Functionalization of Metal Oxide Nanostructures via ...

Functionalization of Metal Oxide Nanostructures via Self-Assembly. Implications and Applications.

Funktionalisierung von Metalloxid-Nanostrukturen mit selbstorganisierenden Monolagen. Implikationen und Anwendungen.

Der Technischen Fakultät der Friedrich-Alexander-Universität Erlangen-Nürnberg zur Erlangung des Doktorgrades Dr.-Ing.

vorgelegt von

Luis Francisco Portilla Berlanga aus Torreon (Coahuila, Mexico)

Als Dissertation genehmigt von der Technischen Fakultät der Friedrich-Alexander-Universität Erlangen-Nürnberg

Tag der mündlichen Prüfung:

19.01.2017

Vorsitzende des Promotionsorgans :

Prof. Dr.-Ing. Reinhard Lerch

Gutachter:

Prof. Dr. rer. nat. Marcus Halik Prof. Dr. rer. nat. Carola Kryschi

Abstract Self-assembled monolayers (SAMs) were employed for the surface modification of a variety of metal oxide nanostructures (NSs). This resulted in an remarkable assortment of inorganic-organic (coreshell) hybrid materials. These materials were fabricated, characterized and employed in different concepts. SAMs allowed to incorporate the endless realm of organic chemistry into the less vast territory of inorganic materials. An in depth, step by step analysis of the aspects involved in the process, of going from a pristine inorganic material to an inorganic-organic (core-shell) hybrid material, is described. In addition, the general theoretical background, which is required to fully grasp the processes and characterization is herein described. The final outcome was, well characterized core-shell building blocks with a solid groundwork regarding their new properties. The latter was of paramount importance, as this hybrid NSs were meant to be further employed in other complex processes.

The pristine NSs cores and core-shell NSs were characterized with a variety of techniques to ensure that a complete and well controlled surface modification has taken place. Chemical information about the composition of the organic shell of the core-shell NSs was obtained via Fourier transform infra red (FTIR) spectroscopy. The amount of organic vs. inorganic material (SAM grafting density) was characterized with thermogravimetric analysis (TGA). The size and zeta potential of NSs were measured using dynamic light scattering techniques (DLS). Surface energies and wettability of the NSs surface were characterized with goniometry via static contact angle (SCA) measurements. A comparison of the robustness between different types of SAM molecules, phosphonic acids, carboxylic acids and catechols was performed as well.

The new properties of the inorganic materials which arose from the effects of surface modification, were carefully tailored and exploited for several applications. Due to the flexibility allowed by incorporating organic chemistry, the resulting applications often had a multidisciplinary character. These applications are also described within this work. Some of them are still work in progress and are therefore described with less detail than others. Example applications include: Solution processing, polymer wrapping, thin films, coatings, self-organization of NSs, polymer composites, nanooncology as well as waste water treatment.

Kurzfassung Selbstorganisierte Monolagen (SAMs) wurden für die Oberflächenmodifizierung einer Vielzahl von Metalloxid-Nanostrukturen (NSs) verwendet. Daraus ging eine bemerkenswerte Anzahl an anorganisch-organischen (Kern-Hülle) Hybridmaterialien hervor. Diese Materialien wurden hergestellt, charakterisiert und in unterschiedlichen Konzepten angewendet. Die Verwendung organischer SAMs ermöglicht die Vielfalt der organischen Chemie mit hoch funktionalen anorganischen Materialien zu verbinden. Eine schrittweise Analyse der in den Prozess eingebunden Parameter wird beschrieben, wie aus einem ursprünglichen anorganischen Material ein anorganisch-organisches (Kern-Hülle) Hybridmaterial wird. Darüber hinaus ist der allgemeine theoretische Hintergrund beschrieben, der erforderlich ist, um die Prozesse und die Charakterisierung zu erfassen. Das Endergebnis waren vollständig charakterisierte Kern-Hülle NSs mit einem Schwerpunkt in Bezug auf ihre einzigartigen Eigenschaften. Dies war von größter Wichtigkeit, da diese Hybrid-NSs weiter in anderen komplexeren Anwendungen eingesetzt wurden.

Die unfunktionalisierten NSs Kerne und Kern-Hülle NSs wurden mit einer Vielzahl von Techniken charakterisiert, um sicherzustellen, dass eine vollständige und gut gesteuerte Oberflächenmodifikation stattgefunden hat. Informationen über die chemische Zusammensetzung der organischen Hülle des Kern-Hülle NSs wurde mittels Fourier-Transformations-Infrarot (FTIR) Spektroskopie erhalten. Das Verhältnis von organischem zu anorganischem Material (SAM Ankerdichte) wurde durch thermogravimetrische Analyse (TGA) bestimmt. Die Größe und das Zeta-Potential von NSs wurden mittels dynamischer Lichtstreuung (DLS) gemessen. Oberflächenenergien und Benetzbarkeit der NSs Oberfläche wurden mit Goniometrie durch die Messung der statischen Kontaktwinkel (SCA) charakterisiert. Ebenfalls wurde ein Vergleich der Robustheit zwischen verschiedenen Arten von SAM-Molekülen, Phosphonsäuren, Carbonsäuren und Katechinen durchgeführt. Die

neuen

Eigenschaften

der

anorganischen

Materialien

konnten,

ermöglicht

durch

Oberflächenmodifizierung, für verschiedenste Anwendungen präzise angepasst und genutzt werden. Die daraus resultierenden Anwendungen besitzen oft einen multidisziplinären Charakter, der durch den flexiblen Einsatz organischer Moleküle ermöglicht wird. Diese Anwendungen werden auch im Rahmen der vorliegenden Arbeit beschrieben, wobei einige noch weiterentwickelt und daher weniger detailliert behandelt werden. Als beispielhafte Anwendungen werden Lösungsprozessierung, dünne Schichten, beschrieben.

Polymerkomposite,

Beschichtungen,

Nanoonkologie

sowie

Abwasserbehandlung

Table of contents Abstract .................................................................................................................................................. iii Kurzfassung ............................................................................................................................................. v 1. Introduction and motivation ................................................................................................................ 1 2. Theoretical background ....................................................................................................................... 3 2.1. Nanostructured materials and nanostructures ............................................................................... 3 2.2. Surface area of nanostructures...................................................................................................... 5 2.3. Self-assembled monolayers .......................................................................................................... 6 2.3.1. Anchor groups ....................................................................................................................... 9 2.3.2. Backbone ............................................................................................................................. 10 2.3.3. Head group .......................................................................................................................... 11 2.3.4. Studied anchor groups ......................................................................................................... 11 2.4. Stability of 0D and 1D NSs in solution ...................................................................................... 15 2.4.1. Electrostatic stabilization..................................................................................................... 16 2.4.2. Steric stabilization ............................................................................................................... 20 2.4.3. Electrosteric ......................................................................................................................... 21 2.4.4. Impact of size and geometry of nanostructures ................................................................... 21 2.5. Thin film transistors ................................................................................................................... 23 2.6. Chapter summary........................................................................................................................ 24 3. Functionalization and characterization of NSs .................................................................................. 25 3.1. Functionalization of 2D materials .............................................................................................. 25 3.2. Functionalization of 0D materials .............................................................................................. 26 3.2.1. Chemical characterization (FTIR-ATR) .............................................................................. 27 3.2.2. Saturation threshold (TGA) ................................................................................................. 33 3.2.3. Mixed monolayers (SCA, FTIR & Zeta potential) .............................................................. 39 3.3. Anchor group stability ................................................................................................................ 43 3.3.1. Desorption ........................................................................................................................... 43 3.3.2. Exchange on 2D NSs ........................................................................................................... 46 3.4. Chapter summary........................................................................................................................ 51 i

ii

Table of contents

4. Applications....................................................................................................................................... 53 4.1. Solution processing .................................................................................................................... 53 4.1.1. Green processing ................................................................................................................. 53 4.1.2. Any medium processing ...................................................................................................... 55 4.1.3. Shell by shell (double shell) ................................................................................................ 56 4.1.4. Polymer wrapping ............................................................................................................... 59 4.2. Thin films. From 0D to 2D ......................................................................................................... 64 4.2.1. Flexible dielectrics............................................................................................................... 64 4.2.2. Coatings ............................................................................................................................... 68 4.3. Self-assembled thin films ........................................................................................................... 70 4.3.1. Regio-selective deposition of nanoparticles ........................................................................ 70 4.3.2. Block co-polymer phase matching ...................................................................................... 74 4.4. Polymer composites.................................................................................................................... 76 4.5. Nanooncology............................................................................................................................. 78 4.6. Magnetic water cleaning............................................................................................................. 83 4.7. Chapter summary........................................................................................................................ 85 5. Conclusion and outlook ..................................................................................................................... 87 6. Characterization methods and materials ............................................................................................ 89 6.1. SCA ............................................................................................................................................ 89 6.2. DLS ............................................................................................................................................ 89 6.3. FTIR-ATR .................................................................................................................................. 89 6.4. TGA ............................................................................................................................................ 90 6.5. Electrical ..................................................................................................................................... 90 6.6. Spray coating .............................................................................................................................. 90 6.7. Materials ..................................................................................................................................... 90 6.8. Functionalization procedures...................................................................................................... 91 6.8.3. AlOx (Sigma A.), ITO (Sigma A.) and TiO2 (Nanograde 30 nm) ....................................... 91 6.8.4. CeO2 .................................................................................................................................... 91 6.8.5. Fe3O4 (Plasmachem ~10 nm) ............................................................................................... 92 6.8.6. TiO2 (Plasmachem 8 nm) ..................................................................................................... 93

Table of contents

iii

6.8.7. Fe3O4 and CoFe3O4 (Prof. Kryschi) ..................................................................................... 94 7. Appendix ........................................................................................................................................... 95 8. List of Figures ................................................................................................................................. 101 9. List of tables .................................................................................................................................... 107 10. Abbreviations ................................................................................................................................ 109 11. Bibliography .................................................................................................................................. 111 12. Acknowledgements ....................................................................................................................... 121 13. Curriculum vitae ............................................................................................................................ 123

1. Introduction and motivation

The advent of tool making marked a point in history which started a continuous technological evolution. Ever since, this understanding has gradually developed and has given shape to our contemporary world and society. In present day technology, manipulation of materials at the nano scale is employed for the creation of new materials and device fabrication. This happens both at industrial and research scales. As the nanorevolution develops, so do our demands for improved devices and technologies. Towards the fulfillment of these requirements, nanoscience provides “plenty of room” to elegantly innovate and expand current designs [1]. The working principle of all modern or primitive devices is based on one simple concept. Essentially, it consists in generating a contrast between two or more materials arranged in a particular configuration. This allows for the differing properties of materials to be beneficially exploited for a specific purpose. In retrospect, the fabrication of a device lacking this contrast would be impractical, or simply impossible to conceive. It can therefore be stated, that material contrast is at the core of any instrument or device that has ever been fabricated. The impact of the previous statement is even more accurate when dwelling with device fabrication at the nanoscale, where properties of materials may differ from those of the bulk. At such scale, material properties can be dependent on their dimensions, geometry, or neighboring materials. This results in additional degrees of freedom when it comes to device fabrication [2], [3]. It is by exploiting these concepts, that this thesis explores the fundamental and technical implications of creating a material contrast via the tailoring of the surface of 0D nanostructures (NSs) with self-assembled monolayers (SAMs). The NSs surface properties are carefully tuned in order to be employed as building blocks for their integration into nanostructured materials (NSMs) and consequentially device fabrication. The research and application fields of NSMs are as broad as the NSMs themselves [2]. The most notable area involving NSMs is the field of semiconductors, being this area the most fruitful of the efforts. However, they very often encompass an interdisciplinary mixture of every natural science. For instance: energy generation and storage [4], [5] enhanced mechanical properties by creating nanocomposite materials [6], nanomedicine (in areas such as targeted drug delivery and nanooncology [7]–[9]), environmental [5], even intricate self-assembly [10] and bottom up fabrication schemes [11], just to name a few.

1

2

Introduction and motivation

Just as NSMs, SAMs encompass a wide area of research and applications themselves [12]–[15]. The field of SAMs and the field of NSMs are closely related, mostly because SAMs offer a versatile bottom-up approach for surface tuning and control (tailoring) of the NSs. Without SAMs, a total and direct surface control of the NSs would often remain out of reach. These SAMs which assemble around the surface of the NSs, create a material contrast between the inorganic core and the organic SAM (core-shell). Naturally, this contrast can be exploited for a specific function. It becomes evident now, that the combination of NSMs and SAMs adds additional degrees of freedom to the already diverse system of NSMs. This in turn, provides a way for new solution processing techniques which involve the self-assembly of "smart" NSs for the formation of NSMs. In addition, this thesis deals with the necessary steps required to provide well characterized and controllable core-shell 0D materials. Lastly, their applications are also discussed. As the prospective applications of the core-shell NSs are described, the multidisciplinary makeup and diverse potential of NSMs and SAMs becomes more and more palpable. A concise description of these applications has been partly published in peer reviewed journals. These applications include: Coatings [11], [16], flexible electronics [17], self-assembly concepts of NSMs [11], [18], as well as some more elementary studies [19]. Other applications which may or may not be under preparation for publication, are also described. However, before discussing these applications in detail, there are a few fundamental concepts that must be elaborated in order to properly discuss these ideas.

2. Theoretical background

In this chapter, we will look into the concept of nanostructures (NSs) and nanostructured materials (NSMs) and their classification. This is followed by the theoretical background of self-assembled monolayers (SAMs), which is concisely described using 2D NSs as an archetypal model. Lastly, the comparatively more intricate consequences of SAMs grafted onto 0D NSs are more exhaustively analyzed.

2.1. Nanostructured materials and nanostructures Primarily, NSMs should be differentiated from NSs. NSMs consist of an arrangement of smaller building blocks. These building blocks (of which NSM are constituted) consist of NSs that can be classified into four classes in function of their dimensionality: 0D, 1D, 2D and 3D (Figure 2.1). As NSs, we distinguish structures in which at least one of its dimensions is of a critical magnitude. This critical magnitude gives rise to size dependent material properties. This critical dimension is frequently in the submicron and nano regimes. However, the exact magnitude of these critical dimensions is difficult to demarcate as they are dependent on various physical phenomena. Therefore, they are typically approached on a case to case basis.

Figure 2.1.: Representation of nanostructures. 0D (red sphere), 1D (green cylinder), 2D (blue rectangular parallelepiped) and in 3D (cylinder 3D matrix).

3

4

Theoretical background

However, a peculiarity needs to be noted in this classification scheme. The discrepancy between a 3D NS and a NSM at this point seems to be vague. Yet, a clear difference between them exists. 3D NSs refer to a 3D matrix comprised of a single material. For example, a phase separated block-copolymer matrix or a porous and hollow structure [20]. Whereas NSMs are comprised of an array of 0D, 1D, 2D or 3D NSs. However, in reality (since the only 3D NS is a matrix) the terms 3D NSs and NSM are used ambiguously in literature [3], [4], [21]. Going into specifics, a systematic classification of NSs has already been proposed by V. Pokropivny in 2006 [21]. It has been adapted in Figure 2.2.

Figure 2.2.: NSs classified on basis of their dimensionality as suggested by V. Pokropivny in 2007 [21]. Reprinted from [21] with permission from Elsevier.


5

2.2. Surface area of nanostructures 0D and 1D NSs are well known for their high surface area to volume ratio when compared to their bulk equivalents. By using a cube, a classical example of this concept is portrayed in Figure 2.3. In this example, as we progress from Figure 2.3a to d, the initial cube is split into eight smaller cubes followed by the splitting of the smaller cubes into eight cubes again and so forth. What we observe is, that by splitting the cube into smaller components we have exponentially increased the total amount of exposed surface, while the initial total volume remains unchanged. Pushing the analogy of Figure 2.3 further into nanoscale proportions, the vast amount of surface area available becomes palpable compared to that of the bulk material. In layman terms this translates to, that a few milligram (mg) of nanoparticles could in fact have more real state than your current shared flat apartment. In terms of SAM formation it means, that a very high quantity of molecules are going to be required if the intent is to fully cover the surface of the nanoparticles (or your entire flat) with SAM molecules. This in great contrast to 2D NSs, were as far as they are concerned, the area involved is simply very much related to the substrate dimensions. How to categorize and measure this surface area is now explained below.

Increase of surface area to volume ratio (SA:V) a)

b)

c)

d)

S. Area = 6 u2

S. Area = 12 u2

S. Area = 24 u2

S. Area = 384 u2

Volume = 1 u3

Volume = 1 u3

Volume = 1 u3

Volume = 1 u3

SA/V ratio = 6:1

SA/V ratio = 12:1

SA/V ratio = 24:1

SA/V ratio = 384:1

Figure 2.3.: Surface area to volume ratio. The increment of surface area vs. volume is graphically represented from a) lowest ratio, to d) highest ratio.

The standard property of solids by which the surface area of a solid (or powder) is defined, is called Specific Surface Area (SSA). Commonly employed units of SSA are usually specified in m2/kg or m2/g. What the SSA property achieves, is to simply correlate a specific amount mass to a specific amount of area. There are two main methods from which the SSA of a nanomaterial can be obtained. By measuring the amount of adsorbate gas to the nanomaterial powder based on the Brunauer– Emmett–Teller isotherm (BET) [22]. Or, geometrically estimated from a particle size distribution of the nanomaterial as it was performed within this work.

6


When there is access to the mean particle size or a particle size distribution, the SSA of a nanopowder can be approximated by simple geometric calculations. This method however, is most accurate when the morphological discrepancies between individual particles are minimal. That is, all particles tend to have the same shape (spheroidal, cuboid, rods, etc.). For the sake of argument, let's assume the case in which all of the particles are spheres with exactly the same size. In this case we can relate the volume of a sphere, density of the material and the mass of material to a finite amount of particles resulting in Equation 2.1, were m = mass, d = density and r = mean radius of particles. 𝐴𝑚𝑜𝑢𝑛𝑡 𝑜𝑓 𝑝𝑎𝑟𝑡𝑖𝑐𝑙𝑒𝑠 =

3𝑚 4𝑑𝜋𝑟 3

2.1

Once the amount of particles is known, simply multiplying the area of a spherical particle by the amount of particles results in the total available area (Equation 2.2). To get some perspective into this, using the proposed scenario, if we calculate the total area of 100 mg of alumina (d = 3.95 g/cm3) particles with a diameter of 50 nm the total area results in 3.03 m2. However, doing the same calculation but with 3 nm particles results in more striking results with an area of 50.63 m2. 𝑇𝑜𝑡𝑎𝑙 𝐴𝑟𝑒𝑎 = 𝐴𝑚𝑜𝑢𝑛𝑡 𝑜𝑓 𝑝𝑎𝑟𝑡𝑖𝑐𝑙𝑒𝑠 × 4𝜋𝑟 2

2.2

Finally, since the SSA unit is in function of mass, diving the total area in m2 by the mass gives the final result (Equation 2.3).

𝑆𝑆𝐴 =

𝑇𝑜𝑡𝑎𝑙 𝐴𝑟𝑒𝑎 𝑚

2.3

Naturally, this method is not restricted to a single particle size. It is also applicable if a nanoparticle size distribution is available. This is achieved by calculating the total area of the different nanoparticle diameters from the distribution and doing a sum of the calculated areas. In conclusion, the SSA is a critical parameter to know before functionalizing 0D or 1D NSs as it will define a theoretical starting point for the amount of required SAM molecules. A major disadvantage of estimating the SSA via geometric calculations is the possibility of a big margin of error, in particular for porous samples or with a rich morphological diversity. Therefore, complementing the SSA calculations with either BET or electron microscopy measurements is highly advisable whenever possible.

2.3. Self-assembled monolayers Concisely, a self-assembled monolayer consists of a spontaneous, highly ordered, self-terminating reaction on a surface. These reacting species usually consist of chain like organic molecular reagents, [12], [14], [23]. The deposition of SAMs typically requires the solvation of the molecules in a liquid or


7

gas medium which allows for the free movement of the SAM molecules [24]. This allows them to selforganize on a surface that is also present within the solvation medium of the SAM molecules. Figure 2.4a shows a graphic representation of the self-assembly process on a 2D NS. It starts with a vacant surface which is subsequently exposed to the SAM molecules. The self-assembly process begins when the free SAM molecules start to move around the vicinity of the NSs surface. At this point they may become physisorbed to the surface. This physisorption, is usually followed by chemisorption onto the surface, however, it is not a requirement for SAM formation. As more molecules bind to the surface, the SAM starts to take shape, this binding of molecules continues until the surface is covered by only one layer of the SAM molecules. This is possible because the surface has a limited amount of reactive sites and the SAM molecules cannot bind between them. In parallel, the intermolecular forces of the SAM molecules can also contribute to the organization and package density of the SAM. By organization of a SAM, we can understand at least two things: SAM tilt angle (Figure 2.4b) and the formation of crystalline domains (Figure 2.4c). The way these parameters are affected depends mainly on the chemical structure of the molecules and the deposition conditions. Eventually, when no more anchoring sites are available at the surface, the SAM molecules are packed as tight as possible and the SAM formation is finished. Analogously depicted in Figure 2.5, the process of SAM formation on 0D NSs follows exactly the process described above. Conversely, as similar as the processes might be, the consequences of SAMs grafted onto 0D NSs diverge in many degrees from those grafted onto 2D NSs. Naturally, it goes without saying, that both systems have complications of their own, which are more thoroughly described later in this work. The structure of self-assembling molecules can be broken down into 3 principal components: The anchor group, the backbone and the head group. Each component plays a crucial role in SAM formation and each one is described in detail throughout this section. Figure 2.6 shows a prototypical graphical representation of the geometrical hierarchy of a SAM grafted onto a 0D NS. Figure 2.6 also describes some commonly employed SAM molecule configurations. At the center of Figure 2.6, there is the NS core, which in our case consists of an inorganic oxide material. The core provides part of the functionality of such building blocks. Core functionalities are diverse, some commonly employed core properties are magnetic, conductive, dielectric, transparent, opaque, used as carriers, etc. As we proceed outwards (in red), we encounter the anchor group which interacts and attaches the molecule to the NS surface. Afterwards in blue, there is the backbone or spacer of the SAM molecule. As its name implies, this component mainly acts as a support for the SAM molecule. However, being the largest component of the molecule, it often provides functionality in an indirect manner rather than on purpose. Finally, there is the head group (in green), which usually defines the main functionality of the SAM molecule. However, the roles of the SAM components are more complex than just described, for this reason they are explained in more detail below.

8


a)

b)

θ SAM tilt angle

2 D S A M f o r m a t i o n

c)

amorphous

crystalline

crystalline

Figure 2.4.: a) Schematic process of the formation of a C18-PA SAM on a 2D NS. b) Side view depiction of SAM tilt angle. c) Top view illustration of the crystalline and amorphous domains that may be present after the SAM formation.

0D SAM formation

Figure 2.5.: Schematic process of SAM formation of onto a 0D NS.


9

Head group (functionality or solubility) •Inert •Reactive •Polar •Ionic •Semiconducting •Etc.

Backbone (solubility and geometry) •Orthogonal •Polar •Ionic •Etc.

Anchor group (stability of SAM) •Phosphonic acid •Carboxylic acid •Catechol

Nanoparticle core (functionality) •Alumina •Titania •Iron oxide •Etc.

Figure 2.6.: Anatomy of a core-shell 0D building block.

2.3.1. Anchor groups The anchor group is the most crucial component of any SAM molecule, it is responsible for the binding of the SAM molecule to the surface. The nature of the bond between the surface and the anchor group is of paramount importance and must be chosen adequately depending on the SAM molecule structure, NS surface properties and the nature of desired application. In order for selfassembly to properly occur, the interaction between the anchor group and the NS surface must be the governing attraction force. This being particularly significant whenever the backbone or the head group of the SAM molecule consist of any number of electrostatic or polar species that could in principle, also interact with the NS surface. Therefore, anchor groups capable of covalent bonding are preferred, especially when a permanent modification of the surface is required. Weaker interactions (electrostatic, hydrogen bonding, orthogonal or Van der Waals) are ideally only sought after when a

10


less permanent or reversible alteration of the surface is required [24], [25]. These weaker interactions can also be employed as permanent modifications, however, they may only be so under more strictly controlled circumstances [18].

2.3.2. Backbone It is very often the case for SAM molecules, that the backbone is the biggest constituent of all components. Usually, the role of the backbone typically comes down to purely geometrical aspects. Intuitively, the backbone acts as a spacer between the anchor group and the head group. In addition, this spacing allows for the head group to be more freely exposed by being further away from the surface. This is predominantly true for surfaces with a pronounced curvature as it is later showcased in Figure 2.17b and d. Nevertheless, the backbone can also have strong influence on the ordering of the SAM as well as the tilt angle. Generally speaking, the stronger the backbone to backbone interactions the more ordered the final structure of the SAM will be. Take for example Figure 2.7. When using these molecules to build a SAM, the bigger the carbon chain backbone is, the stronger the van der Waals interactions between them will be. This results in better ordered SAMs [26]. However, it must be noted that there must be a threshold to this approach, in which increasing the chain length won't necessarily incur in higher ordering, instead the contrary can happen. It is also possible, that the backbone interactions play a negative role on the final ordering of the SAM if they are of a repulsive or other non favorable nature.

Less ordered SAM

More ordered SAM

Figure 2.7.: Phosphonic acid SAM molecules of increasing length. As the backbone (black) length increases, so do the van der Waals interactions between them and therefore also the final ordering of the SAM [26].


11

On a more chemical side of things, the backbone being the biggest component, has a strong influence on its solubility. By solubility we refer to the facility of the SAM molecule to be solvated in the medium of deposition, as well as the dispersibility it may provide by means of steric or electrostatic stabilization when grafted onto a 0D or 1D NSs. These effects are thoroughly described in Section 2.4 As a final point, SAM Backbones are usually of a non-conductive nature. Therefore, very short backbones are usually sought after when electrical conductivity through the SAM is required [27]. However, for the most part the backbone usually plays a highly insulating role in most of applications [17], [28]. More exotic roles also exist, where the backbone is employed as a 2D organic-inorganic hybrid dielectric in self-assembled field-effect transistors (SAMFETs) [29], [30] or even as charge storage layers in thin film transistors memories [31].

2.3.3. Head group The head group being the outermost component (and therefore the most exposed) of the SAM, has the strongest impact in the overall functionality of the SAM. Common effects of head groups are the modification of wettability characteristics of the SAM or adjustment of dipole moment of the SAM molecule [32]. This properties are particularly important when the SAM is grafted onto 0D NSs, as these properties can play a dominating role in the stability of the 0D NSs in solution (Section 2.4). It is also the most diverse component in terms of chemical structures available. Head groups often vary from inert to reactive and even polymerizable groups, as well as polar, electrostatic, semiconducting, etc. (Figure 2.6). Fundamentally, whatever is considered as organic chemistry, can potentially be employed as a head group. Due to these reasons, it is perhaps the most interesting part of the SAM components. To put it another way, the head group is usually the only reason we want to employ a SAM in the first place. Spatial factors tend to be an issue with head groups, especially when the bulkiness of the head group is of a much greater magnitude than the anchor group. Having a SAM with a bulky head group may lead to a poorly packed and disordered SAM. A possibility to overcome this problem is to do a codeposition of mixed SAM. In such an approach, a SAM with a smaller head group fills the gaps left between molecules while consequentially acting as a support for the bulkier head groups [33], [34]. Similarly, a SAM exchange scheme where a bulkier SAM replaces a less bulky and weakly bound SAM is also a successful strategy [35].

2.3.4. Studied anchor groups Commonly employed SAM anchor groups for the modification of metal oxides include phosphonic acids (PAs), carboxylic acids (CAs) and catechols [12]. A study performed during this work revealed that; under ambient conditions (room temperature, neutral pH), phosphonic acids have a far stronger binding affinity to titanium oxide NSs in comparison to carboxylic acids and catechol groups [19]. The

12


same trend is generally true for a variety of other metal oxides we have employed [36]. However, mixed results were often obtained for the CAs and catechols anchor groups. This results are thoroughly described in Section 3.3. Since, it is critical for all of the applications discussed in this work to have a robust and reliable modification of the NSs surface. In this work, the majority of the experiments involve SAM molecules with phosphonic acids acting as anchor groups. Nonetheless, a few experiments involved carboxylic acids and catechols anchor groups and therefore they are described as well.

2.3.4.1. Phosphonic acids The first studies of PAs binding to alumina surfaces date back to the mid-late 1980's [37], [38]. Since then, the covalent binding mechanism of the PAs to a metal oxide surface has remained partly uncontested and is generally accepted as displayed in Figure 2.8. In short, the P-OH groups of the PA molecule react with the OH groups of the metal oxide surface. This results in an acid-base condensation reaction which has as a by-product H2O [39]–[43]. However, the final faith of the double bonded oxygen (P=O) is not so clear. Partly because of this, the final binding configuration of the PA to the metal oxide surface is somewhat of a gray zone. This is mainly due to all the possible binding configurations that arise from having three active motifs (Figure 2.9). The generally accepted trend is that, either a monodentate or bidentate non-chelating configurations are the dominant configurations (Figure 2.9a, c, i and j). The previous statement can be made from several theoretical and experimental studies of PAs grafted onto several metal oxides surfaces [39]– [43]. Moreover, in this thesis, similar observations were obtained via FTIR-ATR of functionalized NP powders with PA molecules (Section 3.2.1). Ultimately, it might as well be that all possible binding configurations co-exist, and that their state merely depends on the accessibility the PA group has to the surface anchoring hydroxyl sites. That being said, other factors such as the metal oxide employed, crystallinity (or lack of) and exposed plane should also be considered, since all of them can have an impact on the amount of hydroxyl groups available. Yet, hydroxyl independent reaction pathways of PA molecules have been reported on SiO2 wafer substrates (T-BAG method) [44], however, this requires harsher reaction conditions (140 °C).

Figure 2.8.: Phosphonic acid binding mechanism to a metal oxide surface [39]–[43].


13

)

)

)

)

)

)

)

)

)

)

)

)

Figure 2.9.: Various phosphonic acid binding modes. M = metal. a) and b) monodentate, c) and d) bridging bidentate, e) bridging tridentate, f) and g) chelating bidentate, h) chelating tridentate, i) j) k) l) other viable hydrogen bonding modes. Reprinted from [41] with permission from the American Chemical Society.

The strong binding affinity of PAs to metal oxides makes them excellent candidates for surface modification. This affinity translates into simpler chemistry, monolayer robustness and overall lower material requirements in order to achieve full surface coverage of the NSs [19], [36]. Lastly, due to their ease of synthesis and storage [41], [43] (long term ambient, no self-polymerization), they are widely commercially available in a variety of different configurations (on a par with thiols and silanes, minus the stench and difficult handling conditions respectively).

2.3.4.2. Carboxylic acids Just as PAs, the first studies of carboxylic acid SAMs on aluminum oxide substrates were reported in the mid-late 1980's [14], [45]. As is often the case with other anchor groups, the coordination of the carboxylic acid with the metal oxide surface is controversial and highly situational. Figure 2.10 displays various carboxylic acid binding modes. In an infrared spectroscopy study by K. Dobson et al. [46], it is stated that coordination of CA group to the metal oxide surface mostly happens as a deprotonated carboxylate (Figure 2.10d, e, f). The study included a variety of metal oxide surfaces such as TiO2, ZrO2, Al2O3, and Ta2O5. The substrates were functionalized in DI water under ambient conditions with several CA molecules containing only one carboxylic acid group. It was found that a strong binding was only present for the ZrO2 substrates, while very weak or not existent for the other substrates. They concluded that the higher affinity of the CA to ZrO2 was attributed to the higher IEP

14


(at pH 9) of the ZrO2 substrates. In contrast, TiO2 substrates exhibited their IEP at pH 5. Meaning the ZrO2 had a surface positive charge at neutral pH, allowing for more favorable reaction conditions. a)

b)

c)

d)

e)

f)

Figure 2.10.: Various carboxylic acid binding modes) M = metal. a) electrostatic attraction, b) H-bonds to bridging oxygen, c) H-bonds to carboxylic oxygen, d) monodentate metal-ester, e) bidentate bridging, f) bidentate chelating. Adapted from [12] with permission from Wiley.

In summary, in order for the carboxylic acid to form a strong bond with the metal oxide surface, it usually requires non-ambient conditions (pH, temperature) according to other studies [47]–[49]. However, it is also purely situational. Take for example the experiments described in Section 0, where CA SAMs formed a stable bond on indium tin oxide (ITO) nanoparticles and zinc oxide nanorods. Whereas, they did not do so on ITO and zinc oxide 2D substrates even when comparable reaction conditions were employed. The ample natural occurrence of carboxylic acids (e.g., fatty acids) is perhaps one of the most attractive characteristics of this anchor group. This natural occurrence results in a more "eco-friendly" accepted approach under the argument that similar molecules already exist in nature [49]. Furthermore, simpler synthetic routes when compared to phosphonic acids or silanes is also a desirable trait [12]. However, due to their weaker binding strength in ambient conditions [19], [36], [46] carboxylic acids have not yet found their way into mainstream commercial availability in terms of SAM molecules. Such status is currently held only by phosphonic acids and silanes for metal oxides, as well as thiols for metals.

2.3.4.3. Catechols The catechol is a relatively new anchor group, it was introduced in 2007 by a bio-mimetics study based on the remarkable adhesion of mussels to organic and inorganic surfaces [50]. Since then the catechol has been employed in numerous surface modification schemes [51], [52]. Despite abundant reports of catechol anchor groups being used for non-reversible modification of metal oxide surfaces via SAMs. In this thesis, we could not obtain credible evidence (Section 3.3) that supports the formation of a covalent bond between the catechol and metal oxide surface under the employed conditions (ambient, neutral pH) [19].


15

The suggested binding configurations obtained from literature [12], [53] of the catechol group to a TiO2 surface are shown in Figure 2.11. Important to note the EWG group, which stands for "electron withdrawing group". The EWG plays two roles. It prevents the oxidation of the catechol group which would otherwise render it useless as acknowledged by M. Rodenstein et al. [54]. Furthermore, according to B. Malisova et al. [55], the EWG plays a critical role on the reactivity of the catechol since it can directly impact the acidity (pKa) of the catechol group. It was concluded by B. Malisova et al. that the ideal pH conditions for catechol reactivity occur when the metal oxide surface is as close to the isoelectric point (IEP) as possible, as well as to the IEP of the catechol, which varies depending on its pKa. Therefore, the EWG can be used to tune the catechol anchor group disassociation constant in order to be employed with a specific metal oxide in function of its IEP. a)

b)

c)

d)

e)

Figure 2.11.: Miscellaneous catechol binding modes on titanium oxide surface. EWG = electron withdrawing group. a) H-bonds, b) monodentate with H-bond, c) bidentate chelating, d) monodentate with bridging H-bond, e) bidentate bridging. Adapted from [12] with permission from Wiley.

As is the case with CAs, the binding of the catechol to the metal oxide surface should be approached on a case by case basis. Often, the reported reaction conditions for the catechols vary greatly and results are often controversial. From a purely practical point of view, there is no particular good reason to employ a catechol anchor group whenever carboxylic acid anchor group is also available. However, the catechol is still a young and perhaps misunderstood anchor group. More research is required to fully understand its employment as a SAM molecule anchor group.

2.4. Stability of 0D and 1D NSs in solution Stability is a pivotal property of 0D and 1D NSs in solution. The primordial reason for this is, that having an unstable dispersion renders it completely useless in terms of solution processability. Therefore, having a stable dispersion is an especially sought-after attribute. By stability, we refer to the ability of the NSs to remain dispersed in solution over a period of time without precipitating to the bottom of the flask. Curiously, the expression "stable dispersion" is frequently used in vague terms and

16


the stability of the dispersion can often vary from years to a mere few hours. Figure 2.12 shows images of nanoparticle dispersions with varying degrees of stability. a)

b)

d)

c)

e)

Figure 2.12.: Photographs of TiO2 and Fe3O4 nanoparticle dispersions with varying degrees of stability. a) Unstable TiO2 dispersion during flocculation. b) Non-transparent stable TiO2 dispersion. c) Transparent stable TiO2 dispersion. d) Unstable, already flocculated Fe3O4 dispersion. e) Transparent stable Fe3O4 dispersion.

NSs dispersed in solution are under constant movement (Brownian motion) which makes them bump into each other all the time. When such collision occurs, the NSs have a propensity to attract each other (agglomerate) in a process known as Ostwald ripening. Essentially, such process occurs simply because bigger particles (agglomerates) are more thermodynamically stable due to their lesser surface area [56]. As more and more NSs collide with each other, eventually their size increases reducing their solubility which causes flocculation to occur rather rapidly. However, this effect can be avoided if supplementary forces are present that directly counteract the attractions of the NSs or even more, avoiding the collisions altogether. If such forces are implemented successfully, they may delay or even completely prevent the flocculation process. To this end, there are two main mechanisms to avoid NSs agglomeration in solution known as electrostatic and steric stabilization.

2.4.1. Electrostatic stabilization Electrostatic stabilization of 0D and 1D NSs is in concept very simple. It consists in charging the surface of the nanoparticles, therefore, forming an electrostatic barrier as shown in Figure 2.13a. These particles having equal charge, when being in proximity of each other they repel themselves due to


17

coulombic interactions as shown in Figure 2.13b. The greater the absolute magnitude of the surface charge, the greater the repulsion, which results in better dispersion stability. The exact magnitude of this surface charge is hard to measure. Therefore, it is measured by a property of colloidal dispersions (and surfaces also), known as zeta (ζ) potential. The zeta potential is directly related to the charge of the surface. However it does not measure the exact charge present at the surface. Examples of dispersions stabilized via electrostatic interactions can be found in Section 4.5.

a)

b)

Charged species on surface

Electrostatic barrier

Electrostatic nanoparticle repulsion

Figure 2.13.: Schematic representation of electrostatic stabilization of nanoparticles. a) A nanoparticle with charged species at its surface. b) Nanoparticles having an equal charge repel each other avoiding agglomeration.

2.4.1.1. Zeta potential The zeta (ζ) potential is a property of any surface in solution. However, under the context of this work, it is purely related to the surface charge of 0D or 1D NSs dispersed in solution. The zeta potential is measured in millivolts (mV). It is generally accepted that dispersions having a zeta potential with an absolute magnitude higher than 20 mV already form moderately stable dispersions. Whereas, outstanding stability is usually found for dispersions with potentials higher than 40 mV. The latter is only true under the concept of an electrostatic stabilization scheme, as highly stable dispersions with a zeta potential of 0 mV can be obtained via a steric stabilization system. To better understand what the zeta potential is, Figure 2.14 depicts the electric double layer (EDL) model. The EDL is the classical model by which the zeta potential is conveniently explained. The model starts with a charged nanoparticle surface and then it proceeds outwards from the nanoparticle surface in arbitrary units. Immediately after the charged surface we find the Stern layer. This shell is comprised of any polar species that have opposing charge from the surface. As we continue outwards, the influence of the surface charge diminishes gradually forming an ever weaker bound shell of polar

18


species. At which point, even polar species of opposite magnitude may exist but in lesser number. Finally, we reach the slipping plane which is defined as the point until the whole system acts "as one", meaning without any influence from the bulk phase (solvent). It is only until this point that there is clear distinction between the phases. It is here, that it is possible to measure a nanoparticle "charge" or more fittingly the zeta potential. Therefore, prior to the slipping plane there is no notable distinction of the systems. As a result, it is difficult to measure the real surface charge. It now becomes clear (in concept) that zeta potential is not completely synonymous to surface charge, instead just related.

Charged surface

Stern layer Slipping plane

Surface potential Stern potential

ζ (mV)

0

ζ potential

Distance from surface

Figure 2.14.: Schematic representation of the electric double layer (EDL) on a nanoparticle. Red and blue spheres represent charged species of opposite magnitude.

Lastly, a brief mention of the most relevant mechanisms by which a zeta potential can arise [57] was compiled. 1. Ionization of surface groups. 2. Difference of electron affinity between the dispersion phase and the nanoparticle surface. 3. Physical entrapment of non-mobile charge. 4. Preference of a phase for ions of a specific charge.


19

Mechanism 1 is one of the most relevant mechanisms when dealing with metal oxides and it is described in more detail in the immediate section below. As to mechanism 2, we can find some evidence suggesting that this effect is the cause of the zeta potential of the nanoparticles described in Section 3.2.3.3. We can also find mechanism 3 very clearly in action in Section 4.5. Whereas, no situation in this work involved mechanism 4. It is possible however, that these effects may combine to either enhance or hinder the overall zeta potential. Therefore, a discrete distinction between the mechanisms that may be at play in a particular situation is often difficult to pin point.

2.4.1.2. Isoelectric point and ionization of surface groups In the absence of any adsorpted species at the surface of a metal oxide nanopowder, the hydroxyl groups provide a system that is often employed for electrostatic stabilization of 0D and 1D pristine metal oxides. The isoelectric point (IEP) of a metal oxide is defined as the point of neutral surface charge. In aqueous solutions, hydroxyl terminated surfaces, such as metal oxides, can exhibit a surface charge in function (mainly) of the pH of the solution (Figure 2.15). The charge arises from the protonation (OH2) or deprotonation (O-) of the hydroxyl groups of the metal oxide surface. Therefore, having a pH level below the IEP of the metal oxide results in the protonation of the surface hydroxyl groups. Whereas, a pH above the IEP results in the deprotonation of the hydroxyl groups. This results in a positively or negatively charged surface. Moreover, a protonated or deprotonated surface can play a crucial role in reactivity of weaker anchor groups such as carboxylic acids and catechols as already explained in Section 2.3.4.

+

ζ pH

(mV)

-

Isoelectric point

Figure 2.15.: Schematic representation of the isoelectric point (IEP) in function of pH.

The IEP between metal oxides varies greatly, since it is heavily influenced by other factors such as: Surface impurities, crystallinity and even physical factors like temperature [58], [59]. Because of this,

20


values for specific materials in the literature often differ from each other [60]. It is not uncommon to obtain the IEP of a metal oxide powder by measuring the zeta potential of aqueous nanoparticle dispersions of varying pH. The pH at which the zeta potential = 0 mV (dispersion will most likely precipitate) is then considered to be IEP.

2.4.2. Steric stabilization Steric stabilization, as its name implies, refers to a stabilization mechanism that relies on spatial hindrance. The stabilization is provided by a shell of less dense material surrounding the far denser 0D or 1D NSs core. This steric barrier (Figure 2.16) prevents collisions between the dispersed NSs cores, effectively preventing any attraction between them. In addition, the shell providing the spatial hindrance at the same time plays a chemical stabilization role via solvation effects [61]. This effect of increased solvation of the NSs can effectively thwart nanoparticle core interactions. In other words, it can provide an intermediate interface, by which the normally insoluble core would now be "more in phase" with the dispersion medium as it would be in its pristine form. The nature of this steric barrier preventing nanoparticle collision is usually an organic layer that surrounds the nanostructures dispersed in solution. To this end, the organic molecules may form monolayers or multilayers around the nanostructure depending on their nature.

Grafted molecules on surface

Steric barrier Figure 2.16.: Schematic representation of steric stabilization of nanoparticles. Physically, the molecules grafted onto the nanoparticle avoid direct nanoparticle collision and nanoparticle core interaction. Chemically, the molecules provide solvation in the dispersion media effectively thwarting nanoparticle core interactions.

It is fairly common that oligomers or polymers are employed for this purpose. Like for example, a myriad of polyethylene glycol (PEG) derivatives [62]. It goes without saying, that SAMs are the perfect toolkit for this purpose as they allow for a precise monolayer deposition onto the NSs surface.


21

An example application of such is presented in this work (Section 4.1.1), in which a phosphonic acid molecule with a small ethylene glycol chain was employed as a steric stabilization agent [17]. A similar example can also be found at Section 4.5 as well.

2.4.3. Electrosteric A combination of both electrostatic and steric stabilization schemes is also possible. This occurs when the "polar nature" of the layer surrounding the nanoparticle is of a non-negligible magnitude. While at the same time this layer also provides a non-negligible level of steric stabilization. Examples of electrosteric stabilization usually involve the usage of polyelectrolytes wrapped around the surface of the NSs providing both steric and coulombic repulsion forces [63]. Not surprisingly, electrosteric stabilization with SAM molecules is also possible. Electrosteric stabilization with SAMs can be achieved in two main ways, the SAM molecules have an ionic moiety of some sort at the backbone or head group, or by dipole moment alignment of the molecules on the surface [64]. Both effects were observed in several occasions during the course of this work, and they are catalogued in Section 3.2.3.3 and 4.5. Lastly, under specific circumstances we hypothesized, that the geometry of the dispersed NSs can have an influence in the stabilization mechanism of the NSs dispersions. This is explained in more detail in the following section.

2.4.4. Impact of size and geometry of nanostructures The size and geometry of 0D and 1D NSs play a critical role on the structural properties of the SAM. These structural differences may also affect the dispersibility of the NSs. In order to understand this, let us take a 5 nm and 50 nm spherical nanoparticles with a grafted SAM as the prototypical examples of 0D NSs. Figure 2.17 represents cross-sectional drawings of the functionalized spherical nanoparticles. In these drawings, the SAM molecules were equally distributed perpendicularly among the spherical surface. The molecules were placed at a grafting density of 6 molecules per nm2 and consist of octadecylphosphonic acid (C18-PA) (Appendix Figure 7.1). At first glance, they do not appear to be much different other than the core size (Figure 2.17a, c), however, when having a closer look, things are bit different (Figure 2.17b, d). The curvature of a sphere is defined by the sphere's radius, therefore, as the nanoparticle radius becomes smaller, the surface curvature of the nanoparticle increases (Figure 2.17b). This is important because as the curvature increases so does the free volume between each SAM molecule [65]. This extra space negatively affects the weaker intermolecular interactions (van der Waals) of the SAM molecules that can only exist when the molecules are in close proximity from each other. Therefore, it becomes increasingly difficult for the SAM molecules to form highly ordered, densely packed monolayers. Furthermore, smaller NSs posses a higher surface area, which increases the free energy of the system and usually leads to a higher degree (faster) of agglomeration [66], [67].

22


a)

b)

5 nm

c)

d)

50 nm Figure 2.17.: Nanoparticle surface curvatures. a) and c) are nanoparticle illustrations (up to scale) of functionalized 5 and 50 nm spherical particles. b) and d) represent a zoomed-in up to scale nanoparticle surface illustration. It becomes apparent in illustrations b) and d) how can surface curvature play a critical role on SAM crystallinity and SAM dipole moment alignment. Also of importance to note, the free space available between the molecules.

In summary, having a high surface curvature leads to the following situations which can strongly impact the solubility of the NSs. 1. Having the SAM molecule backbone and head group more exposed allows for better solvent accessibility. 2. Having more space available between molecules can facilitate surface access and therefore resulting in a higher grafting density (not SAM packing density), especially when compared to a flat surface. 3. The cooperative effect of dipole moment alignment of the SAM molecules may be completely mitigated due to reduced molecule alignment [64]. If there mechanisms are true, a steric stabilization scheme is favored for NSs with a high surface curvature due to the improved solvation provided by the more readily exposed SAM. On the other


23

hand, when employing NSs with a smaller surface curvature, SAM exposure would be limited since the SAM forms a more ordered and denser monolayer. In which case, a steric stabilization scheme is hindered to a certain degree due to the more limited exposure of the organic shell. Furthermore, in NSs with low surface curvature (bigger particles), the effect molecule alignment may have to be factored into the proposed stabilization mechanism. Particularly, for molecules with a strong dipole moment. This is because of the potentially attainable higher order and higher packing density of the SAM. We hypothesized, that under such circumstances the combination of both stabilization mechanisms come into play. Therefore, an electrosteric stabilization seems to be the most viable scenario. In which, the SAM molecules provide certain degree of steric stabilization, while the dipole moment alignment of the SAM polarizes the surface resulting in an electrostatic stabilization at the same time. However, this remains as a hypothesis as the effect of the ligand dipole moment was not fully investigated. Still, the exact implications of SAM dipole moments are a debated topic [64], [68], [69]. Section 3.2.3.3 provides some experimental insight regarding this topic from a nanoparticle perspective.

2.5. Thin film transistors In short, a transistor is an electronic switch which is operated by modulating a voltage. This is the case of field effect transistors (FETs), or by modulating a current in the case of bipolar junction transistors (BJTs). In the matter at hand, which is thin film transistors (TFTs), they are operated by voltage modulation which classifies them as a type of FET. Structurally, a key difference between FETs and TFTs is that all the components of a TFT are built upon an insulating surface which exclusively acts as carrier substrate. Whereas, in traditional FETs, the substrate (typically a silicon wafer) acts as an active component. TFTs as well as FETs can be built in a variety of different configurations and complications, each one with both their pros and cons. Plenty of other technicalities between them exist, however, describing them would be beyond the scope of this thesis [70]. Therefore, it is restricted purely to the configurations employed during this work. Figure 2.18 shows a schematic representation of a bottom gate staggered n-type TFT. The TFT consists of 3 electrodes namely source, drain and gate. When the transistor is in its "off" state, the passing of electric current from drain to source (IDS) is impeded due to the low conductivity of the channel. However, when the transistor is in its "on" state, the conductivity of the channel is increased several orders of magnitude allowing a higher IDS to pass. The amount of current passing through the channel is modulated by the modulation of the VGS voltage. The proportion of VGS vs. IDS begins with a linear behavior known as the "linear regime" and it eventually reaches a non-linear regime called the "saturation regime". These regimes are described by Equation 2.4 for the linear regime and 2.5 for the saturation regime, where µ and C are the carrier mobility within the regime and capacitance of the dielectric respectively. A parameter of particular importance is the carrier mobility (µ) in the

24


saturation regime (Equation 2.6), since this characteristic is used in literature as the benchmark by which transistor performance is categorized.

Channel Source

Drain W

-

L

+

VGS

Gate

Dielectric

Carrier substrate

-

+

ID

VDS Figure 2.18.: Schematic representation of a bottom gate TFT

Linear regime

𝑊𝜇𝐶 𝑉𝐷𝑆 2 )𝑉 𝐼𝐷 = [(𝑉𝐺𝑆 − 𝑉𝑡ℎ 𝐷𝑆 − ] 𝐿 2

Saturation regime

Saturation mobility

𝐼𝐷 =

𝑊𝜇𝐶 (𝑉𝐺𝑆 − 𝑉𝑡ℎ )2 2𝐿

𝜇𝑠𝑎𝑡 =

2𝐿𝑚2 ∙ 𝑊𝐶

2.4

2.5

2.6

2.6. Chapter summary The essential theoretical aspects required for the discussion of the experimental aspects of this thesis have been covered. A clear distinction between nanostructures (NSs) and nanostructured materials (NSMs) was explained as well as the system employed for their classification. The relation of surface area to volume of NSs was summarized as it is one of the most important characteristics of NSs when dealing with functionalization. The concept of self-assembled monolayers (SAMs), their composition and working principle were discussed as well. This was followed by a description of the impact of grafting a SAM onto the surface of NSs in terms of solution dispersibility. The role that the geometry of 0D and 1D NSs might play in NSs dispersibility was also taken into account. The mechanisms behind stable and unstable NSs dispersions were also described. Lastly, the general working principle of thin film transistors (TFTs) and characterization equations employed for their benchmarking were described.

3. Functionalization and characterization of NSs

This chapter focuses on the key factors involved during the modification of the NSs surface, such as: starting material, complete surface coverage, impact of the particle dimensions, deposition of mixed ligand monolayers and anchor group stability. Special attention is given to the high surface to volume ratio of 0D and 1D NSs. Since it plays a critical role during functionalization and is easy to underestimate. But first, as a basis for discussion, the relatively simpler functionalization of 2D NSs is briefly described.

3.1. Functionalization of 2D materials Deposition of SAMs from solution onto 2D substrates is a fairly straightforward procedure (Figure 3.1). First, the metal oxide surface is exposed to an oxygen plasma treatment (5 min, 0.2 mbar, 200 W). This helps to activate the OH groups at the oxide surface as well as cleaning other organic species that may be present. The plasma treated substrate is then immersed into the desired SAM solution (IPA, 0.2 mM) for 24 hr to ensure a tightly bound SAM is formed [36]. a)

O2 plasma

b)

c) Remove from solution

Rinsed (IPA) and dried (N2, hotplate @60 C) substrate with final SAM

AlOx , ITO, ZnO or TiO2 2D substrate Immerse 2D substrate in SAM solution for 24 hours

24 hrs

Figure 3.1.: Steps for SAM deposition on 2D substrates. a) Plasma treatment of the surface. b) immersion of the substrate into the SAM solution. c) Schematic of the finalized SAM.

25

26

Functionalization and characterization of NSs

Finally, the substrate is taken out of the solution and is vigorously rinsed with pure solvent (the same solvent used for deposition) to remove any excess unbound molecules. After rinsing, the substrate is blown dry with nitrogen and heated to 60 °C for a few minutes to remove any remaining solvent at which point the process is finished. As an added step, the quality of the deposited SAM is verified via a few quick SCA measurements.

3.2. Functionalization of 0D materials The deposition of a SAM onto a dispersion of 0D or 1D NSs is also a fairly simple procedure once suitable conditions have been identified (Figure 3.2). However, in order to properly determine these conditions an experimental approach is required. This section describes in detail the step by step characterization from beginning to end as performed during this work. Ideally, the functionalization process begins with a highly stable dispersion of material (Figure 3.2a). A highly stable dispersion assures that the surface of the NSs is exposed to the dispersion medium because little or no agglomeration is present. This guarantees the SAM molecules access to NSs surface. Nevertheless, whenever a dispersion with poor stability must be functionalized, the process of sonication during the functionalization step (Figure 3.2b) can help to compensate for the poor stability of the dispersion. In some cases, the functionalization improves the dispersibility of the NSs, in this case the functionalization process itself will gradually reduce the agglomeration. This has as a consequence, the full exposure of the NSs surface. It must be noted that for the purpose of functionalization and under the same conditions, the sonication process proved to be more valuable than magnetic stirring. Particularly, for magnetic nanoparticles were magnetic stirring is not feasible. The detailed differences between stirring and sonication are not discussed. It suffices to say that stirring often resulted in poorer SAM surface coverage. After the sonication process is finished, any excess unreacted SAM molecules must be washed away (Figure 3.2c). This is achieved by centrifugation of the NSs followed by removal of the supernatant. Afterwards, the addition of a washing solvent to the centrifuged NSs is performed. A washing solvent is a solvent in which the SAM molecules are highly soluble and cannot be isolated by centrifugation. This process is repeated at least twice to guarantee full removal of the excess molecules. When centrifugation is not possible (e.g. highly stable smaller particles), the use of an anti-solvent (technically an anti-dispersant) in excess can be used to encourage the precipitation of the particles. Alternatively, evaporation of the solvent under vacuum and heat in a rotary evaporator (rotavap) can also be used for exchanging the functionalization solvent to a washing solvent. After the washing procedure is finished, the nanoparticles are dispersed in the solvent of choice or kept as a powder for storage or characterization.


a)

27

b)

c) Centrifuge NSs, remove supernatant and redisperse NSs in fresh solvent (x2)

Add SAM molecule in required concentration

Dispersed pristine metal oxide 0D or 1D NSs

Sonicate dispersion for 30 min at room temperature

Functionalized NSs with no free excess SAM molecules

30 min sonication

Figure 3.2.: Steps for SAM deposition on 0D and 1D NSs. a) Pristine nanostructures dispersed in a liquid medium. b) The SAM molecule is added and with the aid of sonication it forms a SAM around the nanostructures surface. c) The final functionalized NSs dispersion (after washing) with no unbound SAM molecules present in solution.

3.2.1. Chemical characterization (FTIR-ATR) Since the sonication process allows for the use of both stable and unstable dispersions. The first real concern is to evaluate the "pristine" NSs surface. The main worry to assess is the presence of any other sort of physisorbed or chemisorbed species to the NSs surface. Some species can hinder or even negate the adsorption of the SAM molecules to the NSs surface. Specially, since plasma cleaning cannot be performed on 0D and 1D NSs dispersions. Therefore, a truly pristine NSs dispersion with no adsorpted species on their surface is highly desirable for functionalization purposes. Alternatively, NSs with weakly bound species that can be removed by the SAM molecules are also suitable for functionalization. Therefore, a pre-functionalization chemical characterization of the NSs is required. This is followed by a post-functionalization characterization by which it can probed whether the SAM molecules are present on the NSs surface or not. To this end, FTIR-ATR is an excellent technique which allows to obtain chemical information regarding the makeup of the NSs. In order to conduct the pre or post functionalization characterization of the NSs via FTIR-ATR, the NSs need to be isolated as a dry powder. Exclusion of the solvent is paramount, as it would interfere

28


with the FTIR measurements. This is achieved by centrifugation of the NSs followed by removal of the supernatant and drying overnight. The drying process was performed in a dry air oven at a temperature close to the boiling point of the solvent. Alternatively, if the solvent has a high vapor pressure, simply leaving the wet centrifuged nanopowder exposed to the negative pressure of a chemical hood is sufficient. Another effective drying route is drying the nanopowder under vacuum

T (%)

and moderate heat with a rotary evaporator.

a)

a)

b)

b)

c)

c)

3800

3600

3400

3200

3000

2800

Wavenumber (1/cm)

2600 2000

1800

1600

1400

1200

1000

800

Wavenumber (1/cm)

d)

Figure 3.3.: FTIR-ATR spectra of several commercial TiO2 nanoparticles. a) Featureless spectrum of pure TiO2 particles. b) Spectrum of HNO3 stabilized TiO2 particles with signals from the HNO3 species present c) Spectrum of allegedly pure TiO2 particles containing signals that are unaccounted for.

Once a dry powder is obtained, the measurement can be performed. Figure 3.3 shows the FTIR-ATR spectra of several commercial TiO2 nanoparticles isolated and dried as received. The ideal case is shown in Figure 3.3a, which is the featureless spectrum of pure TiO2 nanoparticles. In this case, we can be sure that the particles are indeed pure and suitable for functionalization. Likewise, similar flat spectra is also obtained for other pure metal oxide nanopowders. In contrast, Figure 3.3b shows a feature rich spectrum of TiO2 nanoparticles. This particles contain an arbitrary amount of nitric acid (HNO3) on their surface for stabilization purposes (according to manufacturer). Lastly in Figure 3.3c, the spectrum of allegedly pure TiO2 nanoparticles is portrayed. However, as can be observed from the spectrum, the particles contain some unknown adsorpted species. To conclude, it can be safely assumed that the particles from Figure 3.3a can be used for functionalization. Whereas, the particles from Figure 3.3b and Figure 3.3c can conceivably be employed. But only, if the SAM molecule is able to replace the present species on the nanoparticles surface. To this end, all the particles mentioned


29

were functionalized with C16-PA (Figure 3.3d) independently of their purity assessment. After

T (%)

functionalization, the particles were isolated and dried for measurement.

a)

a)

b)

b)

c)

c)

v 3800

3600

3400

3200

3000

2800

Wavenumber (1/cm)

2600 2000

1800

1600

1400

1200

1000

800

Wavenumber (1/cm)

d)

Figure 3.4.: Color coded FTIR-ATR spectra of several commercial TiO2 nanoparticles after being functionalized with C16-PA. a) Spectrum of pure TiO2 particles plus the signals from C16-PA. b) Spectrum of HNO3 stabilized TiO2 particles with the signals from C16-PA. Note that the HNO3 peaks are no longer present. c) Spectrum of allegedly pure TiO2 particles functionalized with C16-PA still containing signals that are unaccounted for, plus the overlapping signals of the C16-PA. d) Chemical structure of C16-PA.

As expected for the pure TiO2 particles, the functionalization was successful. The presence of the bound C16-PA is evident as depicted by the bands present in the post functionalization FTIR spectrum in Figure 3.4a. The blue circles represent the bands related to the methylene groups of the alkyl chain, while the red circle pinpoints the band associated with the methyl group at the end of the chain. The green circle represents the bound phosphonic acid to the TiO2 nanoparticles surface. The FTIR-ATR spectrum of the pure C16-PA SAM molecule can be found at the appendix in Figure 7.2 and Figure 7.3 for a more in depth comparison. The next example is the HNO3 stabilized nanoparticles (Figure 3.4b), in here we can observe that the typical bands of the C16-PA are also present suggesting a successful functionalization. But more importantly, the bands belonging to the HNO3 are barely noticeable after functionalization. This means that the C16-PA was able to replace the nitric acid in its majority, which makes these "impure" particles also a good choice for functionalization with phosphonic acids. Finally, Figure 3.4c shows a partially successful example, in which both the unknown bands of the particles and those of the C16-PA are now overlapped. In this case, the C16-PA was not able to remove the unknown species from the particle surface. Instead, both the C16-PA and the unknown species were present at the surface. Therefore, these particles were never used in any further experiments. Since it is

30


of paramount importance to have full knowledge and control of the nanoparticles surface after functionalization. Finally, we take a look at Figure 3.5, it shows a few other metal oxide NSs characterized in the very same way as explained in this chapter. A clear trend can now be identified between the pure metal oxides and C16-PA functionalized ones. All the functionalized powders exhibit a broad signal from 1200 to 900 cm-1, which is attributed to the overlap of the P-O and P=O vibrations of the phosphonic acid group bonded to the metal oxide surface [71]–[73]. The overlap comprises signals of bonded and non-bonded P-O and P=O groups, owing to the likelihood of mono-dentate, bi-dentate and tri-dentate binding modes co-existing on the particles surface [12], [42]. The two peaks around 2920 and 2850 cm-1 correspond to the methylene groups vibrations of the C16-PA aliphatic carbon chain, followed by a small shoulder appearing at 2970 cm-1 attributed to methyl groups at the tail of the alkyl chain. A less intense peak in the 1470 cm-1 region is recognized as methylene groups scissoring vibrations.

T (%)

Pristine Ceria

CeO2

C16-PA Ceria

T (%)

Pristine Iron oxide

Fe3O4

C16-PA Iron oxide

T (%)

Pristine ZnO Nanorods

ZnO

C16-PA ZnO Nanorods

3800

3600

3400

3200

3000

2800

Wavenumber (1/cm)

2600 1800

1600

1400

1200

1000

800

Wavenumber (1/cm)

Figure 3.5.: Exemplary FTIR-ATR spectra (color coded) of various pure and C16-PA functionalized metal oxide NSs. The general trend of a featureless spectrum for pure metal oxides can be observed. An unmistakable trend is also identifiable after functionalization with C16-PA in all metal oxides. The molecular structure of C16-PA is shown atop.


31

A short hypothesis regarding the binding configuration of the phosphonic acid As mentioned before, the origin of the broad band from 1200 to 900 cm-1 was attributed to the bound phosphonic acids to the metal oxide surface. It was mentioned as well, that it is thought to be an overlap of all the diverse binding configurations. To illustrate this point further, Figure 3.19a shows the spectrum of Fe3O4 nanoparticles functionalized with C16-PA. The characteristic broad band belonging to the bound phosphonic acid can also be observed. In Figure 3.19b the spectrum of the pristine C16-PA is shown, in here the bands corresponding to the free phosphonic acid are clearly defined (1200 to 900 cm-1). The P-O and P=O bands are clearly defined, because these chemical groups can only exist in their free forms within a limited set of configurations. However, a bound phosphonic acid to a metal oxide surface can exist in a variety of different configurations (Section 2.3.4.1). The latter results in a broad band rather than in discrete set of states. This broad band could originate from the overlap of all phosphonic acid configurations as it is hypothetically depicted in Figure 3.19c. c)

a) Fe3O4 C16-PA

b) C -PA 16

2000 1800 1600 1400 1200 1000 800

600

Wavenumber (1/cm) Figure 3.6.: a) FTIR-ATR spectrum of Fe3O4 nanoparticles functionalized with C16-PA. b) Spectrum of pristine C16-PA. c) Hypothetical schematic depicting the diverse vibrations of the phosphonic acid bound to a metal oxide surface.

Etching of metal oxides As it will be explained throughout this chapter, the phosphonic acid is the most promising of the anchor groups due its strong binding to the surface of metal oxides (Section 3.3). However, this strong binding comes with a price, as it often results in the etching of the metal oxide NSs. In general, it is difficult to define the chemical stability of metal oxides. Some metal oxides are more prone to etching than others under a particular set of intricate conditions [74]. However, one particular condition that concerns the functionalization of metal oxides with phosphonic acid SAMs, is the formation of metal salts by the employment of n-alkyl phosphonic acids. The high complexing capability of phosphonic acids with metal ions, which results in the formation metal salts is well documented [75], [76]. Therefore, metal oxides which contain metal ions in their composition are particularly sensitive. To

32


illustrate this phenomena, a case where Fe3O4 nanoparticles were partially etched with phosphonic acids is showcased. FTIR-ATR is valuable tool to detect the etching of metal oxides after being exposed to a phosphonic acid SAM. This is demonstrated in Figure 3.7, which portrays the FTIR-ATR spectra of Fe3O4 functionalized with increasing concentrations of C16-PA. The first spectrum (Figure 3.7) shows the typical featureless signal of pristine Fe3O4 nanoparticles. As we proceed downwards in the spectra, the amount of C16-PA was doubled every time. At the concentrations of 10, 20 and 40 mM, the familiar smooth valley of the bound phosphonic acid is found between the wavelengths of 1200-900 cm-1.

Pristine

10 mM

T (%)

20 mM

40 mM

80 mM

160 mM

3800 3600 3400 3200 3000 2800 2600 2000 1800 1600 1400 1200 1000 800

Wavenumber (1/cm)

600

Wavenumber (1/cm)

Figure 3.7.: Fe3O4 nanoparticles functionalized with increasing concentration of C16-PA. As the concentration increases, the valley corresponding to the anchored phosphonic acid (red square) changes.

However, as the concentration increases the smoothness of the valley changes into a feature rich valley. This change, from a broad band to a set of discrete bands was hypothesized to be caused by the etching of the nanoparticles. The reasoning behind this is, that the metal salt has a defined chemical configuration. This results in the appearance of well defined P-O and P=O vibrations in the FTIR-ATR spectrum, as seen in the 80 and 160 mM concentrations. In contrast, the binding configuration of the phosphonic acid to a metal oxide surface is very diverse. Therefore, resulting in a single broader band


33

as explained before. Experimental proof that band overlapping can have the effect described on FTIR-ATR spectra can be found in Section 3.2.3.2. Previous work, done in partial collaboration with Johannes Hirschmann [77], identified the same trend in FTIR-ATR spectra. In an attempt to functionalize ZnO NSs, he exposed ZnO nanoparticles and nanorods to a phosphonic acid SAM. In parallel, he purposely synthesized the equivalent Zn phosphonate salt. When the spectrum of the Zn salt was compared to that of the presumably phosphonic acid functionalized ZnO. He realized that the spectra were exactly identical. Therefore, the only viable conclusion was the etching of the ZnO NSs into Zn salts. In conclusion for this section, the use of FTIR-ATR to assess the purity of metal oxide materials has been demonstrated. It is highly advisable that before any attempts to perform functionalization, an assessment of the purity of the starting material is made. FTIR-ATR is also a valuable tool for post functionalization evaluation, as it can provide chemical information about the surface of the NSs. Two hypothesis were briefly declared. One regarding the binding configuration of the phosphonic acid as well as one on the subject of the etching of metal oxides with phosphonic acids. However, FTIR-ATR analysis provides a purely qualitative characterization in terms of the nature of material present. Making a quantitative characterization via FTIR-ATR is difficult. However this can be easily measured via thermogravimetric analysis (TGA), which is now explained.

3.2.2. Saturation threshold (TGA) As we have already discussed in section 2.2, the surface area of 0D and 1D NSs is far greater than that of 2D NSs. Therefore, the amount of molecules required for a full surface coverage of 0D and 1D NSs is much larger. The amount of molecules can be estimated by taking the SSA of the NSs into account. However, it is regularly the case that the amount of molecules required for full coverage is the equivalent of several monolayers. This occurs because the volume of the solvent employed can also play a competing role for the NSs surface. Therefore, rather than an amount of molecules, it comes down to the concentration of the SAM molecule vs. the surface area available [19]. Therefore, simply estimating the amount of molecules required is not sufficient to guarantee full surface coverage. Instead, an experimental approach using thermogravimetric analysis (TGA) can be used to determine the conditions at which the surface of the NSs is fully saturated. In these experiments, the SAM molecule C16-PA was employed again (Figure 3.8a). C16-PA is a very good molecule for determination of surface coverage for a few reasons. It contains no bulky backbone or head groups, with the anchor group being the bulkiest. Therefore, a spatial limitation of the grafting density will be limited by the anchor group itself and not by any other groups. Secondly, C16-PA contains no other polar groups which may strongly interact with the NSs surface or have a strong intermolecular interactions. As these interactions can interfere or modify in differing forms the self-assembly process [78], [79]. In addition, the alkyl chain backbone provides enough rigidity to keep the molecule from bending and

34


interfering with the SAM process. Another important quality is that C16-PA has enough molecular weight to be easily detected in TGA measurements. Furthermore, C16-PA is readily and economically available by the hundreds of grams.

a) b)

100 98

Weight (%)

96 94 92

Pristine 2.5 mM 5 mM 10 mM 20 mM 40 mM

90 88 86 0

100

200

300

400

500

600

700

Temperature (°C)

Figure 3.8.: a) Chemical structure of C16-PA. b) TGA under N2 of AlOx NPs functionalized with different concentrations of C16-PA. Adapted from [16] with permission from the American Chemical Society.

Figure 3.8b shows TGA measurements of a fixed amount of AlOx nanoparticles functionalized with increasing amounts of C16-PA. The higher the concentration, the more molecules that attach to the surface. It can be observed from Figure 3.8b how the increasing concentration of C16-PA results in a greater and greater mass loss. This mass loss originates from the decomposition of the grafted C16-PA due to the gradual increase of temperature. In contrast, there is no dramatic mass loss for the pristine nanoparticles (black line). Considerable increments of mass loss are consistent until the 10 mM concentration is used, at which point the higher concentrations of 20 and 40 mM did not result in a significantly greater mass drop. It is at these concentrations (10, 20 and 40 mM) that we can declare that the particles surface is fully saturated and no more molecules can attach to the surface.

3.2.2.1. Grafting density The grafting density of the molecules can be calculated from Equation 3.1. In which wt represents the mass loss measured by TGA, N is Avogadro's constant, MW is the molecular weight of the SAM molecule employed and SSA is the specific surface area of the NSs. This is a particular useful tool, since it allows to verify the success of the functionalization procedure [16]. For example, Table 3.1


35

summarizes the calculated grafting densities from the TGA measurement of Figure 3.8. The resulting grafting densities are in good agreement with literature [80], but most importantly they are of a reasonable magnitude (3-7 molecules per nm2). When the maximum grafting density value is reasonable, unwanted side effects of functionalization such as NSs etching or multilayer coating can be discarded. However, if the calculated grafting densities are out of range, the functionalization process needs to be re-evaluated under more favorable conditions. 𝑤𝑡 𝑁 𝑔𝑟𝑎𝑓𝑡𝑖𝑛𝑔 𝑑𝑒𝑛𝑠𝑖𝑡𝑦 = ( )( ) 100 − 𝑤𝑡 𝑀𝑊 ∙ 𝑆𝑆𝐴

3.1

Table 3.1.: Calculated grafting densities of AlOx NPs functionalized with C16-PA, in accordance with the mass loss from 400 to 650 °C of Figure 3.8 The SSA was calculated from a DLS distribution to be 28.05 m2/g.

grafting density concentration employed. (mM) mass loss (wt)

(molecules per nm2)

2.5

5.27

3.9

5

7.03

5.3

10

7.65

5.8

20

8.01

6.1

40

8.25

6.3

Consideration of anchor group in the calculation It is a fairly common dilemma whether or not to employ the full molecular weight (MW) of the SAM molecule to calculate grafting density with Equation 3.1. The problem in question being, whether or not the totality of the SAM molecule, including the anchor group is removed from the NSs surface during the TGA measurement. This is particularly significant for smaller SAM molecules in which the anchor group can represent more than 50% of the total MW. Therefore, it can contribute heavily to the error of the calculation. In the previous example (Figure 3.8) C16-PA was employed given that it is a fairly "heavy" molecule. Yet, around 25% of the total MW corresponds to the phosphonic acid anchor group. To complicate matters, the PA reaction pathways and binding modes from Figure 2.8 and Figure 2.9 should also be taken into consideration. Since, some of the oxygen atoms might no longer be present after the PA binds with the metal oxide surface. Yet, other oxygen moieties might just be hydrogen bonded. As different anchor groups have different binding modes and affinities it gets even

36


more complex. For the sake of simplicity and consistency, in the entirety of this work the full molecular weight of the SAM molecule was always employed for the calculation of grafting density.

1200 100

1100 1000

90 °C

98

900 800 800 °C

T (%)

Weight (%)

96 94

700 600

1000 °C

500

92

400

1100 °C

90

Temperature (°C)

600 °C

300

Black powder

88

3200

2800

2400

2000

1600

1200

Wavenumber (1/cm)

800

200 100

White powder

86 0

25

50

75

100

125

150

175

200

0 225

Time (min)

Figure 3.9.: TGA under N2 from 25 to 1100 °C of AlOx NPs functionalized with C16-PA. The inset shows FTIRATR spectra of the nanoparticle powder at different stages of the TGA measurement. It can be observed that even after exposing the NP powder to 1100 °C for 2 hours the PA band is still present on the nanopowder. The nanopowder had a black color up to a 1000 °C, past that temperature the powder was white in appearance.

Figure 3.9 provides excellent qualitative data regarding this topic. For this experiment several TGA measurements were made with different final temperatures (90, 600, 800, 1000 and 1100 °C) while all other conditions remained the same (N2, 10 °K/min). At the end of each TGA measurement, a FTIR-ATR analysis of the remaining powder was performed (Figure 3.9 inset). The sample measured until 90 °C was merely used as a control and shows the typical bands of the C16-PA as previously discussed in Section 3.2.1. On the other hand, the FTIR-ATR spectra of the samples measured until 600, 800, 1000 and 1100 °C show no bands related to the methyl and methylene groups of the C16-PA. Yet, the bands pertaining to the anchored PA group clearly remain even after exposing the functionalized NP powder to 1100 °C for 2 hours. The latter, gives a strong hint regarding the remainder of PA groups on the nanoparticle surface after TGA. When comparing the spectra of the 600 and 1100 °C, there seems to be a small decrease in the intensity of the PA band. However, a quantitative determination based on FTIR alone is not possible. As a final remark, the analyzed


37

nanopowders had a dark black color up to a 1000 °C, past that temperature the powder was completely white in appearance. Furthermore, there is a small but steep mass loss around 1050 °C. The origin of this step was not identified. However, it could be speculated that this mass loss can be related to the dissociation of the phosphorous-carbon (P-C) bond or some other sort of carbon remnants on the nanopowder. This latter could explain the color change from black to white, as there would be no more carbon present. Grafting density paradox In the very same way Equation 3.1 can be used to estimate the grafting density, logically, it can also be used to calculate the SSA. If instead of having a fixed value for the SSA, we have a fixed value for the grafting density, the SSA of the particles can be calculated from the measured amount of adsorbed molecules. Fixed values for grafting densities are available from theoretical [42] and experimental literature for all kinds of anchor groups [19], [80], [81]. Even one of the most reliable methods to measure the SSA of a nanomaterial, the Brunauer–Emmett–Teller isotherm (BET) follows the same approach. BET involves measuring the amount of adsorbed gas (usually N2) to the nanopowder surface. Once it is measured, the SSA is calculated by using a fixed grafting density for the gas. Naturally, if the theoretical grafting density was to differ from the actual real grafting density the calculated SSA would be erroneous. To conclude, the methodology presented here to calculate the grafting density of a SAM molecule might be prone to some discrepancies. It is difficult to obtain a "beyond reasonable doubt" grafting density by these methods, since they all involve indirect and complex measurements (BET, DLS, TGA). Yet, as long as the measurement procedure remains the same for all samples a relative evaluation between them can be made. Other techniques that allow for precise measurement of grafting densities of the same systems have been developed. An example of such technique is X-ray reflectivity measurements (XRR) [81]. However, such technique was not applied during this work. Re-functionalization of previously functionalized powder SAM patterning of 2D NSs with 2 or more SAMs, typically involves the full coverage of the substrate with the first SAM. This is then followed by the masking of the SAM via a patterning technique. The exposed SAM is then removed with oxygen plasma, while the masked SAM remains unaffected. Next, the second SAM is deposited on the exposed surface were the removed SAM used to rest. Likewise to the 0D NSs, the question remains whether the phosphonic acid anchor group remains on the surface of the 2D NSs after SAM removal. As well as, the impact that this phosphonic acid groups have on the deposition of the second SAM atop this area. To get some insight into the latter question, AlOx nanoparticles functionalized with C16-PA were subjected to TGA until 1100°C. The particles were then subjected to the same functionalization procedure as before with C16-PA and the FTIR-ATR spectrum was measured at every step.

38


a) Pristine AlO

b) Func. AlO

x

x

with C16-PA

x

with C16-PA after TGA @1100 °C

T (%)

c) Func. AlO

d) Re-func. TGA AlO

e) Re-func. TGA AlO

3800

3600

x

with C16-PA

x

with C16-PA after TGA @1100 °C

3400

3200

3000

2800

Wavenumber (1/cm)

2600 2000

1800

1600

1400

1200

1000

800

Wavenumber (1/cm)

Figure 3.10.: FTIR-ATR spectra (color coded) of pristine and several C16-PA functionalized AlOx nanoparticles. a) Spectrum of pristine AlOx nanoparticles. b) C16-PA functionalized AlOx nanoparticles. c) C16-PA functionalized AlOx nanoparticles after exposure to TGA until 1100 °C. e) The exposed AlOx nanoparticles to TGA until 1100 °C but re-functionalized with C16-PA. e) The C16-PA re-functionalized AlOx nanoparticles after being measured again by TGA until 1100 °C.

Figure 3.10a shows the typical featureless spectrum of pristine alumina particles. On Figure 3.10b the spectrum of nanoparticles functionalized C16-PA can be found. After being exposed to the TGA measurements the spectrum shows no sign of the alkyl chain. However, the phosphonic acid band is still present as expected (Figure 3.10c). Looking at the spectrum from Figure 3.10d, it is clear that the re-functionalization of the particles with C16-PA was successful. However, in this spectrum the phosphonic acid band which occurs between 1000-1200 cm-1 seems to have a slightly modified shape. In addition, there is the appearance of a previously unseen band around 1390 cm-1 (purple). The last mentioned observations, could be an indication of the existence of a different binding configuration of the phosphonic acid to the surface. In contrast, to the observed binding configurations of the phosphonic acid to a pristine alumna surface. Finally, the particles were exposed to the TGA analysis for the last time (Figure 3.10e). As expected, the bands related to the alkyl chain are no longer present, while the phosphonic acid band remains. Interestingly, the unknown band is no longer present after the TGA measurement.


39

It is clear that particles can be re-functionalized as it is evident from the FTIR-ATR spectra above. Furthermore, there were no major discrepancies in the calculated grafting densities of both C16-PA functionalized particles. Therefore, the deposition of a phosphonic acid SAM atop a surface with remaining phosphonic acid groups is possible in a nearly identical fashion as the first one.

3.2.3. Mixed monolayers (SCA, FTIR & Zeta potential) SAM layer formation is not limited to one type of molecule at a time, mixed SAM formation is also a viable and flexible alternative. Stoichiometric mixtures of self-assembled monolayers on 2D AlOx surfaces have been previously confirmed [82], [83]. In here, a facile solution-based procedure for tailoring the surface properties of aluminum oxide nanoparticles by the formation of core-shell systems with mixed self-assembled monolayers will be described [16]. By employing chained molecules with a phosphonic acid anchor group and either hydrophobic or hydrophilic chains (Figure 3.11a), the surface properties of the nanoparticles can be changed dramatically. A mixed monolayer consisting of phosphonic acid molecules with a hydrophobic tail (F17C10-PA) and a hydrophilic tail (H(OC2H4)3-PA) was grafted onto the alumina nanoparticles by employing the previously described procedure (Section 3.2.2). However, in this case the surface properties of the nanoparticles were smoothly tuned by the formation of a mixed ligand monolayer. These layers were possible by using the corresponding stoichiometric mixtures of the SAM molecules during functionalization. To this end, the ratio of the hydrophobic:hydrophilic SAM molecules employed was varied proportionally in the ratios of 0:1, 1:3, 1:1, 3:1 and 1:0. The latter resulted in nanoparticles with differing surface properties, whose characterization is shown ahead.

3.2.3.1. Surface energy Using the aforementioned functionalized nanoparticle dispersions in 2-propanol, nanoparticle films were spray coated onto a Si/SiO2 wafer for static contact angle measurements and surface energy calculations (Figure 3.11b). The film morphology and coverage were measured by AFM (Appendix Figure 7.4). The surface roughness of a 10x10 µm area was measured to be approximately 100 nm RMS which correlates agreeably to the measured DLS particle size distribution. Spray coated films of the 1:0 and 3:1 particles exhibited superhydrophobic properties with presumably DI water contact angles higher than 160°. The DI water contact angle of such films cannot be measured properly, since the adhesion of the water droplet to the dispensing needle is higher than that to the film (Appendix Figure 7.5). As expected, the nanoparticle film hydrophobicity decreased for the films with a lesser degree of F17C10-PA in respect to the H(OC2H4)3-PA, and vice versa in terms of hydrophilicity. However, superhydrophilic properties only manifested at the nanoparticle film in which F17C10-PA was completely absent. The same tendency in wetting was observed when the films were probed using diiodomethane and formamide. Using the contact angle measurements of the probing liquids, the surface energy of the films (not the nanoparticle surface) was calculated. As

40


expected, films consisting of particles functionalized only with F17C10-PA presented an extremely low total surface energy of 0.45 mN/m. Whereas, films with H(OC2H4)3-PA NPs exhibited a much higher surface energy of 67 mN/m.

a)

b) 180

70

SCA (°)

160

not measurable 60

140

50

120

40

100

30

80

Polar Dispersive

60 40

20 10

SE (mN/m)

DI Water Formamide Diiodomethane

1.5 1.0

20

0.5

0

0.0 0:1

1:3

1:1

3:1

1:0

F17C10-PA : H(OC2H4)3-PA (Ratio)

Tunability Figure 3.11.: a) Molecular structure of F17C10-PA and H(OC2H4)3-PA. b) SCA measurements with different liquids and the calculated surface energy of the nanoparticle spray coated films. Reprinted from [16] with permission from the American Chemical Society.

3.2.3.2. FTIR-ATR The FTIR-ATR analysis of the nanoparticle dry powders is shown in Figure 3.12b. The spectrum from the particles functionalized with the H(OC2H4)3-PA ligand (0:1) displays a broad valley from 1000 to 1200 cm-1 corresponding to the P-O and P=O vibrations, which is comparable to the case of the C16-PA. However, in this case the smoothness of the valley is compromised by the overlap of the C-O-C stretching vibrations which are present in the same region [84]. Similarly for the particles functionalized with F17C10-PA (1:0) an overlap between the bound phosphonic acid and the C-F2 and


41

C-F3 vibrations occurs. The carbon fluorine vibrations are represented by the strong peaks at 1150 and 1200 cm-1 and by the smaller adjacent peak at 1250 cm-1 [85], [86]. Further evidence of the presence of a mixed monolayer is shown in FTIR spectra of particles functionalized with the 1:3, 1:1 and 3:1 ratios (Figure 3.12b). In particular for the 1:3 spectrum, a clear superposition of the 1:0 and 0:1 signals can be observed. Equally, for all the spectra, the two small peaks at 2850 and 2920 cm-1 relate to the C-H2 group vibrations. Finally, a trend in intensity is also observable for the methylene groups when comparing the signals of the H(OC2H4)3-PA (0:1), which possesses six methylene groups, against the F17C10-PA (1:0) with only two methylene groups.

a)

0:1

b)

1:3

T (%)

1:1

3:1

1:0

3800

3600

3400

3200

3000

2800

Wavenumber (1/cm)

2600 2000

1800

1600

1400

1200

1000

800

Wavenumber (1/cm)

Figure 3.12.: a) Color coded molecular structure of F17C10-PA and H(OC2H4)3-PA. b) FTIR-ATR spectra of particles functionalized with different ratios of F17C10-PA and H(OC2H4)3-PA. c) Zeta potential measurements of the nanoparticle dispersions in 2-Propanol. Adapted from [16] with permission from the American Chemical Society.

3.2.3.3. Zeta potential The nanoparticles functionalized with H(OC2H4)3-PA exhibited a positive zeta potential in 2propanol of approximately +50 mV (Figure 3.13a). Having such a high zeta potential renders the nanoparticles highly dispersible in solution. Likewise, particles functionalized with F17C10-PA show a negative zeta potential in 2-propanol of -50 mV, making them equally dispersible. However, when a 1:1 mixed ligand monolayer is grafted onto the nanoparticles, the measured zeta potential is nearly zero, which in turn causes the dispersibility of the particles to be very poor (Figure 3.13b). Such effect was hypothesized to be caused by the formation of a randomly ordered mixed monolayer of both ligands, effectively cancelling out the electrostatic potential (or stabilization mechanism) induced by

42


either ligand. Under such conditions, agglomeration of the particles occurred due to the increment of particle to particle interactions. Such effects are of concern regarding the further processability of the nanoparticle dispersion in terms of agglomeration and sedimentation. Therefore, they should be considered when solution processing of the nanoparticles is required.

60 40

Lower stability

Zeta potential (mV)

a)

20 0 -20 -40 -60 0:1

1:3

1:1

3:1

1:0

F17C10-PA : H(OC2H4)3-PA (Ratio)

b)

0:1

1:3

1:1

3:1

1:0

Figure 3.13.: a) Zeta potential measurements of the nanoparticle dispersions in 2-Propanol. b) Photograph of the functionalized nanoparticle dispersions 20 min after re-dispersion by sonication. Adapted from [16] with permission from the American Chemical Society.

Furthermore, it was originally hypothesized that if the 0:1 and 1:0 dispersions were mixed, the dispersions would also precipitate. This was based on their high zeta potentials of opposing charge. The nanoparticles having an opposing charge would agglomerate and eventually precipitate completely. However, this was not the case, when a mixture from the 0:1 and 1:0 nanoparticle dispersions was made, the dispersion remained unaffected as a stable dispersion. This means, that in regards to the stability of the dispersions, other forces must be at play other than the zeta potential. In this particular case, we can hypothesize that the orthogonality of the nanoparticles surface did not allowed for agglomeration to occur, even if their measured zeta potentials were of opposing magnitudes. Therefore, a combination of the steric and electrostatic stabilization schemes must work in parallel, but in an independent form. In this sense, a steric stabilization occurs when particles of differing shells meet, while an electrostatic stabilization prevents the agglomeration of particles of the same shell. The latter hypothesis, serves as a strong indication that the self-assembly of the mixed monolayers is mostly random. Thus, it is not dominated by the orthogonal nature of the backbone and


43

head group of the SAMs. If it were predominantly driven by orthogonality, then the 1:1 nanoparticles would not agglomerate. Since the formed monolayers at the nanoparticle surface would be predominantly hydrophobic or hydrophilic, which would have as consequence a more stable dispersion. However, an unstable dispersion was obtained for the 1:1 nanoparticles. Instead, the process must be random because of the strong interaction of the phosphonic acid with the nanoparticle surface. Different SAM molecules have different effects on the stability of the nanoparticles, it is predictable that different ratios of other SAM molecules can have completely different effects. These beneficial or detrimental effects on dispersibility are not in any way limited to electrostatic or steric stabilization as it is thought to be the case for the latter example. Rather, an intricate relation between the shape and size of the NSs with the highly diverse chemical nature of SAM molecules exists. However, the proper study of such relationship is beyond the scope of this dataset.

3.3. Anchor group stability As mentioned above (Section 2.3.1), the anchor group plays a pivotal role on the stability of the SAM. In the following series of experiments, the attaching affinity of the phosphonic acid, carboxylic acid and catechol anchor groups is demonstrated on a variety of 0D and 2D metal oxide NSs.

3.3.1. Desorption Desorption experiments evaluate the stability of a freshly formed SAM while being exposed to the mildest conditions possible. The desorption experiments as carried out in this work consist in depositing a SAM on a 0D, 1D or 2D NSs. Afterwards, the successful formation of the SAM is controlled via SCA measurements. Finally, the freshly functionalized NSs are exposed to "pure solvent" conditions, without any kind of SAM molecules present. The premise of these experiments relies on the fact that, whenever the SAM is strongly bound to the surface of the NSs, the exposure of the formed SAM to the pure solvent should be negligible. On the other hand, when the formed SAM has a weak interaction with the NSs surface, the impact of solvent exposure is considerable. It is simple experiments like the ones that will be described in this section, which ultimately lead to the selection of the materials for a successful functionalization scheme. Therefore, such experiments are of importance.

3.3.1.1. Desorption on 2D Ns Figure 3.14 portrays the results obtained for desorption experiments of different materials with different molecules. The materials compared were AlOx, ZnO and ITO vs. carboxylic acid, catechol and phosphonic acid anchor groups. The molecules employed in this experiment have one main thing in common, they form a hydrophobic SAM. This facilitates the monitoring of the SAM formation via

44


SCA measurements since the substrates in their starting form are highly hydrophilic. The starting SAMs were deposited over the course of 24 hours on all the substrates as previously described (Section 3.1), however, this time the substrates were then re-immersed into pure solvent (2-propanol) for differing amounts of time. After immersion, the substrates were dried with a N2 blow gun and heated at 60 °C for a few min as per the standard procedure. Then the SCA measurements were conducted.

b) 120

120

110

110

100

100

90

90

DI Water SCA (°)

DI Water SCA (°)

a)

80 70 60 50

AlOx ZnO ITO

40 30

80 70 60 50

AlOx ZnO ITO

40 30

20

20 0

200

400

600

800

0

1000 1200 1400

200

400

Time (min)

600

800

1000 1200 1400

Time (min)

c) 120 110

DI Water SCA (°)

100 90 80 70 60 50

AlOx ZnO ITO

40 30 20 0

200

400

600

800

1000 1200 1400

Time (min)

Figure 3.14.: SCA measurements of functionalized AlOx, ZnO and ITO 2D substrates. The substrates were functionalized and afterwards immersed in pure 2-propanol for different amount of time. Contact angles were measured again after the immersion. a) SCA of substrates functionalized with C17-CA. b) SCA of substrates functionalized with F21-CAT. c) SCA of substrates functionalized with C16-PA.

The first analyzed anchor group is the carboxylic acid in Figure 3.14a. The starting time "0" relates to the starting DI water contact angle after deposition of the SAM. The starting contact angle of the C17-CA SAM is hydrophobic for all substrates (above 90°). The C17-CA functionalized samples were exposed to the pure solvent for 10, 60 and 1440 minutes (24 hours). The results are quite revealing. Even after a 10 minute immersion, we can already observe a noticeable drop in the contact angle for


45

ZnO and ITO, this drop on the contact angle is even more dramatic in the case of the AlO x substrate. The DI water contact angle gradually keeps dropping as evidenced by the 24 hours samples. What we can conclude from Figure 3.14a, is that the C17-CA forms an ordered SAM on the substrates as evidenced by the high starting contact angles. However, this SAM is so weakly bound to the surface that it can be removed by the deposition solvent itself (2-propanol) in all 3 substrates. Particularly for AlOx, where the contact angle drop is higher. It could be argued that, the carboxylic acid forms the weakest bond to this surface. In Figure 3.14b, the F21-CAT molecule shows a similar but less dramatic trend. In this case, however, there were no measurements of 10 and 60 minutes since the contact angle drop of the 24 hours samples was of such lesser magnitude. What we observe with the F 21-CAT is a far slower desorption in all 3 substrates, compared to the C17-CA. Finally, Figure 3.14c shows the measurements for the C16-PA molecule. For this molecule, even after a 24 hour immersion there seems to be no suggestion of any desorption occurring, since the contact angle remained the same for all the substrates. After these observations, we can conclude that under ambient conditions C16-PA is the only molecule that forms a robust SAM compared to the other molecules. Followed by the F21-CAT, with the C17CA having the weakest bond of them all. This information is particularly important, especially when the substrates are going to be exposed to other solution processes, which can undesirably remove the deposited SAM. Therefore, a stable SAM is highly advantageous attribute whenever a permanent functionalization is desired. On the other hand, other instances when SAM removal or exchange is a desired scheme, desorption might prove to be viable strategy [35].

3.3.1.2. Desorption on 0D and 1D NSs 0D and 1D NSs often remain in solution for far longer periods than 2D NSs before their actual usage in processing. As a matter of fact, storage of functionalized 0D and 1D NSs in solution is a common practice. Therefore, it is of essential importance to study the effects of SAM desorption on solution dispersed NSs. For this purpose, AlOx and ITO nanoparticles as well as ZnO nanorods were functionalized with carboxylic and phosphonic acid SAMs namely C17-CA and C16-PA. Due to the higher amounts of SAM material required to functionalize 0D and 1D NSs, the employment of the F21-CAT SAM molecule was not possible as it was limited in availability. However, an in-house study involving all 3 anchor groups on 0D and 2D TiO2 NSs has already been published [19] with similar results. The functionalization procedure with the SAM molecules took place as previously described in Section 3.2. With the exception that this time, the particles were washed between 1 to 3 times in 2-propanol and later spray coated onto a flat substrate as described in Section 3.2.3.1 and 6.6. After deposition of the spray coated films, SCA measurements were conducted on the films (Figure 3.15). As expected (Figure 3.15a), the AlOx nanoparticles functionalized with C17-CA have a strong decrease

46


in the contact angle as they get washed. This originates from the desorption of the C17-CA from the nanoparticle surface, which in turn renders the surface more and more hydrophilic after each washing step. This is in good agreement with the previous 2D desorption studies of the same molecule. However, for the ITO nanoparticles and ZnO nanorods, there is no decrease in the contact angle after washing. This is in disagreement with the previous 2D desorption studies, where desorption occurs for all employed materials. This leads to the conclusion that the binding of the C17-CA is situational and depends not only on the anchor group and material but primarily on the NS surface characteristics. This situation was previously described in Section 2.3.4.2. Therefore, the surface of the ITO NP and ZnO nanorods differs enough from that of the 2D employed substrates that the C17-CA can form a strong bond at the surface. Essentially, a defined set of rules about the relationship of a material vs. anchor group cannot be defined without considering other properties.

b) 160

160

150

150

140

140

DI Water SCA (°)

DI Water SCA (°)

a)

130 120 110 100

AlO ZnO ITO

90 80

130 120 110 100

AlO ZnO ITO

90 80

70

70 1

2

Times washed (counts)

3

1

2

3

Times washed (counts)

Figure 3.15.: SCA measurements of spray coated films of functionalized 0D AlOx and ITO and 1D ZnO NSs. The NSs were washed 1 to 3 times before spray coating. a) Contact angles of spray coated NSs functionalized with C17-CA. b) Contact angles of spray coated NSs functionalized with C16-PA.

In Figure 3.15b, the same experiment was performed but this time with C16-PA. The results are pretty straight forward and convincing, as there is no measurable desorption in any of the materials employed. The higher contact angles of the ZnO rods are related to the higher surface roughness of the film, since they had bigger dimensions than that of the AlOx and ITO NP [87].

3.3.2. Exchange on 2D NSs SAM exchange, consists in replacing a previously deposited SAM with another one. This usually involves replacing a SAM which has a weak attachment to the surface, with a SAM which has a stronger anchor group. Some SAMs are easily replaced by others, while other SAMs are by all practical purposes unexchangeable. The studied anchor groups in this section are again, carboxylic and phosphonic acids as well as catechols on 3 different kinds of 2D NSs AlOx, ITO and ZnO. The


47

exchange experiments were performed by depositing a SAM as described in Section 3.1, followed by the immersion of the substrate under same SAM deposition conditions employed for the first SAM. By exchanging hydrophobic with hydrophilic SAMs and vice versa, a contrast in DI water contact angles can be created and the exchange of the SAM can be monitored rather simply. First example of an exchange reaction is shown in Figure 3.16. This exchange reaction consists in replacing a carboxylic acid SAM with another carboxylic acid SAM. a) 120 110

Exchange

DI Water SCA (°)

100 90 80 70 60 50

AlOx ITO

40 30 20 0

200

400

600

800

1000 1200 1400

Time (min)

b) 120 110

Exchange

DI Water SCA (°)

100 90 80 70 60 50

AlOx ITO

40 30 20 0

200

400

600

800

1000 1200 1400

Time (min)

Figure 3.16.: SCA measurements of SAM exchange between two carboxylic acid SAM molecules on AlO x and ITO 2D substrates. a) Exchange of a C17-CA SAM with a C6-CA SAM. b) Exchange of a C6-CA SAM with a C17-CA SAM.

Figure 3.16a shows the SCA measurements of the exchange of a hydrophobic SAM molecule C17-CA by a smaller and less hydrophobic SAM molecule C6-CA. The contact angle of the substrates with the C17-CA SAM starts at 94 and 102 degrees for AlOx and ITO respectively. After immersion in the C6-CA solution for 10 minutes, an equilibrium state seems to have already occurred. Samples

48


immersed for 60 minutes and 24 hours showed no further considerable decrease in the contact angle. The same occurs vice versa for both the AlOx and ITO substrates when the C6-CA is exchanged by C17-CA (Figure 3.16b). After exchanging the SAM for 10 minutes, an equilibrium point has been reached with no strong increase in the contact angle past that point. This fast exchange behavior is expected, due to the strong desorption of the carboxylic acid molecules that was previously observed. Furthermore, the availability of the new SAM in respect to the of the old one speeds the process even more. When the same exchange experiments are performed with phosphonic acid SAM molecules, no change in contact angle is observed. In Figure 3.17, the measured contact angles of the exchange experiments are displayed. Initially, the molecule C11OH-PA forms hydrophilic SAMs with contact angles around 50 to 60 degrees. After immersion in a C16-PA solution for 24 hours there was no significant change in the contact angle that indicated any SAM exchange occurring. The same occurs (not shown) when the inverse exchange reaction is attempted from C16-PA to C11OH-PA. The surface in this case remains hydrophobic with no considerable change in the contact angle. It can be concluded then, that within the timeline and concentrations employed, phosphonic acid SAMs cannot be significantly replaced by another phosphonic acid SAM. At this stage, it becomes ever more clear that the phosphonic acid anchor group forms a stronger attachment to the surface than carboxylic acids. But before drawing any conclusions, there is a last remaining group to evaluate, the catechol.

120

AlOx ZnO ITO

110

Exchange

DI Water SCA (°)

100 90 80 70 60 50 40 30 20 0

200

400

600

800

1000 1200 1400

Time (min)

Figure 3.17.: SCA measurements of SAM exchange between two phosphonic acid SAM molecules on AlOx, ITO and ZnO 2D substrates. Attempt to exchange a C11OH-PA SAM with a C16-PA SAM.

The catechol exchange experiments are not as clear as the carboxylic vs. phosphonic acid exchange experiments. Yet, the general trend of the phosphonic acids as the superior choice as anchor group is still present. Just as in the desorption experiments, the catechol anchor group lies somewhere in between the carboxylic acid and phosphonic acid. In Figure 3.18a, the DI water contact angles of the


49

exchange between C11OH-PA and F21-CAT are shown. With an increase of only around 4 degrees after 24 hours, it can be concluded that there is no significant exchange of the C11OH-PA with F21CAT. On the other hand, the opposite exchange experiment of F21-CAT with C11OH-PA shows a more favorable exchange rate (Figure 3.18b) of around 15 degrees. Therefore, it can be concluded that the F21-CAT can be exchanged by the C11OH-PA, albeit slower than a carboxylic acid SAM. However, other catechols do no form such stable SAM as the F21-CAT. a) 120

AlOx ZnO ITO

110

Exchange

DI Water SCA (°)

100 90 80 70 60 50 40 30 20 0

200

400

600

800

1000 1200 1400

Time (min)

b) 120

AlOx ZnO ITO

110

Exchange

DI Water SCA (°)

100 90 80 70 60 50 40 30 20 0

200

400

600

800

1000 1200 1400

Time (min)

Figure 3.18.: SCA measurements of SAM exchange between a phosphonic acid and catechol molecules on AlOx, ITO and ZnO 2D substrates. a) Exchange of a C11OH-PA SAM with a F21-CAT SAM. b) Exchange of a F21CAT SAM with a C11OH-PA SAM.

In Figure 3.19, the results from the exchange between a C17-CA SAM and hexanoate-CAT SAM on ZnO are displayed. In this case, both SAMs can be replaced almost completely by the other SAM. This indicates that the binding of both SAMs, is of a weak nature. The latter is consistent with the behavior of the C17-CA, however, not with the previous results of the F21-CAT SAM. The F21-CAT

50


SAM formed a more stable SAM when compared to C17-CA. Yet, in the case of the hexanoate-CAT, the formed SAM appears to have the same robustness as a carboxylic acid SAM. This discrepancy, originates from the impact EWG groups can have on the catechol anchor group, as explained in Section 2.3.4.3. a) 120

ZnO

110

Exchange

DI Water SCA (°)

100 90 80 70 60 50 40 30 20 0

200

400

600

800

1000 1200 1400

Time (min)

b) 120

ZnO

110

Exchange

DI Water SCA (°)

100 90 80 70 60 50 40 30 20 0

200

400

600

800

1000 1200 1400

Time (min)

Figure 3.19.: SCA measurements of SAM exchange between a carboxylic acid and catechol molecules on ZnO. a) Exchange of a C17-CA SAM with a hexanoate-CAT SAM. b) Exchange of a hexanoate-CAT SAM with a C17-CA SAM.

Having a strong anchor group is essential for functionalization. In this section, we have identified phosphonic acids as the most effective anchor group. The phosphonic acid was the only stable anchor group under the mild conditions employed. It was followed in robustness, by the catechol and lastly carboxylic acids. However, due to the more complex chemistry of catechols, some had more affinity to the metal oxide surface than others.


51

3.4. Chapter summary To summarize, this chapter has described the step by step functionalization approach that was performed during this work. The described procedure and characterization are of paramount importance in order to identify a successful functionalization scheme. This successful functionalization scheme consists in obtaining a consistent and reliable process by selecting the appropriate NSs core, SAM molecules and functionalization conditions. This is particularly important since the finalized functionalized NSs are further processed for a variety of applications. These applications, are heavily based upon the NSs functionalization concept. Therefore, a certain robustness of the functionalization must be ensured to avoid further difficulties along the application process chain. For this matter, this section has described a universally relevant pathway for reliable NSs functionalization. First and foremost the characterization of the starting NSs core material was performed with FTIR-ATR. It is advantageous that the starting material surface is free of organic or inorganic components that may interfere with the self-assembled monolayer process. However, in some cases even if the NSs had unknown species at their surface they were removed after the functionalization procedure. The etching of metal oxides with phosphonic acids was also briefly studied via FTIR-ATR. After determining that the starting material is suitable for functionalization, the determination of the conditions required for full coverage of the NSs with a SAM molecule was evaluated by TGA studies. Moreover, the functionalization allowed for the coverage of the nanoparticle surface with a mixed monolayer, only by proportionally varying the phosphonic acid ligands concentrations. The nanoparticle surface was successfully modified to exhibit hydrophobic or hydrophilic properties by employing different ligand mixtures. Furthermore, the wettability of nanoparticle films can be tuned to any degree by employing the mixed monolayer approach. The dispersibility of the particles in solution was demonstrated to be governed by the zeta potential of the particles. This was found to be affected by the proportions and nature of the ligands employed. Lastly, simple SAM desorption and exchange experiments on the NSs provided valuable information of the adsorption affinity of specific anchor groups (catechol, carboxylic and phosphonic acid) to specific materials. The desorption and exchange experiments, proved that the phosphonic acid is the superior anchor group when compared to carboxylic acids and catechols.

4. Applications

In this section, applications and proof of concept ideas that emanated from this work are showcased. This thesis has so far covered, how to tune the surface of commercially available materials by relatively simple procedures. This simplicity is one of the most attractive properties for applications, specially as we move from simpler to more complex schemes. Some of these ideas are still work in progress and collaborative work, therefore the detail of the displayed information varies from project to project. In some applications, a small story concerning the starting point of the project is told in order to give some insight as to why the application was developed.

4.1. Solution processing Processing of 0D and 1D NSs from solution allows for wet chemical methods of a normally solid material. It is somewhat accurate to say that the handling of 0D and 1D NSs is dominated by solution processing. Therefore, the dispersibility of the NSs in solution is of uttermost relevance and it is the first application discussed in this section. Several NSs dispersion schemes are described and are applicable to differing situations that developed during the course of this work.

4.1.1. Green processing We were asked if we could develop highly stable nanoparticle dispersions in water and lower alcohols for the fabrication of dielectric mirrors for solar applications via solution processing. We eventually developed a solution, however, the collaborators lost interest in the project. So here we had, a variety of water and alcohol dispersible metal oxide nanoparticles. Ideally, solution processing of NSs is performed by employing a non-toxic dispersion medium such as water or lower alcohols. Solution processing, particularly from water at neutral pH is a highly sought after attribute which has an impact in many areas of material science [88]–[90]. In here we demonstrate an eco-friendly solution processing concept. It is based on the chemical modification of metal oxide nanoparticles (e.g., TiO2, Fe3O4, AlOx, ITO, and CeO2). It allows for excellent dispersibility in DI water, as well as in methanol, ethanol and 2-propanol [17]. To achieve this dispersibility, the nanoparticles were functionalized with the SAM molecule CH3(OC2H4)3-PA (Figure 4.1a). The molecule is comprised of a phosphonic acid anchor group and a hydrophilic tail. This SAM molecule forms a self-assembled monolayer (SAM) around the nanoparticles surface. The phosphonic acid moiety acts as an anchor while leaving the hydrophilic tail exposed which is responsible for the 53

54

Applications

dispersibility enhancement of the nanoparticles. Figure 4.1 showcases TiO2 and Fe3O4 nanoparticles as the prototypical example of this approach.

a)

c)

b)

DI Water MeOH

EtOH

d) 30

DI Water MeOH

DI water (pristine) DI water Methanol Ethanol 2-Propanol

TiO2 25

Number (%)

IPA

20

e) 30 25 20 15

10

10

5

5

0

0 0

5

10

15

20

Size (nm)

25

30

0

20

IPA

DI water (pristine) DI water Methanol Ethanol 2-Propanol

Fe3O4

15

EtOH

40

60

80

100

Size (nm)

Figure 4.1.: a) Molecular structure of CH3(OC2H4)3-PA employed for nanoparticle functionalization; b) Photograph of dispersed TiO2 nanoparticles in different media before and c) after functionalization. d) Graph of DLS measurements of TiO2 and e) Fe3O4 nanoparticles before and after functionalization. Reprinted from [17] with permission from Wiley.

A qualitative indication of improved dispersibility of the functionalized TiO2 nanoparticles with diameter of ca. 6 nm is showcased in Figure 4.1b and c. From the observed transmittance in the photographs, it can be concluded that the pristine TiO2 particles in DI water exhibit good dispersibility. Opaque dispersions of the pristine TiO2 particles were obtained in alcohols, indicating poor dispersibility in those media. On the other hand, core-shell TiO2 particles functionalized with CH3(OC2H4)3-PA formed a transparent colloidal solution (Figure 4.1c) when dispersed in any of the

Applications

55

four solvents employed. To better discern the degree of improved dispersibility, Figure 4.1d and 4.1e display the particle size distribution obtained by DLS of the TiO2 and Fe3O4 nanoparticles before and after functionalization with CH3(OC2H4)3-PA. Due to particle agglomeration, the size measured for the pristine particles in DI water showed a misleading distribution, which interprets into bigger particles. After functionalization, the nanoparticle size distributions portrayed not only smaller diameters close to the real size of core but also an almost equivalent size distribution in all solvents. Furthermore, the functionalized NPs can be stored as a powder and redispersed in DI water or alcohols and subsequently used to create defined dispersions of various concentrations. The latter approach was employed in water processed flexible nanoparticle dielectrics described in Section 4.2.1. The FTIRATR spectra of the functionalized and pristine nanoparticles can be found in the appendix Figure 7.7. In conclusion, by using a simple functionalization procedure, diverse metal oxide nanoparticles were rendered highly dispersible in DI water and lower alcohols. This allows for solution processing of the materials under a variety of solvents. Furthermore, the same principles could be extended to other metal oxide materials, extending the potential of the concept.

4.1.2. Any medium processing At the very end of the course of this work, I came to the realization that NSs can be well dispersed in any medium as long as you find a proper SAM molecule that fits it. Here is a brief example. Often, the employment of a "not so green" dispersion media (even not encouraged) for solution processing it is in many applications a requirement. This could be merely situational or when no other alternative is available. So, just as NSs can be made highly dispersible in water and alcohols, they can also be so in any particular medium. As proof of this statement, Figure 4.2 showcases a photograph of Fe3O4 nanoparticles functionalized to match a specific phase of orthogonal dispersion media (heptane, water and n-perfluoroheptane). The nanoparticles dispersed in n-heptane were functionalized with C16-PA, while the particles dispersed in DI water, were functionalized as previously described with CH3(OC2H4)3-PA. Lastly, in order to make the particles highly dispersible in n-perfluoroheptane the particles were functionalized with F17C10-PA. Basically, the molecular structure of the SAM molecules employed was matched as close as possible to the dispersion media. Essentially, the popular aphorism "like dissolves like" seems to be also applicable to nanoparticle dispersions. However, the reasons behind their dispersibility might be far more intricate (Section 2.4). The dispersed particles were not characterized thoroughly with DLS and Zeta potential. Instead, by looking at the picture it is clear that by choosing the appropriate SAM molecule, NSs can be tuned for dispersibility in any particular media. This opens up the possibility for solution processing of NSs under an array of different scenarios. Such as, the deposition of materials from orthogonal media, which is a valuable technique when depositing one material atop another. In general, nanoparticle dispersibility can be achieved in any medium, as long as the right SAM is chosen and the nanoparticle surface coverage is well controlled.

56

Applications

a)

b)

c)

Figure 4.2.: Photograph of Fe3O4 nanoparticles functionalized with different molecules (top) and dispersed in different orthogonal media (left). a) Particles dispersed in n-heptane, b) particles dispersed in DI water, c) particles dispersed in n-perfluoroheptane. By tuning the surface of the particles, control on their dispersibility in different media can be achieved.

4.1.3. Shell by shell (double shell) I was a bit bored a Friday afternoon so I decided to go play around in the lab, I met Lukas Zeininger at the hallway and told him if he wanted to a see a "cool trick". The "trick" consisted in adding an amphiphilic molecule to a poorly dispersed hydrophobic nanoparticle dispersion, upon addition, it would turn into a perfect nanoparticle dispersion. He flipped!, he knew some colleagues of him had been synthesizing all kinds of complex amphiphilic molecules that we could try. We tried a few them, some of them worked better than others but overall there was a general tendency, so we were onto something. It eventually got late and I went home, however, Lukas (the nerd) kept at it over the weekend. On Monday we discussed his results and we somewhat came up with the shell by shell stabilization scheme described herein. The shell by shell stabilization scheme offers a dynamic approach for dispersion of NSs in solution. In this approach, hydrophobic forces are exploited towards the formation of a second layer at the surface of an already functionalized NSs. To a certain extent, van der Waals forces may also play a minor and indirect role. However, the hydrophobic force was identified as the major driving force. This noncovalent mechanism for grafting a second layer, is based in the same working principle as a lipid bilayer would form micelles or liposomes. The developed bilayer stabilization concept is schematically depicted in Figure 4.3. It starts by fabricating an hydrophobic shell around the nanoparticles, in this case they were TiO2 nanoparticles fully functionalized with C16-PA (Figure 4.3a).

Applications

57

These particles are then forcefully dispersed (very poorly) in DI water. However, upon the addition of an amphiphilic specie to the dispersion, a bilayer as depicted (prototypically) in Figure 4.3b is potentially formed in a micellar arrangement. To this end, a variety of amphiphilic molecules were synthesized or commercially obtained (Figure 4.3c). However, to illustrate the proof of concept in this section we will focus mainly in the employment of molecule 7. This molecule consists of a twelve carbon long alkyl chain followed by a perylene motif as the hydrophobic component of the amphiphile. The hydrophilic component of the amphiphile consists of a dendritic structure composed of nine pyridinium moieties (Figure 4.4a). As for the other molecules a more thorough description of the results of each amphiphilic molecule can be found online [18].

Figure 4.3.: Shell by shell stabilization concept. a) Nanoparticle is rendered hydrophobic by functionalization with C16-PA. b) The hydrophobic particle is rendered hydrophilic due to the amphiphilic molecules forming a bilayer. c) Example of the amphiphilic molecules employed for this study. Adapted from [18] with permission from Wiley.

58

Applications

The stages of the shell by shell process with the employment of the amphiphilic molecule 7, are clarified with a simple phase transfer experiment showcased with photographs in Figure 4.4b, c, and d. First, the pristine unfunctionalized hydrophilic TiO2 nanoparticles can be seen dispersed in the water phase with a clear toluene phase above them in Figure 4.4b. Then, after the particles have been rendered hydrophobic after functionalization with C16-PA (as previously described in Section 3.2). Naturally, the hydrophobic particles are now dispersible in the toluene phase instead of the water phase (Figure 4.4c). Finally, upon addition of the amphiphilic molecule 7 the dispersibility of the nanoparticles is reversed once more to the water phase due to the formation of a double layer (Figure 4.4d). Additional support for the double layer formation comes from the measured zeta potential at this stage, which was highly positive around +40 mV. this suggests an electrostatic stabilization mechanism provided by the pyridinium motifs engulfing the particles. Also at this point, the particle dispersion has now a strong red color, due to the perylene motif on the amphiphilic molecule. Hydrophilic

a)

Hydrophobic

b)

c)

d)

Toluene DI water

Figure 4.4.: Photograph of TiO2 nanoparticles dispersed in DI-water or toluene. a) Hydrophobic and hydrophilic components of molecule 7. The red coloring of the TiO2 nanoparticles is due to the perylene motif. b) Pristine hydrophilic particles dispersed in DI-water. c) C16-PA functionalized hydrophobic nanoparticles dispersed in toluene. d) Upon addition of the amphiphile molecule 7 the nanoparticles are now dispersible in the water phase due to the formation of a bilayer. Adapted from [18] with permission from Wiley.

The red color arises from both bound and unbound specimens of the molecule 7 to the nanoparticles, as the molecule by itself is highly water soluble. Fascinatingly, we noticed that when the nanoparticle

Applications

59

dispersion was centrifuged at this stage, a clear supernatant and a pink colored particle powder was obtained. Thus, the nanoparticles can sequestrate and isolate in solution what would normally be a highly soluble amphiphilic molecule such as 7. This very same concept, eventually led to the development of nanoparticles that were employed as scavengers as briefly described in Section 4.6. In summary, an approach for generating a controlled second layer on the surface of functionalized nanoparticles has been described. The non-covalent grafting nature of the second layer allows for its later removal if desired. This is possible as the approach described herein relies mainly on hydrophobic forces, which means that the dispersion medium is limited to water, were water would provide the primary force holding the bilayer together [91]. Thus, transferring the double shell nanoparticles from their orthogonal system to a non-orthogonal system (e.g. lower alcohols), should in principle, collapse (at least partially) the second layer into the solution as individual components. Whether this reversibility is advantageous or not is merely situational and application dependant. Lastly, the shell by shell concept is not necessarily limited to hydrophobic interactions, as similar interactions between oleophobic or fluorophobic forces are not unheard of [92], [93]. However, the employment of such driving forcers remains to be investigated.

4.1.4. Polymer wrapping This stabilization scheme was developed when Tobias Rejek required glycol covered nanoparticles that were highly dispersible in toluene, a somewhat contradictory proposition. The particles were needed so that they could be combined in solution with block copolymers, the block copolymers employed were highly soluble in toluene but not in polar solvents. We had a few bumps along the way, but eventually managed to come up with a working solution, which is described below. Wrapping nanoparticles with polymers is a fairly common strategy for nanoparticle stabilization [94]– [96]. In this particular case, we specifically resorted to this stabilization mechanism when the need arose to have hydrophilic iron oxide particles dispersed in toluene. At the same time, the solution should also contain a dissolved block copolymer. The need for these highly stable hydrophilic particles in the conditions described, originated from another project related to nanoparticle selforganization. The driving force for this organization is driven by chemically matching a specific phase of a phase separated block copolymer film. This application is briefly described elsewhere (Section 4.3.2). The best results under the proposed circumstances were obtained when the particles were functionalized with the molecule CH3(OC2H4)4C4H8-PA displayed in Figure 4.5a. The polymers employed consisted of a selected variation of block copolymers and polymers of differing molecular weights, as displayed in Figure 4.5b. To better portrait the proposed approach, a suggested model of a nanoparticle wrapped by the block copolymer is depicted in Figure 4.5c. In such model we can

60

Applications

observe the nanoparticle core surrounded by an hydrophilic SAM composed of CH3(OC2H4)4C4H8-PA (blue). Atop this layer, a far thicker polymer layer lies upon (red and dark blue). The polymer layer is arranged in such a way, that the hydrophilic phase (dark blue) of the polymer remains towards the center of the particle whereas the hydrophobic phase (red) tends to reside around the particle periphery. If this were to happen as described in the model, the hydrophilic particles could be made dispersible in toluene via the steric stabilization provided by the polystyrene phase surrounding the nanoparticles. a)

b)

PS (MW)

PEO (MW)

1

55000

10200

2

42000

11500

3

16400

72000

4

280000

0

c)

Block copolymer wrapping nanoparticle

Nanoparticle hydrophilic tails

Nanoparticle core

Figure 4.5.: Polymer wrapping of nanoparticles. a) Phosphonic acid molecule with hydrophilic tail employed to functionalize the surface of the particles. b) PS-b-PEO block copolymer molecular structure and weights. c) Schematic representation of a nanoparticle wrapped in an orthogonal block copolymer.

Applications

61

Originally, we anticipated that the polymer wrapping would occur simply by dissolving the polymer in toluene and afterwards adding the hydrophilic nanoparticles into the solution. The hydrophilic particles have an extremely poor dispersibility in toluene. Even so, we expected the polymer to pick them up, wrap them and gradually turn them into an stable dispersion. However, this was not the case since the nanoparticles remained heavily agglomerated, even after long stirring or sonication times. At this point, we hypothesized that in order for the particles to be wrapped by the polymer, they first needed to be individualized (non-aggregated). Undoubtedly, the hydrophilic particles could be well dispersed in water, unfortunately the polymers could not. Therefore, we decided to try an approach which involved both the water and the toluene phase (Figure 4.6a). On one hand, the water phase contained highly stable (individualized) hydrophilic particles. On the other hand, the toluene phase contained the dissolved polymer. The two phase solution-dispersion was then vigorously shaken and treated with ultra sonication until it was transformed into an almost homogenous emulsion-like product (Figure 4.6b). Finally, in the last step of the process, the solvents (water and toluene) were evaporated by vacuum at 50 °C in a rotary evaporator. What remained was a brown colored "plastic spider web" like material. Once all the solvent was removed, pure toluene was added to the dried product. This time however, it formed a perfectly stable whisky colored transparent solution-dispersion. The solution-dispersion was now perfectly suited for solution processing.

a)

b)

Dissolved polymer in toluene Hydrophilic iron oxide NPs in DI-water

Before shaking

After shaking

Figure 4.6.: Photograph of the solution before and after shaking vigorously. Photograph b) shows the solution only after shaking, not after sonication.

DLS measurements of the dispersions before and after polymer wrapping were conducted and are displayed in Figure 4.7. The black distribution represents the size measured for the CH3(OC2H4)4C4H8 PA functionalized particles, while the others represent the size measured after the polymer wrapping procedure and were dispersed in toluene. It was a good indication, that after the polymer wrapping, there was a slight increase in the size of all the polymer wrapped nanoparticles.

62

Applications

Since it was expected, that the polymer wrapping would increase the size of the nanoparticles. However, DLS is a complex and indirect measurement technique which can be affected by several factors [97]. Therefore, an increase of a few nanometers in the size distributions (while very consistent) should be taken with skepticism. In search of further evidence to support the polymer wrapping scheme, the zeta potential of the dispersions was measured and it is displayed in Table 4.1. The particles covered by CH3(OC2H4)4C4H8-PA were in theory stabilized via steric stabilization. However, the particles exhibited a clear zeta potential of +30 mV. This suggested an electrosteric stabilization mechanism for this particle dispersion. Interestingly, after the polymer wrapping procedure the measured zeta potentials in toluene were all near zero, yet they formed perfectly transparent dispersions. This lack of zeta potential and perfect stability, was unmistakably due to a steric stabilization which was provided by the wrapping of the polymer to the nanoparticles. Still, at this point we could further argue that the solvent itself can have an impact on the measured zeta potential. In particular, between zeta potential measurements carried out in a polar solvent vs. toluene. For this purpose, the zeta potentials and DLS size distributions of the CH3(OC2H4)4C4H8-PA covered particles were all measured under chloroform. A solvent which is more analogous to toluene. It must be noted, that this measurement was only possible due to the serendipitous discovery that the CH3(OC2H4)4C4H8-PA covered particles were perfectly dispersible in chloroform. As a matter of fact, if in the previously described polymer wrapping procedure the DI-water is replaced by chloroform, it also results in toluene dispersible nanoparticles. With the only difference being that there is no phase separation between toluene and chloroform. So far these experiments have made for a nice story. However, there is still one small troublesome detail left. Surprisingly, when the same procedure is performed by replacing the block copolymer with a long PS polymer 4, similar results were obtained. As a matter of fact, this polymer was only employed to act as a negative control vs. the block copolymers. Unexpectedly, it also made the nanoparticles perfectly dispersible in toluene even when the polymer had no hydrophilic components. Undoubtedly, the polymer must somehow trap or wrap around the nanoparticles making them dispersible. Ultimately, we concluded that the length of the PS polymer (MW = 280k) was too long and therefore wrapped the nanoparticles independently of phase matching or not. This in light of, that using shorter PS polymers (MW = 35k) for dispersing the particles, yielded good negative controls (unstable dispersions), as we originally expected. In conclusion, a procedure for creating stable dispersions of hydrophilic nanoparticles in toluene via polymer wrapping was described. A key factor for the success of the procedure is to first individualize the particles in solution so that the polymer can wrap around them. The individualization was possible as the particles formed stable dispersions in DI water and chloroform. Unexpectedly, in order for polymer wrapping to occur around the nanoparticles a phase matching block copolymer was not required, it sufficed that the polymer was long enough to capture and wrap around the nanoparticles.

Applications

63

40 CH3(OC2H4)4C4H8-PA Fe3O4 PS55k-PEO10.2k Fe3O4

35

PS42k-PEO11.5k Fe3O4

Number (%)

30

PS16.4k-PEO72k Fe3O4 PS280 Fe3O4

25 20 15 10 5 0 0

10

20

30

40

Size (nm) Figure 4.7.: DLS size distribution of CH3(OC2H4)4C4H8-PA nanoparticles before (black) and after polymer wrapping.

Table 4.1.: Zeta potential of Fe3O4 nanoparticles. The hydrophilic CH3(OC2H4)4C4H8-PA nanoparticles have a strong zeta potential due to the electrosteric stabilization caused by the functionalization. After the polymer wrapping procedure the nanoparticles are still stable, yet their zeta potential is near zero. This effect is caused by the steric stabilization provided by the polymer wrapping of the nanoparticles.

Material

Zeta potential (mV)

CH3(OC2H4)4C4H8-PA Fe3O4

+30


+1.87


+2.64

PS16k-PEO72k Fe3O4

+0.9

PS280k Fe3O4

-4.49

Yet, there are still a few open questions regarding this polymer wrapping procedure. For example, when employing similar hydrophilic SAM molecules and following the same protocol only partially stable dispersions were obtained. Therefore, further investigations regarding the exact wrapping mechanisms are required. For the moment however, a simple procedure for polymer wrapping of nanoparticles was described. Furthermore, the same polymer wrapping procedure has been validated for TiO2 nanoparticles with a core size of 6 nm as well.

64

Applications

4.2. Thin films. From 0D to 2D Thin films are corner stone of nanotechnology, their use and application encompass every field of nanotechnology. Often, nanoparticle thin films can be easily fabricated from nanoparticle dispersions via solution processing. It is therefore natural, that having gained such control over nanoparticle dispersibility as previously described, to employ conventional thin film fabrication techniques such as spin coating, spray coating, doctor blading, etc. for the fabrication of nanoparticle thin films. In here, example applications of nanoparticle thin films fabricated using the previously described dispersion methods are presented.

4.2.1. Flexible dielectrics The water and alcohol NP dispersions described in Section 4.1.1 were used to fabricate the dielectric layer for organic thin-film transistors (OTFTs). The dielectric layers consisted of several metal oxide materials namely AlOx, ITO, CeO2, TiO2 and Fe3O4. The films were fabricated by simple spin coating from water or 2-propanol in ambient air, followed by annealing at a maximum temperature of 100 °C for ten minutes to remove the solvent. Interestingly, after annealing, the films showed no signs of degradation (as confirmed by AFM) even after being exposed to a vigorous rinse of DI water or 2-propanol. This potentially allows for spin coating of different layers of different materials atop of each other. Furthermore, it was observed that the film thickness could also be tuned by changing the NP concentration of the spin coating solutions, with higher concentrations resulting in thicker films under the same process conditions. Yet, other factors influencing the NP film thickness include inherent solvent properties, such as vapor pressure or viscosity. Therefore, spin coating of the NP films from either DI water or alcohols enables an extra degree of freedom for thickness tunability. For characterization the core-shell particles were spin coated from aqueous dispersions (0.6 wt-%) onto a silicon oxide wafer, except that the AlOx particles were spin coated from 2-propanol (0.6 wt-%) to avoid etching of the particles in DI water. The AlOx particles when dispersed in DI water tend to agglomerate relatively sooner and permanently, this is not the case when dispersed in 2-propanol. Cross-section images of the spin coated films obtained via scanning electron microscopy (SEM) are shown in Figure 4.8a. The films showed long range and consistent thickness as a function of the particle size. The film surface roughness also relates to the mean nanoparticle diameter of the NPs employed (Table 4.2). The larger the NPs, the larger the surface roughness measured. The NP films were also spin coated onto a silicon oxide wafer that was pre-patterned with capacitor and transistor gate electrodes made of Al with 30 nm thickness. The film quality on top the aluminum electrodes was characterized with atomic force microscopy (AFM), it showed that Al gate electrodes were completely covered by the nanoparticle films (Figure 4.8b).

Applications

65

a)

b)

Aluminum Oxide

2 µm

Indium-Tin Oxide

2 µm

Cerium Oxide

2 µm

Titanium Oxide

2 µm

Iron Oxide

2 µm

Figure 4.8.: a) SEM cross-sections of the spin coated dielectric layers. b) AFM images of the surface morphology of the spin coated films. Reprinted from [17] with permission from Wiley.

Table 4.2.: Summary of film properties and electrical characteristics of the films and devices. Reprinted from [17] with permission from Wiley.

Metal oxide

Particle mean size [nm]

Thickness [nm]

RMS surface roughness [nm]

Capacitance [µF/cm2]

µ sat [cm2/Vs] @ Vds = –3 V

Voltage Threshold [V]

Ion/Ioff

Id/Ig

TiO2

6

8

2.3

1.08

0.14

2.78

25 × 103

12

Fe3O4

10

6

3.3

1.01

0.14

2.80

21 × 103

12

AlOx

67

100

22.7

0.39

0.66

3.12

122 × 103

76

ITO

57

50

18.6

1.12

0.27

2.63

3.8 × 103

17

CeO2

55

40

18.0

0.92

0.24

2.60

5.0 × 103

27

66

Applications

To characterize the dielectric layers, 50 x 50 µm capacitors were fabricated (Figure 4.9a). The breakdown characteristics of the NP-dielectric layers (Figure 4.9b) exhibited reliable values of ±5.5 V. The low leakage current at voltages below breakdown indicate excellent, and almost uniform, insulating behavior with slightly improved behavior for the AlOx layer, due to an increased thickness (Table 4.2) which arises from the use of 2-propanol for spin coating. The corresponding capacitor devices without NP layers (native AlOx) exhibited a higher current density of almost two orders of magnitude. It is also noted that ITO and TiO2 core-shell systems provided uniform insulating properties, indicating that the shell layer significantly limits the transport as obtained previously for core-shell NPs of ZnO [28]. The corresponding capacitance values (measured at 10 kHz) are shown in Table 4.2. In contrast to bulk oxide dielectric layers, the NP-based layers differ in their composition. Considering the core-shell architecture, which provides different -values for core and shell materials, and the particulate structure leading to a certain porosity of the layer, the measured capacitances of the NP layers are larger than those formed from bulk oxide [11], [98], [99].

a)

b) AlOx

10-3 10

ITO CeO2

-5

Fe3O4

AlOx NP dielectric Bottom (Al) Substrate

I (A)

Top (Au)

TiO2

10-7 10-9 10-11 10-13

-8

-6

-4

-2

0

2

4

6

8

Bias Voltage (V)

Figure 4.9.: a) Schematic layout of the fabricated capacitor devices. b) Current density of different NP dielectric materials of 50x50 µm capacitor devices vs. applied voltage. Reprinted from [17] with permission from Wiley.

In order to demonstrate the use of the NP layers as gate dielectrics, they were integrated into thin-film transistor (TFT) devices. Figure 4.10a illustrates the bottom-gate, top-contact architecture of the fabricated devices, which were formed using 2-tridecyl[1]benzothieno[3,2-b][1]benzothiophene (C13-BTBT) as the organic semiconductor [99]. The transfer curves of the devices are shown in Figure 4.10b, and their characteristics summarized in Table 4.2. Devices with AlOx nanoparticles as dielectric showed the best performance, with a saturation mobility of 0.66 cm2V–1s–1, followed by ITO and CeO2. The lower performance of the ITO and CeO2 devices was attributed mainly to the semiconducting nature of these NPs, and, thus, to relatively low ID/IG ratios. The use of ITO and CeO2 as a dielectric was still possible. However, due to the organic insulating shell layer encasing the NPs, which retards

Applications

67

transport between the particles [27], [28]. TiO2 and Fe3O4 devices showed the lowest performance in terms mobility, which can be attributed to the poor wettability of the evaporated C13-BTBT semiconductor films onto the TiO2 and Fe3O4 films. It seems likely that the variation in wettability is linked to the nanoparticle core size, i.e., SAMs grafted onto smaller particles possess additional free space in-between the molecules tails, whereas the ligand to ligand distance is constrained due to a less pronounced surface curvature in bigger particles [65]. This in turn, impacts the free surface energy of the nanoparticle film. Consequently, different semiconductor wetting and device performance behavior was observed, and a direct overall comparison of devices formed from AlOx, ITO, and CeO2 NPs is thus quite difficult [100]. Finally, it is noted that after four months of storage under ambient conditions, the devices performance remained virtually unaffected in terms of charge carrier mobility, VTH and ON/OFF ratio. a)

b)

10-5

AlOx

10-6

AlOx Gate (Al)

Drain (Au) Ig (A), Id (A)

Source (Au)

TiO2

10-8

Fe3O4

10-9 10-10 10-11 10-12

NP dielectric Substrate

10-13 10-14 -5

c)

ITO CeO2

10-7

d)

-4

10-5

-2

-1

0

concave bending no bending convex bending

10-6 10-7 Ig (A), Id (A)

-3

10-8 10-9 10-10 10-11 10-12 10-13 10-14 -5

-4

-3 -2 Vgs (V)

-1

0

Figure 4.10.: a) Molecular structure of the semiconductor molecule C13-BTBT and schematic layout of the fabricated OTFTs devices. b) Transfer curves of the OTFTs with different dielectric materials. c) concave bending of devices during characterization. d) Transfer characteristics of the devices under different bending modes. Reprinted from [17] with permission from Wiley.

68

Applications

To further establish the universality of the developed deposition concept in terms of eco-friendly production methods on bendable substrates, devices were fabricated onto flexible polyethylene naphthalene (PEN) substrates using AlOx nanoparticles as a prototypical dielectric. To evaluate the mechanical robustness, the flexible devices were characterized under different bending conditions as depicted in Figure 4.10c. The bending radius was around 5 mm both in the concave and convex bending modes. Figure 4.10d compares the measured transfer curves of a device under different the bending conditions. The differences between ID and IG are negligible, falling within the same order of magnitude regardless of the bending mode. Due the higher surface roughness of the PEN substrates, the highest mobility measured in the flexible devices was slightly reduced to 0.34 cm2V–1s–1, an effect what is commonly observed [30], [101]. Experience has been obtained in terms of foldability or cyclic bending of the devices. Thus, the feasibility of spin coated nanoparticle dielectrics onto flexible substrates is clearly demonstrated, In conclusion, the flexibility of the procedure in terms of nanoparticle thin film formation has been demonstrated. Unfortunately in terms of transistor fabrication, the performance of the devices was limited. Mainly because of the roughness of the NP films and/or the wetting of the semiconductor. Both are factors which can severely affect the performance of TFTs.

4.2.2. Coatings I realized I can functionalize any nanoparticle core with any ligand and simply spray coat it onto any surface. It would then create some kind of functional surface depending on the core and exposed head groups. Here is a brief example which involves a superhydrophobic coating, and potentially highly oleophobic as well. Once nanoparticles have been modified, they can easily be spray coated onto any surface forming a nanoparticle coating. There is potential in these kind coatings when you consider that the exterior of the nanoparticle core can consist of any organic functional group or even any particular combination of them. As previously described in Section 3.2.3 the nanoparticle surface can be finely tuned with a variety of SAM molecules. Furthermore, by using different nanoparticle cores or even a mixture of cores, they can act as light filters, color pigments or even create a magnetically responsive coating. To demonstrate this concept, superhydrophobic films were fabricated on a glass substrate as well as cardboard and they are shown in Figure 4.11. These samples were prepared by spray coating of F17C10-PA functionalized AlOx nanoparticles onto a glass slide and a piece of cardboard in the same manner as described in Section 3.2.3. The average size of the particles (~50 nm), in conjunction with their "Teflon-like" functionalization, worked in unison to create a superhydrophobic surface wherever they were sprayed. The spray-coated layer created a randomly nanostructured surface with substantial roughness, which is required for creating an superhydrophobic surface [87]. Moreover, the sprayed

Applications

69

nanoparticles have already been rendered hydrophobic, and thus the film required no further treatment. So by providing both requirements of low surface energy and high surface roughness, a superhydrophobic surface was achieved with a single step process of spray coating. Furthermore, the coating was essentially transparent in up-close circumstances (Figure 4.11a). However, in reality it was only semitransparent as evidenced by Figure 4.11b. Additionally in Figure 4.11c several water based liquids were placed upon the coated glass. The same situation can be observed in Figure 4.11d except instead of glass, the coating was sprayed over a piece of cardboard. A video demonstrating the concept can be found online (Appendix Figure 7.6). Interestingly, placing the droplets in an orderly fashion over the coating was a dauntingly impossible task. To facilitate the matter, a small scratch on the film was purposely made wherever a droplet was placed.

a)

Coated

Uncoated

b)

c)

d)

Beer

Au NP

Milk

UV Ink (fluorescein)

Iron oxide NP

Cerium oxide NP

Figure 4.11.: Photographs of spray coated hydrophobic coatings on glass and cardboard. a) Coated and uncoated glass slide for up close transparency comparison. Droplets of different water based liquids dispensed on the top of coated (c) glass slide and (d) cardboard.

Unfortunately, the spray coated nanoparticle films could be removed easily from the glass substrates by weak mechanical means, such as finger tips or a tissue paper. This is expected, as the particles are coated with a "Teflon-like" material, which does not allow for any adhesion to the glass. As for the cardboard coating, a different situation occurred. In this case, the hydrophobic coating showed much robust mechanical stability. This is because of the rougher and fibrous nature of the cardboard surface

70

Applications

which traps the nanoparticles. As a matter of fact, we believe that this roughness acts in a similar way as the strategy that is employed to adhere a Teflon coating onto a metal surface (cooking pans) [102]. In short, the technique consists on increasing the roughness of the metal surface before applying the Teflon coating. Superhydrophobic coatings are currently a highly researched field as it can have many potential applications. The applications are very diverse, a notorious example is the packaging industry [103], automotive, aerospace and naval industry [104], where anti-icing [105], anti-fouling [106], self-cleaning [106] are of particular interest. Urban applications as well, were an hydrophobic coating was applied to street walls to dissuade people from urinating on them. All of these concepts and many more can be easily tackled with the approach presented above. However, further research is still required in terms of coating durability. One possible solution to improve this would be the addition of reactive or polymerizable groups to the nanoparticle shell for increasing the robustness of the film.

4.3. Self-assembled thin films This section presents fabrication of thin films as well, however, these thin films involve a special degree of freedom. These films incorporate the concept of self-assembly at a nanostructure level rather than just at the molecular level as it is the case for SAMs. Via surface functionalization of 2D and 0D NSs with SAMs, the creation of "smart materials" that self-assemble into a thin film in an orderly fashion was achieved.

4.3.1. Regio-selective deposition of nanoparticles A 2D SAM patterning process using photolithography had already been developed for other projects. It occurred to us that this same SAM patterning could be used create a pattern and selectively deposit nanoparticles via a covalent bond to a reactive surface. My undertaking in this project, consisted in the nanoparticles functionalization while Sebastian Etschel took charge of the far harder task of reacting them selectively. The fabrication of regio-selective nanoparticle thin films was achieved by employing reactive nanoparticles and a 2D reactive surface. To this purpose we employed the Huisgen 1,3-dipolar cycloaddition which was introduced in 2001 by Sharpless and Méldal [107], [108], which is also classified as the prototypical click reaction. In summary, the reaction results in a covalent bond between an azide and alkyne group via the formation of a triazole. It has been demonstrated that the click reaction can be used for the immobilization of functional materials on self-assembled monolayer (SAM) substrates [109], [110]. A click reaction is ideal for this purpose as it provides in principle, high yields and almost no byproducts making it ideal for nanoparticle thin film formation.

Applications

71

The regio-selective deposition of nanoparticles was possible by employing pre-patterned substrates with an inert surface in conjunction with a reactive surface. The contrast of these surfaces provided the driving forces for the self-organization. Using SAMs and photolithography, a chemically patterned substrate was fabricated as portrayed in Figure 4.12a [11]. This chemically patterned substrate consisted of SAMs with reactive head groups (red) vs. SAMs with inert groups (purple). The patterned substrate was then immersed in a dispersion of reactive nanoparticles, consisting of the corresponding building block showcased in Figure 4.12b and Figure 4.12c. The nanoparticles were functionalized by following the protocol described previously in Section 3.2. Measured FTIR spectra of the AlOx nanoparticles is shown in the appendix (Figure 7.8). The FTIR spectra of the particles clearly showed the reactive groups were still present after functionalization.

a)

b)

c)

Figure 4.12.: Schematic representation of the building blocks involved in thin film self-assembly. a) Chemically patterned 2D substrate. Red represents the reactive sites provided by the Azide-PA SAM molecule. Purple represents the inert sections of the substrate. the inertness is provided by the F 15C18-PA SAM molecule. b) Reactive Alkyne-PA functionalized nanoparticle. c) Reactive Azide-PA functionalized particle.

A critical parameter during the deposition steps was the stability of the nanoparticle dispersion. A nanoparticle dispersion with poor stability tended to agglomerate and resulted in poor film formation and selectivity. On the contrary, a highly stable dispersion yielded denser film formation and better region selectivity as the nanoparticles can be washed away by the solvent. This situation was of no

72

Applications

surprise, it is simply analogous to attempting to conduct a reaction were one of the reactants is not soluble. Whenever poor dispersibility was obtained after functionalization with the Azide-PA or the Alkyne-PA, a mixed SAM approach can help make the approach feasible. In a mixed SAM situation, a specific SAM provides dispersibility while the other SAM provides reactivity. The compromise for the better dispersibility would be the lesser amount of available reactive groups at the nanoparticles. Fortunately for this example (AlOx and ITO NPs), after functionalization of the nanoparticles with either the Azide-PA or the Alkyne-PA a stable dispersion was obtained. Therefore, there was no need for a mixed SAM approach. Nonetheless, there is work in progress by Sebastian Etschel were the mixed SAM approach was successfully implemented for regio selective deposition with TiO 2 nanorods and other NSs. Figure 4.13a shows a schematic representation as well as reaction conditions for the deposition of the first layer. The first deposited layer, consisted of AlOx nanoparticles with a median size of 50 nm functionalized with Alkyne-PA. Once the first layer was covalently bound, the immobilized nanoparticles still contained un-reacted chemical groups. Therefore, the deposition of a second layer atop the first one was possible. Thus, using the complimentary Azide-PA functionalized ITO nanoparticles a second layer was deposited (Figure 4.13b). Similarly, the ITO nanoparticles also had a median size of around 50 nm. Furthermore, a third layer of Alkyne-PA functionalized CeO2 nanoparticles was deposited [11], however, it is only portrayed in the cross sectional measurements. Theoretically, this process could be repeated many times as long as the complimentary reactive NSs are employed. Strictly speaking however, the regio-selective deposition is limited to a few layers. As with every new layer deposited, the coverage of the film gradually degrades due to lesser and lesser availability of reactive groups. Yet, further tuning the reaction conditions can help overcome difficulties and improve the overall film formation process. However, this was not thoroughly studied in this work. Figure 4.13c shows AFM images of the first and second layers as well as the measured cross-section height of the three consecutively deposited films. The first AlOx nanoparticle layer (depicted in red) had an average thickness of 60 nm which was in good agreement with the nanoparticle core size. This was a good indication, in terms of monolayer formation of the deposited nanoparticles. After deposition of the second layer, an increase of film thickness by an average of 47 nm was measured. This was also in good agreement with the ITO nanoparticle core size and thus also serves as a good indication of a monolayer coverage. Lastly, a layer of CeO2 nanoparticles with core size of 55 nm was deposited atop the ITO layer. Unfortunately, the CeO2 nanoparticle dispersion did not showed as good dispersibility after functionalization as the AlOx and the ITO nanoparticle dispersions. The dispersibility however, was enough for processing. After deposition of the CeO2 nanoparticles, an average increase of thickness around 150 nm was measured. This was not in agreement with the CeO2

Applications

73

nanoparticle core size and was attributed to multilayer formation due the inferior dispersibility of the CeO2 dispersion. Finally, in Figure 4.14 several SEM images of the selectively fabricated nanoparticle thin films are showcased. It becomes clear from the images that the nanoparticles were present only at the reactive surface with striking fidelity. The latter considering the arrangement of the nanoparticles was driven purely by self-assembly. In summary, the presented technique offers plenty of versatility via a dynamic layer by layer nanoparticle thin film formation. The approach can be extended to variety of metal oxide materials which would allow for great variability of thin film stacking combinations. More importantly, the NP films conformed to the chemically pattern on the 2D substrate. In general, the technique leaves open the possibility for the fabrication of other "smart" materials which can potentially self organize by using covalent interactions. c)

a)

1st layer AlOx

1st layer 1st layer AlOx 2nd layer ITO

b)

400 AlOX

350

ITO CeO2

2nd

layer

Height [nm]

300 250 200 150 100 50 0 0

2

4

6

8

10 12 14 16 18 20

Width [m]

Figure 4.13.: a) Schematic representation of the first deposited layer and reaction conditions. b) Schematic representation of the second deposited layer and reaction conditions. c) AFM scans of first and second selectively deposited thin films. AFM cross-sectional height measurement of the first and second layers.

74

Applications

Figure 4.14.: SEM images of selectively deposited nanoparticles.

4.3.2. Block co-polymer phase matching I worked with Johannes Kirschner and Tobias Rejek on this project by providing them with hydrophilic or hydrophobic metal oxide nanoparticles. Here, I briefly write about their work regarding the employment of the functionalized nanoparticles. A block copolymer which exhibits nanophase separation was employed as a 3D matrix in an endeavor to achieve the self-organization of nanoparticles within the matrix. The copolymers in this case, consist of polystyrene as an hydrophobic block and polyethylene oxide as an hydrophilic (Figure 4.15a). The phase separation of the block copolymers occurs due to the orthogonality of the copolymer blocks. First, in order to verify that the phase separation of the copolymer occurs. The block copolymer was dissolved in toluene and spun coated atop a SiO2 wafer. Figure 4.15b shows an AFM image of a spun coated block copolymer film which exhibited phase separation. The brighter areas in the image correspond to the polystyrene phase while the darker ones to the polyethylene oxide phase. The average lateral distance within phases was 20 nm. To achieve the self-assembly of the nanoparticles into the specific phase, the block copolymer was solubilized in toluene. However, before spin coating the copolymer, the hydrophobic or hydrophilic metal oxide nanoparticles were dispersed in conjunction with the copolymer in the same solution. To this end, metal oxide nanoparticles with a size between 4-10 nm were functionalized to exhibit either an hydrophobic or an hydrophilic surface. Naturally, the overall size of the nanoparticles must be smaller than the measured lateral distance between the phases of the film (20 nm). The hydrophobic nanoparticles consisted of nanoparticles with a shell of an alkyl chained phosphonic acid SAM. This

Applications

75

alkyl chained SAM rendered the particles highly dispersible in toluene. The particle and copolymer were then just mixed to desired concentrations resulting in a stable solution-dispersion in toluene. The hydrophilic particles on the other hand, consisted of more elaborated dispersions which were thoroughly described already in Section 4.1.4.

a)

b) Hydrophilic particle

Hydrophobic particle

c)

Hydrophilic particles in the PEO phase

Figure 4.15.: a) Molecular structure of block copolymers. The hydrophobic phase is composed of polystyrene (red) while the hydrophilic phase is composed of polyethylene oxide (blue). b) AFM image of a phase separated spin coated block copolymer thin film. c) SEM image of a phase separated spin coated block copolymer thin film with embedded Fe3O4 hydrophilic nanoparticles in the corresponding phase.

Figure 4.15c shows a SEM image of a spun coated copolymer film with hydrophilic nanoparticles. The film still exhibits phase separation, however, it shows a swollen polyethylene oxide phase. A possible reason for the swelling could be the that the hydrophilic nanoparticles are within the phase. Therefore, causing a thickening of the hydrophilic copolymer phase. However, the exact reason for this swelling is not clear. Understanding how the nanoparticles affect the copolymer phase separation in terms of concentration and whether the nanoparticle size or even organic shell can also have an

76

Applications

impact still remains a challenge [96]. Yet unmistakably, whichever nanoparticles are exposed in Figure 4.15c lie within the polyethylene oxide phase. The latter is a pretty good indication towards the overall validity of the approach. Furthermore, a similar situation occurs for hydrophobic particles. In conclusion, a system with the potential to fabricate via self-assembly a 3D matrix of metal oxide materials was briefly described. In principle, the functionality of the thin-film can be changed simply by changing the core of the metal oxide material, while the location of the nanoparticles can be changed by the organic shell. Still some fine tuning remains to be done regarding the impact of the nanoparticles on the overall phase separation and the ideal nanoparticle vs. copolymer concentrations.

4.4. Polymer composites We had developed particles with a polymerizable head group. Naturally, the idea came to use them during an in situ polymerization in order to improve the mechanical properties of the polymer. This task was performed by Simon Scheiner, while I simply found a way to properly functionalize the nanoparticles. A brief description of the idea is provided below. Nanoparticles were functionalized with a SAM having a polymerizable head group. The SAM molecule employed for this purpose is shown in Figure 4.16a. The MMA-PA SAM contains the equivalent of a methyl methacrylate (MMA) molecule as a head group. This head group can polymerize with other free MMA molecules (Figure 4.16b). Therefore, MMA was employed as monomer for the polymerization reaction. The polymerization of MMA results in the formation of Poly(methyl methacrylate) (PMMA) (Figure 4.16c). The FTIR spectrum of the functionalized particles shows clear evidence of the polymerizable head group being present at the AlOx nanoparticles surface (Figure 4.16d). After functionalization with MMA-PA the AlOx particles formed a moderately stable dispersion under methanol and toluene. Unfortunately, this degree of stability was not good enough to achieve an even distribution of the nanoparticles along the volume of the solution. It is important for this approach, that a stable dispersion is obtained under the polymerization conditions. A stable dispersion is required in order to produce a polymer composite with an even distribution of nanoparticles along its volume. Serendipitously, in a "like dissolves like" situation, the particles were highly dispersible in the liquid monomer (MMA), this served as an excellent dispersion medium for the nanoparticles. For testing, cylindrically shaped specimens were fabricated. These probes consisted of pure PMMA (Figure 4.16e), a PMMA composite with unfunctionalized AlOx nanoparticles (Figure 4.16f) and a PMMA composite with functionalized AlOx MMA-PA nanoparticles (Figure 4.16g). The pure PMMA probes were transparent, while the probes that contained particles were milky white. The polymerization of MMA was carried via radical polymerization. The radical starter was added to the MMA contained in a tube with or without nanoparticles and it was cured in ambient air at 50 °C

Applications

77

overnight. Afterwards, tensile tests were performed on the probes (Figure 4.16h). Consistently, the probes comprised of unfunctionalized particles and PMMA exhibited the worst mechanical properties. Followed by the pure PMMA probes with better results. The probes with the MMA-PA functionalized particles were consistently the most mechanically robust. The latter was attributed to the covalent bond between the particles and the polymer matrix. Furthermore, the nanoparticle core also acts a 3D branching point for the polymer matrix which increases the degree of cross linking of the polymer as graphically depicted in Figure 4.16i. a)

c)

b)

d) T (%)

MMA-PA

3800

3600

3400

3200

3000

2800

2600

2000

1800

1600

Wavenumber (1/cm)

1400

1200

1000

800

Wavenumber (1/cm) h)

1000

Stress [N/mm²]

e)

f)

g)

800 600 400 PMMA Pure PMMA/AlOx PMMA/AlOx-MMA-PA

200 0

0

1

2

3

4

Strain [%] i)

Figure 4.16.: a) Molecular structure of MMA-PA. b) Molecular structure of MMA. c) Molecular structure of PMMA. d) Color coded FTIR ATR spectrum of AlO x nanoparticles functionalized with MMA-PA. e) Cylindrical PMMA probe. f) Cylindrical PMMA probe with unfunctionalized AlO x nanoparticles. g) Cylindrical PMMA probe with MMA-PA functionalized AlOx nanoparticles. h) Strain test curves of PMMA and PMMA composites. i) Artistical rendition of nanoparticles covalently attached to a polymer matrix.

This increased degree of cross linking, improved the mechanical robustness compared to the pure and linear (uncrosslinked) PMMA. Finally, a qualitative proof of the increased degree of cross linking was

78

Applications

observed. The composite was exposed to a flame and non-melting behavior was observed. Instead, the composite burned, turning into a black material with bubbles. The latter was attributed, to the high degree of cross linking modified the thermo plastic behavior as compared to the pure PMMA. Polymers with nanoparticles as a filler have been researched extensively [96], [111]–[113]. However, there are fewer reports about the nanoparticle filler being covalently bound as part of the nanoparticle matrix. Usually, the polymer-filler interactions are of a weaker nature [112], [113]. In here, a nanoparticle filler has been employed successfully to improve mechanical properties of the polymer. Evidence suggesting the covalent attachment of the nanoparticles was also briefly described. However, this project is still work in progress and more research is still required in order to identify the proper polymerization conditions and ideal nanoparticle polymer ratio.

4.5. Nanooncology Water dispersible metal oxide nanoparticles under neutral pH conditions, are highly desirable conditions for bio-applications. Towards this purpose, I was given a set of magnetic nanoparticles which I should try to make dispersible in water. After some back and forth we obtained very good results in terms of nanoparticle stability. This is a brief summary of the results obtained so far from this collaboration with Melek Kizaloglu and Stefanie Klein from Prof. Kryschi group. Fe3O4 and CoFe3O4 nanoparticles were functionalized with a variety of phosphonic acid molecules known to us to improve water dispersibility under neutral pH conditions. The nanoparticles were synthesized by Melek Kizaloglu from Prof. Kryschi group. The synthesis of the particles was performed with an adaptation of the methods described in literature [114], [115]. After synthesis, the nanoparticles were washed thoroughly with pure DI water for removal of any excess reagents. Use of pure DI water is of paramount importance, given that when we employed a phosphate-buffered saline (PBS) buffer solution for washing, the nanoparticle surface appeared to get "functionalized". We were able to confirm this by doing an FTIR analysis of washed particles with PBS, the particles showed the distinct broad valley of the P-O and P=O vibrations. The use of a PBS buffer solution was therefore discontinued for any part of the process. The molecules in employed for functionalization are shown in Figure 4.17a. Of particular attention is the molecule Imidazolium-PA (1) which makes its debut in this section, but more on this particular molecule is described further below. The other molecules (2 and 3) have already been described in Sections 3.2.3 and 4.1.1 for similar purposes and conditions. The obtained nanoparticles were characterized in the same way as all the functionalized NSs in this work. The general functionalization procedure can be reviewed in Section 3.2 and in detail in Section 6.8 as well as supporting FTIR-ATR spectra of the nanoparticles before and after functionalization (appendix Figure 7.9).

Applications

79

First proof of improved nanoparticle dispersibility is Figure 4.17b, which shows DLS measurements in DI water of pristine and functionalized Fe3O4 nanoparticles. The pristine nanoparticles showed a median size around 70 nm. As expected, after functionalization the nanoparticle dispersions stability was greatly improved. Naturally, a smaller median size of 30 nm was measured for the functionalized nanoparticles with all 1, 2 and 3 SAM molecules due to the decreased level of agglomeration after functionalization. A similar situation was found to be the case for the DLS measurements of the pristine and functionalized CoFe3O4 nanoparticles. The stability of the nanoparticles can also be confirmed visually, photographs of the Fe3O4 and CoFe3O4 dispersions can be seen in Figure 4.17c and Figure 4.17d. It can be observed from the photographs how the functionalized nanoparticles form perfectly stable transparent dispersions. Whereas the pristine particles precipitate or form a turbid dispersion. Both photographs were taken 12 hours after redispersion by sonication. It becomes clear from the photographs that all the SAM molecules improved the dispersibility of the nanoparticles under neutral pH conditions. Furthermore, the particles remained stable for several months, especially in the particular case of the Imidazolium-PA were the particles have remained stable over a year. Having access to stable dispersions of the synthesized particles is already an interesting result, as the use of the pristine unstable dispersions is very problematic or even unusable. The zeta potential of the particles was measured to obtain some insight into the stabilization mechanism of the particles. The measured potentials are resumed in Table 4.3. The pristine nanoparticles had a weak positive zeta potential which rendered them mildly stable. Not surprisingly, the nanoparticles functionalized with Imidazolium-PA had positive zeta potentials over 60 mV which explains their remarkable stability. We can be sure then, that the Imidazolium-PA successfully generates a strong positive charge at the nanoparticles surface, as well as that the particles are well dispersed via a strong electrostatic stabilization. Furthermore, the Imidazolium-PA particles have remained stable over several months, with no apparent signs of agglomeration. Positively charged nanoparticles have been demonstrated to possess a better cellular intake time and again [116], [117], [118] as compared to neutral or negatively charged particles. This effect is particularly interesting for this application as not only provides a good stabilization mechanism but also could potentially increase cellular uptake. As to the particles functionalized with the molecules 2 and 3, the measured zeta potential was weaker but positive around 15-30 mV suggesting an electrosteric stabilization scheme. Since the zeta potential remained positive nonetheless, these particles were also interesting in terms of cellular uptake.

80

Applications

b)

a)

30 Pristine Imidazolium-PA H(OC2H4)3-PA

25

1

Number (%)

2

CH3(OC2H4)3-PA

20 15 10

3

5 0

0

20

40

60

80 100 120 140

Size (nm) d)

c)

Pristine

1

2

3

1

Pristine

2

3

Figure 4.17.: a) Molecules used for functionalization of nanoparticles. b) DLS distribution of Fe3O4 pristine and functionalized nanoparticles. c) Photograph of Fe3O4 pristine and functionalized NP dispersions 12 hours after being redispersed on DI water via sonication. d) Photograph of CoFe3O4 pristine and functionalized NP dispersions 12 hours after being redispersed on DI water via sonication.

Table 4.3.: Zeta potentials of the Fe-Fe and the Fe-Co nanoparticle dispersions before and after functionalization with different molecules which portrayed in Figure 4.17.

Metal oxide

Pristine

1

2

3

Fe3O4

+14 mV

+68 mV

+30 mV

+25 mV

CoFe3O4

+20 mV

+60 mV

+12 mV

+15 mV

Biocompatibility of the pristine and functionalized particles was tested on human endothelial cancer cells as well as healthy cells by Stefanie Klein from Prof. Kryschi group. The results of the biocompatibility tests of CoFe3O4 nanoparticles in function of dosage are shown in Figure 4.18. The pristine CoFe3O4 nanoparticles were generally toxic towards both cancer and healthy cells. This effect was observed when a dosage of 10 ug/ml was administered to the cells, which yielded a survival rate of approximately 30%. However, when the CoFe3O4 particles were functionalized with either of the ligands, there was only a small decrease in cell viability observed. This is very interesting as it allows

Applications

81

the CoFe3O4 particles to be present in the cell media without any toxic impact. Thus we hypothesized, that the organic shell surrounding the nanoparticles acts as an isolation barrier, which isolates the toxic particle core by presenting a benign organic shell at each of the nanoparticles surface. This effect in reduced toxicity itself is very interesting, as it can potentially be extended to other toxic metal oxide materials that need to be incorporated into a cell as well. The biocompatibility tests of the Fe3O4 particles were also successful, though less interesting, as there was no dramatic difference between pristine and functionalized particles due to the decreased toxicity of the pristine Fe 3O4 nanoparticle core. These results were placed in the appendix Figure 7.10. Having established that the interactions between the functionalized particles and cells are of a non-toxic nature both in healthy and in cancerous cells, further studies were performed regarding their potential as nanooncology agents.

a) Breast cancer cells

Cell viability (% of control)

150

Uncoated Co-ferrite Pristine CoFe 3O4 NP NP 1Imidazol-Co-ferrite CoFe3O4 2Hydroxy-Co-ferrite CoFe3O4 NP 3Methoxy-Co-Ferrite CoFe3O4 NP

100

50

0 1

5

10

Concentration [g/mL]

b) Healthy cells Uncoated Co-ferrite Pristine CoFe 3O4 NP NP 1Imidazol-Co-ferrite CoFe3O4 2Hydroxy-Co-ferrite CoFe3O4 NP 3Methoxy-Co-Ferrite CoFe3O4 NP


100

50

0 1

5

10


Figure 4.18.: Biocompatibility of pristine and functionalized CoFe3O4 nanoparticles in function of dosage. a) Biocompatibility results on breast cancer cells. b) Biocompatibility results on healthy cells.

The iron and cobalt +2 ions on the Fe3O4 and CoFe3O4 are toxic to cells due the generation of oxygen radicals via the Fenton reaction [119], [120]. This reaction is depicted in Figure 4.19a. Therefore, in order to monitor oxygen radical formation, a labeling dye (dichlorofluorescein) was applied to the cells. Under normal circumstances dichlorofluorescein is non-fluorescent, however, if the dye reacts with an oxygen radical the dye exhibits a strong fluorescence at around 525 nm when excited at around

82

Applications

488 nm. Consequently, this dye is often used in oxygen radical detection and quantification via fluorescence microscopy [121]. It is important to note that a certain amount of oxygen radical generation occurs inside the both cancerous and healthy cells under normal circumstances. Therefore, the amount of fluorescence of the cells must be measured with and without nanoparticles. To this end, Figure 4.19b and Figure 4.19c have been normalized in the vertical axis as the percentage of fluorescence increment compared to the amount of fluorescence exhibited by cells that were not exposed to any CoFe3O4 nanoparticles. Moving on, Figure 4.19b and Figure 4.19c are proof that the particles are present inside the cells, as an increase in oxygen radicals can only be due to the presence of the nanoparticles within the cells membrane. In the same figures a comparison between nonirradiated and radiated samples is shown. Irradiation of the cells was performed with x-rays and consisted of one dosage of 1 Gy, which corresponds to dosage a human patient would get. After irradiation the nanoparticles surface is exposed and the generation of oxygen radicals is dramatically increased in the case of cancer cells (Figure 4.19b). However, in the case of healthy cells there is no impact even after irradiation (Figure 4.19c). The dramatic increase of radicals in the cancer cells arises from their higher metabolic rate (as compared to healthy cells), which speeds up the occurrence of the Fenton reaction eventually resulting in their demise within a shorter timeframe.

a) Fe2+ + H2O2

Fe3+ + OH– + OH•

Co2+ + H2O2

Co3+ + OH– + OH•

200

100

100

50

150 200

100

100 100 50

50

0 0 P P P P P P P P Pristine Pristine N3 N3 N N1 N2 N N1 N2 ir te ir te ir te ir te ir te ir te ir te ir te r r r r r r r r fe fe fe fe fe fe fe fe ooooooooC C C C C -C -C -C y- of iron xand yy- before and after te a) Fenton Figure 4.19.: cobalt. b) Comparisonateof oxygen ol reaction ol radicalxygeneration o ox ox o az az oa o r r h h d d c c i i n n et et yd yd Im cells exposed H H radiationU in cancer to Mpristine and functionalizedUCoFe3O4Imnanoparticles. c) MComparison of oxygen

0

Im

U nc oa te d

non-irradiated irradiated

Im

0

150

250

non-irradiated 300 irradiated 200

C ofe

150

250

non-irradiated 200 irradiated

U nc oa te d

200

c) Healthy cells rr ite id az N P ol -C ofe rr H it e yd ro N U P xy of fluorescence Increase nc -C oa ote control) of (% fe rr M dC i et te o ho -fe N P Im xy rri t id e C az N ool fe P -C rr ite ofe N rr H it e P yd ro N P xy -C ofe rr M ite et ho N P xy C ofe rr ite N P

300

Increase of fluorescence (% of control)

rr ite id az N P ol -C ofe rr H it e yd ro N PIncrease of fluorescence xy -C o(% of control) fe rr M ite et ho N P xy C ofe rr ite N P

250

C ofe

Increase of fluorescence (% of control)

b) Breast cancer cells

radical generation before and after radiation in healthy cells exposed to pristine and functionalized CoFe3O4 nanoparticles.

In summary, we have presented a non-toxic pathway for the introduction of normally toxic nanoparticle species into the cytoplasm of cells. The toxicity of this nanoparticles can later be reactivated by irradiation of the cells. After which, the impact of this toxicity is higher for cancerous

non-irradiated irradiated

Applications

83

cells as it is to healthy cells. This contrast in toxicity is possible as the mechanism exploits the higher metabolism of the cancerous cells which results higher cell mortality within a certain timeframe as compared to healthy cells. Therefore, providing preferential targeting of cancer cells. In addition to that, the same pathway also provides outstanding nanoparticle stabilization under water at neutral pH, as well as providing a positive charge at the nanoparticle surface which is highly beneficial for nanoparticle intake into cells. In addition to all this, this highly positively charged particles are also being studied for siRNA delivery into cells with already some initial promising results. We have also indentified that this approach might have further potential in developing new MRI contrast agents.

4.6. Magnetic water cleaning The surface of magnetic particles (metal oxide based) can be fine tuned to possess a variety of characteristics as it has been demonstrated along the course of this thesis. Understandably, having a magnetic core allows for magnetic extraction of nanoparticles from solution. Based on this, we have devised to employ such particles for decontamination of water sources acting as a sort of magnetic washing agents (Figure 4.20a). In such concept, as the nanoparticles are dispersed in the water, their finely tuned shell will selectively attract a specific contaminant or a variety of contaminants. Once the particles are loaded with a pollutant, they are removed via magnetic means removing the contaminants along with them. Even more, once extracted, the particles could potentially be washed and reutilized. The washing and reutilization is dependent on the nature of the attraction force between the nanoparticle shell and the extracted pollutants. Weak interactions between the shell and the pollutant will render the particles washable and reusable, whereas, strong interactions would not as removal of the pollutant from the shell would be difficult. Two potential applications have been currently identified. Lipophilic nanoparticles for removal of oil from water (Figure 4.20b) and ion caging nanoparticles for removal of metal ions from water (Figure 4.20c). Figure 4.21 showcases the first promising results with magnetic nanoparticles for oil removal. These measurements were performed by Marco Sarcletti and Tobias Luchs. The experiment, consisted in introducing a certain amount of the lipophilic nanoparticles into a water-hydrocarbon mixture. Namely, n-heptane, isooctane and n-cyclohexane. Upon addition of the nanoparticles, the mixture was vigorously shaken. Afterwards, the nanoparticles were removed with a magnet and redispersed in pure toluene. After redispersion in toluene, the particles were magnetically removed again. The remaining toluene was analyzed via gas chromatography and mass spectroscopy (GC-MS). This allowed to determine the exact amount of the magnetically extracted hydrocarbons from the water mixture. The obtained extraction rates varied from hydrocarbon to hydrocarbon, heptane being the most efficiently extracted hydrocarbon, with extraction rates of four times the mass of the nanoparticles.

84

Applications

Furthermore, the simplicity of functionalization process with phosphonic acids, potentially enables the mass production of the core shell magnetic particles. In addition, the high affinity of the phosphonic acid to the metal oxides provides an effective, efficient and robust modification of the nanoparticles. The latter is especially true, when compared to other popular SAM anchor groups such as carboxylic acids and catechols.

a)

Washing of cleaning agent and reuse

Magnet

Polluted water

Mixing of cleaning agent

Gathering of pollutants via magnetic field

b)

Removal of pollutants

c)

Oil removal agent

Ion removal agent Magnetic core with ion caging shell

Adsorbed hydrocarbon shell

Magnetic core with lipophilic shell

Trapped ion

Figure 4.20.: Pollutant extraction scheme from a liquid via functionalized magnetic nanoparticles.

50 45

particles weight hydrocarbon weight

40

Weight (mg)

35 30 25 20 15 10 5 0

Figure 4.21.: Extracted hydrocarbon weight vs. nanoparticle mass.

Applications

85

4.7. Chapter summary By employing a variety of simple schemes, the potential applications of organic-inorganic hybrid NSs was demonstrated. A description of the applications and projects that have been developed over the course of this work were described. Being of primordial importance to solution processing, the applications regarding nanoparticle dispersibility were discussed first. Essentially, this consisted on adequately matching the organic shell of the NSs to the dispersion media. Examples regarding the dispersion of NSs in water, lower alcohols as well as non polar media was addressed. More complex dispersion schemes which involved a two level organic layer (double shell) which allowed for dispersion of NSs in orthogonal media were also described. This double shell systems consisted of small molecules double shells as well as polymeric double shells. Real examples of steric, electrostatic and electrosteric stabilization mechanisms were involved along this chapter. A de facto application of solution processing is thin-films. An example concerning the fabrication of dielectric thin films from NSs dispersions via spin coating was described. Thicker films of NSs were also fabricated via spray coating on a variety of surfaces. These semi transparent films rendered the surfaces extremely hydrophobic, this being is an interesting effect in terms of coating applications. More sophisticated solution based thin film fabrication techniques were also detailed, in such, the NSs themselves self-assembled into specific configurations. One technique exploited covalent reactions between nanoparticles and a 2D patterned surface, whereas another technique relied on orthogonal properties as the main driving force. Other interesting hybrid materials were also described, such as polymer composites in which the NSs themselves formed part of the polymer matrix. The latter was possible by grafting SAMs with polymerizable head groups on the NSs surface. The polymerizable nanoparticles act as a cross linker by providing a 3D branching point for polymer chain formation. This potentially results in stronger polymer materials due to the increased degree of cross linking. Furthermore, the toxicity of cobalt-iron oxide nanoparticles was mitigated by covering the nanoparticles with a bio-compatible organic shell. The latter allowed for their employment in a nanooncology application. Furthermore, this same particles are also being studied for siRNA delivery into cells, with already some initial promising results. This also opened the way for their potential employment as a MRI contrast agent as well, however to this date this has not been investigated. Finally, a proof of concept application were magnetic nanoparticles can be employed as magnetic cleaning agents for contaminants present in water was also described.

5. Conclusion and outlook

This thesis serves as a comprehensible guideline for solution processing and surface functionalization of metal oxide nanostructures with organic SAM molecules. In addition, it also serves to demonstrate the vast assortment of applications enabled by the orderly and controlled incorporation of organic chemistry into inorganic components. In the first part, the theoretical background regarding nanostructures, nanostructured materials, self-assembled monolayers and the solution processing of nanostructures were addressed. This serves as a basis for the clear comprehension of this work. The second part describes a step by step experimental and characterization procedure. The first step consisted in analyzing the starting metal oxide material. The dry nanopowder materials were assessed via a non-destructive technique (FTIR-ATR). This allowed to detect the existence of organic components and other impurities that may have been present in the starting materials. Naturally, after functionalization, a FTIR-ATR characterization was performed as well. This allowed to easily take a glance into the organic composition of the functionalized 0D and 1D NSs. In some instances, the impurities were removed after functionalization, leading to the correct functionalization of the surface. However, in other situations the impurities were not removed, this resulted in poorly controlled NSs functionalization. Consistently, in all cases in which the material was pre-characterized as "pure" the functionalization was successful. Therefore, FTIR-ATR was demonstrated to be a valuable tool to recognize commercial 0D and 1D metal oxide NSs with good functionalization potential. It also proved valuable for the identification of specific chemical groups after functionalization of 0D and 1D NSs. Furthermore, it allowed to easily detect potential anomalies after functionalization, such as NSs etching or multilayer formation. TGA was successfully employed for the quantification of the adsorbed SAM molecules at the surface of the 0D and 1D NSs. This allowed for the determination of reaction conditions which resulted in a fully saturated surface. In combination with a size distribution obtained by DLS, the grafting density of the SAM molecules was calculated. The grafting density was in good agreement with literature and expected values. When obtaining reasonable values, it was possible to fully discard functionalization side effects, such as etching or multilayer formation. On the other hand, if the calculated grafting density values were inconsistent, the process/material was reevaluated/discarded. Formation of mixed SAMs on 0D NSs involving a stoichiometric variation of the SAM molecules was demonstrated. The SAM molecules employed were of a hydrophobic and hydrophilic character. This 87

88

Conclusion and outlook

allowed to create measurable contrast in terms of surface energy. The formation of mixed SAMs was confirmed by the formation of nanoparticle films onto a flat surface via spray coating. SCA measurements with different liquids allowed for the calculation of the surface energy of the films. The surface energy of the films varied from 0.45 to 67 mN/m. The films behavior gradually varied from superhydrophobic to superhydrophilic. FTIR-ATR spectra of the nanopowders confirmed the presence of the mixed SAMs at the surface. Interestingly, the mixed monolayers had a negative impact on the dispersibility of the nanoparticles in solution. The fully hydrophobic nanoparticles had a zeta potential of -50 mV which rendered them highly dispersible in 2-propanol. Whereas, the fully hydrophilic particles exhibited a zeta potential of +50 mV making them highly dispersible in 2-propanol as well. This high zeta potentials are believed to originate from the dipole moments of the SAM molecules. However, mixed monolayers exhibited zeta potentials of decreasing magnitudes, reaching values near 0 mV when a 1:1 ratio of the SAM molecules was employed. The latter rendered the particles completely unstable in 2-propanol. This effect on the zeta potential was attributed to the difference in dipole moments of the molecules, effectively cancelling each other when a 1:1 proportion was used. Therefore, hindering the particles solution processability when employing specific SAM ratios. The functionalization of diverse metal oxide NSs was demonstrated with carboxylic, phosphonic acids and catechols as anchor groups of the SAM molecules. Consistently, the phosphonic acid molecules were the most effective, efficient and robust when compared to the other anchor groups. Carboxylic acids and catechols were also effective, but not as efficient or robust as phosphonic acids. Therefore, the majority of this work is based on phosphonic acid molecules. The final chapter focuses on the applications derived from the functionalization of the metal oxide NSs. The applications encompassed different fields, e.g. ecological, electronics, biology, polymers, coatings as well as bottom up fabrication approaches. All of which, were made possible by the incorporation of phosphonic acid SAMs onto 0D and 1D NSs. In addition, the potential of having access to precise fabrication and tuning of core-shell hybrid materials via self-assembly, was demonstrated. Ultimately, the simplicity of the process necessary for the modification of the NSs, makes it feasible for this work to materialize into the real world, rather than remain strictly as a scientific essay.

6. Characterization and experimental procedures

6.1. SCA Static contact angle (SCA). SCA measurements were investigated by the sessile drop method utilizing DI water, formamide and diiodomethane (1.0 μL) as probe liquids (Dataphysics OCA, Data Physics Instruments GmbH, Germany).

6.2. DLS Dynamic light scattering. Nanoparticle size distribution was obtained by measuring a 0.2 wt. % stable dispersion of nanoparticles in an appropriate solvent (Zetasizer Nano, Malvern, U. K.). Ideally, before measurement particles were passed through a 0.8 µm membrane filter. However, when a filter was not employed most of the times the measurements were the same with or without filtering. Zeta potential. Nanoparticles dispersed in 2-Propanol or DI water (0.2 wt. %) were placed inside a folded capillary cell (DTS1070, Malvern, U. K.) for carrying out zeta potential measurements (Zetasizer Nano, Malvern, U. K.). Zeta potential values were determined by measuring the electrophoretic mobility of the nanoparticles by employing laser Doppler anemometry technique. In both DLS and zeta potential, depending on the material and solvent the concentration sometimes needed to be slightly adjusted. The concentration adjustment was made until a signal of 100-300 kilocounts per second (Kcps) was obtained by the Zetasizer Nano. From experience, this is the "sweet spot" between too much or to less concentrated.

6.3. FTIR-ATR Fourier transform infrared spectroscopy attenuated total reflection (FTIR ATR). FTIR ATR measurements of the nanoparticle dry powder were obtained (IR Prestige-21, Shimadzu, Japan) utilizing an ATR setup with a Diamond/ZnSe crystal plate (MIRacle ATR, Pike Technologies, U.S.A.). Transmission spectra were collected at a resolution of 8 cm-1 (64 scans) by clamping dry nanoparticle powder to the ATR crystal plate. The nanopowder is required to be solvent free, it was dried by heat close to the boiling point of the solvent or by a combination of vacuum and heat.

89

90

Characterization and experimental procedures

6.4. TGA Thermogravimetric analysis (TGA). TGA of the nanoparticle dry powders were carried out under a N2 atmosphere at a heating rate of 10 °C/min (Q500, TA Instruments, U.S.A.) or (TG 209 F1 Libra, Netzsch, Germany)

6.5. Electrical Capacitors and TFTs were fabricated on silicon wafers with 100 nm thermally grown oxide and flexible TFTs onto polyethylene naphthalene ﬁlms (PEN, DuPont Teijin Films, U.K.) with a thickness of 125 µm. Metal electrodes and organic semiconductor were thermally evaporated under vacuum utilizing a shadow mask for patterning. After evaporation of the aluminum capacitor or gate electrodes, the substrate was treated with a 3 min oxygen plasma treatment at 200 W at pressure of 0.2 mbar (Pico, Diener electronic GmbH, Germany). The dielectric layers were spin coated (500 rpm prespin and 2000 rpm final speed) from (0.6 wt-%) solutions in water and from 2-propanol in the case of AlOx to avoid etching of the particles. All electrical characterizations were performed in ambient air (B1500A, Agilent, U.S.A.).

6.6. Spray coating For the spray coating of nanoparticle films. Nanoparticle films were manually spray coated onto Si/SiO2 wafers with a 100 nm thermal oxide layer. Before deposition of the films, the wafers were cleaned with at least a 3 min oxygen plasma treatment at 200 W at pressure of 0.2 mbar (Pico, Diener electronic GmbH, Germany). Substrate was heated up to temperature close to the boiling point of the solvent employed during the spray coating process. The spray coating process was usually performed from 0.2 wt % solutions of NP. However, the concentration of the NP solutions was found to be a rather arbitrary parameter. Therefore, the formation of the films was carried out until the film was optically evident. Afterwards, the film morphology and coverage was studied by AFM to ensure complete coverage of the original substrate surface.

6.7. Materials 

The AlOx, ITO and CeO2 nanoparticles employed were purchased from Sigma Aldrich (702129, 70460 and 643009, U.S.A.).



The TiO2 and Fe3O4 nanoparticles employed were purchased from PlasmaChem GmbH (PLTiO-NO and PL-A-Fe3O4, Germany).



The 30 nm TiO2 nanoparticles were specially ordered from Nanograde AG (Switzerland).



The ZnO nano rods were purchased from Sigma Aldrich (U.S.A.) and employed as received.

Characterization and experimental procedures 

91

The Fe3O4 and CoFe3O4 nanoparticles employed for nanooncology, were synthesized by Prof. Kryschi group.



Phosphonic acid molecules were purchased from SiKEMIA (France), Sigma Aldrich (U.S.A.), PCI Synthesis (U.S.A.) and employed as received.



Carboxylic acid molecules were purchased from Sigma Aldrich (U.S.A.) and employed as received.



Catechol molecules were purchased from Sigma Aldrich (U.S.A.) or synthesized by Prof. Hirsch group. The molecules were employed as received.



Polymers were purchased from Sigma Aldrich (U.S.A.) or Polymer Source (Canada). The polymers were employed as received.



Any other molecule not available commercially displayed during this work was synthesized by Prof. Hirsch group.

6.8. Functionalization procedures 6.8.3. AlOx (Sigma A.), ITO (Sigma A.) and TiO2 (Nanograde 30 nm) The nanoparticles originally come dispersed in a concentrated dispersion in 2-propanol. The dispersions were diluted with 2-propanol into 0.2 wt. % dispersions. 10 ml (0.2 wt %) of nanoparticles dispersed in 2-propanol were used for functionalization. Afterwards, 3 ml of a 10 mM solution of the desired functional molecule dissolved in methanol, ethanol or 2-propanol (depending on molecule solubility) was added. The dispersion was then sonicated for 30 min, yielding a slightly opaque, stable or unstable dispersion of functionalized particles. In order to remove the unreacted excess ligands, the particles were then centrifuged at (10K-14K RPM, 10-20 min) and a clear supernatant was removed, the particles were then redispersed in a washing solvent resulting again in a slightly opaque, stable or unstable dispersion. The washing procedure was repeated at least two more times. After washing the particles, they were dried overnight at room temperature under the negative pressure of a chemical hood. Alternatively, they were immediately redispersed in the solvent of choice.

6.8.4. CeO2 The nanoparticles originally come dispersed in a concentrated dispersion in water. The dispersions were diluted with DI-water into 0.2 wt. % dispersions. 10 ml (0.2 wt %) of nanoparticles dispersed in 2-propanol were used for functionalization. Afterwards, 3 ml of a 10 mM solution of the desired functional molecule dissolved in methanol, ethanol or 2-propanol (depending on molecule solubility) was added. The dispersion was then sonicated for 30 min, yielding a slightly opaque, stable or unstable dispersion of functionalized particles.

92


In order to remove the unreacted excess ligands, the particles were then centrifuged at (10K-14K RPM, 10-20 min) and a clear supernatant was removed, the particles were then redispersed in washing solvent resulting again in a slightly opaque, stable or unstable dispersion. The washing procedure was repeated at least two more times. After washing the particles, they were dried overnight at room temperature under the negative pressure of a chemical hood. Alternatively, they were immediately redispersed in the solvent of choice.

6.8.5. Fe3O4 (Plasmachem ~10 nm) The Fe3O4 nanoparticles originally come dispersed as a 3 wt. % dispersion in water. The dispersion was diluted with DI water in order to obtain a 0.3 wt. % dispersion. 5 ml of the diluted dispersion were employed for functionalization. The diluted nanoparticles form a transparent whisky colored dispersion. At this point, 3 ml of a 40 mM solution of the desired functional molecule dissolved in methanol, ethanol or 2-propanol (depending on molecule solubility) was added. Afterwards, the solution was sonicated for 30 min. After sonication, the impact of the functional molecules on the dispersibility of the nanoparticles becomes visually evident, the effect varies depending on the nature of the ligand employed. Excess unreacted molecules are still present at this point. The washing procedure to remove excess ligands varies depending on whether the dispersed nanoparticles are highly stable or not at this point. To remove the excess molecules from poorly stable dispersions, the dispersion is centrifuged (10K-14K RPM, 10-20 min) until the nanoparticle powder is isolated at the bottom of the container. The supernatant is removed and fresh solvent (methanol, ethanol or 2propanol) is added to the nanoparticles. The nanoparticles are then redispersed via sonication and centrifuged again to add fresh solvent. This procedure was repeated at least 2 times for all experiments. To remove the excess molecules from highly stable dispersions were centrifugation does not isolate the nanoparticles. The dispersions were dried using a rotavapor (@20 mbar 50 °C) to remove the DI water employed during functionalization and then redispersed in 5ml of ethanol. After redispersion in ethanol, the nanoparticles may or may not form a transparent whisky colored stable dispersion depending on the ligand employed. If the nanoparticles formed again a highly stable dispersion, heptane was added to the solutions until a slightly turbid solution was formed, at this point the particles can be centrifuged and the supernatant was removed. Then the nanoparticle powder was redispersed in fresh solvent. This procedure was repeated at least 2 times for all experiments. After washing the particles, they were dried overnight at room temperature under the negative pressure of a chemical hood. Alternatively, they were immediately redispersed in the solvent of choice. Heptane (or an anti-solvent) is required to reduce the solubility of the particles to allow for centrifugation of the particles. Otherwise the particles are so stable that a conventional centrifuge


93

cannot cause the particles to precipitate. Before the addition of heptane removal of all DI water is required since heptane and DI water are not miscible. DI water was used as the starting solvent for functionalization since the pristine nanoparticles showed a considerably higher degree of dispersibility in water than they did in alcohols.

6.8.6. TiO2 (Plasmachem 8 nm) 15 mg of the nanoparticles were immersed into 10 ml of DI water. The dispersions were sonicated until the nanoparticle powder appeared to be thoroughly dispersed with no visual indication of major agglomerates being present. The nanoparticles form a transparent almost colorless dispersion. At this point, 3 ml of a 40 mM solution of the desired functional molecule dissolved in methanol, ethanol or 2propanol (depending on molecule solubility) was added. Afterwards, the solution was sonicated for 30 min. After sonication, the impact of the functional molecules on the dispersibility of the nanoparticles becomes visually evident, the effect varies depending on the nature of the ligand employed. Excess unreacted molecules are still present at this point. The washing procedure to remove excess ligands varies depending on whether the dispersed nanoparticles are highly stable or not at this point. To remove the excess molecules from poorly stable dispersions, the dispersion is centrifuged (10K-14K RPM, 10-20 min) until the nanoparticle powder is isolated at the bottom of the container. The supernatant is removed and fresh solvent (methanol, ethanol or 2-propanol) is added to the nanoparticles. The nanoparticles are then redispersed via sonication and centrifuged again to add fresh solvent. This procedure was repeated at least 2 times for all experiments. To remove the excess molecules from highly stable dispersions were centrifugation does not isolate the nanoparticles. The dispersions were dried using a rotavapor (@20 mbar 50 °C) to remove the DI water employed during functionalization and then redispersed in 5 ml of ethanol. After redispersion in ethanol, the nanoparticles may or may not form a transparent stable dispersion depending on the ligand employed. If the nanoparticles formed again a highly stable dispersion, heptane was added to the solutions until a slightly turbid solution was formed, at this point the particles can be centrifuged and the supernatant was removed. Then the nanoparticle powder was redispersed in fresh solvent. This procedure was repeated at least two times for all experiments. After washing the particles, they were dried overnight at room temperature under the negative pressure of a chemical hood. Alternatively, they were immediately redispersed in the solvent of choice. Heptane (or an anti-solvent) is required to reduce the solubility of the particles to allow for centrifugation of the particles. Otherwise the particles are so stable that a conventional centrifuge cannot cause the particles to precipitate. Before the addition of heptane removal of all DI water is required since heptane and DI water are not miscible. DI water was used as the starting solvent for

94


functionalization since the pristine nanoparticles showed a considerably higher degree of dispersibility in water than they did in alcohols.

6.8.7. Fe3O4 and CoFe3O4 (Prof. Kryschi) Both Fe3O4 and CoFe3O4 nanoparticles were obtained as a dry powder and functionalized using the same procedure. It was found that crushing the nanoparticles prior to dispersion in water greatly speeds up the process. Therefore, using a mortar and pestle the nanoparticles were crushed until a small powder was obtained. Afterwards, 15 mg of the nanoparticles were immersed into 10 ml of DI water. The dispersions were sonicated until the nanoparticle powder appeared to be thoroughly dispersed with no visual indication of major agglomerates being present. At this point, 3 ml of a 40 mM solution of the desired functional molecule dissolved in MeOH for Imidazolium-PA or in IPA for glycol-PAs was added. Afterwards, the solution was sonicated for 30 min. After sonication, the impact of the functional molecules on the dispersibility of the nanoparticles becomes visually evident, the dispersions are now fully transparent whisky colored solutions. Any remaining non-dispersed major agglomerates were removed at this point by extracting the well dispersed nanoparticles with a pipette into a new flask. However, excess unreacted molecules are still present at this point. To remove the excess molecules the dispersions were dried using a rotavapor (@20 mbar 50 °C) and then redispersed in 5 ml of ethanol forming again a stable whisky like colored dispersion. Heptane was added to the solutions until a slightly turbid solution was formed, at this point the particles where centrifuged and the supernatant was removed. Then the nanoparticle powder was redispersed in ethanol again. After washing the particles, they were dried overnight at room temperature under the negative pressure of a chemical hood. Finally, they were redispersed as a dry powder in DI water in the desired concentrations. Heptane (or an anti-solvent) is required to reduce the solubility of the particles to allow for centrifugation of the particles. Otherwise the particles are so stable that a conventional centrifuge cannot cause the particles to precipitate. Before the addition of heptane removal of all DI water is required since heptane and DI water are not miscible. DI water was used as the starting solvent for functionalization since the pristine nanoparticles showed a considerably higher degree of dispersibility in water than they did in alcohols.

7. Appendix

a)

b)

C18 -PA

C18-PA van der Waals radius

c)

C18-PA solvent accessible area

d)

C18-PA hybrid

Figure 7.1.: a) Octadecyl phosphonic acid (C18-PA). b) Van der Waals model of C18-PA. c) Solvent accessible area of C18-PA as calculated by Chemdraw. d) Hybrid model employed in the 3D models thorough this work.

95

Appendix

T (%)

96

4000 3800 3600 3400 3200 3000 2800 2600 2400 2200 2000 1800 1600 1400 1200 1000

800

600

400

800

600

400

Wavenumber (1/cm)

T (%)

Figure 7.2.: FTIR-ATR spectra of several phosphonic acid molecules.

4000 3800 3600 3400 3200 3000 2800 2600 2400 2200 2000 1800 1600 1400 1200 1000

Wavenumber (1/cm)

Figure 7.3.: Continuation of Figure 7.2. FTIR-ATR spectra of several phosphonic acid molecules.

Appendix

97

Figure 7.4.: (Top) 10x10 µm AFM image of a spray coated nanoparticle film. (Bottom) 3D render of the same AFM image. The surface roughness of the measured film was 100 nm (RMS).

https://www.youtube.com/watch?v=wotvGJFQFLw Figure 7.5.: Hyperlink to video of a SCA measurement of a super hydrophobic film.

https://www.youtube.com/watch?v=N_OKvDxOWtQ Figure 7.6.: Hyperlink to video of a super hydrophobic coating on a piece of cardboard.

98

Appendix

Fe3O4 (CH3(OC2H4)3-PA

Fe3O4 Pristine

T (%)

T (%)

Fe3O4 Pristine

3800 3600 3400 3200 3000 2800 2600

Fe3O4 (CH3(OC2H4)3-PA

2000

1800

Wavenumber (1/cm)

T (%)

T (%)

2000

1800

T (%)

T (%)

1400

1200

1000

800

1000

800

1000

800

AlOx (CH3(OC2H4)3-PA

2000

1800

Wavenumber (1/cm)

1600

1400

1200

Wavenumber (1/cm)

CeO2 Pristine

CeO2 Pristine

T (%)

T (%)

1600

AlOx Pristine

3800 3600 3400 3200 3000 2800 2600

Wavenumber (1/cm)

800

Wavenumber (1/cm)

AlOx Pristine

3800 3600 3400 3200 3000 2800 2600

1000

TiO2 (CH3(OC2H4)3-PA

Wavenumber (1/cm)

CeO2 (CH3(OC2H4)3-PA

1200

TiO2 Pristine

3800 3600 3400 3200 3000 2800 2600

AlOx (CH3(OC2H4)3-PA

1400

Wavenumber (1/cm)

TiO2 Pristine

TiO2 (CH3(OC2H4)3-PA

1600

CeO2 (CH3(OC2H4)3-PA

2000

1800

1600

1400

1200

Wavenumber (1/cm)

Figure 7.7.: FTIR-ATR spectra of pristine and CH3(OC2H4)3-PA functionalized nanoparticles. The FTIR-ATR spectrum of the ITO nanoparticles was not possible to obtain due to the lack of transparency of ITO in the infrared region.

Appendix

99

T (%)

Azide-PA AlOx

Alkyne-PA AlOx

3800

3600

3400

3200

3000

2800

2600 2200 2000 1800 1600 1400 1200 1000

800

Wavenumber (1/cm)

Wavenumber (1/cm)

Figure 7.8.: FTIR-ATR spectra (color coded) of Azide-PA and Alkyne-PA AlOx reactive nanoparticles.

T (%)

a) Fe-Fe Pristine

Fe-Fe Pristine

Fe-Fe Imidazolium-PA

Fe-Fe Imidazolium-PA

Fe-Fe H(OC2H4)3-PA

Fe-Fe H(OC2H4)3-PA

Fe-Fe CH3(OC2H4)3-PA

Fe-Fe CH3(OC2H4)3-PA

4000 3800 3600 3400 3200 3000 2800 2600

2000

Wavenumber (1/cm)

1800

1600

1400

1200

1000

800

Wavenumber (1/cm)

T (%)

b) Fe-Co Pristine

Fe-Co Pristine

Fe-Co Imidazolium-PA

Fe-Co Imidazolium-PA

Fe-Co H(OC2H4)3-PA

Fe-Co H(OC2H4)3-PA

Fe-Co CH3(OC2H4)3-PA

Fe-Co CH3(OC2H4)3-PA

4000 3800 3600 3400 3200 3000 2800 2600

Wavenumber (1/cm)

2000

1800

1600

1400

1200

1000

800

Wavenumber (1/cm)

Figure 7.9.: FTIR-ATR spectra (color coded) of pristine and functionalized Fe 3O4 nanoparticles. a) Spectra of Fe-Fe nanoparticles. b) Spectra of Fe-Co nanoparticles.

100

Appendix

a) Breast cancer cells


150

Uncoated Pristine Feferrite 3O4 NP NP 1Imidazol-ferrite Fe3O4 2Hydroxy-ferrite Fe3O4 NP 3Methoxy-ferrite Fe3O4 NP

100

50

0 1

5

10


b) Healthy cells Uncoated Pristine Feferrite 3O4 NP NP 1Imidazol-ferrite Fe3O4 2Hydroxy-ferrite Fe3O4 NP 3Methoxy-ferrite Fe3O4 NP


100

50

0 1

5

10


Figure 7.10.: Biocompatibility of pristine and functionalized Fe3O4 nanoparticles in function of dosage. a) Biocompatibility results on breast cancer cells. b) Biocompatibility results on healthy cells.

8. List of Figures

Figure 2.1.: Representation of nanostructures. 0D (red sphere), 1D (green cylinder), 2D (blue rectangular parallelepiped) and in 3D (cylinder 3D matrix). ............................................. 3 Figure 2.2.: NSs classified on basis of their dimensionality as suggested by V. Pokropivny in 2007 [21]. Reprinted from [21] with permission from Elsevier.................................................. 4 Figure 2.3.: Surface area to volume ratio. The increment of surface area vs. volume is graphically represented from a) lowest ratio, to d) highest ratio........................................................... 5 Figure 2.4.: a) Schematic process of the formation of a C18-PA SAM on a 2D NS. b) Side view depiction of SAM tilt angle. c) Top view illustration of the crystalline and amorphous domains that may be present after the SAM formation...................................................... 8 Figure 2.5.: Schematic process of SAM formation of onto a 0D NS. ................................................... 8 Figure 2.6.: Anatomy of a core-shell 0D building block. ...................................................................... 9 Figure 2.7.: Phosphonic acid SAM molecules of increasing length. As the backbone (black) length increases, so do the van der Waals interactions between them and therefore also the final ordering of the SAM [26]. ................................................................................................ 10 Figure 2.8.: Phosphonic acid binding mechanism to a metal oxide surface [39]–[43]. ....................... 12 Figure 2.9.: Various phosphonic acid binding modes. M = metal. a) and b) monodentate, c) and d) bridging bidentate, e) bridging tridentate, f) and g) chelating bidentate, h) chelating tridentate, i) j) k) l) other viable hydrogen bonding modes. Reprinted from [41] with permission from the American Chemical Society. ........................................................... 13 Figure 2.10.: Various carboxylic acid binding modes) M = metal. a) electrostatic attraction, b) Hbonds to bridging oxygen, c) H-bonds to carboxylic oxygen, d) monodentate metal-ester, e) bidentate bridging, f) bidentate chelating. Adapted from [12] with permission from Wiley. ............................................................................................................................... 14 Figure 2.11.: Miscellaneous catechol binding modes on titanium oxide surface. EWG = electron withdrawing group. a) H-bonds, b) monodentate with H-bond, c) bidentate chelating, d) monodentate with bridging H-bond, e) bidentate bridging. Adapted from [12] with permission from Wiley. .................................................................................................... 15 Figure 2.12.: Photographs of TiO2 and Fe3O4 nanoparticle dispersions with varying degrees of stability. a) Unstable TiO2 dispersion during flocculation. b) Non-transparent stable TiO2 dispersion. c) Transparent stable TiO2 dispersion. d) Unstable, already flocculated Fe3O4 dispersion. e) Transparent stable Fe3O4 dispersion. ......................................................... 16 101

102

List of Figures

Figure 2.13.: Schematic representation of electrostatic stabilization of nanoparticles. a) A nanoparticle with charged species at its surface. b) Nanoparticles having an equal charge repel each other avoiding agglomeration. ........................................................................ 17 Figure 2.14.: Schematic representation of the electric double layer (EDL) on a nanoparticle. Red and blue spheres represent charged species of opposite magnitude. ...................................... 18 Figure 2.15.: Schematic representation of the isoelectric point (IEP) in function of pH. .................... 19 Figure 2.16.: Schematic representation of steric stabilization of nanoparticles. Physically, the molecules grafted onto the nanoparticle avoid direct nanoparticle collision and nanoparticle core interaction. Chemically, the molecules provide solvation in the dispersion media effectively thwarting nanoparticle core interactions. ........................... 20 Figure 2.17.: Nanoparticle surface curvatures. a) and c) are nanoparticle illustrations (up to scale) of functionalized 5 and 50 nm spherical particles. b) and d) represent a zoomed-in up to scale nanoparticle surface illustration. It becomes apparent in illustrations b) and d) how can surface curvature play a critical role on SAM crystallinity and SAM dipole moment alignment. Also of importance to note, the free space available between the molecules. 22 Figure 2.18.: Schematic representation of a bottom gate TFT ............................................................ 24 Figure 3.1.: Steps for SAM deposition on 2D substrates. a) Plasma treatment of the surface. b) immersion of the substrate into the SAM solution. c) Schematic of the finalized SAM. 25 Figure 3.2.: Steps for SAM deposition on 0D and 1D NSs. a) Pristine nanostructures dispersed in a liquid medium. b) The SAM molecule is added and with the aid of sonication it forms a SAM around the nanostructures surface. c) The final functionalized NSs dispersion (after washing) with no unbound SAM molecules present in solution. ..................................... 27 Figure 3.3.: FTIR-ATR spectra of several commercial TiO2 nanoparticles. a) Featureless spectrum of pure TiO2 particles. b) Spectrum of HNO3 stabilized TiO2 particles with signals from the HNO3 species present c) Spectrum of allegedly pure TiO2 particles containing signals that are unaccounted for. ......................................................................................................... 28 Figure 3.4.: Color coded FTIR-ATR spectra of several commercial TiO2 nanoparticles after being functionalized with C16-PA. a) Spectrum of pure TiO2 particles plus the signals from C16PA. b) Spectrum of HNO3 stabilized TiO2 particles with the signals from C16-PA. Note that the HNO3 peaks are no longer present. c) Spectrum of allegedly pure TiO2 particles functionalized with C16-PA still containing signals that are unaccounted for, plus the overlapping signals of the C16-PA. d) Chemical structure of C16-PA. ............................. 29 Figure 3.5.: Exemplary FTIR-ATR spectra (color coded) of various pure and C16-PA functionalized metal oxide NSs. The general trend of a featureless spectrum for pure metal oxides can be observed. An unmistakable trend is also identifiable after functionalization with C16PA in all metal oxides. The molecular structure of C16-PA is shown atop. ..................... 30

List of Figures

103

Figure 3.6.: a) FTIR-ATR spectrum of Fe3O4 nanoparticles functionalized with C16-PA. b) Spectrum of pristine C16-PA. c) Hypothetical schematic depicting the diverse vibrations of the phosphonic acid bound to a metal oxide surface.............................................................. 31 Figure 3.7.: Fe3O4 nanoparticles functionalized with increasing concentration of C16-PA. As the concentration increases, the valley corresponding to the anchored phosphonic acid (red square) changes. ............................................................................................................... 32 Figure 3.8.: a) Chemical structure of C16-PA. b) TGA under N2 of AlOx NPs functionalized with different concentrations of C16-PA. Adapted from [16] with permission from the American Chemical Society. ............................................................................................ 34 Figure 3.9.: TGA under N2 from 25 to 1100 °C of AlOx NPs functionalized with C16-PA. The inset shows FTIR-ATR spectra of the nanoparticle powder at different stages of the TGA measurement. It can be observed that even after exposing the NP powder to 1100 °C for 2 hours the PA band is still present on the nanopowder. The nanopowder had a black color up to a 1000 °C, past that temperature the powder was white in appearance. ........ 36 Figure 3.10.: FTIR-ATR spectra (color coded) of pristine and several C16-PA functionalized AlOx nanoparticles. a) Spectrum of pristine AlOx nanoparticles. b) C16-PA functionalized AlOx nanoparticles. c) C16-PA functionalized AlOx nanoparticles after exposure to TGA until 1100 °C. e) The exposed AlOx nanoparticles to TGA until 1100 °C but re-functionalized with C16-PA. e) The C16-PA re-functionalized AlOx nanoparticles after being measured again by TGA until 1100 °C. ........................................................................................... 38 Figure 3.11.: a) Molecular structure of F17C10-PA and H(OC2H4)3-PA. b) SCA measurements with different liquids and the calculated surface energy of the nanoparticle spray coated films. Reprinted from [16] with permission from the American Chemical Society. ................. 40 Figure 3.12.: a) Color coded molecular structure of F17C10-PA and H(OC2H4)3-PA. b) FTIR-ATR spectra of particles functionalized with different ratios of F17C10-PA and H(OC2H4)3-PA. c) Zeta potential measurements of the nanoparticle dispersions in 2-Propanol. Adapted from [16] with permission from the American Chemical Society. .................................. 41 Figure 3.13.: a) Zeta potential measurements of the nanoparticle dispersions in 2-Propanol. b) Photograph of the functionalized nanoparticle dispersions 20 min after re-dispersion by sonication. Adapted from [16] with permission from the American Chemical Society. . 42 Figure 3.14.: SCA measurements of functionalized AlOx, ZnO and ITO 2D substrates. The substrates were functionalized and afterwards immersed in pure 2-propanol for different amount of time. Contact angles were measured again after the immersion. a) SCA of substrates functionalized with C17-CA. b) SCA of substrates functionalized with F21-CAT. c) SCA of substrates functionalized with C16-PA. ........................................................................ 44 Figure 3.15.: SCA measurements of spray coated films of functionalized 0D AlOx and ITO and 1D ZnO NSs. The NSs were washed 1 to 3 times before spray coating. a) Contact angles of

104

List of Figures spray coated NSs functionalized with C17-CA. b) Contact angles of spray coated NSs functionalized with C16-PA. ............................................................................................. 46

Figure 3.16.: SCA measurements of SAM exchange between two carboxylic acid SAM molecules on AlOx and ITO 2D substrates. a) Exchange of a C17-CA SAM with a C6-CA SAM. b) Exchange of a C6-CA SAM with a C17-CA SAM. ........................................................... 47 Figure 3.17.: SCA measurements of SAM exchange between two phosphonic acid SAM molecules on AlOx, ITO and ZnO 2D substrates. Attempt to exchange a C11OH-PA SAM with a C16-PA SAM. ................................................................................................................... 48 Figure 3.18.: SCA measurements of SAM exchange between a phosphonic acid and catechol molecules on AlOx, ITO and ZnO 2D substrates. a) Exchange of a C11OH-PA SAM with a F21-CAT SAM. b) Exchange of a F21-CAT SAM with a C11OH-PA SAM................... 49 Figure 3.19.: SCA measurements of SAM exchange between a carboxylic acid and catechol molecules on ZnO. a) Exchange of a C17-CA SAM with a hexanoate-CAT SAM. b) Exchange of a hexanoate-CAT SAM with a C17-CA SAM. ............................................ 50 Figure 4.1.: a) Molecular structure of CH3(OC2H4)3-PA employed for nanoparticle functionalization; b) Photograph of dispersed TiO2 nanoparticles in different media before and c) after functionalization. d) Graph of DLS measurements of TiO2 and e) Fe3O4 nanoparticles before and after functionalization. Reprinted from [17] with permission from Wiley. ... 54 Figure 4.2.: Photograph of Fe3O4 nanoparticles functionalized with different molecules (top) and dispersed in different orthogonal media (left). a) Particles dispersed in n-heptane, b) particles dispersed in DI water, c) particles dispersed in n-perfluoroheptane. By tuning the surface of the particles, control on their dispersibility in different media can be achieved. .......................................................................................................................... 56 Figure 4.3.: Shell by shell stabilization concept. a) Nanoparticle is rendered hydrophobic by functionalization with C16-PA. b) The hydrophobic particle is rendered hydrophilic due to the amphiphilic molecules forming a bilayer. c) Example of the amphiphilic molecules employed for this study. Adapted from [18] with permission from Wiley. ..................... 57 Figure 4.4.: Photograph of TiO2 nanoparticles dispersed in DI-water or toluene. a) Hydrophobic and hydrophilic components of molecule 7. The red coloring of the TiO2 nanoparticles is due to the perylene motif. b) Pristine hydrophilic particles dispersed in DI-water. c) C16-PA functionalized hydrophobic nanoparticles dispersed in toluene. d) Upon addition of the amphiphile molecule 7 the nanoparticles are now dispersible in the water phase due to the formation of a bilayer. Adapted from [18] with permission from Wiley. .................. 58 Figure 4.5.: Polymer wrapping of nanoparticles. a) Phosphonic acid molecule with hydrophilic tail employed to functionalize the surface of the particles. b) PS-b-PEO block copolymer molecular structure and weights. c) Schematic representation of a nanoparticle wrapped in an orthogonal block copolymer. ................................................................................... 60

List of Figures

105

Figure 4.6.: Photograph of the solution before and after shaking vigorously. Photograph b) shows the solution only after shaking, not after sonication. ............................................................. 61 Figure 4.7.: DLS size distribution of CH3(OC2H4)4C4H8-PA nanoparticles before (black) and after polymer wrapping. ........................................................................................................... 63 Figure 4.8.: a) SEM cross-sections of the spin coated dielectric layers. b) AFM images of the surface morphology of the spin coated films. Reprinted from [17] with permission from Wiley.65 Figure 4.9.: a) Schematic layout of the fabricated capacitor devices. b) Current density of different NP dielectric materials of 50x50 µm capacitor devices vs. applied voltage. Reprinted from [17] with permission from Wiley. ........................................................................... 66 Figure 4.10.: a) Molecular structure of the semiconductor molecule C13-BTBT and schematic layout of the fabricated OTFTs devices. b) Transfer curves of the OTFTs with different dielectric materials. c) concave bending of devices during characterization. d) Transfer characteristics of the devices under different bending modes. Reprinted from [17] with permission from Wiley. .................................................................................................... 67 Figure 4.11.: Photographs of spray coated hydrophobic coatings on glass and cardboard. a) Coated and uncoated glass slide for up close transparency comparison. Droplets of different water based liquids dispensed on the top of coated (c) glass slide and (d) cardboard. .... 69 Figure 4.12.: Schematic representation of the building blocks involved in thin film self-assembly. a) Chemically patterned 2D substrate. Red represents the reactive sites provided by the Azide-PA SAM molecule. Purple represents the inert sections of the substrate. the inertness is provided by the F15C18-PA SAM molecule. b) Reactive Alkyne-PA functionalized nanoparticle. c) Reactive Azide-PA functionalized particle. ................... 71 Figure 4.13.: a) Schematic representation of the first deposited layer and reaction conditions. b) Schematic representation of the second deposited layer and reaction conditions. c) AFM scans of first and second selectively deposited thin films. AFM cross-sectional height measurement of the first and second layers. .................................................................... 73 Figure 4.14.: SEM images of selectively deposited nanoparticles. ..................................................... 74 Figure 4.15.: a) Molecular structure of block copolymers. The hydrophobic phase is composed of polystyrene (red) while the hydrophilic phase is composed of polyethylene oxide (blue). b) AFM image of a phase separated spin coated block copolymer thin film. c) SEM image of a phase separated spin coated block copolymer thin film with embedded Fe 3O4 hydrophilic nanoparticles in the corresponding phase. .................................................... 75 Figure 4.16.: a) Molecular structure of MMA-PA. b) Molecular structure of MMA. c) Molecular structure of PMMA. d) Color coded FTIR ATR spectrum of AlOx nanoparticles functionalized with MMA-PA. e) Cylindrical PMMA probe. f) Cylindrical PMMA probe with unfunctionalized AlOx nanoparticles. g) Cylindrical PMMA probe with MMA-PA functionalized AlOx nanoparticles. h) Strain test curves of PMMA and PMMA

106

List of Figures composites. i) Artistical rendition of nanoparticles covalently attached to a polymer matrix. .............................................................................................................................. 77

Figure 4.17.: a) Molecules used for functionalization of nanoparticles. b) DLS distribution of Fe 3O4 pristine and functionalized nanoparticles. c) Photograph of Fe3O4 pristine and functionalized NP dispersions 12 hours after being redispersed on DI water via sonication. d) Photograph of CoFe3O4 pristine and functionalized NP dispersions 12 hours after being redispersed on DI water via sonication. ............................................... 80 Figure 4.18.: Biocompatibility of pristine and functionalized CoFe3O4 nanoparticles in function of dosage. a) Biocompatibility results on breast cancer cells. b) Biocompatibility results on healthy cells...................................................................................................................... 81 Figure 4.19.: a) Fenton reaction of iron and cobalt. b) Comparison of oxygen radical generation before and after radiation in cancer cells exposed to pristine and functionalized CoFe3O4 nanoparticles. c) Comparison of oxygen radical generation before and after radiation in healthy cells exposed to pristine and functionalized CoFe3O4 nanoparticles. .................. 82 Figure 4.20.: Pollutant extraction scheme from a liquid via functionalized magnetic nanoparticles. . 84 Figure 4.21.: Extracted hydrocarbon weight vs. nanoparticle mass. ................................................... 84 Figure 7.1.: a) Octadecyl phosphonic acid (C18-PA). b) Van der Waals model of C18-PA. c) Solvent accessible area of C18-PA as calculated by Chemdraw. d) Hybrid model employed in the 3D models thorough this work. ........................................................................................ 95 Figure 7.2.: FTIR-ATR spectra of several phosphonic acid molecules. .............................................. 96 Figure 7.3.: Continuation of Figure 7.2. FTIR-ATR spectra of several phosphonic acid molecules. . 96 Figure 7.4.: (Top) 10x10 µm AFM image of a spray coated nanoparticle film. (Bottom) 3D render of the same AFM image. The surface roughness of the measured film was 100 nm (RMS). .......................................................................................................................................... 97 Figure 7.5.: Hyperlink to video of a SCA measurement of a super hydrophobic film. ....................... 97 Figure 7.6.: Hyperlink to video of a super hydrophobic coating on a piece of cardboard. .................. 97 Figure 7.7.: FTIR-ATR spectra of pristine and CH3(OC2H4)3-PA functionalized nanoparticles. The FTIR-ATR spectrum of the ITO nanoparticles was not possible to obtain due to the lack of transparency of ITO in the infrared region. ................................................................. 98 Figure 7.8.: FTIR-ATR spectra (color coded) of Azide-PA and Alkyne-PA AlOx reactive nanoparticles. ................................................................................................................... 99 Figure 7.9.: FTIR-ATR spectra (color coded) of pristine and functionalized Fe3O4 nanoparticles. a) Spectra of Fe-Fe nanoparticles. b) Spectra of Fe-Co nanoparticles. ................................ 99 Figure 7.10.: Biocompatibility of pristine and functionalized Fe3O4 nanoparticles in function of dosage. a) Biocompatibility results on breast cancer cells. b) Biocompatibility results on healthy cells.................................................................................................................... 100

List of tables

107

9. List of tables

Table 3.1.: Calculated grafting densities of AlOx NPs functionalized with C16-PA, in accordance with the mass loss from 400 to 650 °C of Figure 3.8 The SSA was calculated from a DLS distribution to be 28.05 m2/g. ........................................................................................... 35 Table 4.1.: Zeta potential of Fe3O4 nanoparticles. The hydrophilic CH3(OC2H4)4C4H8-PA nanoparticles have a strong zeta potential due to the electrosteric stabilization caused by the functionalization. After the polymer wrapping procedure the nanoparticles are still stable, yet their zeta potential is near zero. This effect is caused by the steric stabilization provided by the polymer wrapping of the nanoparticles. ................................................. 63 Table 4.2.: Summary of film properties and electrical characteristics of the films and devices. Reprinted from [17] with permission from Wiley............................................................ 65 Table 4.3.: Zeta potentials of the Fe-Fe and the Fe-Co nanoparticle dispersions before and after functionalization with different molecules which portrayed in Figure 4.17. ................... 80

10. Abbreviations

AFM BET CA CAT DI water DLS EDL FTIR-ATR GC-MS MMA MW NP NS NSM OTFT PA PBS PEN PMMA RMS rotavap SAM SAMFET SCA SSA TFT TGA XRR

109

atomic force microscopy Brunauer–Emmett–Teller carboxylic acid catechol de-ionized water dynamic light scattering electric double layer Fourier transform infrared spectroscopy - attenuated total reflectance gas chromatography and mass spectroscopy methyl methacrylate molecular weight nanoparticle nanostructure nanostructured material organic thin film transistor phosphonic acid Phosphate-buffered saline polyethylene naphthalene poly(methyl methacrylate) root mean square rotary evaporator self-assembled monolayer self-assembled monolayer field effect transistor static contact angle specific surface area thin film transistor thermogravimetric analysis x-ray reflectivity

11. Bibliography

[1]

R. Feynman, “There’s Plenty of Room at the Bottom,” Eng. Sci., vol. 23, no. 5, pp. 22–36, 1960.

[2]

P. Moriarty, “Nanostructured materials,” Reports Prog. Phys., vol. 64, no. 3, pp. 297–381, Mar. 2001.

[3]

H. Gleiter, “Nanostructured materials: basic concepts and microstructure,” Acta Mater., vol. 48, no. 1, pp. 1–29, Jan. 2000.

[4]

J. N. Tiwari, R. N. Tiwari, and K. S. Kim, “Zero-dimensional, one-dimensional, twodimensional and three-dimensional nanostructured materials for advanced electrochemical energy devices,” Prog. Mater. Sci., vol. 57, no. 4, pp. 724–803, May 2012.

[5]

Z. Ren, Y. Guo, C.-H. Liu, and P.-X. Gao, “Hierarchically nanostructured materials for sustainable environmental applications.,” Front. Chem., vol. 1, no. November, p. 18, 2013.

[6]

D. Kokkinis, M. Schaffner, and A. R. Studart, “Multimaterial magnetically assisted 3D printing of composite materials,” Nat. Commun., vol. 6, p. 8643, Oct. 2015.

[7]

G. A. Hughes, “Nanostructure-mediated drug delivery,” Nanomedicine Nanotechnology, Biol. Med., vol. 1, no. 1, pp. 22–30, Mar. 2005.

[8]

M. E. Davis, Z. G. Chen, and D. M. Shin, “Nanoparticle therapeutics: an emerging treatment modality for cancer.,” Nat. Rev. Drug Discov., vol. 7, no. 9, pp. 771–782, Sep. 2008.

[9]

O. C. Farokhzad and R. Langer, “Impact of nanotechnology on drug delivery,” ACS Nano, vol. 3, no. 1, pp. 16–20, Jan. 2009.

[10]

Y. Kim and R. J. Macfarlane, “Transmutable nanoparticles with reconfigurable surface ligands,” Science (80-. )., vol. 351, no. 6273, pp. 579–582, Feb. 2016.

[11]

S. H. Etschel, L. Portilla, J. Kirschner, M. Drost, F. Tu, H. Marbach, R. R. Tykwinski, and M. Halik, “Region-Selective Deposition of Core-Shell Nanoparticles for 3D Hierarchical Assemblies by the Huisgen 1,3-Dipolar Cycloaddition,” Angew. Chemie - Int. Ed., vol. 54, no. 32, pp. 9235–9238, Aug. 2015.

[12]

S. P. Pujari, L. Scheres, A. T. M. Marcelis, and H. Zuilhof, “Covalent surface modification of oxide surfaces,” Angew. Chemie - Int. Ed., vol. 53, no. 25, pp. 6322–6356, May 2014.

[13]

J. C. Love, L. A. Estroff, J. K. Kriebel, R. G. Nuzzo, and G. M. Whitesides, “Self-Assembled Monolayers of Thiolates on Metals as a Form of Nanotechnology,” Chem. Rev., vol. 105, no. 4, pp. 1103–1170, Apr. 2005.

[14]

A. Ulman, “Formation and Structure of Self-Assembled Monolayers,” Chem. Rev., vol. 96, no. 4, pp. 1533–1554, Jan. 1996.

[15]

M. Halik and A. Hirsch, “The Potential of Molecular Self-Assembled Monolayers in Organic

111

112

Bibliography Electronic Devices,” Adv. Mater., vol. 23, no. 22–23, pp. 2689–2695, Jun. 2011.

[16]

L. Portilla and M. Halik, “Smoothly Tunable Surface Properties of Aluminum Oxide Core– Shell Nanoparticles By A Mixed-Ligand Approach,” ACS Appl. Mater. Interfaces, vol. 6, no. 8, pp. 5977–5982, Apr. 2014.

[17]

L. Portilla, S. H. Etschel, R. R. Tykwinski, and M. Halik, “Green Processing of Metal Oxide Core-Shell Nanoparticles as Low-Temperature Dielectrics in Organic Thin-Film Transistors,” Adv. Mater., vol. 27, no. 39, pp. 5950–5954, Oct. 2015.

[18]

L. Zeininger, S. Petzi, J. Schönamsgruber, L. Portilla, M. Halik, and A. Hirsch, “Very Facile Polarity Umpolung and Noncovalent Functionalization of Inorganic Nanoparticles: A Tool Kit for Supramolecular Materials Chemistry,” Chem. - A Eur. J., vol. 21, no. 40, pp. 14030–14035, Sep. 2015.

[19]

L. Zeininger, L. Portilla, M. Halik, and A. Hirsch, “Quantitative Determination and Comparison of the Surface Binding of Phosphonic Acid, Carboxylic Acid, and Catechol Ligands on TiO 2 Nanoparticles,” Chem. - A Eur. J., Jul. 2016.

[20]

C. Chen, Y. Kang, Z. Huo, Z. Zhu, W. Huang, H. L. Xin, J. D. Snyder, D. Li, J. A. Herron, M. Mavrikakis, M. Chi, K. L. More, Y. Li, N. M. Markovic, G. A. Somorjai, P. Yang, and V. R. Stamenkovic, “Highly Crystalline Multimetallic Nanoframes with Three-Dimensional Electrocatalytic Surfaces,” Science (80-. )., vol. 343, no. 6177, pp. 1339–1343, Mar. 2014.

[21]

V. V. Pokropivny and V. V. Skorokhod, “Classification of nanostructures by dimensionality and concept of surface forms engineering in nanomaterial science,” Mater. Sci. Eng. C, vol. 27, no. 5–8 SPEC. ISS., pp. 990–993, Sep. 2007.

[22]

S. Brunauer, P. H. Emmett, and E. Teller, “Adsorption of Gases in Multimolecular Layers,” J. Am. Chem. Soc., vol. 60, no. 2, pp. 309–319, Feb. 1938.

[23]

S. A. DiBenedetto, A. Facchetti, M. A. Ratner, and T. J. Marks, “Molecular self-assembled monolayers and multilayers for organic and unconventional inorganic thin-film transistor applications,” Adv. Mater., vol. 21, no. 14–15, pp. 1407–1433, Apr. 2009.

[24]

G. M. Whitesides, “Self-Assembly at All Scales,” Science (80-. )., vol. 295, no. 5564, pp. 2418–2421, Mar. 2002.

[25]

L. E. Depero and M. L. Curri, “Inorganic self-assembly,” Curr. Opin. Solid State Mater. Sci., vol. 8, no. 2, pp. 103–109, Mar. 2004.

[26]

B. O. Acton, G. G. Ting, P. J. Shamberger, F. S. Ohuchi, H. Ma, and A. K.-Y. Jen, “Dielectric Surface-Controlled Low-Voltage Organic Transistors via n -Alkyl Phosphonic Acid SelfAssembled Monolayers on High- k Metal Oxide,” ACS Appl. Mater. Interfaces, vol. 2, no. 2, pp. 511–520, Feb. 2010.

[27]

J.-Y. Kim and N. A. Kotov, “Charge Transport Dilemma of Solution-Processed Nanomaterials,” Chem. Mater., vol. 26, no. 1, pp. 134–152, Jan. 2014.

[28]

J. Hirschmann, H. Faber, and M. Halik, “Concept of a thin film memory transistor based on ZnO nanoparticles insulated by a ligand shell,” Nanoscale, vol. 4, no. 2, pp. 444–447, 2012.

[29]

H. Ma, O. Acton, D. O. Hutchins, N. Cernetic, and A. K.-Y. Jen, “Multifunctional phosphonic acid self-assembled monolayers on metal oxides as dielectrics, interface modification layers and semiconductors for low-voltage high-performance organic field-effect transistors,” Phys. Chem. Chem. Phys., vol. 14, no. 41, p. 14110, 2012.

Bibliography

113

[30]

T. Schmaltz, A. Y. Amin, A. Khassanov, T. Meyer-Friedrichsen, H.-G. Steinrück, A. Magerl, J. J. Segura, K. Voitchovsky, F. Stellacci, and M. Halik, “Low-Voltage Self-Assembled Monolayer Field-Effect Transistors on Flexible Substrates,” Adv. Mater., vol. 25, no. 32, pp. 4511–4514, Aug. 2013.

[31]

A. Khassanov, T. Schmaltz, H.-G. Steinrück, A. Magerl, A. Hirsch, and M. Halik, “Interface Engineering of Molecular Charge Storage Dielectric Layers for Organic Thin-Film Memory Transistors,” Adv. Mater. Interfaces, vol. 1, no. 9, p. n/a-n/a, Dec. 2014.

[32]

O. Zenasni, M. D. Marquez, A. C. Jamison, H. J. Lee, A. Czader, and T. R. Lee, “Inverted Surface Dipoles in Fluorinated Self-Assembled Monolayers,” Chem. Mater., vol. 27, no. 21, pp. 7433–7446, Nov. 2015.

[33]

A. Rumpel, M. Novak, J. Walter, B. Braunschweig, M. Halik, and W. Peukert, “Tuning the Molecular Order of C 60 Functionalized Phosphonic Acid Monolayers,” Langmuir, vol. 27, no. 24, pp. 15016–15023, Dec. 2011.

[34]

C. M. Jäger, T. Schmaltz, M. Novak, A. Khassanov, A. Vorobiev, M. Hennemann, A. Krause, H. Dietrich, D. Zahn, A. Hirsch, M. Halik, and T. Clark, “Improving the Charge Transport in Self-Assembled Monolayer Field-Effect Transistors: From Theory to Devices,” J. Am. Chem. Soc., vol. 135, no. 12, pp. 4893–4900, Mar. 2013.

[35]

Z. Wang, S. Mohammadzadeh, T. Schmaltz, J. Kirschner, A. Khassanov, S. Eigler, U. Mundloch, C. Backes, H.-G. Steinrück, A. Magerl, F. Hauke, A. Hirsch, and M. Halik, “Region-Selective Self-Assembly of Functionalized Carbon Allotropes from Solution,” ACS Nano, vol. 7, no. 12, pp. 11427–11434, Dec. 2013.

[36]

T. Lenz, T. Schmaltz, M. Novak, and M. Halik, “Self-Assembled Monolayer Exchange Reactions as a Tool for Channel Interface Engineering in Low-Voltage Organic Thin-Film Transistors,” Langmuir, vol. 28, no. 39, pp. 13900–13904, Oct. 2012.

[37]

G. D. Davis, J. S. Ahearn, L. J. Matienzo, and J. D. Venables, “Use of hydration inhibitors to improve bond durability of aluminium adhesive joints,” J. Mater. Sci., vol. 20, no. 3, pp. 975– 988, Mar. 1985.

[38]

R. D. Ramsier, P. N. Henriksen, and A. N. Gent, “Adsorption of phosphorus acids on alumina,” Surf. Sci., vol. 203, no. 1–2, pp. 72–88, Sep. 1988.

[39]

M. Textor, L. Ruiz, R. Hofer, A. Rossi, K. Feldman, G. Hähner, and N. D. Spencer, “Structural Chemistry of Self-Assembled Monolayers of Octadecylphosphoric Acid on Tantalum Oxide Surfaces,” Langmuir, vol. 16, no. 7, pp. 3257–3271, Apr. 2000.

[40]

G. Guerrero, P. H. Mutin, and A. Vioux, “Anchoring of Phosphonate and Phosphinate Coupling Molecules on Titania Particles,” Chem. Mater., vol. 13, no. 11, pp. 4367–4373, Nov. 2001.

[41]

P. J. Hotchkiss, S. C. Jones, S. A. Paniagua, A. Sharma, B. Kippelen, N. R. Armstrong, and S. R. Marder, “The Modification of Indium Tin Oxide with Phosphonic Acids: Mechanism of Binding, Tuning of Surface Properties, and Potential for Use in Organic Electronic Applications,” Acc. Chem. Res., vol. 45, no. 3, pp. 337–346, Mar. 2012.

[42]

T. Bauer, T. Schmaltz, T. Lenz, M. Halik, B. Meyer, and T. Clark, “Phosphonate- and Carboxylate-Based Self-Assembled Monolayers for Organic Devices: A Theoretical Study of Surface Binding on Aluminum Oxide with Experimental Support,” ACS Appl. Mater. Interfaces, vol. 5, no. 13, pp. 6073–6080, Jul. 2013.

[43]

S. A. Paniagua, A. J. Giordano, O. L. Smith, S. Barlow, H. Li, N. R. Armstrong, J. E.

114

Bibliography Pemberton, J.-L. Brédas, D. Ginger, and S. R. Marder, “Phosphonic Acids for Interfacial Engineering of Transparent Conductive Oxides,” Chem. Rev., p. acs.chemrev.6b00061, May 2016.

[44]

E. L. Hanson, J. Schwartz, B. Nickel, N. Koch, and M. F. Danisman, “Bonding SelfAssembled, Compact Organophosphonate Monolayers to the Native Oxide Surface of Silicon,” J. Am. Chem. Soc., vol. 125, no. 51, pp. 16074–16080, Dec. 2003.

[45]

D. L. Allara and R. G. Nuzzo, “Spontaneously organized molecular assemblies. 1. Formation, dynamics, and physical properties of n-alkanoic acids adsorbed from solution on an oxidized aluminum surface,” Langmuir, vol. 1, no. 1, pp. 45–52, Jan. 1985.

[46]

K. D. Dobson and A. J. McQuillan, “In situ infrared spectroscopic analysis of the adsorption of aromatic carboxylic acids to TiO2, ZrO2, Al2O3, and Ta2O5 from aqueous solutions,” Spectrochim. Acta Part A Mol. Biomol. Spectrosc., vol. 56, no. 3, pp. 557–565, Feb. 2000.

[47]

B. Kim, S. W. Park, J.-Y. Kim, K. Yoo, J. A. Lee, M.-W. Lee, D.-K. Lee, J. Y. Kim, B. Kim, H. Kim, S. Han, H. J. Son, and M. J. Ko, “Rapid Dye Adsorption via Surface Modification of TiO 2 Photoanodes for Dye-Sensitized Solar Cells,” ACS Appl. Mater. Interfaces, vol. 5, no. 11, pp. 5201–5207, Jun. 2013.

[48]

A. Raman and E. S. Gawalt, “Self-Assembled Monolayers of Alkanoic Acids on the Native Oxide Surface of SS316L by Solution Deposition,” Langmuir, vol. 23, no. 5, pp. 2284–2288, Feb. 2007.

[49]

S. Alexander, J. Eastoe, A. M. Lord, F. Guittard, and A. R. Barron, “Branched Hydrocarbon Low Surface Energy Materials for Superhydrophobic Nanoparticle Derived Surfaces,” ACS Appl. Mater. Interfaces, vol. 8, no. 1, pp. 660–666, Jan. 2016.

[50]

H. Lee, S. M. Dellatore, W. M. Miller, and P. B. Messersmith, “Mussel-Inspired Surface Chemistry for Multifunctional Coatings,” Science (80-. )., vol. 318, no. 5849, pp. 426–430, Oct. 2007.

[51]

J. Sedó, J. Saiz-Poseu, F. Busqué, and D. Ruiz-Molina, “Catechol-Based Biomimetic Functional Materials,” Adv. Mater., vol. 25, no. 5, pp. 653–701, Feb. 2013.

[52]

Q. Ye, F. Zhou, and W. Liu, “Bioinspired catecholic chemistry for surface modification,” Chem. Soc. Rev., vol. 40, no. 7, p. 4244, 2011.

[53]

Y. Wang and I. Zhitomirsky, “Bio-inspired catechol chemistry for electrophoretic nanotechnology of oxide films,” J. Colloid Interface Sci., vol. 380, no. 1, pp. 8–15, Aug. 2012.

[54]

M. Rodenstein, S. Zürcher, S. G. P. Tosatti, and N. D. Spencer, “Fabricating Chemical Gradients on Oxide Surfaces by Means of Fluorinated, Catechol-Based, Self-Assembled Monolayers †,” Langmuir, vol. 26, no. 21, pp. 16211–16220, Nov. 2010.

[55]

B. Malisova, S. Tosatti, M. Textor, K. Gademann, and S. Zürcher, “Poly(ethylene glycol) Adlayers Immobilized to Metal Oxide Substrates Through Catechol Derivatives: Influence of Assembly Conditions on Formation and Stability,” Langmuir, vol. 26, no. 6, pp. 4018–4026, Mar. 2010.

[56]

P. W. Voorhees, “The theory of Ostwald ripening,” J. Stat. Phys., vol. 38, no. 1–2, pp. 231– 252, Jan. 1985.

[57]

R. J. Hunter, R. H. Ottewill, and R. L. Rowell, Zeta Potential in Colloid Science. 1981.

[58]

G. A. Parks, “The Isoelectric Points of Solid Oxides, Solid Hydroxides, and Aqueous Hydroxo

Bibliography

115

Complex Systems,” Chem. Rev., vol. 65, no. 2, pp. 177–198, Apr. 1965. [59]

S. Kittaka and T. Morimoto, “Isoelectric point of metal oxides and binary metal oxides having spinel structure,” J. Colloid Interface Sci., vol. 75, no. 2, pp. 398–403, Jun. 1980.

[60]

M. Kosmulski, “Compilation of PZC and IEP of sparingly soluble metal oxides and hydroxides from literature,” Adv. Colloid Interface Sci., vol. 152, no. 1–2, pp. 14–25, Nov. 2009.

[61]

D. . Napper, “Steric stabilization,” J. Colloid Interface Sci., vol. 58, no. 2, pp. 390–407, Feb. 1977.

[62]

L. A. Lane, X. Qian, A. M. Smith, and S. Nie, “Physical Chemistry of Nanomedicine: Understanding the Complex Behaviors of Nanoparticles in Vivo,” Annu. Rev. Phys. Chem., vol. 66, no. 1, pp. 521–547, Apr. 2015.

[63]

A. Pettersson, G. Marino, A. Pursiheimo, and J. B. Rosenholm, “Electrosteric Stabilization of Al2O3, ZrO2, and 3Y–ZrO2 Suspensions: Effect of Dissociation and Type of Polyelectrolyte,” J. Colloid Interface Sci., vol. 228, no. 1, pp. 73–81, Aug. 2000.

[64]

D. Cahen, R. Naaman, and Z. Vager, “The Cooperative Molecular Field Effect,” Adv. Funct. Mater., vol. 15, no. 10, pp. 1571–1578, Oct. 2005.

[65]

H. Kim, R. P. Carney, J. Reguera, Q. K. Ong, X. Liu, and F. Stellacci, “Synthesis and Characterization of Janus Gold Nanoparticles,” Adv. Mater., vol. 24, no. 28, pp. 3857–3863, Jul. 2012.

[66]

Y. T. He, J. Wan, and T. Tokunaga, “Kinetic stability of hematite nanoparticles: the effect of particle sizes,” J. Nanoparticle Res., vol. 10, no. 2, pp. 321–332, Feb. 2008.

[67]

K. Suttiponparnit, J. Jiang, M. Sahu, S. Suvachittanont, T. Charinpanitkul, and P. Biswas, “Role of Surface Area, Primary Particle Size, and Crystal Phase on Titanium Dioxide Nanoparticle Dispersion Properties,” Nanoscale Res. Lett., Sep. 2010.

[68]

M. Salinas, C. M. Jäger, A. Y. Amin, P. O. Dral, T. Meyer-Friedrichsen, A. Hirsch, T. Clark, and M. Halik, “The Relationship between Threshold Voltage and Dipolar Character of SelfAssembled Monolayers in Organic Thin-Film Transistors,” J. Am. Chem. Soc., vol. 134, no. 30, pp. 12648–12652, Aug. 2012.

[69]

M. Aghamohammadi, R. Rödel, U. Zschieschang, C. Ocal, H. Boschker, R. T. Weitz, E. Barrena, and H. Klauk, “Threshold-Voltage Shifts in Organic Transistors Due to SelfAssembled Monolayers at the Dielectric: Evidence for Electronic Coupling and Dipolar Effects,” ACS Appl. Mater. Interfaces, vol. 7, no. 41, pp. 22775–22785, Oct. 2015.

[70]

H. Klauk, “Organic thin-film transistors,” Chem. Soc. Rev., vol. 39, no. 7, p. 2643, 2010.

[71]

S. Pawsey, K. Yach, and L. Reven, “Self-Assembly of Carboxyalkylphosphonic Acids on Metal Oxide Powders,” Langmuir, vol. 18, no. 13, pp. 5205–5212, Jun. 2002.

[72]

M. C. Zenobi, C. V. Luengo, M. J. Avena, and E. H. Rueda, “An ATR-FTIR study of different phosphonic acids adsorbed onto boehmite,” Spectrochim. Acta Part A Mol. Biomol. Spectrosc., vol. 75, no. 4, pp. 1283–1288, Apr. 2010.

[73]

N. Cinausero, N. Azema, M. Cochez, M. Ferriol, M. Essahli, F. Ganachaud, and J.-M. LopezCuesta, “Influence of the surface modification of alumina nanoparticles on the thermal stability and fire reaction of PMMA composites,” Polym. Adv. Technol., vol. 19, no. 6, pp. 701–709, Jun. 2008.

116

Bibliography

[74]

K. R. Williams, K. Gupta, and M. Wasilik, “Etch rates for micromachining processing-part II,” J. Microelectromechanical Syst., vol. 12, no. 6, pp. 761–778, Dec. 2003.

[75]

G. Cao, H. G. Hong, and T. E. Mallouk, “Layered metal phosphates and phosphonates: from crystals to monolayers,” Acc. Chem. Res., vol. 25, no. 9, pp. 420–427, Sep. 1992.

[76]

J. Zhao, Z. Jian, J. Ma, F. Wang, Y.-S. Hu, W. Chen, L. Chen, H. Liu, and S. Dai, “Monodisperse Iron Phosphate Nanospheres: Preparation and Application in Energy Storage,” ChemSusChem, vol. 5, no. 8, pp. 1495–1500, Aug. 2012.

[77]

J. Hirschmann, “Electrical Characteristics of Functionalized Zinc Oxide Nanoparticles,” Friedrich-Alexander-Universität Erlangen-Nürnberg (FAU), 2014.

[78]

L. Sang, A. Mudalige, A. K. Sigdel, A. J. Giordano, S. R. Marder, J. J. Berry, and J. E. Pemberton, “PM-IRRAS Determination of Molecular Orientation of Phosphonic Acid SelfAssembled Monolayers on Indium Zinc Oxide,” Langmuir, vol. 31, no. 20, pp. 5603–5613, May 2015.

[79]

M. Gliboff, L. Sang, K. M. Knesting, M. C. Schalnat, A. Mudalige, E. L. Ratcliff, H. Li, A. K. Sigdel, A. J. Giordano, J. J. Berry, D. Nordlund, G. T. Seidler, J.-L. Brédas, S. R. Marder, J. E. Pemberton, and D. S. Ginger, “Orientation of Phenylphosphonic Acid Self-Assembled Monolayers on a Transparent Conductive Oxide: A Combined NEXAFS, PM-IRRAS, and DFT Study,” Langmuir, vol. 29, no. 7, pp. 2166–2174, Feb. 2013.

[80]

H. Klauk, U. Zschieschang, J. Pflaum, and M. Halik, “Ultralow-power organic complementary circuits,” Nature, vol. 445, no. 7129, pp. 745–748, Feb. 2007.

[81]

A. Khassanov, H.-G. Steinrück, T. Schmaltz, A. Magerl, and M. Halik, “Structural Investigations of Self-Assembled Monolayers for Organic Electronics: Results from X-ray Reflectivity,” Acc. Chem. Res., vol. 48, no. 7, pp. 1901–1908, Jul. 2015.

[82]

U. Zschieschang, F. Ante, M. Schlörholz, M. Schmidt, K. Kern, and H. Klauk, “Mixed SelfAssembled Monolayer Gate Dielectrics for Continuous Threshold Voltage Control in Organic Transistors and Circuits,” Adv. Mater., vol. 22, no. 40, pp. 4489–4493, Oct. 2010.

[83]

A. Jedaa, M. Salinas, C. M. Jäger, T. Clark, A. Ebel, A. Hirsch, and M. Halik, “Mixed selfassembled monolayer of molecules with dipolar and acceptor character—Influence on hysteresis and threshold voltage in organic thin-film transistors,” Appl. Phys. Lett., vol. 100, no. 6, p. 63302, 2012.

[84]

P. A. Charpentier, K. Burgess, L. Wang, R. R. Chowdhury, A. F. Lotus, and G. Moula, “NanoTiO 2 /polyurethane composites for antibacterial and self-cleaning coatings,” Nanotechnology, vol. 23, no. 42, p. 425606, Oct. 2012.

[85]

S. A. Paniagua, P. J. Hotchkiss, S. C. Jones, S. R. Marder, A. Mudalige, F. S. Marrikar, J. E. Pemberton, and N. R. Armstrong, “Phosphonic Acid Modification of Indium−Tin Oxide Electrodes: Combined XPS/UPS/Contact Angle Studies †,” J. Phys. Chem. C, vol. 112, no. 21, pp. 7809–7817, May 2008.

[86]

X. Gao and J. Chorover, “Adsorption of perfluorooctanoic acid and perfluorooctanesulfonic acid to iron oxide surfaces as studied by flow-through ATR-FTIR spectroscopy,” Environ. Chem., vol. 9, no. 2, p. 148, 2012.

[87]

A. Lafuma and D. Quéré, “Superhydrophobic states,” Nat. Mater., vol. 2, no. 7, pp. 457–460, Jul. 2003.

[88]

Q. Dong, W. Yu, Z. Li, S. Yao, X. Zhang, B. Yang, C. Im, and W. Tian, “All-water-solution

Bibliography

117

processed solar cells based on PPV and TiO2 nanocrystals,” Sol. Energy Mater. Sol. Cells, vol. 104, pp. 75–80, Sep. 2012. [89]

S. H. Jin, S.-K. Kang, I.-T. Cho, S. Y. Han, H. U. Chung, D. J. Lee, J. Shin, G. W. Baek, T. Kim, J.-H. Lee, and J. A. Rogers, “Water-Soluble Thin Film Transistors and Circuits Based on Amorphous Indium–Gallium–Zinc Oxide,” ACS Appl. Mater. Interfaces, vol. 7, no. 15, pp. 8268–8274, Apr. 2015.

[90]

C. Boyer, M. R. Whittaker, V. Bulmus, J. Liu, and T. P. Davis, “The design and utility of polymer-stabilized iron-oxide nanoparticles for nanomedicine applications,” NPG Asia Mater., vol. 2, no. 1, pp. 23–30, Jan. 2010.

[91]

D. Boal, Mechanics of the Cell. Cambridge: Cambridge University Press, 2001.

[92]

M. P. Krafft, F. Giulieri, P. Fontaine, and M. Goldmann, “Reversible Stepwise Formation of Mono- and Bilayers of a Fluorocarbon/Hydrocarbon Diblock on Top of a Phospholipid Langmuir Monolayer. A Case of Vertical Phase Separation,” Langmuir, vol. 17, no. 21, pp. 6577–6584, Oct. 2001.

[93]

L. D. Zarzar, V. Sresht, E. M. Sletten, J. A. Kalow, D. Blankschtein, and T. M. Swager, “Dynamically reconfigurable complex emulsions via tunable interfacial tensions,” Nature, vol. 518, no. 7540, pp. 520–524, Feb. 2015.

[94]

A.-H. Lu, E. L. Salabas, and F. Schüth, “Magnetic Nanoparticles: Synthesis, Protection, Functionalization, and Application,” Angew. Chemie Int. Ed., vol. 46, no. 8, pp. 1222–1244, Feb. 2007.

[95]

B. A. Rozenberg and R. Tenne, “Polymer-assisted fabrication of nanoparticles and nanocomposites,” Prog. Polym. Sci., vol. 33, no. 1, pp. 40–112, Jan. 2008.

[96]

A. C. Balazs, T. Emrick, and T. P. Russell, “Nanoparticle Polymer Composites: Where Two Small Worlds Meet,” Science (80-. )., vol. 314, no. 5802, pp. 1107–1110, Nov. 2006.

[97]

“A basic guide to particle characterization.” http://www.malvern.com/en/support/resourcecenter/Whitepapers/WP120620BasicGuidePartChar.aspx.

[98]

Q. J. Cai, Y. Gan, M. B. Chan-Park, H. Bin Yang, Z. S. Lu, Q. L. Song, C. M. Li, and Z. Li Dong, “Solution-processable organic-capped titanium oxide nanoparticle dielectrics for organic thin-film transistors,” Appl. Phys. Lett., vol. 93, no. 11, p. 113304, 2008.

[99]

A. Y. Amin, A. Khassanov, K. Reuter, T. Meyer-Friedrichsen, and M. Halik, “Low-Voltage Organic Field Effect Transistors with a 2-Tridecyl[1]benzothieno[3,2- b ][1]benzothiophene Semiconductor Layer,” J. Am. Chem. Soc., vol. 134, no. 40, pp. 16548–16550, Oct. 2012.

[Online].

Available:

[100] M. Sarcletti, T. Schmaltz, and M. Halik, “The mutual influence of surface energy and substrate temperature on the saturation mobility in organic semiconductors,” Org. Electron., vol. 15, no. 11, pp. 3082–3086, Nov. 2014. [101] T. Sekitani, U. Zschieschang, H. Klauk, and T. Someya, “Flexible organic transistors and circuits with extreme bending stability,” Nat. Mater., vol. 9, no. 12, pp. 1015–1022, Dec. 2010. [102] R. P. Rowand, “Method of teflon coating of metals,” US 3243321 A, 1962. [103] P. Brown and B. Bhushan, “Durable superoleophobic polypropylene surfaces,” Philos. Trans. A, 2016.

118

Bibliography

[104] A. Pruna, J. Ramiro, and L. Belforte, “Preliminary study on different technological tools and polymeric materials towards superhydrophobic surfaces for automotive applications,” J. Phys. Chem. Solids, vol. 74, no. 11, pp. 1640–1645, Nov. 2013. [105] L. Cao, A. K. Jones, V. K. Sikka, J. Wu, and D. Gao, “Anti-Icing Superhydrophobic Coatings,” Langmuir, vol. 25, no. 21, pp. 12444–12448, Nov. 2009. [106] J. Genzer and K. Efimenko, “Recent developments in superhydrophobic surfaces and their relevance to marine fouling: a review,” Biofouling, vol. 22, no. 5, pp. 339–360, Jan. 2006. [107] V. V. Rostovtsev, L. G. Green, V. V. Fokin, and K. B. Sharpless, “A Stepwise Huisgen Cycloaddition Process: Copper(I)-Catalyzed Regioselective ‘Ligation’ of Azides and Terminal Alkynes,” Angew. Chemie Int. Ed., vol. 41, no. 14, pp. 2596–2599, Jul. 2002. [108] C. W. Tornøe, C. Christensen, and M. Meldal, “Peptidotriazoles on Solid Phase: [1,2,3]Triazoles by Regiospecific Copper(I)-Catalyzed 1,3-Dipolar Cycloadditions of Terminal Alkynes to Azides,” J. Org. Chem., vol. 67, no. 9, pp. 3057–3064, May 2002. [109] T. Lummerstorfer and H. Hoffmann, “Click Chemistry on Surfaces: 1,3-Dipolar Cycloaddition Reactions of Azide-Terminated Monolayers on Silica,” J. Phys. Chem. B, vol. 108, no. 13, pp. 3963–3966, Apr. 2004. [110] J. P. Collman, N. K. Devaraj, and C. E. D. Chidsey, “‘Clicking’ Functionality onto Electrode Surfaces,” Langmuir, vol. 20, no. 4, pp. 1051–1053, Feb. 2004. [111] Z. Guo, T. Pereira, O. Choi, Y. Wang, and H. T. Hahn, “Surface functionalized alumina nanoparticle filled polymeric nanocomposites with enhanced mechanical properties,” J. Mater. Chem., vol. 16, no. 27, p. 2800, 2006. [112] C. L. Wu, M. Q. Zhang, M. Z. Rong, and K. Friedrich, “Tensile performance improvement of low nanoparticles filled-polypropylene composites,” Compos. Sci. Technol., vol. 62, no. 10–11, pp. 1327–1340, Aug. 2002. [113] J.-C. Lin, L. C. Chang, M. H. Nien, and H. L. Ho, “Mechanical behavior of various nanoparticle filled composites at low-velocity impact,” Compos. Struct., vol. 74, no. 1, pp. 30– 36, Jul. 2006. [114] V. Pillai and D. O. Shah, “Synthesis of high-coercivity cobalt ferrite particles using water-inoil microemulsions,” J. Magn. Magn. Mater., vol. 163, no. 1–2, pp. 243–248, Oct. 1996. [115] D. Caruntu, G. Caruntu, and C. J. O’Connor, “Magnetic properties of variable-sized Fe 3 O 4 nanoparticles synthesized from non-aqueous homogeneous solutions of polyols,” J. Phys. D. Appl. Phys., vol. 40, no. 19, pp. 5801–5809, Oct. 2007. [116] L. Chen, J. M. Mccrate, J. C.-M. Lee, and H. Li, “The role of surface charge on the uptake and biocompatibility of hydroxyapatite nanoparticles with osteoblast cells,” Nanotechnology, vol. 22, no. 10, p. 105708, Mar. 2011. [117] E. Fröhlich, “The role of surface charge in cellular uptake and cytotoxicity of medical nanoparticles,” Int. J. Nanomedicine, p. 5577, Nov. 2012. [118] T. Osaka, T. Nakanishi, S. Shanmugam, S. Takahama, and H. Zhang, “Effect of surface charge of magnetite nanoparticles on their internalization into breast cancer and umbilical vein endothelial cells,” Colloids Surfaces B Biointerfaces, vol. 71, no. 2, pp. 325–330, Jul. 2009. [119] J. Imlay and S. Linn, “DNA damage and oxygen radical toxicity,” Science (80-. )., vol. 240, no. 4857, pp. 1302–1309, Jun. 1988.

Bibliography

119

[120] J. Imlay, S. Chin, and S. Linn, “Toxic DNA damage by hydrogen peroxide through the Fenton reaction in vivo and in vitro,” Science (80-. )., vol. 240, no. 4852, pp. 640–642, Apr. 1988. [121] B. Halliwell and M. Whiteman, “Measuring reactive species and oxidative damage in vivo and in cell culture: how should you do it and what do the results mean?,” Br. J. Pharmacol., vol. 142, no. 2, pp. 231–255, May 2004.

12. Acknowledgements

First I would like to extend my gratitude towards Prof. Marcus Halik for giving me the opportunity to be part of his research group. To all the members of the group as well (colleagues and friends) Atefeh Yousefi Amin, Michael Salinas, Johannes Hirschmann, Sebastien Pequeur, Saeideh Grünler, Zhenxing Wang, Thomas Schmaltz, Artoem Khassanov, Sebastian Etschel, Johannes Kirschner, Simon Scheiner, Bastian Gothe, Tobias Rejek, Hyoungwon Park and Judith Wittmann. I would also like to extend my gratitude to Lukas Zeininger from Prof. Hirsch's group for all the fruitful work we did together. It was very enjoyable to sit down over coffee or beer and having useful, useless or plain crazy discussions with you all. Many good and also bad ideas came out of these conversations, yet they were all very constructive for the basis of my Ph. D. work. I would like to specially thank you all for having the patience to answer and correct all of my misguided ideas and irrational thinking regarding chemistry. I hope I was able to contribute to your work as well. I would like to thank as well Melek Kizaloglu and Stefanie Klein from Prof. Kryschi's group for all the interesting research regarding iron oxide nanoparticles for bio-applications. It was a very nice project and I was very happy to make this collaboration. I also would like to thank all the people from the LSP (Lehrstuhl für Polymerwerkstoffe), in particular Jenifer Reiser for some TGA measurements and Alfred Frey for helping me fix the evaporator (ELKE) power supply which was actually not broken, I was just measuring DC instead of AC voltage. Our office neighbors "the Guldis" in particular Ruben Casillas, thanks for all those free beers at the office and being cool about me not bringing anything, even though my house was less than a block away. To Liping Sun as well, thank you for proof reading my thesis and for your invaluable company during this time. To all my friends from Erlangen, in particular my flat mates Girish, Rye and Valeria it was always great fun going out or just staying at home with you guys. Last but not least, I would like thank all of my family, while all of you are very far away, I knew I could always count with your support and love which made this work easier to realize.

121

13. Curriculum vitae

Name:

Luis Francisco Portilla Berlanga

Contact:

[email protected]

Date of birth:

March 22 1984

Nationalities:

Mexican, Spanish

Place of birth:

Torreon, Mexico

Academic and Professional 10/2012 – 12/2015

10/2010 – 02/2012

07/2007-07/2010 10/2002 – 06/2007

Friedrich-Alexander Universität Erlangen-Nürnberg (FAU) Ph. D. in Materials Science Supervisor: Prof. Dr. Marcus Halik Universitat de Barcelona (UB) M. Sc. in Nanoscience and Nanotechnology Supervisor: Prof. Dr. Anna Vila Materials and Technologies (MATECH) Automation Engineer Instituto Tecnológico y de Estudios Superiores de Monterrey (ITESM) B. Sc. in Electronics Engineering

Publications [1]

A. Vilà, A. Gomez, L. Portilla, and J. R. Morante, “Influence of In and Ga additives onto SnO2 inkjet-printed semiconductor,” Thin Solid Films, vol. 553, pp. 118–122, Feb. 2014.

[2]

L. Portilla and M. Halik, “Smoothly tunable surface properties of aluminum oxide core-shell nanoparticles by a mixed-ligand approach.,” ACS Appl. Mater. Interfaces, vol. 6, no. 8, pp. 5977–82, Apr. 2014.

[3]

S. H. Etschel, L. Portilla, J. Kirschner, M. Drost, F. Tu, H. Marbach, R. R. Tykwinski, and M. Halik, “Region-selective Deposition of Core-Shell Nanoparticles for 3 D Hierarchical Assemblies by the Huisgen 1,3-Dipolar Cycloaddition,” Angew. Chemie Int. Ed., vol. 201501957, no. 32, p. n/a–n/a, Aug. 2015.

[4]

L. Portilla, S. H. Etschel, R. R. Tykwinski, and M. Halik, “Green Processing of Metal Oxide Core-Shell Nanoparticles as Low-Temperature Dielectrics in Organic Thin-Film Transistors,” Adv. Mater., vol. 27, no. 39, pp. 5950–5954, Oct. 2015.

123

124

Curriculum vitae

[5]

L. Zeininger, S. Petzi, J. Schönamsgruber, L. Portilla, M. Halik, and A. Hirsch, “Very Facile Polarity Umpolung and Noncovalent Functionalization of Inorganic Nanoparticles: A Tool Kit for Supramolecular Materials Chemistry,” Chem. - A Eur. J., vol. 21, no. 40, pp. 14030–14035, Sep. 2015.

[6]

H. Dietrich, S. Scheiner, L. Portilla, D. Zahn, and M. Halik, “Improving the Performance of Organic Thin-Film Transistors by Ion Doping of Ethylene-Glycol-Based Self-Assembled Monolayer Hybrid Dielectrics,” Adv. Mater., vol. 27, no. 48, pp. 8023–8027, Dec. 2015.

[7]

L. Zeininger, L. Portilla, M. Halik, and A. Hirsch, “Quantitative Determination and Comparison of the Surface Binding of Phosphonic Acid, Carboxylic Acid, and Catechol Ligands on TiO 2 Nanoparticles,” Chem. - A Eur. J., Jul. 2016.

[8]

J. Kirschner, L. Portilla, J. Will, M. Berlinghof, H. Steinrück, T. Unruh, M. Halik "Organic Thin Film Memory Devices Based On Tio2-Loaded Ps-Peo Blockcopolymer Dielectrics," manuscript in preparation.

[9]

S. Klein, M. Kızaloğlu, L. Portilla, L. Distel, M. Halik, C. Kryschi, "Non-toxic water soluble cobalt ferrite nanoparticles for low dose radiation therapy," manuscript in preparation.

Patents 1.

"Kern-Hülle-Partikel," PCT/EP2016/066051, Jul. 2016.

Conference Contributions 1.

Technical session at the MRS 2015 Spring Meeting, with the subject: "Fine Tuning of Mixed Self-Assembled Monolayers Grafted onto 0D and 2D Metal Oxides Nanostructures".

2.

Poster session participation at the Electronic Processes in Organic Materials Gordon Conference 2014, with the subject: “Fine Tuning of Mixed Self-Assembled Monolayers Grafted onto 0D and 2D Metal Oxides Nanostructures”.

3.

Conference paper at the MRS 2013 Spring Meeting, with the title: "Metal Oxides as functional semiconductors. An Inkjet Approach".

4.

Poster session participation at the ITC 2012 8th International Thin-Film Transistor Conference, with the subject: “SnO2-based TFTs by inkjet”.