Despite optimisation of the number of synthetic steps and the quantities of TEOS and base involved in the preparation of the NPs, and the improvement observed from Method A to Method B, the colloidal stability of the system is still not sufficient. This is most likely to be due to a poor coverage of the TEOS shell on the d-UPTES cores. This can arise from the formation of pure TEOS NPs with similar sizes as targeted for the ureasil core-shell ones, which will aggregate either by ageing or Ostwald ripening. In order to create a homogeneous TEOS shell, while avoiding the formation of secondary nuclei upon addition of the reactant to the mixtures, a further modification to the synthetic process seems necessary. The formation of secondary TEOS nuclei is a common problem encountered when trying to prepare monodisperse silica NPs on an industrial scale. Good results have been obtained using the so-called “semi-batch” process, in which TEOS or some other silica precursor is added as a continuous stream into the reaction vessel containing the initial nuclei and the sol-gel process initiators,45-47 and reacts progressively, allowing a better coverage and improved size control of the final NPs. This can be performed on a smaller scale (i.e. in a research laboratory) using a syringe pump. This process was simulated by dividing the addition of the TEOS into a series of steps (Fig. 6.3c) and allowing time between each addition. Two samples were prepared: the first with the TEOS solution (5%v/v, in THF) added as aliquots of 30 μL every
30 minutes and the second through addition of TEOS as two aliquots of 60 μL every 60 minutes. The stability of the resultant NPs was investigated for 50 days (Fig. 6.8). As shown in Fig. 6.8, the size and PdI of the NPs measured immediately after their preparation using Method C are very promising, with an average Dhof ~130-155 nm and PdIs below 0.15. Interestingly, these features
are preserved for samples prepared upon addition of the TEOS solution in two steps of 60 μL every 60 minutes even 50 days after the synthesis.
Figure 6.8. Stability studies of core-shell ureasil NPs prepared with Method C and different rates and volumes of TEOS addition. Evolution of (a) hydrodynamic diameter (Dh) and (b) polydispersity (PdI) as a
function of time. The dashed lines serve only to guide the eyes. Results are the average of the values obtained for three samples.
Given the improved stability and the consistency of the values of both the average Dh and
PdI over time, the synthetic procedure adopted for the preparation of this sample was considered as our optimised method. The ζ-potential of samples prepared using the optimised method was measured to be -51.3 ± 3.3 mV, which is indicative of good colloidal stability.48 The NPs were imaged using AFM (Fig. 6.9). Interestingly, both the phase-contrast and the height-contrast AFM images show that the NPs are characterised by the expected core-shell structure. However, the diameter of the NPs imaged by AFM is ~240 nm, which is larger than that measured with DLS (usually less than 200 nm). This result is somewhat surprising considering that, smaller diameters are usually expected for AFM measurements compared to DLS, due to the absence of the solvent layer around the NPs. Although it can be argued that the observed diameter is still included in the size range measured with DLS (56-344 nm), none of the NPs imaged in the AFM seem to present a diameter that is comparable to the average Dh observed for this sample with DLS. This can be due
to several reasons. Firstly, the preparation method of the sample; for AFM imaging of the NPs, the best images were obtained using a dilute solution of the NP stock (1:50 volume dilution in water). This leads to a decrease in the pH of the solution, which might affect the surface charge and thus the stability of the NPs.49 Moreover, while AFM is a quick and relatively inexpensive imaging method, it can be affected by different artefacts when working with NPs. For example, as the sample dries upon deposition, the retreating liquid meniscus can cause the formation of NP
aggregates.49 Klapetek etal. formulated a theoretical model for the investigation of the deviation of the measured and the real sizes of NPs imaged via AFM in non-ideal conditions (i.e. rough surfaces, aggregated NPs and different tip shapes).50 The simulation revealed that a combination of non-ideal measuring conditions and human error can results in an inaccuracy of ± 7 nm for NPs with a true size of 30 nm. Contact with the substrate or with the tip can also cause deformation of the NPs.51 AFM images performed on another sample prepared and measured in the same conditions, showed how the contact of the tip with the core of the NPs leads to a deflection of the NPs centre, while the shell is left intact (Fig. A.6.3). These NPs however still presented a diameter exceeding that measured by DLS (> 200 nm).
Figure 6.9. Tapping-mode AFM images of a solution of ureasil core-shell NPs prepared using method C: top panels - height-contrast images, lower panels - phase contrast images. The expected core-shell structure is clearly visible.