A novel measurement system for unattended, in situ characterization of carbonaceous aerosols

Abstract

Carbonaceous aerosol is a relevant constituent of the atmosphere in terms of climate and health impacts. Nevertheless, measuring this component poses many challenges. There is currently no simple and sensitive commercial technique that can reliably capture its totality in an unattended manner, with minimal user intervention, for extended periods of time. To address this issue we have developed the fast thermal carbon totalizator (FATCAT). Our system captures an aerosol sample on a rigid metallic filter and subsequently analyses it by rapidly heating the filter directly, through induction, to a temperature around 800 ∘C. The carbon in the filter is oxidized and quantified as CO2 in order to establish the total carbon (TC) content of the sample. The metallic filter is robust, which solves filter displacement or leakage problems, and does not require a frequent replacement like other measurement techniques. The limit of detection of our system using the 3σ criterion is TC =0.19 µg-C (micrograms of carbon). This translates to an average ambient concentration of TC =0.32 µg-C m−3 and TC =0.16 µg-C m−3 for sampling interval of 1 or 2 h respectively using a sampling flow rate of 10 L min−1. We present a series of measurements using a controlled, well-defined propane flame aerosol as well as wood-burning emissions using two different wood-burning stoves. Furthermore, we complement these measurements by coating the particles with secondary organic matter by means of an oxidation flow reactor. Our device shows a good correlation (correlation coefficient, R2>0.99) with well-established techniques, like mass measurements by means of a tapered element oscillating microbalance and TC measurements by means of thermal–optical transmittance analysis. Furthermore, the homogeneous fast-heating of the filter produces fast thermograms. This is a new feature that, to our knowledge, is exclusive of our system. The fast thermograms contain information regarding the volatility and refractoriness of the sample without imposing an artificial fraction separation like other measurement methods. Different aerosol components, like wood-burning emissions, soot from the propane flame and secondary organic matter, create diverse identifiable patterns.