Nanoparticle corona artefacts derived from specimen preparation of particle suspensions

Progress in the implementation of nanoparticles for therapeutic applications will accelerate with an improved understanding of the interface between nanoparticle surfaces and the media they are dispersed in. We examine this interface by analytical scanning transmission electron microscopy and show that incorrect specimen preparation or analysis can induce an artefactual, nanoscale, calcium phosphate-rich, amorphous coating on nanoparticles dispersed in cell culture media. We report that this ionic coating can be induced on five different types of nanoparticles (Au, BaTiO3, ZnO, TiO2 and Fe2O3) when specimen preparation causes a significant rise in pH above physiological levels. Such a pH change reduces ionic solubility in the suspending media to permit precipitation of calcium phosphate. Finally, we demonstrate that there is no indication of a calcium-phosphorus-rich coating on BaTiO3 nanoparticles suspended in culture media when prepared without alteration of the pH of the suspending media and imaged by cryo-STEM. Therefore we recommend that future reports utilising nanoparticles dispersed in cell culture media monitor and report the pH of suspensions during sample preparation.


Nanoparticle synthesis
BaTiO 3 nanoparticles were synthesised at 150 o C for 72 h via a hydrothermal method with particles of diameter 130 nm obtained. [1,2] Similarly ZnO nanoparticles were synthesised via a hydrothermal method at 120 o C for 12 h. [3] Three different sized nanoparticles were synthesised by altering the pH of the reaction solution using NaOH pellets (Reidel-de Haen, Lot# 60100). [4] The pH of the precursor solution was altered as follows: ZnO-A 9.43 (desired 9), ZnO-B 11.08 (desired 11) and ZnO-C pH 12.9 (desired 13). The three sizes measured from the longest length were ZnO-A 840 nm, ZnO-B 400 nm and ZnO-C 150 nm. Fe 2 O 3 nanoparticles <50 nm were purchased from Sigma Aldrich (Lot# 40095/1 22/00). Gold nanoparticles coated with polystyrene sulfate were synthesised via the methods described in [5,6] and were kindly donated by Lucien Roach of the Molecular and Nanoscale Physics group, School of Physics and Astronomy, University of Leeds.

Transmission Electron Microscopy
Transmission electron microscopy (TEM) was undertaken using two microscopes; an FEI Tecnai F20 operated at 200 kV and fitted with a Gatan Orius CCD camera and Oxford instruments 80 mm 2 X-max EDX detector and an FEI Titan 3 Themis G2 operating at 300 kV fitted with 4 EDX silicon drift detectors, a Gatan One-View CCD and a Gatan Quantum 965 ER imaging filter. grid, blotted and rapidly plunged into liquid ethane. The grid was transferred to the microscope using a Gatan 914 TEM cryo-holder and maintained at a temperature of <-165 o C.
For EDX spectroscopy, Oxford instruments Aztec software and Bruker Espirit software was used.
Dual-EELS analysis was carried out using collection and convergence semi -angles of 21 and 8.2 mrad respectively. For all cryo-analytical STEM, a probe current between 40 and 100 pA was used and for EDX a dwell time of 23 µs. For cryo-EELS, to prevent destruction of the vitreous ice during acquisition the electron beam was continually scanned across the image area (5 µs dwell time) and the EEL spectra obtained for the whole area. 7 In all instances the total accumulated fluence used was below the threshold for significant compositional change of hydroxyapatite reported as 100 x 10 6 e -/nm 2 by Eddisford et al. [8] For Ca and P semi-quantification for EDX the K α X-ray lines were used for both Ca and P at 3.69 and 2.013 keV respectively and for EELS the L 2,3 edges were used for both Ca and P. Reference standards of tribasic calcium phosphate (Fluka Chemika) and hydroxyapatite (Sigma Aldrich) were used to ensure accurate quantification and the results are shown in table S1.

pH and temperature change
pH and temperature measurements were taken of cell culture media either bath sonicated, incubated in a water bath at 40 o C or incubated at room temperature ( Figure S1). To confirm that the pH and temperature change were the critical factors in the formation of a calcium phosphate coating rather than incubation time, BaTiO 3 nanoparticles were dispersed in cell culture media for 3 h at room temperature. Figure S2 shows there was no evidence of a calcium phosphate coating around the nanoparticles.

Sonication time series
The calcium-phosphorus-rich coating was monitored after progressive sonication times indicating that the coating establishes after just 30-60 minutes of sonication ( Figure S3).

Nanoparticle size
Three ZnO nanoparticles of average diameter 840 nm (ZnO-A), 400 nm (ZnO-B) and 150 nm (ZnO-C) were dispersed in CCCM at an equal weight per volume of 100 µg/mL. Following extensive bath sonication, a coating was evident on all three particles, which was confirmed as Ca and P rich through EDX analysis (data not shown). No significant different between the thickness of the three coatings was observed with average calcium phosphate coating thicknesses of 65 (±17) nm, 49 (±8) nm and 46 (±13) nm for ZnO-A, ZnO-B and ZnO-C respectively ( Figure S4).

Figure S4: TEM images of ZnO nanoparticles of varying size and shape dispersed in CCCM prepared
via drop casting. Nanoparticle primary particle size increased with decreasing precursor pH during

synthesis. This is confirmed in the DLS plot (D) where there is an increase in size between ZnO-A (blue), ZnO-B (red) and ZnO-C (black). All three dispersions showed a calcium phosphate coating
could be formed around the nanoparticles independent of size and/or shape.

Cryo-analytical STEM
Confirmation that using in situ cryogenic techniques eradicates drying artefacts (previously described in Ilett et al. [9] ) was demonstrated by cryo-analytical STEM of ZnO and Fe 2 O 3 nanoparticles dispersed in CCCM by sonication ( Figure S5) and BaTiO 3 ( Figure S6).

Ca/P ratio quantification
The compositions and in particular the concentrations of Ca 2+ and PO 4 3vary significantly between DMEM and RPMI with each having Ca 2+ and PO 4 3concentrations of 1.8 mM and 0.92 mM (DMEM) and 0.42 mM and 5.63 mM (RPMI) respectively. [10 ] However a calcium phosphate coating was observed for BaTiO 3 and BaTiO 3 -PLL nanoparticles dispersed in both media after >1 h bath sonication and, although there was a suggestion from EDX relative semi-quantification that there may be a reduced Ca/P ratio when using RPMI, this was not significant when standard deviations were taken into account (Table S3). Since the calcium phosphate coating established around nanoparticles suspended in either media this would suggest that it is changes in solubility that is the key factor in coating formation rather than the absolute levels of calcium and phosphorus present in solution.