Assessment of Mixed Monolayer-protected Gold Nanoparticles

Assembly in Solution: Study and Characterization

Ana Isabel Martins Tiago Fernandes

in fulfillment with thesis requirement for the degree of

Master of Science in Biological Engineering

Examining Committee

Chair: Prof. Dr. Luís Joaquim Pina da Fonseca (IST)

Advisors: Prof. Dr. João Pedro Rodrigues Estrela Conde (IST)

Prof. Dr. Molly M. Stevens (IC)

Reviewer: Prof. Dr. Duarte Miguel de Franca Teixeira dos Prazeres (IST)

September 2010

2 of 94

Assessment of Mixed Monolayer-protected Gold Nanoparticles

Assembly in Solution: Study and Characterization

Ana Isabel Martins Tiago Fernandes

in fulfillment with thesis requirement for the degree of

Master of Science in Biological Engineering

Examining Committee

Chair: Prof. Dr. Luís Joaquim Pina da Fonseca (IST)

Advisors: Prof. Dr. João Pedro Rodrigues Estrela Conde (IST)

Prof. Dr. Molly M. Stevens (IC)

Reviewer: Prof. Dr. Duarte Miguel de Franca Teixeira dos Prazeres (IST)

September 2010

3 of 94

Acknowledgements

Firstly I would like to express my deepest gratitude to Prof. Molly Stevens, my supervisor at Imperial

College, for giving me the opportunity to work on this interesting project. Her constant encouragement,

interest and insightful support had a considerable impact on my success in the course of my work. She

gave me the freedom to explore new ideas while keeping me focused on my goals. It was a privilege to

have worked in Prof. Molly Stevens ’ group.

I am also thankful to Vanessa LaPointe for her brilliant ideas and continuous support and to Nia Bell, my

synthesis mate, for all her availability and sympathy given at laboratory!

I also want to express my sincere gratitude to Dr. Nicolas Schaeffer he was la lumière au bout du tunnel

when I most needed it. Merci beaucoup!

My experience at IC was definitely improved by the constant sense of mutual aid and friendship with the

greatest office mates (e.g. Air, Benji, David, Kristy, Jess, John, Maria, Mathew, Pinyuan, Stuart…!)

I want to thank Prof. Stellacci and Dr. Jeffrey Kuna at MIT for their availability and for their encouraging

and helpful advices.

I would also like to thank Prof. João Pedro Conde from IST, for his support, patience and

encouragement.

I am short of words for the immeasurable emotional and financial support of my family, especially of my

parents, brother, sister and grandmother. To all my friends who helped me in one way or another.

Obrigada Gonçalo, Carmo e Rui.

Finally, I dedicate this work to Francisco, he is still too young to understand the importance he had and

the inspiration and strength he gave during the past year.

4 of 94

“Things on a very small scale behave like nothing that you have any direct experience about. They do

not behave like waves, they do not behave like particles, they do not behave like clouds or billiard balls,

or weights on springs, or l ike anything that you have ever seen.”

Richard Feynman

5 of 94

Abstract

With the recent progress in nanoscience, gold nanoparticles (NP) have found widespread use in many

areas of scientific research ranging from biology to physics and medicine. These particles are easily

synthesized and can be readily coated with a self-assembled monolayer (SAM) of thiolated ligand

molecules that, depending on the synthesis conditions, phase-separate into ordered domains (rippled

domains) that encircle the gold core. This ligand shell confers stability against coalescence and controls

the particle's interactions with its environment (e.g. assembly, electron transfer ability). Furthermore,

the ability to manipulate and assemble these nanomaterials through the controlled functionalization of

their ligand shell is crucial for their incorporation into and development of new nanoparticle based

materials and devices. The engineering process behind the creation of those new nanostructures is

strongly dependent on the methods used to position these particles in specific locations in relation to

their neighbouring counterparts. Besides the individual NP properties, the gl obal properties and

functions of the assembled nanostructured systems are also ruled by the interparticle distances and

interactions between each of the cons tituent nanoparticle.

Herein, we describe the synthesis and characterization of near monodisperse homo and mixed

monolayer-protected gold NPs. Crucial information about these particles spontaneous assembly in

solution is also provided and a procedure to prevent this unwanted aggregation is presented.

Furthermore, in order to investigate how, depending on their ligand shell morphology (rippled or not),

these particles assemble, studies (UV-Vis and NMR spectroscopies and transmission electron microscopy

(TEM)) conducted on five different series of monolayer-protected metal nanoparticles (MPMNs) in the

presence of dithiol containing molecules are reported. High linker to gold NPs ratios (~2500) induce fast

and high rates of place-exchange reactions between the ligands in the SAM and the incoming linker

molecules, and three-dimensional ball -type aggregates form. In contrast, lower ratios (~125) do not

induce NPs assembly. We conclude it is very challenging to have a precise control over the number of

functional groups attached to each particle, therefore further studies with different solvents, surfactants

and cross-linker molecules should be performed.

Keywords: Gold Nanoparticles, Self-assembled monolayer, Phase-separated ordered domains,

Nanoparticles assembly, Dithiol, Trasmission electron microscopy

6 of 94

Resumo

Com os recentes progressos em nanociência, as nanopartículas de ouro têm sido alvo de estudo

intensivo apresentando uma ampla util ização em diversos campos da investigação científica que se

estendem por áreas como a biologia, a física e a medicina. Estas partículas são facilmente sintetizadas,

sendo espontaneamente revestidas por uma monocamada auto-agregada de moléculas ligantes que,

dependendo das condições de síntese, se separam e organizam formando domínios ordenados à volta

da partícula de ouro. Este revestimento de ligandos confere estabilidade contra coalescência e controla

as interacções destas partículas com o ambiente local ( i .e. agregação, capacidade de transferência de

electrões). Além disso, a capacidade de manipular e agrupar estes nanomateriais através da

funcionalização controlada da monocamada de ligandos é crucial para o desenvolvimento e/ou sua

incorporação em novas nano-estruturas e dispositivos. O processo de criação destas nano-estruturas

está dependente dos métodos utilizados para posicionar estas partículas em localizações específicas em

relação às partículas vizinhas. Para além das propriedades individuais de cada nanopartícula, as

características e propriedades globais das nano-estruturas agregadas dependem igualmente das

distâncias e interacções estabelecidas entre cada uma das partículas constituintes.

Na presente tese, a síntese e caracterização de nanopartículas protegidas por monocamadas compostas

por um (homogéneas) ou mais tipos (mistas) de ligandos é descrito. Informação crucial sobre a

agregação espontânea destas partículas em solução é fornecida e um processo para prevenir esta

agregação indesejada é apresentado.

A fim de investigar o modo de agregação destas partículas na presença de um linker (ditiol), estudos

(Espectroscopia de absorção UV-Vis, espectroscopia RMN e microscopia de transmissão electrónica) em

cinco séries de nanopartículas com diferentes morfologias na monocamada de ligandos (ordernada ou

não ordenada) são descritos. Rácios linker por nanopartícula de ouro elevados (~2500) induzem rápidas

trocas entre os l igandos da monocamada e as moléculas do l inker, formando-se agregados tri -

dimensionais de forma circular. Por outro lado, rácios baixos (~125) não induzem a agregação das

partículas. Conclui-se que é muito difícil controlar o número de grupos funcionais na superfície de cada

partícula e estudos com outros solventes, surfactantes e l inker deverão ser seguidos.

Palavras-chave: Nanopartículas de ouro, Monocamada auto-agregada, domínios separados

ordenadamente, Organização de nanopartículas, Ditiol, Microscopia de transmissão electrónica

7 of 94

Contents

Acknowledgements ................................................................................................................3

Abstract................................................................................................................................5

Resumo ................................................................................................................................6

Contents ...............................................................................................................................7

List of Tables .........................................................................................................................9

List of Figures ...................................................................................................................... 10

List of Abbreviations and Symbols........................................................................................... 13

1. Introduction ................................................................................................................. 15

2. Literature Review .......................................................................................................... 16

2.1. Nanoscience ...................................................................................................... 16

2.2. Nanoparticles (NPs): Nanoscience Building Blocks .................................................... 17

2.2.1. Gold Nanoparticles...................................................................................... 18

2.2.1.1. Historical Background on Gold Nanoparticles ............................................... 18

2.2.1.2. Stabilization of Gold Nanoparticle Dispersions .............................................. 19

2.2.1.3. Properties and Characteristics of Gold Nanoparticles. .................................... 19

2.2.1.4. Surface Plasmon Resonance ...................................................................... 20

2.3. Monolayer-Protected Metal Nanoparticles.............................................................. 20

2.4. Self-assembled Monolayers (SAMs) ....................................................................... 20

2.4.1. Phase-Separated Ordered Domains ................................................................ 22

2.4.1.1. Experimental Observations ....................................................................... 22

2.4.1.2. Properties of rippled MPMNs .................................................................... 25

2.4.2. Place-Exchange Reactions. ............................................................................ 25

2.5. MPMNs Assembly............................................................................................... 27

2.5.1. Polar Defects.............................................................................................. 27

2.5.2. Dithiols ..................................................................................................... 29

2.5.3. Other types of Assembly .............................................................................. 30

3. Thesis Outline ............................................................................................................... 31

4. Materials and Methods .................................................................................................. 32

4.1. Characterization Techniques for the Nanoscale ........................................................ 32

4.1.1. UV/Visible Spectroscopy............................................................................... 32

4.1.2. Nuclear Magnetic Resonance (NMR) Spectroscopy............................................ 34

4.1.3. Transmission Electron Microscopy ................................................................. 35

4.2. NP Synthesis ...................................................................................................... 36

4.3. NP Characterization ............................................................................................ 37

4.3.1. UV/Visible Absorption Spectroscopy............................................................... 37

8 of 94

4.3.2. Transmission Electron Microscopy (TEM) ......................................................... 38

4.3.3. Nuclear Magnetic Resonance (NMR) Spectroscopy............................................ 38

4.4. NP Assembly...................................................................................................... 38

4.4.1. UV/Visible Absorption Spectroscopy............................................................... 39

4.4.2. Transmission Electron Microscopy (TEM) ......................................................... 39

5. Results and Discussion.................................................................................................... 40

5.1. Synthesis and Analysis of Monolayer-Protected Metal Nanoparticles ........................... 40

5.2. MPMNs Molarity ................................................................................................ 48

5.2.1. Number of Gold Atoms: ............................................................................... 48

5.3. Cross-linking of MPMNs in solution........................................................................ 52

5.3.1. UV/Visible Analysis...................................................................................... 52

5.3.2. TEM Analysis .............................................................................................. 56

5.3.2.1. Hypothesis ............................................................................................. 56

5.3.2.2. TEM Sample Preparation .......................................................................... 58

5.3.2.3. Stability of Particles in Solution .................................................................. 59

5.3.2.4. Assembly Variations with Time .................................................................. 66

5.3.2.5. Assembling Behaviours with Cross -linker Concentration ................................. 70

5.3.2.6. Assembly after Purification Steps ............................................................... 75

5.3.2.7. New cross-linker approach ........................................................................ 77

6. Conclusions and Future Work .......................................................................................... 81

7. References ................................................................................................................... 83

Appendix ............................................................................................................................ 92

9 of 94

List of Tables

Table 1 Amounts of l igands used in the synthesis of NPs with various compositions .......................... 37

Table 2 Table of average core size ± Standard deviation for MH/OT MPMN series: ........................... 44

Table 3 Total yield and efficiency of MPMNs synthesis................................................................. 45

Table 4 Plasmon peak values and intensities for each of the MH/OT MPMNs series. ......................... 46

Table 5 Estimated gold clusters characteristic dimensions. ........................................................... 49

Table 6 Linear formula, molecular weight, molar concentration and average number of nanoparticles

per gold cluster (NNPs) for each set of MPMNs. ........................................................................... 50

Table 7 TEM grid and TEM image considerations. ....................................................................... 50

Table 8 1,9-nonanedithiol and 1,16-hexadecanedithiol chemical structure and molecular weight ........ 51

Table 9 Number of mols (N) and number of cross-linker molecules (Ncross-linker) for different molar

concentrations considering a final volume of 0.5 ml of cross -linker solution..................................... 51

Table 10 Molar concentration (C), number of mols (N) and number of NPs and thiols (NNanoparticles and

Nthiols respectively) correspondent to two different MPMNs final concentrations, considering a final

MPMNs solution volume of 0.5 ml. .......................................................................................... 51

Table 11 Cross-linker to gold NP and cross-linker to thiol ratios ..................................................... 52

Table 12 Different amounts of time samples were left under stirring after the purification steps......... 61

Table 13 Plasmon peak wavelengths and intensity variations with time for MPMNs solution .............. 65

Table 14 Table with the calculated linker to particle and linker to thiol ratio values........................... 70

Table 15 Different conditions used in the assembly study, when the solutions analysed were submitted

to the purification protocol..................................................................................................... 75

10 of 94

List of Figures

Figure 1 Growth of manufacturer-identified, nanotechnology products. ......................................... 17

Figure 2 The Lycurgus Cup ...................................................................................................... 18

Figure 3 Faraday’s colloidal ruby gold ....................................................................................... 19

Figure 4 Representative diagram of an SAM of alkanethiolates supported on a gold NP .................... 21

Figure 5 STM image of MPMNs showing phase-separated ordered domains .................................... 22

Figure 6 Surface plot of the ligand shell presenting the contours of the ripples................................. 23

Figure 7 Three dimensional rendering of gold NPs STM height images ........................................... 23

Figure 8 Schematic representation of the free volume available to l igands on curved surfaces ............ 24

Figure 9 Equilibrium arrangements of binary SAMs adsorbed on NPs with different degrees of curvature

determined through mesoscale simulations............................................................................... 24

Figure 10 Illustration of the processes involved in ligand place-exchange reactions on gold MPMNs .... 26

Figure 11 Illustration of a rippled MPMN showing one of the two diametrically opposed polar defects . 28

Figure 12 TEM images of MPMN chains. ................................................................................... 28

Figure 13 TEM images of l inear hierarchical assemblies of MPMNs ................................................ 29

Figure 14 Schematics of MPMN ligand shell chemistry and morphology.......................................... 31

Figure 15 Schematic depicting the promotion of an electron from an orbital in the ground state (π) to an

unoccupied orbital at a higher energy level (π*) ......................................................................... 33

Figure 16 Light passing through the cuvette containing the sample. ............................................... 33

Figure 17 Components of a transmission electron microscope (TEM). Adapted from [122] ................. 36

Figure 18 Chemical structures of 1-octanethiol and 6-mercapto-1-hexanol...................................... 40

Figure 19 Representative TEM images obtained for four different MH/OT MPMNs synthesized showing

the size of the gold core of the different particles. ...................................................................... 41

Figure 20 Representative TEM images obtained for three different MH/OT MPMNs synthesized showing

the size of the gold core of the different particles. ...................................................................... 42

Figure 21 Histograms obtained for five different MH/OT MPMNs synthesized de picting core size

distributions. ....................................................................................................................... 43

Figure 22 Histograms obtained for two different MH/OT MPMNs synthesized depicting core size

distributions. ....................................................................................................................... 44

Figure 23 UV-Vis absorption spectra of MPMNs suspended in EtOH ............................................... 46

Figure 24 1H NMR spectrum obtained for 2:1 MH/OT in CD3 OD ..................................................... 47

Figure 25 Fi lm of MH/OT 2:1 nanoparticles imaged by AM-AFM in ultrapure water........................... 47

Figure 26 Graph showing the relation between the number of layers of a gold NP and the total number

of gold atoms. ...................................................................................................................... 48

Figure 27 Change of absorbance in the 450-900 nm region of UV/Vis spectra for 1:1 MH/OT MPMN

solution upon addition of 1, 5 and 10 mM solution of cross -linker.................................................. 53

Figure 28 Change of absorbance in the 450-900 nm region of UV/Vis spectra for 5:1, 2:1, 1:2 and 1:5

MH/OT MPMNs upon addition of 5 mM of NDT solution. ............................................................. 54

11 of 94

Figure 29 SPR peak shifts for all the batch of the 5 sets of MPMNs tested at different times. .............. 55

Figure 30 Stabilization time .................................................................................................... 56

Figure 31 Observed assembly of the 1:1 MH/OT gold MPMNs. ...................................................... 57

Figure 33 Predicted structures of the assembled 2:1 and 1:2 MH/OT gold MPMNs ........................... 57

Figure 33 Expected structures of the assembled 5:1 and 1:5 MH/OT gold MPMNs ............................ 58

Figure 34 TEM images of 2:1 MH/OT MPMNs solution ................................................................. 59

Figure 35 TEM images of 2:1 MH/OT MPMNs solution after sonication........................................... 60

Figure 36 TEM images of 2:1 and 1:1 MH/OT MPMNs solutions 2 hours after purification steps . ......... 60

Figure 37 TEM pictures of the 5:1 MH/OT MPMNs solution 3 days after purification steps ................. 61

Figure 38 TEM pictures of the 1:1 MH/OT MPMNs solution 1 hour after purification steps. ................ 62

Figure 39 TEM pictures of the 1:1 MH/OT MPMNs solution 5 hours after purification steps................ 62

Figure 40 TEM pictures of the 1:1 MH/OT MPMNs solution 3 and 14 days after purification steps ....... 63

Figure 41 TEM images of the 1:1 MH/OT MPMNs solution 5 days after purification steps................... 63

Figure 42 UV-Vis absorption spectra of 1:1 MH/OT MPMNs dissolved in EtOH The sample was kept

untouched during a period of twelve weeks, and during this time interval several UV/Vis samples were

taken periodically ................................................................................................................. 64

Figure 43 1:1 MH/OT in Methanol............................................................................................ 65

Figure 44 TEM images of 1:1 MH/OT MPMNs solution 1 hour after adding the cross -linker, NDT. ........ 66

Figure 45 TEM images of 1:2 MH/OT A MPMNs solution 1 hour after adding the cross-linker, NDT ...... 66

Figure 46 TEM images of 1:2 MH/OT B MPMNs solution 1 hour after adding the cross-linker, NDT ...... 67

Figure 47 TEM images of 5:1 MH/OT MPMNs solution 1 hour after adding the cross -linker, NDT ......... 67

Figure 48 TEM images of 1:1 A and 1:2 B MH/OT MPMNs solutions 1 minute after adding the cross-

linker (NDT) solution ............................................................................................................. 68

Figure 49 TEM images of 1:2 B (same grid as figure 48 (b) but analysed two weeks after), 2:1 A and 5:1 A

MH/OT MPMNs solutions 1 minute after adding the cross-linker (NDT) solution............................... 69

Figure 50 TEM images of 1:2 MH/OT MPMNs solution 1 and 5 minutes after adding of cross-linker (NDT)

solution (1 mM).................................................................................................................... 70

Figure 51 TEM images of 1:2 MH/OT MPMNs solution 1, 5 and 15 minutes after adding of cross-linker

(NDT) solution (0.5 mM)......................................................................................................... 71

Figure 52 TEM images of 2:1 MH/OT MPMNs 1 minute after adding the 1, 0.5 and 0.1 mM solution of

cross-linker (NDT). Scale bars: 50 nm ........................................................................................ 72

Figure 53 TEM images of 2:1 MH/OT sonicated MPMNs prepared 1 minute after adding the 0.5 and 0.1

mM solution of cross-linker, NDT. ............................................................................................ 72

Figure 54 TEM images of 2:1 MH/OT MPMNs 1 and 5 minutes after adding the “old” 1 mM solution of

cross-linker, NDT .................................................................................................................. 73

Figure 55 TEM images of 2:1 MH/OT MPMNs 1 and 5 minutes after adding the “fresh” 1 mM solution of

cross-linker, NDT .................................................................................................................. 73

Figure 56 TEM pictures of the 1:1, 2:1 and 5:1 MH/OT purified MPMNs solutions prepared 1 and 10

minutes after adding the 1mM cross-linker solution, NDT............................................................. 76

12 of 94

Figure 57 TEM pictures of the 1:1 MH/OT purified MPMNs solution prepared 10 minutes, 1 and 12 hours

after adding the 0.01 mM cross-linker solution (NDT) .................................................................. 77

Figure 58 Schematic illustrating the cross-linking process for the two different dithiol containing

molecules used: NDT and HDDT. .............................................................................................. 78

Figure 59 TEM images of 1:1 MH/OT MPMNs solution 1 and 15 minutes, 1, 3 and 18 hours after adding a

0.01 mM cross-linker solution (HDDT)....................................................................................... 79

Figure 60 Schematic illustrating the incoming cross -linker molecule coiling around in the gold NP

forming two thiol -gold bonds in the same particle. ..................................................................... 79

Figure 61 TEM images of 1:1 MH/OT MPMNs solution 1 minute after adding a 5 mM cross -linker

solution (HDDT) .................................................................................................................... 80

13 of 94

List of Abbreviations and Symbols

AM-AFM Amplitude modulation - Atomic force microscopy

CCD Charge-coupled device

DCM Dichloromethane

DNA Deoxyribonucleic acid

Et2O Diethyl ether

EtOH Ethanol

FTIR Fourier transform infrared spectroscopy

HDDT 1,16-hexadecanedithiol

MH 6-mercapto-1-hexanol

MPMN Monolayer-protected metal nanoparticles

MW Molecular Weight

NDT 1,9-nonanedithiol

NMR Nuclear magnetic resonance

NP Nanoparticle

OT 1-octanethiol

PDI Polydispersity index

PMMA Poly(methyl methacrylate)

SAM Self-assembled monolayer

SPR Surface plasmon resonance

STM Scanning tunnelling microscopy

TEM Transmission electron microscopy

UV Ultraviolet

Vis Visible

I0 Intensity of l ight entering the sample

It Intensity of l ight exiting the sample

A Absorbance

T Transmittance

ε Absorptivity

c Concentration

l Path length

r Microscope resolution

λ Wavelength

µ Medium refractive index

α Semi-angle of microscopes aperture

14 of 94

NAu Total number of gold atoms

d Particles’ average diameter

VAu Particles’ volume

Nthiols Number of l igands per gold nanoparticle

LTEM image Square side of TEM image

ATEM image Observable area per TEM image

dTEM grid TEM grid total diameter

ATEM grid TEM grid total area

NP/ATEM image Total number of particles per TEM image

NP/ATEM grid Total number of particles per TEM grid

Ncross-linker Number of cross-linker molecules

15 of 94

1. Introduction

The idea of building structures from the bottom-up, molecule by molecule or atom by atom, is one of

the core concepts of nanoscience [1]. However, the success and potential behind this versatile “bottom-

up” approach is dependent on two main points: the first being the synthesis, at the nanoscale, of

building units and the second being the organization of these nanopieces together into a device or

material with predefined and sometimes sophisticated structures and properties. Generally

nanomaterials’ properties are not only dependent on each unit that is forming the final structure, but

also on the space and type of interaction existent between them. In the past thirty years a remarkable

success has been attained in the development of synthesis protocols for several types of nanobuilding

units, e.g. nanoparticles (NPs), nanofibers and nanorods [2]. In fact, these nanosized materials have

already been introduced in some commercial applications such as clothing and footwear (i .e. stain-

resistant trousers) [3, 4], titanium dioxide present in anti-aging cosmetic creams [5], and carbon

nanotubes present in stronger but very l ight tennis rackets and bicycle frames [3, 6].

A large effort was specifically focused on the study of metal -core nanoparticles because of their optical,

electronic, and surface properties [7]. Metal nanoparticles have shown various properties, such as single

electron transistor behaviour and surface plasmon resonance tuning and sensing [8, 9]. In the case of

monolayer-protected metal nanoparticles (MPMNs), the ligand shell that coats the particle surface

prevents coalescence when in solution and sometimes in solid state [10]. Moreover, this organic

molecules coating provides most of the particles' surface related properties, including assembly and

sensing properties as well as solubility [7, 10, 11]. Depending on the nature of the coating they can be

dried from and re-dissolved in solvents many times, or can be purified via dialysis, chromatography or

filtration [11]. However, controlled and directional assembly of these particles is still a bottleneck for

present nanotechnoloy research. Due to their considerably small dimensions, the association of these

nanobuilding units into complex and sophisticated materials, with the expected structure, properties

and functions represents an extremely complex scientific challenge.

16 of 94

2. Literature Review

2.1. Nanoscience

In the past, macro and microsized structures of well-defined nature, shapes and functionalities have

been well documented. However over the last decades, there has been a considerable increase of

interest in the field of nanoscience and na notechnology [12] due to the development of new tools for

analysing, imaging and manipulating nanometer scaled objects, such as scanning probe and electron

microscopes [10]. This current tendency to shrink materials’ dimensions is strongly encouraged because

of the unique material properties and performance advantages (compared to the bulk) that arise when

their dimensions are decreased to the nanometer length scales [13].

Nanoscience deals with the control and manipulation of systems and objects in which at least one

dimension is between 1 – 100 nm [10]. It is a highly inter-disciplinary field where chemistry plays a

fundamental role in the development of nanostructures and new synthetic methods, physics is used to

explain and characterize changes in the properties of matter with size, engineering is essential in

applying the understanding of nano-scale materials into useful devices and biology often acts as the

source of inspiration, with biological systems offering many examples of sophisticated nanostructures

interacting in complex networks suggesting new strategies with which to build artificial nanosystems

[14].

Their nano-scaled dimensions impart these nanostructured systems with many exclusive properties and

therefore they represent promising candidates for a variety of applications such as drug delivery and

sensing in medicine, catalysts in materials science, electronic and optical devices in electronics and

magnetic storage media [15]. Additionally, the capacity to systematically modify their properties by

controlling the structure and the chemical properties of these nanostructured systems makes them

well-suited for uses in more fundamental scientific studies of nano-scale interactions. This is of special

interest considering that most of the biological processes occur at the nano-scale (e.g. interactions

between water molecules, cells and proteins).

As dimensions diminish to nanoscale, the properties of matter become scale dependent and materials

exhibit different properties from that of the bulk material leading to interesting physical behav iour

based on quantum-mechanical phenomena, such as electron affinity, optical effects, conductivity,

ionization potential, superparamagnetism, electron tunnelling and surface plasma resonance (SPR) [13,

16, 17].

Those unique optical, electronic, and thermal properties deliberately pursued in the nanometer length

scale are strongly related to the surface-to-volume ratio that is drastically increased compared to the

bulk material [13]. This size-dependent behaviour leads, in the most extreme situation, to structures

17 of 94

where almost every atom in the structure is interfacial. Atoms or molecules at the surface of a material

experience a different environment to that of the atoms situated in the bulk of the material and present

boundary properties that can be amplified in any interaction involving these nanosystems as the ratio of

atoms on the surface to atoms in the bulk becomes greater [18].

Over the last twenty years, several nanomaterials have been developed, some of them being already

industrially produced and commercially available (i.e. sunscreen, cosmetics, food packaging, clothing,

disinfectants and fuel catalyst [3]). Figure 1 shows the evolution of the “nanomarket” growth, depicting

the number of “manufacturer–identified, nanotechnology–enabled products” inventoried by the official

US Consumer Product Safety Commission and the Project on Emerging Nanotechnology [3].

Figure 1 Growth of manufacturer-identified, nanotechnology products listed on Project on Emerging

Nanotechnologies Consumer Products Inventory from 2005 to 2009 (in grey) showing products under possible

Consumer Product Safety Commission jurisdiction (blue) [3].

2.2. Nanoparticles (NPs): Nanoscience Building Blocks

Nanoparticles (NPs) represent an attractive category of nano-scaled materials, therefore they have been

intensively studied over the past few years.

They can be considered as zero-dimensional nanostructures. Several types of NP systems (i.e. metal,

metal-oxide, or semiconductor colloids and nanocrystals) have been prepared and studied [19]. They are

becoming an important group of nanomaterials due to the simplicity of their synthesis [11, 20, 21] and

their unique size- and shape-dependent [8, 22-24], electronic [25], optical [19, 26, 27] and catalytic [7]

properties. Moreover, chemical compositions and dimensions (~2 – 100 nm) [21, 28] makes them of

comparable size as biomolecules and biomolecular assemblies (e.g. proteins, nucleic acids) which makes

them well suited to investigate the biological interactions at the molecular -scale.

0

200

400

600

800

1000

2005 2006 2007 2008 2009

18 of 94

2.2.1. Gold Nanoparticles

In the field of nanosciences, a considerable effort was focused on the study of gold nanoparticles which

are made of the aggregation of a few metallic gold atoms (Au0), ranging from one to a few hundred

nanometers. To be stable in solution, these aggregates are generally surrounded by a protective layer

that can be a polymer, an organic or biological molecule, which prevents further aggregation or

coagulation between particles.

2.2.1.1. Historical Background on Gold Nanoparticles

Gold particles have been widely studied and used since the 5th

century B.C for coloring glass and to cure

illness, first appearing in China, Egypt and in the Roman Empire. One of the most well -known

demonstrations that gold NP exclusive properties were popular is their insertion into the famous

Lycurgus cup (Figure 2) in the 4th

century B.C.. This cup has a dichroic effect that makes it shine red in

transmitted light and green in reflected light [29].

The therapeutical effects of gold colloids were described for the first time in “panacea aurea auro

portabile” by Francisci Antonii in 1618[30]. Later, in 1676, the german chemist Johann Kunckels reported

the use of “drinkable gold” in his book “Nuetliche Observationes oder Anmerkungen von Auro und

Argento Potabili” in which he states that “drinkable gold containing metallic gold in a neutral, slightly

pink solution exert curative properties for several diseases” [31, 32].

A more complete review on gold colloids was published in 1718 by Hans Heinrich Helcher, who

remarkably found the need for gold colloids to be stabilized with boiled starch [32, 33].

Figure 2 The Lycurgus Cup in (a) reflected and (b) transmitted light. Department of Prehistory and Europe, The

British Museum. Reproduced from [29].

However, it was not until the 19th

century that research on this nano-scaled material has evolved

tremendously with the pioneering work carried by Michael Faraday, who by using a two phase synthesis

protocol (reduction of a chloroaurate (AuCl 4-) with phosphorous in CS2) concluded that the ruby-red

color of certain tainted glass was a result of the presence of small gold colloids [34, 35] (see Figure 3).

19 of 94

Figure 3 Faraday’s colloidal ruby gold. Reproduced from [34]

Then, in 1994, Brust et al. showed that combining Faraday’s two-phase colloid synthesis with the known

self-assembly of thiol molecules on gold surfaces creates neutral small gold NPs relatively monodisperse

in size (~3.4 nm) coated with an alkanethiolate monolayer [20]. The resulting particles can easily be

isolated from and redissolved repeatedly in solvents without experiencing irreversible aggregation or

decomposition and are stable in atmospheric conditions.

2.2.1.2. Stabilization of Gold Nanoparticle Dispersions

The chemical stabilization of particles is essential to prevent degradation processes such as oxidation or

undesired sintering of NPs and to avoid agglomeration. One of the key aspects about colloidal chemistry

has to do with the means used to stabilize the particle suspensions in the medium in which they are

dispersed [28, 36].

NPs dispersion behavior is essentially dependent on the Van der Waals attraction and on the Brownian

motion [36]. Van der Waals forces alone are only signi ficant for short inter-particle distances, however

when combined with the Brownian motion, which ensures the continuous collision of particles in the

medium, these two forces lead to irreversible aggregation. This aggregation can be controlled and

stopped in the presence of repulsive forces capable of counteracting these attractive forces [28]. This

can be attained or by electrostatic stabilization which creates a distribution of charged species in the

system and/or by steric stabilization which involves the adsorption of molecules or m acromolecules

onto the particle surfaces.

2.2.1.3. Properties and Characteristics of Gold Nanoparticles.

Gold NPs are distinct from many other types of NPs because gold possesses some practical and

distinguishing advantages. First and foremost it is a reasonably inert material. Gold does not undergo

oxidation at temperatures below its melting point, it does not react with atmos pheric O2 and with most

chemicals. These properties make possible to handle and control it under atmospheric conditions [10].

Gold is biocompatible, and binds thiols with high affinity, which is determinant for the formation of self -

assembled monolayers (SAMs) (vd. Section 2.4 – Self-Assembled Monolayers (SAMs)). Furthermore, gold

nanoparticles are amenable to “mixed monolayer” coverages , where different ligands can be affixed to

the particle surface in well-defined ratios [37].

20 of 94

Another reason for the common use of gold for the synthesis of nanoparticles is the well-defined

synthetic methods for their fabrication and manipulation. Indeed, they can be quickly synthesized by

several ways: either in aqueous [38-41] or organic solvent [20, 42] and following mono- [19, 42, 43] or

biphasic methods [20].

2.2.1.4. Surface Plasmon Resonance

Small metallic NPs exhibit strong optical absorption at one particular frequency due to a collective

oscillation of the free electron gas in the NP [44]. This frequency known as surface plasmon resonance

frequency is a result of the propagation of electromagnetic waves along the surface of a conductor [45,

46]. When the dimensions of the conductor are reduced, boundary and surface e ffects become very

important [47] and since the wave is on the boundary of the metallic surface and the external medium,

the resonance frequency becomes extremely sensitive to any change on this boundary. This frequency

can be tuned by varying the composition of the NP core [48] and the dielectric constant of the

surrounding material which is dependent on the solvent and molecules adsorption to the metal surface

[46, 49]. This sensitivity of the plasmon resonance to the environment has driven the use of NPs as

biological and chemical sensors [15, 50].

2.3. Monolayer-Protected Metal Nanoparticles

Monolayer-Protected Metal Nanoparticles (MPMNs) are supramolecular assemblies consisting of a

nanoscale, crystalline, metallic core surrounded by an outer ligand shell, a sel f-assembled monolayer

(SAM see below) composed of thiol-containing molecules bond to the surface through a sulphur-metal

bond [11, 51]. These particles exhibit numerous useful and unique properties many of which arise and

are adjusted by the close spatial contact between the core and the shell. These particles characteristics

are conferred by their metallic core (e.g. surface plasmon absorption) [8], their SAM (e.g. solubility and

sensing) [11], and to both of the components (e.g. single electron transistor) [9] and can be easily

synthesized [52]. They have been recently exploited for many applications in all most of science fields

ranging from material science [53], to medicine [54], biology [55] , physics [56] and chemistry [57].

2.4. Self-assembled Monolayers (SAMs)

Understanding the relationship between the nanostructure of a material and its macroscopic properties

has always been a major goal for interfacial science [50]. Self-assembled monolayers (SAMs) constitute a

particularly good way to explore these interactions as they are structurally well defined and offer the

prospect of creating organic interfaces that can be tailored for different mechanistic studies.

SAMs are ordered monomolecular assemblies that adsorb spontaneously on the metal substrate due to

a surface energy minimization of the metal NP (a planar surface or highly curved nanostructure) [10, 58].

21 of 94

These molecules consist of a “head-group” which shows a special affinity for a substrate and, in the

other end of the molecule, a “terminal -group” a tail with a functional group that controls the surface

properties of the SAM (Figure 4) [59]. SAM adsorption on gold NPs involves the initial physisorption of

the molecules on the metal and the subsequent chemisorption of the “head -groups”. This strong

covalent sulphur-gold (S-Au) bond (typically ~50 kcal mol-1

) is stable over a large range of temperatures,

solvents and pH [10, 60, 61]. Once adsorbed on the metal lattice, by losing the mercaptan’s proton, H,

SAM molecules can adopt energetically more favourable conformations, which allow high degrees of

van der Waals interactions (and in s ome cases hydrogen bonding) [62, 63] staying tightly packed. SAMs

composition can be easily and deliberately altered, for example, functional groups of the assembled

ligands can be tailored to exhibit for insta nce hydrophobic (e.g. methyl groups) or hydrophilic (e.g.

hydroxyl) ends [64].

It is also noteworthy that since the thiol molecules efficiently smooth the faceted and highly anisotropic

surface energy of the gold particles, once wrapped with this thiolated monolayer gold NPs acquire a

round and smooth shape [65, 66].

Figure 4 Representative diagram of an SAM of a lkanethiolates supported on a gold NP. Light grey circles represent

chemisorbed “head-groups”, which are thiol groups; dark grey circles represent the “terminal-groups”, which can

be a variety of chemical functionalities. Adapted from [10]

This protecting monolayer plays a critical role, controlling and ruling all the particles’ interactions with

the outside molecular environment and providing them with a long series of properties, such as stability

22 of 94

against aggregation into the thermodynamically preferable bulk state, solubility in many solvents,

assembly properties and sensing of specific biomolecules [51]. However, a detailed understanding of the

dynamics of the assembly still lacks of a complete and proper characterization [67].

2.4.1. Phase-Separated Ordered Domains

SAMs composed of a mixture of l igands on flat surfaces have long been known to phase-separate into

randomly sized and shaped domains [68] or into worm-like structures [48] as observed by scanning

tunnelling microscopy (STM) [69-71].

Phase-separation also occurs on NPs. However the structure of SAMs on NPs is complicated by the fact

that the two-dimensional monolayer (the SAM) must be adsorbed onto a three-dimensional structure

(the NP). This additional complication leads to a unique situation of phase-separation into organized

(rippled) domains.

2.4.1.1. Experimental Observations

In 2004, Jackson et al. first confirmed the presence of ordered phase-separated domains in the ligand

shell of MPMNs. They found that those SAMs of alkanethiol l igands with varying tail groups phase

separate into stripes or ripples that encircle or spiral around the me tal nanoparticle (see Figures 5 and

6) [52]. The width of those ribbon-like domains was found to be as small as < 1 nm (often no more than

two molecules) and their presence was first confirmed by STM [52] and later using Fourier transform

infrared spectroscopy (FTIR) [72]. With FTIR, due to the perturbed intermolecular forces noticeable in

phase-separated domains an upward shift in the CH2 symmetric stretching frequency was observed

confirming the presence of phase-separation. Unfortunately, those results could not completely prove

that the phase-separation was ordered.

Figure 5 a) STM image of MPMNs showing phase-separated ordered domains on b) STM image of a sing le gold

nanoparticle, arrows indicate ripple spacing. c) Schematic depiction helping to v isualise the arrangement of

molecules on the NP surface where the ra ised yellow regions symbolize the octanethiol molecules and the red ones

depict mercaptopropionic acid. Reproduced from [52]. Scale bar: 10 nm

23 of 94

Figure 6 Surface plot of the ligand shell presenting the contours of the ripples (scales in nm). The distance between

the two arrows shows the ripple spacing. Reproduced from [52].

The same group have also shown that changing the stoichiometric ratio of the ligands during synthesis

affects the SAM shell morphology, ranging from discrete phase -separated domains to highly ordered

ripples (see Figure 6). This phenomenon of ordered phase separation was then confirmed to form for a

wide variety of binary l igand shell compositions, including molecules with variable chain lengths and

molecules with different end groups and backbone structures (e.g. aliphatic vs. aromatic).

Figure 7 Three dimensional rendering of gold NPs STM height images. (a) (b) Ordered phase-separated domains (2:1

molar ratio of decanethiol/mercaptopropionic acid) and the schematic drawing respectively, (c) (d) Discrete packed

phase-separated domains of the less abundant component (10:1 molar ratio of OT/ mercaptopropionic acid) and

the schematic drawing repectively. Reproduced from [52].

In 2007, Glotzer et al. carried out atomistic and mesoscale simulations in order to assess the origin of

the experimentally reported stripe formation [73].They predicted that ligands nanophase-separation

into striped patterns on NPs surfaces will depend on a balance between the enthalpic losses and

entropic gains [73].

24 of 94

As presented on Figure 8, phase-separation contributes to an increase in conformational entropy.

Basically, the longer or bulkier surfactant tails can occupy more space when surrounded by shorter or

less bulky molecules. When this entropy gain overcomes (a) the energy reduction that would be

observed in the case of bulk separation and (b) the energetic penalty of generating additional interfaces,

then nanophase-separated stripes will form. In this situation, the domain width of the ripples will

increase with the ligand chain length [73].

Figure 8 Schematic representation of the free volume (indicated by the shaded cones) available to ligands on curved

surfaces, and respective simulated cross-sectional v iews. (a) indicates ligands with s imilar lengths; (b) shows ligands

with cons iderable differences in lengths. The free volume and the consequent ga in in conf igurational entropy for

ligands with different lengths is clearly larger than that for ligands with the same length. Adapted from [73]

In addition, simulations predicted bulk phase separation for mixtures of short enough surfactants (e.g.

three carbon chains), or for l igands with small bulkiness difference. Bulk phase -separation was also

observed in the case of extremely high degrees of curvature (very small NPs) (see Figure 9 (a)). The

spheres, if small when compared to the surfactants length will not gain significant entropy by generating

extra interfaces (and therefore ripples) as the tails already possess enough conformational entropy by

moving radially outward in the sphere [73-75]. Increasing the NP radius results in ordered stripes as

mentioned above (Figure 9 (b)). In larger spheres (Figure 9 (c)), disordered stripes and irregular domains

will form as predicted by atomistic simulations. Extending the radius to infinite (flat surfaces) the SAM

will form wormlike stripes for the referred tail length ratio - 4:7 (Figure 9 (d)) [73].

Figure 9 Equilibrium arrangements of binary SAMs adsorbed on NPs with different degrees of curvature determined

through mesoscale s imulations. Yellow and red lobes symbolize head terminals of longer and shorter ligands,

respectively. Sphere radii: (a) 3 σ – small NPs: binary mixtures of surfactants separate into two bulk phases , (b) 5 σ –

increas ing NP radius: phase-separation into ordered stripes occurs, (c) 10 σ - further increase in NP radius:

disordered stripes and patchy domains, (d) ∞ - f lat surface: same as (c). Sphere radius not in sca le. Reproduced

from [73]

25 of 94

Then, in 2008, Centrone et al. confirmed that when ordered phase-separated domains (ripples) are

present, the solubility of MPMNs in ethanol shows an unexpected behaviour, and does not increase

monotonically with the concentration of hydrophilic ligand [52, 76]. In this case, the longer of the two

ligands determines the solubility. On the other hand, as should be expected, when the particles have no

ripples the concentration of the hydrophilic ligand determines the solubility. These findings prove that

ligand shell morphology influences the solubility of these NPs almost as much as the chemical and

molecular composition of the SAM [76].

At the same time, in a combined STM and TEM study, Hu et al. showed that when ripples are present in

the ligand shell, they enhance the particle-particle interactions, leading to a stronger degree of

interdigitation when compared to their homoligand counterparts and are considerably less able to form

ordered supracrystals [77].

2.4.1.2. Properties of rippled MPMNs

This unprecedented molecularly defined arrangement imparts a variety of singular properties to a NP

system such as non-monotonic dependence of solubility [52, 78] on ligand shell composition and

demonstrated good resistance to protein nonspecific adsorption [78]. The latter occurs for NPs coated

with a mixture of hydrophobic- and hydrophilic-ligands. Once formed, the inter-distance between the

hydrophilic and hydrophobic domains in the ligand shell will be too small to allow the protein to find a

suitable conformation to adsorbe onto the NP [79]. This is the functional basis behind the dolphin skin

resistance to biofouling [80].

2.4.2. Place-Exchange Reactions

The development of versatile strategies to functionalize alkanethiolate gold MPMNs is vital in the

development of these materials as potential chemical reagents and catalysts [20]. In this context, l igand

place-exchange reactions are a key step in opening up MPMNs functionalization [11, 67].

MPMNs with alkanethiolate monolayers (R-S) can be changed by exposing them to a solution of another

kind of molecule (R’-S). This type of reaction is schematized below:

where x and n represent the numbers of new and former ligands, respectively.

Although the microscopic details remain unclear, reports of l igand exchange on two-dimensional-SAMs

on gold surfaces often show that many sites exchange extremely slowly or not at all, whereas others are

relatively reactive [81-84]. Similarly for gold MPMNs Hostetler et al. found that exchange occurs

26 of 94

preferentially at minority sites such as defects and pinholes (e.g. terrace edges, vertexes) as well as grain

boundaries and that terrace-bond ligands are both less reactive and, at best, only slightly mobile (see

Figure 10). These assumptions suggest the existence of a hierarchy of the different core surface binding

sites with associated susceptibility to place exchange [67].

The same group also indicated two different exchange regimes to explain such a site preference: (a) the

incoming ligand enters the monolayer in order to undertake place -exchange. This associative

mechanism would favour less crowded, nonterrace sites [67]; or (b) l igands will be desorbed

(preferentially at defect sites) followed by attachment of new thiol to the newly created surface vacancy

[85, 86]. Also important to mention is that these easily exchanged sites are not a static population (as

confirmed by a serial exchange experiments [87]) and are l iable to migration within the ligand shell . (see

Figure 10) [67].

The same study also revealed that the MPMN monolayer only possess a limited number of ligands that

are bond weakly enough to be lost as disulfides. The kinetic re sults demonstrate that the rate of l igand

exchange on gold MPMNs depends on the concentration of the entering and exiting l igands, a rate that

is initially rapid but slows dramatically. Longer entering ligands and chain lengths in the protecting

monolayer will both decrease the rate of l igand exchange. Study of l igand place-exchange dynamics and

mechanism show that exchange has a 1:1 stoichiometry, which means that one molecule is adsorbed for

each molecule desorbed [67].

Figure 10 Representative illustration of the processes involved in ligand place-exchange reactions on gold MPMNs.

(a) Exchange of vertex thiolates (1) with solution thiol; (b) Exchange of edge and near-edge thiolates (2) with

solution thiol; (c) Exchange of terrace thiolates (3) with solution thiol; (d) surface migration among vertice and edge

thiolates; (e) Surface migration within edge (and near-edge) and terrace thiolates; (f) Surface migration within the

terrace thiolates. Adapted from [67].

27 of 94

2.5. MPMNs Assembly

To take advantage of the useful optical and electronic properties of MPMNs it is of big interest to be

able to predict and control the assembly of NPs into specific structures since this would expand the

range of potential applications of these particles [88]. In order to use NPs as nano blocks to build

complex devices, their positions relative to each other and to their environment must be precisely

controllable. Therefore, extensive research is evolving in order to try to organize MPMNs in a

predictable and ordered way.

NPs are inherently isotropic and therefore will not spontaneously form into stable anisotropic one

dimensional assemblies without experiencing some external driving force [89, 90]. Thus breaking the

symmetry of the inter-NP interactions represents one of the key challenges in nanomaterials research

[91]. In fact, efforts to control the assembly of MPMNs, based mainly on biomolecules [92-94] and other

templating molecules [95], have been delayed due to a lack of control in the number of receptors

interacting with the templating agent. Herein some examples of chemically directed assembly will be

reviewed.

2.5.1. Polar Defects

SAMs on flat gold surfaces form a two-dimensional crystal in which each ligand can be represented by a

vector corresponding to its projection in the surface normal (determined by the tilt angle) [10, 96]. In

the case of the SAM on a NP core, it is necessary to consider the assembly of a vectorial order

(projection of SAM ligands) onto a topological sphere [97, 98], which inevitably requires the formation

of two diametrically opposed defect points on the NP. This is consistent with the known “Hairy Ball

Theorem”, which states that it is impossible to arrange a vector field onto a sphere without the

formation of at least two diametrically opposed singularities. This theorem also explains quotidian

phenomena like the whirl present on our hair and also the existence of at least one cyclone in the

atmosphere at any given time [99].

Considering the last statements, the ordered domains present in the rippled MPMNs result in two

profoundly demarcated polar singularities [91]. Molecules in the poles, being not entirely stabilized by

intermolecular interactions with their neighbours , manifest themselves as highly reactive defect points

(Figure 11) and consequently are more vulnerable to be displaced by place-exchange reactions.

28 of 94

Figure 11 Schematic illustration of a rippled MPMN showing one of the two diametrically opposed polar defects

that need to exist to enable the alternation of concentric rings. This defect is highly reactive due to a lack of

intermolecular stabilizing forces. Reproduced from [91].

Stellacci and co-workers used these unstable singularities to functionalize MPMNs at two diametrically

opposed points [99] in order to form directional and controlled chains of NPs [91]. They positioned

carboxylic acid-terminated molecules at these unstable poles to create divalent NPs and then reacted

them with a diamine linker generating the linear chains of NPs. TEM was performed to assess and

characterize the chains of pole-functionalized rippled MPMNs (see Figure 12). The specificity of the

place-exchange reaction was demonstrated since a small number of branched chains and three-

dimensional aggregates were observed.

Figure 12 TEM images of MPMN chains. Chains were formed after adding a cross-linker to the solution. Most of the

chains do not have branches or 3D structures which is consistent with the fact that the two polar defects are the

most reactive points of the particles. Scale bars 200 nm, inset: left 25 nm; right 50 nm. Reproduced from [91].

They have also shown the ability to vary the distance between each particle in the linear chain by

altering the length of the linking diamine molecules [91].

In a subsequent study, Carney et al . demonstrated that the formation of l inear chains occurs only with

MPMNs within a determined size range (in the referred case from approximately 2.5 to 8.0 nm) which,

29 of 94

by extension means that the ordered striped organization of the assembled monolayer only exists in

that specified size range as it was previously predicted by simulation (section 2.4.1.2) [73, 75].

Later on, in a UV-Vis study Vanessa LaPointe also found that in the case of rippled MPMNs (1:1 MH/OT)

a clear change in the absorbance spectrum was visible only 5 minutes after starting the reaction with

the linker molecules (1,9-nonanedithiol) and appeared to be finished approximately at the end of 20

minutes. However, in the 5:1 MH/OT series, which does not have these diametrically opposed and

highly reactive defects , the cross-linking reaction and the subsequent aggregation took much longer. It

took more than one hour before the reaction appeared to have ended and the aggregation level was

lower.

2.5.2. Dithiols

The simplest way of inducing controlled MPMNs aggregation is to use bi -dentate thiol ligands that cross-

link the gold particles together as a result of the strong sulphur-gold interaction. The common feature of

these materials is that they are completely insoluble because of the high degree of three-dimensional

cross-linking [56, 100, 101]. Brust group explored this method and demonstrated that the central

property behind the structures of the generated assemblies is the number of l inking molecules per gold

NP. Adding 1,9-nonanedithiol molecules to the tetraoctylammonium bromide-stabilized gold NPs

(within a defined range of gold-dithiol molar ratios) assembled the particles together because thiols

form a stronger bond in the gold surface than the bromide molecules [102]. The resultant three-

dimensional ball -type assemblies precipitate at both high and low linker ratios, however at an

intermediate ratio (~60 - 14 000 ligands per MPMN) they remain dissolved in toluene. Curiously, the

round aggregates line up and generate relatively straight lines when ethanol is mixed with the toluene

solution (Figure 13) [102].

Figure 13 TEM images of linear hierarchical assemblies of MPMNs formed upon addition of ethanol to the toluene

solution. Reproduced from [102].

30 of 94

2.5.3. Other types of assembly

There is a wide range of methods to organize gold NPs besides the specific examples explained above.

Using the known covalent and strong non-covalent interactions between the ligands attached to the

NPs, various nanostructures, including mono- and multilayers, as well as composites with oligomers or

conducting polymers have been engineered [7, 103, 104]. Gold NPs have also been used as multivalent

cores to build dendritic structures [7] but the isotropic character of this type of binding represents a

problem for the bottom up fabrication of more complex nanoparticle -based structures.

Other groups studied monofunctionalized NPs bound to multidentate molecules with set up direc tional

interactions which permits a controlled assembl y. The required single valency in the NPs is prepared via

reaction with a functional thiol linked to a solid support [105-107]. For example, using the extreme

specificity of the interaction between DNA strands used as ligands, cyclic, linear, and discrete branched

structures of monofunctionalized gold NPs have been obtained [108, 109]. However, the introduced

single valency renders NPs spectator pendant groups, ra ther than building blocks for these nano-

engineering challenges.

Another strategy to achieve one-dimensional aggregates of NPs is the use of pre-existing one

dimensional structures such as nanotubes, as templates onto which NP arrangments can be formed.

The one-dimensional assembly of nanorods is also viable since the ends of these structures are made of

a different material in relation to the center, therefore these ends can be functionalized with molecules

with reactive end groups and generate l inear as semblies.

31 of 94

3. Thesis Outline The possibility to understand, control and predict particles assembly has become a very attractive

research area since directional and isotropic assembly of this nanoscale building blocks could impart

upon them properties that enable their use for diverse ends such as sensing and electronic devices.

During the last decade a large effort was put into the study of gold NPs assembly. However the recent

finding of ordered phase-separated domains, ripples, and the consequent properties this morphology

confers (non-monotonic dependence of solubility and protein resistance properties) provided a new tool

for the study of these particles’ interactions within the solvents [76] and their neighbouring MPMNs

[91].

In this context and based on the work developed by Stellacci’s group (see Section 2.5.1 – Polar Defects)

the aim of this thesis is focused on the assembly study of gold NPs coated with an organic layer

assembly in the presence of a cross-linker in solution. This monolayer is composed by the same

molecules (6-mercapto-1-hexanol (MH) and 1-octanethiol (OT)) but in different ratios (as illustrated on

Figure 14). 2:1, 1:1 and 1:2 MH/OT present ordered phase-separated domains, ripples, in their l igand

shell while 5:1 and 1:5 MH/OT present unordered domains.

Figure 14 Schematics of MPMN ligand shell chemistry and morphology. Ratios indicate stoichiometric ratios of

thiolated ligands (6-mercapto-1-hexanol and 1-octanethiol) during synthesis. (Schematics courtesy of of Dr. Steve

Mwenifumbo

The present project was divided in two main stages. First, seven different types of gold MPMNs were

synthesized and characterized. This was attained by direct observation (the formation of a brownish

powder is related with the presence of smaller particles), solubility testing in EtOH (0:1 M H/OT was

tested in toluene), UV-Vis absorption spectroscopy, 1H NMR spectroscopy and TEM analysis for size

distribution. An adequate characterization of gold nanoparticles is preponderant for the future use of

the same. Gold NPs characterization was developed along with Miss Nia Bell and Miss Vanessa LaPointe.

The second part of this project was focused on the study of these MPMNs assembly. The assemblies

were studied by UV-Vis spectroscopy to determine the rate of assembly for each type of MPMNs and

TEM to observe the morphology of the aggregates. The state of aggregation observ ed on the TEM

images was used to determine the parameters that were used in the subsequent TEM analysis, such as

cross-linker molar concentration, time and MPMN concentration.

32 of 94

4. Materials and Methods All reagents were purchased from Sigma-Aldrich Inc., UK (Dorset, UK) and used as received without

further purification, unless otherwise specified.

Solvents and acids such as absolute ethanol (EtOH), acetone, isopropanol, diethyl ether, (Et 2O),

dichlorometane (DCM), toluene, benzene, hexane, nitric acid (HNO3) and hydrochloric acid (HCl) were

purchased from VWR, UK (Lutterworth, UK)

Carbon coated Copper TEM grids (300 mesh) were obtained from Agar Scientific.

TEM Images were obtained using a JEOL 2010. The images were analyzed using NIH Image Software

ImageJ. UV-Vis absorbance measurements were acquired using a Perkin-Elmer spectrometer.

4.1. Characterization Techniques for the nanoscale

Nanostructures are “inconveniently small” [110] – too small to be observed and studied directly. For this

reason a “spy” versatile enough to report on a wide range of molecules and capable of relaying the

information on the structures, motions and chemical reactions of these systems without significantly

altering those properties is needed [110].

These nanostructures chemical and physical properties are mostly governed by their composition, size,

structure, shape and surface properties, thus their precise characterization is essential to ass ess these

parameters and understand their effect on the intrinsic properties of the materials. For these purposes ,

a series of analytical tools have been developed. Examples like electron microscopy [111], atomic force

microscopy (AFM) [112], scanning probe microscopy (STM) [113] are used for structural analysis of these

nanostructures. Spectroscopic techniques, such as UV-Vis absorption spectroscopy or Raman

spectroscopy are used to obtain information on the optical properties of the material [114-116]. The

different techniques used to characterize the nanoparticles mentioned in this thesis are described

below.

4.1.1. UV-Visible spectroscopy

A UV-Visible (UV-Vis) spectrophotometer is used to determine the absorption or transmission in both

the UV and visible wavelength [20]. Absorption of photons results in the promotion of an electron from

an orbital of a molecule in the ground state to an unoccupied orbital at a higher energy level (see Figure

15).

33 of 94

Figure 15 Schematic depicting the promotion of an electron from an orbital of a molecule in the ground state (π) to

an unoccupied orbital at a higher energy level (π*)

In general, UV-Vis spectroscopy is used to study how a sample responds to l ight as it measures the

attenuation of a beam of l ight after it passes through a sample (Figure 16). Ultraviolet and visible light

are energetic enough to promote outer electrons to higher energy levels, and UV/Vis is usually applied

to molecules or inorganic complexes in solution [117].

Figure 16 Light passing through the cuvette containing the sample.

A UV-Vis spectrophotometer consists of a radiation source which uses an incandescent bulb for visible

wavelengths and a deuterium lamp in the ultra -violet wavelengths, a sample holder, a monochromator

or diffraction grating (which enables the selection of a narrow band of wavelengths), a photodetector to

measure the intensity of l ight which is transmitted through the sample, and an output device [117].

When the beam passes through a sample, some light gets absorbed, while some continues through the

sample to the transmitter. The ratio of the intensity of l ight beam entering the sample (I 0) and coming

out (It) at a specific wavelength is defined as transmittance (T).

Absorbance (A) is the negative logarithm of T

34 of 94

The Beer-Lambert law states that for a given ideal solution, there is a linear relationship between

concentration and absorbance provi ded that the path length is kept constant. The extinction coefficient

(ε) is a constant for each molecule for each wavelength.

where ε, an intrinsic property of the species, represents the extinction coefficient of the substance (L

mol-1

cm-1

), c is concentration of absorbing species (mol L-1

), and l is the absorption path length (cm)

through the sample. Therefore, provided that ε and l are kept constant, there is a linear relationship

between concentration and absorbance.

Gold NPs, specifically, exhibit distinct colors (and thus SPR bands) that are a characteristic of particles

nature, shape, size or assembly state on the refractive index of the surrounding medium and on the

nature of their protective layer [114, 118] The SPR peak intensity and broadness is dependent on the

NPs solution concentration (and extinction coefficient) and its size dispersion respectively. [114, 118]

For instance, particles smaller than 3 nm in diameter do not show any SPR peak.

4.1.2. Nuclear Magnetic Resonance (NMR) Spectroscopy

Nuclear magnetic resonance is a property that magnetic nuclei have in a magnetic field, which cause the

nuclei to absorb energy from the electromagnetic pulse and radiate this energy back out at a specific

resonance frequency which, among other factors, depends on the strength of the magnetic field.

NMR Spectroscopy is a powerful technique that can provide useful and detailed information on the

topology, dynamics and structure of molecules and biomolecules in solution. This spectroscopic

technique relies on the magnetic properties of the atomic nucleus and can be used to investigate the

quantum mechanical properties of molecules. When placed in a strong magnetic field, certain nuclei

(e.g., 1H,

13C,

15N – the ones with an even number of electrons ) resonate at a characteristic frequency

(characteristic of the isotope and electronic/electromagnetic environment) in the radio frequency range

of the electromagnetic spectrum. Slight variations i n this resonant frequency give detailed information

about the molecular structure in which the atom resides, providing us structural information about the

global system [119]. NMR always provides “local” information; i . e. the world is studied from the

perspective of a single atom in a molecule where this atom can only “s ee” 5 Å (approximately three

bonds away). However this perspective can be tuned so we could “see” the world from each constituent

atom in the molecule. The NMR data consists of a series of relationships between the atoms of the

molecule, where the intensities of the signals are directly proportional to the concentration. With the

right information about these relationships we can construct an unambiguous model of the molecular

structure [119].

35 of 94

4.1.3. Transmission Electron Microscopy

Transmission Electron Microscopes today are capable of achieving a point to point resolution of better

than 0.1 nm. In fact, a major attraction to the early developers of the TEM was that, since electrons are

smaller than atoms, it would be possible, at least theoretically, to build a microscope capable of

“seeing” details below the atomic level [120].

In this context, transmission electron microscopy (TEM) is a powerful imaging technique used to

determine the size and shape of materials at the nanometer length scale that uses, in essence, the same

working principles of the light microscopy technique, both based on the foc us of an electron or a l ight

beam through magnetic or optical lenses, respectively. The resolution of the microscopy, r, is defined

by:

where λ represents the wavelength of the incident beam, µ is the medium refractive index and α stands

for the semi-angle of the microscope aperture [121, 122]. Since the wavelength of electrons is about ten

thousand times smaller than the photons wavelength, theoretically the TEM resolution will be

approximately ten thousand ti mes smaller than an optical microscope resolution, enabling the

observation of nanometer scale objects [121].

Figure 17 shows the components of a conventional TEM. The electron gun provides an intense beam of

high energy electrons that pass through an ultra thin specimen (10 to 100 nm), interacting with it. An

image is formed from the interaction of the electrons transmitted through the specimen. The electron

beam can be generated either by thermionic discharge or by field emission and TEMs use an

accelerating voltage between 100kV to 400 kV. To avoid scattering of electrons by air, TEMs are

operated in vacuum. The first condenser lens is used to create a de -magnified image of the gun

crossover and to control the minimum spot size available in the rest of the condenser system. The

second condenser lens affects the convergence of the beam at the specimen and the diameter of the

illuminated area of the sample. The condenser aperture controls the fraction of the electron beam

which is allowed to hit the specimen. The objective lens then forms an inverted initial image of the

sample. The objective aperture is used to select the electrons which will contribute to the image i.e. to

control the contrast of the initial image. The project lens is used to magnify the initial image formed by

the objective lens to a desired magnification for viewing. A viewing screen is an integral part of any

electron microscope. It can be coated with materials such as ZnS, which translates electron intensity to

l ight intensity, which we observe and record on a layer of photographic fi lm or on a CCD camera [123].

36 of 94

Figure 17 Components of a transmission electron microscope (TEM). Adapted from [123]

4.2. NP synthesis

Seven types of MH and OT coated gold NP were synthesized following a previously published [43] one-

phase method and the general procedure was as described below.

All the glassware was previously cleaned with aqua regia (3:1 HCl to HNO3), deionized water, acetone,

toluene, acetone, dried with N2 gas and finally rinsed with the respective reaction solvent

(dichloromethane for all particles except the 0:1 MH/OT which was rinsed using benzene).

In a 300 ml round bottom flask, 80 ml of the reaction solvent was stirred in a water bath held at 55°C in

which 496 mg (~1 mmol) of AuPPh3Cl was dissolved. Once totall y dissolved (10 min), the thiol -containing

ligands were added (see Table 1) and the solution was then stirred for 5 minutes before adding 870 mg

(~10 mmol) of the reducing agent, in this case a borane tert-butylamine complex. The reaction vessel

was then left for one hour and the condenser system was turned on.

After one hour the solution was allowed to cool at room temperature. The particles were precipitated

overnight at room temperature in 80 ml of Et2O. The supernatant was removed and the particles were

collected and washed in Et2O. This purification process was repeated until particles were fully

precipitated and the solution was clear. The remaining supernatant was then removed. The particles

were dried in air, weighed and stored for future use.

37 of 94

Table 1 Amounts of ligands used in the synthesis of NPs with various compositions

Ratio Volume (µl)1

(MH/OT) MH OT

1:0 273 0

5:1 227 58

2:1 181 116

1:1 136 174

1:2 91 231

1:5 45 290

0:1 0 347

1A 1:2 ligand to gold salt molar ratio was used in

the synthesis of the MH/OT coated particles.

Surfactants are added to the reaction vessel during nanoparticle formation in order to control the rate

of growth and limit aggregation [43]. The adsorption of surfactant-like molecules to nucleated

nanocrystals lowers the free energy of the surface and, therefore, the reactivity of the particles. The

ratio of surfactant to metal precursor can control the size distribution of the particles.

The use of amineborane complexes is essential for the syntheses of monodisperse metallic

nanoparticles because compared to other reducing agents such as sodium borohydride (NaBH4) and

lithium borohydride (LiBH4), amineborane complexes have a weaker reducing ability, which can slow the

reducing rate of gold cations and allow control over the growth of nanoparticles [43].

4.3. NP Characterization

In order to characterize the size distribution and composition of the nanoparticles a series of techniques

were applied: UV-Vis spectroscopy, transmission electron microscopy and nuclear magnetic resonance

spectroscopy respectively. Besides the transmission electron microscopy (TEM) imaging, all

characterization was performed in solution and thus was an ensemble average of all the present

structures.

4.3.1. UV-Visible absorption spectroscopy

UV-Vis spectroscopy measurements were carried out to rapidly estimate the NPs size by ensuring that

all the MPMNs had similar surface plasmon resonance (SPR) peaks .

Measurements were acquired using a PerkinElmer Lambda25 UV-Vis spectrophotometer operating at

room temperature and at a scanning rate of 240 nm.min-1

from 350 to 900 nm with 1 nm resolution.

38 of 94

UV-Vis spectra were prepared in disposable poly(methyl methacrylate) (PMMA) cells for the particles

dissolved in EtOH or a 1 cm path-length quartz cuvette for 0:1 MH/OT NPs dissolved in toluene. NPs

solutions were prepared by dissolving 3 mg of nanoparticles in 12 ml of absolute EtOH or toluene for the

OT homo NPs, sonicating for 20 min and stirring for 2 or more days before performing the UV-Vis

spectroscopy. The UV-Vis samples were prepared by diluting 500 µl of the NP solution in 500 µl of EtOH

(or toluene in the case of 0:1 MH/OT particles).

4.3.2. Transmission Electron Microscopy (TEM)

Characterization of particle size and size distribution was carried out using transmission electron

microscopy (TEM). 7 µl of 0.125 mg/ml nanoparticle solution were pipetted onto a 300 mesh carbon-

coated copper TEM grid, wicked with a KimWipe and allowed to dry in air. Microscopy was performed

on a JEOL 2010 (JEOL I) operating at an accelerating voltage of 200 kV. NP size distributions were

obtained by analyzing a minimum of 300 parti cles from several TEM images using ImageJ software and

were plotted as % of Frequency vs Size.

4.3.3. Nuclear Magnetic Resonance (NMR) Spectroscopy

1H NMR spectroscopy was performed on the NP solutions to determine the purity of the solution and

whether unbound ligands were present. The binding leads to a broadening of the peaks in the NMR

spectrum associated to the ligands (MH – R-CH2OH at 3.4 ppm; OT – R-CH3 at 0.70 ppm). Because the

analysis must be performed in deuterated solvents, the product must be carefully dried before this

analysis in order to prevent product loss [37]. The sample tubes were cleaned following the same

glassware procedure described for the MPMNs synthesis except in the final where they were rinsed with

δ3CD3OD (deuterated methanol), then 10 mg of NPs were dissolved in 0.6 ml of CD3OD. The NMR was

performed on a Varian 500 MHz NMR (Brucker).

4.4. NP Assembly

To characterize the assembly process two different techniques were used. To assess the state of

assembly based on the optical properties of the nanostructures UV-Vis analysis was performed. To

observe the structural morphology of these assembled structures , TEM imaging was carried out.

39 of 94

4.4.1. UV-Visible absorption spectroscopy

To assess the time of aggregation, UV-Vis spectroscopy measurements were carried out. 0.5 ml of a 0.25

mg/ml NPs solution were mixed with 0.5 ml of different cross-linker solutions: 1, 5 and 10 mM. The

samples were left to run in the spectrophotometer taking UV-Vis spectra over 1 hour period. Each

sample was mixed at the end of each 10 min. A marked change in the absorbance spectrum is indicative

of aggregation, and the time they take to stabilize after adding the li nker will be taking into account.

4.4.2. Transmission Electron Microscopy (TEM)

To image the gold nanoparticles assemblies with the addition of the cross -linker, a 7 µl drop with the

gold nanoparticles solution mixed with the cross -linker solution was pippetted onto the TEM grid. Some

parameters were varied during the experiments like the gold nanoparticles solution concentration (0.25

g/l, 0.05 g/l and 0.025 g/l), cross-linker concentration (10, 5, 1, 0.1 and 0.01 mM), time (0, 1, 5, 10, 1

hour, 12 hour, 24 hour time) and also cross-linker type molecules. A more detailed description is

provided in the next session.

40 of 94

5. Results and Discussion

5.1. Synthesis and Analysis of Monolayer-Protected Metal Nanoparticles

Seven types of MPMNs were synthesized using hydrophobic OT and hydrophilic MH as protective

ligands (see Figure 18). These capping agents were chosen due to their strong binding affinity to gold

surfaces because of the presence of a pending thiol group and their small head group size, which

prevents charged electrostatic interactions associated with large terminal groups of other typically used

ligands.

Figure 18 Chemica l structures of a) 1-octanethiol (OT) the hydrophobic ligand and b) 6-mercapto-1-hexanol (MH)

the hydrophilic ligand. MH/OT carbon chain ratio 6:8

The method, described in section 4.2, allowed the synthesis of nanoparticles coated with a homogenous

or heterogeneous SAM. The term homogeneous layers hereby refers to MPMNs functionalized with only

one type of thiol containing molecule (MH or OT), and heterogeneous particles to MPMNs coated with

various ratios of MH and OT. The terminology that will be used throughout this work is as follow: x:y

MH/OT, with x and y representing the molar ratios of MH and OT used during the synthesis of the

particles (i.e. 1:1 MH/OT implies 1:1 molar ratio of the MH and OT ligands; 1:0 MH/OT and 0:1 MH/OT

MPMNs represent gold NPs entirely coated with MH and OT molecules respectively). Seven series of

MH/OT MPMNs have been synthesized (1:0, 5:1, 2:1, 1:1, 1:2, 1:5 and 0:1 MH/OT respectively) and to

ensure some reproducibility, three different batch of each MPMNs synthesis have been produced.

By controlling the reaction temperature (55°C) and keeping the reagent ratios constant ( i .e. 1:2 l igand to

gold salt and 10:1 reducing agent to gold salt molar ratios) the synthesis produced MH /OT MPMNs with

relatively narrow size distributions.

Figures 19 and 20 present typical TEM images of the seven different MH/OT MPMNs sets. Histograms of

the respective core size distributions are depicted in Figures 21 and 22. They provide critical information

about the size dispersion of the samples. Average MH/OT MPMNs series sizes ranged from 4.2 ± 0.9 to

4.9 ± 0.8 nm (see Table 2). Core sizes and size distribution were obtained by measuring at least 300

particles from several TEM images for each of the MPMN sets.

41 of 94

Figure 19 Representative TEM images obta ined for the three batch of four different MH/OT MPMNs synthesized

showing the size of the gold core of the different particles. Scale bars - 20 nm.

42 of 94

Figure 20 Representative TEM images obta ined for the three batch of three different MH/OT MPMNs synthesized

showing the size of the gold core of the different particles. Scale bars - 20 nm.

43 of 94

Figure 21 Histograms obtained for the three batch of f ive different MH/OT MPMNs synthesized depicting core size

distributions. Each MPMN system exhibit relatively monodisperse s ize distributions. NP core s izes were obta ined by

analyzing at least 300 particles from different TEM images for each of the MPMN systems.

44 of 94

Figure 22 Histograms obtained for the three batch of two different MH/OT MPMNs synthes ized depicting core size

distributions. Each MPMN system exhibit relatively monodisperse s ize distributions. NP core s izes were obta ined by

analyzing at least 300 particles from different TEM images for each of the MPMN systems.

Table 2 Table of average core size ± Standard deviation for MH/OT MPMN series:

MPMN Core size (nm)

1:0 MH/OT 4.6 ± 0.8

5:1 MH/OT 4.7 ± 0.8

2:1 MH/OT 4.9 ± 0.9

1:1 MH/OT 4.4 ± 0.7

1:2 MH/OT 4.6 ± 0.8

1:5 MH/OT 4.2 ± 0.9

0:1 MH/OT 4.9 ± 0.8

Global average diameter 4.6 ± 0.8

Overall, the synthesis produced a relatively high yield of particles which ranged from 97.3 to 309.2 mg

which corresponds to efficiencies between 19.5 and 61.8% Table 3 shows the yield and efficiency data

for each of the MH/OT MPMN series.

45 of 94

Table 3 Total yield and efficiency of MPMNs synthesis.

Nanoparticles Yield (mg) Efficiency (%)

1:0 MH/OT

Batch A 271.5 54.3

Batch B 282.9 56.6

Batch C 309.2 61.8

5:1 MH/OT

Batch A 279.5 55.9

Batch B 258.4 51.7

Batch C 226.6 45.3

2:1 MH/OT

Batch A 124.6 24.9

Batch B 207.3 41.5

Batch C 153.5 30.7

1:1 MH/OT

Batch A 203.8 40.8

Batch B 175.8 35.2

Batch C 159.1 31.8

1:2 MH/OT

Batch A 118.1 23.6

Batch B 121.2 24.2

Batch C 166.9 33.4

1:5 MH/OT

Batch A 156.2 31.2

Batch B 206.3 41.3

Batch C 97.3 19.5

0:1 MH/OT

Batch A 185.8 37.2

Batch B 160.5 32.1

Batch C 166.0 33.2

2:1 MH/OT batch A were the first MPMNs synthesized which justifies the low yield obtained. Also a

decrease in the yield is noticeable for the particles with a higher ratio in the hydrophobic (OT) l igand.

This could be due to a loss of particles during the washing steps of the synthesis protocol or as a

consequence of OT solubility in EtOH.

UV-Vis absorption spectroscopy was used in the characterization of MPMNs. These particles SPR band

(and thus their colour in solution) is characteristic of their nature, size and shape, and also depends on

the refractive index of the surrounding medium and on the nature of their protective layer. Typically,

particles smaller than ca. 2 nm do not show any plasmon peak at all [7]. Thus, UV-Vis spectroscopy of

the prepared MPMN suspensions was carried out; the results are depicted in Figure 23. Surface plasmon

peaks are fairly similar with a maximum absorbance ranging from 504.2 to 514.0 nm. Due to limited

solubility of the different MPMN seri es, UV/Vis measurements were performed after at least a 2 days

settling period to enable slow floculation of the insoluble NPs, leaving in solution only the colloidaly

stable ones. As explained in section 2.4.1, the solubility of mixed ligand shell MPMNs (mixtures of

hydrophilic (MH) and hydrophobic (OT) l igands) does not increase linearly when increasing the fraction

of hydrophilic l igand in the ligand shell in a hydrophilic solvent such as EtOH. Since the intensity of the

46 of 94

peak of the plasmon band is directly proportional to the particles concentration and can be related to

the critical colloidal solubility, MPMNs solubility in EtOH was ranked as follows: 2:1 MH/OT (100%); 1:2

MH/OT (95 %); 1:5 MH/OT (81 %); 1:1 MH/OT (71 %); 1:0 MH/OT (56 %); 5:1 MH/OT (53 %). The relative

solubilities reported here have been normalized to highest intensity of the plasmon peak values: 2:1

MH/OT – 1.391. (0:1 MH/OT MPMNs were prepared in toluene since they are not soluble in EtOH so

they have not been considered for these calculations). Since all the NPs are roughly the same size the

variations in plasmon intensity at the given concentration have not been considered.

Figure 23 UV-Vis absorption spectra of MPMNs suspended in EtOH (or toluene in the case of the OT Homo NPs) at

an initia l concentration of 0.125 mg/ml. For the MPMNs dissolved in EtOH, solubility follows a non-monotonic trend

with composition.

Table 4 Plasmon peak values and intensities for each of the MH/OT MPMNs series.

MPMN Plasmon peak Plasmon peak intensity

1:0 MH/OT 507.0 ± 3.0 0.81 ± 0.03

5:1 MH/OT 510.1 ± 3.2 0.77 ± 0.07

2:1 MH/OT 512.9 ± 5.1 1.46 ± 0.24

1:1 MH/OT 507.9 ± 4.6 1.03 ± 0.16

1:2 MH/OT 508.9 ± 6.7 1.39 ± 0.26

1:5 MH/OT 504.2 ± 5.6 1.19 ± 0.09

0:1 MH/OT 514.0 ± 2.0 0.20 ± 0.07

1H NMR analysis confirmed the presence of unbound ligands, denoted by the presence of a sharp peak

in the NMR spectrum associated with the proton from the pending thiol of the ligand (MH – R-CH2OH at

3.4 ppm) however the absence of a peak at 0.70 ppm confirms the absence of OT in solution [61]. This

excess of thiols in solution will have repercussions in the MPMNs stability in solution as discussed in

below (see section 5.3.2.5). Figure 24 shows a representative spectrum and respective analysis obtained

47 of 94

for one of the sets of MPMNs (2:1 MH/OT Batch B). Measurements were done for all MPMNs,

respective spectra and analysis can be seen in Appendix .

Figure 24 1H NMR spectrum obta ined for 2:1 MH/OT Batch B in CD3OD. The presence of a peak in the NMR

spectrum associated to the the hydrophilic ligand (MH at 3.4 ppm) shows the presence of free ligands in solution.

In order to achieve information on the structure of the ligand shells, Amplitude modulation – Atomic

Force Microscopy (AM – AFM) (performed and analysed by Dr. Kislon Voitchovsky at MIT) was

performed on the 2:1 MH: OT Batch A fi lm. When high resolution was possible (see Figure 25) MPMNs

appeared normal and rough which citing Prof. Stellacci means that the few isolated particles present in

the fi lm are indeed striped. A more detailed analysis was not possible since the fi lms were co mposed of

~30-40 nm NPs aggregates which make it difficult to achieve a high imaging quality mandatory for a

proper analysis of the ligand shell stripes and their inter -distance.

Figure 25 Film of MH:OT 2:1 nanoparticles imaged by AM-AFM in ultrapure water. Right: XY: 40 nm, Z: 11 nm; Left:

Scale bar: 5 nm

48 of 94

AM-AFM was not performed for the other MPMNs series however morphological differences are

expected between homogeneous and heterogeneous-ligand particles shells. Pure MH and OT NPs were

assumed to have homogeneous hydrophilic and hydrophobic ligand shells. The 1:2, 1:1 and 2:1 MH/OT

particles displayed classical ripples which are attributed to hydrophilic/hydrophobic striated ligand shell s

(schematics of l igand shell structure are presented on Figure 14, section 3). The 1:5 and 5:1 MH/OT

MPMN were assumed to have homogeneous ligand structures, interspersed with discrete domains of

the minority l igand.

5.2. MPMNs Molarity

In order to determine the best gold NPs and cross -linker concentrations to start the MPMNs assembly

study and vary these parameters in a controlled way, the particles molarity has to be determined. The

next section shows how we have assessed our MPMNs and cross -linker solutions molarity.

5.2.1. Number of Gold Atoms:

The gold atoms that form the gold NPs, are strongly and tightly packed, with each of the constituent

atom being surrounded by twelve neighbouring atoms [7]. Therefore, the smallest cluster has thirteen

atoms and the following contain 10 n2

+ 2 atoms, where n represents the layer number. The increase of

the total number of atoms with the number of layers is represented in figure 26 [7].

Figure 26 Graph showing the relation between the number of layers of a gold NP and the tota l number of gold

atoms.

0

500

1000

1500

2000

2500

3000

3500

4000

4500

1 2 3 4 5 6 7 8 9 10 11

n of layers

n of layer atoms

total n of gold atoms

49 of 94

A facetted NP is considered a topological sphere [7] thus, in theory, the total number of gold atoms, NAu,

existent in a NP can be calculated knowing the average diameter, d, of the core. Of course, the presence

of defects will allow NPs with an intermediate number of atoms to exist (i .e. not following the 10 n2 + 2

distribution).

Where VAu is the gold NP volume.

As specified on table 2, the average diameter (d) for all the considered NPs is 4.6 nm, We should note

that in the TEM images we can only see the gold core shell, so the ligand shell will not be considered.

The grafting density of thiols (Nthiols) bound to the NP surface depends on the surface area and t he space

occupied per each thiol on it (see Eq. 7)

Where the surface area can be calculated if we have the average diameter value (see Eq. 8) and the

space occupied per thiol in the surface of the NP which is known to be 22 Å2 [124].

Table 5 synthesises the results obtained:

Table 5 Estimated gold clusters characteristic dimensions.

d (nm) 4.6

VAu (nm3) 0.51

NAu 2970

Surface Area (Å2) 6647.6

Nthiols 302

Once estimated the total number of gold atoms (NAu), 2970, it was possible to estimate the molecular

weight of the five different sets of MPMNs. Then, considering the final volume and concentration of 12

ml and 0.25 mg/ml respectively, the total number of MPMNs in solution can be determined. The results

are presented in the table below:

50 of 94

Table 6 Linear formula, molecular weight, molar concentration and average number of nanoparticles per gold

cluster (NNPs) for each set of MPMNs.

MPMNs Linear Formula Molecular

Weight (g/mol)

Molar concentration

(M) NNPs

5:1 MH/OT Au2 9 7 0(HS(CH2)6OH)25 1,7(CH3(CH2)6CH2SH)50,3 6.26E+05

3.99E-07 2.88E+15

2:1 MH/OT Au2 9 7 0(HS(CH2)6OH)20 1,3(CH3(CH2)6CH2SH)100,7 6.27E+05

1:1 MH/OT Au2 9 7 0(HS(CH2)6OH)1 5 1(CH3(CH2)6CH2SH)1 5 1 6.27E+05

1:2 MH/OT Au2 9 7 0(HS(CH2)6OH)10 0,7(CH3(CH2)6CH2SH)201,3 6.28E+05

1:5 MH/OT Au2 9 7 0(HS(CH2)6OH)50 ,3(CH3(CH2)6CH2SH)251,7 6.29E+05

Average 6.27E+05

Taking in consideration our TEM grid, it is a fair estimate to consider that of the 7 µl MPMNs solution

drop only 1 µl (0.125 mg/ml) stays in the carbon grid, and that at a 60 K magnification there is an

average of 300 particles per TEM image.

Knowing this and the details provided in the following table one can d etermine the average number of

MPMNs per TEM grid following two different ways: a) considering the molar concentration, the volume

that stays in the TEM grid (1 µl) and the Avogadro number; and b) the average number of particles per

image; Thus the values can be compared in order to verify the consistency of the approximations used

(a) molar concentration and final volume on the grid and b) number of particles per image).

Table 7 TEM grid and TEM image considerations.

LTEM image (nm) 150

ATEM image (m2) 2.25E-14

dTEM grid (mm) 3

ATEM grid (m2) 7.07E-6

NP/ATEM image ~300

NP/ATEM grid1

9.5E+10

NP/ATEM grid2 1.2E+11

1Considering an average of 300 particles per TEM image 2Considering that only 1 µl (0.125 mg/ml) stays in the TEM grid

Where LTEM image is the square side of the TEM image, ATEM image is the observable area of the TEM image,

dTEM grid and ATEM grid represent the carbon grid’s total diameter and area respectively. NP/ATEM image and

NP/ATEM grid correspond to the total number of nanoparticles per TEM image area and TEM grid area

respectively.

Both values are in the same order of magnitude which indicates the approximation taken is valid and

significant.

51 of 94

Next, the ratio of cross-linker molecules used was determined. Two types of dithiol molecules (1,9-

nonanedithiol (NDT) and 1,16-hexadecanedithiol (HDDT)) were used at different stages of the work.

Their structure and characteristics are presented in the table below.

Table 8 1,9-nonanedithiol and 1,16-hexadecanedithiol chemical structure and molecular weight

1,9 – nonanedithiol HS(CH2)9SH

MW (g/mol) 192.4

1,16 – hexadecanedithiol HSCH2(CH2)1 4CH2SH

MW (g/mol) 290.6

Considering a 0.5 ml volume of cross -linker solution and the different molar concentrations used C (C=

xM, yM, ...) the total number of mols (N) and the number of cross -linker molecules (Ncross-linker) can be

calculated in the respective 0.5 ml solutions. The results are summarized in table 9.

Table 9 Number of mols (N) and number of cross-linker molecules (Ncross-linker) for different molar concentrations

considering a final volume of 0.5 ml of cross-linker solution.

C (mM) N (mol) Ncross-linker

10 5.0E-06 3.0E+18

5 2.5E-06 1.5E+18

1 5.0E-07 3.0E+17

0.1 5.0E-08 3.0E+16

0.01 5.0E-09 3.0E+15

During the TEM analysis two different concentrations (c) of MPMN solution were used (0.25 and 0.05

mg/ml). Considering a 0.5 ml volume for the NPs solution, the molar concentration (C), the number of

mols (N) the total number of NPs and thiols (NNanoparticles and Nthiols respectively) in solution can be

determined for each of the MPMN concentration:

Table 10 Molar concentration (C), number of mols (N) and number of NPs and thiols (NNanoparticles and Nthiols

respectively) correspondent to two different MPMNs final concentrations, considering a final MPMNs solution

volume of 0.5 ml.

c (mg/ml) C (M) N (mol) N Nanoparticles Nthiols

0.25 4.0E-07 2.0E-10 1.2E+14 3.6E+16

0.05 7.97E-08 4.0E-11 2.4E+13 7.2E+15

52 of 94

Finally, knowing the total number of cross -linker molecules present in solution as well as the total

number of NPs and number of thiols in the ligand shell available to place exchange we can determine

the ratio of cross-linker molecules per a) each gold NP and b) each thiol bound to the gold NP surface:

Table 11 Cross-linker to gold NP ratio and cross-linker to thiol ratio for five different cross-linker molar

concentrations.

Ccross-linker

(mM)

Cross-linker to gold NP ratio Cross-linker to thiol ratio

0.25 mg/ml 0.05 mg/ml 0.25 mg/ml 0.05 mg/ml

10 2.50E+04 1.25E+05 8.33E+01 4.17E+02

5 1.25E+04 6.25E+04 4.17E+01 2.08E+02

1 2.50E+03 1.25E+04 8.33E+00 4.17E+01

0.1 2.50E+02 1.25E+03 8.33E-01 4.17E+00

0.01 2.50E+01 1.25E+02 8.33E-02 4.17E-01

5.3. Cross-linking of MPMNs in solution

5.3.1. UV-Visible Analysis

Since it is impossible to know exactly the number of NPs in solution (the molecular weight calculations

represent an estimation based on the TEM imaging) achieving the right cross-linker to gold NPs ratio is

very challenging. Having this in consideration for a first experimentation of the cross -linker effect, three

solutions with different molar concentrations of the linker molecules, 1,9 – nonanedithiol (NDT) were

added to all, but 1:0 and 0:1 MH/OT, NPs series and aggregation was observed for all types of MPMNs.

Soon after adding the cross-linker solution an insoluble black precipitate began to form and after one to

two hours the EtOH solution had lost most of its colour and the precipitation process reached

completion. These precipitates are indicative of big insoluble NPs agglomerates formation.

The extent and the kinetics of these processes can be studied with UV-Vis spectroscopy and are

reported on Figures 27 and 28 where the change of absorbance for the aggregation in the 450-900 nm

region is shown.

Figure 27 shows the changes observed in the UV-Vis spectra when different concentrations of dithiol

were added.

53 of 94

Figure 27 Change of absorbance in the 450-900 nm reg ion of UV-Vis spectra for 1:1 MH/OT Batch A MPMN solution

upon addition of (a) 1 mM, (b) 5 mM and (c) 10 mM solution of cross-linker respectively. A dithiol alkane chain was

added to the solution of MPMNs causing the particles to link together as a place exchange reaction was taking

place. The legend represents a timecourse from 0 to 60 minutes after addition of cross-linker, curves were taken at

6 minutes intervals.

Following the cross-linker addition, the intensity of the SP band starts to decrease and a broadening of

the SPR band is observed. Furthermore the SP band maximum also slightly shifts to longer wavelengths

which is associated with the formation of NP clusters. However UV-Vis results are inconclusive in

determining whether this shift is caused by the formation of defined structures or non-specific

aggregation, a proper analysis is only achievable with the TEM (see section 5.3.2)

Figure 28 compares the different behaviours between the remaining particl es (5:1, 2:1, 1:2 and 1:5

MH/OT Batch A, respectively) upon addition of the cross-linker solution (5 mM). (The UV-Vis

measurements and analysis were repeated for all batches and sets of MPMNs for the three different

concentrations of cross-linker although here only Batch A is represented and for a representative

concentration of cross-linker, 5 mM)

54 of 94

Figure 28 Change of absorbance in the 450-900 nm reg ion of UV/Vis spectra for (a) 5:1, (b) 2:1, (c) 1:2 and (d) 1:5

MH/OT Batch A MPMNs upon addition of 5 mM of 1,9-nonanedithiol solution respectively. A dithiol a lkane chain

was added to the solution of MPMNs caus ing the particles to link together as a place exchange reaction was taking

place. The legend represents a timecourse from 0 to 60 minutes after addition of cross -linker, the curves were

taken at 6 minutes intervals. The arrows point to the SPR peak, 0 and 60 minutes after adding the cross-linker

molecules and in a) and d) a smooth appearance of a second peak is also signed.

All graphs show MPMNs aggregate with time, however no clear difference between the three

concentrations of 1,9-nonanedithiol tested could be observed. On the other hand each set of these

particles shows a different behaviour.

2:1, and 1:2 MH/OT exhibit a straighter peak at 0 min (Figure 28, b) and c)) while 5:1 and 1:5 MH/OT

show a broader peak (Figure 28, a) and d)). For 2:1 and 1:2 MH/OT particles one can only denote a

decrease in the peak intensity whereas along with that 5:1 and 1:5 MH/OT particles also show the arise

of a flatter curve. In fact for the 1:5 MH/OT particles at the end of one hour the plasmon peak almost

completely disappeared. It is also clear there is a shift of the SP band maximum to higher wavelengths

in the case of 5:1 MH/OT particles. It is also noteworthy the smooth appearance of a second peak in the

case of 5:1 and 1:5 MH/OT particles (Figure 28 a) and d)). These results meet the established theoretical

descriptions of Mie scattering for aggregates. This theory suggests that if the SPR absorption peak is still

present in the absorption spectrum of the cross -linked solution, small aggregate clusters of similar sizes

55 of 94

are probably forming. On the other hand, if the plasmon band completely disappears and a flatter

spectrum forms, aggregates are probably bigger and have a broad size distribution [125, 126].

The SPR peak variations with time for each set of MPMNs upon the addition of a cross -linker solution

are illustrated in Figure 29. A gradual increase of the SPR maximum value with increasing MH

concentrations is observed with this trend being clearer for the 5:1 MH/OT set.

Figure 29 SPR peak shifts for all the batches of the 5 sets of MPMNs tested at different times. Dark blue line: SPR

peak of MPMNs in EtOH; Medium and light blue lines: SPR peak of MPMNs solution measured immediately and 30

minutes after adding the 1,9-nonanedithiol solution, respectively; a) 1 mM b) 5 mM and c) 10 mM 1,9-

nonanedithiol concentration. The Figure clearly demonstrates a dependence on the ligand shell morphology. When

the ratio of hydrophilic ligand increases, more the SPR band shifts.

The advancement of the aggregation process was also followed with the naked eye as the solution

gradually changed from a light pink-red colour to a browner one, and finally, upon standing, to a dark

precipitate.

Since the differences between the five distinct MPMN solutions were not too evident (as shown in

Figures 27 and 28 (a-d)) we opted to measure the stabilization time (Figure 30), which corresponds to

the time the difference between the curves at 800 nm takes to stabilize, i .e. reach zero. The stabilization

time indicates that the place exchange reaction rate is slowing down and the reaction is reaching the

equilibrium state.

56 of 94

Figure 30 Stabilization time, i.e. time, the difference between the curves at 800 nm along time, takes to stabilize

(reach zero), reaching the equilibrium.

Once again we can verify that the stabilization time is also dependent on the ligand shell composition

and morphology. Increasing MH concentration, the hydrophilic ligand, in the monolayer shell

composition appears to result in slower the stabilization time. This trend can arguably be explained

considering that the hydrophilic character of the ligand confers more stability in solution, prolonging the

place-exchange reaction. However, l igand place-exchange reactions are dependent on many

parameters, such as the ligands relative affinities to gold surface, their size and their solubility. Thus, the

detailed mechanisms of the ligand place exchange could not be fully rationalized.

5.3.2. TEM Analysis

After studying the optical properties of the different MPMNs assembl ies, their morphology was

examined by TEM. During the TEM analysis, several approaches have been used. TEM data is hereby

presented according to the different variables studied, i .e. colloidal suspension’s stability, time, cross-

linker concentration and cross-linker type.

5.3.2.1. Hypothesis

TEM analysis was based on several assumptions based on the different behaviours and self -assembly

driving forces for each set of MPMNs.

DeVries et al. previously reported that when both the ligands are present in the same ratio in the ligand

shell, the resulting SAM will phase-separate into ordered (rippled) domains and therefore will present

fewer stable molecules in the poles. These ligands being not optimally stabilized by in termolecular

interactions will be the first molecules to be replaced in place -exchange reactions. Thus after

57 of 94

functionalizing these specific points, divalent NPs can, in theory, be generated and linear chains ranging

from three to twenty MPMNs in length wil l form (Figure 31) [91].

Figure 31 Observed assembly of the 1:1 MH/OT gold MPMNs. The unstable polar defects will be place-exchanged

with the dithiol molecules in solution originating linear assemblies. TEM image reproduced from [91].Scale bar: 50

nm.

In the case of 1:2 and 2:1 MH/OT the polar singularities will react faster than other defects that might

exist in the ligand shell. However these diametrically opposed polar defects are in this case more

stabilized by intermolecular interactions with their neighbouring counterparts which results in still l inear

but smaller chains with occasional branched structures (see figure 32)

Figure 32 Predicted structures of the assembled 2:1 and 1:2 MH/OT gold MPMNs, respectively. The existence of less

reactive poles will be the driving force for the formation of linear but a few branched structures. Scale bar: 20 nm

The remaining sets of particles, 5:1 and 1:5 MH/OT gold MPMNs supposed to have an almost-

homogenous ligand structure with discrete domains of the minority component are expected to form

unordered three-dimensional assemblies (Figure 33) since the linker molecules will place-exchange

randomly or at the occasional defect and pinhole sites that might exist anywhere in the ligand shell.

58 of 94

Figure 33 Expected structures of the assembled 5:1 and 1:5 MH/OT gold MPMNs, respectively. Instead of ripples,

the lesser abundant ligand will form discrete domains in the monolayer. There are no polar defects in these types of

ligand arrangements so the cross-linker molecules will bond to the meta llic surface randomly or at the occasional

defects that might exist so unordered and three-dimensional assemblies are expected. Scale bar: 10 nm

Considering their l igands ratio, 2:1 and 5:1 MH/OT series of MPMNs present a SAM with less amount of

OT, which is the longer molecule. According to Hostetler et al. studies, the rate of l igand exchange

decreases when the chain length of the protecting monolayer increases, because, according to the

previously explained associative mechanism it will be more difficult for the incoming molecules (dithiol)

to penetrate the monolayer and undertake place-exchange [67]. In this case the more abundant l igand

in the SAM is the MH (the smaller molecule in length) which favors the place exchange with the

incoming dithiol resulting in a quicker process and bigger structures.

5.3.2.2. TEM Sample preparation

Imaging of self-assembled NPs can be challenging, thus, TEM sample preparation is established

according to several parameters:

1. If the NPs solution is too concentrated, the density of NPs in the grids will be too high and will

make it difficult to discern whether the structures formed were in fact a result of the cross -

linker action or formed by interparticle interactions upon drying.

2. If the NPs solution is allowed to sit for too long on the TEM grid, it will be possible for the

MPMNs in solution to arrange or re-arrange themselves into ordered or random structures in

the TEM grid.

3. The solvent used must be pure and miscible to avoid artifacts that are a result from the

nanoscale bubbles of phase separation (see figure 39 section 5.3.2.3.1).

For these reasons the solution was cast onto a carbon coater copper mesh grid and placed on a clinical

tissue (KimWipeTM

) in order to rapidly wick away the excess of solvent and place the MPMNs onto the

59 of 94

TEM grid in the same arrangement as they were in solution, preventing the formation of unspecific

aggregation patterns that might artificially cause size selection and affect the size distribution results.

5.3.2.3. Stability of particles in solution

5.3.2.3.1. Control experiments

Throughout the project several control experiments under different conditions have been tested. The

control experiments were made in order to find out the agglomeration state of the nanoparticles in

solution along time without the presence of cross -linker molecules. This was determinant to conduct the

work and to follow the approaches taken during this project.

Therefore, TEM grids were prepared with NP solutions in the absence of cross-linker as controls.

Representative images are shown below:

Figure 34 TEM images of 2:1 MH/OT batch A MPMNs solution (0.125 mg/ml). Scale bars: 100 nm

A general view over the TEM grid confirmed the presence of several small assemblies everywhere on the

grid. This was the first indication that, so far, suggests these particles are not stable in solution and are

spontaneously self-assembling.

In an attempt to eliminate the spontaneously formed small aggregates of NPs, s olutions were then

sonicated for 1 hour. The samples were then left to stir for a period of 24 hours, to let the eventual

persisting agglomerates to sediment. However, no significant improvement could be observed on the

TEM pictures, and small aggregates were still visible (Figure 35).

60 of 94

Figure 35 TEM images of 2:1 MH/OT batch A MPMNs solution (0.125 mg/ml). Grid prepared after sonicating the

sample for 1 hour period. Scale bars: 100 nm

Based on previous results, fresh solutions of nanoparticles were prepared. They were sonicated for two

hours and stirred for at least two days. Then, in order to effectively eliminate or at least reduce the

number and size of the spontaneous agglomerates, the solution was purified by fi ltration-centrifugation

using a syringe fi lter (ANATOP – 0.1 µm pore size at 15 000 rpm for 8 minutes).

Fresh solutions of 2:1, 1:1 and 5:1 MH/OT Batch B were thus diluted (1:10 to 0.025 mg/mL), fi ltered and

centrifuged. The TEM grids were prepared 2 hours after the purification for the first two types of

MPMNs (2:1 and 1:1 MH/OT). In the case of the 5:1 MH/OT MPMNs the solution was left standing for 3

days and the TEM grids were prepared at the end of the third day after the purification steps.

A general look over the TEM grid of the first two sets of NPs revealed that there were no aggregates in

solution (see figure 36). However samples analysed three days after purification a lready started to show

signs of aggregation, once again suggesting that the synthesized NPs were spontaneously self-

assembling in solution with their neighbouring particles (figure 37). Nevertheless the different SAM

shells of the particles should also be considered as a reason for the spontaneous self-assembly of the 5:1

MH/OT molecules.

Figure 36 TEM images of (a) 2:1 and (b) 1:1 MH/OT batch B MPMNs solutions (0.025 mg/ml). Grids prepared two

hours after filtering and centrifuging the MPMNs fresh solutions. Scale bars: 50 nm.

61 of 94

Figure 37 TEM pictures of the 5:1 MH/OT batch B MPMNs solution (0.025 mg/ml). Grids prepared 3 days after the

filtration and centrifugation steps. Scale bars: 50 nm

In order to study the time-dependence of the aggregation process, a diluted batch of 1:1 MH/OT coated

particles (1:5 – 0.05 mg/ml) were fi ltered and centrifuged at different times (see table 12).

Table 12 Different amounts of time samples were left stirring after the purification steps and prior to TEM grid

preparation. Each figure letter corresponds to a figure, presented below, and represents the images obtained for

the different solutions of 1:1 MH/OT.

figure Settling time

38 1 hour

39 5 hours

40 Left - 3 days

Right - 14 days

41 5 days

62 of 94

Figure 38 TEM pictures of the 1:1 MH/OT Batch B MPMNs solution f iltered and centrifuged 1 hour before the TEM

grid preparation. Upper images: 0.125 mg/ml; down images: 0.05 mg/ml. Scale bars: 50 nm.

Figure 39 TEM pictures of the 1:1 MH/OT Batch B MPMNs solution (0.125 mg/ml) filtrated and centrifuged 5 hours

before the TEM grid preparation. Scale bars: 50 nm.

63 of 94

Figure 40 TEM pictures of the 1:1 MH/OT Batch B MPMNs solution (0.05 mg/ml) f iltered and centrifuged 3 and 14

days before the TEM grid preparation, left and right images, respectively. Scale bars: 50 nm.

Figure 41 TEM images of the 1:1 MH/OT Batch B MPMNs solution (0.05 mg/ml) f iltered and 5 days before the TEM

grid preparation. Scale bars: 50 nm.

TEM grids prepared hours after the purification steps present a considerable number of small l inear

assemblies (figure 38), which is interesting considering that these are the rippled particles. After three

days two and small three-dimensional structures could be observed (figure 39) however on the grids

prepared using “older” solutions no defined structure could be imaged. Those results are not consistent

with the previous results (figures 34 to 37). Working with different types of MPMNs that have different

ligand shell morphologies can make them have occasionally unexpected behaviours towards solvents. As

an example, the nanoscale rings highlighted in figure 39 could be explained as (a) a result of holes

opening up in the liquid film and pushing particles into their borders [127], (b) a result of water droplets

which condense on the surface of nonpolar solvents from humid air [128] or (c) impurities in the

solvent, such as other l iquids droplets.

Several groups have focused on the study of the induced and spontaneous assembly of nanoparticles in

solution. Sidhaye et. al found that the size of the three dimensional lattices spontaneously formed in

solution depend on the chain length of the alkanethiol molecules that wrap around the NPs, i .e. when

the chain length of the ligands that cap the NP metal core increases, the interparticle attractive forces

decrease and so decreases the size and overlapping of the structures. [129]. The thiol-containing

64 of 94

molecules used in the present project are relatively short which can explain the formation of these big

structures.

Sidhaye et.al also proved that the formation of three-dimensional structures was only possible in the

presence of excess thiol molecules in solution, which are in dynamic equilibrium with those

chemisorbed onto the NP surface. Our NMR analysis revealed the presence of unreacted MH in solution

which can explain the instability of MPMNs in solution [67, 129].

5.3.2.3.2. UV-Vis study

To test and confirm the MPMNs stability in solution over time, a solution of NPs was prepared as

described previously, sonicated for 20 minutes and left to stand for a period of twelve weeks. SPR peak

evolution was recorded periodically by UV-Vis spectrophotometry during this time interval. The results

are summarized in the graph presented below:

Figure 42 UV-Vis absorption spectra of 1:1 MH/OT Batch B MPMNs dissolved in EtOH at an initial concentration of

0.125 mg/ml. The sample was kept untouched during a period of twelve weeks, and during this time interval several

UV-Vis samples were taken periodically (2, 4, 6, 8 and 12 weeks after, respectively). A decrease in the intensity and

a slight shift of the SPR peak can be observed.

65 of 94

Table 13 Plasmon peak wavelengths and intensity variations with time for MPMNs solution (1:1 MH/OT Batch B –

0.125 mg/ml). This solution was kept untouched for twelve weeks period after a 20 minute - sonication step.

Analysis date Plasmon Peak

Plasmon peak intensity

12th Nov 2009 508 0.88

26th Nov 2009 508 0.85

10th Dec 2009 509 0.81

21s t Dec 2009 510 0.76

15th

Jan 2010 512 0.66

As shown in Table 13 and Figure 42, there is a considerable decrease of the SPR peak intensities as well

as a slight shift of the SPR peak value, suggesting slow MPMN aggregation as described in section 5.3.1.

When particles aggregate they tend to sediment at the bottom leading to a reduction in the number of

particles in solution and thus a decrease in the peak intensity. This parallel study was important to

confirm the image-based finding that this type of particles were self-assembling in the ethanolic

solution.

5.3.2.3.3. Solvent Dependence

MPMNs solubility was also studied using other solvents of various hydrophobicity (toluene,

dichloromethane (DCM) and methanol ). None but methanol confers stability to the particles (Figure 43).

All the MPMNs series are soluble in methanol and seem to be more st able than their counterparts in

EtOH in the first hours after solution preparation. However an increase in the number of smaller

aggregates (20-50 nm) in solutions aged two days was observed.

Figure 43 1:1 MH/OT Batch A in Methanol a) 2 and b) 48 hours after solution preparation. a) When the TEM sample

was prepared 2 hours after preparing the solution a homogeneous grid was observed. b) 48 hours after preparing

the MPMNs solution the grid presented a considerable number aggregates approximately 50 nm in size. The TEM

analysis was repeated a lso for the 2:1 and 5:1 MH/OT particles ( images not shown) revealing that both these

particles were stable in methanol. Scale bars: 20 nm

66 of 94

5.3.2.4. Assembly variations with time

In the first approach followed the assembly process was explored and analyzed at different times.

1 hour after adding the cross-linker solution - i=6

Initially TEM grids prepared 1 hour after adding the cross -linker molecules (5 mM solution of NDT in

EtOH) to the MPMN solutions have been analyzed. Representative results are presented in the images

below:

Figure 44 TEM images of 1:1 MH/OT Batch A MPMNs solution (0.125 mg/ml) 1 hour ( i=6) after adding the cross -

linker, NDT. Ba ll-type aggregates are evidenced in the images. These ball-type aggregates were further cross-linked

due to a large excess of time and dithiol molecules or because of phase-segregation due to poor solubility. Scale

bars: 50 nm

Figure 45 TEM images of 1:2 MH/OT Batch A MPMNs solution (0.125 mg/ml) 1 hour ( i=6) after adding the cross -

linker, NDT. Small ball-type aggregates are evidenced in the images. Scale bars: 50 nm

67 of 94

Figure 46 TEM images of 1:2 MH/OT Batch B MPMNs solution (0.125 mg/ml) 1 hour ( i=6) after adding the cross-

linker, NDT. Small ball-type aggregates further assembled are shown in the images. Arrows indicate the magnified

zones. Scale bars: 50 nm

Figure 47 TEM images of 5:1 MH/OT Batch A MPMNs solution (0.125 mg/ml) 1 hour after adding the cross-linker,

NDT. Ball-type structures are linked together forming larger structures. Arrows indicate the magnified structures.

Scale bars: 50 nm

TEM images show that one hour after adding the cross -linker solution large assemblies were formed

suggesting that many NPs have dithiol molecules at random sites in the ligand shell. These big structures

appear to have a somewhat chain-like character and interestingly they all seem to form large networks

of ball -type assemblies that link together originating chains. This is more evident on figures 44 and 47.

These spherical assemblies may be a result of (a) interparticle hydrogen bonding between the

hydrophilic alcohol termini of the ligands (so it depends on the MH/OT ratio in the ligand shell), (b) they

may be composed of long chains that are more stable when coiled together than when extended in an

all-trans configuration or (c) may be a result of the fully encapsulation of t he gold NPs assemblies by

dithiols. These ball -type aggregates are likely to be further cross -linked because of the large excess of

l inker molecules.

Similar results have been achieved by Hussain et. al. In their work they have obtained uniformly sized

spherical assemblies of gold colloids in toluene by adding alkanedithiol solutions in different gold NPs to

dithiol molar ratios [102]. Moreover, they were able to further organize these ball -type assemblies into

relatively l inear chains through the addition of EtOH.

1:2 MH/OT Batch A appears to behave differently when the cross -linker is added as no chains or big

assemblies were veri fied one hour after adding the linker solution. Instead small isolated circular

68 of 94

assemblies were formed (figure 45). This resistant behaviour of the mentioned particles towards self-

assembly was also previously predicted by UV-Vis analysis (see Appendix). Batch B of the same particles

was also analysed confirming the presence of l inear chains of ball -like aggregates.

Overall, TEM analysis strongly suggests that NDT molecules are efficiently cross-linking the NPs.

However, focusing on the final aim of this project, how the different MPMNs start the assembly process,

it is necessary to observe intermediate structures. In order to try to assess this stage of the assembly

process TEM grids have been prepared one minute after pippeting the cross-linker solution. The 5 mM

solution of NDT was mixed with the 1:1 and 5:1 MH/OT Batch A and 1:2 MH/OT Batch B. Analysis of 1:2

MH/OT TEM grids was repeated two weeks later to check if the NPs stay stable and in the same

arrangement over time on the TEM grids. Representative images are presented below.

1 minute after adding the cross-linker solution - i=0

Figure 48 TEM images of (a) 1:1 A and (b) 1:2 B MH/OT MPMNs solutions (0.125 mg/ml) one minute ( i=0) after

adding the cross-linker (NDT) solution. Three-dimens ional structures are still v isible especia lly on f igure (a). Scale

bars: 50 nm

a)

b)

69 of 94

Figure 49 TEM images of (a) 1:2 B (same grid as f igure 48 (b) but analysed two weeks after), (b) 2:1 A and (c) 5:1 A

MH/OT MPMNs solutions (0.125 mg/ml) one minute (i=0) after adding the cross-linker (NDT) solution. Three-

dimensional structures are still visible especially on figures (a) and (b). Scale bars: 50 nm

One minute after adding the cross-linker solution, there were already assemblies and small clusters of

MPMNs present in the TEM grid, except in the case of the 1:2 MH/OT batch B where only occasional two

and one-dimensional structures were seen. The large number of two and three-dimensional aggregates

is probably a result of the high degree of interdigitation which can be caused by the proximity of the

particles due to high concentrations (0.125 mg/ml).

a)

b)

c)

70 of 94

5.3.2.5. Assembling behaviours with cross-linker concentration

Besides the specific chemical reactivity or the cross -linker structure, another important parameter that

determines the way the assembly process develops is the linker and NP relative molar concentrations.

Initially the assembly process was too fast, precluding the view of intermediate structures in the TEM

grids, so the cross-linker to gold NPs molar ratio was reduced. Instead of 12500 NDT/gold NP (5 mM) the

MPMN samples were tested with a less concentrated solution of cross-linker molecules (1 and 0.5 –

2500 and 1250 NDT/gold NP, respectively (see table 14)).

Table 14 Table with the calculated linker to particle and linker to thiol ratio values

Ccross-linker

(mM)

Cross-linker to gold NP ratio Cross-linker to thiol ratio

0.25 mg/ml 0.25 mg/ml

5 1.25E+04 4.17E+01

1 2.50E+03 8.33E+00

0.5 1.25E+03 4.17E+00

0.1 2.50E+02 8.33E-01

The TEM grids analysed were prepared 1 (i=0), 5 (i=1) and 15 minutes (i=2) after adding the cross -linker

solution.

1 mM 1,9-nonanedithiol solution

Figure 50 TEM images of 1:2 MH/OT batch B MPMNs solution 1 ( i=0) and 5 minutes (i=1) after adding the cross-

linker (NDT) solution (1 mM). Scale bars: 50 nm

71 of 94

0.5 mM 1,9-nonanedithiol solution

Figure 51 TEM images of 1:2 MH/OT batch B MPMNs solution a) 1 ( i=0), b) 5 (i=1) and c) 15 minutes ( i=2) after

adding the cross-linker (NDT) solution (0.5 mM). Scale bars: 50 nm

TEM images show that one minute after adding the 1 m M cross -linker solution, mainly isolated MPMNs,

and occasional two-dimensional aggregates were observed. At the end of five minutes, big three-

dimensional structures (300-3000 nm) were formed but no intermediate arrangements were seen, i.e.

one and two-dimensional structures.

Opposed to what should be expected, TEM micrographs show that one minute after adding the 0.5 mM

cross-linker solution, two- and three-dimensional NPs were present in the sample. Five (i=1) and fifteen

(i=2) minutes after massive three-dimensional aggregates (300-3000 nm) were present in the grid. Ball -

l ike arrangements were not evident in these samples.

A different set of particles (2:1 MH/OT batch A) was analysed with the same solutions of cross -linker

plus a less concentrated one (1, 0.5 and 0.1 mM – 2500, 1250 and 250 NDT/gold NP, respectively (see

table 14)). The solutions pippeted onto the TEM were prepared one minute after adding the cross-linker

solutions (i=0). Results obtained before (Figure 52) and after (Figure 53) exposing the solution to

sonication for a period of one hour are shown.

a)

b)

c)

72 of 94

Figure 52 TEM images of 2:1 MH/OT batch A MPMNs 1 minute (i=0) after adding the (a) 1 (b) 0.5 and (c) 0.1 mM

solution of cross-linker (NDT). Scale bars: 50 nm

Figure 53 TEM images of 2:1 MH/OT batch A sonicated MPMNs prepared 1 minute ( i=0) after adding the (a) 0.5 and

(b) 0.1 mM solution of cross-linker, NDT. Upper images: 1, 2 and 3 – deta iled images showing ordered assemblies. 4

– a magnif ied example of structure lacking the degree of order observed on the previous examples . Lower images:

blue arrows pointing to linear assembly based arrangements; grey arrow: unordered three-dimens ional assemblies.

Scale bars: 50 nm

a)

b)

c)

a)

b)

73 of 94

A general overview over the TEM grid shows that the solution cross -linked with a concentration of 1 mM

of dithiol has small assemblies like the ones observed without the presence of cross-linker (see Figure 33

– Control experiments). However, despite the lower molar concentration (0.5 and 0.1 mM), one minute

after adding the cross-linker solution, the TEM images show for both the concentrations random

aggregates of aggregates (300 - 3000 nm), with inhomogeneous form, dispersed in the microgrid.

Taking in consideration the last unexpected results, a fresh 1 mM solution of 1,9 -nonanedithiol was

prepared. The solution previously used was two months old so it was hypothesized that the cross -linker

solution could deteriorate with time which could explain the fact that less concentrated dithiol solutions

were forming faster big three-dimensional structures. However the results (see Figures 54 and 55)

indicate that the cross-linker stays stable in solution since there were no differences between the TEM

grids prepared with the two 1 mM NDT solutions synthesized with two months of difference.

Representative TEM micrographs are presented below:

Figure 54 TEM images of 2:1 MH/OT batch A MPMNs (a) 1 minute (i=0) and (b) 5 minutes after adding the 1 mM

solution of cross-linker, NDT (prepared on the 4th of December 2009). Blue arrows point to details of the three-

dimensional assemblies where a linear-based structural arrangement is observed. Scale bars: 50 nm

Figure 55 TEM images of 2:1 MH/OT batch A MPMNs (a) 1 minute (i=0) and (b) 5 minutes after adding the fresh 1

mM solution of cross-linker, NDT (prepared on the 4th of February 2010). The blue arrow indicates a detail of the

assembly where a linear-based structural arrangement is observed. Scale bars: 50 nm

All the TEM images (Figures 44-55) show randomized chain-like structures. There are a variety of

justifications for the existence of such networks in the TEM grid: (a) secondary bonding occurs between

a)

b)

a) b)

74 of 94

the alcohol groups in the ligand shells of neighboring NPs. (b) In the case of the rippled NPs, the pole

functionalization reaction conditions are not optimized and consequently place -exchange occurs not

only at the polar defects but also at other defect s ites in the ligand shell [91] and (c) the presence in

solution of excess thiol which was proven to play a major role in the formation of superlattice structures

[129]. However these explanations are not enough to clarify why are the 0.1 and 0.5 cross -linker

solutions making the MPMNs assemble faster than the 1 and 5 mM solutions.

It is also noteworthy to mention the two different types of arrangement that can be observed in the

aggregated structures. Blue arrows on Figures 53, 54 and 55 and images 1, 2 and 3 on Figure 53 point to

assembly structures where a very organized linear –based structure can be observed which suggests

that these structures have actually a high degree of organization. These observations suggest that this

type of aggregates observed on the TEM grids are composed of covalently bonded one- and two-

dimensional chains that are clustered into larger structures by van der Waals interdigitation forces. On

the other hand grey arrows on Figures 53, 54 and 55 and image 4 on figure 53 highlight structures

lacking the degree of organization observed on the examples mentioned on the last paragraph. Instead

unordered three-dimensional assemblies were formed.

Hussain et al. found that when the amount of cross -linker is reduced below a certain ratio (60 NDT/gold

NPs) the ligands at the gold NP surface will be easily replaced by dithiols, and the partial capping of gold

NPs with l inker molecules would then result in cross -linking leading to insoluble and irregular aggregates

[102]. The ratios of cross-linker solution applied and the kinetics of the place-exchange reaction will

result in a fast assembling process. For these reasons it is very challenging to control the aggregation

and assess an intermediate stage of the process where small one- and two-dimensional structures can

be observed.

Based on the previous results, and on the control experiments made witho ut the cross-linker (section

5.3.2.3) fresh solutions of MPMNs were prepared, diluted (0.125 mg/ml), sonicated for two hours and

left to stir for a minimum period of two days.

In concentrated solutions (e.g. 0.25 mg/ml) particles are already too close so can easily interdigitate and

bind to the neighbouring equivalents. Thus, initially, 1:100 and 1:1000 dilutions of the source solutions

(0.25 mg/ml) were prepared and analysed however, for these concentrations there were only a few

particles in solution which means it was difficult to see them in the TEM. (Images not shown)

75 of 94

5.3.2.6. Assembly after Purification Steps

Once completed the stabilization study of the MPMN solutions (see section 5.3.2.3) and after verifying

that NPs were spontaneously assembling in solution it was decided to start doing purification steps

before preparing the TEM grids. After confirming the absence of big three dimensional structures in the

TEM samples prepared with the purified MPMN solutions, the effect of the cross-linker in these samples

was then studied.

The first set of TEM grids was prepared three days after the purification steps and the second one was

prepared just one hour after the purification protocol. For these two systems the cross-linker and source

solution concentrations were also varied. The different conditions used are described in the following

table:

Table 15 Different conditions used in the assembly study, when the solutions analysed were submitted to the

purification protocol

Purification steps Syringe Filtration (0.1 µm) and centrifugation (15 000 rpm – 8 min)

1:1, 2:1 and 5:1 MH/OT Batch B TEM grids prepared 3 days after the purification steps

Cross-linker NDT – 1mM in EtOH

Concentration of AuNPs solution 1:10 dilution –0.025 mg/ml

Cross-linker to gold NP ratio 2.5E+04:1

Time after adding the cross-linker i=0 (few seconds after) i=1 (10 mins after)

1:1 MH/OT Batch B TEM grids prepared 1 hour after the purification steps

Cross-linker NDT – 0.01mM in EtOH

Concentration of AuNPs solution 1:5 dilution –0.05 mg/ml

Cross-linker to gold NP ratio 125:1

Time after adding the cross-linker i=1 (10 mins after) i=2 (1 hour after) and i=3 (12 hour after)

76 of 94

Figure 56 TEM pictures of the (a) 1:1 (b) 2:1 and (c) 5:1 MH/OT batch B purif ied MPMNs solutions (0.025 mg/ml)

prepared 1 minute ( i=0) after adding the 1mM cross-linker solution, NDT and (d) 1:1 (e) 2:1 and (f) 5:1 MH/OT batch

B prepared 10 minutes (i=1) after adding the same solution of cross-linker respectively.

In all cases, after a general view of the TEM grids there were a fe w aggregates. It was actually possible to

see these aggregates were slightly bigger 10 minutes after adding the cross-linker.

This slower kinetics of aggregation can be a result of the lower concentration of MPMNs solution used.

Decreasing the solution concentration increases the interparticle distance which leads to a reduced

probability of interdigitation or van der Waals interactions between neighbouring particles.

However, considering the ratio of 2.5E+04 molecules of cross-linker per gold NP which means 83 NDT to

thiol ratio, a higher disparity between the two different analysed times should be expected. This fact

combined with the lack of intermediate structures in the samples and the time factor (purification steps

performed three days before) drove the present work to the same questions: Are the existing

aggregates actually an effect of the cross-linker? Are the gold NPs self-assembling in solution? Or are

these structures a result of both mechanisms?

77 of 94

On the other hand, the results obtained after adding a low concentration cross -linker solution to a

MPMNs solution purified one hour before (see Figure 57) show that the assembly was only visible 12

hours after mixing these solutions, however intermediate structures were not observable in the samples

leading again to the same question: are the aggregates a result of spontaneous assembly or a result of

the cross-linker addition?

Figure 57 TEM pictures of the 1:1 MH/OT batch B purified MPMNs solution (0.05 mg/ml) prepared 10 minutes, 1

and 12 hours after adding the 0.01 mM cross-linker solution (1,9-nonanedithiol). Aggregation only visible after 12

hours. Scale bars: 50 nm

Interpreting these images is further complicated by the high interdigitation of gold NPs, in that it is

possible that the aggregates observed in the TEM grids are composed of cross-linked one- and two-

dimensional aggregates that are further assembled into larger structures by van der Waals forces and

cross-linker action.

5.3.2.7. New cross-linker approach

Taking in consideration the last results a new approach involving a different cross-linker was followed.

Considering that molecules in the ligand shell, MH and OT, are 6 and 8 carbons in length, respectively

and that the previous cross -linker (NDT) is 9 carbons, it is difficult to discern if the cross -linker is in fact

responsible for linking the neighboring particles together or is just helping or being an intermediate of

interdigitation and interparticle interactions between the ligands bond to the surface of each NP (see

Figure 58). For this reason a bigger cross-linker particle – 1,16-hexadecanedithiol (HDDT) was tested, in

order to achieve a faster place exchange reaction, using a lower ratio of cross-linker per gold NP. This

longer cross-linker will enable the assembly of the particles together without needing to form any extra

interactions or interdigitation between the ligands.

78 of 94

Figure 58 Schematic illustrating the cross-linking process for the two different dithiol containing molecules used: (a)

NDT and (b) HDDT (9 and 16 carbons in length respectively). Since the ligands in the SAM are 8 and 6 carbons in

length (a) a 9 carbon in length dithiol molecule will only link 2 particles together by chang ing the conformation of

the ligand shell, which will require a high degree of interdigitation and intermolecular interactions between the

ligands of the neighboring particles. On the other hand, (b) a 16 carbon in length dithiol molecule will link the

particles together without needing any extra interactions or interdigitations, which translates in a more

energetically favorable reaction, and thus a faster and more effective assembly process.

It is evident that when a dithiol-containing molecule solution in a ratio of 2500:1 NDT:gold NP is added,

it is going to make particles aggregate. This fact is due to the direct competition between the cross -

linker solution and the thiols assembled to the surface of the gold NPs. In such a high ratio it will place-

exchange imprecisely in the gold NP surface and not just on the defects. Thus, in order to see a

preferential place-exchange reaction, e.g. the place-exchange reaction occurring only in the defect spots

existent in the ligand shell we need to add the cross -linker in a ratio of 25 dithiol/gold NP (no more than

that) and be able to observe the reaction taking place in a countable time.

a)

b)

79 of 94

Figure 59 TEM images of 1:1 MH/OT Batch B MPMNs solution (0.05 mg/ml) 1 and 15 minutes, 1, 3 and 18 hours

after adding a 0.01 mM cross-linker solution (HDDT) (25:1 HDT/gold NP ratio) Scale bars: 50 nm

The TEM images revealed that aggregation was only visible 18 hours after adding the cross -linker

solution. Once again, this can be a result of spontaneous assembly of the particles in solution and not a

result of the cross-linker itself. The low HDDT/gold NP ratio used can be a cause for such a slow place

exchange kinetic, because it will not be energetically favourable for the SAM to lose its conformational

and stable morphology to let the cross -linker molecules bond to the particles’ surface. Also considering

that this cross-linker molecule is 16 carbons in length, the loop of it should be considered. In this case

the cross-linker molecule will form two thiol -gold bonds in the same gold NP and derail the further

l inking with other particles (see Figure 60).

Figure 60 Schematic illustrating the incoming cross-linker molecule coiling around in the gold NP forming two thiol-

gold bonds in the same particle.

80 of 94

In order to test if the cross -linker was effectively linking the particles and also confirm that not all the

cross-linker molecules were coiling around the particle, a new 5 mM sol ution of HDDT (2500:1 HDT/gold

NP ratio) was added.

Aggregation was observed 1 minute after adding the cross -linker solution. The assembly process was

also followed with the naked eye. In approximately 2 hours the solution changed from a light pink -red

colour to a browner one and finally to a dark precipitate.

Figure 61 TEM images of 1:1 MH/OT Batch B MPMNs solution (0.05 mg/ml) 1 minute after adding a 5 mM cross -

linker solution (HDDT) Scale bars: 50 nm.

81 of 94

6. Conclusions and Future Work

With the advent of nanotechnology and nanofabrication techniques and the development of new tools

for analysing and imaging nanoscaled objects, nanostructured systems can now be easily created and

modified at both structural and chemical levels, and represent the subject of fundamental scientific

studies of nano- and molecular-scale interactions. However, the device application of nanoscaled

systems lacks of suitable methods to organize them into well-defined structures.

Founded on the Stellacci’s group studies [51, 52, 72, 75-77, 91], which have explored the induced

assembly into linear chains of MPMNs coated with a binary mixture of ligands that self-assemble in the

ligand shell forming sub-nanometer-ordered domains, the present thesis was developed regarding the

relationship between the ligand shell structure (presence or absence of sub-nanometer-ordered

domains in the SAM) and their assembly using a dithiol as cross-linker.

In the first part of this thesis, the systematic synthesis, using a relatively fast and cost -effective one-

phase method [43], and further characterization of seven series of differently coated MPMN systems is

reported. The method followed allows the production of MPMNs with a narrow size distribution on a

gram scale.

In the second part of this project, important information regarding these particles spontaneous

assembly in solution is provided. Both in MPMNs assembly studies and i n MPMN on surfaces for protein

and cells interaction studies, stable solutions of MPMNs are required ( i .e. without the presence of

aggregates), therefore a purification method was developed, allowing the production of isolated

MPMNs in solution.

The third part describes the different approaches followed in the study of the particle assembly into

larger structures through induced place exchange reactions using a dithiol -containing molecule. Several

parameters were varied including the MPMNs solution and cross -linker concentrations, time and cross -

linker type molecule.

It was extremely challenging to control the number of functional groups attached to each parti cle. Given

the example of MPMNs with a core diameter around 4.6 nm, there are in total about 300 surface gold

atoms, and the total number of organic thiol ligands that may be attached to the particles is also close to

300 [20, 67]. Thus, discarding the hypothesis of place-exchange reaction at the, if present, polar defects

due to the high cross-linker/gold NPs ratios employed, these particles will present multiple unknown

numbers of functional groups .

When the cross-linker concentration was decreased, evident NP assemblies were hardly observable due

to the poor place exchange reaction kinetics. In such low concentrations it was not energetically

82 of 94

favourable to change the stable conformational morphology of the ligand shell to place-exchange with

an incoming molecule that forms the same bond (sulfur -gold) in the gold NP surface.

Due to the lack of a precise control over the number of functional groups, any chemical reactions

conducted on such nanoparticles most likely will lead to the formation of large aggregates with

unknown and irreproducible structures and properties or no clear aggregation at all [130].

As a first step to overcome this challenge, one needs to be able to produce nanoparticles with a

controlled number of chemical functional group attached to the surface that can later l ink directly or

exchange with a cross-linker molecule inducing assembly in predictable and controlled way. The

development of such regioselective chemistry on NPs can then enable the production of wide variety of

complex nanostructures.

Improvements can still be made in the synthesis and assembly study procedures, and there are s everal

potential directions for this research to proceed. A few options are discussed here:

The choice of the solvent and incoming molecules should be carefully studied and analyzed in order to

shift the reaction dynamic equilibrium favouring the adsorbed state of the incoming molecule. For

instance, an incoming molecule that is only slightly soluble in the reaction solvent would exchange onto

poles or other eventual defects in the ligand shell but feel only a weak driving force to desorb. In this

situation a more effective and faster place exchange reaction rate would take place allowing us to work

with lower ratios of cross-linker/gold NP, and in the case of rippled NPs observe the formation of one -

dimensional chains.

The composition of the ligand shell could also be altered in order to favour the place exchange reaction

rates. For instance, tetraoctylammonium bromide coated gold NPs exchange faster and more easily with

the dithiol-containing incoming molecules because thiols form a stronger bond in the gold surface than

the bromide molecules [102, 131]. Thus, using a low ratio of cross -linker/gold NP molecules

(approximately 20) we should be able to observe preferential and controlled place exchanging at the

gold NP surface.

Moreover alternative linking chemistries should be further studied to identify any potential advantages

they may present over the thiol-bond coupling used in this work. For instance biological molecules

represent ideal candidates for use as pole functionalization molecules [132] due to the specificity of

their chemistry that would allow precise control over the chain composition.

Finally, an alternative to form structures following a particle by particle trend, focused on the control of

chemical and inter-particle reactivity at the single NP level, is to simultaneously arrange several particles

into self-assembled materials. This approach yielded a bigger scope of new structures at the nanometer

scale with a multitude of potential applications such as optical coatings and gas sensors [102].

83 of 94

7. References

1. Feynman, R.P., Plenty of room at the bottom. J. Microelectromechanical Syst, 1992. 1: p. 60.

2. Qun, H.W., J. G., Monofunctional gold nanoparticles: synthesis and applications. Journal of

Nanoparticle Research, 2006. 9(6): p. 1013-1025.

3. Rejesky, D., Cpsc fy2010 agenda and priorities. , in Consumer Product, T. Report, Editor. 2009,

Safety Comission.

4. Diegoli, S.M., A. L.; Begum, S.; Jones, I. P.; Lead, J. R.; Preece, J. A., Interaction between

manufactured gold nanoparticles and naturally occurring organic macromolecules. Science,

2008. 402(1): p. 51-61.

5. Colvin, V.L., The Potential Environmental Impact of Engineered Nanomateri als. Nature

Biotechnology, 2003. 21(10): p. 1166-70.

6. Kanellos, M. Carbon nanotubes enter Tour of France. CNET news.com 07/07/2006.

7. Astruc, D., DM-C. , Gold nanoparticles: Assembly, supramolecular chemistry, quantumsize-

related properties, and applications toward biology, catalysis, and nanotechnology. Chemical

Reviews, 2004. 104(1): p. 293-346.

8. Link, S.E.-S., M. A. , Spectral properties and relaxation dynamics of surface plasmon electronic

oscillations in gold and silver nanodots and nanorods. Journal of Physical Chemistry B 1999.

103(40): p. 8410-26.

9. Andres, R.P.e.a., ‘‘Coulomb staircase’’ at room temperature in a self-assembled molecular

nanostructure. Science, 1996. 272(5266): p. 1323-1325.

10. Love, J.C.E., L. A.; Kriebel, J. K.; Nuzzo, R. G.; Whitesides, G. M., Self-Assembled Monolayers of

Thiolates on Metals as a form of Nanotechnology. Chemmical Reviews, 2005. 105(4): p. 1103-

1170.

11. Templeton, A.C.W., W.P.; Murray, R. W. , Monolayer-Protected Cluster Molecules. Accounts of

Chemical Research, 2000. 33(1): p. 27-36.

12. Mirkin, C.A., The beginning of a small revolution. Small, 2005. 1(1): p. 14-16.

13. Graigner, D.W.C., D. G.,, Nanobiomaterials and Nanoanalysis: Opportunities for Improving the

Science to Benefit Biomedical Technologies. Advanced Materials, 2008. 20(5): p. 867-877.

14. Ratner, M.R., D. , Nanotechnology: a gentle introduction to the next bigidea, ed. I. BooksCraft.

2003, Indianopolis: Bernard Goodwin.

15. Yang, Y.Y., Y.; Wang, W.; Li, J.;, Precise Size Control of Hydrophobic Gold Nanoparticles using

Cooperative Effect of Refluxing Ripening and Seeding Growth. Nanotechnology, 2008. 19(17): p.

175603.

16. Alivisatos, A., Semiconductor clusters, nanocrystals, and quantum dots. . Science, 1996.

271(5251): p. 933-37.

84 of 94

17. Daniel M, A.D., Gold nanoparticles: Assembly, supramolecular chemistry, quantumsize-related

properties, and applications toward biology, catalysis, and nanotechnology. Chemical Reviews,

2004. 104(1): p. 293-346.

18. Adamson, A.W.G., A. P. , Physical Chemistry of Surfaces. 6th ed. Catalytic Chemistry, ed. J.W.

Sons. 1997, New York.

19. Yee, C.K.J., R.; Ulman, A.; White, H.; King, A.; Rafailovich, M.; Sokolov, J., , Novel One-phase

Synthesis of Thiol -Functionalized Gold, Palladium and Iridium Nanoparticles using Superhydride.

Langmuir, 1999. 15(10): p. 3486-3491.

20. Brust, M.W., M.; Bethell, D.; Schiffrin, D. J.; Whyman, R., Synthesis of thiol -derivatised Gold-

Nanoparticles in a Two-phase Liquid-Liquid System. Journal of Chemical Society, 1994: p. 801-2.

21. Niemeyer, C.M., Nanoparticles, proteins, and nucleic acids: Biotechnology meets materials

science Angewandte Chemie-International Edition 2001. 40(22): p. 4128-58.

22. Feldheim, D.L., Keating, C. D., , Self-assembly of single electron transistors and related devi ces

Chemical Society Reviews 1998. 27(1): p. 1-12.

23. Hostetler, M.J.G., S. J.; Stokes, J. J.; Murray, R. W., Monolayers in three dimensions: Synthesis

and electrochemistry of omega-functionalized alkanethiolate-stabilized gold cluster compounds.

Journal of the American Chemical Society 1996. 118(17): p. 4212-3.

24. Link, S.E.-S., M. A., Size and temperature dependence of the plasmon absorption of colloidal

gold nanoparticles. . Journal of Physical Chemistry B 1999. 103(21): p. 4212-7.

25. Collier, C.P.S., R. J.; Shiang, J. J.; Henrichs, S. E.; Heath, J. R., Reversible tunning of silver

quantum-dot monolayers through the metal -insulator transistor. Science, 1999. 277(5334): p.

1978-81.

26. Malinsky, M.D.K., K. L.; Schatz, G. C.; Van Duyne, R. P.,, Chain length dependence and sensing

capabilities of the localized surface plasmon resonance of silver nanoparticles chemically

modified with alkanethiol self-assembled monolayers Journal of the American Chemical Society

2001. 123(7): p. 1471-82.

27. Thomas, K.G.K., P. V.,, Chromophore-functionalized gold nanoparticles Accounts of Chemical

Research, 2003. 36(12): p. 888-98.

28. Hayat, M.A., Colloidal Gold - Principles, Methods, and Applications, ed. I. Academic Press. 1989,

San Diego: Harcourt Brace Jovanovich, Publishers.

29. Freestone, I., Meeks, N., Sax, M., Higgitt, C., The Lycurgus Cup - A Roman Nanotechnology. Gold

Bulletin, 2007. 40(4): p. 270-277.

30. Antonii, F., Panacea Aurea-Auro Potabile. Bibliopolio Frobeniano, 1618.

31. Kunckels, J., Nuetliche Observationes oder Anmerkungen von Auro und Argento Potabili.

Schutzens, 1676.

32. Klickstein, H.S.L., H. M., A Source Book in Chemistry, 1400 - 1900, ed. O.U. Press. 1968.

33. Helcher, H.H., Aurum Potabile oder Gold Tinstur. J. Herbord Klossen, 1718.

85 of 94

34. Thompson, D., Michael Faraday's Recognition of Ruby Gold: the Birth of Modern

Nanotechnology. Gold Bulletin, 2007. 40(4): p. 267-269.

35. Faraday, M., The Bakerian Lectura: Experimental Relations of Gold (and Other Metals) to Light.

Philosophical Transactions Royal Society of London 1857. 147: p. 145-181.

36. Schmid, G., Clusters and Colloids - From Theory to Applications, ed. VCH. 1994, Weinheim.

37. Glogowski, E.H., J.; Russel, T. P.; Emrick, T. , Mixed monolayer coverage on gold nanoparticles

for interfacial stabilization of immiscible fluids. Chemical Communications, 2005. 32: p. 4050-

52.

38. Foos, E.E.S., A. W.; Twigg, M. E.; Ancona, M. G, Thiol-Terminated Di-, Tri-,and Tetraethylene

Oxide Functionalized Gold Nanoparticles: A Water-Soluble, Charge-Neutral Cluster. . Chemistry

of Materials, 2002. 14(5): p. 2401-2408.

39. Kanaras, A.G.K., F. S.; Schaumburg, K.; Kiely, C. J.; Brust, M. , Thiolalkylated tetraethylene glycol:

a New Ligand for Water Soluble Monolayer Protected Gold Clusters. . Chemical

Communications, 2002. 20: p. 2294-2295.

40. Yonezawa, T.S., M.; Kunitake, T. , Practical Preparation of Size-Controlled Gold Nanoparticles in

Water. Chemistry Letters 1997. 26(7): p. 619-620.

41. Warner, M.G.R., S. M.; Hutchison, J. E., Small, Water-Soluble, Ligand-Stabilized Gold

Nanoparticles Synthesized by Interfacial Ligand Exchange Reactions. . Chem. Mater. , 2000.

12(11): p. 3316-3320.

42. Rowe, M.P.P., K. E.; Kim, K.; Kurdak, Ç.; Zellers, E. T.; Matzger, A. J., , Single-Phase Synthesis of

Functionalized Gold Nanoparticles. Chemistry of Materials, 2004. 16(18): p. 3513-17.

43. Zheng, N.F., J.; Stucky, G. D., One-Step One-Phase Synthesis of Monodisperse Noble-Metallic

Nanoparticles and their Colloidal Cristals, . Journal of the American Chemical Society, 2006.

128(20): p. 6550-1.

44. Mie, G., Beiträge sur Optik trüber Medien, speziell killoidaler Metallösungen. Annalen der

Physik, Vierte Folge, 1908. 25(3): p. 376-445.

45. Barnes, W.L.D., A.; Ebbesen, T. W. , Surface plasmon subwavelength optics. Nature, 2003.

424(6950): p. 824-830.

46. Du, G.X.M., T.; Suzuki, M.; Saito, S.; Fukuda, H.; Takahashi, M. , Evidence of localized surface

plasmon enhanced magneto-optical effect in nanodisk array Applied Physics Letters 2010. 96(8):

p. 081915.

47. Liz-Marzán, L.M., Tailoring Surface Plasmons through the Morphology and Assembly of Metal

Nanoparticles. Langmuir, 2006. 22(1): p. 32-41.

48. Link, S.W., Z. L.; El -Sayed, M. A., Alloy formation of Gold-silver Nanoparticles and the

Dependence of the Plasmon Absorption on their Composition. Journal of Physical Chemistry B,

1999. 103(18): p. 3529-33.

49. Schasfoort, R.B.M.T., A. J. , Handbook of Surface Plasmon Resonance, ed. T.R.S.o. Chemistry.

2008, Cambridge: RSC publishing.

86 of 94

50. Mrksich, M.W., G. , Using self-assembled monolayers to understand the interactions of man-

made surfaces with proteins and cells. Annual Review of Biophysics and Biomolecular Structure,

1996. 25: p. 55-78.

51. Jackson, A.M.H., Y; Silva, P. J.; Stellacci, F From Homoligand- to Mixed-Ligand- Monolayer-

Protected Metal Nanoparticles: A Scanning Tunneling Microscopy Investigation Journal of the

American Chemical Society, 2006. 128(34): p. 11135-49.

52. Jackson, A.M.M., J. W.; Stellacci, F. , Spontaneous assembly of subnanometre ordered domains

in the ligand shell of monolayer-protected nanoparticles Nat Matter, 2004. 3(5): p. 330-336.

53. Akthakul, A.H., A. I.; Stellacci, F.; Mayes, A. M. , Size Fractionation of Metal Nanoparticles by

Membrane Filtration. Adv. Mater, 2005. 17(5): p. 532-535.

54. Georganopoulou, D.G.C., L.; Nam, J. M.; Thaxton, C. S.; Mufson, E. J.; Klein, W. L.; Mirkin, C. A.,

Nanoparticle-based Detection in Cerebral Spinal Fluid of a Soluble Pathogenic Biomarker for

Alsheimer’s Disease Proceedings of the National Academy of Sciences U.S.A. , 2005. 102(7): p.

2273-2276.

55. Wu, Z.Z., B.; Yan, B. , Regulation of Enzyme Activity through Interactions with Nanoparticles.

International Journal of Molecular Sciences, 2009. 10(10): p. 4198-4209.

56. Andres, R.P.B., J. D.; Henderson, J. I.; Janes, D. B.; Kolagunta, V. R.; Kubiak, C. P.; Mahoney, W.

J.; Osifchin, R. G. , Self-Assembly of a two-dimensional superlattice of Molecularly linked Metal

Clusters. Science, 1996. 273(5282): p. 1690-1693.

57. Ingram, R.S.M., R. W., Electroactive Three-Dimensional Monolayers: Anthraquinone ω-

Functionalized Alkanethiolate-Stabilized Gold Clusters. . Langmuir, 1998. 14(15): p. 4115-4121.

58. Ulman, A., Formation and Structure of Self-assembled monolayers. Chemical Reviews, 1996.

96(4): p. 1533-54.

59. Schwartz, D.K., Mechanisms and Kinetics of Self-Assembled Monolayer Formation Annual

Review of Physical Chemistry, 2001. 52: p. 107-137.

60. Vos, J.G.F., R. J.; Keyes, T. E. , Formation and Characterization of Modified Surfaces, in

Interfacial Supramolecular Assemblies, L. John Wiley & Sons, Editor. 2003, Wiley-Interscience:

West Sussex, England. p. 88-94.

61. Hasan, M.B., D.; Brust, M., The Fate of Sulfur-Bound Hydrogen on Formation of Self-Assembled

Thiol Monolayers on Gold: 1H NMR Spectroscopic Evidence from Solutions of Gold Clusters.

Journal of the American Chemical Society, 2002. 124(7): p. 1132-3.

62. Bareman, J.P.K., M. L. , Collective ti lt behavior in dense, substrate-supported monolayers of

long-chain molecules: a molecular dynamics study. J. Phis. Chem., 1990. 94(13): p. 5202-5.

63. Vericat, C.V., M. E.; Salvarezza, R. C., Self-assembled monolayers of alkanethiols on Au(111):

Surface structures, defects and dynamics. Journal of Physical Chemistry B, 2005. 7(18): p. 3258-

68.

64. Prime, K.W., G.;, Self-Assembled Organic Monolayers - Model Systems for studying adsorption

of proteins at surfaces. Science, 1991. 252(5009): p. 1164-7.

87 of 94

65. Zanchet D, H.B., Ugarte D. , Structure population in thiol -passivated gold nanoparticles. J. Phys

Chem B. , 2000. 104(47): p. 11013-11018.

66. Gutierrez, W.C.A., J.; Perez-Alvarez, M.; Marin-Almazo M.; Yacaman, J. M. , On the structure and

formation of self-assembled lattices of gold nanoparticles. J Clust Sci. , 1998. 9(4): p. 529-545.

67. Hostetler, M.J.T., A. C.; Murray, R. W.; , Dynamics of Place-Exchange Reactions on Monolayer-

Protected Gold Cluster Molecules. Langmuir, 1999. 15(11): p. 3782-89.

68. Georgakilas, V.G., D.; Tzitzios, V.; Pasquato, L.; Guildi, D. M.; Prato, M., Decorating Carobon

Nanotubes with Metal or Semiconductor Nanoparticles. Journal of Materials Chemistry, 2007.

17(26): p. 2679-94.

69. Stranick, S.A., S.; Parikh, A.; Wood, M. , Nanometer-scale phase separation in mixed

composition self-assembled monolayers. Nanotechnology, 1996. 7(4): p. 438-442.

70. Folkers, J.L., P.; Whitesides , G. , Self-assembled monolayers of alkanethiols on gold:

comparisons of monolayers containing mixtures of short - and long-chain constituents with

methyl and hydroxymethyl terminal groups. . Langmuir, 1992. 8(5): p. 1330-1341.

71. Smith, R.K.R., S. M.; Lewis, P. A.; Monnell, J. D.; Clegg, R. S.; Kelly, K. F.; Bumm, L. A.; Hutchison,

J. E.; Weiss, P. S., Phase separation within a binary self-assembled monolayer on Au{111} driven

by an amide-containing alkanethiol Journal of Physical Chemistry B, 2001. 105(6): p. 1119-1122.

72. Centrone, A.H., Y.; Jackson, A. M.; Zerbi, G.; Stellacci, F. , Phase separation on mixed-

monolayer-protected metal nanoparticles: a study by infrared spectroscopy and scanning

tunneling microscopy. . Small, 2007. 3(5): p. 814-817.

73. Singh, C.G., P. K.; Horsch, M. A.; Jackson, A. M.; Larson, R. G.; Stellacci, F.; Glotzer, S. C.,

Entropy-Mediated Patterning of Surfactant-Coated Nanoparticles and Surfaces Phys Rev L ett .,

2007. 99(22): p. 226106.

74. Casagrande, C.V., C.R. , Janus Beads - Realization and 1st Observation of Interfacial Properties.

Comptes Rendus de L'Academie des Sciences, 1988. 306: p. 1423-25.

75. Carney, R.P.D., G.A.; Dubois, C.; Kim, H.; Kim, J.Y.; Singh, C.; Ghorai, P. K.; Tracy, J. B.; Stiles, R.

L.; Murray, R. W.; Glotzer, S. C.; Stellacci, F. , Size l imitations for the formation of ordered

striped nanoparticles. Journal of the American Chemical Society, 2008. 130(3): p. 798-799.

76. Centrone, A.P., E.; Sharma, M.; Myerson, J.W.; Jackson, A.M.; Marzari , N.; Stellacci, F., The role

of nanostructure in the wetting behavior of mixed-monolayer-protected metal nanoparticles.

Proc Natl Acad Sci USA, 2008. 105(29): p. 9886-9891.

77. Hu, Y.U., O.; Dubois, C.; Stellacci, F. , Effect of Ligand Shell Structure on the Interaction between

Monolayer-Protected Gold Nanoparticles. Journal of Physical Chemistry C, 2008. 112(16): p.

6279-6284.

78. Uzun, O.H., Y.; Verma, A.; Chen, S.; Centrone, A.; Stellacci, F., Water-Soluble Amphiphilic Gold

Nanoparticles with Structured Ligand Shells,. Journal of the Chemical Society, Chem. Commun. ,

2007. 2: p. 196-198.

88 of 94

79. McFarland, A.D.V.D., R. P., Single Silver Nanoparticles as Real -Time Optical Sensors with

Zeptomole Sensibity. Nano Letters, 2003. 3(8): p. 1057-62.

80. Wooley, K., New nanoparticle coating mimics dolphin skin prevents 'biofouling' of ship hulls, in

Innovations Report - Forum for Science, Industry and Business. 2002, Materials Sciences.

81. Brown, L.O.H., J. E. , Convenient Preparation of Stable, Narrow-Dispersity, Gold Nanocrystals by

Ligand Exchange Reactions. Journal of the American Chemical Society, 1997. 119(50): p. 12384-

12385

82. Dubois, L.H.N., R. G., The Synthesis, Structure, and Properties of Model Organic Surfaces. Annual

Reviews of Physical Chemistry 1992. 43: p. 437-463.

83. Collard, D.M.F., M. A. , Use of electroactive thiols to study the formation and exchange of

alkanethiol monolayers on gold. Langmuir, 1991. 7(6): p. 1192-1197.

84. Scott, J.R.B., L. S.; Everett, W. R.; Wilkins, C. L.; Fritsch,I. , Laser Desorption Fourier Transform

Mass Spectrometry Exchange Studies of Air-Oxidized Alkanethiol Self-Assembled Monolayers on

Gold. Analytical Chemistry, 1997. 69(14): p. 2636-2639.

85. Schlenoff, J.B.L., M.; Ly, H. , Stability and Self-Exchange in Alkanethiol Monolayers. Journal of

the American Chemical Society, 1995. 117(50): p. 12528-12536.

86. Nishida, N.H., M.; Sasabe, H.; Knoll, W. , Formation and Exchange Processes of Alkanethiol Self-

Assembled Monolayer on Au(111) Studied by Thermal Desorption Spectroscopy and Scanning

Tunneling Microscopy. Japanese Journal of Applied Physics, 1996. 36: p. 2379-2385.

87. Ingram, R.S.H., M. J.; Murray, R. W. , Poly-hetero-ω-functionalized Alkanethiolate-stabilize Gold

Cluster Compounds. Journal of the American Chemical Society, 1997. 119(39): p. 9175-8.

88. Westerlund, F.B., T. , Directed assembly of gold nanoparticles. Current Opinion in Colloid &

Interface Science 2009. 14(2): p. 126-134.

89. Maier, S.A.B., M. L.; Kik, P. G.; Meltzer, S.; Requicha, A. A. G.; Koel, B. E.; Atwater, H. A.,

Plasmonics - a route to Nanoscale Optical Devices. Advanced Materials, 2001. 13(19): p. 1501-5.

90. Thomas, K.G.B., S.; Ipe, B. I.; Joseph, S. T. S.; Kamat , P. V., Unidirectional Plasmon Coupling

through Longitudinal Self-assembly of Gold Nanorods. Journal of Physical Chemistry B, 2004.

108(35): p. 13066-68.

91. DeVries, G.A.B., M.; Hu, Y.; Jackson, A. M.; Long, B.; Neltner, B. T.; Uzun, O.; Wunsch, B. H.;

Stellacci, F. , Divalent Metal Nanoparticles Science, 2007. 315(5810): p. 358-361.

92. Rosi, N.L.T., C. S.; Mirkin, C. A., Control of Nanoparticles Assembly using DNA-modified Diatom

templates. Angewandte Chemie - International Edition, 2004. 43(41): p. 5500-3.

93. Bachand, G.D.R., S. B.; Boal, A. K.; Gaudioso, J.; Liu, J.; Bunker, B. C. , Assembly and transport of

nanocrystal CdSe quantum dot nanocomposites using microtubules and kinesin motor proteins.

Nano Lett., 2004. 4(5): p. 817-21.

94. Wang, G.L.M., R. W., Controlled assembly of Monolayer-Protected Gold Clusters by Dissolved

DNA. Nano Lett., 2004. 4(1): p. 95-101.

89 of 94

95. Tang, Z.Y.K., N. A., One-Dimensional Assemblies of Nanoparticles: Preparations, Properties and

Promise. Adv. Mater. , 2005. 17(8): p. 951-62.

96. Schreiber, F., Structure and growth of self-assembling monolayers Prog. in Surface Science,

2000. 65(5-8): p. 151-257.

97. Luedtke, W.D.L., U., Structure and Thermodynamics of Self-Assembled Monolayers on Gold

Nanocrystall ites Journal of Physical Chemistry B, 1998. 102(34): p. 6566-6572.

98. Ghorai, P.K.G., S. C. , Molecular Dynamics Simulation Study of Self-Assembled Monolayers of

Alkanethiol Surfactants on Spherical Gold Nanoparticles. Journal of Physical Chemistry C, 2007.

111(43): p. 15857-62.

99. Nelson, D.R., Toward a Tetravalent Chemistry of Colloids. Nano Letters, 2002. 2(10): p. 1125-

1129.

100. Brust, M.B., D.; Kiely, C.; Schiffrin, D. , Self-assembled Gold Nanoparticle Thin Films with

Nonmetallic Optical and Electronic Properties Langmuir, 1998. 14(19): p. 5425-5429.

101. Brust, M.B., D.; Schiffrin, D. J.; Kiely, C. , Novel Gold-Dithiol Nano-networks with Non-metallic

Electronic Properties. J. Adv. Mater 1995. 7(9): p. 795-797.

102. Hussain, I.W., Z.; Cooper, A. I.; Brust, M. , Formation of Spherical Nanostructures by the

Controlled Aggregation of Gold Colloids. Langmuir, 2006. 22(7): p. 2938-2941

103. Sih, B.C.W., M. O. , Metal Nanoparticle-Conjugated Polymer Nanocomposites. Chemical

Communications, 2005: p. 3375-84.

104. Perepichka, D.F.R., F. , Metal Nanoparticles: From “Artificial Atoms” to “Artificial Molecules”

Angewandte Chemie - International Edition, 2007. 46(32): p. 6006-6008.

105. Sung, K.-M.M., D. W.; Peelle, B. R.; Zhang, S.; Jacobson, J. M. , Synthesis of Monofunctionalized

Gold Nanoparticles by Fmoc Solid-Phase Reactions. J. Am. Chem. Soc., 2004. 126(16): p. 5064-

65.

106. Worden, J.G.S., A. W.; Huo, Q. , Controlled Functionalization of Gold Nanoparticles through a

Solid Phase Synthesis Approach. Chemical Communications, 2004(5): p. 518-9.

107. Huo, F.L.-J., A. K. R.; Mirkin, C. A. , Asymmetric functionalization of Nanoparticles based on

Thermally addressable DNA Interconnects. Advanced Materials, 2006. 18(17): p. 2304-6.

108. Deng, Z.T., Y.; Lee, S. H.; Ribbe, A. E.; Mao, C. , DNA-Encoded Self-Assembly of Gold

Nanoparticles into One-Dimensional Arrays. Angewandte Chemie - International Edition, 2005.

44(23): p. 3582-85.

109. Aldaye, F.A.S., H. F. , Sequential Self-Assembly of a DNA Hexagon as a Template for the

Organization of Gold Nanoparticles. Angew. Chem., 2006. 45(14): p. 2204-09.

110. Hore, P.J., Nuclear Magnetic Resonance. Oxford Chemistry primers, ed. O.U. Press. 1995,

Oxford: Oxford University Press Inc.

111. Hoppert, M., Microscopic Techniques in Bioltechnology, ed. W.-. VCH. 2003, Verlag.

112. Ricci, D.B., P. C., Atomic Force Microscopy: Biomedical Methods and Applications. Methods in

Molecular Biology, ed. H. Press. Vol. 242. 2004, New Jersey.

90 of 94

113. Vilarinho, P.M.R., Y; Kingon, A., Scanning Probe Microscopy: Characterization, Nanofabrication

and Device Application of Functional Materials. NATO Science Series, ed. K. academic. 2005,

Dordrecht.

114. Kreibig, U.Q., M., Optical properties of aggregates ofsmall metal particles. Surface Science,

1986. 172(3): p. 557-577.

115. Hyningl, D.L.V.Z., C. F., Formation Mechanisms and Aggregation Behaviour of Borohydride

Reduced Silver Particles. Langmuir, 1998. 14(24): p. 7034.

116. Natan, M.J.B., K. R., Hydroxylamine Seeding of Colloidal Au Nanoparticles in Solution and on

Surfaces. Langmuir, 1998. 14(4): p. 726.

117. Owen, T., Fundamentals of modern UV-visible spectroscopy, ed. Primer. 2000, Germany.

118. Collier, C.P.S., R. J.; Shiang, J. J.; Henrichs, S. E.; Heath, J. R., Reversible tunning of silver

quantum-dot monolayers through the metal -insulator transistor. Science, 1997. 277(5334): p.

1978-81.

119. Jacobsen, N.E., NMR Spectroscopy Explained - Simplified Theory, Applications and Examples for

Organic Chemistry and Structural Biology, ed. Wiley-Interscience. 2007, New Jersey: John Wiley

& Sons, Inc.

120. Zhang, X.-F.Z., Z., Progress in Transmission Electron Microscopy 1 - Concepts and Techniques.

Springer Series in Surface Sciences, ed. G.G. Ertl, R.; Lüth,H.; Mills, D. L. 1999, New York:

Springer.

121. Goodhew, P.J.H., F. J.; Beanland, R., Electron Microscopy and Analysis, ed. T.a. Francis. 2001,

London.

122. Yao, N.W., Z. L., Handbook of Microscopy for Nanotechnology. 2005, New York: Kluwer

Academic Publishers

123. Williams, D.B.C., C.B., Transmission Electron Microscopy - Basics. A Textbook for Materials

Science. 1996, New York: Plenum Press.

124. Labande, A.R., J.; Astruc, D., Supramolecular Gold Nanoparticles for the Redox Recognition of

Oxoanions: Synthesis, Titrations, Stereoelectronic Effects, and Selectivity. Journal of the

American Chemical Society, 2002. 124(8): p. 1782-9.

125. Liao, J.Z., Y.; Yu, W.; Xu, L.; Ge, C.; Liu, J.; Gu, N., Linear Aggregation of Gold Nanoparticles in

Ethanol. Colloids and surfaces. A, Physicochemical and engineering aspects, 2003. 223(1-3): p.

177-183.

126. Galletto, P.B., P. F.; Girault, H. H.; Antoine, R.; Broyer, M., Size Dependence of the Surface

Plasmon Enhanced Second Harmonic Response of Gold Colloids:Towards a New Calibration

Method. Chemical Communications, 1999(7): p. 581-582.

127. Ohara, P.C.H., J. R.; Gelbart, W. M., Self-Assembly of Submicrometer Rings of Particles from

Solutions of Nanoparticles. Angewandte Chemie - International Edition, 2003. 36(10): p. 1078-

1080.

91 of 94

128. Khanal, B.P.Z., E. R., Rings of Nanorods. Angewandte Chemie - International Edition, 2007.

46(13): p. 2195-2198.

129. Sidhaye, D.S.P., B. L. V., Melting Characteristics of Superlattices of Alkanethiol-Capped Gold

Nanoparticles: The "Excluded" Story of Excess Thiol. Chemistry of Materials, 2010. 22(5): p.

1680-5.

130. Shaffer, A.W.W., J. G.; Huo, Q., Comparison Study of the Solution Phase versus Solid Phase Place

Exchange Reactions in the Controlled Functionalization of Gold Nanoparticles. Langmuir, 2004.

20(19): p. 8434-8351.

131. Shon, Y.-S.C., S.; Voundi, P., Stability of Tetraoctylammonium Bromide-Protected Gold

Nanoparticles: Effects of Anion Treatments. Colloids and surfaces A: Physicochemical and

Engineering aspects, 2009. 353(1-3): p. 12-17.

132. Mann, S.S., W.; Li, M.; Connolly, S.; Fitzmaurice, D., Biologically Programmed Nanoparticle

Assembly. Advanced Materials, 1999. 12(2): p. 147-150.

92 of 94

Appendix

Figure A.1 1H NMR spectrum obtained for 1:0 MH/OT in CD3OD. The presence of a peak in the NMR spectrum

associated to the the hydrophilic ligand (MH at 3.4 ppm) shows the presence of free ligands in solution.

Figure A.2 1H NMR spectrum obtained for 5:1 MH/OT in CD3OD. The presence of a peak in the NMR spectrum

associated to the the hydrophilic ligand (MH at 3.4 ppm) shows the presence of free ligands in solution.

1:0 MH/OT

5:1 MH/OT

93 of 94

Figure A.3 1H NMR spectrum obtained for 1:1 MH/OT in CD3OD. The presence of a peak in the NMR spectrum

associated to the the hydrophilic ligand (MH at 3.4 ppm) shows the presence of free ligands in solution.

Figure A.4 1H NMR spectrum obtained for 1:2 MH/OT in CD3OD. The presence of a peak in the NMR spectrum

associated to the the hydrophilic ligand (MH at 3.4 ppm) shows the presence of free ligands in solution.

Figure A.5 1H NMR spectrum obtained for 1:5 MH/OT in CD3OD. The presence of a peak in the NMR spectrum

associated to the the hydrophilic ligand (MH at 3.4 ppm) shows the presence of free ligands in solution.

1:1 MH/OT

1:2 MH/OT

1:5 MH/OT

94 of 94

Figure A.6 UV/Vis spectra of 1:2Batch A MH/OT particles. Only a very slight decrease of the SPR band was observed

suggesting that the aggregation was taking more time