Ultrafine particle number concentrations and size distributions were measured on the platform of a metro station in Athens, Greece, and compared with those recorded at an urban background station. The volatility of the sampled particles was measured in parallel, providing further insights on the mixing state and composition of the sampled particles. Particle concentration exhibited a mean value of 1.2 × 10⁴ # cm⁻³ and showed a weak correlation with train passage frequency, but exhibited a strong correlation with urban background particle concentrations. The size distribution appears to be strongly influenced by outdoor conditions, such as the morning traffic rush hour and new particle formation events observed at noon. The aerosol in the metro was externally mixed throughout the day, with particle populations being identified (1) as fully refractory particles being more dominant during the morning traffic rush hours, (2) as core-shell structure particles having a non-volatile core coated with volatile material, and (3) fully volatile particles. The evolution of particle volatility and size throughout the day provide additional support that most nanoparticles in the metro station originate from outdoor urban air. Ultrafine particles concentration in the metro were within the range of those in urban air, and are originated mainly from outdoors, even though the metro is naturally ventilated.

Volatility Tandem Differential Mobility Analysers (VTDMAs) are widely used for determining the volatile and refractory fractions and thus the mixing state of aerosols particles. A three-channel VTDMA consisting of two thermodenuders (TDs) with distinct designs (i.e., the NanoTD, having a straight tube design, and a coiled TD; cTD) and a by-pass line was built and fully characterized. Both TDs were tested using laboratory-generated aerosol particles (single compound and core-shell particles) as well as atmospheric aerosols observed at an urban background station. The NanoTD exhibited high particle penetration efficiency and negligible thermophoretic losses, making it advantageous for ultrafine particle analysis, especially in environments with low particle concentration. The cTD allows longer particle residence time for the same flow rate, resulting in higher particle volatilization in some cases. Higher particle losses in this TD, both thermophoretic and diffusional, pose a limitation when dealing with low particle concentrations.The difference in the performance between the thermodenuders was only noticed at intermediate temperatures, at which particle volume loss becomes more pronounced. These temperatures vary among aerosols, since the volatilization rate depends on the chemical complexity and size of the particles sampled. Differences in the aerosol volume fraction remaining after heating with the two TD designs exhibited a maximum of 20% for single-compound particles and 12% for urban background aerosols. Measurements using core-shell particles yielded differences of up to 21% in particle volatilization, independently of particle size, when comparing the system using either of the two TD designs. Similar results were obtained with the two TD designs at higher operating temperatures (e.g., 230. °C), indicating that at this temperature most of the material on the particles was evaporated.