Single-stage NBI sampler for PM1 mass and five-stage NBI-NMCI sampler for PM1 mass distribution measurements

Abstract

In recent years, concerns about the health and environmental risks associated with PM₁ particles have increased. However, existing PM₁ sampling instruments remain to be improved mainly because their PM₁ inlets have particle loading and bounce issues. To address these challenges, PM₁ inlets based on the Non-Bouncing Impactor (NBI) technique were developed. These inlets employ vacuum oil-wetted glass fiber filter (GFF) substrates to eliminate particle bounce and incorporate a daily vacuum oil injection to prevent particle loading. The 16.7 L/min PM₁ NBI, modified from the PM_2.5 M-WINS design with a reduced diameter of the single nozzle, was designed to adapt readily with current standard sampling and monitoring systems. The cut-size (D_pa50) of 0.99 ± 0.02 μm was determined by considering the effects of Reynold number and the ratio of jet-to-plate distance and nozzle diameter. Field tests comparing the 16.7 L/min PM₁ NBI sampler to the 9-stage NCTU Micro-orifice Cascade Impactor (NMCI₉) revealed small sampling biases, with a mean difference of +0.26 ± 2.28 μg/m³ for PM₁ measurements when silicone oil-coated aluminum foil (AF) and GFF-AF were used in the NMCI₉. The NBI with the oil-wetted GFF substrate effectively removed particles larger than 1.0 μm, resulting in more accurate PM₁ mass concentration measurements. Additionally, the development of a 5-stage NMCI with the 30 L/min PM₁ NBI as the first stage enabled detailed measurement of PM₁ mass distribution in two modes, which was challenging when measuring the entire mass distribution of ambient aerosols. The distribution of size-dependent water-soluble inorganic ions in PM₁ showed dominance of SO₄²⁻ and NH₄⁺ in PM₁ compared to PM_2.5. In summary, the PM₁ NBI enhances accuracy for long-term atmospheric sampling by addressing particle loading and bounce, making it a more reliable standard sampling instrument.