in delta O-18 values of biogenic silica and should be removed prior to isotope analysis. This problem is particularly significant in the fraction containing clay particles and therefore the fraction %26lt;10 mu m is generally avoided. Here, we present an updated preparation protocol for the %26lt;10 mu m fraction based on known methods, developed using test material from short cores of Lake El%26apos;gygytgyn, NE Russia. New contamination assessment and correction techniques are discussed. Best results (low initial weight/high precision and throughput rate) for the %26lt;10 mu m fraction were achieved using geochemical mass-balancing and determining the %26quot;percentage of contamination%26quot; by analysing the sample%26apos;s and end-member%26apos;s Al2O3 with improved Inductively Coupled Plasma Optical Emission Spectroscopy (ICP-OES) and Energy-Dispersive X-ray Spectroscopy (EDS) protocols or by determining the factor for calculating the percentage of contamination from the Al2O3 percentage of clay assemblages. Despite using small initial weights for ICP-OES (10 mg) and EDS analysis (%26lt;0.5 mg) a sufficiently high precision (for low Al2O3 percentages between 0.1 and 1.5%) could be achieved. %26lt;br%26gt;For the %26gt;10 mu m fraction, point-counting under the light microscope was the preferred option. The different size-fractions from Lake El%26apos;gygytgyn samples show a distinct and significantly different species composition (average counts: %26lt;10 mu m: %26gt;90% Cyclotella ocellata complex: %26gt;10 mu m: %26gt;60% Pliocaenicus seczkinae). Differences in the oxygen isotope composition of these different size fractions could be explained by a species-effect on the delta O-18 values. However, this species-effect disappeared after the correction (avg delta O-18 %26lt;10 pm: 19.8 +/- 0.6%.; %26gt;10 mu m: 19.8 +/- 0.4 parts per thousand). By comparing these delta O-18 values with additional recent water samples, an average 1000 In alpha of 39.6 parts per thousand for Lake El%26apos;gygytgyn sub-recent diatoms was calculated. This %26quot;tool-kit%26quot; provides support for the presented improved contamination assessment and corrections and allows utilising smaller initial weight (app. 10 mg) and small fractions (%26lt;10 mu m) for gaining corrected, higher resolution delta O-18 records.

• 出版日期2012-3-18