Stacking micro-hole filters traps particles and sorts them by size

What is it about?

Air filtration systems are critical for protecting human health, maintaining industrial cleanrooms, and capturing dangerous airborne pollutants. Most commercial air filters consist of random arrangements of fibers, which can be difficult to model, clean, or use for analyzing captured pollutants. In this study, we investigated a novel "sandwiched monolith filter" consisting of thin, highly structured layers made of microscopic channels. By stacking two or three of these patterned filter layers in series with small gaps in between—and systematically decreasing the hole diameters from layer to layer (from 40 µm down to 20 µm)—we designed a system that cleans the air while automatically separating particles based on their size. Using three-dimensional computer simulations, we examined how air and floating microscopic particles move through these stacked stages under two main flow configurations: cross flow (where air sweeps sideways across the filter face) and normal flow (where air flows straight through perpendicular to the filter). We tested particle diameters from 2 to 10 µm across various operating pressure drops. Our modeling showed that cross-flow filtering achieves higher overall trapping efficiency because particles are intercepted on both the front surfaces and inner channel walls. Furthermore, the stacked architecture creates a self-sorting effect: larger particles hit the first front layer, while smaller particles pass through and are caught by downstream layers, leaving the gaps between layers open for particle sampling or easy cleaning.

Featured Image

Photo by Bekky Bekks on Unsplash

Photo by Bekky Bekks on Unsplash

Why is it important?

This research introduces a predictable, micro-engineered alternative to traditional fibrous filters that combines aerosol cleaning with size-selective particle separation. Standard fiber filters are prone to rapid clogging from fine dust and typically require high-pressure compressed air pulses to clean. In contrast, our sandwiched monolith design creates open spaces between filter stages that function as built-in flow channels. This enables gentle side-currents or tiny suction pumps to sweep away captured dust for real-time analysis without taking the filter apart or interrupting its operation. Our quantitative simulation results also revealed unexpected physical behaviors that challenge standard multi-layer filter design assumptions. For cross-flow systems, adding layers significantly boosts the capture of hard-to-trap smaller particles (under 3 µm) without needing higher fan pressures. Surprisingly, under normal-flow conditions, a two-layer filter actually outperforms a three-layer filter of the exact same total thickness. Because fluid velocity stays higher in the two-layer setup, particles retain greater momentum and cannot bend around air streamlines to escape, leading to direct surface impact. These practical design insights allow engineers to optimize bio-aerosol detectors, air purifiers, and environmental monitoring devices for maximum energy efficiency and targeted particle sorting.

Perspectives