Nanoparticle size distribution quantification from transmission electron microscopy (TEM) of ruthenium tetroxide stained polymeric nanoparticles

Hypothesis: Dynamic Light Scattering (DLS) generated particle size distributions (PSD) of polymer-stabilized nanoparticles are dependent on the optimization parameters used to generate an inversion solution fit to the measured autocorrelation function. The accuracy of the DLS PSD average and polydispersity can be determined by comparing analyzed Transmission Electron Microscopy (TEM) images with the DLS results if the TEM measured sizes can be corrected for the thickness of the hydrated polymer corona that impacts particle hydrodynamics but is a collapsed, desiccated shell in the TEM images. Experiments: Nanoparticles were prepared by Flash NanoPrecipitation with either poly(ethylene glycol) (PEG) or hydroxypropyl methylcellulose acetate succinate (HPMCAS) stabilizing polymers. Solvated nanoparticle size distributions were measured by DLS in aqueous media. The same nanoparticle dispersions were lyophilized onto TEM grids and stained by ruthenium tetroxide (RuO₄) vapor to improve electron contrast. Desiccated particle size distributions were generated by measuring a minimum of 300 particle diameters in the stained TEM images. Findings: Using our protocol for staining soft matter nanoparticles in TEM measurements, we have quantitatively analyzed the correlation between DLS and TEM generated PSDs. Average diameters disagree by the hydrated polymer corona thickness for each stabilizer due to the high-vacuum TEM environment, with 21.4 nm for PEG and 51.2 nm for HPMCAS. While corrected average diameter agrees within 10% for each technique, DLS consistently over-estimates the standard deviation of the PSD by 100% compared to the TEM measurement.