# Probing the Properties of the Molecular Adlayers on Metal Substrates: Scanning Tunneling Microscopy Study of Amine Adsorption on Gold(111) and Graphene Nanoislands on Cobalt(0001)

Contents

1

Introduction to STM and Applications of STM ………………………… … … 1

1.1

Principles of the STM ………………… .. …………………………………2

1.1.1

The Concept of Tunneling ……………… .. …………………………

2

1.1.2

The

Theory

of STM ……………… .. ………………………………… 6

1.2

The Introduction

of STS ……………… .. ………………………………… 8

1.2.1

The Concept

o f STS ……………… .. ………………………………… 8

1.2.2

The Theory of STS ……………… .. ………………………………… 10

1.3

Atom ic and Molecular

Manipulation

with the STM ……………… .. …… 11

1.3.1

Sliding Process ……………… .. ………………………………… .. … 12

1.3.2

Field - Assisted Diffusion ……………… .. ……………… ... ………… 14

1.4

Inelastic - Tunneling - Induced Manipulati on ……………… .. … . ……… . … 18

1.5

Experiment Setup ……………… .. ………………………… . …………… 19

1.5.1

STM

O perations ……………… .. ……………………… .. ………… ..19

1.5.2

Molecule D osing ……………… .. ……………………………… . … ..24

1.5.3

Tip P reparation ………… . …… .. …………………………………… .26

1.6

Bibliography ……………… .. ………………………………………… ..29

2

Au(111) Substrate and Pr eparation ………………………………………… 33

2.1

Introduction of G old and Au(111) S urface ……………………………… 34

ii

2.2

Au(111) Preparation …………………………………………………… ...41

2.3

Bibliography …………………………………………………………… ..44

3

Study of 1,4 - benzenedia mine Adsorption on Au (111) …………………… ..47

3.1

Abstract ……………………… ..

…………… ……………… ... ………… 47

3.2

Introduction ………………………… ..

…………………… . …………… 48

3.2.1

Introduction to

SAMs ………………………… ..

………………… ..48

3.2.2

Introduction to the Break - junction

C onductance M easurement … ..49

3.2.3

Introduction to 1,4 - benzenedimaine (BDA) ……………………… ..52

3.3

Experiment Methods ………………………… ..

………… …………… ..55

3.4

Results and Discussion ………………………… ..

…………………… ...56

3.4.1

Morphology of BDA on Au(111)

………………………………… ...56

3.4.2

Tip - induced movement

of BDA on Au(111) ……………………… ..69

3.5

Summary ………………………… ..

…………………………………… 76

3.6

Bibliography ……………………… ..

……………………… ... ………… 76

4

Graphene N ano isl ands on C o (0001) ………………………………………… 81

4.1

Introductions ……………… ..

…………………………… ..

…………… 82

4.1.1

Introduction to G raphene

and Graphene N anoribbons ……… . …… . 82

4.1.2

Introduction to Cobalt ( Co )

……………… ..

……………………… . 87

4.2

Experimental and Calculation M ethods ……………… ..

……………… ..91

iii

4.3

R esults

and D iscussion ……………… ..

…………………………… ..

… 93

4.3.1

Structure s

of Graphene N anoislands ……………… ..

……………… 93

4.3.2

Electronic P roperties ……………… ..

……………………………… 99

4.3.3

Edge S tates ……………… ..

…………………………… …

……… 105

4.4

Summary ………………… ..

…………………………… ..

…………… 106

4.5

Bibliography … .. ………… ..

…………………………… ..

…………… 1 06

iv

List of Tables

Table 1: Sele cted physical properties of BDA…………………… 53

v

List of Figures

1.1 The difference between classical theory and quantum theory. ……………… .. … 2

1.2: A one - dimensional metal - v acuum - metal tunneling j unction . …………… .. …… 4

1.3 Schematic picture of tunneling junction in Tersoff and Hamann analysis … … … .7

1.4 Energy diagram for the sample and the tip. . ………………… . …………… .. … .9

1.5 Schematic picture of the sliding process for a xenon atom. . ……………… … ..13

1.6 Schema tic of the potential energy of an adsorbate on the surface as a fun ction of the lateral position . ………………………………………………………………… 16

1.7 Schematic of the behavior of the polarized adsorbate in the nonuniform electric

field induced by the STM tip . ……………………… . …………… …… . ………… 17

1.8 Schematic depicting the essential steps in the theoretically modeling of single O2 mole cule dissociation on Pt(111).

. ………………… . …………………………… .18

1. 9: Schematics of STM operation.

. ………………… . ………………………… ...21

1.10: A front view and a side view of

Omicron LTSTM. . ………… ... …………… 23

1.11: Picture of 1,4 - benznediamine (BDA) dosing line. . ………………………… .26

1.12: Electrochemical etching of tungsten tips. . ………………… . ……………… 28

2.1: Schematic of the crystal structure of bulk gold and the surface of Au (111). …… … ……………………………………………………… . ………… ..34

2.2: Schematic of the top three layers of the fcc and hcp structures. …… ………… 36

vi

2.3: Hard sphere model of one stacking fault region in Au(111). …… …………… .37

2.4: STM topography of Au(111) surface. …………… ... ……………… ... ……… ..38

2.5: A tomic resolution STM topography of Au(111) herringbone elbow sites and ridge sites. ………………………………………………………………… .. ……… 39

2.6: Triangle reconstruction STM topography of Au(111). ………… . …………… .40

2.7: Large - scale STM topography of Au surface transition through sput tering and annealing …… ……………………………………………………………………… 43

3.1: Schematic of the modified STM in Venkataraman’s measurements. … ... …… .50

3.2: Conductance measurement of clean Au, BDA, 1,4 - benzenedithiol (BDT), and 1,4 - benzenediisonitrile (BDI) in Venkataraman’s measurements. …… ... ………… .52

3.3: Diagrams of BDA and two isomers. ……………………………… . ………… .53

3.4: STM topographic images of BDA on Au(111) at different coverages. ……… .58

3.5: Experimental results and models of individual BDA molecules on the Au(111) surface. ………… ……………………………………………………… . ………… ..59

3.6: STM images of the preferred adsorption sites of BDA molecules …… .. … .. … ..61

3.7: STM images of the preferred adsorption sites of BDA chain structures at low coverage. ………………………………………………………………… . …… . … ..63

3.8: Diagram of BDA molecule in the line structure. …………………… ... ……… .64

3.9: STM images of the preferred adsorption sites of BDA network structure ..........66

3.10: XPS spectra of N 1s and C 1s of BDA on Au(111) form a low to high

vii

coverage. ………………………………………………………………………… . . ..68

3. 11: Sequential STM images of the same area under different bias voltage and current. . …………………………………………………………………… . ……… ..70

3.12: Sequential STM images of the same area of the line lattice. . …… .. ………… .71

3.13: Sequential STM images of the same area before and af ter a voltage pulse . . .. . .72

4.1: Schematic of the band structure of graphene near Fermi level…………………………………………………………………………………85 4.2:

Schematic of the zigzag edges and armchair edges of graphene…………………………………………………………………………… . 86

4.3: Schematic of th e unit vect ors in hcp crystals……………………………… . … .88

4.4: Schematic of hcp(0001) structures ………………………… . ………………… .89

4.5: STM topographic image of Co(0001) ........................................ ……………… 92

4.6: STM topographic images of as - deposited and after thermal annealing

of HBC/Co(0001)……………… …………………………… . ……………………… ..94

4.7: STM topographic image of one graphene island and the corresponding height profile on Co(0001)……………………… ………………………………………… 96

4.8: STM topographic image of one graphene island and the schematics of possible e pit axial growth models on Co(0001)……………………………………………… 98

4.9: Experimental differential conductance spectra, dI/dV, of Co(0001) and graphene island/C o

……………………… ………………………………………………… . 100

viii

4.10: Calculated energy bands are density of states of graphene on Co(0001 ) in the AC geometry…………………………… ……………………………………… … 102

4.11:

Comparison of calculated PDOS and experimental STS data of graphene/Co(0001)………… ………………………………………………… ... … 103

4.12: STM topographic images, conductance map, and STS of graphene islands on Co(0001)……………… … ……………………………………………………… . ..104

ix

Acknowledgement

Seven years. Looking back, I found the journey as the most demanding task in my life. I was lucky to have so many people ’ s hands on my way to the mountain

top. Without them, I don ’ t think I w ould have finally made it.

I owe my greatest gratitude

to

my advisor, Tony Heinz. He has been a guiding landmark along the whole way, from the moment he admitted me to Columbia University to the moment he saw

me leave

after

defense. H e not only set

an ex ample of hard - working, but also showed the greatest

patience with his students. Moreover, I would not been able to find my life path without his encouragement and help during my final year as a student.

I also want to thank my husband, Dr. Jie Lin, for hi s biggest heart and support for whatever decision I make. He is my emotional pillow and he is always there for me.

There are two people who taught me the most in experiment techniques. Dr. Daejin Eom opened the door of STM to me and provided a continuous help with my experiment. Dr. Zonghai Hu, whom I still owe a paper, showed self less devotion into my projec t .

I am also grateful to a lot of o t h e r

professors and

colleagues in research. Prof. Flynn, who has been my unofficial advisor in the Chemistry Departm ent, provided a great sight

x

and continuous support for my project. Prof. Morgante taught me a lot of vaccum techniques and helped me settle down on my trip to Italy. Prof. Venkataramen and her student, Masha Kamenetska , along with Prof. Morgante ’

student,

Dr. Martina De ll'angela

filled me in with knowledge

of benzenediamine. Dr. Kwang Taeg Rim set a good example of consistency in research

and his optimi sm often encouraged me.

Prof. Abhay Pasupath y often provided me some insightful discussions.

Liuyan Zhao

has been a best friend and a great inspiration

inside and outside of the lab.

I wish to thank the following teachers and colleagues for their wisdom and expertise:

Prof. Andrew Millis,

Dr. Li Liu, Dr. Hao Wang, Dr. Hayn Park, Dr. Hugen Yan, Dr. Yang Wu, Dr . Daohua Song, Dr. Joanna Atkin, Dr. Kin Fai Mak, Dr. Joshua Lui, Dr. Sami Rosenblatt, Dr. G

ina Florio,

Dr. Mingyuan Huang

and

Dr. Scott Goncher.

Looking back, I realized that there are so many people I owe gratitude to. Thanks to all of them, my seven y ears become so precious. I will take the attitude and skills they taught me and continue my life journey.

1

Chapter 1

Introduction to STM and Applications of STM

The Scanning Tunneling Microscope (STM) was invented by Binnig and Rohrer in 1981 who shared the 1986 N obel Prize in Physics [6] . It is a powerful instrument to image surfac es at the atomic level ,

achieving a lateral resolution of 0.1nm and vertical resolution of 0.01nm [12] . At this scale, individual atoms on the material surfaces can be resolved. In this c hapter,

I will review the principles of the STM and its applications, focusing on Scanning Tunneling Spect roscopy (STS) and atom /molecule

manipulation with STM.

Chapter 1. Introduction to S TM and Applications of STM

2

1.1.

Principle of the STM

1.1.1.

The Concept of Tunneling

The operation of the STM is based on the phenomenon of quantum mechanical tunneling. Imagine a potential barrier and a microscopic particle, e.g. an electron, with

energy

smaller than the potential barrier , a s illustrated

in Figure 1.1 .

I n c lassical mechanics, the electron is con fined by the barrier region; while in quantum mechanics, Figure 1.1 The difference between classical theory and quantum theory.

When U>E , in c lassical m e chanics, the electron

cannot overcome a barrier; in q uantum m echanics, the electron has a non - zero probability to tunnel through the barrier

[2]

.

Chapter 1. Introduction to S TM and Applications of STM

3

the electron can tunnel through the barrier. To illustrate the tunneling, we consider a simplified one - dimen sional model shown in Figure 1. 2 . In this case, the electron is described by a wavefunction , which satisfies the one - dimensional Schrödinger Equation.

.

( 1 . 1 )

Its solution in the classical forbidden region is:

,

( 1 . 2 )

where

( 1 . 3 )

is the decay constant. Thus, the probability densi ty through the barrier is proportional to , where L is the barrier width. Here we define a transmission coefficient T, as the ratio of the transmitted current density and the incident current density [10] . In a simplified junction with equal potentials on the tip and sample In Figure 1.2b, T is given as :

.

( 1 . 4 )

Chapter 1. Introduction to S TM and Applications of STM

4

A typical value for 1/ is about 0.1nm. Thus, according to Eq 1.4 , the

current decays about 10 times, or one order of magnitude,

when the tip - sample distance changes by about 0.1nm. This gives STM spatial sensitivity on the atomic scale.

Figure 1.2 : A one - dimensional metal - vacuum - metal tunneling junction.

(a ) Electrons can tunnel through vacuum (modeled with a potential barrier) between the sample and the t ip with a decayed wavefunctio n [2] . (b) A simplified one - dimension model with equal potentials on both ends and a square potential barrier.

Chapter 1. Introduction to S TM and Applications of STM

5

In 1957, a more quantitative study was provided by La n d au er [13]

who used the semiclassical WKB approximation to calculate the tunneling probability. It has two important assumptions: First, the electrodes can be described by a one - dimensional free - elec tron gas in an ideal square potential well. Second, the current is ballistic ,

which means the current is the product of the density of the electron and the classical velocity

in the electrodes . When no

bias voltage

is applied ,

the electrons from either the

sample or the tip have equal

tunneling probability to th e other side through the vacuum

which leads to zero

net tunneling current. When a negative sample bias is applied, a difference of the Fermi levels of two electrodes is generated and a

net current fl ows . As a result, we observe a single - channel current of :

,

( 1 . 5 )

where T is the tunneling coefficient ,

which can be calculated in the WKB approximation.

T he tunneling conductance G is

defined as:

,

( 1 . 6 )

where the conductance quantum

is given by:

.

( 1 . 7 )

If we assume the vacuum can be approximated by a square potential barrier, then , and

Chapter 1. Introduction to S TM and Applications of STM

6

.

( 1 . 8 )

In practice, the transmission coefficient T is complicated by the tip geometry and the surface electronic structure [8, 9] . But the general tendency of remains the same.

1.1.2.

The

Theory

of STM

One widely used theory for the tunneling phenomenon in solids ,

as well as in STM is Barde e n Tunneling Theory. Bardeen first introduced the

following

tunneling approximation [14] :

,

( 1 . 9 )

where f(E) is the Fermi distribution function, V is the applied voltage, and is the tunneling matrix element between the states of the probe and those of the surface

(

denotes a state of one elect rode and denotes a state of the other) . Note that the reference point

of zero energy is the Fermi level.

is given

by

,

( 1 . 10 )

where and are the energies corresponding to and , respectively, and the integral is over the entire surface within the vacuum barrier region.

Chapter 1. Introduction to S TM and Applications of STM

7

By introducing the density of states of both electrodes, Eq 1.9 can be rewritten as:

,

( 1 . 11 )

where s denotes the sample, and t denotes the tip. M is the tunneling matrix element defined in Equation 1.10. This equation indicates that t he tunneling currently is directly related to the local density of states (LDOS) of the surface at the Fermi level.

Later on, further refinements were made for the theory

of STM . For example, Tersoff and Hamann [8, 9]

modeled the tip with an s - wa ve approximation ( ) for a spherical

Figure 1.3 Schem atic picture of tunneling junction in Tersoff and Hamann

analysis [8, 9] .

The tip is modeled as a spherical object with radius R and the center r o . The tip and sample distance is d.

Chapter 1. Introduction to S TM and Applications of STM

8

object and the surface with a general expansion o f

Bloch functions ( )

(See Fig. 1.3). They obtained the following simplified formula:

,

( 1 . 12 )

where is the density of states per unit volume of the tip. Through thes e advances, STM theory can be applied towards the understanding of some experimental data.

1.2.

The Introduction

of STS

1.2.1.

The Concept

of STS

Lang [15]

proposed the following picture at zero temperature

(Fig 1.4).

At

zero bias, the Fermi levels of the sample and tip are equal. When a positive sample bias is applied, a

net tunneling current arises from tunneling electrons from th e occupied states of the tip to the unoccupied states of the surface, whereas when a negative sample bias applied, there is a

net current tunnel from occupied states of the sample into unoccupied states of the tip. As a result, the polarity of the sample b ias determines whether the occupied states or the unoccupied states of the sample are probed. This is the idea of Scanning Tunneling Spectroscopy (STS).

Chapter 1. Introduction to S TM and Applications of STM

9

Figure 1.4 Energy diagram for the sample and the tip.

(a) Independent sample and tip. (b) Sample and tip at equilibrium, separated by small vacuum gap. (c) Positive sample bias: electrons tunnel from tip to sample. (d) Negative sample bias: electrons tunnel from sample into tip. [10, 11]

Chapter 1. Introduction to S TM and Applications of STM

10

In practice, we can obtain STS by observing the variation of the tunnel ing current I as a function of the bias V, while the tip is held at a fixed height over the sample. It provides an I - V curve which can be numerically differentiat ed to a dI/dV curve. Or we can also directly obtained dI/dV by using a modulation technique wi th a lock - in amplifier.

1.2.2.

The Theory of STS

Here we follow the analysis of Chen [2] . For simplicity, we only con sider the c ondition when the bias is small. Therefore, the tunneling matrix element can be approximated by a constant. We further assume that the density of states of the tip is independent of energy. Therefore, in Eq. 1.11, we can take both the matrix ele ment and the density of states of the tip out of the integral to obtain:

.

( 1 . 13 )

If the temperature is not very high, the distribution function satisfies

a step function, thus:

( 1 . 14 )

Combining Eq. 1.13 and Eq. 14, we can get:

,

( 1 . 15 )

Chapter 1. Introduction to S TM and Applications of STM

11

or,

.

( 1 . 16 )

Thus, within those approximations, the d ifferential conductance is directly proportional to the LDOS of the surface.

1.3.

Atom ic and Molecular

Manipulation

with the STM

Since the early 1990s, the application of STM has made possible manipulation of individual atoms and molecules on the surface. An

early demonstration of these capabilities was provided by Eigler and Schwerzer [1]

who constructed an IBM logo with individual Xe atoms on Ni(110). Using STM manipulation techniques, one can construct quantum structures on an atom - by - atom basis , synthesize

single molecules on a one - molecule - at - a - time - basis, and access single atom/molecule properties, as reviewed in Hla and Bai [12, 16] .

A variety of different atomic/molecular manipulation processes with ST M have now been realized. We can divide the processes into two classes in terms of the location of the manipulated atoms/molecules: parallel

processes and vertical processes [4] . In parallel processes, the motion of the adsorbed atoms/molecules is parallel to the surface, i.e., they remain on the surface. While in vertical

processes, atoms/molecules can be

Chapter 1. Introduction to S TM and Applications of STM

12

transferred

from the tip to surface, or vice versa. Here I will only focus on the parallel processes.

The parallel processes can be further divided into three categories by the mechanisms

of the motion: s liding

processes,

field - assisted diffusion, and inelastic - electron tunneling (IET) induced manipulation. The corresponding manipulation mechanisms are chemical forces, electric field s , and tunneling

electrons, respectively.

1.3.1.

Sliding Process

The STM tip always exerts a forc e on an adsorbate bound

to the surface. By adjusting the position of the tip, we can tune the magnitude and direction of the force. Thus, there is a potential to manipulate the adsorbate by pulling it across the surface. This is called a sliding process [4, 17] .

A typical sliding process involves three steps: (1) vertica lly approaching the tip toward the adsorbate until the tip - adsorbate is close enough, (2) sliding the tip across the surface with the same tip - adsorbate distance, dragging the adsorbate along, and (3) retracting the tip away from the

surface, leaving the a dsorbate on the surface. (See Figure 1.5) Since the first observed sliding process reported on Xenon/ Ni(110) [1] , it has been extended to many other systems, e.g. Fe/Cu(111) [18] , Co/Cu(111) [19] , and Pb on Cu(111) [20] .

Chapter 1. Introduction to S TM and Applications of STM

13

In most of those sliding process, the motion of the adsorbates is not sensitive to the sign or magnitude of the electric field, the voltage, or the current. It is only dependent on th e tip - sample separation. In the case of a silver atom sliding on Ag(111) [21] , the distance is 1.9Å between the edges of van der Waa ls radii of tip apex and the manipulated

adsor bate . At this distance, the atomic orbitals of the tip apex and the adsorbate are overlapping and a weak chemical bond is formed, which is believed to be the mechanism of the atom/molecule motion in the

sliding process.

Manipulation of single atoms/ molecul e s on surfaces allows the construction of artificial quantum structures and the study of the novel phenomena associated with these structures. In addition, the possibility

of constructing microscopic circuits has been Figure 1. 5

S c h e ma t i c p i c t u r e o f t h e s l i d i n g p r o c e s s f o r

a xenon atom [1] .

(a ) The adsorbate is located on the surface and the tip is placed directly over it. The tip is lowered to the position (b), where the adsorbate - tip attractive force is sufficient to keep the adsorbate located beneath the tip whe n the tip is subsequently moved across the surface (c) to the desired location (d) .

Finally the tip is withdrawn to position (e).

Chapter 1. Introduction to S TM and Applications of STM

14

demonstrated . For example, Heinrich co nstructed a type of logic gate with CO/Cu(111) [19] .

1.3.2.

Field - Assisted Diffusion

There is a strong ele ctric

field between the tip and the surface in the normal STM scanning. This may induce field - assisted diffusion. Let us

start by examin ing

how close the tip is located near the sample in our case. In the previous section (Eq 1.8), we obtained a simple re lation between conductance and the tip - sample distance: with

, where is a constant

of

order 1

in the limit of [2] .

is the tip - sample distance where

is defined when

the tip is in a single - atom contact with the sam ple .

is the decay constant . Take the vacuum level as the reference point of energy, , where

is the work function. For simplicity, we take the average of the work functions of th e tip and the sample, which is

typical around 5 eV . U is the applied bias and much smaller than the work functions. So

.

( 1 . 17 )

Under

scanning conditions of and , .

Then

.

( 1 . 18 )

Chapter 1. Introduction to S TM and Applications of STM

15

The tip - sample distance z obtained by this simple analysis

matches

STM measurements

[22 - 24] .

If we use a r ough estimation of , the electric field is roughly o n the order of 0.2 . Under

the real STM imaging condition s , due to the asymmetrical geometry, the electric field is inhomogeneous and concentrated in the vicinity of the tip. With a similar tip - sample distance and a bias of 1 to 10V, the electrical field strength is in the range from 0.2 to 2 , as obtained by simulation [22] .

Such

large electrical

fields approach those

required for the ionization or desorption

of an atom, which is around 3 to 5 . The fields may also be large enough for field - assisted migration of adsorbates.

When an atom/molecule is adsorbed on the surface, it

may have a static dipole

or a

dipole

induced

by the electric

filed. As a first approximation,

,

( 1 . 19 )

w here is the permanent dipole moment, is the polarizability

of the adsorbate .

As a result, the electric potential energy is given as

.

( 1 . 20 )

Chapter 1. Introduction to S TM and Applications of STM

16

This potential energy is added to the periodic potential of the surface (Fig 1.6).

Note that there are two terms in the equation. I f the second term dominates, then the adsorbate will a lways

to be attracted to the tip. When the first term dominates, the orientation of the dipole moment remains unchanged, causing the direction of the adsorbate motion to reverse with the bias polarity, as shown in Figure 1.7.

Figure 1. 6

Schematic of the potential energy of an adsorbate on the surface as a function of the lateral position [4] .

The tip is shown as the red solid object on the top. The interaction o f the adsorbate and the surface atoms gives rise to a periodic potential energy (shown in blue). The interaction of the tip - induced electric field and the adsorbate gives rise to a broad potential well. Adding these two potentials together leads to a broad

potential well located beneath the tip

with periodic

oscillations. The adsorbate is drawn as the green ball.

Chapter 1. Introduction to S TM and Applications of STM

17

A number of field - assisted diffusion experiments with STM have been reported. Depending on the relative magnitudes of the first (static dipole) and second (induced dipole) term in Eq 1.20, different

types of behavior have been observed [4, 7, 16, 23 - 30] . When the induced dipole dominates, the adsorbate will be attracted to the tip

irrespective of the polarity [29 ] . When the static dipole term dominates, such as thallium/Si(100) and In/Si(100) [7, 25] , the adsorbates diffuse in opposite directions at different bias.

For Cs atoms on GaAs and InSb(110) surfaces, the two terms

are of comparable magnitude and the diffusion

towards

the tip only takes place at positive bias [26] .

Figure 1. 7

S c h e m a t i c o f t h e b e h a v i o r o f t h e p o l a r i z e d a d s o r b a t e i n t h e n o n u n i f o r m e l e c t r i c f i e l d i n d u c e d b y t h e S T M t i p [ 7 ] .