SURFACE TAILORING OF OXIDE-BASED NANOSYSTEMS FOR THE DESIGN OF ADVANCED COMPOSITE MATERIALS AND SMART DEVICES

Abstract My Ph.D. project focused on the tailoring of the surface properties of oxide substrates for the preparation of advanced composite materials and devices. The initial focus of my research activity was the surface modification of halloysite nanotubes (HNT), a natural material with unique structural features. I then extended my investigation to other oxides (titanium dioxide, TiO2, and Superparamagnetic Iron Oxide Nanoparticles, SPION), applying the surface modification approaches developed for HNT substrates. The resulting oxide-based hybrid systems showed promising properties as stimuli-responsive devices for health and environmental applications, and as fillers for polymeric nanocomposites with enhanced durability. The main results obtained for each oxide material will be presented in the following sections. 1. Halloysite nanotubes (HNT) Halloysite is a polymorph of kaolinite which naturally wraps itself to form tubular structures (Figure 1A). It is one of the few nanotubular systems presenting an inner lumen and an outer surface characterized by different surface charge and structural composition: the internal surface exposes aluminium hydroxyl groups, while the outermost layer is silica[1]. Among its many potential applications, its tubular dual structure has sparked interest in the field of nanomedicine and polymer nanocomposites[2]. However, only few reports investigated the possibility of a selective functionalization of the inner and outer HNT surfaces. During my first year of PhD, I investigated the selective functionalization of HNT with phosphonates, hetero-organic compounds bearing a C PO(OH)2 group, as potential site-specific linkers for the grafting of active molecules. I took advantage of the phosphonic acid selectivity towards certain oxides (especially aluminium oxides) to achieve surface-specific functionalization of HNT. An in-depth comprehension of such hybrid systems is no trivial matter, as the inner and outer HNT surfaces possess different accessibility and reactivity but are not separable. For this reason, beside HNT, I used purposely prepared model oxides, mimicking the inner and outer nanotube surfaces, to better study the actual selectivity towards SiO2 and AlOOH exhibited by the phosphonate moiety. Octylphosphonic acid (OPA) was chosen as functionalizing agent, as its alkyl chain allowed me to monitor the surface modification through changes in the water contact angle (θ). I found that the oxide isoelectric point (pHIEP) plays a major role in determining a stable OPA adsorption: while AlOOH showed good reactivity towards OPA, SiO2, which is negatively charged at the impregnation pH (pH 4), did not react with the phosphonate heads. The functionalization reversibility was also assessed: Samples showed OPA release at pH values more alkaline than the oxide isoelectric point, when the surface charge is negative. Overall, these results support an OPA-oxide bond governed both by electrostatic and covalent forces. The selective functionalization of HNT inner lumen was also demonstrated via a combination of characterization techniques, including FTIR spectroscopy, ζ-potential measurement and water dispersibility assay. Another scantly investigated topic regarding HNT nanosystems is their covalent modification. Covalent grafting allows for a superior control over the release kinetics of active principles, as the initial release burst observed in the case of electrostatically bound compounds is greatly reduced. In this respect, I studied the covalent attachment of a biologically active molecule to HNT via an imine bond. This type of covalent bond was chosen to open up the possibility for a controlled release of the bioactive molecule activated by pH changes. To this purpose, HNT was functionalized with (3 aminopropyl)triethoxysilane (APTES), a bifunctional linker which can cover the oxide surface with amino groups. Despite being non-selective for either of the two phases, it was chosen based on the relative ease with which it can be grafted to the substrate. Then, tetrathia[7]helicene aldehyde (7-THA) was bound to HNT via APTES, exploiting imine chemistry. 7-THA belongs to a group of polyaromatic molecules capable, thanks to their peculiar helicoidal shape, of intercalating DNA strands, which is at the basis of anti-sense therapy[3]. Nonetheless, HNT abysmal water solubility highly reduces their bioavailability and makes the use of hydrophilic nanocarriers necessary. I tested the relative release efficiency of 7-THA under slightly different pH conditions, representative of lysosomal, tumoral and healthy extracellular pH values. The resulting HNT-7TH system was studied in detail via XPS and angle-resolved near edge X-ray absorption fine structure spectroscopy (NEXAFS) at the Material Science Beamline of the Elettra synchrotron facility. The latter characterization technique required the synthesis and functionalization of oxide films replicating HNT surfaces, in order to probe the orientation of a population of adsorbed molecules with respect to the surface plane. NEXAFS results suggested a preferential orientation of the aromatic rings of helicene normal to the oxide surface, possibly as a result of π-π stacking interactions (Figure 1B). XPS analyses were performed at each stage of the preparation process of HNT-7TH and after the release test at pH 5 (Figure 1C): the observed change in the elemental composition indicated a release of 70% of the loaded 7-THA as a result of pH change. Given the interest for tumour therapy applications, in vitro tests were carried out to assess the selectivity of this system on cancer cells. The effect of HNT-7TH on the viability of bladder cancer cells was tested by Dr Riccardo Vago at the IRCCS San Raffaele Scientific Institute. Two different cell lines, named 5637 and HT-1376, with an extracellular pH of 7.2 and 6.8 respectively, were subjected to increasing concentrations of HNT 7TH and bare HNT as control. HNT 7TH was found to cause a more marked reduction in cell viability on the HT-1376 cell line, suggesting a faster release of the cytotoxic 7-THA at slightly acidic pH values. Release kinetics also supported this hypothesis: a new model system utilizing benzodithiophene (BDT), a more water-soluble mimic of 7-THA, was prepared analogously to the HNT-7TH powders. The release of BDT was monitored via UV-vis spectroscopy at pH 5.0, 6.8 and 7.4. The amount of released BDT at pH 7.4 was negligible even after 48 h; decreasing the pH value to 6.8 visibly increased release rates, while the release efficiency was highest for the treatment at pH 5.0. The natural origin of HNT causes a marked variability in its physicochemical features, such as its morphology and surface charge, depending on the extraction site[4]. I investigated this aspect, often overlooked in the literature, with respect to the integration of HNT as nanofillers in polymer composites. Halloysite has been investigated as filler owing to its low cost, thermal and mechanical resistance, and high aspect ratio, which is crucial to guarantee strong polymer-filler interactions[5]. In this regard, I investigated also the effect of surface functionalization in promoting HNT compatibility with the chosen polymer matrix, Polyamide 6 (PA6). PA6 has a broad range of applications, from the automotive sector to the textile industry, owing to its good mechanical performances and high thermal resistance. However, when exposed to humid conditions, it suffers from degradation in a matter of few weeks[6], as water can interpenetrate within the hydrogen bonds between –NH and C=O groups and disrupt them, leading to the loss of tensile strength and elasticity[6–9]. The addition of nanotubular fillers represents a viable strategy for overcoming this issue, although the additive/polymer interface at high filler content can become a privileged site for moisture accumulation[10]. For this reason, HNT were added to PA6 in very low amounts (< 5%w). The roles played in the reinforcement of the polymer by the physicochemical properties of HNT from two different sources and their functionalization with APTES were investigated in composites prepared by two different dispersion techniques (in situ polymerization vs. melt blending). The aspect ratio (5 vs. 15) and surface charge (−31 vs. −59 mV) of the two HNT samples proved crucial in determining their distribution within the polymer matrix: both in situ and melt blending dispersion techniques showed that lower surface area, higher aspect ratio and greater surface charge enhance filler incorporation and improve the final composite performance. Finally, filler surface modification with APTES played a major role in the durability of the PA6-HNT nanocomposites: after 1680 h of hydrothermal ageing, functionalized HNT reduced the diffusion of water into the polymer, lowering water uptake after 600 h up to 90%, increasing the materials durability. Positive effects could also be measured regarding the molecular weight distribution and rheological behaviour. These improvements could be related to the presence of amino groups on the HNT surface, which lowered the filler surface energy and prevented the diffusion of water molecules into the nanocomposites[11]. 2. Titanium dioxide TiO2 is arguably the most extensively investigated semiconductor for photocatalytic applications, from solar cells to pollutant abatement. However, applications of TiO2 for the preparation of smart surfaces are comparatively less common. My research interest in TiO2 started with the preparation of surfaces with controlled wetting features. First, the surface of commercial TiO2 powders was functionalized with perfluorinated alkylsilanes. Then, I employed the photocatalytic features of TiO2 films to produce patterned surfaces with superhydrophilic/superhydrophobic contrast by means of photocatalytic lithography. During the course of my second year of PhD, I focused on the deposition of TiO2 films with controlled porosity followed by their surface modification to impart them functional properties. These systems were applied as photo renewable coatings for electrochemical sensors and stimuli-responsive surfaces for the controlled release of active compounds. Control over film porosity was achieved by a hard-template approach using polystyrene (PS) nanospheres of different sizes, both commercial and synthesized in-house following a classical procedure[12]. Different film deposition strategies were investigated. First, I deposited a TiO2 layer on top of a porous SiO2 coating on a FTO electrode. Extensive characterization via cyclic voltammetry showed that the addition of a TiO2 layer increased peak currents due to the promoted diffusion of the analyte in the porous structure, driven by capillary effects[13] (Figure 2A). Furthermore, the TiO2 layer promoted the light-activated regeneration of the electrode surface after having been fouled. Starting from there, I investigated the possibility of pure TiO2 mesoporous films, a task made difficult by the intrinsic incompatibility between an alcohol-based TiO2 sol, extremely prone to hydrolysis, and a water-based PS suspension. This issue was solved by either adopting an aqueous TiO2 sol, slowly evaporated in presence of the particle templates (Figure 2B), or by performing a solvent exchange procedure on the PS suspension. The latter procedure resulted in pure TiO2 films with easily tuneable thickness and homogeneous porosity, opening the door to a fine tuning of the cyclovoltammetric response. The self-cleaning features of the pure TiO2 films were also tested by purposely fouling their surface with long-alkyl chain substituents: a fast and complete regeneration of the surface was achieved upon irradiation with UV light. The prepared pure TiO2 porous films were also used as substrates for the loading of bioactive substances. Cinnamaldehyde, a natural substance known for its antimicrobial properties but unstable in environmental conditions, was anchored to the film surface via APTES linkers through an imine bond, using a protocol similar to the one developed for 7-THA-loaded HNT. The immobilized cinnamaldehyde proved stable to environmental conditions for months and tests of pH triggered release performed at pH 5.5 and 7.4, showed a faster release at lower pH values. Finally, the photoactive nature of the oxide substrate could be used to promote the self-cleaning of the fouled surface after usage: after UV-light regeneration, the TiO2 film could be functionalized anew and reused. 3. Superparamagnetic Iron Oxide Nanoparticles (SPION) During my third year of PhD, I spent six months at the Technische Universiteit Delft, Netherlands, in the Advanced Soft Matter group, under the supervision of Dr Laura Rossi. There my research focused on the synthesis and surface modification of Ultrasmall SPION. SPION have gained increasing attention thanks to their peculiar behaviour: being smaller than a single magnetic crystal domain, they are free to rotate unless a specific orientation is induced by an external field. Due to their magnetic properties, they can be adopted in hyperthermia, drug delivery and as contrast agents (CA) for magnetic resonance imaging (MRI)[14]. Contrast agents are commonly used to speed up either T1 or T2 relaxation, enhancing the local contrast in pathological tissue to produce more detailed images. T1 CA are commonly represented by gadolinium complexes, while SPION are generally adopted as T2 CA. The first are preferred by radiologists, while the latter are less favoured because the darker tones they provide can be mistaken with low resolution and background interference[15,16]. Nonetheless, Gd-based CA present a serious health risk for those patients unable to efficiently remove these heavy metal complexes due to pre-existent kidney or liver pathologies[17]. While most SPION act as T2 CA, several papers report their potential use as T1 CA if their size is sufficiently small, indicatively less than 4 nm[18,19]. These materials are known as Ultrasmall SPION, or USPION. My aim was to develop a synthetic protocol to prepare USPION suitable as T1 contrast agents via co-precipitation, to minimize synthetic requirements. To this purpose, the influence of several parameters such as reaction temperature, base type, purification procedure, stabilizing agent and precursor concentration was investigated. Particles were synthesised at room temperature (RT) and 50°C, using NH4OH, N(CH3)4OH or NaOH as base. Particle purification was performed via magnetic decantation, centrifugation and dialysis against different solutions (water, citric acid and sodium citrate solutions). Sodium oleate, (3-aminopropyl)triethoxysilane (APTES) and citric acid were tested as stabilizing agents and precursor concentration was varied between 1 M and 0.5 M. It was found that the best results were obtained at room temperature and that the peptizing effect of the tetramethylammonium ion is crucial to guarantee an optimal colloidal stability, making N(CH3)4OH the base of choice. The concentration of starting precursor solutions proved to be the determining factor acting on particle size, as halving it led to a narrow particle size distribution centred around 3 nm, a significant shift from the starting 7 nm (Figure 3). Centrifugation was ineffective when adopted to wash the nanoparticles, but it proved a promising size-selection tool that could be combined with dialysis in an efficient work up protocol. Dialysis proved to be the most efficient technique to remove potentially toxic impurities, but it negatively impacted the colloidal stability, which could be mitigated by the use of a proper stabilizing agent. To preserve a high colloidal stability even after the removal of N(CH3)4OH, surface modification with several stabilizing agents was tested. Among the tested molecules, citric acid was the only one to show positive effects on particle size and aggregation, more so when added before the start of particle nucleation. These results represent a promising advance in the development of efficient T1 contrast agents based on USPION in terms of lowering the synthetic requirements: monodisperse magnetic nanoparticles were prepared through a simple co-precipitation procedure, performed at room temperature, without the aid of any polymeric additive. References [1] Y. Lvov, W. Wang, L. Zhang, R. Fakhrullin, Halloysite Clay Nanotubes for Loading and Sustained Release of Functional Compounds, Adv. Mater. 28 (2016) 1227–1250. doi:10.1002/adma.201502341. [2] E. Abdullayev, V. Abbasov, A. Tursunbayeva, V. Portnov, H. Ibrahimov, G. Mukhtarova, Y. Lvov, Self-healing coatings based on halloysite clay polymer composites for protection of copper alloys, ACS Appl. Mater. Interfaces. 5 (2013) 4464–4471. doi:10.1021/am400936m. [3] E. Licandro, S. Cauteruccio, D. Dova, Thiahelicenes, in: 2016: pp. 1–46. doi:10.1016/bs.aihch.2015.12.001. [4] E. Joussein, S. Petit, J. Churchman, B. Theng, D. Righi, B. Delvaux, Halloysite clay minerals – a review, Clay Miner. 40 (2005) 383–426. doi:10.1180/0009855054040180. [5] M.R. Ayatollahi, S. Shadlou, M.M. Shokrieh, M. Chitsazzadeh, Effect of multi-walled carbon nanotube aspect ratio on mechanical and electrical properties of epoxy-based nanocomposites, Polym. Test. 30 (2011) 548–556. doi:10.1016/j.polymertesting.2011.04.008. [6] I. Ksouri, O. De Almeida, N. Haddar, Long term ageing of polyamide 6 and polyamide 6 reinforced with 30% of glass fibers: physicochemical, mechanical and morphological characterization, J. Polym. Res. 24 (2017) 133. doi:10.1007/s10965-017-1292-6. [7] H. Shinzawa, J. Mizukado, Water absorption by polyamide (PA) 6 studied with two-trace two-dimensional (2T2D) near-infrared (NIR) correlation spectroscopy, J. Mol. Struct. 1217 (2020) 128389. doi:10.1016/j.molstruc.2020.128389. [8] K.R. Rajeesh, R. Gnanamoorthy, R. Velmurugan, Effect of humidity on the indentation hardness and flexural fatigue behavior of polyamide 6 nanocomposite, Mater. Sci. Eng. A. 527 (2010) 2826–2830. doi:10.1016/j.msea.2010.01.070. [9] D.P.N. Vlasveld, J. Groenewold, H.E.N. Bersee, S.J. Picken, Moisture absorption in polyamide-6 silicate nanocomposites and its influence on the mechanical properties, Polymer (Guildf). 46 (2005) 12567–12576. doi:10.1016/j.polymer.2005.10.096. [10] K.P. Pramoda, T. Liu, Effect of moisture on the dynamic mechanical relaxation of polyamide-6/clay nanocomposites, J. Polym. Sci. Part B Polym. Phys. 42 (2004) 1823–1830. doi:10.1002/polb.20061. [11] K. Prashantha, M.F. Lacrampe, P. Krawczak, Processing and characterization of halloysite nanotubes filled polypropylene nanocomposites based on a masterbatch route: effect of halloysites treatment on structural and mechanical properties, Express Polym. Lett. 5 (2011) 295–307. doi:10.3144/expresspolymlett.2011.30. [12] J.W. Goodwin, J. Hearn, C.C. Ho, R.H. Ottewill, Studies on the preparation and characterization of monodisperse polystyrene latices, Colloid Polym. Sci. 252 (1974) 464–471. [13] L. Rimoldi, V. Pifferi, D. Meroni, G. Soliveri, S. Ardizzone, L. Falciola, Three-dimensional mesoporous silica networks with improved diffusion and interference-abating properties for electrochemical sensing, Electrochim. Acta. 291 (2018) 73–83. doi:10.1016/j.electacta.2018.08.131. [14] J. Dulińska-Litewka, A. Łazarczyk, P. Hałubiec, O. Szafrański, K. Karnas, A. Karewicz, Superparamagnetic Iron Oxide Nanoparticles—Current and Prospective Medical Applications, Materials (Basel). 12 (2019) 617. doi:10.3390/ma12040617. [15] Y. Okuhata, Delivery of diagnostic agents for magnetic resonance imaging, Adv. Drug Deliv. Rev. 37 (1999) 121–137. doi:10.1016/S0169-409X(98)00103-3. [16] J.-C. Brisset, M. Sigovan, F. Chauveau, A. Riou, E. Devillard, V. Desestret, M. Touret, S. Nataf, J. Honnorat, E. Canet-Soulas, N. Nighoghossian, Y. Berthezene, M. Wiart, Quantification of Iron-Labeled Cells with Positive Contrast in Mouse Brains, Mol. Imaging Biol. 13 (2011) 672–678. doi:10.1007/s11307-010-0402-1. [17] M. Rogosnitzky, S. Branch, Gadolinium-based contrast agent toxicity: a review of known and proposed mechanisms, BioMetals. 29 (2016) 365–376. doi:10.1007/s10534-016-9931-7. [18] H. Wei, O.T. Bruns, M.G. Kaul, E.C. Hansen, M. Barch, A. Wiśniowska, O. Chen, Y. Chen, N. Li, S. Okada, J.M. Cordero, M. Heine, C.T. Farrar, D.M. Montana, G. Adam, H. Ittrich, A. Jasanoff, P. Nielsen, M.G. Bawendi, Exceedingly small iron oxide nanoparticles as positive MRI contrast agents, Proc. Natl. Acad. Sci. 114 (2017) 2325–2330. doi:10.1073/pnas.1620145114. [19] Y. Bao, J.A. Sherwood, Z. Sun, Magnetic iron oxide nanoparticles as T 1 contrast agents for magnetic resonance imaging, J. Mater. Chem. C. 6 (2018) 1280–1290. doi:10.1039/C7TC05854C.