Colloidal Mechanisms of Gold Nanoparticle Loss in Asymmetric Flow Field-Flow Fractionation

Flow field-flow fractionation is a powerful method for the analysis of nanoparticle size distributions, but its widespread use has been hampered by large analyte losses, especially of metal nanoparticles. Here, we report on the colloidal mechanisms underlying the losses. We systematically studied gold nanoparticles (AuNPs) during asymmetrical flow field-flow fractionation (AF4) by systematic variation of the particle properties and the eluent composition. Recoveries of AuNPs (core diameter 12 nm) stabilized by citrate or polyethylene glycol (PEG) at different ionic strengths were determined. We used online UV-vis detection and off-line elementary analysis to follow particle losses during full analysis runs, runs without cross-flow, and runs with parts of the instrument bypassed. The combination allowed us to calculate relative and absolute analyte losses at different stages of the analytic protocol. We found different loss mechanisms depending on the ligand. Citrate-stabilized particles degraded during analysis and suffered large losses (up to 74%). PEG-stabilized particles had smaller relative losses at moderate ionic strengths (1-20%) that depended on PEG length. Long PEGs at higher ionic strengths (≥5 mM) caused particle loss due to bridging adsorption at the membrane. Bulk agglomeration was not a relevant loss mechanism at low ionic strengths ≤5 mM for any of the studied particles. An unexpectedly large fraction of particles was lost at tubing and other internal surfaces. We propose that the colloidal mechanisms observed here are relevant loss mechanisms in many particle analysis protocols and discuss strategies to avoid them.