MÀSTER EN NANOTECNOLOGIA

PROGRAMA OFICIAL DE POSTGRAU D’ENGINYERIA MEMÒRIA DEL TREBALL DE MASTER

Especialitat Nanociència i Nanotecnologia

Synthesis of Cobalt Nanoparticles and their Self-Assembly on Highly

Oriented Pyrolitic Graphite (HOPG)

Autora: Miriam Varón Izquierdo Director: Victor Puntes Departament/Institut del director: Inorganic Nanoparticles Group/Institut Català de Nanotecnologia Data: 20/02/2009

CONTENTS

1. ABSTRACT.................................................................................................... 1

2. INTRODUCTION ............................................................................................ 3

2.1. MAGNETIC PROPERTIES OF SOLIDS.................................................. 3

2.2. MAGNETIC NANOPARTICLES............................................................... 4

2.3. SELF-ASSEMBLY ................................................................................... 8

2.4. CHARACTERIZATION .......................................................................... 12

2.4.1. Optical Microscope.......................................................................... 12

2.4.2. Transmission Electron Microscopy.................................................. 12

2.4.3. Scanning Electron Microscope - EDX ............................................. 13

2.4.4. SQUID Magnetometry ..................................................................... 13

2.4.5. Atomic Force Microscope (AFM)/Magnetic Force Microscope (MFM)

.................................................................................................................. 14

3. RESULTS and DISCUSSION ...................................................................... 17

3.1. SYNTHESIS OF COBALT NANOPARTICLES ...................................... 17

3.2. Deposition onto HOPG .......................................................................... 18

3.3. Regarding the formation mechanism ..................................................... 22

4. CONCLUSIONS ........................................................................................... 29

5. REFERENCES ............................................................................................. 33

5. ACKNOWLEDGMENTS .............................................................................. 35

6. ANNEX I....................................................................................................... 37

1

1. ABSTRACT

Superparamagnetic single crystal Cobalt nanoparticles (Co NPs), with single

magnetic domain, of 6 nm and 8 nm in diameter evaporated onto Highly

Pyrolytic Oriented Graphite (HOPG), spontaneously self-assemble into super

structures with an elongated shape. These structures have been studied by

optical and scanning electron microscopies (SEM), atomic and magnetic force

microscopy (AFM and MFM), electron dispersive x-ray (EDX) analysis and

SQUID magnetometry. It was proposed that the weak dipolar interactions

between superparamagnetic dipoles of the Co NPs are responsible for the

formation of these structures. This could be explained because dipolar magnetic

interactions are strong enough to influence the general process of self assembly

dominated by Van der Waals forces between nanoparticles and between

nanoparticles and substrate, and also due to the evaporation dynamics of the

experiment.

2

3

2. INTRODUCTION

2.1. MAGNETIC PROPERTIES OF SOLIDS

Depending on their response towards external applied fields, materials

can be classified as diamagnetic, paramagnetic, ferromagnetic, ferrimagnetic

and antiferromagnetic, according to the arrangement of their magnetic dipoles

in the absence or presence of an external magnetic field.

Diamagnetic material: It is a material with no unpaired electrons in the orbital

shells (magnetic dipoles) resulting in no net magnetic moment in the absence of

an external field. The magnetization of a diamagnet responds in the opposite

direction to the external field.

Paramagnetic material: Some of the atoms or ions in the material have a net

magnetic moment due to the unpaired electrons in partially filled orbitals, even

in the absence of an applied magnetic field. When a magnetic field is applied,

the dipoles tend to align with the applied field, resulting in a net magnetic

moment in the direction of the applied field (attraction).

Ferromagnetic material: A ferromagnetic material has unpaired electrons

(magnetic dipoles) resulting in a permanent magnetic dipole (both in presence

or absence, after being magnetized of an external field) and exhibits a long

arrangement at the atomic level.

Ferrimagnetic and antiferromagnetic materials: Like ferromagnetic ones, these

materials have permanent magnetic dipoles. The magnetic structure is

composed of two magnetic sublattices (called A and B). In ferrimagnets the

magnetic moment of A and B are not equal and result in a net magnetic

moment. If the A and B sublattices are exactly equal but opposite, the net

moment is zero and this type of magnetic ordering is called antiferromagnetism.

4

Superparamagnetic material: A superparamagnetic material is composed of

small ferromagnetic clusters (crystallites), being so small that these can

randomly flip direction under thermal fluctuations. As a result, the material as a

whole is not magnetized except under an externally applied magnetic field (in

that respect, it is like paramagnetism).

2.2. MAGNETIC NANOPARTICLES

A magnetic nanoparticle is a structured material formed at least by

hundreds or thousands of atoms. These atoms are organized within a

crystalline structure that determines their size and magnetic properties.

A magnetic nanoparticle has a net moment that is the sum of the spin of all

atoms. Therefore, the magnetic moment depends of the size and shape of the

particle (Figure 2.1).

The magnetic moment of the particle aligns spontaneously in one direction,

corresponding to the easy magnetization axis and the direction that minimizes

the anisotropy energy. Some materials possess any kind of anisotropy that

affects the behavior of the magnetization.

Spherical Particles

Diameter (nm) Volume (nm3) Magnetic Moment (J/T)

3 14.1 2.0 e-20

5 65.4 9.3 e-20

8 268 3.8 e-19

10 524 7.4 e-19

15 1770 2.5 e-18 Figure 2.1: Magnetic moment for cobalt nanoparticles depending on the size.

Inside a material the spins form domains, where the individual moments of the

atoms are aligned with one another. The regions separating different magnetic

domains are called domain walls where the magnetization rotates coherently

5

from the direction in one domain to that in the next domain. The existence of

magnetic domains is a result of the energy minimization.

The formation of domains is highly related with the size of the particle. In one

hand, in big particles the energy considerations suits the formation of domains.

On the other hand, when the size of the particle decreases the number of

domains also decreases and becomes a single domain. If the particle size is

reduced, there is a critical volume below which it costs more energy to create a

domain wall than to support the external magnetostatic energy (stray field) of

the single-domain state. This critical diameter typically lies in the range of a few

tens of nanometers, depends on the material and it is influenced by the

contribution from various anisotropic energy terms.

Figure 2.2: Different magnetic responses in the hysteresis curves when varying the particle size.

Magnetic nanoparticles have been the subject of extensive research because

they have the potential to be utilized in several applications such us ultra-high-

density recording media,1,2 contrast agent in magnetic resonance imaging,3

drug delivery4 and as single electron transistors.1,2

The development, characterization and exploitation of novel materials based on

the assembly of molecular components (such us nanoparticles) is highly active

and rapid expanding field. Their properties arise from the large amount of atoms

present on the surface of the nanoparticle and from the finite number of atoms

in each crystalline core. It was in magnetic materials that such finite-size effects

were recognized, with more subtle effects noted, later in non-magnetic metals

such as semiconductors. The focus in this area is still on building solids with

tailored magnetic properties and to make a breakthrough in the principal

Super-paramagnetic

Ferromagnetism

Superparamagnetism

Hysteresis curves

6

challenge facing work in low dimension structures; i.e., how to make the

transition from individual nanoscale behaviour to bulk macroscopic phenomena

and properties.

Recently it has been shown that nanocrystals can self assemble in 2D and 3D

arrays with long range translational and even orientational order. Nanocrystals

can be organized into close packed arrays simply by evaporating the solvent

from a hydrophobic sterically stabilized dispersion, provided that the size

distribution of the particles is sufficiently narrow (i.e., a standard deviation about

the mean diameter of less than 10%). Thus, these arrays may reproduce the

nanoscale properties in macroscopic materials through homogeneity and

periodicity, and elicit collective electronic, magnetic, and optical behavior

resulting from the relative positioning of the nanocrystals in the array.

In ferromagnetic materials, the dipolar magnetic interactions between particles

add a new term in the energy balance, and they are not screened or affected by

the surrounding materials (solvent, substrate, surfactants, etc.). In particular, as

the mean dipolar energy increases with the particle volume, the thermal energy

at RT becomes comparable to the dipolar energy of two sticking Co NPs with

diameter size of about 12 nm. A collective behavior due to dipolar interactions

has been observed in the low susceptibility measurements corresponding to a

highly ordered fine particles system.

Magnetic metal particles experience strong Van der Waals attractions which,

combined with their magnetic dipole interactions, makes the stabilization of

these systems very challenging. Highly ordered structures of nanocrystals such

as two or three dimensional assemblies are also interesting because these

nanocrystals could act as building blocks in future devices for potential

applications in nanoelectronics and spintronics.

In this work it was observed the power of spontaneous SA when multiple forces

are balanced (including isotropic and anisotropic) which reminds the observed

complex biological self organization able to reconstruct a living cell after

duplication of its components during mitosis. This is relevant, not only to build

up super-structures for specific applications, like compact monolayers for

7

magnetic recording media, but also to elucidate the behaviour of magnetic

nanoparticles in complex liquid media, that are often out of equilibrium, as in

biology. Thus, a minimal amount of solvent has to be present even in the late

stages of the SA, at whose moment the liquid film breaks into droplets that later

evaporate. Even if this has an impact on the SA, it is clearly not the driving

mechanism in this case nor in the other observed formation of super-crystals of

nanoparticles.

Synthesis

Various methods have been developed for the synthesis of colloidal

nanocrystals.5-7 In particular, the thermal decomposition of metal carbonyls is

known to produce well-defined metallic nanocrystals.8 The properties of the

individual particles as well as their mutual interactions determine important

features of the nanoparticles systems. Since magnetic properties are highly

dependent on the size, shape, crystallinity and surface state of the

nanocrystals, their controlled synthesis with a narrow size distribution and

uniform shape remains an important issue. By systematically increasing the

particle size, TEM images show an abrupt transition from SA monolayers to

randomly oriented linear aggregates and branched chains or networks,8

witnessing the rise of stability and intensity of the magnetic moment (Figure

3.1).

Figure 2.3: General scheme of nanoparticles synthesis by thermal descomposition

In this work, spherical Co NPs in the epsilon (ε) crystalline face coated with

oleic acid were synthesized. Synthesis of ε-Co NPs was performed under argon

atmosphere at high temperature, which involved the thermal decomposition of

di-cobalt octa-carbonyl Co2(Co)8 in a mixture of a hot organic solvent

(dichlorobenzene) and surfactants (oleic acid and trioctylphosphine oxide),

according to a method previously reported.8,9

organometallic reagents

Nucleation Growth of particle by condensation

ATOMS

8

Co

Reagents

• Organometallic reagent: Co2(CO)8

• Surfactants: Oleic Acid (OA) (hydrophobic), Trioctylphosphine oxide

(TOPO)

• Organic Media: Dichlorobenzene (DCB)

Figure 2.4: General scheme of cobalt nanoparticles

2.3. SELF-ASSEMBLY

One of the most important reasons for the study of SA is its central role in

life. For example, the components of a cell replicate and assemble into another

cell during mitosis.

9

SA is a spontaneous organization of components into patterns and structures

without the human intervention, and is common throughout nature and

technology. This is one of the few practical strategies for making ensembles of

nanostructures and it will therefore be an essential part of nanotechnology. The

concept of SA is increasingly used in many disciplines with a different nuances

in order to break nature’s code of SA.10

SA structures are highly present in the nature. Peptide chains fold to form

proteins and enzymes, single-stranded DNA finds its complementary strand and

forms a double-stranded helix, and phospholipids align themselves in order to

make up cell walls. The process of SA is facilitated by both specific molecular

interactions and the drive to minimize the interaction energy between

molecules. In addition, it is the only practical approach for building a wide

variety of nanostructures.

Figure 2.5: A) A self-assembled peptide amphiphile nanofibers. B) A Self-assembly of gold nanoparticles.

Dispersed colloidal nanoparticles (NPs) self-assemble into complex structures

when segregated from the solvent either by evaporation or precipitation.

Different micro and macroscopic structures like opals, fractals, mixed structures

and other forms based in NPs have been observed11 as a result of the balance

between electrostatic forces, surface tension, entropy, topography, substrate

affinity and, most importantly, the size, shape and concentration of the

A B

10

particles.12 This is a key issue for applications and for fundamental

investigations.

Different effects involved in the SA process of magnetic nanoparticles such as

proximity (matrix), long-range interactions, long-range order, domain walls,

hysteresis, collective behavior, percolation, correlated local and global

perturbations, among others, are subjects of current investigation. In this

context, magnetic dipolar interactions that are not screened or interfered by the

medium are an interesting piece of puzzle to understand and control such

processes. Making sure that the components assemble themselves correctly is

not an easy task, because the forces at work are so small. Self-assembling

particles can get trapped in undesirable conformations, making defects that are

impossible to avoid.

Figure 2.6.: Example to understand the SA process.

Most synthesized particles contain a surfactant layer, which maintain them

separated some nanometers. This means that interactions between

nanoparticles, even in these dense assemblies, is negligible and their mutual

interactions are, therefore, dominated by long-range dipole-dipole interactions.

For nanostructured materials built from magnetic nanometric units, dipolar

interactions play an important role in determining cooperative phenomena at the

molecular scale and the final properties of the material.13 Dipolar materials

display a for arranging themselves into highly non homogeneous structures.

This is a consequence of the very strong anisotropy of the dipole-dipole

interaction, which couples the orientations of the dipole moments with that of

11

the interparticle vector that is different from the isotropic Van der Waals

interaction.

For the past few years, various methods have been developed for the synthesis

and the SA of magnetic nanocrystals. For example, Gao et al.14 produced

uniform 3D super-structures of spherical cobalt nanocrystals by the interplay

between dipolar interaction and applied magnetic field. Furthermore, Yin et al.15

reported three different kinds of self-assemblies of cobalt nanocrystals forming

different structures attributed to the magnetic properties of cobalt nanoparticles

with different size. Zeng et al.13 showed how SA of mixtures of magnetic

nanoparticles make it possible to produce materials with excellent magnetic

properties. Legrand et al.16 reported the formation of large-scale 3D

superlattices of cobalt nanocrystals, applying a magnetic field perpendicular to

the substrate during the deposition onto HOPG substrates. Sevchenko et al.17,18

obtained faceted triangular and hexagonal platelets, pyramids and tetrahedron

with sizes between 10-30 µm, consisting of individual monodisperse FePt

nanocrystals aligned in a 3D superlattice. Other examples include the formation

of micrometric supercubes from nanometric iron cubes,19 all of them yielding

different structures.

In this study it was observed the influence of the substrate during the

evaporation process, the NP-Substrate interactions and the formation of SA

super-structures of magnetic NPs in the absence of an applied magnetic field.

The study of the SA processes of cobalt nanoparticles coated with oleic acid

onto HOPG and comparing with the SA on a carbon coated TEM grid, gives us

the opportunity to study in detail the balance between NP-NP and NP-substrate

interactions in the particular case where both particles and substrate are highly

hydrophobic. The micro self-assembled structures resulting from the

evaporation of a solution of superparamagnetic (SPM) Co NPs onto a HOPG

substrate create a framework to study the magnetic properties of the material

since self-assembled structures reflect information coded (as shape, surface

properties, charge, polarizability, magnetic dipole, mass, etc.) residing in its

individual components.

12

2.4. CHARACTERIZATION

2.4.1. Optical Microscope The optical microscope, often referred to as the "light microscope", is a

type of microscope which uses visible light and a system of lenses to magnify

images of small samples (micron sizes).

An optical microscope Zeiss observer z1m was used in our studies.

2.4.2. Transmission Electron Microscopy The transmission electron microscope (TEM) operates on the same basic

principles as the light microscope but uses electrons instead of light. What you

can see with a light microscope is limited by the wavelength of light. TEM

overcomes the limitations from the wavelength of light by using electrons as

"light source" making it possible to get a resolution a thousand times better than

with a light microscope. Objects can be observed down to a few angstroms (10-

10 m).

TEM is a microscopy technique whereby a beam of electrons is transmitted

through an ultra thin specimen, interacting with the specimen as they pass

through. An image is formed from the interaction of the electrons transmitted

through the specimen, which is magnified and focused onto an imaging device,

such as a fluorescent screen, as is commonly used in most TEMs, on a layer of

photographic film, or to be detected by a sensor such as a CCD camera. TEM

images show contrast in color due to the absorption of electrons in the material,

the thickness of the film and the composition of the material.

A Transmission Electron Microscopy JEOL 1010 with a Bioscan (Gatan) digital

imaging system is used throughout our studies.

Figure 2.7: Transmission Electron Microscopy scheme

13

2.4.3. Scanning Electron Microscope - EDX The scanning electron microscope (SEM) is a type of electron

microscope that images the sample surface by scanning it with a high-energy

beam of electrons in a raster scan pattern. The electrons interact with the atoms

that make up the sample producing signals that contain information about the

sample's surface topography, composition and other properties such as

electrical conductivity.

X-rays, which are also produced by the interaction of electrons with the sample,

may also be detected in an SEM equipped for energy-dispersive X-ray

spectroscopy.

The morphology and chemical composition of the Rice-Grain like structures

were characterized with an Environmental Scanning Electron Microscope. The

images were taken on an ESEM Quanta 200 FEI, XTE 325/D8395 with a beam

energy of 10 kV and 15 kV.

Figure 2.8: Scanning Electron Microscopy scheme

2.4.4. SQUID Magnetometry SQUIDs, or superconducting quantum interference devices, measure

extremely small magnetic fields; They are very sensitive vector magnetometers,

with noise levels as low as 3 fT·Hz−0.5 in commercial instruments and 0.4

fT·Hz−0.5 in experimental devices. A SQUID magnetometer (MPMS, Quantum Design) was used to record the

magnetic signal (in-plane and out-of-plane) of a Co NPs monolayer. In order to

14

check if the particles were oxidized, and the effect of that oxidation on the

magnetic behaviour, hysteresis loops were performed after 1-T field cooling at

50 K, a temperature lower than the Neel temperature for cobalt oxide (CoO)

(corresponding to the ordering of the antiferromagnetic phase). CoO–Co

coupling would result in either an exchange coupling, which would increase the

anisotropy and thus the coercivity, and/or exchange bias, which would induce

an x-shift in the loops after field cooling.

Figure 2.9: SQUID Magnetometer scheme

2.4.5. Atomic Force Microscope (AFM)/Magnetic Force Microscope (MFM) The Atomic Force Microscope (AFM) or scanning force microscope

(SFM) is a very high-resolution type of scanning probe microscope with

demonstrated resolution of fractions of a nanometer, more than 1000 times

better than the optical diffraction limit.

The AFM is one of the foremost tools for imaging, measuring and manipulating

matter at the nanoscale. The information is gathered by "feeling" the surface

with a mechanical probe. Piezoelectric elements that facilitate tiny but accurate

and precise movements on (electronic) command enable the very precise

scanning.

The AFM consists of a microscale cantilever with a sharp tip (probe) at its end

that is used to scan the specimen surface. The cantilever is typically silicon or

silicon nitride with a tip radius of curvature on the order of nanometers. When

the tip is brought into proximity of a sample surface, forces between the tip and

the sample lead to a deflection of the cantilever according to the Hooke law.

15

Normally, the deflection is measured using a laser spot reflected from the top

surface of the cantilever into an array of photodiodes.

Figure 2.10: Atomic Force Microscopy scheme

The MFM is a type AFM. Unlike typical AFM, magnetic materials are used for

the sample and tip, so that not only the atomic force but also the magnetic

interaction is detected. Many kinds of magnetic interactions are measured by

MFM, including magnetic dipolar interaction.

16

17

A B

C D

3. RESULTS and DISCUSSION

3.1. SYNTHESIS OF COBALT NANOPARTICLES A concentrated solution of Co2(CO)8 (0.54 g in 3 ml of anhydrous DCB)

was injected into a mixture of OA in DCB in anhydrous DCB (~15 ml) at reflux

temperature (181 °C). The decomposition and nucleation occurs

instantaneously upon injection. The lifetime of atoms in solution is short leading

to the simultaneous formation of many small metal clusters. The surfactant is

present in the reaction mixture at concentrations of about 1%. Mixture of OA

and TOPO were used. The control over the temperature and the surfactant

concentration modifies the strength of the metallic particle–organic molecule

bonding. Thus, by controlling the precursor/surfactant ratio, the reaction

temperature and the injection time, the size of the spherical particles can be

tailored and varied between 3 and 17 nm. This method produces macroscopic

quantities of Co single crystals that are nearly monodisperse. As shown in

Figure 3.1, Co NPs of 6, 8, 14 and 17 nm were obtained and characterized by

TEM.

Figure 3.1: TEM images of Co nanoparticles with sizes of 6 nm (A), 8 nm (B), 14 nm (C) and 17 nm (D).

18

Given that the majority studies of self-assembly of such magnetic nanoparticles

have been developed onto carbon coated TEM substrates, the objective of this

work was the study the SA onto a more technological interesting substrate.

HOPG presents a set of favorable properties: smooth, flat and clean surface

and hydrophobic character that make it compatible with OA and DCB that

increases the affinity between the solvent and the substrate.

The Co NPs are spherical (no shape anisotropy) and with epsilon (cubic) crystal

anisotropy (six easy magnetization axes). In these assemblies, dipolar

interactions are strong, as they are systems of strong magnetic moment and

low and degenerated anisotropy. In this context, gravity force for a 10 nm Co

nanoparticle is 0.04 fN, while the force associated to the Brownian motion at

300 K is 4 fN20.

3.2. DEPOSITION ONTO HOPG

The deposition of NPs onto a HOPG substrate was performed by drop

casting. Were deposited a few drops of the initial solution (1016 particles per ml)

on a 5x5 mm of freshly cleaved HOPG. The excess of the solution was

removed with the aid of an absorbent paper, leaving the entire surface evenly

coated. The substrate was covered with a petri dish during 30 minutes to

facilitate a slow initial evaporation under a DCB atmosphere. After this time the

substrate was uncovered and the evaporation process continued until total

dryness of the solvent.

Spherical Co NPs self-assembled during evaporation from a colloidal solution

onto graphite. While the NPs on the substrate were still mobile, they tend to

arrange according their lowest energy sites. Equilibrium structures of NPs are

determined by a balance between Van der Waals attractive forces on one side

and vibrational entropy of the particles and steric repulsion of surfactant

molecules on the other side. Dipolar magnetic interactions can significantly

modify the balance to promote SA of NPs into nanometric or microscopic

objects of different shape and size.

19

Slow evaporation at room temperature of the DCB (181 ºC) resulted to a

spontaneous formation of micrometric cobalt rice-like structures in the case of 6

nm NPs (Figure 3.2) and wire-like structures in the case of 8 nm NPs (Figure

3.3), which were first observed by optical microscopy.

Figure 3.2: A) Optical microscope image of 4-8 μm cobalt like-rice structures. Insert is a zoom view of the image. B) Size distribution analysis of cobalt like-rice structures.

Figure 3.3: Optical microscope image of cobalt like-wire structures. Insert is a zoom view of the image. The size distribution analysis of cobalt like-rice structures reveals a relative

narrow size distribution, around 1-2 μm and a length that varies between 4-8

μm (figure 3.2B). The observed (Gaussian and narrow) size distribution recalls

the LaMer description of homogeneous nucleation and growth of monodisperse

NPs in solution known as the “burst nucleation” of a supersaturated solution.21

When the size of the particles was increased the formation of chains of particles

were observed, due to the fact that are moving to the ferromagnetic regime. It is

20

A B

0 2 4 6 8 10 120

5

10

15

20

25

30

Cou

nts

of N

PD istance (µm )20 µm

20

A

C C D

B

well known that for large ferromagnetic NPs, dipolar interactions favour

formation of chains, which are observed in the wire-like structures on the HOPG

substrate.

SEM images (Figures 3.4 and 3.5) show the cobalt structures in white color.

The corresponding EDX analysis of the structures indicates that they are

formed by CoNP. The background of the picture corresponds to the HOPG

substrate.

Figure 3.4: Back Scattered SEM images of 4-8 μm cobalt like-rice structures (A and B). EDX analysis of the cobalt like-rice structures (C) and the HOPG substrate (D).

Figure 3.5: A) SEM image of 4-8 μm cobalt wire-like structures. B) EDX analysis of the cobalt like-rice structures.

A B

21

Figure 3.4 shows a distribution of random orientations of the micrometric grains.

The fact that the majority of the structures are isolated suggest that these

assemblies did not form first in solution and then sedimented. Rather, it

indicates that these structures formed on the substrate at the last stages of

evaporation. In areas where the density of structures is higher, they appeared

as intersected structures rather than collapsed. These observations together

with their flat top surface suggest growth from the substrate.

It could be proposed that the particles self-assemble in solution at very high

concentrations (prior to deposition) when the majority of the solvent has been

evaporated but still loosely bind to the substrate, and later the structure is

transferred to the substrate when the evaporation is completed. However, this

simple scenario was discarded by the work of Liangshi et al.,22 in which different

structural patterns and dynamics were obtained with different substrates,

indicating that, apart from concentration increasing, substrate affinity also plays

a crucial role.23 This issue can be further corroborated by investigating the effect

of NPs mobility reduction by depositing them onto substrates with different

affinities. In the work of Puntes et al.,24 a similar sample was evaporated onto a

(111) single silicon crystal functionalized with an amine self-assembled

monolayer obtaining disordered Co monolayers. Amines are known to strongly

bind to Co and, therefore, reduce particle mobility onto such a substrate. These

results suggest that in order to form ordered structures a certain mobility is

needed once the nanoparticle reaches the substrate or the growing crystal. The

mobility of the particles is related to its size.33 Therefore, in the case of bimodal

size distributions, particles self-assemble with similar ones and the smaller ones

appear to be beneath the larger, as a result of the different mobility.

Regarding the wire-like structures, we can observe that they are not uniform

structures. SEM images at high magnifications show us that they are

constituted by small substructures forming the chain.

The use of high boiling solvents allows slow evaporation at high temperature.

This added thermal energy permits the particles to diffuse to their lowest energy

sites during evaporation, producing a well-defined superlattice structure.

22

3.3. REGARDING THE FORMATION MECHANISM

It is obvious that SA does not occur either in the dispersed colloid nor in

the colloid vessel walls (these colloids are stable for months to years). Besides,

a minimal amount of solvent is needed in order to start self-assembling, which

lasts until there is no more solvent to allow particles to move and then, the

structure freezes. It is also likely that particles feel an effective attraction by the

substrate. Therefore, as particles arrive near the substrate, they temporally

attach on it before going back to the solution. When the concentration is high

enough, the mobility of the particles on the substrate surface is reduced and

then SA starts, due to both a reduction of unavailable positions for the NP and

an increase in the self-assembled structure stability.

Smooth grains supports the idea that particles are able to move during structure

formation, i.e., there is still solvent left in contact with the substrate, which is

consistent with the previous observations and the flat tops observed by AFM

(Figure 3.6)

4.0µm

Figure 3.6: AFM image of rice-like structures integrated by cobalt NP.

SA requires the components to be able to move with respect to one another.

Their steady state positions balance both the attraction and repulsion. Because

SA requires that the components be mobile, it usually takes place in fluid

phases and on smooth surfaces. The environment can modify the interactions

between the components. If components stick together irreversibly when they

collide, they form a glass rather than a crystal or other regular structure. SA

23

requires that the components equilibrate between aggregated and non-

aggregated states, or adjust their position relative to one another once they

aggregate.

These facts, together with a high affinity of the particles to attach on the

substrate, favor 2D over 3D SAs enough to construct flat objects with their

larger surface in contact with the substrate. This balance between NP-NP and

NP-substrate affinity may be modified by varying the hydrophobic/hydrophilic

character of the substrate. In our case, both particles and substrate are highly

hydrophobic, although the presence of surfactant molecules results in a steric

repulsion between particles and a lower NP-NP affinity than NP-substrate,

favoring flat structures. Same experiments onto freshly cleaved mica (a clean

and flat, but hydrophilic substrate) did not yield any particular SA structure

(Figure 3.7).

Figure 3.7: Optical image of cobalt nanoparticles deposited on MICA (scale bar 20µm).

Isolated medium size Co NPs (6 nm) are superparamagnetic at RT. Figure 3.9

shows susceptibility curves obtained after a zero field cooling (ZFC)-field

cooling (FC) process in a magnetic field of 100 Oe. A maximum in the ZFC

branch is observed around the blocking temperature TB≈130 K. The broadness

of the peak indicates that NPs present a wide size distribution function. A

second peak is observed in both the ZFC and FC branches at T≈45 K. We

tentatively assign this peak to the presence of some cobalt oxide at the NP

surface that has the Neèl transition close to this temperature. Despite the

existence of the protecting oleic acid layer, a small degree of oxidation

attributed to the surface can not be avoided when samples are exposed to air,

as previously described.9 In fact, AC susceptibility measurements (Figure 3.8)

do not show any frequency dependence of the position of this peak thus,

20

24

indicating that it corresponds to a true thermodynamic transition. The

paramagnetic upturn observed in the low temperature regime may be due to

the existence of very small Co NPs with a very low blocking temperature or to

the existence of paramagnetic Co atoms coordinated to OA or TOPO.

Figure 3.8: Temperature dependence of AC susceptibility for 6 nm colloidal NP

Figure 3.9: ZFC-FC measurements. A) Dispersed 6 nm Co nanoparticles and B) rice-shape grains

Figure 3.10 shows the M(H) curves of dispersed NP at RT and at T=10 K. At

RT, M(H) exhibits the typical features of superparamagnetic behavior and can

be described by using the Langevin paramagnetic function M(x)=coth x-1/x,

being x=µH/kBT, for a classical Heisenberg spin of moment µ in a magnetic

field H. The values obtained for µ are around 104 µB, that assuming the density

and moment per Co atom corresponding to bulk fcc Co,25 allows an estimation

of the particle size of about 6 nm, similar to that obtained from TEM that also

includes the antiferromagnetic (AFM) cobalt oxide layer at the surface of the

particles. At T=10 K M(H) curve exhibits a well developed hysteresis loop with a

coercive field HC≈ 2 kOe. It is also worth mentioning that after a FC in a high

25

field the loop is off-set from zero indicating the existence of an exchange bias

field (HEX≈ 0.5 kOe) likely generated by the CoO layer at the surface of the Co

NP.26 On the other hand, the remanence to saturation ratio is about ½ in

agreement with the Stoner-Wolfart model for non interacting particles.

-1

-0.5

0

0.5

1

-2 104 -1 104 0 1 10 4 2 104

C olloida l NP

Self-Assem bledM

/Ms

H (O e)

R T

B

-1

-0.5

0

0.5

1

-30 -20 -10 0 10 20 30

Self-Assembled

Colloidal NP

M/M

s

H (kOe)

10KA

Figure 3.10: Hysteresis loop for colloidal and self-assembled Co NP. A) low temperature (10ºK) and B) room temperature (RT)

The magnetic characterization of self assembled NP is shown in Figures 3.9. In

this figure we depict the ZFC-FC susceptibility curves. They look very much

alike to that of the dispersed NP. A broader peak is observed at the blocking

temperature TB. It is worth mentioning that in this case TB is slightly smaller that

for dispersed NP, which is contrary to that found in other cases in which a small

increase of TB use to be observed due to interparticle interactions. Figure 3.10

shows M(H) curves a RT and T=10 K. As in de case of dispersed NP, at RT

M(H) exhibits the typical features of superparamagnetic behavior and can be

described by using the Langevin paramagnetic function with similar values of

the magnetic moment per NP (µ≈104 µB). Nevertheless at low temperature the

hyteresis loop is quite different, the coercive field has been drastically reduced

to HC≈ 380 Oe as well as the exchange bias field HEX≈ 100 Oe. These effects

are usually attributed to interparticle interactions. NPs in self assembled objects

are tightly packed and interparticle interactions may strongly affect the

magnetic behavior. In particular, dipolar interactions between NP can frustrate

the orientation of the moments along the NP’s easy axis at low temperature

promoting the appearance of a collective glassy behavior that drastically

reduces the coercive field. The existence of strong magnetic interactions

26

between NPs in self assembled nanoobjects is also evidenced by the

remanence to saturation ratio that is clearly smaller than ½.

At 10 K smaller coercivity was observed in the case of the Co structures

probably due to the demagnetizing character of the dipolar interactions27-29. The

hysteresis observed at 10 K is only in part due to the particles themselves, as it

is larger than that observed in diluted nanoparticle colloids, indicating that

interactions are partially responsible for the coercivity. In disordered systems,

dipolar interactions tend to reduce the coercivity, whereas in ordered systems

the dipolar interactions increase it.

For an isolated Co nanoparticle with diameter of 6 nm and anisotropy KCo(fcc) =

8 x 104 J/m3 the estimated blocking temperature TBSP = KV/25kB (where V is the

volume) should be about 40 K.

Considering the extremely short-range character of exchange interactions and

the fact that the cores are isolated by at least twice the surfactant thickness, the

possible coupling of Co through such interactions can be ruled out.

However when such a particle is in the vicinity of other particles, interparticle

dipolar interactions develop. The maximum dipolar field, 3

0 00 3 3 32 2 (2 )

at atdip

N RHr a R l

μ μ μ μμπ π

= =+

(where μ0 is the permeability of vacuum, N is

the number of atoms per particle, μat is the atomic magnetic moment, R is the

particle radius, a is the apparent atomic radius, l/2 is the surfactant thickness),

created along the axis between two particles in contact amounts to about 0.7 T.

The energy of 6 nm Co particle (particle moment Nμat ≈ 1.10-4 μB) in a field of

0.7 T is much higher than room temperature. The above oversimplified

considerations suggest possibility for thermally (enough) stable magnetic

moments of the particles during the solvent evaporation and hence for their self

assembling. Thus, the intuitive picture of self organization scenario is in the

Figure 3.11. Since the dipolar interactions favor formation of chains of particles,

“nucleus” of such chains will spontaneously appear during the initial stage of

solvent evaporation and they will further grow by attracting particles which are in

27

the vicinity of the chains ends. At the same time, other particles which are

positioned far from the chain ends won’t be able to join the chain. Instead they

will try to move towards the nearest particle of the chain, orienting their

magnetic moments in opposite direction. Thus chains with progressively shorter

lengths will be eventually attached to the original “nucleus” resulting in 3-D “rice-

shaped” objects. Naturally, the coercivity and the remanence of such “rice-

shaped” objects formed by dipolar forces, being macroscopic

“antiferomagneets”, will be close to zero. Of course, the interplay of several

parameters (like particle concentration, particle magnetic moment, solvent

viscosity and evaporation conditions) will all play a role in determining whether

the particles will self assemble into “rice” or chains, confine in layers or they will

not assemble at all.

Figure 3.11: A) Scheme of the self-organization process. B) For large ferromagnetic nanoparticles dipolar interactions favor formation of chains of particles.

High

Low field

Low field

Low field

Low field

Low field

Low fieldLow field

A

B

28

Magnetic Force Microscopy (MFM) measurements show that the observed Co

structures are magnetic (Figure 3.12). No magnetic structures are observed

inside the microstructures (black corresponds to attraction). Neither variation of

the contrast is observed varying the orientation of the MFM tip magnetization

corroborating the absence of remanence and coercivity. One also has to bear in

mind that before the MFM scan, where the tip approaches the sample at

approximately 50 nm, the tip is scanned at 5 nm to record the surface structure.

Thus, the sensing magnet of the tip is first scanned very close to the particles in

what remains a write (AFM)–read (MFM) process.

Figure 3.12: MFM image of some of these structures showing the magnetic character at room temperature.

29

4. CONCLUSIONS There have been numerous experimental studies of closed-packed

magnetic NPs arrays, yet proof of dipolar ferromagnetism has been eluded

because the patterns have not previously been observed experimentally. In

previous works, ferromagnetic-like (collective) behavior in magnetization

measurements of a 2D self-assembled monolayer of 9 and 12 nm diameter Co

NPs has been observed.27-29 They demonstrated the formation of micron-sized

magnetically ordered regions and transverse domain walls, with a spatially

varying local order parameter.

Dipolar ferromagnetism was first predicted by Luttinger and Tisza,30 who found

a ferromagnetic ground state for a face-centered-cubic (fcc) lattice of point

dipoles. We have previously observed that monodisperse Cobalt nanoparticles

with sizes that lead to a blocking temperature close to RT (therefore showing

significant dipolar interactions) tend to display correlated magnetic areas in a

domain-like type structure. In the present work, it is studied how such structures

are formed and how those dipolar interactions drive the SA into elongated

super-structures with a narrow size distribution, from micrometric rice-grain-like

structures to micrometric wire-like structures. The properties of the individual

particles as well as their mutual interactions determine important features of the

nanoparticle systems, such as storage capacity and switching field. A Full

understanding of nanoparticle SA would allow the design and construction of

complex modular materials with customized properties and would bring material

science to a higher stage of potentiality. A complete magnetic characterization

of both individual particles and self assembled structures as a function of

magnetic field and temperature is reported here. The results also indicate new and unique features of purely dipolar

ferromagnets. There is a small local order parameter even at RT, 200 K higher

than the temperature where the apparent coercivity drops to zero. The minimal

hysteretic losses and the ability to tune the ordering temperature with particle

size may be beneficial for designing nanocomposites. At the same time, if a

30

dipolar ferromagnetic state exists it may prompt further investigation of two-

dimensional critical behavior, and of spin waves, which would be important for

high-density patterned media.

NPs coated with organic surfactants are able to self organize into ordered

arrays due to the interplay between Van der Waals attractive forces and steric

repulsion. When the NPs are magnetic, magnetostatic forces due to dipolar

interactions, which are anisotropic, may play a relevant role influencing and

then prevailing over isotropic interactions. When magnetic NPs have a stable

enough magnetic moment (note that the superparamagnetic behaviour of an

isolated particle may decrease when dipole-dipole interactions appear)

depending on the relation between magnetostatic forces, and dispersive forces,

different self-assembled arrays may be generated. For instance, when

magnetostatic forces prevail over dispersion forces, particles form chains. In

contrast, when magnetic interactions are weaker than the other isotropic

interactions, close-packed structures are formed.31,32 Van der Waals

interactions and steric forces can be modulated by using different types of

surfactant while dipolar magnetic interactions can be modified by changing the

size of the magnetic NPs.

In the case of these studies, the aim is to modulate the contribution of dipolar

magnetic interactions to the self assembling process onto HOPG by using Co

NPs of different sizes. Smaller NPs, around 4 nm in diameter, do not show any

ordered SA. Medium size NPs around 6-8 nm in diameter self-assemble into

regular ellipsoidal grains of about 6 µm in length (for diameters ≈ 6 nm) and

chains of about 2 µm wide and variable length (for diameters ≈ 8 nm). Finally

large NPs of around 13-14 nm in diameter do not show any type or regular

arrangement. The focus of this work is addressed to medium size NPs since

only they exhibit clear evidence of SA.

Even when isolated NPs are superparamagnetic, above TB, groups of NPs can

have a non-zero remanence and local ferromagnetic ordering. Thus, once the

SA has started, NPs close to the growing objects will experience significant

dipolar interactions. The theoretical SA process to generate objects similar to

that found in our case can be found in reference 3131. Dipolar interactions

31

between NPs would initially favor parallel alignment of the moments forming a

monolayer. Nearby NPs will experience a field favoring parallel alignment with

the monolayer moment. Once a second small layer is formed on the top of the

first one, the magnetic moment direction of the second layer of NPs is locked in

by the large in-plane Lorentz cavity field. Additional particles arriving at the

surface of the bilayer will experience a ferromagnetic field from the layer below

and antiferromagnetic from the first layer. In this way, the formation of the

ellipsoidal objects, due to the existence of dipolar magnetic interactions, can be

understood. Obviously, the appearance of such objects and their size depends

critically on the balance between the different forces acting on the NPs

assembly. In fact we have observed that a small variation of the NPs size

(variation of the NPs mean diameter from ≈ 6 nm to ≈ 8 nm) ends up in the

formation of chains instead of ellipsoidal objects. Some of the chain structures

seem an intermediate state where the grains started thinning and aligning

nucleating a wire, leaving curly elongated structures, suggesting that the chain

forming process is based on the coalescence of elongated structures, all

consistent on the increase of magnetization per particle.

32

33

5. REFERENCES (1) Black, C. T.; Murray, C. B.; Sandstrom, R. L.; Sun, S. H. Science 2000, 290, 1131-1134. (2) Sun, S. H.; Murray, C. B.; Weller, D.; Folks, L.; Moser, A. Science 2000, 287, 1989-1992. (3) Pankhurst, Q. A.; Connolly, J.; Jones, S. K.; Dobson, J. Journal of Physics D-Applied Physics 2003, 36, R167-R181. (4) Dobson, J. Drug Development Research 2006, 67, 55-60. (5) Sun, S. H.; Murray, C. B. Journal of Applied Physics 1999, 85, 4325-4330. (6) Peng, X. G.; Manna, L.; Yang, W. D.; Wickham, J.; Scher, E.; Kadavanich, A.; Alivisatos, A. P. Nature 2000, 404, 59-61. (7) Jana, N. R.; Gearheart, L.; Murphy, C. J. Langmuir 2001, 17, 6782-6786. (8) Puntes, V. F.; Krishnan, K. M.; Alivisatos, A. P. Science 2001, 291, 2115-2117. (9) Puntes, V. F.; Zanchet, D.; Erdonmez, C. K.; Alivisatos, A. P. Journal of the American Chemical Society 2002, 124, 12874-12880. (10) Whitesides, G. M.; Grzybowski, B. Science 2002, 295, 2418-2421. (11) Murray, C. B.; Kagan, C. R.; Bawendi, M. G. Annual Review of Materials Science 2000, 30, 545-610. (12) Puntes, V. F.; Bastus, N. G.; Pagonabarraga, I.; Iglesias, O.; Labarta, A.; Batlle, X. International Journal of Nanotechnology 2005, 2, 62-70. (13) Zeng, H.; Li, J.; Liu, J. P.; Wang, Z. L.; Sun, S. H. Nature 2002, 420, 395-398. (14) Youhui, G.; Yuping, B.; Michael, B.; Akira, Y.; Daisuke, S.; Kannan, M. K.; AIP: 2004; Vol. 84, p 3361-3363. (15) Yin, J. S.; Wang, Z. L. Nanostructured Materials 1999, 11, 845-852. (16) Legrand, J.; Ngo, A. T.; Petit, C.; Pileni, M. P. Advanced Materials 2001, 13, 160-160. (17) Shevchenko, E.; Talapin, D.; Kornowski, A.; Wiekhorst, F.; Kotzler, J.; Haase, M.; Rogach, A.; Weller, H. Advanced Materials 2002, 14, 287-+. (18) Rogach, A. L.; Talapin, D. V.; Shevchenko, E. V.; Kornowski, A.; Haase, M.; Weller, H. Advanced Functional Materials 2002, 12, 653-664. (19) Dumestre, F.; Chaudret, B.; Amiens, C.; Renaud, P.; Fejes, P. Science 2004, 303, 821-823. (20) Gelbart, W. M.; Sear, R. P.; Heath, J. R.; Chaney, S. Faraday Discussions 1999, 112, 299-307. (21) LaMer, V. K.; Dinegar, R. H. 1950; Vol. 72, p 4847-4854. (22) Li, L. s.; Walda, J.; Manna, L.; Alivisatos, A. P. 2002; Vol. 2, p 557-560. (23) Rabani, E.; Reichman, D. R.; Geissler, P. L.; Brus, L. E. Nature 2003, 426, 271-274. (24) Puntes, V. F.; Gorostiza, P.; Aruguete, D. M.; Bastus, N. G.; Alivisatos, A. P. Nature Materials 2004, 3, 263-268. (25) Suzuki, T.; Weller, D.; Chang, C. A.; Savoy, R.; Huang, T.; Gurney, B. A.; Speriosu, V. Applied Physics Letters 1994, 64, 2736-2738.

34

(26) Skumryev, V.; Stoyanov, S.; Zhang, Y.; Hadjipanayis, G.; Givord, D.; Nogues, J. Nature 2003, 423, 850-853. (27) Luo, W.; Nagel, S. R.; Rosenbaum, T. F.; Rosensweig, R. E. Physical Review Letters 1991, 67, 2721. (28) Dormann, J. L.; Cherkaoui, R.; Spinu, L.; Noguès, M.; Lucari, F.; D'Orazio, F.; Fiorani, D.; Garcia, A.; Tronc, E.; Jolivet, J. P. Journal of Magnetism and Magnetic Materials 1998, 187, L139-L144. (29) Kechrakos, D.; Trohidou, K. N. Physical Review B 1998, 58, 12169. (30) Luttinger, J. M.; Tisza, L. Physical Review 1946, 70, 954-964. (31) Yamamuro, S.; Farrell, D. F.; Majetich, S. A. Physical Review B 2002, 65, 224431. (32) Korgel, B. A.; Fitzmaurice, D. Physical Review B 1999, 59, 14191. (33) Root Mean Velocity (RMS): V (RMS) = (3RT/M)0.5 R = 8.314 Joule/(Mole x Kelvins) T = Kelvin Temperature, M = molecular weight x 0.001 (to give kilograms/mole).

35

5. ACKNOWLEDGMENTS I want to tank Dr. Victor Puntes for his valuable discussions and

experimental support.

Also, to the members of the comitee for their time in reading this work.

The Inorganic Nanoparticles Group lab-mates, for the support and discussions.

Finally, I want to thank the Magnetic Materials and Functional Oxides group of

the Institut de Ciència de Materials de Barcelona (ICMAB) for all the magnetic

measurements and discussion of results.

This work is supported by the Misnisterio de Educación y Ciencia (MAT2006-

13572-C02-02).

36

37

6. ANNEX I We have studied three cases doing the deposition of 8 nm nanoparticles

under different magnetic fields.

A) The HOPG on top of a strong magnetic field (NdFeB)

B) The HOPG on top of a On a weak magnetic field (ferrite magnet)

38

C) The HOPG in a homogeneous magnetic field (between two ferrite

magnets)

These experiments confirm the magnetic character of the particles and that

cobalt NPs can self assemble into large wires depending on the substrate. The orientation of the wires can be manipulated with magnetic fields during the

evaporation process.

Picture of the edge of the surface, near the magnet.

Picture of the center of the surface, where there are homogenous field.

39