Flow field-flow fractionation (FlFFF) is widely employed for the size separation of macromolecules, including those of biological origins, polymeric and particulate materials [1], [2], [3], [4], [5]. The separation in FlFFF occurs within a thin and unobstructed channel space, utilizing two perpendicular flow streams: migration flow and crossflow [6,7]. FlFFF's unobstructed channel provides an advantage over typical chromatographic systems by avoiding possible sample interactions with the packing materials. The use of aqueous buffer solution as a carrier liquid enhances the biocompatibility of FlFFF, leading to diverse applications including water-soluble polymers, proteins, cells, subcellular species, extracellular vesicles, and virus-like particles [4,[8], [9], [10], [11], [12], [13]].

The original channel design of FlFFF featured a fixed breadth with permeable frits in both depletion and accumulation walls, creating a symmetrical channel [14]. In the symmetrical channel the crossflow rate into the channel via the upper frit ideally equals the crossflow rate out through the membrane and bottom frit. The channel inlet flow rate then equals the channel outlet flow rate. FlFFF gained more popularity with an asymmetrical channel design replacing the depletion wall with an impermeable glass or plastic wall [15]. However, this resulted in decreasing migration flow velocity along the channel axis, resulting in increased retention time and poor recovery. To address this, the asymmetrical channel system adopted a trapezoidal channel design where channel breadth decreases along the channel axis, compensating for the decrease in migration flow rate [16]. While the mean flow velocity in an asymmetrical channel with the trapezoidal design continuously decreases along the channel length, the mean flow velocity in an exponential channel design, where breadth decreases exponentially, can be maintained constant for the entire channel length under certain run condition. The performance of an exponential channel for the separation of submicron particles was found to be similar to or slightly better than that of a trapezoidal one [17]. Attempts to enhance separation efficiency through alteration of channel geometry aimed to miniaturize the dimensions of asymmetrical FlFFF channel by reducing channel length and breadth about a maximum of one-third scale of the conventional channel [18], [19], [20], [21]. Certainly, miniaturization of AF4 channel offered several advantages, including enhanced separation speed and resolution as well as reduced consumption of sample injection amount. Furthermore, reduced channel flow rates in miniaturized AF4 channels enabled direct hyphenation to mass spectrometry for the online characterization of proteins and lipids in biological particles [22,23]. While enhancing separation efficiency in flow FFF can be achieved through modifications in channel geometry, effective separation of particles or macromolecules across a broad size range requires a systematic increase in migration speed. This can be achieved by either decreasing crossflow rate or increasing migration flow velocity through programmed decay of field strength (crossflow rate) or gradual ramping up of the migration flow velocity (outflow rate) [24], [25], [26].

A recent study introduced a thickness-tapered channel [27] (Fig. 1) where channel thickness decreases along the channel axis, enhancing migration flow velocity without the need for programmed decrease of crossflow rate or programmed ramping of outflow rate. The thickness-tapered channel was constructed without using a plastic channel spacer, instead carving the channel into the surface of the depletion wall with a decrease of channel depth from the inlet to the outlet. Initial evaluations showed that the thickness-tapered channel significantly improved the capability to separate a broad diameter range (0.05 ∼ 10 μm) of polystyrene latex standards in steric/hyperlayer mode, especially for long-retaining particles, while increasing speed and elution recovery [27].

In this study, we compare the particle separation performance of the thickness-tapered channel with that of a conventional channel of uniform thickness, ensuring a similar void volume, employing field programming or outflow rate programming. The evaluation involves the separation of polystyrene latex beads in normal mode of FlFFF by varying field strength and migration flow rate, along with an examination of the migration flow velocity profiles along the channel axes based on theory.