**1. Introduction**

In the last decade metal oxide semiconducting nanoparticles (NPs) have been receiving grea<sup>t</sup> interest in the field of biological applications due to their intriguing optical properties, low toxicity, good biocompatibility and their low cost [1]. Among them, zinc oxide (ZnO) nanoparticles have shown to be particularly promising due to their peculiar chemical and physical properties, which can be specifically tailored on the basis of the particles' size and shape [2,3]. As an example, the wide bandgap typical of ZnO (about 3.37 eV at room temperature, RT) entails a fluorescence excitation situated in the

ultraviolet (UV) region [4], which allows ZnO to be successfully employed for optical cell imaging. More in general, it has been demonstrated how ZnO could be a promising material for therapeutic and diagnostic applications [2], showing high levels of drugs loading and a quite easy control over the following release [5,6]. In this context, the optical, targeting and drug delivery properties of ZnO can be more specifically addressed by the combination of various synthetic procedures (sol-gel, sputtering, hydro-solvothermal, etc.) [7] and ZnO morphologies (nanowires, nanorods, nanobelts, desert roses and spherical nanoparticles) [8,9], together with surface functionalization approaches [2,10]. To this purpose, it is important to observe that crystal density, morphology and defects could be critical factors in determining the material properties, and as a consequence, the final applications. Although ZnO nanostructures featuring dimensions lower than 100 nm are nowadays considered the most promising for biomedicine [11–13], the most part of the literature generally neglects the influence of the synthesis processes and parameters on the characteristics of the resulting NPs and, finally, on the reproducibility of their biological response. Actually, this is an essential point for further strengthening the biological application of crystalline nanoparticles. In this regard, it is worth mentioning that many articles deal with the use of commercial particles (with the limit about their morphologies and size distribution) or did not report any specific detail about important synthesis aspects like the synthesis precursors, the surface chemistry, and sometimes the hydrodynamic size and z-potential [14,15].

In this work, we report a novel synthetic approach of ZnO nanocrystals (NCs) based on a microwave-assisted solvothermal process that allows us to reach a greater control over the morphology and dimensional dispersion of the NPs. At the same time this new method guarantees a higher reproducibility level of experimental data with respect to those obtained by using ZnO NCs synthesized with a more conventional wet approach. The microwave-assisted synthesis also presents additional advantages, like shorter reaction time and lower energy consumption. Actually, one of the main characteristics of this technique is to guarantee a uniform heating of the precursors and to present outstanding reaction rates, moving from several hours to a few minutes [16–18]. Furthermore, this synthetic method assures high reliability and high reaction yields. This explains the growing popularity and diffusion of microwave-assisted synthetic approaches over a wide range of different nanomaterials synthesis [19,20], but to our knowledge very few reports are dedicated to the preparation of ZnO-based nanomaterials [21–24]. Indeed, thanks to this synthetic route we obtain a simultaneous nucleation of nanocrystals that leads to a uniform and reproducible dimensional dispersion of ZnO (about 20 nm in diameter), with outstanding colloidal stability both in ethanol and in water. In this way, we demonstrate how the synthetic route should be an important factor to be considered for the achievement of reproducible and reliable results, also when evaluating the cell viability when in contact with these nanocrystals.

### **2. Materials and Methods**

### *2.1. Synthesis, Functionalization and Labelling of ZnO Nanocrystals.*

All the chemicals were used as purchased without further purification. ZnO nanocrystals were synthesized through two different synthetic routes: a traditional solvothermal way (sample named ZnO-st) and a microwave-assisted synthesis (sample named ZnO-mw). As zinc precursors we chose zinc acetate di-hydrate (Zn(CH3COO)2·2H2O Puriss. p.a., ACS Reagent, ≥99.0% Fluka) and a hydroxide as mineralizing agen<sup>t</sup> (Sigma-Aldrich) both dissolved in methanol (Reag. Ph Eur Grade VWR Chemicals). The reaction path, in both cases, is based on the hydrolysis of the zinc precursor due to the presence of the hydroxide, as shown in the following reactions scheme:

$$\text{KOH} \rightarrow \text{K} + \text{(aq)} + \text{OH} \text{--(aq)}\tag{1}$$

$$\text{Zn(Ac)}\_2 \rightarrow \text{Zn}^{2+}\text{ (aq)} + 2\text{Ac}^-\text{ (aq)}\tag{2}$$

$$2\text{Zn}^{2+} \text{ (aq)} + 2\text{OH}^- \text{ (aq)} \rightarrow 2\text{n(OH)}\_2 \text{ (aq)}\tag{3}$$

$$\text{Zn(OH)}\_{2}\text{ (aq)} + 2\text{H}\_{2}\text{O} \rightarrow \text{Zn(OH)}\_{4}\text{4}^{2-} + 2\text{H}^{+}\text{ (aq)}\tag{4}$$

$$\text{Zn(OH)}\_{4}^{2-} + 2\text{H}^{+} \text{ (aq)} \rightarrow \text{ZnO (s)} + 3\text{H}\_{2}\text{O} \tag{5}$$

The microwave-assisted synthesis was carried out as follows. A solution containing the zinc precursor in methanol (0.09 M, 60 mL) was prepared and stirred directly in the Teflon reactor vessel. In order to better initiate the zinc oxide nucleation, 480 μL of double-distilled water was added and then the potassium hydroxide solution (0.2 M, 35 mL) (KOH ≥ 85% pellets, Sigma-Aldrich) was mixed together in a 270 mL Teflon reactor vessel, equipped with pressure and temperature probes, connected with the microwave furnace (Milestone START-Synth, Milestone Inc, Shelton, Connecticut). The resulting solution was put into a microwave oven for 30 min at 60 ◦C. After the completion of the reaction, the solution was cooled down to room temperature and followed by two washing steps to change the reaction solvent and to remove any unreacted compound. To do that, the colloidal solution was collected and centrifuged for 10 min at 3500 g (Mega Star 600R, VWR), the supernatant was then removed, and the precipitate was dispersed and washed twice in 15 mL of ethanol (Sigma-Aldrich, 99%). The as-obtained ZnO NCs pellet was suspended through sonication (LABSONIC LBS2, FALC Instruments SRL) in fresh ethanol to give the final colloidal suspension.

The traditional solvothermal synthesis process was carried out as already reported [25] with the same zinc precursor at the same concentrations in a round-bottom glass flask (100 mL). In detail, zinc acetate di-hydrate (0.09 M) was directly dissolved in the reaction flask with methanol (42 mL) and heated under continuous stirring in reflux conditions since the temperature of 60 ◦C was reached and the double-distilled water was added (318 μL). The methanol solution of sodium hydroxide (0.31 M, 23 mL) (NaOH BioXtra, ≥98% acidimetric, pellets anhydrous, Sigma-Aldrich) was then added dropwise to the zinc acetate solution (in about 20 min). The reaction conditions were maintained for 2.5 h and after this time the as obtained suspension was cooled to RT. ZnO NCs (named ZnO-st) were collected and washed with fresh ethanol (Sigma-Aldrich, 99%) as previously reported for the microwave-assisted synthesis. Reaction yields were evaluated for both the synthetic procedures by weighing the dried NCs from a known volume of the obtained colloidal solutions.

Both the typologies of as synthesized ZnO NCs were functionalized in order to proceed with the in vitro cell culture studies, according to a previously reported method [25,26]. In particular, the ZnO NCs surface was decorated with the amino-propyl group (ZnO-NH2 NCs). Approximately 50 mg of ZnO NCs, dispersed in 20 mL of ethanol (Sigma-Aldrich 99%), were heated to 80 ◦C in a 25 mL round glass flask under continuous stirring and nitrogen gas flow. A 10 mol% ratio of 3-aminopropyltrimethoxysilane (H2N(CH2)3Si(OCH3)3 APTMS 97%, Sigma Aldrich, 10 μL), with respect to total ZnO amount, was added to the NCs suspension and the reaction was carried out for 6 h. The excess of unreacted APTMS was then removed by washing twice the ZnO NCs with fresh ethanol, separating them from the reaction medium by centrifugation (10 min, 10,000 g).

Only for the internalization of nanocrystals into cancer cells, the ZnO-NH2 NCs were coupled with ATTO633-NHS ester dyes (Thermofischer), by adding 2 μg of dye each mg of NCs in ethanol suspension. The as obtained solution was dark-stirred overnight and then washed twice by centrifuging (10 min, 10,000 g) and resuspending the pellet in fresh ethanol to remove unbounded dye molecules [25]. To minimize the effects of particles aggregation and sedimentation under biological tests, the suspension of dye-labelled nanocrystals was always freshly prepared and shortly sonicated (10 min) before each experiment.
