Thermal-optical analysis of water insoluble carbon - how to handle interferences caused by mineral dust

Abstract

Introduction Thermal-optical analysis is commonly applied to quantify water insoluble organic carbon and elemental carbon in snow samples. Mineral dust contains Fe-oxides and changes the transmittance of the filter, leading to a systematic bias or even hindering the discrimination of the carbonaceous fractions. Solving this cross-sensitivity and a combined determination of mineral dust and elemental carbon are of great interest, since both components are known as light absorbing snow impurities, reducing the albedo of the snow cover. We present an approach to correct for the change of transmittance caused by Fe-oxides and a method to approximate Fe loadings via the temperature dependent change of the transmittance signal.

Methods Within thermal-optical analysis (TOA) the sample filter is heated up in an inert and subsequently an oxidizing atmosphere and cooled down during the calibration phase. The transmittance of the filter at 660 nm is logged during the analysis, offering an accurate discrimination of organic and elemental carbon. After microwave-assisted digestion, we quantified the Fe content of the samples using inductively coupled plasma-optical emission spectroscopy (ICP-OES) or inductively coupled plasma-mass spectrometry (ICP-MS). We calculated the attenuation caused by mineral dust from the transmittance data of TOA at two temperatures in the calibration phase, correlated it with the Fe loading on the filter and used the signal in the calibration phase to correct the discrimination of organic and elemental carbon.

Results We observed a linear correlation for Fe loadings between 10 and 100 µgFe/cm² and the attenuation in the calibration phase, where all carbonaceous compounds are removed. This allows to estimate Fe loadings without digestion of samples and use of an additional analytical method. The fit computed for snow samples is also applicable for particulate matter samples collected in the same high alpine environment. For samples containing Fe originating from other sources than mineral dust no correlation was observed [1]. We propose a correction of the apportionment of organic carbon and elemental carbon when marked loads of mineral dust are present. This was done using data of the calibration phase of TOA and via a correction utilizing data of a repeated analysis. While the first approach led to an increase of elemental carbon between 27 and 118 % (median 92 %), the second approach led to even higher results. The influence on the more abundant organic carbon was negligible.

Innovative aspects • Determination of Fe loading directly from TOA data • Correction of biased transmittance signal caused by mineral dust • Improved determination of elemental carbon in snow samples containing mineral dust

References [1] Kau, D. et al., Thermal-optical analysis of quartz fiber filters loaded with snow samples – determination of iron based on interferences caused by mineral dust. Atmospheric Measurement Techniques 2022, 15(18), 5207 - 5217.