b.sc. and m.sc. - eprints.qut.edu.au · b.sc. and m.sc. submitted in fulfilment of the requirements...

MOLECULAR REACTIONS ON SURFACES: TOWARDS THE GROWTH OF SURFACE-

CONFINED POLYMERS

Maryam Abyazisani

B.Sc. and M.Sc.

Submitted in fulfilment of the requirements for the degree of

Doctor of Philosophy

School of Chemistry, Physics and Mechanical Engineering

Science and Engineering Faculty

Queensland University of Technology

2019

To love of my life

Molecular reactions on surfaces: towards the growth of surface-confined polymers i

Abstract

The development of surface-confined nanostructures is a promising approach

towards developing new electronic and optoelectronic devices. Molecular assembly

under ultra-high vacuum (UHV) conditions offers precise control over the reaction

conditions. Despite all the control provided by the UHV, achieving high quality and

well-ordered nanostructures in both one and two dimensions is still a challenging

problem. A detailed understanding of the reaction of a range of molecules on surfaces

will be the key to developing targeted strategies for on-surface synthesis.

In this regard, the Ullmann reaction is very well-studied and constitutes an aryl-

aryl coupling of halogenated moieties. This reaction produces chemisorbed metal

halides as a byproduct and has consequently been a hotly-debated issue owing to the

presumed detrimental effects on the quality of the products. To find a solution for this

challenge, this thesis aims to contribute to the understanding of the absence of metal

halide byproducts in two ways. Firstly, employing the decarboxylative coupling

known as a “clean” reaction which produces volatile byproducts, and secondly in

removing the metal halide byproducts resultant from the Ullman reaction by using

hydrogen etching following the polymerisation step.

To investigate the decarboxylative reaction, the adsorption and reaction of 3,5-

pyridinedicarboxylic acid (PDC) and isophthalic acid (IPA) have been studied using

photoemission spectroscopy (PES) and near edge x-ray absorption fine structure

(NEXAFS) at the Australian Synchrotron facility. PE spectra of PDC reveal the

presence of nitrogen on the central group of the precursor result in different adsorption

configurations of the molecule on the surface: (a) via the nitrogen, and (b) via the

deprotonated carboxyl group. The result from NEXAFS supports that molecules are

adsorbed tilted with respect to the surface. The extracted activation energy for PDC

decarboxylation is (1.93 ± 0.17) eV which is large compared to the activation energy

for decarboxylation of a planar aromatic molecule, suggesting the tilted adsorption

configuration may change the activation energy. These results reveal a competition

between decarboxylation and molecular fragmentation at temperatures near

decarboxylation temperature.

ii Molecular reactions on surfaces: towards the growth of surface-confined polymers

Photoemission spectra of IPA reveal that the molecule partially deprotonated

upon adsorption on Cu(111). In addition, the beam energy dependence and angular

dependence spectroscopy of the IPA illustrates the relative geometry of the carboxyl

group, deprotonated carboxyl group and phenyl ring. The deprotonated carboxyl group

is buried beneath the phenyl ring, suggesting the molecule is anchored to the surface

via the deprotonated carboxyl group, and the NEXAFS data support the PES result.

Furthermore, STM illustrates that decarboxylative coupling of IPA successfully

proceeds on Cu(111). The IPA and PDC studies represent a step forward in

understanding the decarboxylation reaction and highlight the importance of the

chemistry of the building block in the adsorption geometry of the molecule and

therefore in the polymerization reaction.

To study the removal of by-product halogens from the surface, 1,4

dibromobenzene was deposited on Cu(111) and Cu(110). Atomic hydrogen was dosed

to the surface after the polymerization reaction completed. STM and XPS data

confirmed that halogens have been removed from the surface. This characteristic opens

up the possibility of employing the Ullmann reaction while a flux of atomic hydrogen

may offer effective way for removing unwanted halide by-products and a possibility

for improving the network qualities as well as polymer’s coverage on the surface.

Molecular reactions on surfaces: towards the growth of surface-confined polymers iii

Table of Contents

Abstract ..................................................................................................................................... i

Table of Contents .................................................................................................................... iii

List of Publication .....................................................................................................................v

List of Figures ....................................................................................................................... viii

List of Tables ..........................................................................................................................xv

List of Abbreviations ............................................................................................................ xvi

Statement of Original Authorship ........................................................................................ xvii

Acknowledgements ............................................................................................................. xviii

Chapter 1: Introduction ...................................................................................... 1

1.1 Background .....................................................................................................................1

1.2 Context ............................................................................................................................2

1.3 Purposes ..........................................................................................................................3

1.4 Thesis Outline .................................................................................................................4

1.5 References ......................................................................................................................5

Chapter 2: Literature Review ............................................................................. 7

2.1 Historical Background ....................................................................................................7

2.2 Ullmann Reaction ...........................................................................................................9

2.3 Decarboxylation Reaction ............................................................................................23

2.4 Pyridine-Based Polymer ...............................................................................................25

2.5 Summary and Implications ...........................................................................................29

2.6 References ....................................................................................................................32

Chapter 3: Research Design .............................................................................. 39

3.1 Important Parameters In Surface-Confined Reactions .................................................39

3.2 Surface Sensitive Analysis Techniques ........................................................................45

3.3 References ....................................................................................................................52

Chapter 4: Adsorption and Reactivity of Pyridine Dicarboxylic Acid on

Cu(111) .......................................................................................................... 55

4.1 Abstract .........................................................................................................................57

4.2 Introduction ..................................................................................................................58

4.3 Experimental Methods ..................................................................................................59

4.4 Results ..........................................................................................................................62

4.5 Discussion .....................................................................................................................73

4.6 Conclusions ..................................................................................................................75

4.7 References ....................................................................................................................77

iv Molecular reactions on surfaces: towards the growth of surface-confined polymers

4.8 Supporting Information ................................................................................................ 82

Chapter 5: Adsorption, deprotonation and decarboxylation of isophthalic

acid on Cu(111) ......................................................................................................... 87

5.1 Abstract ........................................................................................................................ 89

5.2 Introduction .................................................................................................................. 90

5.3 Experimental ................................................................................................................ 91

5.4 Results and discusion ................................................................................................... 92

5.5 Discussion .................................................................................................................. 103

5.6 Conclusions ................................................................................................................ 105

5.7 References .................................................................................................................. 106

5.8 Supporting Information .............................................................................................. 109

Chapter 6: Cleaning up after the Party: Removing the Byproducts of On-

surface Ullmann Coupling ..................................................................................... 117

6.1 Abstract ...................................................................................................................... 119

6.2 Introduction ................................................................................................................ 119

6.3 Experimental .............................................................................................................. 120

6.4 Results ........................................................................................................................ 122

6.5 Discussion .................................................................................................................. 133

6.6 Conclusion ................................................................................................................. 134

6.7 References .................................................................................................................. 135

6.8 Supporting Information .............................................................................................. 138

Chapter 7: Conclusions.................................................................................... 143

7.1 Conclusions ................................................................................................................ 143

7.2 Outlook....................................................................................................................... 145

Appendices .............................................................................................................. 147

Appendix A Comparison of dehalogenation temperature for different molecule on different

surfaces ................................................................................................................................. 147

Molecular reactions on surfaces: towards the growth of surface-confined polymers v

List of Publication

The results chapters in this thesis are based on articles that have been published

or submitted in peer-reviewed journals and presented in a conventional form of the

publication and contains an introduction, experimental details, results and discussions

making them self-contained.

Chapter 4: Abyazisani, M.; Bradford, J.; Motta, N.; Lipton-Duffin, J.;

MacLeod, J., Adsorption and Reactivity of Pyridine Dicarboxylic Acid on

Cu (111). JPC C 2018, 122 (31), 17836-17845.

DOI: 10.1021/acs.jpcc.8b04858.

Chapter 5: Abyazisani, M.; Bradford, J.; Motta, N.; Lipton-Duffin, J.;

MacLeod, J., Adsorption, deprotonation and decarboxylation of isophthalic

acid on Cu(111),

DOI: 10.1021/acs.langmuir.8b04233.

Chapter 6: Abyazisani, M.; MacLeod, J.; Lipton-Duffin, J., Cleaning up

after the party: removing the byproducts of on-surface Ullmann coupling,

under review for publication.

Publications not presented in this thesis:

Lipton-Duffin, J.; Abyazisani, M.; MacLeod, J., Periodic and nonperiodic

chiral self-assembled networks from 1, 3, 5-benzenetricarboxylic acid on

Ag (111). Chem. Commun. 2018, 54 (60), 8316-8319.

Abyazisani, M.; Jayalatharachchi, V.; MacLeod, J., 2.14 - Directed On-

Surface Growth of Covalently-Bonded Molecular Nanostructures. In

Comprehensive Nanoscience and Nanotechnology (Second Edition),

Andrews, D. L.; Lipson, R. H.; Nann, T., Eds. Academic Press: Oxford,

2018; Vol. 2, pp 299-326,

DOI: 10.1016/B978-0-12-803581-8.09239-0

http://pubs.acs.org/doi/abs/10.1021/acs.jpcc.8b04858

http://pubs.acs.org/doi/abs/10.1021/acs.jpcc.8b04858

http://pubs.acs.org.ezp01.library.qut.edu.au/doi/abs/10.1021/acs.langmuir.8b04233

https://doi.org/10.1016/B978-0-12-803581-8.09239-0

vi Molecular reactions on surfaces: towards the growth of surface-confined polymers

Conference Presentations

Maryam Abyazisani, Jonathan Bradford, Nunzio Motta, Josh Lipton-

Duffin, Jennifer MacLeod: “Growth of 1D polymers through on-surface

reactions”; The International Conference of Young Researchers on

Advanced Materials (ICYRAM), Adelaide Australia, November 2018

(oral presentation).


Duffin, Jennifer MacLeod: “Growth of graphene-like materials through on-

surface reactions”, The 6th International Symposium on Graphene

Devices (ISGD-6), Petersburg Russia , May 2018.



reactions”; International Conference on Nanoscience and

Nanotechnology (ICONN), Wollongong Australia, January 2018 (oral

presentation).



reactions”; The 42nd Condensed Matter and Materials Meeting, Wagga

Wagga Australia, February 2018 (oral presentation).

Maryam Abyazisani, Vishakya Jayalatharachchi, Nunzio Motta, Josh

Lipton-Duffin, Jennifer MacLeod: “Growth of 1D Polymers Through On-

Surface Reactions”; Nanostructures for Sensors, Electronics, Energy

and Environment (NanoS-E3), Brisbane Australia, March 2017 (poster

presentation).

Maryam Abyazisani, Nunzio Motta, Josh Lipton-Duffin, Jennifer

MacLeod: “Growth of 1D and 2D Polymers Through On-Surface

Reactions“, Australian Microbeam Analysis Society (AMAS), Brisbane

Australia, February 2017 (oral presentation).

Molecular reactions on surfaces: towards the growth of surface-confined polymers vii



Reactions” Nanotechnology and Molecular Science HDR Symposium

(NMS), Brisbane Australia, July 2017 (oral presentation).



Reactions”, 22ND Australian Institute of Physics (AIP) Congress,

Brisbane Australia, August 2016 (poster presentation).


MacLeod: “Growth of Graphene-Like Materials Through On-Surface

Reactions”; The 5th International Symposium on Graphene Devices

(ISGD-5), Brisbane Australia, July 2016, (poster presentation).


MacLeod: “Benchmarking Temperatures for Catalyzed Decarboxylation of

Trimesic Acid on Metal Surfaces”; Nanotechnology and Molecular

Science HDR Symposium (NMS), Brisbane Australia, February 2016,

(poster presentation).

viii Molecular reactions on surfaces: towards the growth of surface-confined polymers

List of Figures

Figure 2-1: Schematic reaction for Ullmann coupling. .............................................. 10

Figure 2-2: Overview STM images representing transformation of DMTP

molecules from a) halogen-bonded network to b) DMTP−Cu

coordination network and on Ag c) to organometallic networks.

Reprinted with permission from [30]. Copyright (2018) American

Chemical Society. ........................................................................................ 11

Figure 2-3 a) As-deposited at RT and b) thermally activated hexaiodo-

substituted macrocycle cyclohexa-m-phenylene (CHP) on Cu(111).

The white circle highlights radicals surrounded by iodine atoms.

Reprinted with permission from [32]. Copyright (2010) American

Chemical Society, c) organometallic and biphenyl formed from

bromobenzene on Cu(111). Reproduced from [28] with permission

from The Royal Society of Chemistry. ........................................................... 12

Figure 2-4: STM image of 1,4-diiodobenzene lines after annealing at 500K on

Cu(110). Reproduced with permission from [43]. ....................................... 13

Figure 2-5: Fast-XPS C 1s spectra during annealing of each precursor on

Cu(110) dosed at RT. Reproduced with permission from [35]. ................... 14

Figure 2-6: Tracking the desorption of different masses during annealing

DBBA multilayer using TPD, the desorbing fragments assigned in the

figure nest to related curve and the correspond schematic cartoon for

desorption process are shown in the bottom of image. Reprinted with

permission from [33]. Copyright (2015) American Chemical Society. ....... 16

Figure 2-7: Confirming the halogen removal of surface resulted from

polymerization of 4,5,9,10-tetrabromo-2,7-di-tertbutylpyrene

precursor on Au(111), using molecular hydrogen dosing and thermal

annealing. Republished with permission from [49]. Copyright (2017)

Royal Society of Chemistry. ........................................................................ 17

Figure 2-8: a,b) STM image presenting the OM and polymer network of 1,3,5-

tris(4-bromophenyl)benzene (TBB) on Cu(111), and c,d) the

corresponding models, respectively. The red circle in (a) shows the

coper atom and the arrow shows the backbone of (TBB) molecule.

Adapted with permission from [50] Copyright (2014) American


Figure 2-9: C1s core level XPS for dBB on Cu(110), peak 1 is attributed to the

OM species, peak at 286.4 eV corresponding to carbon-bromine bonds

in intact molecule at LT vanishes at RT, peak 2 and 3 are assigned to

C2 carbons in the aromatic ring of the dehalogenated molecules.

Adapted with permission from [53] Copyright (2013) American


Figure 2-10: a) The C1s and b) the Br3d and I4d core-level shifts simulation,

for the systems depicted above each plot. For the methine C-atoms (C-

atoms with an H-atom) the average core level shifts are indicated.

Molecular reactions on surfaces: towards the growth of surface-confined polymers ix

Adapted with permission from [29]. Copyright (2013) American


Figure 2-11: transformation of a) self-assembled BIB molecules to b) a well-

ordered hexagonal OM and to C) covalently-linked networks.

Published by The Royal Society of Chemistry [54]. ....................................... 21

Figure 2-12: Scheme of the sequential activation mechanism and

corresponding STM images of hierarchical Ullmann reaction of BIB

precursor onto Au(111). Adapted with permission from [59]

Copyright (2014) American Chemical Society............................................ 22

Figure 2-13: Scheme of the sequential activation mechanism and

corresponding STM images of hierarchical Ullmann reaction of trans-

Br2I2TPP molecules on Au(111).Reprinted with permission from [58].

Copyright (2012) Springer Nature. .............................................................. 22

Figure 2-14: a) Reaction pathways for decarboxylation of NDCA; b) STM

images of organometallic of NDCAs surface; c) poly-2,6 naphthalene

on Cu(111). Adapted with permission from [72]. Copyright (2014)

American Chemical Society. ....................................................................... 24

Figure 2-15: STM image of a) self-assembled b) organometallic (Cu adatoms

are visible as circular protrusions) and c) covalently-bonded network

obtained from decarboxylative coupling of TCPB. The inset shows a

high-resolution image with a superimposed molecular structure. d-f)

XPS spectra of the C 1s and O 1s. Adapted with permission from [73].

Copyright (2016) American Chemical Society............................................ 25

Figure 2-16: a,b) STM images of two types of single pyridine molecules

adsorbed on Ag(1 1 0) at 13 K, and c,d) Equilibrium atomic structures

of vertically upright (stand-up) and flat-lying configuration (Pf)

configurations of pyridine on Ag(1 1 0) surfaces. Top views of the

two topmost layers and cross-section views of the four topmost layers

of the 3 × 4 Ag(1 1 0) surface in a supercell are shown in each

configuration. Large and small gray spheres represent Ag atoms on

the surface layer, and on a layer below it, respectively. The black,

green, and red circles represent carbon, hydrogen, and nitrogen atoms,

respectively. Reprinted with permission from [83]. .................................... 26

Figure 2-17: a-c) STM images of 0.04, 0.07, and 0.17 nm2 for pyridine

coverage. The height of feature a was always higher than that of

feature b. d) Plot showing the dependency of the configuration to the

coverage surface by comparing relative population ratio of pyridine

configurations as a function of the total surface concentration.

Reprinted with permission from [83]. .......................................................... 27

Figure 2-18: The possible adsorption geometries of pyridine on a metal surface.

Reprinted with permission from [84]. .......................................................... 28

Figure 2-19: C1s and N1s XPS spectra for 2,4-dibromopyridine on Cu(100) at

different temperature. Reprinted with permission from [85]. Copyright

(2015) American Chemical Society. ............................................................ 29

x Molecular reactions on surfaces: towards the growth of surface-confined polymers

Figure 2-20: Proposed adsorption configurations for mono-debrominated a, b)

2,4-dibromopyridine and c,d) 2,3-dibromopyridine, Reprinted with

permission from [85]. Copyright (2015) American Chemical Society. ....... 29

Figure 3-1: Schematic dependence of relative substitution pattern of precursors

and their products. X represents any possible reactive group in the

precursor. ...................................................................................................... 40

Figure 3-2: STM images of (a) TIPB in which covalent-bonded oligomers,

including dimers and trimers mostly near step-edges and (b) TIB

which result in covalent networks and iodine by-products on Au(111).

Adapted with permission from [3]. .............................................................. 41

Figure 3-3: Scheme of two different reaction pathways of DN on Ag(111) and

Au(111) illustrating C−C or C−H coupling of DN on different

surfaces. Reprinted with permission from [15] Copyright (2017)

American Chemical Society. ........................................................................ 43

Figure 3-4: Top view of atoms in the first layer of different planes. ......................... 44

Figure 3-5: Comparison of reactivity effect on polymerization of DBBA

between Cu(111) and Cu(110). Reprinted with permission from [17]. ....... 45

Figure 3-6: XP spectrum for as-deposited TMA on Cu(111) surface. In addition

to the copper peaks, C 1s and O 1s peaks originating from TMA are

discernible. The insets show the Cu 2p and C 1s highlighted region of

the spectrum in more detail. ......................................................................... 47

Figure 3-7: left: Sketch of tunnelling current through sample to tip. Efs and Eft

are the Fermi level of the sample and tip, respectively. Right:

Schematic view of the scanning tunnelling microscope.25 .......................... 48

Figure 3-8: Energy diagram representing electron transition in NEXAFS. ............... 50

Figure 3-9: scheme of X-ray absorption in a diatomic molecule (bottom), and

the corresponding k-shell absorption spectrum (top). Reprinted with

permission from [29]. ................................................................................... 50

Figure 3-10: Schematic representation of photon incident with normal angle

(left) and grazing angle (right). The diatomic molecule adsorbed

parallel to the surface. Reprinted with permission from [29]. ..................... 51

Figure 4-1: (a) molecular structure of 3,5-pyridinecarboxylic acid (PDC). (b)

Decarboxylation reaction schematic for PDC molecule. ............................. 61

Figure 4-2: C 1s, O 1s and N 1s spectra for PDC deposited onto Cu(111) at RT

and annealed up to 270°C. Line plot with black circle markers

represents the acquired data, and solid gray lines show the envelope of

the fitted peaks. Fitted components are color-coded and detailed in the

text. ............................................................................................................... 63

Figure 4-3: Calculated adsorption geometries and energies for configurations

for PDC on Cu(111) when (a) intact, (b) singly deprotonated and (c)

completely deprotonated. ............................................................................. 64

Figure 4-4: Stacked plots of the C 1s region for PDC on Cu(111) with the

sample held at a constant temperature of 225 (a), 230 (b) and 235°C

(c). Each spectrum was collected over ~30 s. .............................................. 66

Molecular reactions on surfaces: towards the growth of surface-confined polymers xi

Figure 4-5: (a) Decarboxylation rate and (b) Arrhenius plot for the

decarboxylation reaction of PDC on Cu(111) derived from analysis of

PES spectra of the C 1s region at 225, 230 and 235°C, where k is the

reaction rate determined from a. .................................................................. 67

Figure 4-6: NEXAFS spectra collected at the carbon (a), nitrogen (b) and

oxygen (c) K-edges for an as-deposited multilayer sample (sample 1)

with = 20°. Indicative peak-fitting analyses are shown through the

multicoloured components ascribed to each spectrum. (Inset) Angular

dependence of each set of NEXAFS spectra is shown. The inset of (a)

shows the experimental geometry and the definition of the angle θ,

which relates the orientation of the electric field polarization to the

sample surface. ............................................................................................. 68

Figure 4-7: Evolution of the carbon (a) and oxygen (b) K-edges with annealing

of the PDC film. RT corresponds to an as-deposited multilayer, 100°C

to a film of partially deprotonated molecules, 155°C to a film of fully

deprotonated molecules, and 280°C to the remnants of a broken,

decarboxylated molecule (carbon K-edge) and to near-complete

desorption of all oxygen from the surface (oxygen k-edge). All spectra

were collected at = 20°. ............................................................................ 70

Figure S4-8: C 1s spectra for PDC deposited onto Ag(111) kept at -130°C and

annealed up to 100°C. The beam energy is 486 eV. .................................... 82

Figure S4-9: Beam damage test for PDC on Ag(111). The beam energy is 486

eV and each scan was acquired over 50 seconds. The as-deposited

sample is at the bottom, with increasing beam damage moving up the

stack. ............................................................................................................ 83

Figure S4-10 C 1s and N 1s spectra for PDC deposited onto Cu(111) at 270°C

and annealed up to 480°C. ........................................................................... 84

Figure 5-1: (a) Decarboxylation reaction schematic for IPA molecule in the

presence of copper and (b) three possible structural products of

polymerization of IPA, from left to right: zig-zag, crinkled and rosette

macrocycles respectively. ............................................................................ 93

Figure 5-2: (a) C 1s and (b) O 1s spectra for IPA deposited onto Cu(111) at 300

K, followed by two annealing steps at 453 K and 488 K. The black

markers represent the acquired data, and the coloured peaks show

synthetic fits. ................................................................................................ 94

Figure 5-3: Angular dependence of the apparent stoichiometry of IPA

deposited at room temperature collected at hυ=380 eV using four

different angles with respect to the surface normal. The intensity of

carboxylate, carboxyl and phenyl components have been normalized

with repect to the total C 1s. Inset shows a schematic of the adsorption

geometry. ..................................................................................................... 96

Figure 5-4: Angular dependence of the apparent stoichiometry of fully

deprotonated IPA collected at hυ=380 eV using four different angles

with respect to the surface normal The intensity of carboxylate,

carboxyl and phenyl peak has been normalized with respect to the

total C 1s. Inset shows a schematic of the adsorption geometry. ................ 97

xii Molecular reactions on surfaces: towards the growth of surface-confined polymers

Figure 5-5: NEXAFS spectra collected (a) at the carbon K-edge for half-

deprotonated, fully-deprotonated, and decarboxylated IPA on Cu(111)

with s-polarized (normal incidence), magic angle (55° incidence) and

p-polarized (glancing incidence, 20°), and (b) at the oxygen K-edge

for the half- and fully-deprotonated samples using the same incident

beam geometries. ......................................................................................... 98

Figure 5-6: STM image of an IPA layer deposited onto Cu(111) at room

temperature. Image parameters: U=-360 mV, I=0.3 nA, 72 nm × 72

nm. .............................................................................................................. 100

Figure 5-7: STM image of the IPA layer after annealing to 373 K, U=-470 mV,

I=0.01 nA. The inset shows the tentative adsorption configuration of

mono-deprotonated IPA superimposed on the high resolution STM of

chains, 19.7 nm × 19.7 nm, inset: U=-760 mV, I=0.01 nA, 3.7×1.9

nm2. ............................................................................................................ 101

Figure 5-8: STM image of the fully-deprotonated IPA layer after annealing to

433 K. Image parameters: U=-1100 mV, I=0.2 nA, 49 nm × 49 nm.

Inset: U=-1100 mV, I=0.1 nA. The inset is a tentative model. .................. 102

Figure 5-9: STM images showing the polymer resulting from IPA annealed at

513 K at different magnification, (a) U=-1100 mV, I=0.1 nA,

80.9×80.9 nm2, (b) U=-140 mV, I=0.2 nA, 22.6×22.6 nm2. ...................... 103

Figure S5-10: C 1s spectra for IPA deposited onto Cu(111) at RT and annealed

up to 240°C. ............................................................................................... 109

Figure S5-11: Apparent stoichiometry of IPA deposited at room temperature

collected with different beam energies: 380, 486, 650 and 908 eV. .......... 110

Figure S5-12: The red data points (with error bars) indicate the component

area. The blue line shows the fit to the vector symmetry formula Stöhr

9.16a.4 The resulting inclination angles are reported in each panel of

the figure. (a) Half-deprotonated carbon phenyl, (b) half-deprotonated

carbon carboxylic/carboxylate, (c) half-deprotonated oxygen

carboxylic/carboxylate, (d) deprotonated carbon phenyl, (e)

deprotonated carbon carboxylate, (f) deprotonated oxygen

carboxylate. ................................................................................................ 111

Figure S5-13: STM image of IPA annealed at 100°C, a) U=-480 mV, I=0.01

nA, 14.5×14.5 nm2, b) U=-480 mV, I=0.01 nA, 19.7×19.7 nm2. .............. 112

Figure S5-14: (a) STM image showing macrocycles and other oligomeric

products. The blue line indicates the location of the line profile shown

in (b). Image parameters: U=-140 mV, I=0.2 nA, 4.4 nm × 4.4 nm. (b)

The blue line is the profile indicated in (a). An indicative

measurement of the pore diameter, 0.76 nm, is shown; the values in

the main manuscript comprise multiple measurements and uncertainty

propagation. At the top of the figure is a molecular mechanics

(MMFF94) optimized structure for the macrocycle produced via

Avogadro, which has a diameter of 0.76 nm. ............................................ 112

Figure S5-15: The carbon to copper ratio of IPA at room temperature and

annealed at 180°C ...................................................................................... 113

Molecular reactions on surfaces: towards the growth of surface-confined polymers xiii

Figure S5-16: The carbon to copper ratio of IPA from deprotonated to

decarboxylated state. .................................................................................. 113

Figure S5-17: Monitoring of deprotonation of low coverage IPA. .......................... 114

Figure 6-1: On-surface Ullmann coupling of dibromobenzene and subsequent

removal of bromine by exposure to atomic hydrogen. .............................. 120

Figure 6-2: STM images of the organometallic assemblies resulting from dBB

on Cu(110) (a: I=0.5 nA, U=-0.5 V, 98x98 nm2, b: I=0.5 nA, U = 0.5

V, 28x28 nm2, c: I=0.05 nA, -0.01 V, 10 x 10 nm2). ................................. 123

Figure 6-3: Photoelectron spectroscopy of a) the carbon 1s and b) the bromine

3p core levels of dBB on Cu(110) as a function of atomic hydrogen

etching time. ............................................................................................... 124

Figure 6-4: Evolution of Br 3p:C 1s and C 1s:Cu 2p as a function of hydrogen

etching time. ............................................................................................... 125

Figure 6-5: STM images of Cu(110) surface after growth of PPP polymers

from DBB precursor (ac), etching procedure (d-f), and after final

annealing (h-j). Image dimensions are identical by column, and a z-

height colour scale is given beside each image. a, c) I=630 pA, U=-

1.26 V, b)I=630 pA, U=-280 mV , d) I=50 pA, U=-1.14V, e, f) I=100

pA, U=-180 mV, g,i) I=20 pA, U=-980 mV, h) I=100 pA, U=-300

mV.............................................................................................................. 127

Figure 6-6: C1s XPS as a function of atomic hydrogen etching. a) After

deposition, b) After H-etching for 16 minutes, c) After annealing to

275°C. ........................................................................................................ 128

Figure 6-7: STM micrographs of organometallic chains on Cu(111) (: 100 x

100 nm2, I = 10 pA, U = -0.5 V, b: 39x39 nm, I=5 pA, U=-100 mV, c:

10 x 10nm2, I = 20 pA, U=-1.0V). ............................................................. 130

Figure 6-8: STM images of dBB on Cu(111) annealed for 5 minutes at 275°C,

with local spatial characteristics. a) I=10 pA, U=-300 mV, 200 x 200

nm2 b) I= 0.1nA and U= -250 mV, 7.2 x 7.2 nm2, c) I=0.2 pA, U = -

710 mV, 50x50 nm2. .................................................................................. 131

Figure 6-9: a) Carbon 1s and b) bromine 3p core levels during etching of dBB

on Cu(111). ................................................................................................ 132

Figure S6-10: Summary of XPS experiments for etched aC:H film. Presented

are survey spectra (top), high resolution of C 1s (left inset, top) and Si

2p (left inset, bottom). At the bottom right, the intensity vs time trend

is shown for both these core levels, along with the extrapolated linear

fit. ............................................................................................................... 139

Figure S6-11: Long-range image (100×100 nm2) of Cu(110) surface following

hydrogen etching treatment and annealing. The surface presents a near

100% orientation of the polymers along the <1-10> direction. I=50

pA, U=-340 mV ......................................................................................... 140

Figure S6-12: (a) Wide and (b) close range STM images of PPP chains on

Cu(110) after completely removing bromine from a submonolayer

surface, U=-1V, I=0.005nA, 73.7×73.7 nm2, U=-1.8V, I=0.002nA,

25.3×25.3 nm2. ........................................................................................... 140

xiv Molecular reactions on surfaces: towards the growth of surface-confined polymers

Figure S6-13: STM images of atomic hydrogen-etched the PPP on Cu(111). I=

0. 1 nA, U= -1.5 V, 100×100 nm2, and I=0.01 nA, U=-0.05V, 11×11

nm2. ............................................................................................................ 141

Figure S6-14: Two STM images of PPP on Cu(110) after partial bromine

removal, collected sequentially over a period of 450 seconds (each).

The white circles are meant as a guide to the eye to indicate changes

in the PPP island morphology scan-to-scan. .............................................. 142

Molecular reactions on surfaces: towards the growth of surface-confined polymers xv

List of Tables

Table 2-1: Comparison of atomic and molecular desorption energies as well as

energy barriers for the dehalogenation for bromine and iodine on

Cu(111), Ag(111) and Au(111).29 All values are in eV. .............................. 15

Table 3-1 Comparisons of structural symmetries of Cu, Ag and Au, reproduced

from.10 .......................................................................................................... 44

Table 4-1. NEXAFS average adsorption angles (relative to the surface plane)

for submonolayer and Multilayer PDC on Cu(111) α .................................. 72

Table 4-2: NEXAFS calculated average adsorption angles for multilayer PDC

on Cu(111). Based on deposition time, with no consideration of the

reduced sticking coefficient after monolayer completion, coverage on

this sample should be slightly more than double the coverage on the

multilayer sample reported in the main manuscript. .................................... 84

Table 5-1: NEXAFS-derived tilt angles for the phenyl and

carboxylic/carboxylate of IPA in different states of reaction. Angles

are averages, and specify the angle of the π* orbital with respect to the

surface normal. ............................................................................................. 99

Table 6-1: Summary of binding energies of carbon and bromine core levels on

Cu(110) and Cu(111) surfaces. .................................................................. 132

Table A-1: Some results of Ullmann coupling reported in the literature,

classified as a function of surface and halogen .......................................... 147

Table A-2 Abbreviation definition of molecules ..................................................... 148

xvi Molecular reactions on surfaces: towards the growth of surface-confined polymers

List of Abbreviations

0D zero dimensional

1D one dimensional

2D two dimensional

BE binding energy

CLS core level shifts

COFs covalent-organic frameworks

CVD chemical vapor deposition

dBB dibromobenzene

DFT density functional theory

DOS density of state

FCC face centered cubic

FWHM full width at half maximum

GNRs graphene nanoribbons

HOPG

HOMO

highly oriented pyrolytic graphite

highest occupied molecular orbital

IMFP

IPES

inelastic mean free path

inverse photoemission spectroscopy

IPA isophatalic acid

LDOS local density of states

LEED

LUMO

low energy electron diffraction

lowest unoccupied molecular orbital

MBE molecular beam epitaxy

MOF metal–organic framework

NEXAFS near edge x-ray absorption fine structure

OM organometallic

OMBE organic molecular beam epitaxy

PDC 3,5-Pyridinedicarboxylic acid

PDMS Polydimethylsiloxane

PES Photoemission spectroscopy

PMMA poly-methyl methacrylate

PMP poly meta-phenylene

PPP poly-para-phenylene

RT room temperature

RAIRS reflection−absorption infrared spectroscopy

STM scanning tunnelling microscopy

TMA trimesic acid = 1,3,5 benzenetricarboxylic acid

TPD temperature-programmed desorption

TPR temperature-programmed reaction

UHV

UPS

ultra-high vacuum

ultraviolet photoelectron spectroscopy

UV ultraviolet

WF work function

XPS X-ray photoelectron spectroscopy

Molecular reactions on surfaces: towards the growth of surface-confined polymers xvii

Statement of Original Authorship

The work contained in this thesis has not been previously submitted to meet

requirements for an award at this or any other higher education institution. To the best

of my knowledge and belief, the thesis contains no material previously published or

written by another person except where due reference is made.

Signature:

Date: June 2019

QUT Verified Signature

xviii Molecular reactions on surfaces: towards the growth of surface-confined polymers

Acknowledgements

I am grateful that I was part of a unique research environment with supportive

and inspiring group members. I would like to recognise and greatly appreciate Dr

Jennifer MacLeod for her wisdom, expertise and continued support throughout my

research journey. I was completely supported by her in every possible way to achieve

my scientific goals.

I thank Dr Josh Lipton-Duffin for his mentorship and guidance during this work.

I enjoyed working with him and I was inspired by the environment that he has created

for discovery and learning. I would also like to acknowledge the important

contributions to this work made by Professor Nunzio Motta.

I would also like to acknowledge the financial support provided by CPME school

and the tuition fee waiver from QUT for this PhD project.

My thanks also go to the heart-warmed students and friends in M202 and my

friend Jonathan Bradford.

I would like to express my deepest gratitude to my parents Gholamreza and

Mansureh, for the unconditional supports and enthusiasm they provided for me to

reach my goals.

Last but not least, I owe this to my beloved husband. Hossein thanks for all your

endless support in the completion of this work.

Chapter 1: Introduction 1

Chapter 1: Introduction

This chapter outlines the background, motivation, history and context of the

present research (section 1.1 and 1.2). Section 1.3 describes the significance and scope

of this research and provides definitions of terms used. Finally, section 1.4 includes an

outline of the remaining chapters of the thesis.

1.1 BACKGROUND

Low dimensional materials formed through on-surface synthesis are fascinating

due to their potential impact on our fundamental understanding of natural science and

in particular nanotechnology.1-3 Although direct manipulation of individual atoms and

engineering of molecular structure is widely-used these days, the idea of this

possibility can be traced back to the prominent lecture of Prof. Richard Feynman

“There's Plenty of Room at the Bottom” in 1959,4 wherein he described the use of

atomic and molecular building blocks to construct the desired structure. However, the

concept of manipulating individual molecules and atoms remained almost unnoticed

until the development of scanning tunnelling microscopy (STM) in 19815 which won

its inventors, Gerd Binnig and Heinrich Rohrer, the Physics Nobel prize in 1986. STM

provided a very powerful technique for visualizing structures on the atomic scale and

exploring surface-confined reactions and fulfilled scientists’ dream of manipulating

molecular structures with atomic precision.

The discovery of low dimensional carbonaceous materials such as fullerene,

carbon nanotube and in particular graphene6 has inspired researchers to investigate a

range of carbon-based low-dimensional materials. Graphene has extraordinary

properties and potential applications in optical and electronic nanotechnological

devices.7 Graphene or other natural sheets,8 such as BN or MoS2, are available in

nature and, for example, graphene can be isolated from graphite using a top-down

approach. In fact, top-down approaches can only provide the precision control to a

certain degree over the product,9-11 while the electronic properties of these

nanomaterial structure strictly (critically) influenced by their design, e.g. graphene

nanoribbon (GNRs) needs “atomic” precision control over the doping level,12 width13-

14 or/and edge shape15 of the product. Fabrication of functional organic architectures

2 Chapter 1: Introduction

using surface science, which is based on a bottom-up approach, capitalizes on the

unique catalytic and crystallographic properties of solid surfaces and provides the

possibility for designing low-dimensional materials with adjustable electrical and

optical properties.

Various on-surface reactions have been used on different surfaces to fabricate

zero-dimensional (0D) quantum dots or one-dimensional (1D) chains and two-

dimensional (2D) covalent nanostructures. The dimension of the products can be

controlled by the number of reactive groups in combination with their positions in the

precursors. These polymers could allow researchers to achieve functionality not

available from graphene.16 Thus, the surface science approach has attracted

tremendous interest over the last decade and continues to provide an important

approach to the production of nanostructures.

1.2 CONTEXT

Molecular nanostructures confined to surfaces can be classified into self-

assembled and polymer categories based on the nature of the chemical bonds in the

organic nanostructures. Self-assembled structures are stabilized by non-covalent

interactions such as hydrogen bonds17-19 van der Waals interactions20 and metal–ligand

interactions.21 These non-covalent interactions are reversible and therefore lead to

well-defined and long-range ordering of molecules on the surface. However, the weak

nature of these interactions causes the self-assembled molecular networks to have poor

mechanical and thermal stability and also affect the charge transport of the stucture.22

The lack of robustness of self-assembled architectures can be overcome using covalent

coupling. Covalent coupling of targeted precursors can also be used to synthesize

conjugated polymers allowing efficient charge transport through their bonds.23-24

These properties have made covalently-connected organic networks a core focus of

the bottom-up approach for research in nanotechnology. Here, I limit the term

“polymers” to architectures that involve covalently-linked structures with repeating

units of the same building block.

It must be noted that most covalent bonds are irreversible in contrast to their non-

covalent counterparts. In a non-covalent reaction, post-correcting and modification of

defects is possible via breaking and reforming bonds. However, covalent coupling,

which provides a highly robust structure, can minimise the ability for optimising the


structure or modification of morphology. As a result, both the spatial extent and

ordering of surface-polymerized architectures tend to be very poor.25 This creates a

challenge, since the geometrical dimensions and topology of products, which depends

on different parameters e.g. entropic effect, are inherently linked to the electronic

properties. Thus, a detailed understanding of the reaction is necessary to optimise the

growth and produce high-quality polymers.

1.3 PURPOSES

Surface-confined covalent coupling in ultra-high vacuum (UHV) is a promising

approach to achieve molecular devices by fabricating well-ordered polymers directly

on a surface. However, since this topic is relatively new, it is poorly explored in terms

of a detailed understanding of the reaction mechanisms and the substrate influence in

the reaction process. There is a particular lack of understanding around molecules

incorporating heteroatoms; while heteroatoms are commonly used to adjust the

electronic properties of organic semiconducting molecules, they can perturb

adsorption geometries on the surface and can introduce reaction pathways that

compete with the polymerization process.26-27

To bridge the gap between fundamental studies of covalent 2D polymers and

practical technological applications, it is necessary to develop a thorough

understanding of surface-confined coupling reactions by which well-ordered and large

molecular networks can be grown. This work is targeting two specific reactions for

on-surface molecular coupling to produce useful and well–ordered networks and

chains. Additionally, some part of research has been devoted to investigating

heteroatoms in the molecular structure of polymers. Heteroatom-containing molecules

are organic compounds in which one or more of their hydrogen or carbon atoms are

replaced with other atoms such as nitrogen, boron etc. The presence of, e.g. nitrogen,

can modify the bandgap in a predictable manner.28-30 However, nitrogen-containing

molecules rarely have been studied.31 The other key challenge in adopting 2D

polymers in device applications is their structural order, which is generally poor, and

does not allow the polymers to be grown large enough for many applications. In order

to be able tailor the band gap of structure for nitrogen-containing polymers, the

polymer’s network need to be spatially extended. To address the challenge of both

structural order and spatial extension of the polymers, it is critical to expand our


fundamental understanding on reaction, intermediate state of product. Therefore, this

study could contribute to two main objectives:

1. Investigating of decarboxylative coupling, an emerging polymerization

reaction that has a number of advantages for maximizing polymer

coverage on the surface.

2. Study the reaction and the possible resultant product from a nitrogen-

containing carboxyl monomers.

3. Optimizing the polymer structural quality by removing the by-product in

the Ullmann reaction.

1.4 THESIS OUTLINE

This chapter has introduced the research problem and thesis objectives, followed

by literature review in chapter 2 which expands the description of on-surface

synthesised polymers and highlights the research gap introduced in chapter 1. In

chapter 3, the methodology used in this research, the key parameters in on-surface

reactions as well as a brief introduction of surface sensitive analysis techniques

including X-ray photoemission spectroscopy (XPS), scanning tunnelling microscopy

(STM) and near edge x-ray adsorption fine structure (NEXAFS) is given. The

following two chapters present studies of the decarboxylation reaction of a

heteroatom-containing molecule as a building block against its carbon-containing

counterpart. Chapter 6 tackle the question to improve the order of Ullmann-derived

polymers by removing the halogens.


1.5 REFERENCES

1. Joachim, C.; Gimzewski, J. K.; Aviram, A. Electronics using hybrid-molecular

and mono-molecular devices. Nature 2000, 408 (6812), 541-548.

2. Solís-Fernández, P.; Bissett, M.; Ago, H. Synthesis, structure and applications

of graphene-based 2D heterostructures. Chemical Society Reviews 2017, 46 (15),

4572-4613.

3. Chen, Z.; Molina-Jirón, C.; Klyatskaya, S.; Klappenberger, F.; Ruben, M. 1D

and 2D Graphdiynes: Recent Advances on the Synthesis at Interfaces and Potential

Nanotechnological Applications. Annalen der Physik 2017, 529 (11), 1700056-n/a.

4. Feynman, R. P. There's plenty of room at the bottom [data storage]. Journal of

Microelectromechanical Systems 1992, 1 (1), 60-66.

5. Binnig, G.; Rohrer, H. Scanning tunneling microscopy. Surface Science 1983,

126 (1-3), 236-244.

6. Novoselov, K. S.; Geim, A. K.; Morozov, S. V.; Jiang, D.; Zhang, Y.;

Dubonos, S. V.; Grigorieva, I. V.; Firsov, A. A. Electric field effect in atomically thin

carbon films. Science 2004, 306 (5696), 666-669.

7. Novoselov, K. S.; Falko, V. I.; Colombo, L.; Gellert, P. R.; Schwab, M. G.;

Kim, K. A roadmap for graphene. Nature 2012, 490 (7419), 192-200.

8. Novoselov, K.; Jiang, D.; Schedin, F.; Booth, T.; Khotkevich, V.; Morozov,

S.; Geim, A. Two-dimensional atomic crystals. Proceedings of the National Academy

of Sciences of the United States of America 2005, 102 (30), 10451-10453.

9. Kim, M.; Safron, N. S.; Han, E.; Arnold, M. S.; Gopalan, P. Fabrication and

characterization of large-area, semiconducting nanoperforated graphene materials.

Nano letters 2010, 10 (4), 1125-1131.

10. Eroms, J.; Weiss, D. Weak localization and transport gap in graphene antidot

lattices. New Journal of Physics 2009, 11 (9), 095021.

11. Bai, J.; Zhong, X.; Jiang, S.; Huang, Y.; Duan, X. Graphene nanomesh. Nature

Nanotechnology 2010, 5 (3), 190.

12. Terrones, H.; Lv, R.; Terrones, M.; Dresselhaus, M. S. The role of defects and

doping in 2D graphene sheets and 1D nanoribbons. Reports on Progress in Physics

2012, 75 (6), 062501.

13. Son, Y.-W.; Cohen, M. L.; Louie, S. G. Energy gaps in graphene nanoribbons.

Physical Review Letters 2006, 97 (21), 216803.

14. Han, M. Y.; Özyilmaz, B.; Zhang, Y.; Kim, P. Energy band-gap engineering

of graphene nanoribbons. Physical Review Letters 2007, 98 (20), 206805.

15. Tao, C.; Jiao, L.; Yazyev, O. V.; Chen, Y.-C.; Feng, J.; Zhang, X.; Capaz, R.

B.; Tour, J. M.; Zettl, A.; Louie, S. G. Spatially resolving edge states of chiral graphene

nanoribbons. Nature Physics 2011, 7 (8), 616.

16. Bouju, X.; Mattioli, C.; Franc, G.; Pujol, A.; Gourdon, A. Bicomponent

supramolecular architectures at the vacuum–solid interface. Chemical Reviews 2017,

117 (3), 1407-1444.

17. Barth, J. V.; Weckesser, J.; Trimarchi, G.; Vladimirova, M.; De Vita, A.; Cai,

C.; Brune, H.; Günter, P.; Kern, K. Stereochemical Effects in Supramolecular Self-

Assembly at Surfaces: 1-D versus 2-D Enantiomorphic Ordering for PVBA and

PEBA on Ag(111). Journal of the American Chemical Society 2002, 124 (27), 7991-

8000.

18. Nath, K. G.; Ivasenko, O.; MacLeod, J. M.; Miwa, J. A.; Wuest, J. D.; Nanci,

A.; Perepichka, D. F.; Rosei, F. Crystal Engineering in Two Dimensions: An


Approach to Molecular Nanopatterning. The Journal of Physical Chemistry C 2007,

111 (45), 16996-17007.

19. MacLeod, J. M.; Ivasenko, O.; Fu, C.; Taerum, T.; Rosei, F.; Perepichka, D. F.

Supramolecular Ordering in Oligothiophene−Fullerene Monolayers. Journal of the

American Chemical Society 2009, 131 (46), 16844-16850.

20. De Feyter, S.; Miura, A.; Yao, S.; Chen, Z.; Würthner, F.; Jonkheijm, P.;

Schenning, A. P. H. J.; Meijer, E. W.; De Schryver, F. C. Two-Dimensional Self-

Assembly into Multicomponent Hydrogen-Bonded Nanostructures. Nano Letters

2005, 5 (1), 77-81.

21. Dmitriev, A.; Spillmann, H.; Lin, N.; Barth, J. V.; Kern, K. Modular Assembly

of Two‐Dimensional Metal–Organic Coordination Networks at a Metal Surface.

Angewandte Chemie 2003, 115 (23), 2774-2777.

22. Gourdon, A. On‐Surface Covalent Coupling in Ultrahigh Vacuum.

Angewandte Chemie International Edition 2008, 47 (37), 6950-6953.

23. Perepichka, D. F.; Rosei, F. Extending polymer conjugation into the second

dimension. Science 2009, 323 (5911), 216-217.

24. Champness, N. R. Surface chemistry: Building with molecules. Nature

Nanotechnology 2007, 2 (11), 671-672.

25. El Garah, M.; MacLeod, J. M.; Rosei, F. Covalently bonded networks through

surface-confined polymerization. Surface Science 2013, 613, 6-14.

26. Dinca, L. E.; Fu, C.; MacLeod, J. M.; Lipton-Duffin, J.; Brusso, J. L.; Szakacs,

C. E.; Ma, D.; Perepichka, D. F.; Rosei, F. Unprecedented Transformation of

Tetrathienoanthracene into Pentacene on Ni (111). ACS Nano 2013, 7 (2), 1652-1657.

27. Dinca, L. E.; MacLeod, J. M.; Lipton-Duffin, J.; Fu, C.; Ma, D.; Perepichka,

D. F.; Rosei, F. Tailoring the Reaction Path in the On-Surface Chemistry of

Thienoacenes. The Journal of Physical Chemistry C 2015, 119 (39), 22432-22438.

28. Bronner, C.; Stremlau, S.; Gille, M.; Brauße, F.; Haase, A.; Hecht, S.; Tegeder,

P. Aligning the band gap of graphene nanoribbons by monomer doping. Angewandte

Chemie International Edition 2013, 52 (16), 4422-4425.

29. Cai, J.; Pignedoli, C. A.; Talirz, L.; Ruffieux, P.; Söde, H.; Liang, L.; Meunier,

V.; Berger, R.; Li, R.; Feng, X. Graphene nanoribbon heterojunctions. Nature

Nanotechnology 2014, 9 (11), 896-900.

30. Kawai, S.; Saito, S.; Osumi, S.; Yamaguchi, S.; Foster, A. S.; Spijker, P.;

Meyer, E. Atomically controlled substitutional boron-doping of graphene

nanoribbons. Nature Communications 2015, 6, 8098.

31. Basagni, A. Covalent stabilization of 2D self-assembled nanostructures on

surfaces. 2016.

Chapter 2: Literature Review 7

Chapter 2: Literature Review

This chapter begins with a historical background of low-dimensional materials

(section 2.1) and reviews the literature on the different polymerisation approaches as

well as strategies employed so far for improving the quality of polymers (section 2.2

to 2.4). Section 2.5 highlights the implications from the literature and develops the

conceptual framework for the study.

2.1 HISTORICAL BACKGROUND

Although the 0D and 1D family of graphene i.e. fullerenes1 and nanotubes2-3

were synthesised in 1985 and 1991, respectively, it took until 2004 for the

experimental discovery of the first 2D material, graphene,4-5 to revolutionise solid state

physics. This discovery won the 2010 Nobel Prize in Physics "for groundbreaking

experiments regarding the two-dimensional material graphene" 6 for Andre Geim and

Konstantin Novoselov. In fact, the existence of graphene invalidated Peierls’7 and

Landau’s8 theories which argued that 2D material could not exist due to

thermodynamic instability. So, it is not exaggerated to say that graphene opened the

door of experimental research to synthesising new engineered low-dimensional

materials and caused an exponential growth in the field of surface science.9-11

However, the electronic and optoelectronic applications of graphene are limited by its

inherent zero energy-gap.6 For example, a band gap in the range of 0.4 eV (or more)

is required to make a field-effect graphene transistor.12 The band gap in this range

provides excellent switching capabilities, as well as an acceptable on–off ratio (Ion/Ioff,

of between 104 and 107) for the transistor.

A few years after the discovery of graphene many other 2D structures were

obtained by the mechanical exfoliation of layered materials that occur in nature like

BN, MoS2 WS2, and many others.13-16 Nowadays more than 5000 2D materials have

been identified. Amongst these natural 2D crystals many have an electronic gap and

could potentially be used for electronic and optoelectronic applications, but important

advantage in growing customized structures using surface science tools: the possibility

to fabricate novel low-dimensional materials with the desired physical and chemical

properties by tailoring the topology and chemistry of the structures.

8 Chapter 2: Literature Review

Polymers are typically fabricated by exploiting solution-based organic chemistry

approaches.17 These flask type approaches provide polymerisation in solution, but the

structures produced with these approaches tend to be folded and limited in lateral size

owing to insufficient solubility of the products (unless they specifically functionalized

for solubility) which occurs as result of the increasing size of the structures.

Transferring the classical reactions to a solid surface in ultrahigh vacuum (UHV)

conditions provides control over the reactions and also offers the possibility to

fabricate complex polymers with tunable bandgap. UHV offers control over the

thickness of deposited organic films down to the monolayer range.18 Deposition on an

atomically clean substrate by evaporation under UHV conditions potentially offers

advantages over the solution-based growth in liquid environments for the following

reasons:

• On-surface synthesis in UHV widens the available reaction temperature

range.

• The substrate employed in the UHV approach also offers an addition control

over the reaction due to the different reactivity and catalytic properties of

different surfaces.

• Eliminates the chance of out-of-plane polymer aggregation compared to

solution synthesis owing to the templating role of surface as well as the self-

limiting catalytic effect of the surface.

• The precursors do not need to be functionalized for solubility.

The main distinction between a flask-type reaction and a surface-confined

reaction is a difference in the reaction pathway as well as in the kinetics on the surface.

This difference could arise from either the reaction environment and presence of

surface atoms (catalysis role of surface) or confinement of the monomers to the surface

(template role of the surface). Despite all the advantages afforded by the presence of

the surface, the available approaches comprise some intrinsic limitations due to the

irreversible nature of many reactions under UHV condition which prevent the self-

repairing process. Therefore, polymers formed by surface-confined reactions tend to

be irregular and defective. Thus, fabricating high-quality and extended low-

dimensional materials is a demanding objective in surface science. Apart from this

challenge, it is not possible to use polymer architectures directly on metal surfaces for


electronic and optoelectronic applications, as this metal surface is conductive.

Recently some decoupling approaches have been developed to address this issue

(section 2.5). Identifying and/or implementing a solution to separate the polymer from

the metal substrate is out of the scope of this thesis, though it is very important to find

an answer for this. In this thesis, I am focusing on understanding new surface-confined

reactions and investigating new technique to grow high-quality polymers, and in the

following section, I will review the main publications that demonstrate the

considerable progress made to date in the synthesis of covalent architectures through

on-surface reactions and their limitations.

2.2 ULLMANN REACTION

In this section, I review the basic concepts behind the Ullmann reaction. Then, I

highlight the progress made in understanding the key parameters that influence the

reaction mechanism and the quality of the products. Additionally, the challenges in the

Ullmann reaction and the corresponding approaches for overcoming them will be

critically discussed.

2.2.1 Ullmann reaction: an important on-surface reaction

The term “Ullmann reaction” typically refers to aryl-aryl coupling and involves

catalytic activation of halogenated moieties which leads to the formation of C-C bonds.

This reaction was named after Fritz Ullmann, who in 1901 reported the synthesis of

symmetric biphenyls from aryl halides in sulphuric acid and in the presence of fine

copper powder.19 It took nearly 100 years before Bent et al. successfully translated the

Ullmann reaction into UHV conditions.20-21 Soon after, in 2000, the pioneering work

by Hla et al.22 illustrated the coupling of biphenyls via the Ullmann reaction using

STM tip manipulation. The STM tip was used to cleave the halogen from iodobenzene,

lift the phenyl radicals into proximity with one another and finally to trigger the C-C

coupling. This work provided an important proof-of-principle as well as a fundamental

insight into on-surface reaction mechanisms. However, this approach is not promising

for up-scaling.

It was after the seminal work of Grill et al. in 200723 that numerous research

groups began to widely employ the Ullmann reaction to synthesise 1D and 2D

polymers. Grill et al. reported the covalent linkage of tetraphenylporphyrin (TPP)

molecules. Their experiments showed that control over the dimension of the product


can be imposed by carefully selecting the chemical structure of the building blocks.

The authors used n-bromo TPP molecules (with n=1, 2, 4 for synthesis of 0D, 1D and

2D structure, respectively) and deposited them onto Au (111) at room temperature.

Subsequently, thermal activation of the sample at 607 K was performed to trigger the

reactions including both the dissociation of C-Br and association of C-C. Alternatively,

they found that the molecules could be activated prior to deposition by annealing them

in the evaporator. The desired covalently-bonded structures were achieved by thermal

activation in both approaches.

Grill’s report was a significant development in using the Ullmann method for

producing both 1D and 2D structures on different surfaces.24-25 Since then, the

Ullmann reaction has been widely used to fabricate GNRs.26 GNRs are an important

class of 1D polymers with a number of attractive electronic characteristics like

excellent carrier transport properties in presence of a controllable bandgap introduced

through confinement.27

2.2.2 Reaction mechanism, limitation and developments

Figure 2-1 shows the reaction scheme for Ullmann coupling. In the first step, the

halogen (X=I, Br or Cl) is removed from the precursor and phenyl radicals are

produced. The abstraction of halogens can be triggered by thermal annealing, however

in some cases- depending the surface and halogen- the catalytic effect of the surface

at room temperature is enough for cleavage of the halogen. The C-X bond can be

cleaved without risking decomposition of the phenyl ring (C-H or C-C cleavage)

because the C-C bond is much stronger than the C-X bond. The radicals produced in

the first step diffuse on the surface and subsequently form an organometallic (OM)

intermediate with carbon–metal–carbon (C–M–C) bonds.28 In the second step, an

external stimulus such as thermal activation is employed to eject the metal atom from

the intermediate, and consequently, C-C coupling occurs to form covalently cross-

linked networks.29 The C-C coupling at the last step is irreversible, and this prevents

any self-healing processes in the networks, i.e. scission and reformation of bonds in

the defect and irregular networks to form defect-free and regular 1D or 2D structures.

Figure 2-1: Schematic reaction for Ullmann coupling.


Recent work by Fan et al. attempted to shed light on the mechanism of the

Ullmann reaction by studying 4,4′′-dibromo-m-terphenyl (DMTP) precursor at low

temperature.30 DMTP was deposited onto Cu(111) held at 90 K and also on Ag(111)

held at 88 K followed by annealing at different temperatures to investigate the

structural transformation of the sample described below and shown in Figure 2-2.

• Annealing at 97 K led to a large-area ordered 2D network stabilised by

Br···Br halogen bonds is formed (Figure 2-2a).

• Annealing up to 113 K resulted in densely packed DMTP−Cu(Ag)

coordination network (Figure 2-2b), which is stabilized by the weak

Coulomb attraction between the precursor molecule and metal adatoms.

XPS revealed that the molecule remained intact during this step.

• In the case of Ag, the molecule dehalogenated at 333 K and

Br−(MTP)−Ag−(MTP)−Br organometallic dimers formed (Figure 2-2c).

The halogen remained bonded to a coordination network. The DMTP−Ag

organometallic network counts as direct intermediate phase to transfer to C-

C covalently-bonded product for the Ullmann reaction.

Figure 2-2: Overview STM images representing transformation of DMTP molecules from a) halogen-

bonded network to b) DMTP−Cu coordination network and on Ag c) to organometallic networks.

Reprinted with permission from [30]. Copyright (2018) American Chemical Society.

Additionally, this study highlights that the C-Br bonds cleavage is likely

catalysed by metal adatoms, not by the metal atoms in the terraces, since metal adatoms

were found in close proximity to C-Br bonds.

But what happens to the cleaved halogen on the surface? Normally, the halogen

remains as a by-product chemisorbed on the surface, and the presence of surface-

adsorbed reaction by-products may play a role in the formation of the polymer product.


For example, following Ullmann dehalogenation, surface-adsorbed halogens may

have a negative influence on the formation of high-quality extended polymers, since

they can disturb radical diffusion31 and block the surface from further polymerization

of the desirable product,32 as shown in Figure 2-3, or even re-bond to the activated

monomer or intermediate.29, 33 It is debatable whether the halogen presence on the

surface is detrimental to the long-range ordering of the polymer or not, but it appears

that the halogens have some influence on the structure of polymer.34-35

Figure 2-3 a) As-deposited at RT and b) thermally activated hexaiodo-substituted macrocycle

cyclohexa-m-phenylene (CHP) on Cu(111). The white circle highlights radicals surrounded by iodine

atoms. Reprinted with permission from [32]. Copyright (2010) American Chemical Society, c)

organometallic and biphenyl formed from bromobenzene on Cu(111). Reproduced from [28] with

permission from The Royal Society of Chemistry.

Additionally, the electronegativity of the halogen can create the possibility to

release metal atoms; in the case of Au(111), this can lead to a reconstruction of the

surface.36-37 Hence, removing the halogens has been a major pursuit in this field, and

to circumvent the challenge of halogen adsorption, studies have been carried out to

find a solution for the so-called by-product problem.38 One alternative approach is

using a reaction that either does not produce any by-products (e.g. cycloaddition

reactions) or alternatively, produces volatile by-products (reactions such as

dehydrogenation, decarboxylation and Glaser coupling). These low-by-product

reactions are known as “clean reactions”. While these reactions remove the possibility

for polymer/by-product interactions, this is itself not a panacea and will not cure the


defect and irregularity of polymers. Furthermore, they can pose their own drawbacks.

For example, in the case of Glaser coupling, a variety of reaction pathways such as

cross-coupling, homocoupling, cyclotrimerization and trimerization can occur

simultaneously and thus lead to non-uniformity in the resultant products.39 Similarly,

the Bergman reaction suffers from simultaneous isomerization of monomers during

cyclization.40 Amongst these clean reactions, the decarboxylation reaction, which has

volatile by-products leaving the surface (H2 41

and CO2 42), has rarely been studied. A

discussion about this reaction has been provided in section 2.3, since part of this thesis

is devoted to an investigation on decarboxylation with the goal of enhancing our

understanding of this reaction.

On the other hand, the negative influence of halogens on product quality is still

debated. In fact, recent papers argued that the retention of the halogen on the surface

could have a positive effect. For example, Lipton-Duffin et al.43 found that the

detached halogens adsorb between polymer lines, and both the halogens and the

polymers exhibit a high degree of order, suggesting a symbiotic ordering (see Figure

2-4).

Figure 2-4: STM image of 1,4-diiodobenzene lines after annealing at 500K on Cu(110). Reproduced

with permission from [43].

Recently the contribution of halogens to the structure of OM and polymers has

been studied, revealing that the type of halogen can also affect the final product

structure.35, 44 Galeotti et al. reported a comprehensive comparison of five different

halogenated-molecules para-positioned on Cu(110) to reveal the halide by-product35

1,4-dibromobenzene (dBB), 1,4-diiodobenzene (dIB), 1,4-dichlorobenzene (dCB), 1-

bromo-4-chlorobenzene (BCB) and 1-bromo-4-iodobenzene (BIB). The authors used

fast-XPS spectra of the C 1s core level as a function of temperature to study the

evolution from the OM to polymeric phase for the five precursors. As shown in Figure


2-5, the transition of OM to the polymer is halogen-dependent. The fully

dehalogenated phase of the OM formed for dIB, dBB and BIB precursors form

immediately upon deposition due to C–I and C–Br dissociation occurring below RT,

whereas the partially dehalogenated dCB and BCB precursors, which are stable up to

393 K, form short OM chains upon deposition. The fast-XPS maps not only reveal the

temperature range of the polymerization reaction, they also illustrate the behaviour of

the C–C coupling step of the reaction; i.e. dIB shows a gradual shift in C1s towards

higher binding energy (BE) with temperature in contrast to the sharp and sudden shift

for bromine and chlorine-containing molecules. A difference was seen in STM images

where the OM and polymer for dIB followed the same direction on the surface whereas

this is not the case for all other molecules. This is due to the difference of preferred

adsorption site for different halogens, i.e. iodine adsorbed on hollow sites whereas Br

and Cl adsorbed on short bridge sites on Cu(110). Thus, authors inferred that the

transition of OM to polymer in dIB occurs without disturbances to neighbouring

chains, and therefore, the reaction occurs gradually, but for other precursors an

additional barrier are required due to a rotation diffusion needed for the phenyl groups.

Figure 2-5: Fast-XPS C 1s spectra during annealing of each precursor on Cu(110) dosed at RT.

Reproduced with permission from [35].

However, to further complete the understanding of the halogen impact on

polymerization, one must be mindful that other factors such as substrate and precursor

etc. need to be considered, because a combination of factors can act in concert to

determine the reaction pathway and/or the resultant products. Section 3.1 briefly

reviews the monomer and substrate effects on the reaction and products.


2.2.3 Using post-reactions to remove by-products from the surface

To improve the quality and spatial extension of the polymer, in particular 2D,

halogens may need to be removed from the surface to limit their interference with the

product. Halogens are chemisorbed on the surface, and the halogen-halogen

interactions are much weaker than the halogen−surface interactions. Therefore, it is

more likely that two halogens will desorb atomically rather than molecular halogen

due to the energy cost of desorbing of them, however, the surface is also important in

the desorption form of halogen. For example, atomic halogens desorb from Ag and Au

surfaces,29, 45 while desorb from Cu surface as CuX, where X is halogen.46

To desorb halogens from the surface using thermal treatment, it should be

considered that the halogen binding energies with the surface are much higher than

dehalogenation barriers. Thus, desorption of halogens from the surface can require

considerably higher temperatures than the dissociation temperature of the polymer, as

indicated by the energies in Table 2-1. For example, the halide by-products from the

polymerization of CHP were desorbed after annealing at 825 K47 while the

polymerization temperature is 570 K. This is challenging owing to the fact that the

required temperature for desorbing halogen might surpass of dissociation temperature

of the polymer.

The prerequisite of such high thermal energy for desorption of the halogen raises

a number of problems: a) the physisorbed reaction product (oligomer/polymer) can be

co-desorbed or even destroyed,36 b) the release of metal adatoms which can bond with

the dehalogenated molecules and form metal-organic architectures48 and c) the

reconstruction of the surface owing to the high temperature annealing.36 Therefore,

removing the halogen is strongly dependant on the type of halogen and the surface,

and it is a challenging undertaking.

Table 2-1: Comparison of atomic and molecular desorption energies as well as energy barriers for the

dehalogenation for bromine and iodine on Cu(111), Ag(111) and Au(111).29 All values are in eV.

Bromine Iodine

eV Eatom Emol Edehalogenation Eatom Emol Edehalogenation

Cu(111) 3.19 3.83 0.66 2.97 3.64 0.40

Ag(111) 3.23 3.93 0.81 3.01 3.74 0.52

Au(111) 2.80 3.07 1.02 2.76 3.24 0.71


In 2014, Bronner et al. tracked the halogen during the on-surface formation of

GNRs using a Br-substituted molecular precursor by employing temperature-

programmed desorption (TPD).33 They found that the halogens left the surface as HBr

(not Br2) after initiating the second step of the reaction called cyclodehydrogenation.

The cyclodehydrogenation reaction produces molecular hydrogen as a by-product, and

exposure of the sample to this molecular hydrogen led to removal of the halogen from

the surface. Figure 2-6 shows TPD curves for various masses during desorption of a

multilayer of 10,10′-dibromo-9,9′-bianthryl (DBBA).

Figure 2-6: Tracking the desorption of different masses during annealing DBBA multilayer using

TPD, the desorbing fragments assigned in the figure nest to related curve and the correspond

schematic cartoon for desorption process are shown in the bottom of image. Reprinted with

permission from [33]. Copyright (2015) American Chemical Society.

This work inspired other groups to study the possibility for the removal of

halogens using hydrogen dosing by XPS and STM as well49 (see Figure 2-7). In 2017,

Fasel’s group also proposed that the iodine desorption temperature from Au(111),


which occurs after annealing up to 683 K for iodine alone, was reduced to 646 K, if

atomic hydrogen resulting from the cyclodehydrogenation is available on the surface.44

Figure 2-7: Confirming the halogen removal of surface resulted from polymerization of 4,5,9,10-

tetrabromo-2,7-di-tertbutylpyrene precursor on Au(111), using molecular hydrogen dosing and

thermal annealing. Republished with permission from [49]. Copyright (2017) Royal Society of

Chemistry.

However, this alternative approach has been rarely studied and, for example, was

not examined on more reactive surfaces such as Cu. I investigated the bromine removal

from the Cu(111) and Cu(110) using atomic hydrogen in chapter 5.

2.2.4 Error correction by exploiting the metal-organic intermediate

As previously discussed in the section describing the Ullmann mechanism, the

intermediate product obtained from dehalogenation can bond with surface metal atoms

or diffusing adatoms on the surface and thus form surface-stabilized radicals. The

signature of the metal atoms in the organometallic intermediate can be tracked by both

STM and XPS. STM shows the presence of an organometallic in two ways. Firstly, it

can be visualized through a different contrast in a STM image, sometimes discernible

as bright features, as shown in Figure 2-8. In addition, the length of intermolecular


linkage from the image can reveal if the chemical identity of the bond is covalent C-C

or C-M-C.

Figure 2-8: a,b) STM image presenting the OM and polymer network of 1,3,5-tris(4-

bromophenyl)benzene (TBB) on Cu(111), and c,d) the corresponding models, respectively. The red

circle in (a) shows the coper atom and the arrow shows the backbone of (TBB) molecule. Adapted

with permission from [50] Copyright (2014) American Chemical Society.

In 2004, Weiss et al.51 found that there is a strong intermolecular interaction that

aligned the partially-reacted para-diiodobenzene monomer units. They named these

structure “protopolymer chains”. Although these authors did not visualize the metal

adatoms and the subsequent carbon-metal linkage, they suggested that the appropriate

choice of substrate and monomer is a key factor for components to align in this orderly

way. A few years later, the chemical nature of these strong intermolecular interaction

was attributed to the C-Cu-C,43 where the spacing of periodicity of the protopolymer

measured 5.5 Å, 30% larger than the one for covalent bonded poly-para-phenylene.

Annealing this metal-organic polymer reduced the periodicity to 4.1 Å, consistent with

the covalently-bonded polymer. In the same year, Lackinger’s group were able to

capture the metal adatoms in STM image as bright spots between adjacent molecules.52

XPS also provides the possibility to confirm the presence of an organometallic

phase. XP spectra of the carbon core level include an extra feature peak at low binding

energy (with respect to the carbon peak of the aromatic ring of monomer)


corresponding to the C-M chemical signature. Figure 2-9 shows the organometallic

peak at lower binding energy.53 The transformation from organometallic to covalently-

bonded polymers is accompanied by a shift in the main C 1s peak to higher binding

energy.

Figure 2-9: C1s core level XPS for dBB on Cu(110), peak 1 is attributed to the OM species, peak at

286.4 eV corresponding to carbon-bromine bonds in intact molecule at LT vanishes at RT, peak 2 and

3 are assigned to C2 carbons in the aromatic ring of the dehalogenated molecules. Adapted with

permission from [53] Copyright (2013) American Chemical Society.

Bjork et al. simulated the expected core level shifts for the carbon and halogen

for both bromobenzene and iodobenzene on the 111 facets of Cu, Ag and Au, and

provided a reference for XPS experiments.29


Figure 2-10: a) The C1s and b) the Br3d and I4d core-level shifts simulation, for the systems depicted

above each plot. For the methine C-atoms (C-atoms with an H-atom) the average core level shifts are

indicated. Adapted with permission from [29]. Copyright (2013) American Chemical Society.

Within the general groundwork established above, and considering the reversible

nature of the C-M linkage, error correction of the OM structure and that could

potentially enhance the quality of the final products,54-55 i.e. once the C-M bond

scission occurring, the radicals can rearrange to impair the errors and defects of the

networks. One main prerequisite to take advantage of the reversible nature of the OM

is that the temperature at which the OM converts to the covalent structure is

sufficiently higher than the temperature at which the C-M linkage is reversible. In this

aspect, the Ag substrate is a better candidate amongst the coinage metals, due to its

intermediate reactivity compare to Au and Cu. The C-M bond is very strong on Cu,

which reduces the reversibility of C-Cu bond of OM. On the other hand, the OM are

not very stable on Au and to date only a few studies have observed OM structures on

Au.37, 56-57

Lackinger’s group investigated the possibility of the improving the structural

quality of 2D covalent polymers using metastable organometallic networks.54 1,3-

bis(p-bromophenyl)-5-(p-iodophenyl) benzene (BIB) was deposited on Ag held at

different temperatures to investigate equilibration of the OM before forming the

covalent structure. BIB adsorbed partially dehalogenated upon RT deposition and thus

formed irregular OM networks. The OM structures rearrange into hexagonal networks

after annealing up to 398 K. This is useful, owing to the fact that the OM structure is

topologically similar to covalent network structure and thus during transformation of

the OM to covalent network the topology of networks remains unchanged. The


conversion of OM to covalent structure started at about 443 K and completed at 523

K, as illustrated in Figure 2-11.

Figure 2-11: transformation of a) self-assembled BIB molecules to b) a well-ordered hexagonal OM

and to C) covalently-linked networks. Published by The Royal Society of Chemistry [54].

2.2.5 Hierarchical fabrication: Sequential activation

The ideal conditions for the formation of well-ordered 2D networks are largely

dictated by the radical recombination step of the polymerization process, which

includes diffusion and coupling. Achieving high-quality and extended covalently

bonded networks in 2D is even more challenging than in 1D, owing to the increased

complexity of monomers. 2D networks suffer from lack of uniqueness of product, i.e.

usually there are variety of unwanted polygons in the network. These defects can be

attributed to two major issues: formation of undesired C-C connections and the

incomplete coupling of related radicals. Although these issues are not arising from the

by-product obstacle which was discussed for Ullmann in section 2.2.2, the Ullmann

reaction can offer a nice trick to address the latter issue in some cases. Improving the

quality of 2D networks can be possible by exploiting the difference in bond

dissociation energy for Cl, Br and I. This difference in energy implies that the

dehalogenation temperature is different for different C-X bonds. Thus, sequential

growth of 2D polymers is possible using rational pre-defined building block with two

different halogens. In 2012, hierarchical Ullmann coupling, established by Lafferentz

et al., exploited the different activation temperatures of the different halogens, and as

such, has become a promising methodology for enhancing the structural quality of 2D

polymers.58

A comparison between the hierarchical polymerization pathway of 5,15-bis(4′-

bromophenyl)-10,20-bis(4′-iodophenyl)porphyrin (trans-Br2I2 TPP)58 and the BIB

precursor59 illustrates an instructive lesson about importance of symmetric monomer

effects on the product. The deiodination of BIB was followed by the formation of a


dimer (called TBQ), and finally, sequential debromination led to the formation of 2D

networks with relatively high defect densities, see Figure 2-12. The defects occurred

due to the unsymmetrical substitution of bromine and iodine in the precursor.

Figure 2-12: Scheme of the sequential activation mechanism and corresponding STM images of

hierarchical Ullmann reaction of BIB precursor onto Au(111). Adapted with permission from [59]

Copyright (2014) American Chemical Society.

In contrast, trans-Br2I2 tetraphenyl porphyrin is a fourfold-symmetric monomer.

The deiodination of trans-Br2I2 tetraphenyl porphyrin resulted in the formation of 1D

chains followed by subsequent interlinking into 2D networks, as shown in Figure 2-13.

Changing from a low-symmetry isomer to the higher symmetry trans-Br2I2 tetraphenyl

porphyrin significantly increased the structural quality as well as the spatial extension

of the products due a zipper effect of the 1D chains.

Figure 2-13: Scheme of the sequential activation mechanism and corresponding STM images of

hierarchical Ullmann reaction of trans-Br2I2TPP molecules on Au(111).Reprinted with permission

from [58]. Copyright (2012) Springer Nature.


This approach has also been adopted for different halogens.60-62 The sequential

activation temperature idea has also been implemented through the combination of

different types of reactions using different activation groups, such as halogen and

hydrogen63 or bromine functionalization group and halogen.64 This approach has been

a valuable tool for improving the increasing length of GNR.44 In addition, the

sequential reaction concept was adapted to pairing Glaser coupling with a subsequent

dehydrogenative coupling reaction.65 The domino reactions of a bifunctional precursor

with carboxyl and alkyne groups (6-ethynyl-2-naphthoic acid (ENA) molecule)

resulted in chains with lengths up to 100 nm. In addition, conceptually similar

hierarchical reaction pathway has been employed in a number of studies.55, 66-67

2.3 DECARBOXYLATION REACTION

On-surface C-C coupling through decarboxylative polymerization is a relatively

recent achievement. The formation of 2D metal−organic coordination networks

(MOCNs) on Cu(100) substrate was published by Kern's group in 2002.68 This work

and the following studies on carboxylic acids69-71 produced MOCNs based on

deprotonation of carboxylic group, however probably inspired to the decarboxylation

reaction as a new polymerization approach. This approach is based on the removal of

a –COOH group to form reactive moieties for polymerization.72 The covalent coupling

of 2,6-naphthalenedicarboxylic acid (NDCA) reported by Gao is completed in a three-

step process:

Dehydrogenation, which leads to the formation of metal carboxylate (-

COO-M).

Decarboxylation, which leads to the corresponding organometallic

intermediate species.

Polymerization of chains, which happens at higher temperature via

elimination of metal atoms and covalent coupling of the radicals.

A schematic of the reaction pathway and corresponding STM images of OM and

poly-2,6 naphthalene from the decarboxylative polymerization of NDCA are shown in

Figure 2-14.


Figure 2-14: a) Reaction pathways for decarboxylation of NDCA; b) STM images of organometallic

of NDCAs surface; c) poly-2,6 naphthalene on Cu(111). Adapted with permission from [72].

Copyright (2014) American Chemical Society.

The decarboxylation reaction has also been employed by Kern’s group to grow

2D networks using the 1,3,5-tris(4-carboxyphenyl)benzene (TCPB) precursor.73 They

used photoemission spectroscopy in addition to the microscopy to study the

decarboxylation reaction including deprotonation, decarboxylation and covalent

coupling. During deposition on Cu(111), the TCPB fully deprotonated and STM

revealed a well-ordered self-assembled structure linked by strong ionic hydrogen

bonds. Thermal treatment up to 453 K triggered the coupling reaction to form an

organometallic network with different pore sizes and geometries including pentagonal,

hexagonal and heptagonal structures. STM confirmed the formation of the covalent

structure after annealing up to 493 K. However, only 5-10% of the product constituted

an ordered hexagonal structure. The variation in the pore size of the structure was

attributed to the hydrogen abstraction on the phenyl due to the annealing treatment at

493 K. This is demonstrated in Figure 2-15. The resulting network is not well-ordered.


Figure 2-15: STM image of a) self-assembled b) organometallic (Cu adatoms are visible as circular

protrusions) and c) covalently-bonded network obtained from decarboxylative coupling of TCPB. The

inset shows a high-resolution image with a superimposed molecular structure. d-f) XPS spectra of the

C 1s and O 1s. Adapted with permission from [73]. Copyright (2016) American Chemical Society.

Covalent coupling via decarboxylation has not yet been well-studied (only the

two above mentioned studies). Many of the details of the process of the

decarboxylation reaction need to be further studied to shed light on the mechanism of

the reaction. An investigation of the decarboxylation reaction of isophthalic acid (IPA)

is provided in chapter 5.

2.4 PYRIDINE-BASED POLYMER

As previously mentioned, chemical control is a promising route for opening the

band gap in polymers. In this aspect, azabenzenes (azines) are very good candidates.

They are organic molecules consisting of a benzenic ring where one or more carbon

atoms are replaced by nitrogen atoms. In the past few years, pyridine and its derivatives

have been studied for fabricating metal-organic frameworks (MOFs) 74 or self-

assembled structures.75 However the C-C coupling reaction of these important ligands

has been rarely studied.76 There are a few studies investigating the growth of nitrogen

doped graphene via different approaches, as well as looking at N-graphene’s

applications.77-81 Here, I provide an overview of the main studies using pyridine-based

molecules as precursors, and in chapter 4 I present my study, aimed at understanding

whether the presence of nitrogen changes the adsorption of molecules as well as the

reaction pathway.


When compared to benzenes, azines can have different adsorption behaviour on

metal surfaces owing to the presence of the nitrogen lone pair.82 Pyridine can be either

in a flat-lying configuration affected by π orbitals of the ring or vertical/tilted with

respect to the surface owing to the nitrogen lone pairs. For example, the study of single

pyridine molecules on Ag(110) using STM and DFT revealed that:

a) molecule adsorbs on the surface either in the stand-up or flat-lying

configuration at low temperature (see Figure 2-16),

b) the configuration strongly depends on the surface coverage, and the stand-up

configuration is favoured with increasing the coverage due to increasing

molecule–molecule interactions,83 as shown in Figure 2-17,

c) both configurations (in particular the flat-lying one) were influenced by the

electrostatic interaction between pyridine (the donor) and metal atoms of the

surface (the acceptor).

Figure 2-16: a,b) STM images of two types of single pyridine molecules adsorbed on Ag(1 1 0) at 13

K, and c,d) Equilibrium atomic structures of vertically upright (stand-up) and flat-lying configuration

(Pf) configurations of pyridine on Ag(1 1 0) surfaces. Top views of the two topmost layers and cross-

section views of the four topmost layers of the 3 × 4 Ag(1 1 0) surface in a supercell are shown in

each configuration. Large and small gray spheres represent Ag atoms on the surface layer, and on a

layer below it, respectively. The black, green, and red circles represent carbon, hydrogen, and nitrogen

atoms, respectively. Reprinted with permission from [83].


Figure 2-17: a-c) STM images of 0.04, 0.07, and 0.17 nm2 for pyridine coverage. The height of feature

a was always higher than that of feature b. d) Plot showing the dependency of the configuration to the

coverage surface by comparing relative population ratio of pyridine configurations as a function of the

total surface concentration. Reprinted with permission from [83].

In addition to the coverage, the temperature and the type of metal surface also

influence the adsorption behaviour of the pyridine-based molecule. This is because the

competitive bonding of the nitrogen lone pair and π-aromatic ring of the molecule is

strongly affected by the molecule-molecule or substrate-molecule interaction. Lee et

al. 84 proposed four possible adsorption configurations for pyridine on metal surfaces:

a) a stand-up configuration via the nitrogen lone pair, b) tilted adsorption owing to

interaction of both the nitrogen lone pair and π orbital with the surface, c) flat-lying

adsorption via interaction of π orbital of ring and metal surface, and finally, d) edge-

on adsorption through the N and its neighbouring C atom, as shown in Figure 2-18.


Figure 2-18: The possible adsorption geometries of pyridine on a metal surface. Reprinted with

permission from [84].

Recent reports of the reactions of heterocyclic aromatic halides highlighted the

importance of the influence of the adsorption configuration of pyridine-based molecule

for on-surface reactions.85-86 For example, Lin et al. employed reflection−absorption

infrared spectroscopy (RAIRS), XPS, NEXAFS and temperature programmed

reaction/desorption (TPR/D) to investigate the adsorption configuration of 2,4-

dibromopyridine and 2,3-dibromopyridine on their debromination reaction on

Cu(100).85 Their XPS and RAIRS results show that 2,4-dibromopyridine adsorbed

mainly intact at 100 K and then mono-dehalogenated upon annealing 250 K.

Annealing up to 450 K increased the debromination, but the molecule started to

decompose upon 400 K annealing and as a result evolved H2, as shown in Figure 2-19.

TPR/D confirmed that H2 desorption mainly occurred above 550 K, which is consistent

with XPS showing multiple peaks in the N1s and C1s core levels.

The relative geometry of the bromines and nitrogen led to the distinct C-Br

scission kinetics, and therefore the authors conducted the same experiments for the

2,3-dibromopyridine precursor. In this case, the molecule adsorbed intact at 110 K and

mono-debrominated at 180 K but underwent full debromination at lower temperature

compared to that of 2,4-dibromopyridine molecule. The difference in debromination

of the molecules has been attributed to the different vicinity of the Br atom to the

surface, as shown in the Figure 2-20.


Figure 2-19: C1s and N1s XPS spectra for 2,4-dibromopyridine on Cu(100) at different temperature.

Reprinted with permission from [85]. Copyright (2015) American Chemical Society.

Figure 2-20: Proposed adsorption configurations for mono-debrominated a, b) 2,4-dibromopyridine

and c,d) 2,3-dibromopyridine, Reprinted with permission from [85]. Copyright (2015) American

Chemical Society.

In chapter 4, I investigate the adsorption and decarboxylation reaction of 2,5-

dicarboxylic acid on Cu(111).

2.5 SUMMARY AND IMPLICATIONS

On-surface reactions provide a versatile route to the synthesis of covalent

organic nanostructures using simple precursor molecules as building blocks. Amongst


the different available reactions, the Ullmann reaction has been intensively studied to

synthesize 1D and 2D polymers. Although this reaction usually includes an annealing

treatment to induce C-C coupling, it leaves metal-halide as a by-product on the surface.

The presence of the by-product has been a hotly-debated issue owing to presumed

detrimental effects on the spatial extension and quality of the product. There have been

a few approaches suggested to overcome this problem:

• Using post-reactions to remove by-products from the surface

• Error- correction of the structure by exploiting of the OM intermediate

• Stepwise activation of precursor to improve the structural-quality

• Using clean reactions which produce no or volatile by-product.

In the case of clean reactions, up to now, only two papers employed the

decarboxylation reaction to synthesize 1D and 2D structures (Gao and Kern’s groups,

respectively), and this reaction is still considered as an unexplored synthetic approach.

In the case of 1D chains, Gao achieved 50 nm extension of the polymer chain. In the

present work, I have focused on the decarboxylation reaction of IPA and PDC

molecules on Cu(111) and Ag(111) surfaces.

As previously discussed using heteroatom in molecular scaffold is an elegant

way to control the physical properties of the products. The PDC molecule has a

nitrogen atom in the core of its structure and is a very interesting candidate for studying

the decarboxylation reaction of nitrogen containing molecules. But it has been reported

that the presence of nitrogen for nitrogen-containing molecule with halogen group

affect the adoption geometry. What is the effect of nitrogen for a molecule with

carboxylic acid group? What is the effect of nitrogen in the reaction pathway of

decarboxylative coupling? These questions are discussed and addressed in chapter 4

where PDC molecule has been studied.

In order to better understand the interplay of carboxyl/carboxylate group and

surface and its influence the adsorption geometry of IPA, an analogous molecule that

does not contain nitrogen has been studied in chapter 5. This provides an interesting

comparison to PDC study in chapter 4

As discussed previously, work is being done towards removing the unwanted

halogen from the surface after reaction. Halogens can be removed by dosing hydrogen


together with using thermal annealing during NGRs formation on a gold surface.

However, this approach has not been widely studied and needs to be investigated on

different surfaces to clarify its impact on the products. It is particularly interesting to

remove halogens from copper due to the strong adhesion of halogen to Cu surfaces in

comparison to Ag and Au. To contribute to this understanding, I have studied the

removal of halogen by-products resulting from the polymerization of dBB on Cu(111)

and Cu(110), using STM and XPS and the results are discussed in chapter 6.

Finally, it should be noted that even if a perfect covalent structure can be made,

there is still an important question to address: what is the point of having a tailor-made

polymer structure on a metal surface? Although this topic is out of the scope of this

thesis, it is important to consider the next steps after having ideal polymers, since the

metal substrate may preclude certain applications of the polymer. It may be useful to

consider adopting the approaches that have been used to transfer chemical vapour

deposition (CVD)-grown graphene to arbitrary substrates.87-88 For example, poly-

methyl methacrylate (PMMA) or polydimethylsiloxane (PDMS) can be spin-coated

on top of the polymer structure, and then the underlying metal is removed via different

approaches, like chemical etching. In this case, thin metal films instead of single

crystals are used as the substrate. This facilitates the decoupling of products from the

surface without compromising the catalytic role of the metal in the reaction. For

example, thin Au(111) or Ag(111) films on mica (or similarly Cu(111) on sapphire)

can be used as a substrate, and the product polymer can be decoupled from the metal

film via standard etching approaches. For example, Au grown on mica can be etched

away using an aqueous potassium iodide (KI) solution and the polymer network

transfer to a dielectric surface. The first successful transformation of the porphyrin

polymer grown by Ullmann coupling on Au(111) has been reported by Beton et al..89

The main drawback of this strategy is the possible contaminations which can be

introduced to the structure.

Although the above-mentioned approaches for transferring the polymer have

been successful, there is still more work to do in this regard. Additionally, it would be

interesting to pursue further studies attempting an alternative approach based on the

direct growth of the polymer on semiconductor or insulator surfaces.


2.6 REFERENCES

1. Krato, H.; Heath, J. O'Brien, SC, Curl, RF & Smalley, RE. Nature 1985, 318,

162-163.

2. Iijima, S. Helical microtubules of graphitic carbon. Nature 1991, 354 (6348),

56.

3. Iijima, S.; Ichihashi, T. Single-shell carbon nanotubes of 1-nm diameter.

Nature 1993, 363 (6430), 603.

4. Geim, A. K.; Novoselov, K. S. The rise of graphene. Nature Materials 2007, 6

(3), 183-191.

5. Ferrari, A. C.; Bonaccorso, F.; Fal'Ko, V.; Novoselov, K. S.; Roche, S.;

Bøggild, P.; Borini, S.; Koppens, F. H.; Palermo, V.; Pugno, N. Science and

technology roadmap for graphene, related two-dimensional crystals, and hybrid

systems. Nanoscale 2015, 7 (11), 4598-4810.

6. Novoselov, K. S.; Geim, A. K.; Morozov, S. V.; Jiang, D.; Zhang, Y.;

Dubonos, S. V.; Grigorieva, I. V.; Firsov, A. A. Electric field effect in atomically thin

carbon films. Science 2004, 306 (5696), 666-669.

7. Peierls, R. Quelques proprietes typiques des corpses solides. Ann. IH Poincare

1935, 5, 177-222.

8. Landau, L. D. On the theory of phase transitions. Ukrainian Journal of Physics

1937, 11, 19-32.

9. Nacci, C.; Hecht, S.; Grill, L., The emergence of covalent on-surface

polymerization. In On-Surface Synthesis, Springer: 2016; pp 1-21.

10. Barton, D.; Gao, H.-Y.; Held, P. A.; Studer, A.; Fuchs, H.; Doltsinis, N. L.;

Neugebauer, J. Formation of Organometallic Intermediate States in On-Surface

Ullmann Couplings. Chemistry – A European Journal 2017, 23 (25), 6190-6197.

11. Shen, Q.; Gao, H.-Y.; Fuchs, H. Frontiers of on-surface synthesis: From

principles to applications. Nano Today 2017, 13, 77-96.

12. Schwierz, F. Graphene transistors. Nature Nanotechnology 2010, 5 (7), 487.

13. Novoselov, K.; Jiang, D.; Schedin, F.; Booth, T.; Khotkevich, V.; Morozov,

S.; Geim, A. Two-dimensional atomic crystals. Proceedings of the National Academy

of Sciences of the United States of America 2005, 102 (30), 10451-10453.

14. Xu, M.; Liang, T.; Shi, M.; Chen, H. Graphene-like two-dimensional materials.

Chemical Reviews 2013, 113 (5), 3766-3798.

15. Miró, P.; Audiffred, M.; Heine, T. An atlas of two-dimensional materials.

Chemical Society Reviews 2014, 43 (18), 6537-6554.

16. Solís-Fernández, P.; Bissett, M.; Ago, H. Synthesis, structure and applications

of graphene-based 2D heterostructures. Chemical Society Reviews 2017, 46 (15),

4572-4613.

17. Sakamoto, J.; van Heijst, J.; Lukin, O.; Schlüter, A. D. Two‐Dimensional

Polymers: Just a Dream of Synthetic Chemists? Angewandte Chemie International

Edition 2009, 48 (6), 1030-1069.

18. Forrest, S. R. Ultrathin organic films grown by organic molecular beam

deposition and related techniques. Chemical Reviews 1997, 97 (6), 1793-1896.

19. Ullmann, F.; Bielecki, J. Ueber synthesen in der biphenylreihe. Berichte der

Deutschen Chemischen Gesellschaft 1901, 34 (2), 2174-2185.

20. Xi, M.; Bent, B. E. Iodobenzene on Cu (111): formation and coupling of

adsorbed phenyl groups. Surface Science 1992, 278 (1-2), 19-32.

21. Xi, M.; Bent, B. E. Mechanisms of the Ullmann coupling reaction in adsorbed

monolayers. Journal of the American Chemical Society 1993, 115 (16), 7426-7433.


22. Hla, S.-W.; Bartels, L.; Meyer, G.; Rieder, K.-H. Inducing all steps of a

chemical reaction with the scanning tunneling microscope tip: towards single molecule

engineering. Physical Review Letters 2000, 85 (13), 2777.

23. Grill, L.; Dyer, M.; Lafferentz, L.; Persson, M.; Peters, M. V.; Hecht, S. Nano-

architectures by covalent assembly of molecular building blocks. Nature

Nanotechnology 2007, 2 (11), 687-691.

24. Lindner, R.; Kühnle, A. On‐Surface Reactions. ChemPhysChem 2015, 16 (8),

1582-1592.

25. Lackinger, M. Surface-Assisted Ullmann coupling. Chemical Communications

2017, 53 (56), 7872-7885.

26. Talirz, L.; Ruffieux, P.; Fasel, R. On‐Surface Synthesis of Atomically Precise

Graphene Nanoribbons. Advanced Materials 2016, 28 (29), 6222-6231.

27. Han, M. Y.; Özyilmaz, B.; Zhang, Y.; Kim, P. Energy band-gap engineering

of graphene nanoribbons. Physical Review Letters 2007, 98 (20), 206805.

28. Lewis, E. A.; Murphy, C. J.; Liriano, M. L.; Sykes, E. C. H. Atomic-scale

insight into the formation, mobility and reaction of Ullmann coupling intermediates.

Chemical Communications 2014, 50 (8), 1006-1008.

29. Björk, J.; Hanke, F.; Stafström, S. Mechanisms of Halogen-Based Covalent

Self-Assembly on Metal Surfaces. Journal of the American Chemical Society 2013,

135 (15), 5768-5775.

30. Fan, Q.; Liu, L.; Dai, J.; Wang, T.; Ju, H.; Zhao, J.; Kuttner, J.; Hilt, G.;

Gottfried, J. M.; Zhu, J. Surface Adatom Mediated Structural Transformation in

Bromoarene Monolayers: Precursor Phases in Surface Ullmann Reaction. ACS Nano

2018, 12 (3), 2267-2274.

31. Batra, A.; Cvetko, D.; Kladnik, G.; Adak, O.; Cardoso, C.; Ferretti, A.; Prezzi,

D.; Molinari, E.; Morgante, A.; Venkataraman, L. Probing the mechanism for

graphene nanoribbon formation on gold surfaces through X-ray spectroscopy.

Chemical Science 2014, 5 (11), 4419-4423.

32. Bieri, M.; Nguyen, M.-T.; Groning, O.; Cai, J.; Treier, M.; Ait-Mansour, K.;

Ruffieux, P.; Pignedoli, C. A.; Passerone, D.; Kastler, M. Two-dimensional polymer

formation on surfaces: insight into the roles of precursor mobility and reactivity.

Journal of the American Chemical Society 2010, 132 (46), 16669-16676.

33. Bronner, C.; Björk, J.; Tegeder, P. Tracking and Removing Br during the On-

Surface Synthesis of a Graphene Nanoribbon. The Journal of Physical Chemistry C

2015, 119 (1), 486-493.

34. Fan, Q.; Gottfried, J. M.; Zhu, J. Surface-Catalyzed C–C Covalent Coupling

Strategies toward the Synthesis of Low-Dimensional Carbon-Based Nanostructures.

Accounts of Chemical Research 2015, 48 (8), 2484-2494.

35. Galeotti, G.; Di Giovannantonio, M.; Lipton-Duffin, J.; Ebrahimi, M.; Tebi, S.;

Verdini, A.; Floreano, L.; Fagot-Revurat, Y.; Perepichka, D.; Rosei, F. The role of

halogens in on-surface Ullmann polymerization. Faraday Discussions 2017.

36. Pham, T. A.; Song, F.; Nguyen, M.-T.; Li, Z.; Studener, F.; Stöhr, M.

Comparing Ullmann Coupling on Noble Metal Surfaces: On-Surface Polymerization

of 1,3,6,8-Tetrabromopyrene on Cu(111) and Au(111). Chemistry – A European

Journal 2016, 22 (17), 5937-5944.

37. Saywell, A.; Gren, W.; Franc, G. g.; Gourdon, A.; Bouju, X.; Grill, L.

Manipulating the conformation of single organometallic chains on Au (111). The

Journal of Physical Chemistry C 2013, 118 (3), 1719-1728.


38. Pham, T. A.; Song, F.; Nguyen, M.-T.; Stöhr, M. Self-assembly of pyrene

derivatives on Au (111): substituent effects on intermolecular interactions. Chemical

Communications 2014, 50 (91), 14089-14092.

39. Gao, H.-Y.; Wagner, H.; Zhong, D.; Franke, J.-H.; Studer, A.; Fuchs, H. Glaser

Coupling at Metal Surfaces. Angewandte Chemie International Edition 2013, 52 (14),

4024-4028.

40. Sun, Q.; Zhang, C.; Li, Z.; Kong, H.; Tan, Q.; Hu, A.; Xu, W. On-surface

formation of one-dimensional polyphenylene through Bergman cyclization. Journal

of the American Chemical Society 2013, 135 (23), 8448-8451.

41. Schmitt, T.; Hammer, L.; Schneider, M. A. Evidence for On-Site

Carboxylation in the Self-Assembly of 4, 4′-Biphenyl Dicarboxylic Acid on Cu (111).

The Journal of Physical Chemistry C 2016, 120 (2), 1043-1048.

42. Franke, M.; Marchini, F.; Zhang, L.; Tariq, Q.; Tsud, N.; Vorokhta, M.;

Vondracek, M.; Prince, K.; Rockert, M.; Williams, F. J. Temperature-Dependent

Reactions of Phthalic Acid on Ag (100). The Journal of Physical Chemistry C 2015,

119 (41), 23580-23585.

43. Lipton‐Duffin, J.; Ivasenko, O.; Perepichka, D.; Rosei, F. Synthesis of

Polyphenylene Molecular Wires by Surface‐Confined Polymerization. Small 2009, 5

(5), 592-597.

44. Di Giovannantonio, M.; Deniz, O.; Urgel, J. I.; Widmer, R.; Dienel, T.; Stolz,

S.; Sánchez-Sánchez, C.; Muntwiler, M.; Dumslaff, T.; Berger, R. On-Surface Growth

Dynamics of Graphene Nanoribbons: The Role of Halogen Functionalization. ACS

Nano 2017, 12 (1), 74-81.

45. Szulczewski, G.; White, J. Thermal and photon-stimulated reactions of

iodobenzene on Ag (111). Surface Science 1998, 399 (2-3), 305-315.

46. Blake, M. M.; Nanayakkara, S. U.; Claridge, S. A.; Fernández-Torres, L. C.;

Sykes, E. C. H.; Weiss, P. S. Identifying reactive intermediates in the Ullmann

coupling reaction by scanning tunneling microscopy and spectroscopy. The Journal of

Physical Chemistry A 2009, 113 (47), 13167-13172.

47. Bieri, M.; Treier, M.; Cai, J.; Aït-Mansour, K.; Ruffieux, P.; Gröning, O.;

Gröning, P.; Kastler, M.; Rieger, R.; Feng, X. Porous graphenes: two-dimensional

polymer synthesis with atomic precision. Chemical Communications 2009, (45),

6919-6921.

48. Villagómez, C. J.; Sasaki, T.; Tour, J. M.; Grill, L. Bottom-up assembly of

molecular wagons on a surface. Journal of the American Chemical Society 2010, 132

(47), 16848-16854.

49. Tran, B. V.; Pham, T. A.; Grunst, M.; Kivala, M.; Stohr, M. Surface-confined

[2+ 2] Cycloaddition Towards One-dimensional Polymers Featuring Cyclobutadiene

Units. Nanoscale 2017, 9 (46), 18305-18310.

50. Chen, M.; Xiao, J.; Steinruck, H.-P.; Wang, S.; Wang, W.; Lin, N.; Hieringer,

W.; Gottfried, J. M. Combined photoemission and scanning tunneling microscopy

study of the surface-assisted Ullmann coupling reaction. The Journal of Physical

Chemistry C 2014, 118 (13), 6820-6830.

51. McCarty, G. S.; Weiss, P. S. Formation and Manipulation of Protopolymer

Chains. Journal of the American Chemical Society 2004, 126 (51), 16772-16776.

52. Gutzler, R.; Walch, H.; Eder, G.; Kloft, S.; Heckl, W. M.; Lackinger, M.

Surface mediated synthesis of 2D covalent organic frameworks: 1, 3, 5-tris (4-

bromophenyl) benzene on graphite (001), Cu (111), and Ag (110). Chemical

Communications 2009, (29), 4456-4458.


53. Di Giovannantonio, M.; El Garah, M.; Lipton-Duffin, J.; Meunier, V.;

Cardenas, L.; Fagot Revurat, Y.; Cossaro, A.; Verdini, A.; Perepichka, D. F.; Rosei,

F.; Contini, G. Insight into Organometallic Intermediate and Its Evolution to Covalent

Bonding in Surface-Confined Ullmann Polymerization. ACS Nano 2013, 7 (9), 8190-

8198.

54. Eichhorn, J.; Strunskus, T.; Rastgoo-Lahrood, A.; Samanta, D.; Schmittel, M.;

Lackinger, M. On-surface Ullmann polymerization via intermediate organometallic

networks on Ag (111). Chemical Communications 2014, 50 (57), 7680-7682.

55. Lin, T.; Shang, X. S.; Adisoejoso, J.; Liu, P. N.; Lin, N. Steering On-Surface

Polymerization with Metal-Directed Template. Journal of the American Chemical

Society 2013, 135 (9), 3576-3582.

56. Zhang, H.; Franke, J. H.; Zhong, D.; Li, Y.; Timmer, A.; Arado, O. D.; Mönig,

H.; Wang, H.; Chi, L.; Wang, Z. Surface Supported Gold–Organic Hybrids: On‐Surface Synthesis and Surface Directed Orientation. Small 2014, 10 (7), 1361-1368.

57. Basagni, A.; Ferrighi, L.; Cattelan, M.; Nicolas, L.; Handrup, K.; Vaghi, L.;

Papagni, A.; Sedona, F.; Di Valentin, C.; Agnoli, S. On-surface photo-dissociation of

C–Br bonds: Towards room temperature Ullmann coupling. Chemical

Communications 2015, 51 (63), 12593-12596.

58. Lafferentz, L.; Eberhardt, V.; Dri, C.; Africh, C.; Comelli, G.; Esch, F.; Hecht,

S.; Grill, L. Controlling on-surface polymerization by hierarchical and substrate-

directed growth. Nature Chemistry 2012, 4 (3), 215-220.

59. Eichhorn, J.; Nieckarz, D.; Ochs, O.; Samanta, D.; Schmittel, M.; Szabelski, P.

J.; Lackinger, M. On-surface ullmann coupling: the influence of kinetic reaction

parameters on the morphology and quality of covalent networks. ACS Nano 2014, 8

(8), 7880-7889.

60. Shi, K. J.; Zhang, X.; Shu, C. H.; Li, D. Y.; Wu, X. Y.; Liu, P. N. Ullmann

coupling reaction of aryl chlorides on Au (111) using dosed Cu as a catalyst and the

programmed growth of 2D covalent organic frameworks. Chemical Communications

2016, 52 (56), 8726-8729.

61. Steiner, C.; Gebhardt, J.; Ammon, M.; Yang, Z.; Heidenreich, A.; Hammer,

N.; Görling, A.; Kivala, M.; Maier, S. Hierarchical on-surface synthesis and electronic

structure of carbonyl-functionalized one- and two-dimensional covalent

nanoarchitectures. Nature Communications 2017, 8, 14765.

62. Shi, K. J.; Yuan, D. W.; Wang, C. X.; Shu, C. H.; Li, D. Y.; Shi, Z. L.; Wu, X.

Y.; Liu, P. N. Ullmann Reaction of Aryl Chlorides on Various Surfaces and the

Application in Stepwise Growth of 2D Covalent Organic Frameworks. Organic

Letters 2016, 18 (6), 1282-1285.

63. Ferrighi, L.; Píš, I.; Nguyen, T. H.; Cattelan, M.; Nappini, S.; Basagni, A.;

Parravicini, M.; Papagni, A.; Sedona, F.; Magnano, E. Control of the Intermolecular

Coupling of Dibromotetracene on Cu (110) by the Sequential Activation of C-Br and

C-H Bonds. Chemistry – A European Journal 2015, 21 (15), 5826-5835.

64. Faury, T.; Clair, S.; Abel, M.; Dumur, F. d. r.; Gigmes, D.; Porte, L. Sequential

linking to control growth of a surface covalent organic framework. The Journal of

Physical Chemistry C 2012, 116 (7), 4819-4823.

65. Held, P. A.; Gao, H. Y.; Liu, L.; Mück‐Lichtenfeld, C.; Timmer, A.; Mönig,

H.; Barton, D.; Neugebauer, J.; Fuchs, H.; Studer, A. On‐Surface Domino Reactions:

Glaser Coupling and Dehydrogenative Coupling of a Biscarboxylic Acid To Form

Polymeric Bisacylperoxides. Angewandte Chemie International Edition 2016, 55 (33),

9777-9782.


66. Zhang, C.; Sun, Q.; Chen, H.; Tan, Q.; Xu, W. Formation of polyphenyl chains

through hierarchical reactions: Ullmann coupling followed by cross-dehydrogenative

coupling. Chemical Communications 2015, 51 (3), 495-498.

67. Li, Q.; Yang, B.; Björk, J.; Zhong, Q.; Ju, H.; Zhang, J.; Cao, N.; Shi, Z.;

Zhang, H.; Ebeling, D.; Schirmeisen, A.; Zhu, J.; Chi, L. Hierarchical

Dehydrogenation Reactions on a Copper Surface. Journal of the American Chemical

Society 2018, 140 (19), 6076-6082.

68. Lin, N.; Dmitriev, A.; Weckesser, J.; Barth, J. V.; Kern, K. Real-Time Single-

Molecule Imaging of the Formation and Dynamics of Coordination Compounds.


69. Kley, C. S.; Cechal, J.; Kumagai, T.; Schramm, F.; Ruben, M.; Stepanow, S.;

Kern, K. Highly adaptable two-dimensional metal–organic coordination networks on

metal surfaces. Journal of the American Chemical Society 2012, 134 (14), 6072-6075.

70. Cechal, J.; Kley, C. S.; Kumagai, T.; Schramm, F.; Ruben, M.; Stepanow, S.;

Kern, K. Convergent and divergent two-dimensional coordination networks formed

through substrate-activated or quenched alkynyl ligation. Chemical Communications

2014, 50 (69), 9973-9976.

71. Seitsonen, A. P.; Lingenfelder, M.; Spillmann, H.; Dmitriev, A.; Stepanow, S.;

Lin, N.; Kern, K.; Barth, J. V. Density functional theory analysis of carboxylate-

bridged diiron units in two-dimensional metal− organic grids. Journal of the American

Chemical Society 2006, 128 (17), 5634-5635.

72. Gao, H.-Y.; Held, P. A.; Knor, M.; Mück-Lichtenfeld, C.; Neugebauer, J.;

Studer, A.; Fuchs, H. Decarboxylative Polymerization of 2,6-Naphthalenedicarboxylic

Acid at Surfaces. Journal of the American Chemical Society 2014, 136 (27), 9658-

9663.

73. Morchutt, C.; Björk, J.; Straßer, C.; Starke, U.; Gutzler, R.; Kern, K. Interplay

of Chemical and Electronic Structure on the Single-Molecule Level in 2D

Polymerization. ACS Nano 2016, 10 (12), 11511-11518.

74. Liu, J.; Lin, T.; Shi, Z.; Xia, F.; Dong, L.; Liu, P. N.; Lin, N. Structural

transformation of two-dimensional metal–organic coordination networks driven by

intrinsic in-plane compression. Journal of the American Chemical Society 2011, 133

(46), 18760-18766.

75. Martínez-Galera, A. J.; de la Torre, B.; Ugeda, M. M.; Gómez-Rodríguez, J.

M. Adsorption and Growth of 1,3,5-Triazine on Cu(111) at Low Temperature under

Ultrahigh Vacuum Conditions. The Journal of Physical Chemistry C 2012, 116 (17),

9568-9574.

76. Matena, M.; Riehm, T.; Stöhr, M.; Jung, T. A.; Gade, L. H. Transforming

Surface Coordination Polymers into Covalent Surface Polymers: Linked

Polycondensed Aromatics through Oligomerization of N-Heterocyclic Carbene

Intermediates. Angewandte Chemie International Edition 2008, 47 (13), 2414-2417.

77. Wang, Y.; Shao, Y.; Matson, D. W.; Li, J.; Lin, Y. Nitrogen-Doped Graphene

and Its Application in Electrochemical Biosensing. ACS Nano 2010, 4 (4), 1790-1798.

78. Usachov, D.; Vilkov, O.; Grüneis, A.; Haberer, D.; Fedorov, A.; Adamchuk,

V. K.; Preobrajenski, A. B.; Dudin, P.; Barinov, A.; Oehzelt, M.; Laubschat, C.;

Vyalikh, D. V. Nitrogen-Doped Graphene: Efficient Growth, Structure, and Electronic

Properties. Nano Letters 2011, 11 (12), 5401-5407.

79. Wang, H.; Zhang, C.; Liu, Z.; Wang, L.; Han, P.; Xu, H.; Zhang, K.; Dong, S.;

Yao, J.; Cui, G. Nitrogen-doped graphene nanosheets with excellent lithium storage

properties. Journal of Materials Chemistry 2011, 21 (14), 5430-5434.


80. Li, Q.; Zhang, S.; Dai, L.; Li, L.-s. Nitrogen-doped colloidal graphene quantum

dots and their size-dependent electrocatalytic activity for the oxygen reduction

reaction. Journal of the American Chemical Society 2012, 134 (46), 18932-18935.

81. Wang, H.; Maiyalagan, T.; Wang, X. Review on recent progress in nitrogen-

doped graphene: synthesis, characterization, and its potential applications. ACS

Catalysis 2012, 2 (5), 781-794.

82. Furukawa, M.; Fujisawa, H.; Katano, S.; Ogasawara, H.; Kim, Y.; Komeda, T.;

Nilsson, A.; Kawai, M. Geometrical characterization of pyrimidine base molecules

adsorbed on Cu (1 1 0) surfaces: XPS and NEXAFS studies. Surface Science 2003,

532, 261-266.

83. Hahn, J. R.; Kang, H. S. Role of molecular orientation in vibration, hopping,

and electronic properties of single pyridine molecules adsorbed on Ag (1 1 0) surface:

A combined STM and DFT study. Surface Science 2010, 604 (3-4), 258-264.

84. Lee, J.-G.; Ahner, J.; Yates Jr, J. The adsorption conformation of chemisorbed

pyridine on the Cu (110) surface. The Journal of Chemical Physics 2001, 114 (3),

1414-1419.

85. Lin, J.-L.; Ye, C.-H.; Lin, B.-C.; Li, S.-H.; Yang, Z.-X.; Chiang, Y.-H.; Chen,

S.-W.; Wang, C.-H.; Yang, Y.-W.; Lin, J.-C. Thermal Reaction of 2, 4-

Dibromopyridine on Cu (100). The Journal of Physical Chemistry C 2015, 119 (47),

26471-26480.

86. Cheng, F.; Ji, W.; Leung, L.; Ning, Z.; Polanyi, J. C.; Wang, C.-G. How

Adsorbate Alignment Leads to Selective Reaction. ACS Nano 2014, 8 (8), 8669-8675.

87. Suk, J. W.; Kitt, A.; Magnuson, C. W.; Hao, Y.; Ahmed, S.; An, J.; Swan, A.

K.; Goldberg, B. B.; Ruoff, R. S. Transfer of CVD-grown monolayer graphene onto

arbitrary substrates. ACS Nano 2011, 5 (9), 6916-6924.

88. Bennett, P. B.; Pedramrazi, Z.; Madani, A.; Chen, Y.-C.; de Oteyza, D. G.;

Chen, C.; Fischer, F. R.; Crommie, M. F.; Bokor, J. Bottom-up graphene nanoribbon

field-effect transistors. Applied Physics Letters 2013, 103 (25), 253114.

89. Wieland, M. B.; Slater, A. G.; Mangham, B.; Champness, N. R.; Beton, P. H.

Fullerenes as adhesive layers for mechanical peeling of metallic, molecular and

polymer thin films. Beilstein Journal of Nanotechnology 2014, 5, 394.

Chapter 3: Research Design 39

Chapter 3: Research Design

This chapter is devoted to introducing the research design used in this thesis to

achieve an understanding of the decarboxylation reaction, the effect of using nitrogen-

containing precursors and to investigate an approach to removing the halogens after

covalent coupling in the Ullmann reaction. Section 3.1 provides an overview of some

key parameters including monomer and substrate which play vital role in surface-

confined reactions.

Molecular beam epitaxy (MBE) growth of molecular layer and XPS and STM

characterisation of the sample have been done in the UHV system at QUT.

Synchrotron radiation-based techniques including photoemission spectroscopy (PES)

and NEXAFS have also been used for some experimental development of this thesis.

Section 3.2 provides an overview of all the instruments used in the study and the basic

principles of these techniques.

3.1 IMPORTANT PARAMETERS IN SURFACE-CONFINED REACTIONS

3.1.1 Precursor molecules*

In general terms, the idea of on-surface synthesis is to use a predesigned

monomer with a carefully-selected type and number of reactive groups. These groups

will leave the monomer after triggering the appropriate reaction and the resultant free

radicals will diffuse on the surface to covalently bind to each other and saturate the

coordination. To form a 1D polymer the precursor needs two reactive groups while to

form a 2D polymer the precursor needs three or more reactive groups. *Additional

control can be introduced via the relative position of substituents within the precursor,

which can be used to define the shape of the polymer. For example, while linear chains

can be produced by a para-substituted precursor, meta-substituted precursors can result

in the 1D chains and 0D macrocyclic structures shown Figure 3-1. Additionally, the

obtained 1D chains from meta-substituted precursors does not have a unique structure,

i.e. there are multiple possible chain structures including zig-zag and armchair.

Therefore, when comparing the para-substituted precursor with the meta-substituted

counterpart, the former tends to produce well-ordered 1D polymers more consistently

owing to the uniqueness of the product. For instance, Lipton-Duffin et al.

The paragraphs in this chapter denoted by the * sign have been published in the following book chapter:

doi.org/10.1016/B978-0-12-803581-8.09239-0. The author is the main contributor to the sections of the book

chapter that appear here.

40 Chapter 3: Research Design

demonstrated the first synthesis of surface-confined conjugated polyphenylene chains

using 1,4- and 1,3-diiodobenzene as precursors, which shows the effect of the relative

position of substituents on the final products.1 Using 1,4-diiodobenzene as the

precursor led to linear polymers (poly-para-phenylene, PPP), whereas 1,3-

diiodobenzene results in a zig-zag chain (poly-meta-phenylene, PMP) as well as

macrocycle oligomers.*

Figure 3-1: Schematic dependence of relative substitution pattern of precursors and their products. X

represents any possible reactive group in the precursor.

In a manner conceptually similar to 1D, the 2D structure can be predefined using

an appropriate monomer. For example, tri-, tetra-, and hexa-functional monomers

ideally result in structures with hexagonal, square and triangle-shape pore,

respectively. However, the product’s topology is not only dependent in the structure

of monomer and the conditions of experiment results to non-optimal bonding and

consequently various polygonal products. For example, a variety of products with the

tetragonal, pentagonal, heptagonal or octagonal pores can formed from a three-fold

symmetric monomer2 owing to different deposition rates, surface temperatures, etc.

In addition, the monomer size can affect the structure quality and morphology

of the products3 as well. For example, Lackinger et al. compare the resulting products

grown from 1,3,5-triiodobenzene (TIB) and 1,3,5-tris(4′-iodophenyl) benzene (TIPB)

precursors on Au(111). TIB and TIPB are three-fold symmetric iodinated molecules

which differ in size. The schematic molecular structure and corresponding STM

images for TIPB and TIB are shown in Figure 3-2. The small precursor resulted in

irregular networks whereas the larger precursor produced only patches of immobile

XX

XX


oligomers. Post-annealing of the TIPB led to an oligomeric covalently-linked network.

The main distinction between using two different sizes of precursor is the ratio of split-

off halogen (by-products) to the resulting organic network due to abundance of

halogen atoms abstracted from the small molecules compare to large ones. Therefore,

the ratio of surface covered by iodine to the surface covered with organic network for

TIB is 1.4 while it is 0.2 for TIPB. This ratio could be a determining factor in the

quality of the product and polymerization reaction, in particular in Ullmann reaction

in which by-product passivate active sites on the surface and might interfere with the

diffusion and polymerization of products.

Figure 3-2: STM images of (a) TIPB in which covalent-bonded oligomers, including dimers and

trimers mostly near step-edges and (b) TIB which result in covalent networks and iodine by-products

on Au(111). Adapted with permission from [3].

Finally, the rigidity of molecular structural also might enhance the uniformity of

the product.4 For example, the 3,5-tris(4-bromophenyl) benzene (TBPB) monomer on

Au(111) led to various polygonal networks (4 to 8 edges). The diversity of the products

can be attributed to the intrinsic flexibility of backbone of the monomer.5 In contrast,

a rigid monomer such as hexaiodo-substituted cyclohexa-m-phenylene (CHP) result in

a more uniform hexagonal network, which was attributed to the stiff and interlinked

skeleton of CHP.6

3.1.2 Substrate

In general, the ordering of a molecular layer on a metal surface is influenced by

both molecule-molecule and molecule-surface interactions.7 Therefore, the substrate

can play a vital role in the resultant structure. Substrates can have two significant

influences on surface-confined reaction, i.e. the substrate can act as a catalyst for the

reaction in addition to acting as a template for molecules during the reaction process.


This section is devoted to the comparison of different metal surfaces and their effect

on the organic product.

The most frequently used surfaces are coinage (noble) metals with a face

centered cubic (fcc) structure including copper, silver and gold. The reactivity of noble

metal surfaces is closely correlated to their electronic structures. The electronic

structure of noble metals depends on coexistent of s and d states in the valence band.

In contrast to the electrons in s state which are localized, the electrons in the d band

are spatially localized and lead to complex hybridization at the surface. The electronic

properties of the surface such as work function, surface energy and reactivity can be

explained by the hybridization of d band.8-9 Work function (WF) is the required energy

to transfer an electron from the Fermi level to vacuum, and is one of the valuable

indicator of electronic properties of surfaces and thus the reactivity of surface.10 WF

is determined by the both chemical composition and the crystallographic orientations

of the surface.10

Substrate Chemistry

The physical definition of noble metals is strict, only including metals that have

a filled d band. Using this definition only copper, silver and gold are noble metals in

which the highest occupied energy level is the d band. Some other precious metals (Pt,

Rh, Pd) are catalytically important and have also been employed in surface-confined

reaction. For example, platinum has a 5d band -in addition to a 4d band- which crosses

the Fermi level and therefore imparts a higher reactivity compared to noble metals.

Hammer et al. 8 suggested that the reactivity of a metal surface is strongly depends on

the entire valance band and not on the Fermi level or empty d band individually. The

Hammer–Nørskov d-band model nicely explains the correlation of chemisorption

bond strengths to the d-band density of state (DOS).11 For noble metals, the valence d

band is several eV below the Fermi level. This means that the anti-bonding states of

an adsorbate are below the Fermi level and therefore filled, i.e. both bonding and anti-

bonding states of surface -molecule are occupied. This interaction generates only a

repulsive energy contribution which increases with orbital size. This explains the order

of strength of adsorption bonds of molecule with noble metal: Cu˃ Ag ˃ Au.

Moreover, if adsorbates strongly interact with the substrate, they anchor where they

hit the surface and cannot diffuse easily on the surface and leave a disordered film.12


In a study by Lackinger’s group the dehalogenation reaction of 1,3,5-tris 4-

bromophenyl benzene (TBB) was investigated for different substrates; Cu(111),

Ag(111), and Ag(110).13 The reaction proceeded on Cu(111) and Ag(110) but not on

Ag(111). This reveals the interesting reactivity differences for various substrate which

is dependent on both material (Cu(111) versus Ag(111)) and also the surface

orientation (Ag(110) versus Ag(111)). The influence of molecule- substrate interaction

result from different metal surfaces (Cu(111) versus Au(111)) for dehalogenation and

cyclodehydrogenation of 10,10ʹ-dibromo-9,9ʹ bianthracene (DBBA) has been reported

as well.14 While DBBA adsorbed intact on Au, it fully dehalogenated on Cu. Similarly,

the cyclodehydrogenation reaction which result in formation of GNRs occurred at two

different temperature for theses surface: 400 and 200℃ on Au and Cu, respectively.

The variety in substrate chemistry also can be exploited to trigger a selective

reaction to precisely synthesise selective products. For example, as illustrated on while

C-C coupling dehalogenated (R)-(−)-6,6′-dibromo-1,1′-bi-2-naphthol (DN) molecule

are predominant on Au(111), new organic molecules formed on Ag(111) owing to the

selective C-H coupling of dehalogenated DN.15

Figure 3-3: Scheme of two different reaction pathways of DN on Ag(111) and Au(111) illustrating

C−C or C−H coupling of DN on different surfaces. Reprinted with permission from [15] Copyright

(2017) American Chemical Society.

Substrate Structure

Different surface facets can be exposed at the surface by orienting and cutting

fcc metal single crystals. Atoms on the obtained surface have fewer neighbouring

atoms (known as coordination number, CN) than those in the bulk crystal. The atoms


on the surface can also rearrange so that the energy of the atoms is reduced. This

usually occurs through vertical or lateral relaxation of surface atoms.

The metal (111), (100) and (110) facets for FCC are the most commonly studied

in surface science. The (111) facet is the closest-packed layer of atoms and has a three-

fold symmetry (or six-fold symmetry considering mirror symmetry). All atoms in the

111 facet of fcc crystals are equivalent and formed a surface that is atomically smooth

(see Figure 3-4) while this is not the case for (100) and (110) facets. For example, the

surface layer atoms in (110) are in a rectangular symmetry so that atoms are close-

pack along the row and have substantial spacing of √2 a0 along the orthogonal

direction. This result in surface atoms that are more open and thus have a lower

coordination number, i.e. the second layer of atoms are exposed to the surface. The top

views of the (111), (100) and (110) surfaces are shown in Figure 3-4.

Figure 3-4: Top view of atoms in the first layer of different planes.

The difference in the orientation of metal surfaces can dictate significant change

in the adsorption topology and geometry of molecules and the corresponding reactions.

Table 3-1 compares the unit cell and possible adsorption site for (111), (100) and (110)

surfaces of noble metals. The coordination number of surfaces determines their

chemical reactivity of surfaces10 in which the surface reactivity decreases with

increasing of surface coordination number16: (100), (110) and (111) respectively.

Table 3-1 Comparisons of structural symmetries of Cu, Ag and Au, reproduced from.10

Surface Area unit cell of 2D /Å2 Possible adsorption sites

(111) √3/4 a02 on-top sites, bridging sites and hollow sites

(100) 1/2 a02 on-top sites, bridging sites, hollow sites

(110) √2/2 a02 on-top sites, short (in a single row) and long bridging sites

(in adjacent rows)


To gain a detailed understanding of substrate template effects and how reactions

can be steered by the substrate lattice, different groups have provided comparison

between different facets of surface for single molecular precursors. For example, a

study revealed that while 10,10ʹ-dibromo-9,9ʹ bianthracene (DBBA) transformed to

GNRs on Cu(111), only 0D nanographene structures were formed on Cu(110) (Figure

3-5).17 This has been attributed to the strong molecular-substrate interaction surpassed

the intermolecular interaction on Cu(110) which limits the diffusion of adsorbate on

the surface and prohibit the lateral interaction between them.

Figure 3-5: Comparison of reactivity effect on polymerization of DBBA between Cu(111) and

Cu(110). Reprinted with permission from [17].

A conceptually similar study showed that 4,4ʺ-dibromo-meta-terphenyl (DMTP)

forms zigzag oligophenylene chains on Cu(110)18 while the product of polymerisation

of the same molecule on Cu(111) is hydrocarbon macrocycles with 18 phenyl rings19

due to substrate effect on the products.

It should be noted that intrinsic adatoms and defects on surfaces can play

important roles as “active site” and are favourable for chemical reactions. In addition,

stepped surfaces can be used as a template because of diffusion barriers.20 They are

also more reactive because of their high density of undercoordinated atoms.20

3.2 SURFACE SENSITIVE ANALYSIS TECHNIQUES

As previously has been implied, surface science is a multidisciplinary field and

employs different techniques to study the physical and structural properties of the

monolayer/multilayer of the material on a solid surface. Surface sensitive techniques

enable us to study the surface-confined reaction pathway and electronic and structural


properties of the resultant polymers. This section briefly describes the theory of XPS,

STM and NEXAFS as three important probes in the analytical tool box used in this

thesis.

3.2.1 X-ray Photoelectron Spectroscopy (XPS)

The photoelectric effect describes the emission of the photoelectrons from a

metal surface induced by photons. The discovery of the photoelectric effect (Nobel

Prize–Einstein 1921) underpins XPS. The Nobel Prize in Physics 1981 was awarded

to Kai Siegbahn21 because of his contribution to the development of electron

spectroscopy.

Electrons bound to atoms (in the core level) are excited by using an X-ray

photon, generating free electrons, commonly termed as photoelectrons. By using a

hemispherical analyser, the ejected photoelectron is filtered based on its energy, and

is recorded by a detector. Identification of the elements in the sample can be made

directly by determining the binding energy based on measurement of the kinetic

energies of these ejected photoelectrons according to the following equation:

Ebinding = EPhoton – (EKinetic+ φ), (1)

where Ephoton is the energy of the X-ray photons being used (often Al Kα X-rays with

Ephoton = 1486.7 eV or Mg Kα X-rays with Ephoton = 1253.6 eV), Ekinetic is the kinetic

energy of the electron as measured by the instrument, and φ is the work function which

is dependent on the spectrometer. Synchrotron-based radiation sources provide a

tuneable energy for Ephoton. The binding energy of the electron is an intrinsic property

of the element from which it originated,22 and identifies elements on the sample. The

relative concentrations of elements can also be determined from the photoelectron

intensities.

Different materials can be distinguished based on the peaks visible in the XPS

spectrum. As an example, an XP survey spectrum for trimesic acid (TMA) on a copper

surface shown in Figure 3-6. The spectrum was collected with an Al Kα x-ray source.

The spectrum identified three elements: copper, oxygen and carbon. The copper peak

originates from the substrate, while oxygen and carbon peaks result from the TMA

molecule deposited on the surface. The Cu 2p peaks demonstrates the spin-orbit

splitting effect, i.e. the energy level splits into two different energies induced by two

possible spin direction of electron with orbital quantum number bigger than zero (l >


0). The carbon and oxygen also can be deconvolved to reveal different chemical states

in the molecule. The inset shows the fitted carbon fit as well as the schematic structure

of the TMA.

Figure 3-6: XP spectrum for as-deposited TMA on Cu(111) surface. In addition to the copper peaks, C

1s and O 1s peaks originating from TMA are discernible. The insets show the Cu 2p and C 1s

highlighted region of the spectrum in more detail.

Resolution of characteristic core level shifts due to changes in the local chemical

environment is one of the strong features of the XPS technique.23

3.2.2 Scanning Tunnelling Microscopy (STM)

Gerd Binnig and Heinrich Rohrer won the Physics Nobel Prize in 1986 for the

invention of the scanning tunnelling microscope.24 The invention of STM, which

visualizes molecules in real space, has greatly aided nanotechnology research, and has

become a very powerful atomic-scale tool in surface science. This probe has been

applied in a diverse range of environments; for instance, in UHV, at the liquid–solid

interface and under ambient conditions.

The STM concept is based on quantum tunnelling whereby a particle can tunnels

through a barrier larger than total energy of particle. This can only be explained by

quantum mechanics owing to the particle-wave nature of electrons.

In this technique, the tip is moved close to a flat surface with an accuracy of

better than 0.1 nm using a piezoelectric positioner. The piezoelectric scanner provides


a precise control in any of the three perpendicular axes (x,y,z) of sample. Schematic

view of the scanning tunnelling microscope is shown in Figure 3-7.

Figure 3-7: left: Sketch of tunnelling current through sample to tip. Efs and Eft are the Fermi level of

the sample and tip, respectively. Right: Schematic view of the scanning tunnelling microscope.25

When the tip is brought very close to the surface of the sample, electrons will

tunnel between tip and surface, i.e. a finite tunnelling conductance. If a voltage bias

(V) is applied between tip and the sample, a tunnelling current is generated. The

solution of Schrӧdinger's equation for one-dimensional rectangular barrier reveals that

the wave function 𝜓 decays exponentially within the barrier distance (d).26 The current

of these tunnelling current (I) at the bias voltage is a function of the transmission

probability of electrons:

𝐼 =4𝜋𝑒

ћ∫ 𝜌𝑠 (𝐸𝐹 − 𝑒𝑉 + 𝜀)𝜌𝑇

𝑒𝑉

0(𝐸𝐹 + 𝜀)|𝑀|2 𝑑𝜀, (3)

where f is the Fermi function the ρs and ρT are the density of states (DOS) of

sample and tip, respectively. M is the tunnelling matrix element in the Bardeen

tunnelling theory and given by27

𝑀 = −ℎ2

2𝑚∫ (𝜒∗∇𝜓 − 𝜓∇𝜒∗). 𝑑𝑆

∑

,

where ψ and χ are the wavefunction of the sample and tip, respectively.

Accordingly, the tunnelling current is exponentially dependent on distance between

the tip and the surface (d) and therefore, a small change in d dramatically change the

tunnelling current. Thus, STM is very sensitive to the tip-sample spacing.

In order to map the surface with atomic resolution, the tunnelling current is

precisely monitored as a function of lateral position. The obtained STM images contain


information about both topography and LDOS, which together define the contrast in

the images. Vibration isolation is essential to achieve atomic resolution.

There are two possible modes for scanning an STM image: constant height mode

and constant current mode. The spacing between tip and sample is unchanged for the

constant height mode and the therefore the tunnelling current changes during imaging.

The resultant image from this mode is a function of the variations of the current signal.

In the current constant mode, which is more common, the tunnelling current is held

constant using a feedback loop. In this mode the tip-sample spacing d is controlled by

a feedback system to maintain the tunnelling current constant. The voltage on the

piezoelectric in z direction represents the local height of the topography. For example,

molecules on a flat single crystal, corrugation of the sample, plotted on STM image

due to variation of d over scanning.

STM is a challenging technique and to get high resolution images (especially on

the atomic scale), an atomically clean surface and stable and sharp tip and excellent

vibration isolation are necessary. STM is not able to directly characterize the charge

state or local chemical environment of neither the molecular network components nor

the surface.

3.2.3 Near Edge X-Ray Absorption Fine Structure (NEXAFS)

NEXAFS is a synchrotron-based spectroscopic technique which was first

established in the early 1980s. This technique is sensitive to the chemistry and bonding

in materials and can reveal orientation of adsorbed molecules.28 NEXAFS cannot be

performed in the home laboratory owing to two important required properties of the

X-ray source which are only provided in synchrotron facilities: the X-ray light must

be tunable and polarized.29

In this technique an electromagnetic beam (X-ray beam in which its energy

varies around/less an ionization edge) is shone on the sample. Absorption of photons

promotes core level electrons (for instance, the K shell) into unoccupied states. This

generates a core hole and leaves the molecule in an excited state.

The created hole can relax in two different ways: (a) an electron from higher

level goes to the core level, this is accompanied with a fluorescence emitted (see Figure

3-8b). Alternatively, relaxation process will occur via emission an Auger electron,

shown in Figure 3-8c. Therefore, the existence of the core hole and therefore the


absorption cross section ()*, can be detected either by measuring the florescence

photon (by a fluorescence detector) or by employing a channeltron to measure the

Auger electrons. The electron detection offers higher surface sensitivity (as well as

higher intensity in the case of low-Z molecule, where Z is the number of the proton)

than fluorescence detection.

Figure 3-8: Energy diagram representing electron transition in NEXAFS.

In general, NEXAFS spectra reveals the unoccupied orbitals that electrons can

occupy over excitation. The transition of the core electron to unfilled molecular orbital

for a diatomic molecule is shown in Figure 3-9.

Figure 3-9: scheme of X-ray absorption in a diatomic molecule (bottom), and the corresponding k-

shell absorption spectrum (top). Reprinted with permission from [29].


Combining atomic orbitals from each atom of the molecule leads to the

formation of molecular orbitals. For example, pairing of two s orbitals or px orbitals

constructs and * orbitals, and interactions of two equivalent pys (or pzs) leads to π

and π* molecular orbitals. Considering that molecular orbitals have a specific

directionality and the incident beam is polarized, this can be used to extract

information on the orientation of the orbitals with respect of the substrate. An electric

field of the light parallel to the direction of a molecular orbital has a maximum of an

absorption transition probability. While an electric field of light is perpendicular to

direction of a molecular orbital, the probability of adsorption transition will be

minimized. This characteristic can be used to determine the orientation of molecules

with respects to surface, see Figure 3-10.

Figure 3-10: Schematic representation of photon incident with normal angle (left) and grazing angle

(right). The diatomic molecule adsorbed parallel to the surface. Reprinted with permission from [29].


3.3 REFERENCES



(5), 592-597.

2. Blunt, M. O.; Russell, J. C.; Champness, N. R.; Beton, P. H. Templating

molecular adsorption using a covalent organic framework. Chemical Communications

2010, 46 (38), 7157-7159.

3. Schlögl, S.; Heckl, W. M.; Lackinger, M. On-surface radical addition of triply

iodinated monomers on Au (111)—the influence of monomer size and thermal post-

processing. Surface Science 2012, 606 (13), 999-1004.

4. Zhang, Y.-Q.; Kepcija, N.; Kleinschrodt, M.; Diller, K.; Fischer, S.;

Papageorgiou, A. C.; Allegretti, F.; Björk, J.; Klyatskaya, S.; Klappenberger, F.;

Ruben, M.; Barth, J. V. Homo-coupling of terminal alkynes on a noble metal surface.

Nature Communications 2012, 3, 1286.

5. Faury, T.; Clair, S.; Abel, M.; Dumur, F. d. r.; Gigmes, D.; Porte, L. Sequential

linking to control growth of a surface covalent organic framework. The Journal of






7. Rosei, F.; Schunack, M.; Naitoh, Y.; Jiang, P.; Gourdon, A.; Laegsgaard, E.;

Stensgaard, I.; Joachim, C.; Besenbacher, F. Properties of large organic molecules on

metal surfaces. Progress in Surface Science 2003, 71 (5–8), 95-146.

8. Hammer, B.; Nørskov, J. Electronic factors determining the reactivity of metal

surfaces. Surface Science 1995, 343 (3), 211-220.

9. Fall, C.; Binggeli, N.; Baldereschi, A. Work-function anisotropy in noble

metals: Contributions from d states and effects of the surface atomic structure.

Physical Review B 2000, 61 (12), 8489.

10. Pedio, M.; Cepek, C.; Felici, R., Organic molecules on noble metal surfaces:

the role of the interface. In Noble Metals, InTech: 2012.

11. Hammer, B. Special sites at noble and late transition metal catalysts. Topics in

catalysis 2006, 37 (1), 3-16.

12. Tautz, F. Structure and bonding of large aromatic molecules on noble metal

surfaces: The example of PTCDA. Progress in Surface Science 2007, 82 (9-12), 479-

520.

13. Walch, H.; Gutzler, R.; Sirtl, T.; Eder, G.; Lackinger, M. Material- and

Orientation-Dependent Reactivity for Heterogeneously Catalyzed Carbon−Bromine

Bond Homolysis. The Journal of Physical Chemistry C 2010, 114 (29), 12604-12609.

14. Simonov, K. A.; Vinogradov, N. A.; Vinogradov, A. S.; Generalov, A. V.;

Zagrebina, E. M.; Mårtensson, N.; Cafolla, A. A.; Carpy, T.; Cunniffe, J. P.;

Preobrajenski, A. B. Effect of Substrate Chemistry on the Bottom-Up Fabrication of

Graphene Nanoribbons: Combined Core-Level Spectroscopy and STM Study. The


15. Kong, H.; Yang, S.; Gao, H.; Timmer, A.; Hill, J. P.; Díaz Arado, O.; Mönig,

H.; Huang, X.; Tang, Q.; Ji, Q.; Liu, W.; Fuchs, H. Substrate-Mediated C–C and C–H

Coupling after Dehalogenation. Journal of the American Chemical Society 2017, 139

(10), 3669-3675.


16. Peljhan, S.; Kokalj, A. DFT study of gas-phase adsorption of benzotriazole on

Cu (111), Cu (100), Cu (110), and low coordinated defects thereon. Physical

Chemistry Chemical Physics 2011, 13 (45), 20408-20417.


Zagrebina, E. M.; Svirskiy, G. I.; Cafolla, A. A.; Carpy, T.; Cunniffe, J. P.; Taketsugu,

T. From graphene nanoribbons on Cu (111) to nanographene on Cu (110): critical role

of substrate structure in the bottom-up fabrication strategy. ACS Nano 2015, 9 (9),

8997-9011.

18. Dai, J.; Fan, Q.; Wang, T.; Kuttner, J.; Hilt, G.; Gottfried, J. M.; Zhu, J. The

role of the substrate structure in the on-surface synthesis of organometallic and

covalent oligophenylene chains. Physical Chemistry Chemical Physics 2016, 18 (30),

20627-20634.

19. Fan, Q.; Wang, C.; Han, Y.; Zhu, J.; Hieringer, W.; Kuttner, J.; Hilt, G.;

Gottfried, J. M. Surface‐Assisted Organic Synthesis of Hyperbenzene Nanotroughs.


20. Saywell, A.; Schwarz, J.; Hecht, S.; Grill, L. Polymerization on stepped

surfaces: alignment of polymers and identification of catalytic sites. Angewandte


21. Siegbahn, K.; Nordling, C.; Fahlman, A.; Nordberg, R.; Hamrin, K.; Hedman,

F.; Johansson, G.; Bergmark, T.; Karlsson, S.; Lindgren, I. Nova Acta Regiae Soc. Sci.

Ups., Ser. IV. 1967.

22. Lee, S. M., Reference book for composites technology. CRC Press: 1989; Vol.

1.

23. Baraldi, A.; Comelli, G.; Lizzit, S.; Kiskinova, M.; Paolucci, G. Real-time X-

ray photoelectron spectroscopy of surface reactions. Surface Science Reports 2003, 49

(6), 169-224.

24. Elemans, J. A.; Lei, S.; De Feyter, S. Molecular and Supramolecular Networks

on Surfaces: From Two‐Dimensional Crystal Engineering to Reactivity. Angewandte


25. Schmid, M. Schematic diagram of a scanning tunneling microscope.

https://commons.wikimedia.org/wiki/File:ScanningTunnelingMicroscope_schematic.

png.

26. Meyer, E.; Hug, H. J.; Bennewitz, R., Scanning probe microscopy: the lab on

a tip. Springer Science & Business Media: 2013.

27. Chen, C. J., Introduction to scanning tunneling microscopy. Oxford University

Press New York: 1993; Vol. 2.

28. Stöhr, J., NEXAFS spectroscopy. Springer Science & Business Media: 2013;

Vol. 25.

29. Hähner, G. Near edge X-ray absorption fine structure spectroscopy as a tool to

probe electronic and structural properties of thin organic films and liquids. Chemical

Society Reviews 2006, 35 (12), 1244-1255.

https://commons.wikimedia.org/wiki/File:ScanningTunnelingMicroscope_schematic.png

https://commons.wikimedia.org/wiki/File:ScanningTunnelingMicroscope_schematic.png

Chapter 4: Adsorption and Reactivity of Pyridine Dicarboxylic Acid on Cu(111) 55

Chapter 4: Adsorption and Reactivity of

Pyridine Dicarboxylic Acid on

Cu(111)

As previously discussed, decarboxylation reaction is a promising approach to

fabricate byproducts-free polymer. At the same time, precursors with heteroatoms in

their scaffold offer an elegant way to engineer the band gap of the products. In this

regard, chapter 4 provides results from the study regarding the influence of nitrogen in

the aromatic core of a molecule undergoing a decarboxylation reaction. The findings

reported in this chapter could be interpreted from two viewpoints: (a) investigating a

nitrogen-containing molecule as an important class of precursor in surface-confined

reactions and (b) studying the decarboxylation reaction and pathway of a nitrogen-

containing molecule for the first time. The primary experiment has been done on

Omicron (Scienta-Omicron GmbH) at QUT and PES and NEXAFS have been done at

the Australian Synchrotron.

The results of this study has been published in The Journal of Physical Chemistry

C: Abyazisani, M.; Bradford, J.; Motta, N.; Lipton-Duffin, J.; MacLeod, J. Adsorption

and Reactivity of Pyridine Dicarboxylic Acid on Cu (111). The Journal of Physical

Chemistry C 2018, 122 (31), 17836-17845.

56 Chapter 4: Adsorption and Reactivity of Pyridine Dicarboxylic Acid on Cu(111)

Statement of Contribution of the Co-Authors for Thesis by

Published paper

The authors listed below have certified that:

• They meet the criteria for authorship in that they have participated in the

conception, execution, or interpretation, of at least that part of the publication in their

field of expertise;

• They take public responsibility for their part of the publication, except for the

responsible author who accepts overall responsibility for the publication;

• There are no other authors of the publication according to these criteria;

• Potential conflicts of interest have been disclosed to (a) granting bodies, (b) the

editor or publisher of journals or other publications, and (c) the head of the responsible

academic unit, and

• They agree to the use of the publication in the student’s thesis and its publication

on the QUT’s ePrints site consistent with any limitations set by publisher requirements.

In the case of this chapter:

Adsorption and reactivity of pyridine dicarboxylic acid on Cu(111):

J. Phys. Chem. C, 2018,122(31), pp 17836–17845, DOI: 10.1021/acs.jpcc.8b04858.

Contributor Statement of contribution

Maryam Abyazisani Conducted lab-based experiments, conducted synchrotron

experiments, analysed data, drafted manuscript. Signature

Jonathan Bradford Conducted synchrotron experiments, revised manuscript

Nunzio Motta Revised manuscript, supervision.

Josh Lipton-Duffin Revised manuscript, supervision.

Jennifer MacLeod Conceived the project, conducted synchrotron experiments,

analysed data, revised manuscript, supervision.

Date Signature Name

I have sighted email or other correspondence from all co-authors confirming their

certifying authorship.

Principal Supervisor Confirmation

Jennifer MacLeod 6/12/2018



http://www.mopp.qut.edu.au/F/F_01_03.jsp

https://pubs.acs.org/doi/abs/10.1021/acs.jpcc.8b04858


4.1 ABSTRACT

A detailed understanding of the reaction of a range of molecules on surfaces will

be key to developing targeted strategies for on-surface synthesis. Here, we have the

deprotonation and decarboxylation reactions of 3,5-pyridine dicarboxylic acid (PDC)

on Cu(111) using synchrotron radiation photoelectron spectroscopy (SRPES), near-

edge X-ray absorption fine structure (NEXAFS) and density functional theory (DFT).

PDC partially deprotonates upon deposition on Cu(111) at room temperature and

adsorbs with the plane of its aromatic ring inclined at an average of ∼45° with respect

to the surface. By heating to 100°C, the deprotonation of the molecule increases. When

the PDC is partially deprotonated, the plane of its aromatic ring adopts a more upright

orientation with respect to the surface. Additional heating to 160°C causes complete

deprotonation, upon which the molecule returns to a more planar molecular adsorption

geometry. By examining both the N 1s core level and the NEXAFS results, we ascribe

these changes in adsorption to a reaction-induced change in the predominant

molecule−surface interaction, which is driven by the nitrogen lone pair prior to

deprotonation, and by the carboxylate groups after deprotonation of the −COOH

groups. These interaction channels and adsorption geometries are supported by our

DFT calculations. Heating above 200°C induces decarboxylation of the molecule; by

observing the rate of reaction over a range of fixed temperatures, we extract an

activation energy of 1.93 ± 0.17 eV for the decarboxylation reaction. Around this

temperature we also begin to observe the ring opening of the molecule, suggesting that

if PDC is to be used as a building block for on-surface synthesis of polymers, careful

control of temperature is necessary for obtaining decarboxylation and covalent

coupling of the molecule without destroying the aromatic core.


4.2 INTRODUCTION

On-surface covalent polymerization has grown into an active research domain1-

4 owing to promising applications in nanotechnology and molecular electronics.5-6 A

variety of surface-confined reactions performed under ultrahigh vacuum (UHV)

conditions have been translated from established solution-chemistry methodologies.

In particular, the Ullmann reaction for C–C coupling of aryls7 has been thoroughly

studied, and many of the details of the process are understood.8 However, this reaction

inevitably produces halogen by-products which tend to remain chemisorbed on the

surface, resulting in spatial restriction of the polymer as well as in catalyst poisoning

through a reduction of active sites on the surface. To address this problem, alternative

reactions that do not produce any byproducts or that have volatile byproducts are

actively being pursued. Typical examples include the Glaser reaction,9-10 Bergman

reaction,11 and dehydrogenation.12 More recently, on-surface decarboxylation has been

explored as a “clean” reaction.13

Decarboxylative coupling is appealing since the byproducts, H2 and CO2, are

unlikely to remain adsorbed on the surface under reaction conditions; CO2 is volatile,

and molecular hydrogen remains adsorbed on Cu(111) only until 330 K.14 On-surface

covalent coupling via decarboxylation was demonstrated for the first time when poly-

2,6-naphthalene was formed using 2,6-naphthalenedicarboxylic acid as a precursor.13

Scanning tunnelling microscopy (STM) confirmed that molecules first undergo

deprotonation upon annealing, followed by decarboxylation with further increase of

the reaction temperature, and subsequent coupling to form polynaphthalenes. More

recently, decarboxylative coupling of 1,3,5-benzenetribenzoic acid (BTB) was used to

fabricate a 2D network.15

To date, studies of decarboxylative coupling have focused only on benzene-

based precursors. However, incorporation of heteroatoms in the aromatic ring is a

powerful approach for tuning the physical properties of polymers in a controlled

way.16-19 For example, nitrogen substitutions have been used to modify the electronic

properties of nano-graphenes.20-21 The presence of nitrogen in the precursor may also

have implications for the adsorption geometry of the monomer.22-23 Interaction

between the substrate and the nitrogen lone pair can lead to adsorption with the

molecular plane tilted or vertical with respect to the surface in small molecules.24-27

On the other hand, the π orbitals also present an interaction channel between the


molecule and the surface and can result in flat adsorption of pyridine under some

circumstances, e.g. electrochemical conditions.28 The addition of carboxylic groups to

small molecule precursors add some complication to this scenario, since the copper–

carboxylate bond strength is comparable to copper–copper bonding of the surface,29

and therefore, the carboxylate group can also influence adsorption geometries.30

Moreover, additional effects on adsorption are known to relate to molecular

coverage,31 number of COOH groups,32 and/or availability of metal adatoms.29

This is a matter of some importance for on-surface coupling, since the adsorption

geometry of the monomer has been shown to critically affect molecular reactions on

surfaces.33 A surface-catalysed reaction requires proximity of the molecular reaction

site to the metal surface, and consequently, the presence of nitrogen in an aromatic

ring can result in an adsorption-based reaction barrier. For example, dibromobenzene

fully dehalogenates upon room temperature adsorption on Cu(110),34-35 whereas only

one of the two C−Br bonds of dibromopyridine is broken at room temperature on

Cu(100); this difference is attributed to the adsorption geometry induced by the

presence of the nitrogen in the aromatic ring, which orients the nitrogen into the

surface, leaving the unreacted C−Br oriented out of the surface.36

In this work, we aim to contribute to the understanding of on-surface

decarboxylation by studying the reaction of 3,5-pyridinedicarboxylic acid (PDC) on

Cu(111). Studying this molecule gives us the opportunity to understand how

decarboxylation progresses in a nitrogen-containing molecule. It also enables us to

understand the interplay between the different parts of the molecule as well as its

interaction with surface. The changes in chemistry and adsorption geometry with

annealing were investigated through synchrotron radiation photoelectron spectroscopy

(SRPES), near-edge X-ray absorption fine structure (NEXAFS) spectroscopy, and

density functional theory (DFT), and the kinetics of decarboxylation were studied by

measuring the decarboxylation rate at different temperatures.

4.3 EXPERIMENTAL METHODS

All experiments were carried out under UHV conditions (base pressures of 2

×10-10 mbar) at the Soft X-ray beamline at the Australian Synchrotron. Cu(111) and

Ag(111) single crystals were cleaned by repeated cycles of Ar+ sputtering (E = 1 keV)

and annealing (~400°C) until no impurities were detected by SRPES. 3,5-


Pyridinedicarboxylic acid (PDC) (98%, Alfa Aesar, shown schematically in Figure

4-1a) was thoroughly degassed at 120°C in UHV. Results from Ag(111) are not

reported in the main manuscript but selected results are included in the SI. Briefly,

PDC deposited onto Ag(111) held at −130°C adsorbed intact, then partially

deprotonated and finally desorbed from the surface with heating to 100°C (Figure S1).

We additionally performed a beam damage test of PDC on Ag(111) and found that a

486 eV beam caused deprotonation of the PDC (Figure S2). Although the same test

was not performed on Cu(111), we assume a similar effect is present, and hence that

the level of deprotonation of any sample may be dependent on the exposure to the

beam. To minimize this effect, we moved the sample frequently so that we were not

exposing any one region for more than one spectrum. Every spectrum took 30 s to

acquire. The PDC was deposited onto the room-temperature Cu(111) substrate from a

Knudsen cell (Eberl-MBE Komponenten) held at 135°C. The sample was annealed at

different temperatures in order to trigger the deprotonation and decarboxylation

reactions (shown in Figure 4-1b). The presented SRPES and NEXAFS data were

acquired in separate experiments. For SRPES experiments, the sample was annealed

at 100, 160, 240 and 270°C, with each step performed for 10 min. For NEXAFS

experiments, the sample was annealed at 100, 160 and 280°C, with each step

performed for 5 min. SRPES spectra were collected with beam energies of 486 eV for

C 1s and N 1s regions and 908 eV for the O 1s region. The pass energy of the analyser

was set to 10 eV for an overall energy resolution of 0.29 eV. For quantitation, survey

spectra were recorded using a beam energy of 1487 eV with the pass energy set to 20

eV. All PES spectra were analysed using CasaXPS software.37 Spectra were calibrated

with respect to the Fermi level and were fitted using a Shirley background. For time-

resolved experiments, the sample was kept at fixed temperatures (220, 230 and 240°C,

all ±1°C) and C 1s spectra were sequentially collected using a beam energy of 486 eV

over a period of 30 s per spectrum. Lineshape fitting of the C 1s region for each of the

spectra permits a determination of the proportion of reacted molecules and was used to

find the activation energy of the decarboxylation reaction. An estimate for the uncertainty

of the fitted species was obtained by applying a Monte Carlo method to the fitted

intensities. NEXAFS spectra were collected using linearly polarized light at three

different incidence angles with respect to the surface plane: glancing incidence at =

20°, the so-called “magic” (tilt-independent) angle at = 55°, and normal incidence at


= 90°. All spectra were analyzed using the QANT software package.38 NEXAFS

spectra were double normalized against the incoming photon flux determined by

measuring the current, I0, collected from a gold mesh in the path of the beam and

against spectra collected over the same energy range and at the same angle from a

clean Cu(111) substrate.

DFT calculations were performed using the projector-augmented wave (PAW)

pseudopotentials in Quantum ESPRESSO v6.2.139 under the generalized gradient

approximation (GGA) with Perdue‒Burke‒Ernzerhof (PBE) parameterization for the

exchange-correlation functional.40 The exchange correlation was modified by adding

an ab initio nonlocal van der Waals correlation contribution (vdW-DF), which is

known to improve the description of the long-range dispersive forces of the standard

GGA.41-43 After relaxing the bulk Cu face-centered cubic (FCC) cell, the Cu(111)

surface was modeled using an eight-layer slab, holding the bottom three layers fixed

and allowing the top five layers to relax using a 12121 k-point grid. The top three

layers were then used to model the absorption of PDC in a 55 supercell at high-

symmetry sites in a variety of starting geometries, allowing the top two copper layers

to relax along with the molecule while holding the bottom layer fixed in the geometry

from the bare metal slab relaxation. A vacuum gap of at least 10 Å (more for the more

planar starting geometries) between the slabs was used to prevent communication of

the molecules in the top layer with the Cu atoms of the bottom layer. The k-point

sampling for these geometry optimizations was restricted to the gamma point.

Figure 4-1: (a) molecular structure of 3,5-pyridinecarboxylic acid (PDC). (b) Decarboxylation

reaction schematic for PDC molecule.


4.4 RESULTS

4.4.1 SRPES Study of the Annealing of 3,5-Pyridinedicarboxylic Acid/Cu(111)

Figure 4-2a-c shows the photoemission spectra of the C 1s, O 1s and N 1s core

levels of PDC on Cu(111) deposited at RT, annealed to 100°C, and annealed to 270°C.

The C 1s spectra for the as-deposited sample exhibit two well-separated peaks centered

at 285.2 and 289.3 eV which we assign to the aromatic ring and carboxylic group

(−COOH) in the PDC molecule, respectively. The binding energy (BE) splitting

between the carboxyl and the aromatic peaks, 4.1 eV, is in agreement with the

previously reported values for carboxylated phenyls.30, 44-49 The peak assigned to the

aromatic ring contains two contributions due to the two inequivalent types of carbon

atoms in the ring: two carbons adjacent to nitrogen, C−N (285.8 eV), and three carbons

that bond only to other carbons, C−C (285.4 eV).50 The ratio of these components is

fixed at 2:3 based on the stoichiometric ratio between C−C and C−N in the pyridine.

This is in contrast to previous work, where the five carbon atoms in the pyridine ring

have been treated as indistinguishable.36 Three additional components contribute to

the spectrum: the carboxylate group (−COO−) at a BE of 288.1 eV, C−Cu at a BE of

284.1 eV, and finally the π−π shakeup transition of the aromatic system at 291.7 eV.44

The carboxylate signature implies that PDC has already partially deprotonated upon

adsorption on Cu(111) at room temperature. This is in agreement with the O 1s

spectrum for the as-deposited sample, where the peak at 531.3 eV originates from the

two equivalent oxygen atoms in the −COO associated with deprotonated molecules.51

Two additional contributions arise from carbonyl (C=O at 532.3 eV) and hydroxyl

(−OH at 533.6 eV)44-45 (see Figure 4-2b). Additionally, for a very precise fit of the

spectra one small feature at 532.5 eV is required. The ratio of deprotonated to intact

molecule determined from the O 1s core level is 0.86, which is reasonably close to the

ratio derived from the C 1s core level (0.75). The spacing between the −COO peak and

the carbonyl is 1 eV, which is consistent with previous reports.44 This is indicative of

the simultaneous presence of deprotonated and intact carboxylic groups, in agreement

with previously reported studies for carboxylated precursors that adsorb partially

deprotonated onto different Cu surfaces.15, 44, 46, 51


Figure 4-2: C 1s, O 1s and N 1s spectra for PDC deposited onto Cu(111) at RT and annealed up to

270°C. Line plot with black circle markers represents the acquired data, and solid gray lines show the

envelope of the fitted peaks. Fitted components are color-coded and detailed in the text.

Following room-temperature deposition, two main regions of intensity are

observed in the N 1s core level: a feature at lower binding energy that we attribute to

the nitrogen in the first layer and a higher binding energy feature consistent with

multilayer adsorption24 in which molecules interact with each other through hydrogen

bonding between the carboxyl/carboxylate group and nitrogen. The peak at lower BE

can be deconvolved into two components, suggesting the existence of different

adsorption configurations,52 which imply different orientations of the nitrogen with

respect to the surface. Figure 4-3 shows the calculated adsorption geometries and

energies for the configurations suggested for PDC on Cu(111) when (a) intact, (b)

singly deprotonated and (c) completely deprotonated. In Figure 4-3a, we show the

minimum-energy adsorption geometry for a single molecule in isolation, which is

inclined slightly from the plane of the surface and stabilized by an adsorption energy

of −0.92 eV. The peak at 399.8 eV is assigned to the nitrogen atom for molecules

chemisorbed to the surface via the lone pair24, 52 consistent with the calculated

geometry for the intact molecule, and the peak at 399.1 eV can be assigned to non-

surface-adsorbed nitrogen, consistent with a monopodal adsorption geometry of PDC

via carboxylate−Cu bonding with the nitrogen oriented away from the surface. The

calculated geometry corresponding to this state is shown in Figure 4-3b and

corresponds to an adsorption energy of −3.18 eV.


Figure 4-3: Calculated adsorption geometries and energies for configurations for PDC on Cu(111)

when (a) intact, (b) singly deprotonated and (c) completely deprotonated.

After annealing the sample to 100°C for 10 min the integrated intensity of the

phenyl peak is reduced by about 50%, suggestion that the multilayer has desorbed,

leaving only one monolayer. This desorption produces a shift to lower binding energy

by 0.4 eV for all peaks. The shift arises from enhanced core-hole screening for the

surface monolayer. In addition, the carboxyl peak at 289.2 eV almost vanishes,

indicating that deprotonation has nearly completed at this point. However, we cannot

exclude the scenario in which some fraction of molecules in the first layer were

completely deprotonated immediately upon adsorption, and that desorption of the

multilayer reveals the previously attenuated signal from these molecules. The

corresponding O 1s spectrum corroborates the complete deprotonation of the PDC: the

hydroxyl and carbonyl peaks have disappeared, and the only remaining oxygen peak

at 531.1 eV corresponds to ‒COO‒. This is consistent with reported XPS studies on

other carboxylic acids on copper.15, 44 Deconvolution of N 1s spectra reveals that the

configuration of the molecule changes upon completion of the deprotonation reaction:

the integrated intensity of the peak assigned to the chemisorbed nitrogen at 399.5 eV

decreases in conjunction with an increase in peak intensity at 398.8 eV, reflecting that

the carboxylate-dominated geometry is the favored geometry compared to the

chemisorbed nitrogen geometry for deprotonated PDC. This is consistent with a

bipodal adsorption configuration via carboxylate−Cu bonding, as shown in Figure


4-3c, which corresponds to an adsorption energy of −5.67 eV. Deprotonated trimesic

acid (TMA) on Cu(100) has also been reported to adsorb with the carboxylate group

oriented into the surface.53 The higher binding energy peaks at 401.4 and 402.1 eV

have almost disappeared following desorption of the multilayer, as expected for

second-layer species.

Annealing the sample to 270°C causes the peak at 531.1 eV in the O 1s to vanish,

confirming that the carboxylate has been fully cleaved from the molecule.

Decarboxylation is also indicated by the C 1s region, where the peak at 288.1 eV has

vanished as well, as shown in Figure 4-2a. There is an obvious decrease in the intensity

of the peak corresponding to the pyridine ring as a result of annealing at 270°C, in

conjunction with the appearance of two peaks at 284.1 and 286.3 eV. One possible

explanation for these new states is that PDC undergoes ring opening due to scission of

a C−N bond. It has similarly been reported that annealing pyridine at 260°C on Ni(111)

leads to decomposition of the molecule.52

The N 1s core level spectra shown in Figure 4-2c reveal that after annealing PDC

at 270°C the two molecular adsorption peaks are no longer observable, while three

new peaks appear: the peak at 398.6 eV, assigned to nitrogen in a fully decarboxylated

molecule, and two peaks at 398.1 and 400.5 eV, which are attributed to broken

molecules which arise from N−C bond scission.

The thermal stability of this product was studied by annealing the sample to

480°C. The C 1s and N 1s spectra suggest that after this high-temperature treatment

the PDC has fully decomposed (see, Figure S3). The two N 1s peaks assigned to the

decomposed molecule at 270°C dramatically increase after full decomposition of the

molecule at 480°C, in agreement with other works.36 Those studies reported that

decomposition of the molecule produces fragments such as CH2, (CH)n52 and (CN)2,

36

suggesting that the decomposed molecule will leave behind diminished and non-

stoichiometric on-surface residue, consistent with our measurements.

4.4.2 Activation Energy for the Decarboxylation Reaction of PDC

The decarboxylation reaction was further studied to examine the reaction

kinetics. The Arrhenius equation can be applied to determine the activation energies

for thermally-activated reactions;54-55


𝑘 = 𝐴 𝑒−𝐸𝑎𝑘𝐵𝑇,

where k is the reaction rate, A is the pre-exponential factor which is dependent

on the rate of transfer of energy to a decomposition site, Ea is the activation energy, kB

is Boltzmann’s constant and T is the temperature. By assuming that A and Ea are

independent of temperature,56-57 linearization of the equation yields;

ln(𝑘) =−𝐸𝑎

𝑘𝐵 (

1

𝑇) + 𝑙𝑛 (𝐴),

such that the activation energy can be calculated from the slope of a plot of ln(k)

versus 1/T.

To determine the decarboxylation rate, spectra were acquired continuously over

the C 1s region with the sample held at a fixed temperature.58 The spectra could then

be deconvolved to determine how the spectral intensity in the −COO− region varied

with time. Data sets acquired at 225, 230 and 235°C are shown in Figure 4-4.

Figure 4-4: Stacked plots of the C 1s region for PDC on Cu(111) with the sample held at a constant

temperature of 225 (a), 230 (b) and 235°C (c). Each spectrum was collected over ~30 s.

From the fitted components of the C 1s region, the proportion of reacted

molecules can be defined as 1 − (3ICOO/2IC-C). A plot of the proportion of reacted

molecules as a function of time for each of the three data sets is shown in Figure 4-5a.

The plots are linear, suggesting zero-order kinetics, and the slope of the linear fit for each

data set gives the decarboxylation rate at the respective temperature. The plot of the

natural logarithm of the obtained rates as a function of 1/T, shown in Figure 4-5b,


allows calculation of the activation energy for decarboxylation directly from the slope.

In this case, Ea = 1.93 ± 0.17 eV.

Figure 4-5: (a) Decarboxylation rate and (b) Arrhenius plot for the decarboxylation reaction of PDC

on Cu(111) derived from analysis of PES spectra of the C 1s region at 225, 230 and 235°C, where k is

the reaction rate determined from a.

NEXAFS Study of the Annealing Treatment of 3,5-Pyridinedicarboxylic Acid on

Cu(111)

The adsorption geometry of the molecule on the surface was further investigated

using near-edge X-ray absorption fine structure (NEXAFS) spectroscopy. Typical

NEXAFS spectra for an as-deposited multilayer film are shown along with indicative

peak fitting in Figure 4-6.


Figure 4-6: NEXAFS spectra collected at the carbon (a), nitrogen (b) and oxygen (c) K-edges for an

as-deposited multilayer sample (sample 1) with = 20°. Indicative peak-fitting analyses are shown

through the multicoloured components ascribed to each spectrum. (Inset) Angular dependence of each

set of NEXAFS spectra is shown. The inset of (a) shows the experimental geometry and the definition

of the angle θ, which relates the orientation of the electric field polarization to the sample surface.

The carbon K-edge has two distinct spectral components that can be associated

with the PDC molecule: the signal originating from the aromatic ring, apparent

between 283 and 287 eV, and the signal from the carboxylic groups, apparent between

287 and 291 eV. NEXAFS of pyridine has previously been studied with a combined


experimental and theoretical approach59, and the contributions to the spectral structure

are well understood. The split peak centered at 285 eV arises from C 1s → 1π*

transitions, with the splitting attributable to the chemical difference between the

proximity of carbon atoms to the nitrogen atom in the pyridine ring. The ortho-carbons

give rise to the higher energy component at 285.5 eV, and the carbons in the meta and

para positions give rise to the lower energy component at 284.8 eV. NEXAFS of small

carboxylated aromatics have also been previously studied experimentally and

theoretically60 allowing us to interpret the spectral region between 287 and 289 eV.

Here, the peaks at 288.3 and 291 eV can be attributed to C 1s → π* transitions

originating from the carboxylic carbon. At higher energies, the spectral features arise

from C 1s → * transitions.

The primary spectral component at the nitrogen K-edge (399.5 eV) can be

assigned to an N 1s → 1π* transition, with a high-energy shoulder (400.8 eV) arising

from a vibrationally-assisted N 1s → 2π* transition. The peak at 402.4 eV occurs due

to a N 1s → 3π* transition.59 At the oxygen K-edge, well-defined peaks arising from

carbonyl O 1s → π* transitions are evident at 532.1 and 534.3 eV, with an extended

structure due to both carbonyl- and hydroxyl-related transitions extending through the

region above 535 eV.60

Previous work on carboxylated aromatics at metal surfaces suggests that neither

the carbon nor oxygen-related π* resonances show a significant dependence on the

chemical state of the carboxylic/carboxylate group.30, 47 By taking NEXAFS data

through our annealing range, we have obtained data that directly address this point.

Figure 4-6 shows the carbon and oxygen K-edges for PDC films annealed to different

temperatures. We find that the primary resonances are not significantly changed. In

the carbon K-edge, the aromatic-related feature stays at a constant position. A slight

reduction in the relative intensity of the high energy split peak with respect to the low-

energy split peak can be seen in the deprotonated molecule, but we cannot

unequivocally ascribe this to deprotonation. In the carboxylic/carboxylate region, the

peaks at 288.3 and 291 eV are relatively unchanged in position and shape as the

molecule goes from intact to partially deprotonated to deprotonated. A small shoulder

peak at ~289 eV loses intensity with reaction; this peak is related to the carbon in the

carboxylic/carboxylate group60 and may be sensitive to the deprotonation. The spectral

contributions above 290 eV are associated with * orbitals.60 A small peak at ~290 eV


is present at RT, disappears after annealing to 100°C, and is restored after annealing

to 155°C. Data taken at 55° radiation incidence through the same annealing steps

appear to show the same trend but are much noisier so do not unequivocally support

this behavior. At the oxygen K-edge, the primary carbonyl-related resonances at 532.1

and 534.3 eV stay constant in position and shape as the molecule deprotonates.

Spectral changes, mostly in the form of intensity loss, occur throughout the region

above 536 eV with deprotonation of the molecule. Since there are hydroxyl-related

contributions in this range, this spectral loss is expected.60 In both the carbon and the

oxygen spectra, the molecule is clearly changed after annealing to 280°C. In oxygen,

this corresponds to the loss of all signal intensity, as expected for a decarboxylated

molecule. In carbon, this corresponds to the loss of well-defined spectral features

associated with the aromatic ring and carboxylic region. The sharp double peak

indicative of the pyridine ring has been replaced by a broad, poorly-defined resonance

that we ascribe to molecular remnants following the disintegration of the pyridine ring.

Figure 4-7: Evolution of the carbon (a) and oxygen (b) K-edges with annealing of the PDC film. RT

corresponds to an as-deposited multilayer, 100°C to a film of partially deprotonated molecules, 155°C

to a film of fully deprotonated molecules, and 280°C to the remnants of a broken, decarboxylated

molecule (carbon K-edge) and to near-complete desorption of all oxygen from the surface (oxygen k-

edge). All spectra were collected at = 20°.

Through consideration of the selection rules for the electric dipole transitions

associated with photon absorption, NEXAFS spectra acquired at different incident

radiation angles can be analyzed to determine the orientation of molecular orbitals with

respect to the surface. The formalism for this process is described by Stöhr,61 with the

geometry dependence of the intensity of a given NEXAFS transition described by the


following relation for a 3-fold-symmetric surface and a polarization factor of 1, as is

typical for an undulator beamline

𝐼(𝛼, 𝜃) = 𝐴 [1 +1

2(3 cos2𝜃 − 1)(3 cos2𝛼 − 1)]

where A is a constant, is the angle between the incident radiation and the

sample plane, and α is the angle between the sample normal and the transition dipole

moment of the molecular orbital, which for a π* orbital is perpendicular to the plane

of the aromatic ring or –COO–/COOH group.

We acquired NEXAFS spectra at three different incident angles: glancing

incidence at = 20°, magic (tilt-independent) angle at = 55°, and normal incidence

at = 90°. For planar adsorption of an aromatic molecule like PDC, the selection rules

dictate that that intensity associated with π* transitions should vanish for normal

incidence radiation and should be maximized as → 0°. The insets in Figure 4-6a–c

clearly show that for all three absorption edges measured a pronounced angular

dependence (dichroism) is observed. Similar considerations can be made for the *

orbitals,61 but we focus here on the π* transitions since these transitions are clearly

defined by sharp peaks and can be used to fully characterize the adsorption geometry

of the molecule.

Table 4-1 summarizes the calculated adsorption geometries for the PDC

molecule in a submonolayer film, a multilayer film, and thorough annealing at 100

(leading to partial deprotonation) and 155°C (full deprotonation). Similar results were

obtained on a second sample, described in Table S1 in the SI. In the submonolayer

film, the molecules adsorb at a relatively low inclination with respect to the surface

(∼35°), whereas in the multilayer film the molecules assume a more upright geometry

(∼45°). In both sub-ML and multilayer films the inclination of the aromatic ring and

the carboxylic group agree within uncertainty and to high precision, so we assume that

the molecule remains untwisted in its intact form.


Table 4-1. NEXAFS average adsorption angles (relative to the surface plane) for submonolayer and

Multilayer PDC on Cu(111) α

Pyridine Carboxylic/carboxylate C K-edge N K-edge calcd C K-edge O K-edge calcd

as-deposited sub-

monolayer 31° ± 3° 25.2° 36° ± 3° 17.2° / 16.9°

as-deposited

multilayer

45° ± 3° 41° ± 1°

44° ± 1° 44° ± 1°

annealed to 100°C

(partially

deprotonated)

52° ± 5° 58° ± 2° 60.0° 51° ± 7° 44° ± 1° 65.1° / 59.4°

annealed to 155°C

(fully deprotonated) 31° ± 10° 24° ± 7° 24.9° 30° ± 20° 25° ± 4° 40.5° / 38.4°

α Following annealing to 100°C, the uncertainty on the inclination angles calculated from the carbon K-

edge increase but the uncertainties associated with the nitrogen and oxygen edges remain relatively low.

The partially deprotonated molecule assumes a more upright geometry in the aromatic ring (from ∼52°

to ∼ 58°), whereas data from the oxygen K-edge suggest that the carboxylic/carboxylate groups may be

slightly out-of plane with respect to the ring, with an inclination of only 44°. However, the carbon K

edge/nitrogen K-edge data put the inclination of the ring and carboxylic/carboxylate in agreement with

one another, suggesting that any distortion of the molecule is minimal. This is borne out by the

calculation results for a singly deprotonated PDC molecule, which is found to have a minimum

adsorption energy when the deprotonated group points down into the substrate, leaving the ring tilted at

an angle of 60° with respect to the surface normal and with the carboxylic/carboxylate essentially

aligned with the aromatic ring.

Annealing the film to 155°C produced a completely deprotonated PDC layer.

NEXAFS measurements from all edges provide a consistent result, with calculated

adsorption tilt geometries from ∼25° to ∼30°. The uncertainties on all calculated

values are quite large (4−20°), which may arise from inhomogeneities in the film due

to adsorption at step edges and defects. These results are in partial agreement with the

calculated adsorption geometry of a fully deprotonated molecule, where the ring is

found to tilt by ∼25°. In the calculated geometry, the carboxylate groups twist out of

plane to form backbonds to the substrate, but our experimental data do not reveal this

twist, which may be obscured by the large uncertainties and/or by the existence of a

minority adsorption phase, as suggested by the spectral signature of surface-adsorbed

nitrogen in the SRPES data (Figure 4-2b). As shown in Figure 4-7, spectra collected

after annealing to 280°C are indicative of a broken molecule.


4.5 DISCUSSION

We learn two important things from our SRPES data: (1) the PDC molecule

assumes different adsorption geometries on the surface, with the nitrogen coordinated

both into and out of the surface, even as it deprotonates completely, and (2) the process

of molecular fragmentation is competitive with the process of decarboxylation. The

former finding indicates that additional adsorption geometries may be energetically

similar to the ones identified in Figure 4-3. The latter finding suggests that obtaining

a polymer product from PDC may be challenging, since fragmentation may occur

before the on-surface diffusion of decarboxylated moieties that is required to form

oligomers/polymers. Similar competitive activation/fragmentation reactions have been

observed in other heteroatom-containing molecules,62-63 where they can preclude the

formation of the desired polymer product.64

From our kinetic study, we obtain a reaction barrier that is higher than the free

energy barrier predicted from a DFT-based calculation of the transition state involved

in the decarboxylation of biphenyl-4-carboxylic acid on Cu(111).15 This difference can

be rationalized in the context of the variable adsorption geometry of PDC on Cu(111).

As evidenced in our study, the PDC molecule adopts a semi-upright configuration

during deprotonation, and a similar dependence of adsorption geometry with state of

reaction could occur during decarboxylation. If the adsorption geometry reduces the

proximity of the molecular reaction site to the catalytic metal surface or places the

−COO− in some other unfavorable geometry, we would expect to see a corresponding

increase in the reaction barrier as compared to the case of planar adsorption. The DFT-

calculated structure for biphenyl-4-carboxylic acid suggests that it is adsorbed

relatively flat on Cu(111), putting the carboxylic group close to the Cu surface. This

is consistent with the experimental decarboxylation activation barrier for PDC on

Cu(111), Ea = 1.93 ± 0.17 eV, exceeding the one calculated for biphenyl-4-carboxylic

acid on Cu(111), Ea = 1.63 eV.

As shown in Figure 4-4, annealing time is a determining factor for the state of

reaction of the molecule. PES data suggest that the molecule is partially decomposed

after annealing for 10 min at 270°C, whereas the NEXAFS data show the molecule is

fully decomposed after annealing for 5 min at 280°C. This molecular fragmentation

result suggests that ring opening could onset at temperatures below 270°C and may be


a significant factor in samples that are held for a long time at temperatures in the range

where decarboxylation occurs.

The NEXAFS-observed changes in adsorption geometry of the PDC could arise

from two different sources: (1) the coverage of the molecule on the surface, and (2)

the chemical state of the molecule. The coverage clearly has an impact on the

adsorption geometry: the sub-ML and multilayer films of unreacted molecules have

different adsorption geometries, due to intermolecular interactions within the film.

This effect has previously been observed for other systems, e.g., benzene on Pd(111),

which evolves from a flat adsorption geometry at low coverage (< 0.16 ML) to a more

upright geometry with increasing coverage,65 and pyridine on Ag(111) at 100 K, where

the molecular tilt angle increases from 45° ± 5° to 70° ± 5° with increasing coverage.66

However, as shown in Figure 4, even annealing at temperatures above 225°C does not

cause significant desorption. Deprotonation of the molecule is likely much more

important for this change in adsorption geometry, since coordination of the –COO–

groups into the surface provides a favorable charge compensation and a strong

surface–molecule interaction, as indicated by our DFT calculations. A transition from

planar to upright adsorption has previously been associated with the deprotonation of

a similar small carboxylated aromatic molecule, trimesic acid, on Cu(100).53 Here, the

modification of the adsorption geometry is consistent with a transition from the

nitrogen-dominated surface interaction of the intact molecule to a carboxylate-

dominated interaction for the deprotonated molecule, where simultaneous surface

adsorption of both carboxylated groups in a low-density film results in a lower tilt

angle than the nitrogen-dominated interaction in a dense film. Between these two

points, the molecule is adsorbed in a more upright position, consistent with the

monopodal adsorption shown in Figure 4-3b, which accommodates one surface-bound

carboxylate group and one –COOH oriented out of the surface. Hence, our results

indicate that the molecule may assume a variety of adsorption geometries during the

reaction process, with each of the configurations in Figure 4-3 defining the majority

adsorption geometry at some step of the process, and the molecule-surface interaction

increasing successively as the molecule deprotonates first once, and then twice.

Finally, we note that in low-temperature studies of pyridine on Pt(111),

annealing increases the tilt angle of the molecule to normal adsorption on the surface,

a phenomenon attributed to C–H dehydrogenation and the formation of a carbon–metal


bond at an ortho position.67 We do not anticipate that this phenomenon is relevant here,

since our observed tilt angle does not consistently increase with annealing, and since

we do not see an increase in spectral intensity in the C–Cu region following annealing

to 155°C.

4.6 CONCLUSIONS

Following room temperature deposition, SRPES indicates that PDC is partially

deprotonated and adsorbs in two different configurations on Cu(111): (1) via the

nitrogen lone pair and (2) via a carboxylate group. NEXAFS confirms that the

molecules are tilted with respect to the surface, and that the angle of inclination

depends on coverage. Following annealing at 100°C, which causes further

deprotonation of the molecules, N 1s SRPES indicates that the molecules shift away

from a lone-pair-driven adsorption geometry. DFT suggests that the molecules reorient

into a predominantly monopodal adsorption geometry in which carboxylate–Cu

bonding prevails, and NEXAFS at this point indicates that the molecules are more

upright, consistent with monopodal adsorption. With full deprotonation, the molecules

assume a lower inclination with respect to the surface, as expected for the bipodal

adsorption that orients both carboxylate groups into the surface. Continuous

observation of the C 1s core level during annealing at different temperatures allowed

us to investigate the kinetics of the decarboxylation reaction, for which we extract an

activation energy of 1.93 ± 0.17 eV. This is larger than the previously calculated

activation energy for decarboxylation of a planar aromatic molecule15 and suggests

that adsorption geometry may influence this reaction barrier. The statistical

implications of this activated process are quite apparent in our data, where the time at

temperature, rather than just the highest annealing temperature reached, dictates the

final state. This is particularly evident in the high-temperature range (i.e. at

temperatures greater than 220°C), where decarboxylation and molecular

fragmentation, through C−N bond scission, are competitive processes. This

competition means that achieving a polymeric product through decarboxylation of

PDC may be more challenging than obtaining a polymer from a molecule where

fragmentation does not significantly occur in the same temperature range. Vibrational

spectroscopy can be quite useful in identifying these fragments and could be

illustrative in future work. The results presented in this paper contribute to our

fundamental understanding of how decarboxylation proceeds in molecules that are


subject to a variety of evolving interactions with the surface during the reaction

process.


4.7 REFERENCES

1. El Garah, M.; MacLeod, J. M.; Rosei, F. Covalently bonded networks through

surface-confined polymerization. Surface Science 2013, 613, 6-14.

2. Lindner, R.; Kühnle, A. On‐Surface Reactions. ChemPhysChem 2015, 16 (8),

1582-1592.




4. Franc, G.; Gourdon, A. Covalent networks through on-surface chemistry in

ultra-high vacuum: state-of-the-art and recent developments. Physical Chemistry

Chemical Physics 2011, 13 (32), 14283-14292.

5. Chen, Z.; Molina-Jirón, C.; Klyatskaya, S.; Klappenberger, F.; Ruben, M. 1D

and 2D Graphdiynes: Recent Advances on the Synthesis at Interfaces and Potential

Nanotechnological Applications. Annalen der Physik 2017, 529 (11), 1700056-n/a.



7. Ullmann, F.; Bielecki, J. Ueber synthesen in der biphenylreihe. Berichte der

Deutschen Chemischen Gesellschaft 1901, 34 (2), 2174-2185.


2017, 53 (56), 7872-7885.

9. Eichhorn, J.; Heckl, W. M.; Lackinger, M. On-surface polymerization of 1, 4-

diethynylbenzene on Cu (111). Chemical Communications 2013, 49 (28), 2900-2902.

10. Gao, H.-Y.; Franke, J. r.-H.; Wagner, H.; Zhong, D.; Held, P.-A.; Studer, A.;

Fuchs, H. Effect of metal surfaces in on-surface glaser coupling. Journal of Physical

Chemistry C 2013, 117 (36), 18595-18602.

11. de Oteyza, D. G.; Gorman, P.; Chen, Y.-C.; Wickenburg, S.; Riss, A.;

Mowbray, D. J.; Etkin, G.; Pedramrazi, Z.; Tsai, H.-Z.; Rubio, A. Direct imaging of

covalent bond structure in single-molecule chemical reactions. Science 2013,

1238187.

12. Sattler, J. J.; Ruiz-Martinez, J.; Santillan-Jimenez, E.; Weckhuysen, B. M.

Catalytic dehydrogenation of light alkanes on metals and metal oxides. Chemical

Reviews 2014, 114 (20), 10613-10653.




9663.

14. Lloyd, P. B.; Swaminathan, M.; Kress, J. W.; Tatarchuk, B. J. Temperature

programmed desorption study of the adsorption and absorption of hydrogen on and in

Cu (111). Applied Surface Science 1997, 119 (3-4), 267-274.




16. Terrones, H.; Lv, R.; Terrones, M.; Dresselhaus, M. S. The role of defects and

doping in 2D graphene sheets and 1D nanoribbons. Reports on Progress in Physics

2012, 75 (6), 062501.

17. Wang, H.; Maiyalagan, T.; Wang, X. Review on recent progress in nitrogen-

doped graphene: synthesis, characterization, and its potential applications. ACS

Catalysis 2012, 2 (5), 781-794.


18. Cai, J.; Pignedoli, C. A.; Talirz, L.; Ruffieux, P.; Söde, H.; Liang, L.; Meunier,

V.; Berger, R.; Li, R.; Feng, X. Graphene nanoribbon heterojunctions. Nature

Nanotechnology 2014, 9 (11), 896-900.

19. Niu, T.; Wu, J.; Ling, F.; Jin, S.; Lu, G.; Zhou, M. Halogen-Adatom Mediated

Phase Transition of Two-Dimensional Molecular Self-Assembly on a Metal Surface.

Langmuir 2018, 34 (1), 553-560.

20. Bronner, C.; Stremlau, S.; Gille, M.; Brauße, F.; Haase, A.; Hecht, S.; Tegeder,

P. Aligning the band gap of graphene nanoribbons by monomer doping. Angewandte


21. Basagni, A.; Vasseur, G.; Pignedoli, C. A.; Vilas-Varela, M.; Pena, D.;

Nicolas, L.; Vitali, L.; Lobo-Checa, J.; de Oteyza, D. G.; Sedona, F. Tunable band

alignment with unperturbed carrier mobility of on-surface synthesized organic

semiconducting wires. ACS Nano 2016, 10 (2), 2644-2651.

22. Davies, P.; Shukla, N. The adsorption of pyridine at clean, oxidised and

hydroxylated Cu (111) surfaces. Surface Science 1995, 322 (1), 8-20.

23. Cheng, F.; Ji, W.; Leung, L.; Ning, Z.; Polanyi, J. C.; Wang, C.-G. How

Adsorbate Alignment Leads to Selective Reaction. ACS Nano 2014, 8 (8), 8669-8675.

24. Furukawa, M.; Fujisawa, H.; Katano, S.; Ogasawara, H.; Kim, Y.; Komeda, T.;

Nilsson, A.; Kawai, M. Geometrical characterization of pyrimidine base molecules

adsorbed on Cu (1 1 0) surfaces: XPS and NEXAFS studies. Surface Science 2003,

532, 261-266.

25. Bandy, B.; Lloyd, D. R.; Richardson, N. V. Selection rules in photoemission

from adsorbates: Pyridine adsorbed on copper. Surface Science 1979, 89 (1-3), 344-

353.

26. Hahn, J. R.; Kang, H. S. Role of molecular orientation in vibration, hopping,

and electronic properties of single pyridine molecules adsorbed on Ag (1 1 0) surface:

A combined STM and DFT study. Surface Science 2010, 604 (3-4), 258-264.

27. Hahn, J.; Ho, W. Imaging and vibrational spectroscopy of single pyridine

molecules on Ag (110) using a low-temperature scanning tunneling microscope. The

Journal of Chemical Physics 2006, 124 (20), 204708.

28. Wang, D.; Xu, Q.-M.; Wan, L.-J.; Wang, C.; Bai, C.-L. Adlayer structures of

pyridine, pyrazine and triazine on Cu (111): An in situ scanning tunneling microscopy

study. Langmuir 2002, 18 (13), 5133-5138.

29. Perry, C.; Haq, S.; Frederick, B.; Richardson, N. Face specificity and the role

of metal adatoms in molecular reorientation at surfaces. Surface Science 1998, 409 (3),

512-520.

30. Cebula, I.; Lu, H.; Zharnikov, M.; Buck, M. Monolayers of trimesic and

isophthalic acid on Cu and Ag: the influence of coordination strength on adsorption

geometry. Chemical Science 2013, 4 (12), 4455-4464.

31. Frederick, B.; Leibsle, F.; Haq, S.; Richardson, N. Evolution of lateral order

and molecular reorientation in the benzoate/Cu (110) system. Surface Review and

Letters 1996, 3 (04), 1523-1546.

32. Han, B.; Li, Z.; Wandlowski, T. Adsorption and self-assembly of aromatic

carboxylic acids on Au/electrolyte interfaces. Analytical and Bioanalytical Chemistry

2007, 388 (1), 121-129.

33. Williams, F. J.; Cropley, R. L.; Vaughan, O. P.; Urquhart, A. J.; Tikhov, M. S.;

Kolczewski, C.; Hermann, K.; Lambert, R. M. Critical Influence of Adsorption

Geometry in the Heterogeneous Epoxidation of “Allylic” Alkenes: Structure and

Reactivity of Three Phenylpropene Isomers on Cu (111). Journal of the American







8198.

35. Di Giovannantonio, M.; Tomellini, M.; Lipton-Duffin, J.; Galeotti, G.;

Ebrahimi, M.; Cossaro, A.; Verdini, A.; Kharche, N.; Meunier, V.; Vasseur, G.; Fagot-

Revurat, Y.; F. Perepichka, D.; Rosei, F.; Contini, G. Mechanistic Picture and Kinetic

Analysis of Surface-Confined Ullmann Polymerization. Journal of the American





26471-26480.

37. Fairley, N., CasaXPS Manual 2.3. 15: Introduction to XPS and AES. Casa

Software: 2009.

38. Gann, E.; McNeill, C. R.; Tadich, A.; Cowie, B. C.; Thomsen, L. Quick AS

NEXAFS Tool (QANT): a program for NEXAFS loading and analysis developed at

the Australian Synchrotron. Journal of Synchrotron Radiation 2016, 23 (1), 374-380.

39. Giannozzi, P.; Baroni, S.; Bonini, N.; Calandra, M.; Car, R.; Cavazzoni, C.;

Ceresoli, D.; Chiarotti, G. L.; Cococcioni, M.; Dabo, I. QUANTUM ESPRESSO: a

modular and open-source software project for quantum simulations of materials.

Journal of Physics: Condensed Matter 2009, 21 (39), 395502.

40. Perdew, J. P.; Burke, K.; Ernzerhof, M. Generalized gradient approximation

made simple. Physical Review Letters 1996, 77 (18), 3865.

41. Dion, M.; Rydberg, H.; Schröder, E.; Langreth, D. C.; Lundqvist, B. I. Van der

Waals density functional for general geometries. Physical Review Letters 2004, 92

(24), 246401.

42. Thonhauser, T.; Cooper, V. R.; Li, S.; Puzder, A.; Hyldgaard, P.; Langreth, D.

C. Van der Waals density functional: Self-consistent potential and the nature of the

van der Waals bond. Physical Review B 2007, 76 (12), 125112.

43. Román-Pérez, G.; Soler, J. M. Efficient implementation of a van der Waals

density functional: application to double-wall carbon nanotubes. Physical Review

Letters 2009, 103 (9), 096102.

44. Stepanow, S.; Strunskus, T.; Lingenfelder, M.; Dmitriev, A.; Spillmann, H.;

Lin, N.; Barth, J. V.; Wöll, C.; Kern, K. Deprotonation-Driven Phase Transformations

in Terephthalic Acid Self-Assembly on Cu(100). The Journal of Physical Chemistry

B 2004, 108 (50), 19392-19397.

45. Payer, D.; Comisso, A.; Dmitriev, A.; Strunskus, T.; Lin, N.; Wöll, C.; DeVita,

A.; Barth, J. V.; Kern, K. Ionic Hydrogen Bonds Controlling Two‐Dimensional

Supramolecular Systems at a Metal Surface. Chemistry – A European Journal 2007,

13 (14), 3900-3906.

46. Classen, T.; Lingenfelder, M.; Wang, Y.; Chopra, R.; Virojanadara, C.; Starke,

U.; Costantini, G.; Fratesi, G.; Fabris, S.; de Gironcoli, S.; Baroni, S.; Haq, S.; Raval,

R.; Kern, K. Hydrogen and Coordination Bonding Supramolecular Structures of

Trimesic Acid on Cu(110). The Journal of Physical Chemistry A 2007, 111 (49),

12589-12603.

47. Shen, C.; Cebula, I.; Brown, C.; Zhao, J.; Zharnikov, M.; Buck, M. Structure

of isophthalic acid based monolayers and its relation to the initial stages of growth of

metal–organic coordination layers. Chemical Science 2012, 3 (6), 1858-1865.


48. Cebula, I.; Shen, C.; Buck, M. Isophthalic acid: a basis for highly ordered

monolayers. Angewandte Chemie International Edition 2010, 49 (35), 6220-6223.

49. Kanninen, L.; Jokinen, N.; Ali-Löytty, H.; Jussila, P.; Lahtonen, K.; Hirsimäki,

M.; Valden, M.; Kuzmin, M.; Pärna, R.; Nõmmiste, E. Adsorption structure and

bonding of trimesic acid on Cu(100). Surface Science 2011, 605 (23–24), 1968-1978.

50. Jhuang, J.-Y.; Lee, S.-H.; Chen, S.-W.; Chen, Y.-H.; Chen, Y.-J.; Lin, J.-L.;

Wang, C.-H.; Yang, Y.-W. Adsorption and Reaction Pathways of 1H-Pyrazole on

Cu(100) and O/Cu(100). The Journal of Physical Chemistry C 2018, 122 (11), 6195-

6208.

51. Lin, N.; Payer, D.; Dmitriev, A.; Strunskus, T.; Wöll, C.; Barth, J. V.; Kern, K.

Two-Dimensional Adatom Gas Bestowing Dynamic Heterogeneity on Surfaces.


52. Cohen, M.; Merrill, R. Adsorption of pyridine on Ni (111): A high-resolution

electron energy loss spectroscopy, angular-resolved UV photoemission, and x-ray

photoelectron spectroscopy study. Langmuir;(USA) 1990, 6 (7).

53. Dmitriev, A.; Lin, N.; Weckesser, J.; Barth, J. V.; Kern, K. Supramolecular

Assemblies of Trimesic Acid on a Cu(100) Surface. The Journal of Physical Chemistry

B 2002, 106 (27), 6907-6912.

54. Starink, M. J. The determination of activation energy from linear heating rate

experiments: a comparison of the accuracy of isoconversion methods. Thermochimica

Acta 2003, 404 (1), 163-176.

55. Laidler, K. J. The development of the Arrhenius equation. Journal of Chemical

Education 1984, 61 (6), 494.

56. Flynn, J. H.; Wall, L. A. A quick, direct method for the determination of

activation energy from thermogravimetric data. Journal of Polymer Science Part C:

Polymer Letters 1966, 4 (5), 323-328.

57. Flynn, J. H.; Wall, L. A. General treatment of the thermogravimetry of

polymers. J Res Nat Bur Stand 1966, 70 (6), 487-523.

58. Papp, C.; Steinrück, H.-P. In situ high-resolution X-ray photoelectron

spectroscopy – Fundamental insights in surface reactions. Surface Science Reports

2013, 68 (3), 446-487.

59. Kolczewski, C.; Püttner, R.; Plashkevych, O.; Ågren, H.; Staemmler, V.;

Martins, M.; Snell, G.; Schlachter, A.; Sant’Anna, M.; Kaindl, G. Detailed study of

pyridine at the C 1 s and N 1 s ionization thresholds: The influence of the vibrational

fine structure. The Journal of Chemical Physics 2001, 115 (14), 6426-6437.

60. Zhang, W.; Nefedov, A.; Naboka, M.; Cao, L.; Wöll, C. Molecular orientation

of terephthalic acid assembly on epitaxial graphene: NEXAFS and XPS study.

Physical Chemistry Chemical Physics 2012, 14 (29), 10125-10131.


Vol. 25.

62. Dinca, L. E.; Fu, C.; MacLeod, J. M.; Lipton-Duffin, J.; Brusso, J. L.; Szakacs,

C. E.; Ma, D.; Perepichka, D. F.; Rosei, F. Unprecedented Transformation of

Tetrathienoanthracene into Pentacene on Ni (111). ACS Nano 2013, 7 (2), 1652-1657.

63. Dinca, L. E.; MacLeod, J. M.; Lipton-Duffin, J.; Fu, C.; Ma, D.; Perepichka,

D. F.; Rosei, F. Tailoring the Reaction Path in the On-Surface Chemistry of

Thienoacenes. The Journal of Physical Chemistry C 2015, 119 (39), 22432-22438.

64. Di Bernardo, I.; Hines, P.; Abyazisani, M.; Motta, N.; MacLeod, J.; Lipton-

Duffin, J. On-Surface Synthesis of Polyethylenedioxythiophene. Chemical

Communications 2018, 54 (30), 3723-3726.


65. Lee, A. F.; Wilson, K.; Lambert, R. M.; Goldoni, A.; Baraldi, A.; Paolucci, G.

On the coverage-dependent adsorption geometry of benzene adsorbed on Pd {111}: a

study by fast XPS and NEXAFS. The Journal of Physical Chemistry B 2000, 104 (49),

11729-11733.

66. Bader, M.; Haase, J.; Frank, K.-H.; Puschmann, A.; Otto, A. Orientational

phase transition in the system pyridine/Ag (111): A near-edge x-ray-absorption fine-

structure study. Physical Review Letters 1986, 56 (18), 1921.

67. Johnson, A. L.; Muetterties, E.; Stohr, J.; Sette, F. Chemisorption geometry of

pyridine on platinum (111) by NEXAFS. The Journal of Physical Chemistry 1985, 89

(19), 4071-4075.


4.8 SUPPORTING INFORMATION

4.8.1 Desorption prior to complete deprotonation for PDC on Ag(111)

PDC was deposited onto Ag(111) held at -130°C, and was then stepwise

annealed up to 100°C. A stack of spectra, shown in Figure S4-8, reveals that the

molecule desorbs before the deprotonation reaction is completed.

Figure S4-8: C 1s spectra for PDC deposited onto Ag(111) kept at -130°C and annealed up to 100°C.

The beam energy is 486 eV.

4.8.2 Beam-induced deprotonation of PDC on Ag(111)

PDC was deposited onto Ag(111) at room temperature and spectra were

collected continuously as the sample was held in front of the synchrotron beam (h =

486 eV). A stack of C1s PES spectra acquired from sample is shown in Figure S4-9

and illustrates that the beam deprotonates the sample. The molecule adsorbed intact on

Ag, however a small contribution of COO- is observable in the spectrum indicating very

small fraction of the molecule deprotonated which is consistent with beam-induced

damage. Beam exposure to an identical spot for 5 minutes doubles the proportion of

deprotonated molecule.


Figure S4-9: Beam damage test for PDC on Ag(111). The beam energy is 486 eV and each scan was

acquired over 50 seconds. The as-deposited sample is at the bottom, with increasing beam damage

moving up the stack.

4.8.3 PES data at N1s for partially and completely decomposed PDC

The signature of the decomposition of the molecule is the red and orange peaks

in N1s peaks. At 270°C, the molecule is partially broken, and the red and yellow peaks

appear at binding energies of 398.1 eV and 400.4 eV. Further annealing up to 480°C

leads to full molecular decomposition, and a concominant increase in these two peaks.

Assigning these peaks as decomposition products is in agreement with reported results

for a simlar pyridine-based molecule after ring opening.1


Figure S4-10 C 1s and N 1s spectra for PDC deposited onto Cu(111) at 270°C and annealed up to

480°C.

4.8.4 NEXAFS data for an additional sample with over one ML of PDC

Table 4-2: NEXAFS calculated average adsorption angles for multilayer PDC on Cu(111). Based on

deposition time, with no consideration of the reduced sticking coefficient after monolayer completion,

coverage on this sample should be slightly more than double the coverage on the multilayer sample

reported in the main manuscript.

Pyridine Carboxylic/carboxylate

C K-edge N K-edge C K-edge O K-edge

As-deposited multilayer 59±5 58±1 57±7 51±1

Annealed to 80°C (partially deprotonated) 61±3 60±1 59±6 56±1

Annealed to 150°C (fully deprotonated) 46±10 32±3 40±10 43±40


References




26471-26480.

Chapter 5: Adsorption, deprotonation and decarboxylation of isophthalic acid on Cu(111) 87

Chapter 5: Adsorption, deprotonation and

decarboxylation of isophthalic

acid on Cu(111)

The findings from the last chapter show that 3,5-pyridine dicarboxylic acid

decarboxylated on Cu(111), however decarboxylation and molecular fragmentation,

through C−N bond scission, are competitive processes for this molecule. The

following chapter discusses research that I carried out in collaboration with my

research group and that was a study of the decarboxylation of isophthalic acid, a similar

molecule to 3,5-pyridine dicarboxylic acid but without nitrogen in the molecular

scaffold. In fact, replacing nitrogen with carbon in the central group of the precursor

may help to simplify the challenge for completing the decarboxylation reaction. The

deprotonation and decarboxylation of isophthalic acid on Cu(111) has been studied

using PES and NEXAFS collected at the Australian Synchrotron, as well as by using

STM on the Omicron (Scienta-Omicron GmbH) available at QUT.

The results of this study have been published in Langmuir: Abyazisani, M.;

Bradford, J.; Motta, N.; Lipton-Duffin, J.; MacLeod, J. Adsorption, Deprotonation,

and Decarboxylation of Isophthalic Acid on Cu(111). Langmuir 2019

88 Chapter 5: Adsorption, deprotonation and decarboxylation of isophthalic acid on Cu(111)


Published paper




field of expertise;






academic unit, and




Adsorption, deprotonation and decarboxylation of isophthalic acid on Cu(111): DOI:

10.1021/acs.langmuir.8b04233


Maryam Abyazisani Conducted lab-based experiments, conducted synchrotron

experiments, analysed data, drafted manuscript. Signature

Jonathan Bradford Conducted synchrotron experiments, revised manuscript

Nunzio Motta Revised manuscript, supervision.

Josh Lipton-Duffin Revised manuscript, supervision.

Jennifer MacLeod Conceived the project, conducted synchrotron experiments,

analysed data, revised manuscript, supervision.

RSC, Level 4, 88 Musk Ave, Kelvin Grove Qld 4059

Date SignatureName








https://doi.org/10.1021/acs.langmuir.8b04233

https://doi.org/10.1021/acs.langmuir.8b04233


5.1 ABSTRACT

The surface-assisted reaction of rationally designed organic precursors is an

emerging approach towards fabricating atomically precise nanostructures. Recently,

on-surface decarboxylation has attracted attention due to its volatile by-products which

tend to leave the surface during the reaction, which means that only the desired

products are retained on the surface. However, in addition to acting as the reactive site,

the carboxylic acid groups play a vital role in the adsorption configuration of small-

molecule molecular precursors and therefore in the reaction pathways. Here, scanning

tunnelling microscopy (STM), synchrotron radiation photoelectron spectroscopy

(SRPES) and near-edge x-ray absorption fine structure (NEXAFS) spectroscopy have

been employed to characterize the mono-deprotonated, fully deprotonated and

decarboxylated products of isophthalic acid (IPA) on Cu(111). IPA is partially reacted

(mono-deprotonated) upon adsorption on Cu(111) at room temperature. Angular

dependent x-ray photoelectron spectroscopy reveals that IPA initially anchors to the

surface via the carboxylate group. After annealing, the molecule fully deprotonates

and reorients so that it anchors to the surface via both carboxylate groups in a bipodal

configuration. NEXAFS confirms that molecule is tilted upon adsorption as well as

after full deprotonation. Following decarboxylation, the flat-lying molecule forms into

oligomeric motifs on the surface. This work demonstrates the importance of molecular

adsorption geometry for on-surface reactions.


5.2 INTRODUCTION

The fabrication of surface-confined covalent polymers from molecular precursors has

drawn attention in the past decade due to potential relevance to nanoscale electronic

and optical devices.1-2 Different on-surface reactions and precursors are being

explored with the goal of optimizing the resultant nanostructures. In this regard,

precursors functionalized with carboxylic acid3-5 have recently attracted attention as

they present a route to surface-confined polymer formation that may offer advantages

over Ullmann coupling, a reaction that has been widely used to grow 1D and 2D

polymers.6 The Ullmann reaction produces halogen by-products, which can remain

chemisorbed on the surface and might have a negative effect on the diffusion of

surface-stabilized radicals and the spatial extension of the polymers. In contrast, the

by-products resulting from decarboxylation, H2 and CO2 leave the surface.7-10

Gao et al. reported that the polymerization of 2,6-naphthalenedicarboxylic acid

via decarboxylative coupling proceeds in a three-step process: deprotonation to

convert the carboxylic groups to carboxylate, followed by cleavage of the carboxylate

groups and finally diffusion of the activated moieties to allow the formation of C-C

coupling. Each of these processes can be triggered through thermal activation.3 On

reactive surfaces, the presence of carboxylic acid can affect the adsorption geometry

of small molecules, since the molecule can deprotonate on deposition and the

molecular plane can be oriented upright or tilt with respect to the substrate.9, 11-13

Benzoic acid, a simple monocarboxylic acid, deprotonates upon adsorption on copper

surfaces and assumes an upright configuration with respect to the surface, i.e., the

molecules are oriented with their phenyl rings perpendicular to the substrate.14-21 This

adsorption geometry arises from the strong carboxylate-copper interaction.19 The

addition of more than one carboxylic group to a molecule introduces some

complication to this scenario owing to competition between copper-carboxylate and

carboxylate-carboxylate interactions.9, 22-23 For example, the copper-carboxylate

interaction suppresses hydrogen bonding between molecules in terephthalic acid

(TPA), a dicarboxylated phenyl, since TPA adsorbs in a perpendicular orientation with


respect to the Cu(110) surface.9 In the TPA study, one of the carboxyl groups of

molecule deprotonates upon adsorption and forms a carboxylate group which anchors

to the copper surface and leaves the other acid group intact and oriented away from

the surface.

A central theme in these studies of aromatic carboxylic acid-functionalized

molecules has been the importance their adsorption configuration on the surface.

Systematic investigations of the adsorption of the various aromatic carboxylate acid

precursors will ultimately result in recipes for a deliberate design of precursors with a

rational number of carboxylate acid groups in a specific geometry to fabricate ordered

polymers via decarboxylative coupling. Understanding of molecule-substrate

interactions is a key step towards controlling the growth of polymer films at surfaces.24

Here, we present a study of the adsorption, deprotonation and decarboxylation

of 1,3-benzenedicarboxylic acid (isophthalic acid, IPA) on Cu(111). Photoemission

spectroscopy (PES) was used to monitor the chemistry of the IPA molecules.

Measurements collected at different angles with respect to the surface provide insight

into the orientation of the different parts of the molecule during the reactions. We

obtained additional insight into the adsorption geometry using near-edge x-ray

absorption fine structure (NEXAFS) spectroscopy, and by imaging the molecular films

using scanning tunnelling microscopy (STM). Together, these investigations provide

a complete picture of the adsorption, deprotonation, decarboxylation, and subsequent

C-C coupling of IPA on Cu(111).

5.3 EXPERIMENTAL

The ultrahigh vacuum (UHV) system at Queensland University of Technology

has a base pressure better than 2 × 10−10 mbar and comprises two chambers: the first

is equipped with a dual-anode x-ray lamp (ScientaOmicron GmbH DAR 400) and a

hemispherical electron energy analyzer (iSphera), as well as standard facilities for

sample preparation, and the second chamber houses a scanning tunnelling microscope

(ScientaOmicron GmbH VT-AFM/XA). All STM images for this study were recorded

in constant current mode and at room temperature with an electrochemically etched

tungsten tip. Bias voltages are measured with respect to the tip. For every STM data

set, corresponding XP spectra were measured with the Al Kα x-ray laboratory source

to check the chemistry of the surface before STM scanning.


Synchrotron radiation photoelectron spectroscopy (SRPES) experiments were

performed on the soft x-ray (SXR) Beamline of the Australian Synchrotron. Survey

spectra were recorded using a beam energy of 1487 eV with a pass energy of 100 eV.

Core level spectra of the C 1s and O 1s regions were acquired using photon energies

of 486 and 980 eV, respectively. The pass energy was set at 20 eV during high

resolution scans for an overall energy resolution of 0.29 eV. CasaXPS software was

used to analyze the PES spectra 25 and all spectra were energy-calibrated by rigidly

shifting to set the Fermi level to zero binding energy. Shirley and linear backgrounds

were used for the C 1s and O 1s regions, respectively. The PE spectra of C 1s core

level were acquired with the surface normal at different angles with respect to the

analyzer to elucidate the carboxyl/carboxylate group location with respect to phenyl

group. PE spectra of the C 1s core level were also collected with different beam

energies between 380 eV and 908 eV.

NEXAFS spectra were collected using linearly polarized light at glancing

incidence (Ө = 20°), magic angle (Ө = 55°) and normal incidence (Ө = 90°) with

respect to the surface plane. The QANT software package were used to analyze all

spectra.26

Sample Preparation: a Cu(111) crystal was cleaned by repeated cycles of Ar+

ion sputtering (1 keV) at room temperature and flash annealing to 780 K. Low-energy

electron diffraction (LEED) and PES of C 1s and O 1s verified the sample cleanliness.

Commercially purchased isophthalic acid (IPA) (98%, Sigma Aldrich Corp.) was

thoroughly degassed at 383 K in UHV and was deposited by organic molecular beam

epitaxy (OMBE) from a Knudsen cell held at 393 K onto a room temperature (RT)

substrate for 10 minutes. The sample was subsequently heated stepwise until the

molecule fully decarboxylated.

5.4 RESULTS AND DISCUSION

The polymerization of IPA via decarboxylative coupling is shown in Figure

5-1a. In the decarboxylative reaction, bonds are successively cleaved at the carboxylic

group: deprotonation as a result of O-H bond scission and then decarboxylation as a

cleavage of COO– from the ring. Following decarboxylation, the activated moieties

can form the organometallic through participation of surface atoms/adatoms. Scission

of the C-Cu bond and coupling of the resultant activated moieties results in formation


of polymeric structures. The carboxylate groups in the IPA are in the meta positions,

thus three well-defined products are possible: zig-zag chains, crinkled chains and

rosette macrocycles (see Figure 5-1b).

Figure 5-1: (a) Decarboxylation reaction schematic for IPA molecule in the presence of copper and (b)

three possible structural products of polymerization of IPA, from left to right: zig-zag, crinkled and

rosette macrocycles respectively.

5.4.1 XPS study

The C 1s and O 1s photoemission spectra of IPA on Cu(111) are compiled in

Figure 5-2. The experimental data points are shown as black circles and the

corresponding fit with a gray line. The main feature consists of two peaks 284.9 eV

and 285.3 eV, which are assigned to the carbon atoms of phenyl rings in the first and

second layers respectively. Peaks at 288.0 and 289.4 eV originate from the carboxylate

and carboxylic groups, respectively, of molecules in the first layer.27-28 Additional

spectral intensity mandates the use of a second feature at 289.4 eV to account for

carboxylic groups of intact molecules in the second layer, which forms due to

deposition of slightly more than 1 ML of molecules. The small peak at 291.5 eV is

assigned to the π−π* shake-up transition of the aromatic system.27. It is well-known

that carboxylic acids either fully or partially deprotonate upon adsorption on copper

surfaces.4, 27, 29 Here, both intact (–COOH) and deprotonated (–COO–) groups are

evident in the spectrum originating from the first layer, with the deprotonation

contribution 40% larger than the intact contribution. After annealing to 373 K the

phenyl and carboxylic acid peaks originating from molecules in the second layer

disappear, consistent with desorption of intact molecules from the second layer, and

the –COOH and –COO– peaks become equal in magnitude (See Figure S1). The

presence of equal COO– and COOH in the first monolayer could arise from two

different scenarios: (a) fully deprotonated and intact IPA molecules coexisting on the


surface, (b) mono-deprotonated molecules. We believe that scenario (b) is more likely

and the first monolayer of IPA mono-deprotonates upon adsorption on Cu(111) at

room temperature. The adsorption geometry of the mono-deprotonated molecule will

be further discussed later, based on angular and energy dependent PES, which

illustrates the location of carboxyl (–COOH) and carboxylate groups (–COO–) respect

to the surface and phenyl ring.

The O 1s spectrum confirms the coexistence of chemical states for carboxylate

and carboxylic groups for the as-deposited sample: a peak at 531.6 eV is assigned to

the carboxylate group,28 while the hydroxyl (533.8 eV) and carbonyl (532.4 eV) peaks,

locked to a 1:1 ratio, are attributed to the two chemical state of oxygen in a carboxylic

group. This is in accordance with the results from the C 1s core level. Furthermore, a

pair of extra peaks at 532.9 and 534.4 eV is attributed to the carboxylic groups of

molecules in the second layer.

Figure 5-2: (a) C 1s and (b) O 1s spectra for IPA deposited onto Cu(111) at 300 K, followed by two

annealing steps at 453 K and 488 K. The black markers represent the acquired data, and the coloured

peaks show synthetic fits.

Before discussing the structural model of molecule in different stages, we turn

to spectroscopic data that can provide insight into the adsorption geometry of the

molecules. Comparing PE spectra collected at different angles with respect to the


surface normal provides information about the relative geometry of –COOH, –COO–

and phenyl with respect to the surface. All spectra were fitted with the same peak set

determined from normal-incidence data collected at 380 eV. The intensity ratios of –

COOH, –COO– and phenyl normalized to the total C 1s region for the as-deposited

sample are plotted in Figure 5-3. The change in the ratio between the phenyl and the

functional groups depends on the escape depth of the electrons.30-31 The distance that

a photoelectron has to travel to escape the surface is shorter when the optical axis of

the detector lies along the surface normal and increases when the emission angle is

off-normal, meaning that the chemical species closest to the substrate will decrease in

intensity with increased angle.

The comparison of the intensities of different signals reveals that –COOH

intensity increases with off-normal measurements, while the intensities of both phenyl

and –COO– decrease. This indicates that the molecule is not adsorbed planar, and that

the carboxylate group is buried beneath the aromatic ring. On the other hand, both the

–COOH and phenyl intensity are reduced at off-normal angles, suggesting that the

carboxyl group is in the outer-most plane surface of as-deposited sample. This is

consistent with the molecule being anchored to the copper via the –COO– group in a

monopodal configuration while the phenyl ring is perpendicular/tilted with respect to

the surface and –COOH pointing out of the surface.

A variation in intensity for spectra collected at different angles can arise from

photoelectron diffraction effects.32-33 To verify that the observed variation was due to

the adsorption geometry of the molecules and not photoelectron diffraction, we

confirmed our findings by varying the photon energy of the probe beam, which

provides similar information to the angle-resolved measurements, but without the

possibility for photoelectron diffraction effects (see Figure S2). This measurement

agrees with our angle-dependent results, supporting that the observed variation in the

angle-resolved measurements arises from the adsorption geometry of the molecules.

This interpretation is further supported by STM and NEXAFS measurements, which

will be discussed later. The same adsorption geometry has been reported for mono-

deprotonated IPA on Au(111) molecules on Au(111).34-35.


Figure 5-3: Angular dependence of the apparent stoichiometry of IPA deposited at room temperature

collected at hυ=380 eV using four different angles with respect to the surface normal. The intensity of

carboxylate, carboxyl and phenyl components have been normalized with repect to the total C 1s.

Inset shows a schematic of the adsorption geometry.

Annealing the sample at 453 K results in the disappearance of the carboxyl peak

in C 1s (c.f. Figure 5-2), demonstrating that the molecule is fully deprotonated at this

stage. This agrees well with the O 1s spectrum shown in the same figure, which

confirms that the hydroxyl and carbonyl peaks have vanished. We also note that the C

1s peak shifts towards lower binding energy. The shift is attributed to desorption of

second-layer molecules after the first annealing step, which enhances core-hole

screening for the surface monolayer.36

PE spectra of the fully deprotonated sample acquired at different angles and the

corresponding normalized intensities of –COO– and phenyl are shown in Figure 5-4.

Comparison of the two trends reveals that the relative intensity of –COO– signal is

highest at angles close to surface normal and in contrast the phenyl signal is stronger

for the off-normal measurements. This implies that both –COO– groups are beneath the

phenyl ring suggesting that the molecule has re-oriented so that both –COO– groups

coordinate to the surface. This implies a bipodal configuration where two –COO– are

facing the surface while the phenyl ring is tilted/perpendicular to the surface.


Figure 5-4: Angular dependence of the apparent stoichiometry of fully deprotonated IPA collected at

hυ=380 eV using four different angles with respect to the surface normal The intensity of carboxylate,

carboxyl and phenyl peak has been normalized with respect to the total C 1s. Inset shows a schematic

of the adsorption geometry.

Annealing up to 488 K induces decarboxylation of the molecule. Both the C 1s

and O 1s spectra shown in Figure 5-2 confirm that the majority of molecules have lost

their carboxylate groups.

5.4.2 NEXAFS study

The adsorption geometry of the molecule on the surface was further studied

using NEXAFS spectroscopy. Figure 5-5 shows the carbon K-edge and oxygen K-

edge NEXAFS spectra for surfaces with partially and fully deprotonated IPA, and for

a surface with decarboxylated IPA. Comparing the spectra obtained with s-polarized

(normal incidence), p-polarized (glancing incidence at 20°) and magic-angle (55°)

incidence light reveals the dichroism of the π* resonance of aromatic ring and

carboxylate/carboxylic acid groups, as shown in Figure 5-5. Consistent with previous

work on similar systems, we assign the two sharp peaks at 285 and 288 eV in the

carbon spectra to the π* resonance of the aromatic ring and carboxylate/carboxylic

acid group in the molecule, respectively, and the peaks at 532.7 and 535.3 eV in the

oxygen spectrum to the π* resonance of the carboxylic/carboxylate groups.30-31, 37 The

wide peaks at 293 and 300 eV in the carbon spectrum originate from the * resonance

of the aromatic ring and carboxylic acid group, respectively, with the * resonance of

the carboxylic/carboxylate group also responsible for the broad feature starting at ~540

eV in the oxygen spectrum.


Figure 5-5: NEXAFS spectra collected (a) at the carbon K-edge for half-deprotonated, fully-

deprotonated, and decarboxylated IPA on Cu(111) with s-polarized (normal incidence), magic angle

(55° incidence) and p-polarized (glancing incidence, 20°), and (b) at the oxygen K-edge for the half-

and fully-deprotonated samples using the same incident beam geometries.


We are able to characterize the adsorption geometry of the partially and fully

deprotonated molecules by creating synthetic (Gaussian) fits to the NEXAFS spectra

obtained at different incident radiation angles, considering the selection rules for the

electric dipole transitions associated with photon absorption, and applying the

formalism laid out in the book by Stöhr for π* resonances.38 Plotted intensities are

provided in the Figure S3. This analysis reveals that the half-deprotonated molecules

adsorb with their π* orbitals at 52°±3° with respect to the surface normal, whereas the

fully deprotonated molecules assume a less upright geometry ( = 46°±3°). In the half-

deprotonated molecules the carboxyl/carboxylic group tilt matches the phenyl ring tilt

( = 51°±3°), but in the fully-deprotonated molecules the carboxylates are less upright

than the ring ( = 33°±7°). The oxygen K-edge provides redundant information for the

carboxylic/carboxylate groups, and corroborates these findings. For the

decarboxylated molecule, the π* resonance at 289 eV has disappeared, and the average

inclination calculated from the phenyl-derived π* resonance is = 40°±3°. The

oxygen K-edge data shown in Figure 5-5b are similar to the carboxyl/carboxylic

orientations obtained from the carbon K-edge data. The extracted tilt angles are shown

in Table 5-1.

Table 5-1: NEXAFS-derived tilt angles for the phenyl and carboxylic/carboxylate of IPA in different

states of reaction. Angles are averages, and specify the angle of the π* orbital with respect to the

surface normal.

Reacted state of

molecule

Carbon K-edge Oxygen K-edge

Phenyl ring (°)

Carboxylic/carboxylate

(°)

Carboxylic/carboxylate

(°)

Half-deprotonated 52±3 51±3 46±1

Deprotonated 46±3 33±7 40±1

Decarboxylated 40±3 - -

5.4.3 STM study

Figure 5-6 shows STM images of IPA deposited on Cu(111) at RT. The as-

deposited self-assembled IPA is hard to scan, lacks lateral order, and the molecules

cannot be imaged individually. The corresponding XP spectrum reveals that 39% of

the carboxylate groups are deprotonated at this stage (see figure S2), which we


interpret as being due to a small amount of unreacted molecules in a second layer, on

top of a layer of singly-deprotonated molecules.

Figure 5-6: STM image of an IPA layer deposited onto Cu(111) at room temperature. Image

parameters: U=-360 mV, I=0.3 nA, 72 nm × 72 nm.

After annealing the sample to 373 K for 5 minutes the molecules rearrange into

small domains of chain-like lines aligned with their long axes along <2 -1 -1>, <-1 2 -

1> and ˂ -1 -1 2˃, shown in Figure 5-7. XPS reveals that 48% of the carboxylate groups

are deprotonated, which we take as an indication that the second layer has been

desorbed and that the layer comprises only singly-deprotonated molecules in contact

with the copper surface. Angle-resolved PES (see previous) indicated that these mono-

deprotonated molecules adsorb in an upright orientation with the carboxylate group

facing into the Cu(111) and the carboxyl group pointing up, i.e. molecules anchor to

Cu in a monopodal adsorption geometry via the carboxylate group. The intact carboxyl

groups are free to engage in hydrogen bonding with adjacent molecules, however, if

present, this interaction may not be highly stabilizing since scanning with the STM tip

easily perturbs the structure (see Figure S4). Previously, a disturbance caused by a

tunnelling current of more than 30 pA has also been reported for IPA in a monodentate

configuration on Au(111).34 Although additional stabilization of the structure is

provided via − stacking of the tilted phenyl rings, the structure of monopodal IPA

on the surface is fragile, as the intact carboxyl group hinders molecule from optimal

adsorption on the surface.34


Figure 5-7: STM image of the IPA layer after annealing to 373 K, U=-470 mV, I=0.01 nA. The inset

shows the tentative adsorption configuration of mono-deprotonated IPA superimposed on the high

resolution STM of chains, 19.7 nm × 19.7 nm, inset: U=-760 mV, I=0.01 nA, 3.7×1.9 nm2.

A tentative model for the adsorption structure IPA at 373 K on Cu(111) is shown

Figure 5-7. The carboxylate group anchors to the surface via an adatom while the intact

carboxylic group orients out of the surface and towards neighboring molecules.

Annealing the sample to 433 K gives rise to a distinctly different pattern of

striped structures. In Figure 5-8, the STM image illustrates single-feature rows that are

aligned in three domains, reflecting the symmetry of the substrate. At domain

boundaries between these striped regions, both amorphous and locally ordered

structures can be seen. Under different tip conditions (bias and/or tip termination), the

single-row structure appears as a double-stacked row. The periodicity measured

perpendicular to the row direction is 0.93 ± 0.20 nm and measurement along the row

is 0.48 ± 0.10 nm. These values are in agreement with value of row and intermolecular

distances for IPA solution-deposited onto Cu/Au(111).39 In the proposed model for the

IPA/Cu/Au(111) system, IPA is anchored to the substrate via two carboxylate groups,

leading to the formation of a highly crystalline stripe-like arrangements of molecules.

This is consistent with what we observed for deprotonated IPA on Cu(111).


Figure 5-8: STM image of the fully-deprotonated IPA layer after annealing to 433 K. Image

parameters: U=-1100 mV, I=0.2 nA, 49 nm × 49 nm. Inset: U=-1100 mV, I=0.1 nA. The inset is a

tentative model.

After annealing at 513 K for 160 minutes, XPS showed that the molecule was

fully decarboxylated (see Figure S1). Figure 5-9 shows the overview and detailed STM

images of the corresponding structures, which are disordered and scattered on the

surface. Small fragments of both zig zag and crinkled polymers, as well as the

macrocycle product, are observable in the STM image Figure 5-9. The diameter

(distance between diametrically opposed phenyl rings) of the macrocycle is found to

be 0.75 ± 0.16 nm, which is consistent with the formation of a covalent bond between

phenyls and is in agreement with previous reported values4, 40 and with molecular

mechanics calculations of the pore diameter (see Figure S5). Inspection of the bonding

geometries of the macrocycles suggests that the molecule may have experienced some

C-H bond scission, as threefold bonding geometries are apparent in the images;

activation at only the meta-sites would lead to isolated macrocycles, but the observed

macrocycles appear as substituents of extended oligomeric structures.


Figure 5-9: STM images showing the polymer resulting from IPA annealed at 513 K at different

magnification, (a) U=-1100 mV, I=0.1 nA, 80.9×80.9 nm2, (b) U=-140 mV, I=0.2 nA, 22.6×22.6 nm2.

5.5 DISCUSSION

PES reveals that IPA adsorbs partially deprotonated following deposition onto

Cu(111) at RT. Carboxylated molecules are known to partially/fully deprotonate on

different facets of copper at RT,23, 27-28, 41 and our result is consistent with these

previous findings. From the angular and beam energy dependence of the C 1s core

level components we find that IPA adopts a monopodal upright adsorption geometry,

wherein the molecule is bound to the surface via the carboxylate group. This is in

agreement with our expectations that the deprotonation/decarboxylation reactions are

metal-catalyzed, and consistent with a previous study that revealed that proximity of

the copper surface is required for the deprotonation of carboxylic acid.4 The carboxylic

groups in the molecules present in a second layer appear to be intact, as expected for -

COOH not in contact with the metal surface. Following desorption of the multilayer

with annealing at 373 K, the ratio of the carboxyl and carboxylate intensities becomes

1:1, and is accompanied by a 59% decrease of the phenyl intensity. The integrated

intensity of phenyl is only negligibly reduced following annealing to 488 K. This

corroborates that the multilayer desorbs after annealing at 373 K, leaving the

monolayer IPA on the surface and suggests that only a negligible amount of molecules

desorb after annealing the surface-bound layer at higher temperatures.

At the monolayer molecular coverage presented here, the deprotonation of IPA’s

two carboxylic groups occurs in two steps: the first carboxyl group deprotonates upon

adsorption of the IPA on the surface, whereas the second group deprotonates only after

annealing at 453 K, suggesting that the lack of proximity between the second


carboxylic group and the surface, which arises due to the molecular adsorption

geometry, increases the activation barrier for the second deprotonation. This second

deprotonation event is associated with a change in the adsorption configuration from

monopodal to bipodal geometry. The NEXAFS spectra demonstrate that this

reorientation of the molecule does not significantly change the inclination of the

molecule with respect to the surface. As has been observed previously,30 the

carboxylate groups maintain a slightly different orientation angle with respect to the

surface, likely to accommodate coordination to surface atoms. It should be noted that

the molecular coverage also has influence on the adsorption geometry, i.e., a flat

adsorption geometry is generally favorable at low coverages while more upright

standing geometries are reported for high coverages.42-43 This adsorption geometry in

turn affects the deprotonation reactions - our PES data collected from a low-coverage

surface show that the molecule is fully deprotonated at 300 K (see Figure S7).

We attribute the chain-like structure of the monodeprotonated molecules to a

motif in which one Cu adatom sits between the carboxylate groups of four IPA

molecules in each unit cell. A somewhat similar cloverleaf structure has been observed

for flat-laying 1,3,5-benzenetricarboxylic acid (trimesic acid, TMA) on Cu(110), with

a bright protrusion at the center indicative of formation of coordinative bonds between

four carboxylate ligands and one Cu adatom [Cu(TMA)4]n-.41 Important differences

arise in our study, since the molecules are oriented upright with respect to the surface.

The second deprotonation of the molecule creates a well-ordered, metal-coordinated

linear structure consistent with one observed previously.39

Decarboxylation of the IPA leads to covalent coupling of the molecules, but the

resulting oligomeric structures suffer from poor crystallinity and limited spatial

expansion. NEXAFS spectra acquired on the decarboxylated product indicate that the

phenyl rings may be inclined with respect to the surface; since we expect the oligomer

product to adsorb relatively flat on the surface, we attribute this apparent non-planar

adsorption to molecules trapped at defects and step edges, which may account for a

significant proportion of the retained molecular product after high-temperature

annealing and the accompanying desorption (c.f. the low signal to noise for the

NEXAFS scan of the decarboxylated sample in Figure 5-5a, which is consistent with

low molecular coverage relative to the previous annealing step.) Further optimization

of the experimental conditions may lead to increased spatial extent of the structures,


but is unlikely to improve the crystallinity due to the number of polymorphs allowed

by the precursor molecule, and by the possibility for additional activation via

dehydrogenation at C-H bonds.44

5.6 CONCLUSIONS

We found that isophthalic acid mono-deprotonates upon adsorption on Cu(111)

at room temperature. By collecting PE spectra at different angles and also at different

beam energies, we reveal the relative geometry of carboxyl group, carboxylate group

and the phenyl ring, and find that the molecule stands up via the surface-bound

carboxylate group upon adsorption. Annealing at 453 K causes full deprotonation of

the molecule, which re-orientates and anchors to the surface via both carboxylate

groups. NEXAFS confirms that the molecule is inclined from the surface in both

mono-deprotonated and fully-deprotonated forms, and that the inclination angle with

respect to surface is larger for the monopodal configuration compared to the bipodal

configuration. STM demonstrates the structure of molecules at different steps of

deprotonation. The tentative structural model for mono-deprotonated molecules

consists of four mono-deprotonated molecules anchor to the surface via carboxylate

group with one copper adatom siting between them.

Following annealing at 513 K, the molecule fully decarboxylates and forms C-

C bonds that define oligomeric structures on the surface. The results presented in this

work add to our understanding of the important balance between adsorbate-adsorbate

and substrate-adsorbate interactions as a key factor in the defining the pathway through

which molecular reactions proceed on a surface. In this case, we find successive

deprotonation steps for the meta-configured carboxylic groups. This suggests that the

number and relative geometry of the carboxylic acid groups in a molecule will affect

both adsorption and reactivity on surfaces, and should be carefully considered in

selecting molecules for on-surface materials design and synthesis.


5.7 REFERENCES









9663.




5. Abyazisani, M.; Bradford, J.; Motta, N.; Lipton-Duffin, J.; MacLeod, J.

Adsorption and Reactivity of Pyridine Dicarboxylic Acid on Cu (111). The Journal of



2017, 53 (56), 7872-7885.

7. Schmitt, T.; Hammer, L.; Schneider, M. A. Evidence for On-Site

Carboxylation in the Self-Assembly of 4, 4′-Biphenyl Dicarboxylic Acid on Cu (111).

The Journal of Physical Chemistry C 2016, 120 (2), 1043-1048.

8. Franke, M.; Marchini, F.; Zhang, L.; Tariq, Q.; Tsud, N.; Vorokhta, M.;

Vondracek, M.; Prince, K.; Rockert, M.; Williams, F. J. Temperature-Dependent

Reactions of Phthalic Acid on Ag (100). The Journal of Physical Chemistry C 2015,

119 (41), 23580-23585.

9. Martin, D.; Cole, R.; Haq, S. Creating a functionalized surface: The adsorption

of terephthalic acid onto Cu (110). Physical Review B 2002, 66 (15), 155427.

10. Lee, J.; Dougherty, D. B.; Yates, J. T. Phenyl species formation and

preferential hydrogen abstraction in the decomposition of chemisorbed benzoate on

Cu (110). The Journal of Physical Chemistry B 2006, 110 (20), 9939-9946.

11. Wühn, M.; Weckesser, J.; Wöll, C. Bonding and Orientational Ordering of

Long-Chain Carboxylic Acids on Cu(111): Investigations Using X-ray Absorption

Spectroscopy. Langmuir 2001, 17 (24), 7605-7612.

12. Tait, S. L.; Lim, H.; Theertham, A.; Seidel, P. First layer compression and

transition to standing second layer of terephthalic acid on Cu (100). Physical

Chemistry Chemical Physics 2012, 14 (22), 8217-8223.

13. Tekiel, A.; Prauzner-Bechcicki, J. S.; Godlewski, S.; Budzioch, J.; Szymonski,

M. Self-assembly of terephthalic acid on rutile TiO2 (110): Toward chemically

functionalized metal oxide surfaces. The Journal of Physical Chemistry C 2008, 112

(33), 12606-12609.

14. Frederick, B.; Ashton, M.; Richardson, N.; Jones, T. Orientation and bonding

of benzoic acid, phthalic anhydride and pyromellitic dianhydride on Cu (110). Surface

Science 1993, 292 (1-2), 33-46.

15. Pascal, M.; Lamont, C. L.; Kittel, M.; Hoeft, J. T.; Terborg, R.; Polcik, M.;

Kang, J.; Toomes, R.; Woodruff, D. P. Quantitative structural determination of the

high coverage phase of the benzoate species on Cu (1 1 0). Surface Science 2001, 492

(3), 285-293.

16. Woodruff, D.; McConville, C.; Kilcoyne, A.; Lindner, T.; Somers, J.; Surman,

M.; Paolucci, G.; Bradshaw, A. The structure of the formate species on copper


surfaces: new photoelectron diffraction results and sexafs data reassessed. Surface

Science 1988, 201 (1-2), 228-244.

17. Frederick, B.; Leibsle, F.; Haq, S.; Richardson, N. Evolution of lateral order

and molecular reorientation in the benzoate/Cu (110) system. Surface Review and

Letters 1996, 3 (04), 1523-1546.

18. Frederick, B.; Chen, Q.; Leibsle, F.; Lee, M.; Kitching, K.; Richardson, N.

Long-range periodicity in c (8× 2) benzoate/Cu (110): a combined STM, LEED and

HREELS study. Surface Science 1997, 394 (1-3), 1-25.

19. Perry, C.; Haq, S.; Frederick, B.; Richardson, N. Face specificity and the role

of metal adatoms in molecular reorientation at surfaces. Surface Science 1998, 409 (3),

512-520.

20. Chen, Q.; Perry, C. C.; Frederick, B. G.; Murray, P. W.; Haq, S.; Richardson,

N. V. Structural aspects of the low-temperature deprotonation of benzoic acid on

Cu(110) surfaces. Surface Science 2000, 446 (1), 63-75.

21. Lennartz, M. C.; Atodiresei, N.; Muller-Meskamp, L.; Karthauser, S.; Waser,

R.; Blugel, S. Cu-adatom-mediated bonding in close-packed benzoate/Cu (110)-

systems. Langmuir 2008, 25 (2), 856-864.

22. Dmitriev, A.; Spillmann, H.; Stepanow, S.; Strunskus, T.; Wöll, C.; Seitsonen,

A. P.; Lingenfelder, M.; Lin, N.; Barth, J. V.; Kern, K. Asymmetry Induction by

Cooperative Intermolecular Hydrogen Bonds in Surface‐Anchored Layers of Achiral

Molecules. ChemPhysChem 2006, 7 (10), 2197-2204.

23. Dmitriev, A.; Lin, N.; Weckesser, J.; Barth, J. V.; Kern, K. Supramolecular

Assemblies of Trimesic Acid on a Cu(100) Surface. The Journal of Physical Chemistry

B 2002, 106 (27), 6907-6912.

24. Sandoval, T. E.; Bent, S. F. Adsorption of Multifunctional Organic Molecules

at a Surface: First Step in Molecular Layer Deposition. 2013.

25. Fairley, N., CasaXPS Manual 2.3. 15: Introduction to XPS and AES. Casa

Software: 2009.

26. Gann, E.; McNeill, C. R.; Tadich, A.; Cowie, B. C.; Thomsen, L. Quick AS

NEXAFS Tool (QANT): a program for NEXAFS loading and analysis developed at

the Australian Synchrotron. Journal of Synchrotron Radiation 2016, 23 (1), 374-380.

27. Stepanow, S.; Strunskus, T.; Lingenfelder, M.; Dmitriev, A.; Spillmann, H.;

Lin, N.; Barth, J. V.; Wöll, C.; Kern, K. Deprotonation-Driven Phase Transformations

in Terephthalic Acid Self-Assembly on Cu(100). The Journal of Physical Chemistry

B 2004, 108 (50), 19392-19397.

28. Classen, T.; Lingenfelder, M.; Wang, Y.; Chopra, R.; Virojanadara, C.; Starke,

U.; Costantini, G.; Fratesi, G.; Fabris, S.; de Gironcoli, S.; Baroni, S.; Haq, S.; Raval,

R.; Kern, K. Hydrogen and Coordination Bonding Supramolecular Structures of

Trimesic Acid on Cu(110). The Journal of Physical Chemistry A 2007, 111 (49),

12589-12603.

29. Payer, D.; Comisso, A.; Dmitriev, A.; Strunskus, T.; Lin, N.; Wöll, C.; DeVita,

A.; Barth, J. V.; Kern, K. Ionic Hydrogen Bonds Controlling Two‐Dimensional

Supramolecular Systems at a Metal Surface. Chemistry – A European Journal 2007,

13 (14), 3900-3906.








32. Poon, H.; Tong, S. Focusing and diffraction effects in angle-resolved x-ray

photoelectron spectroscopy. Physical Review B 1984, 30 (10), 6211.

33. Smith, N.; Farrell, H.; Traum, M.; Woodruff, D.; Norman, D.; Woolfson, M.;

Holland, B. Photoelectron diffraction effects in core-level photoemission from Na and

Te atoms adsorbed on Ni (001). Physical Review B 1980, 21 (8), 3119.

34. Li, Z.; Wandlowski, T. Structure Formation and Annealing of Isophthalic Acid

at the Electrochemical Au (111)− Electrolyte Interface. The Journal of Physical

Chemistry C 2009, 113 (18), 7821-7825.

35. Han, B.; Li, Z.; Wandlowski, T. Adsorption and self-assembly of aromatic

carboxylic acids on Au/electrolyte interfaces. Analytical and Bioanalytical Chemistry

2007, 388 (1), 121-129.

36. Tait, S. L.; Wang, Y.; Costantini, G.; Lin, N.; Baraldi, A.; Esch, F.; Petaccia,

L.; Lizzit, S.; Kern, K. Metal− organic coordination interactions in Fe− terephthalic

acid networks on Cu (100). Journal of the American Chemical Society 2008, 130 (6),

2108-2113.

37. Zhang, W.; Nefedov, A.; Naboka, M.; Cao, L.; Wöll, C. Molecular orientation

of terephthalic acid assembly on epitaxial graphene: NEXAFS and XPS study.

Physical Chemistry Chemical Physics 2012, 14 (29), 10125-10131.


Vol. 25.






6919-6921.

41. Lin, N.; Dmitriev, A.; Weckesser, J.; Barth, J. V.; Kern, K. Real-Time Single-

Molecule Imaging of the Formation and Dynamics of Coordination Compounds.


42. Lee, A. F.; Wilson, K.; Lambert, R. M.; Goldoni, A.; Baraldi, A.; Paolucci, G.

On the coverage-dependent adsorption geometry of benzene adsorbed on Pd {111}: a

study by fast XPS and NEXAFS. The Journal of Physical Chemistry B 2000, 104 (49),

11729-11733.

43. Bader, M.; Haase, J.; Frank, K.-H.; Puschmann, A.; Otto, A. Orientational

phase transition in the system pyridine/Ag (111): A near-edge x-ray-absorption fine-

structure study. Physical Review Letters 1986, 56 (18), 1921.

44. Li, Q.; Yang, B.; Björk, J.; Zhong, Q.; Ju, H.; Zhang, J.; Cao, N.; Shi, Z.;

Zhang, H.; Ebeling, D.; Schirmeisen, A.; Zhu, J.; Chi, L. Hierarchical

Dehydrogenation Reactions on a Copper Surface. Journal of the American Chemical

Society 2018, 140 (19), 6076-6082.



Beam energy dependence PES, additional XPS data, additional STM images and

measurement of macrocycle diameter.

5.8.1 XPS data for IPA before STM imaging

Here, both intact (–COOH) and deprotonated (–COO–) groups are evident in the

spectrum, with the intact contribution 1.6 times larger than the deprotonated

contribution. This could arise from fully deprotonated and intact IPA molecules

coexisting on the surface, or singly deprotonated molecules coexisting with intact

ones. After annealing to 100°C the C 1s peak intensity decreases and the –COOH and

–COO– peaks become equal in magnitude. These changes indicate desorption of the

multilayer upon annealing leaving a monolayer of mono-protonated IPA on Cu(111)

at room temperature, while intact molecules existed only in the second layer.

Figure S5-10: C 1s spectra for IPA deposited onto Cu(111) at RT and annealed up to 240°C.

5.8.2 Orientation of singly-deprotonated molecules: beam-energy dependence

We additionally compared spectra acquired at different photon energies where

lower beam energies produce lower energy electrons which have a different electron

attenuation length.1-3 The dependence of the intensity of the different moieties with

beam energy shown in Figure S2 further supports the picture obtained from angular

dependence experiment for the singly-deprotonated molecule. The –COO– signal

normalized with respect to the whole C 1s region is increasing with increasing the


beam energy from 380 eV to 908 eV, this is contrast with the trend for phenyl and –

COOH group. Increasing the beam energy increases the corresponding kinetic energy

of the photoelectrons ejected from the surface, which in this energy range means the

electrons have a larger escape depth. This confirms again that –COO– signal, which

increases with the increasing beam energy, originates from beneath the phenyl and –

COOH group.

Figure S5-11: Apparent stoichiometry of IPA deposited at room temperature collected with different

beam energies: 380, 486, 650 and 908 eV.

5.8.3 Additional NEXAFS fitting data.

These plots show the peak areas corresponding to each of the components of

interest (carbon phenyl, carbon carboxylic/carboxylate, oxygen

carboxylic/carboxylate) for the half-deprotonated and fully-deprotonated samples.


Figure S5-12: The red data points (with error bars) indicate the component area. The blue line shows

the fit to the vector symmetry formula Stöhr 9.16a.4 The resulting inclination angles are reported in

each panel of the figure. (a) Half-deprotonated carbon phenyl, (b) half-deprotonated carbon

carboxylic/carboxylate, (c) half-deprotonated oxygen carboxylic/carboxylate, (d) deprotonated carbon

phenyl, (e) deprotonated carbon carboxylate, (f) deprotonated oxygen carboxylate.


5.8.4 STM images showing the easily perturbed structure during scanning

Figure S5-13: STM image of IPA annealed at 100°C, a) U=-480 mV, I=0.01 nA, 14.5×14.5 nm2, b)

U=-480 mV, I=0.01 nA, 19.7×19.7 nm2.

5.8.5 Measurement of macrocycle diameter

Figure S5-14: (a) STM image showing macrocycles and other oligomeric products. The blue line

indicates the location of the line profile shown in (b). Image parameters: U=-140 mV, I=0.2 nA, 4.4


nm × 4.4 nm. (b) The blue line is the profile indicated in (a). An indicative measurement of the pore

diameter, 0.76 nm, is shown; the values in the main manuscript comprise multiple measurements and

uncertainty propagation. At the top of the figure is a molecular mechanics (MMFF94) optimized

structure for the macrocycle produced via Avogadro, which has a diameter of 0.76 nm.

5.8.6 Estimation of coverage based on carbon to copper ratio

Figure S5-15: The carbon to copper ratio of IPA at room temperature and annealed at 180°C

5.8.7 Plot of the ratio of C1s intensity to Cu 2p as a function of annealing

temperature

Figure S5-16: The carbon to copper ratio of IPA from deprotonated to decarboxylated state.


5.8.8 Monitoring deprotonation of low coverage of IPA deposited on a cold

surface

A small coverage (<0.5 ML) of IPA was deposited into surface held at -120°C.

The sample was then annealed up to 10°C in increments of 10°C, and after that the

increment step was reduced to 5°C to monitor the deprotonation reaction. The

molecule is fully deprotonated at 35°C.

Figure S5-17: Monitoring of deprotonation of low coverage IPA.


References










Vol. 25.

Chapter 6: Cleaning up after the Party: Removing the Byproducts of On-surface Ullmann Coupling 117

Chapter 6: Cleaning up after the Party:

Removing the Byproducts of On-

surface Ullmann Coupling

In the two previous chapters, I have framed the decarboxylation reaction as a

“clean reaction” to achieve by-product free polymers. However, the results from the

decarboxylation reaction for both PDC and IPA molecules show obstacles that arise

from this clean reaction. As has been previously discussed, the unwanted halogen by-

product can be removed after Ullmann coupling. In this chapter the Ullmann coupling

of 1,4-dibrombenzene (dBB) has been investigated to fabricate poly-paraphenylene

(PPP) chains, which coexist with surface-bound bromine. The surface has been

exposed to a flux of atomic hydrogen at room temperature and remove the bromine

from the surface. This alternative approach can be applied to achieve halogen-free

polymers on surface.

XPS and STM (Omicron (Scienta-Omicron GmbH), QUT) have been employed

to study the sample before and after hydrogen etching. The results are under review

for publication.

118 Chapter 6: Cleaning up after the Party: Removing the Byproducts of On-surface Ullmann Coupling


Published paper




field of expertise;






academic unit, and




Cleaning up after the party: removing the byproducts of on-surface Ullmann

coupling: submitted in ACS NANO, ID: nn-2018-09581m


Maryam Abyazisani Conducted experiments, analysed data, and drafted manuscript. Signature

Jennifer MacLeod Revised manuscript, supervision.

Josh Lipton-Duffin Conceived the project, conducted experiments, analysed data,

revised manuscript, supervision.

RSC, Level 4, 88 Musk Ave, Kelvin Grove Qld 4059

Date Signature Name









6.1 ABSTRACT

Ullmann coupling is one of the most frequently employed methodologies for

producing π-conjugated surface-confined polymers. One unfortunate side product of

the reaction is the creation of metal halide islands formed from liberated halogen

atoms. Following the coupling reaction, these halide islands can account for a large

proportion of the substrate surface area, and thus inhibit domain growth and effectively

poison the catalyst. Here, we describe an efficient and reliable methodology for

removing the halogen byproduct at room temperature by etching with a beam of atomic

hydrogen; this action removes the halogen atoms in a matter of minutes, with minimal

impact to the polymer structure. We also find that it is possible under certain

circumstances to preserve the pre-etch epitaxy after removal of the halogen. This

finding provides a convenient and straightforward technique for addressing the most

oft-cited drawback of the on-surface Ullman coupling methodology, and provides

access to a previously inaccessible parameter space for these types of experiments.

6.2 INTRODUCTION

The field of surface-confined polymerization has been the subject of intense

study for the last ten years since its inception in the mid-late 2000s.1-6 The concept of

building new materials from the bottom-up, using rationally designed building blocks,

is an alluring approach to generate new materials that address challenges in electronics,

sensing, energy storage and other areas.7 The use of a flat surface as a template for the

material bypasses the entropic penalties against forming planar materials that are

inherent in solution synthesis. A wide variety of coupling methodologies have been

studied in this context, including Glaser-Hay coupling,8-9 Sonogashira coupling,10

Bergman cyclization,11 Schiff-Base coupling,12 and decarboxylative coupling,13-14 all

of which have been demonstrated to produce 1D or 2D materials on surfaces. However

the most popular methodology to date has been on-surface Ullmann coupling (Figure

6-1).15 Ullmann coupling is a particularly attractive approach as the substrate acts as a

catalyst for the coupling reaction, meaning that only moderate thermal inputs are

required to initiate the reaction. The reactants and products are readily identifiable both

by morphological characterizations by scanning tunnelling microscopy (STM), as well

as chemical characterizations like x-ray photoelectron spectroscopy (XPS).


Figure 6-1: On-surface Ullmann coupling of dibromobenzene and subsequent removal of bromine by

exposure to atomic hydrogen.

One drawback of surface-confined Ullmann reaction is that the byproducts,

halogen atoms abstracted from the aryl-containing building blocks, remain on the

surface after the reaction. These halogen atoms are often chemisorbed to the surface,

where they inhibit the diffusion of radicals,16-19 which in turn inhibits polymer

growth.20 The polymers grow surrounded by halogen networks,21 which effectively

block catalytic sites of the surface,22 preventing additional dehalogenation and

coupling events.18, 20 This problem is often exarcerbated by the tendency of small

building blocks to undergo dehalogentative desorption during the deposition process,23

leaving a surface that is far richer in halogen content than would be suggested purely

by the stoichiometry of the precursors.

There is therefore a concerted effort to either identify "clean" methodologies for

producing surface-confined polymers,13-14, 24 or to find reliable procedures for

removing the adsorbed halogen atoms after the polymerisation step from the substrate.

A recent publication by Stöhr and coworkers25 demonstrated the effect of annealing

Ullmann-coupled systems in a molecular hydrogen background: after a ~2 hour

annealing in a background of 10-7 mbar H2 the halogen content of this surface was

significantly reduced. The most likely mechanism is that the halogen is removed by

splitting of the hydrogen molecule to form and subsequently desorb as HBr.26 Here,

we report on a significantly more efficient method of using a beam of atomic hydrogen

to achieve halogen-free poly-paraphenylene (PPP) from 1,4-dibrombenzene (dBB) on

both Cu(111) and (110) after Ullmann coupling. Notably, the procedure may be

performed at room temperature, with a light annealing subsequent to halogen removal

to improve ordering on the surface.

6.3 EXPERIMENTAL

All experiments were performed in an ultrahigh vacuum chamber with a base

pressure of 10-11 mbar (Scienta-Omicron GmbH). The system consists of two

chambers, the first containing a dual-anode x-ray lamp (DAR 400) and a hemispherical


electron energy analyzer (iSphera). The second chamber contains a scanning tunneling

microscope (VT-AFM/XA), which was operated at room temperature for this study.

High resolution XPS core level spectra were collected in fixed analyzer transmission

mode using Al K- radiation with a pass energy of 20 eV and a step size of 0.1 eV.

Intensity ratios were computed using the area beneath full entire core level regions

(both peaks in the case of doublets), with Scofield sensitivity factors and the

empirically determined transmission function for the analyser. STM measurements

were acquired in constant current mode, with the quoted bias voltages measured with

respect to the tip.

Sample Preparation: 111 and 110-oriented single crystals of copper (Mateck

GmbH and MTI Corp) were cleaned by repeated cycles of Ar+ sputtering (1 keV) and

annealing (450°C) until no carbon or oxygen surface contaminants were detectable by

XPS. The 1,4-dibromobenzene (98%, Sigma Aldrich Corp) was deposited by allowing

a room temperature vapor of the molecules to leak into the chamber via a drift tube

pointed at the sample. Typical apparent pressures were in the mid 10-8 mbar range,

with depositions lasting 1 to 5 minutes. This corresponds to an estimated coverage of

1-10 Langmuir, though we have not corrected the pressure reading for

dibromobenzene from the default tuning for nitrogen. The substrates were annealed

with a resistive element embedded in the sample manipulator, and temperatures

controlled via a Eurotherm 2048 PID controller feeding back on a k-type thermocouple

mounted on the manipulator in proximity to the sample. Atomic hydrogen etching was

achieved with a hot-capillary-type thermal H2 cracker (EFM-H, Focus GmbH)

operated at 40 watts, with an inlet pressure of approximately 1 bar and an in-vacuum

pressure reading of 5E-7 mBar (equivalent N2 pressure as read by an ionization gauge).

The source was mounted at an angle of 52° with respect to the sample normal at a

working distance of 62.5 mm. The flux of hydrogen on the surface was estimated to

be (6±3)x1017 /s/m2 based on the erosion rate of an amorphous carbon film (see SI).

Density functional theory calculations: Gas-phase ab initio calculations were

performed at the B3LYP/6-31G(d,p) using Gaussian 09.27 The 1-dimensional PPP

conformation was determined by allowing all atoms to relax in a 1D periodic boundary

condition (PBC), with no geometrical constraints. Single point energy calculations

were performed from this relaxed geometry by incrementing the size of translation


vector, but without further relaxation. HOMO and LUMO energies were extracted

from the resultant energy calculation and are referenced to the vacuum level.

6.4 RESULTS

Cu(110): Figure 6-2 shows an STM image of dBB deposited on Cu(110) held at

200°C. The molecule dehalogenates upon adsorption and forms arrays of

organometallic (OM) chains. The structure depicted is well-known from previous

work; the phase appearing in Figure 6-2 corresponds to a saturated single monolayer

of dBB OM assemblies.28-29 The bright features in these arrays correspond to the metal

atoms separating each phenyl ring, and the smaller features are halogen atoms.

Chemically similar but structurally distinct features are observed when the experiment

is repeated using Cu(111) as the substrate (see below).


Figure 6-2: STM images of the organometallic assemblies resulting from dBB on Cu(110) (a: I=0.5

nA, U=-0.5 V, 98x98 nm2, b: I=0.5 nA, U = 0.5 V, 28x28 nm2, c: I=0.05 nA, -0.01 V, 10 x 10 nm2).

Annealing to 275°C converts the OM chains to PPP. On Cu(110) this results in

ordered arrays of PPP,28-29 with each polymer strand interdigitated with a row of

bromine. The polymers are epitaxially matched to the substrate, running along the <1

-1 2> direction (a subset of polymers may also run along <1 -1 0>, which has been


correlated with lower starting coverages).29-30 DFT calculation relating the spacing of

the phenyl rings as a function of lattice constant suggests that epitaxial matching may

drive significant modulation of the electronic properties of surface-confined polymers

(See Figure S1).

Exposing the surface to a flux of atomic hydrogen at room temperature

eventually completely removes the bromine from the surface, as seen in the

photoelectron spectra in Figure 6-3. The removal rate follows 1st-order kinetics, as

evident from the logarithmic decrease of the ratio of the Br 3p signal to C 1s with

respect to etch time, presented in Figure 6-4. No substantive change in the intensity of

the C 1s core level is observed after hydrogen etching, but a progressively increasing

shift towards higher binding energy is observed as the exposure continues. We

presume that the removal rate is tied both to the binding energy of the bromine to the

metal surface as well as the interaction between bromine and the organic species

themselves. Since there appears to be a higher cohesive interaction between polymers

and bromine on Cu(110) we expect the time required to remove Br to be substantially

shorter on Cu(111), due to the absence of forces associated with halogen-polymer

interaction. This is consistent with the results from our single etch experiments where

all the Br was removed on Cu(111) after just five minutes (see below).

Figure 6-3: Photoelectron spectroscopy of a) the carbon 1s and b) the bromine 3p core levels of dBB

on Cu(110) as a function of atomic hydrogen etching time.


Figure 6-4: Evolution of Br 3p:C 1s and C 1s:Cu 2p as a function of hydrogen etching time.

The removal of a substantial amount of Br from Cu(110) is correlated with a

dramatic change in the surface morphology, , as summarized in Figure 6-5. STM

images collected from successively etched surfaces show polymer features with a

progressively fuzzier appearance and a reduction in lateral cohesion and epitaxial

alignment. While pristine polymer surfaces (Figure 6-4a-c) show short oligomers

aligned either along surface <1,0> or <1,±1> directions, this ordering is progressively

lifted as the etching continues. By the time that all of the bromine has been removed

from the surface (corresponding to an etch time of 960 s) the surface is nearly

impossible to image with STM, with substantial tip-fouling and diffusion present at

room temperature, and an apparent loss of ordering in regions where imaging was

possible (see Figure 6-5d-f). We also note a progressive up-shifting in binding energy

of the C 1s core level with etching time, suggestive of introduction of sp3-type defects

in the polymer framework by superhydrogenation.31-33 We anticipate that this induces

an upward shift of the binding energy for those affected carbons, eg cyclohexene

manifests a higher energy (284.8 eV)34 than benzene (284.5 eV)35 on the same Pt(111)

surface. The original position of the carbon core level is restored by annealing for 5

minutes to 275°C. This change is accompanied by a ~20% reduction of intensity in the

C/Cu ratio, and a sharpening appearance of the STM images, and a reorientation of the

polymers to lie along the surface <10> azimuth (Figure 6-5g-i). This suggests


desorption of the sp3-like carbon species that may have arisen from interaction with

the atomic hydrogen, leaving only well-formed PPP structures on the surface. This

intuitive description is supported by deconvolution of the C 1s core level in Figure 6-6

(details of the fitting procedure are found in SI), where a sp3-like component (blue,

BE=285.2 eV) is found to grow as a function of etching time, and is subsequently

removed by annealing leaving only the original sp2-like component (yellow,

BE=284.8 eV). This spectroscopic interpretation along with the appearance of the

STM images post-annealing (Figure 6-6g-i) suggests that the polymers are mostly left

intact by the hydrogen etching treatment. The overall decrease in intensity of the C 1s

peak after annealing implies that the rehybridised carbon species formed during the

removal of the bromine are desorbed, but the majority of the surface sp2 carbon

character is recovered subsequent to heating at the synthesis temperature.

On surfaces with saturated monolayer coverage we find that the PPP lines

predominantly run along the <1 -1 0> direction, the close-packing orientation of the

Cu atoms on the 110 surface, and is generally known to be the more facile direction

for diffusion of surface species. The surface maintains this preferential direction over

regions larger than (100 nm2) (see SI), which is a substantial departure from the starting

geometry of short segments running along <1 -1 0> and <1 -1 2> in Figure 6-5 a-c.

This implies that the epitaxy observed upon growth may in some part be driven by a

co-operative assembly in which the halogens play a critical role in positioning and

aligning the polymers. Cu(110) surfaces with submonolayer starting coverage show a

preferential orientation of the polymers along the <1 -1 0> direction, but the dense

arraying observed for the monolayer surfaces is not present, and in certain cases entire

islands of polymer are observed to exhibit a diffusive motion (see SI).


Figure 6-5: STM images of Cu(110) surface after growth of PPP polymers from DBB precursor (ac),

etching procedure (d-f), and after final annealing (h-j). Image dimensions are identical by column, and

a z-height colour scale is given beside each image. a, c) I=630 pA, U=-1.26 V, b)I=630 pA, U=-280

mV , d) I=50 pA, U=-1.14V, e, f) I=100 pA, U=-180 mV, g,i) I=20 pA, U=-980 mV, h) I=100 pA,

U=-300 mV.


Figure 6-6: C1s XPS as a function of atomic hydrogen etching. a) After deposition, b) After H-etching

for 16 minutes, c) After annealing to 275°C.

Cu(111): As with Cu(110), the organometallic assembly on Cu(111) grows

epitaxially in very large defect-free domains. It consists of bright chains spaced by

6.78±0.05 Å along the crystal <3 -2 -1> directions (substrate <-1 2> directions), which

corresponds to an epitaxial spacing of √7𝑎 = 6.773 Å, where a=2.56 Å is the nearest-

neighbour spacing on Cu(111). This provides a reasonably close match to the expected

gas-phase dimension of the OM structure (6.74 Å), and is qualitatively similar to arrays

of OM chains on Cu(110) at ~0.5 ML coverage. On Cu(110) the chains kink every five

units to relieve internal strain caused by the epitaxial mismatch.3 By contrast no

kinking is observed on Cu(111), consistent with adsorption along a substrate direction

that yields a periodicity very close to that predicted for the gas phase. The chains on

Cu(111) pack into dense arrays described by epitaxy matrices of the form

(−1 25 2

),


with very large domains reflecting mirror symmetry along sample <-1 1 0>

azimuths simultaneously present on the surface. Next to the arrays of OM lines we

observe regions bearing a √3×√3R30° lattice (left of Figure 6-7c), which is the

expected arrangement for Br on Cu(111).36 As on Cu(110), we also observe features

consistent with Br atoms between every OM chain (see SI),3, 29 though these are not

visible at all STM biases. Together these regions of Br amount to substantially more

Br on the surface than would be expected by the stoichiometry of the molecule,

supporting the hypothesis of dehalogenative desorption during the preparation of the

surface.


Figure 6-7: STM micrographs of organometallic chains on Cu(111) (: 100 x 100 nm2, I = 10 pA, U = -

0.5 V, b: 39x39 nm, I=5 pA, U=-100 mV, c: 10 x 10nm2, I = 20 pA, U=-1.0V).


Figure 6-8: STM images of dBB on Cu(111) annealed for 5 minutes at 275°C, with local spatial

characteristics. a) I=10 pA, U=-300 mV, 200 x 200 nm2 b) I= 0.1nA and U= -250 mV, 7.2 x 7.2 nm2,

c) I=0.2 pA, U = -710 mV, 50x50 nm2.

Unlike on the 110 surface (see prior work),3, 28-29 the polymers on Cu(111) are

generally not ordered on the surface. After heating to 275°C disordered aggregates of

‘spaghetti’ are found, with patches of organometallic chains still present on the surface,

as shown in Figure 6-8a. In close vicinity of the OM island edges we find morphologies

similar to those in Figure 6-8b, where individual nucleate from the ends of OM lines

and split off from the edge of the islands, leading to partially-converted OM chains,

with polymers nucleating out of the ends of OM chains, and regions of Br-√3×√3

between the polymer and OM regions. This image suggests that the conversion of

organometallic to polymer is non-dissociative, with the reaction progressing step-wise

from the end of the OM, ejecting the metal atoms from the OM without significantly

disrupting the chains. In this localized region the conversion to polymer is

accompanied with a reorientation along a new surface direction.

However, most of the non-OM surface comprises disordered and meandering

polymers with no evidence of ordering, as shown in Figure 6-8c, with the localized

regions of order between Br, OM and polymer presented in Figure 6-8b being the

exception rather than the norm. . This suggests that once formed the mediating action

of bromine atoms found on the Cu(110) surface plays no significant role on Cu(111).


Figure 6-9: a) Carbon 1s and b) bromine 3p core levels during etching of dBB on Cu(111).

The removal rate of halogen from Cu(111) is substantially faster than on

Cu(110), with complete removal of the bromine occurring after just 5 minutes of

atomic hydrogen etching, as shown in the XPS spectra in Figure 6-9. We note an slight

offsets in both the Br 3p and C 1s core levels when comparing spectra taken on the

two different surfaces, which are summarized in xx. In the case of C1s, this can

plausibly correspond to differences in workfunction between the 110 and 111 surfaces;

the 110 surface is known to have a ~400 meV shallower workfunction,37 which should

correspond to higher observed binding energies for a physisorbed species.

Table 6-1: Summary of binding energies of carbon and bromine core levels on Cu(110) and Cu(111)

surfaces.

The surface morphology is largely unchanged from the as-polymerised sample,

with strings of PPP following random directions on the surface with no sign of

epitaxial ordering. Further images of the polymer phase post hydrogen etching can be

Cu(111) Cu(110)

C 1s (sp3) 285.0 285.2

C 1s (sp2) 284.6 284.8

Br 3p 3/2 182.3 182.2


found in the SI. We also note that there is no apparent shifting of the carbon 1s core

level after complete removal of the bromine from the 111 suface. It is presumed that

the etching time is not long enough to induce defects on the scale of those observed on

Cu(110) and as such the perceived change to the PPP on the surface is vanishingly

small.

6.5 DISCUSSION

Our STM images reveal that the PPP chains show a strong degree of epitaxial

alignment to the <1 -1 0> direction (surface <1 0> direction) on Cu(110) subsequent

to bromine removal and annealing. This suggests that the domain orientations along

<1 -1 2> (surface <1 1> direction), cf Figure 6-5a observed for monolayer surfaces

subsequent to deposition are driven largely by cooperative assembly effects with

halogens, and that the ordering observed in arrays of products produced by Ullmann

coupling is significantly influenced by the interaction of the halogen byproducts with

the polymers. This phenomenon has been implied by previous works where the species

of halogen (i.e. iodine, bromine, chlorine) is found to have a direct relation to the

topology of the resultant surfaces.30 However, it is clear that this is not a result of

chemical bonding, as the XPS signature of each of the polymers produced by this

method is identical: no halogen-carbon association is discernible in the C 1s or Br 3p

core levels, and as such we propose that the interaction proceeds a modulation of the

local potential energy landscape surrounding the surface containing each halogen-

monomer triplet; as nucleation and growth of the organometallics proceeds, there is

apparently a net attraction between the detached halogen atoms and their parent

molecule that drives the assembly based on the preferred adsorption site for each

species. This produces a cooperative assembly effect that drives self-assembly of the

islands of alternating organometallic chains and rows of bromine. This effect must

persist during the conversion to polymer, which as illustrated above, is nucleated at

the end of organometallic chains and proceeds step-by-step down each of the rows,

with surrounding halogens driving the epitaxy into the new energetic minimum for the

PPP and Br system. The complete removal of the bromine from the surface suppresses

the attractive interactions between adjacent chains and releases the PPP from its

epitaxial constraints, and the alignment observed on Cu(110) is likely driven by

diffusion considerations. As suggested by our calculations, the straining of the PPP

likely also produces an electronic modification to the molecular orbitals, as the strain-


induced gap modification will be released through this process. Furthermore, it appears

that the energetic input required to remove the halogen from the 110 surface is larger

than for the 111 surface, implying local increase of adsorption energy between halide

and substrate that is enhanced by the presence of the polymer.

6.6 CONCLUSION

In conclusion, we found that room temperature etching with atomic hydrogen

effectively removes the byproduct of Ullmann dehalogenation of dBB, surface-bound

bromine. Prior to etching, the bromine aligns in interstitial rows between the PPP lines.

The pre-etch surface is highly ordered and clearly adheres to epitaxy. Following

etching, when no trace of Br can be detected on the surface, the PPP polymers are

observed to exist in disordered clumps. These results suggest that the surface-bound

bromine drives the ordering of the PPP polymer, and that this ordering is compromised

by removal of the bromine. We further found that the slight epitaxial strain induced by

commensurability may have a significant effect on the HOMO-LUMO gap energy. It

may therefore be instructive to interrogate the orbital energies by spectroscopic

methods (i.e scanning tunnelling spectroscopy, ultraviolet photoemission

spectroscopy, inverse photoemission spectroscopy) to correlate the measured gap with

strain effects induced by surface commensurability; since this commensurability is

released by atomic hydrogen etching, we should presumably be able to measure an

electronic difference between pre- and post-etch PPP. This will be the focus of a future

study.


6.7 REFERENCES

1. McCarty, G. S.; Weiss, P. S. Formation and manipulation of protopolymer

chains. Journal of the American Chemical Society 2004, 126 (51), 16772-16776.

2. Grill, L.; Dyer, M.; Lafferentz, L.; Persson, M.; Peters, M. V.; Hecht, S. Nano-

architectures by covalent assembly of molecular building blocks. Nat Nanotechnol

2007, 2 (11), 687-691.





8198.

4. Lackinger, M. On‐surface polymerization–a versatile synthetic route to two‐dimensional polymers. Polymer International 2015, 64 (9), 1073-1078.

5. Franc, G.; Gourdon, A. Covalent networks through on-surface chemistry in

ultra-high vacuum: state-of-the-art and recent developments. Physical Chemistry

Chemical Physics 2011, 13 (32), 14283-14292.

6. Abyazisani, M.; Jayalatharachchi, V.; MacLeod, J., 2.14 - Directed On-Surface

Growth of Covalently-Bonded Molecular Nanostructures. In Comprehensive

Nanoscience and Nanotechnology (Second Edition), Andrews, D. L.; Lipson, R. H.;

Nann, T., Eds. Academic Press: Oxford, 2018; Vol. 2, pp 299-326.

7. Colson, J. W.; Dichtel, W. R. Rationally synthesized two-dimensional

polymers. Nature Chemistry 2013, 5, 453.

8. Zhang, Y.-Q.; Kepcija, N.; Kleinschrodt, M.; Diller, K.; Fischer, S.;

Papageorgiou, A. C.; Allegretti, F.; Björk, J.; Klyatskaya, S.; Klappenberger, F.;

Ruben, M.; Barth, J. V. Homo-coupling of terminal alkynes on a noble metal surface.

Nature Communications 2012, 3, 1286.

9. Bakker, A.; Gao, H. Y.; Fuchs, H., Glaser–Hay Coupling on Surfaces A2 -

Wandelt, Klaus. In Encyclopedia of Interfacial Chemistry, Elsevier: Oxford, 2018; pp

272-284.

10. Sanchez-Sanchez, C.; Orozco, N.; Holgado, J. P.; Beaumont, S. K.; Kyriakou,

G.; Watson, D. J.; Gonzalez-Elipe, A. R.; Feria, L.; Fernandez Sanz, J.; Lambert, R.

M. Sonogashira cross-coupling and homocoupling on a silver surface: chlorobenzene

and phenylacetylene on Ag (100). Journal of the American Chemical Society 2015,

137 (2), 940-947.

11. de Oteyza, D. G.; Gorman, P.; Chen, Y.-C.; Wickenburg, S.; Riss, A.;

Mowbray, D. J.; Etkin, G.; Pedramrazi, Z.; Tsai, H.-Z.; Rubio, A. Direct imaging of

covalent bond structure in single-molecule chemical reactions. Science 2013,

1238187.

12. Weigelt, S.; Busse, C.; Bombis, C.; Knudsen, M. M.; Gothelf, K. V.; Strunskus,

T.; Wöll, C.; Dahlbom, M.; Hammer, B.; Lægsgaard, E. Covalent interlinking of an

aldehyde and an amine on a Au (111) surface in ultrahigh vacuum. Angewandte

Chemie 2007, 119 (48), 9387-9390.




9663.






2017, 53 (56), 7872-7885.









18. Fan, Q.; Wang, T.; Liu, L.; Zhao, J.; Zhu, J.; Gottfried, J. M. Tribromobenzene

on Cu (111): Temperature-dependent formation of halogen-bonded, organometallic,

and covalent nanostructures. The Journal of Chemical Physics 2015, 142 (10), 101906.

19. Björk, J.; Hanke, F.; Stafström, S. Mechanisms of Halogen-Based Covalent

Self-Assembly on Metal Surfaces. Journal of the American Chemical Society 2013,

135 (15), 5768-5775.

20. Fan, Q.; Wang, C.; Liu, L.; Han, Y.; Zhao, J.; Zhu, J.; Kuttner, J.; Hilt, G.;

Gottfried, J. M. Covalent, organometallic, and halogen-bonded nanomeshes from

tetrabromo-terphenyl by surface-assisted synthesis on Cu (111). The Journal of





Nano 2017, 12 (1), 74-81.




Journal 2016, 22 (17), 5937-5944.




24. Abyazisani, M.; Bradford, J.; Motta, N.; Lipton-Duffin, J.; MacLeod, J.

Adsorption and Reactivity of Pyridine Dicarboxylic Acid on Cu (111). The Journal of


25. Tran, B. V.; Pham, T. A.; Grunst, M.; Kivala, M.; Stohr, M. Surface-confined

[2+ 2] Cycloaddition Towards One-dimensional Polymers Featuring Cyclobutadiene

Units. Nanoscale 2017, 9 (46), 18305-18310.

26. Bronner, C.; Björk, J.; Tegeder, P. Tracking and Removing Br during the On-

Surface Synthesis of a Graphene Nanoribbon. The Journal of Physical Chemistry C

2015, 119 (1), 486-493.

27. Frisch, M. J.; Trucks, G. W.; Schlegel, H. B.; Scuseria, G. E.; Robb, M. A.;

Cheeseman, J. R.; Scalmani, G.; Barone, V.; Mennucci, B.; Petersson, G. A.;

Nakatsuji, H.; Caricato, M.; Li, X.; Hratchian, H. P.; Izmaylov, A. F.; Bloino, J.;

Zheng, G.; Sonnenberg, J. L.; Hada, M.; Ehara, M.; Toyota, K.; Fukuda, R.; Hasegawa,

J.; Ishida, M.; Nakajima, T.; Honda, Y.; Kitao, O.; Nakai, H.; Vreven, T.; J. A.

Montgomery, J.; Peralta, J. E.; Ogliaro, F.; Bearpark, M.; Heyd, J. J.; Brothers, E.;

Kudin, K. N.; Staroverov, V. N.; Keith, T.; Kobayashi, R.; Normand, J.; Raghavachari,

K.; Rendell, A.; Burant, J. C.; Iyengar, S. S.; Tomasi, J.; Cossi, M.; Rega, N.; Millam,

J. M.; Klene, M.; Knox, J. E.; Cross, J. B.; Bakken, V.; Adamo, C.; Jaramillo, J.;

Gomperts, R.; Stratmann, R. E.; Yazyev, O.; Austin, A. J.; Cammi, R.; Pomelli, C.;


Ochterski, J. W.; Martin, R. L.; Morokuma, K.; Zakrzewski, V. G.; Voth, G. A.;

Salvador, P.; Dannenberg, J. J.; Dapprich, S.; Daniels, A. D.; Farkas, O.; Foresman, J.

B.; Ortiz, J. V.; Cioslowski, J.; Fox, D. J. Gaussian 09, Revision B.01; Gaussian, Inc.:

Wallingford CT, 2010, 2010.



(5), 592-597.

29. Vasseur, G.; Fagot-Revurat, Y.; Sicot, M.; Kierren, B.; Moreau, L.; Malterre,

D.; Cardenas, L.; Galeotti, G.; Lipton-Duffin, J.; Rosei, F. Quasi one-dimensional band

dispersion and surface metallization in long-range ordered polymeric wires. Nature

Communications 2016, 7, 10235.




31. Rauls, E.; Hornekær, L. Catalyzed routes to molecular hydrogen formation and

hydrogen addition reactions on neutral polycyclic aromatic hydrocarbons under

interstellar conditions. The Astrophysical Journal 2008, 679 (1), 531.

32. Skov, A. W.; Andersen, M.; Thrower, J. D.; Jørgensen, B.; Hammer, B.;

Hornekær, L. The influence of coronene super-hydrogenation on the coronene-

graphite interaction. The Journal of Chemical Physics 2016, 145 (17), 174708.

33. Sánchez-Sánchez, C.; Martínez, J. I.; Ruiz del Arbol, N.; Ruffieux, P.; Fasel,

R.; López, M. F.; de Andres, P. L.; Martín-Gago, J. Á. On-Surface Hydrogen-Induced

Covalent Coupling of Polycyclic Aromatic Hydrocarbons via a Superhydrogenated

Intermediate. Journal of the American Chemical Society 2019.

34. Rodriguez, J.; Campbell, C. T. Cyclohexene adsorption and reactions on clean

and bismuth-covered Pt (111). Journal of Catalysis 1989, 115 (2), 500-520.

35. Zhang, R.; Hensley, A. J.; McEwen, J.-S.; Wickert, S.; Darlatt, E.; Fischer, K.;

Schöppke, M.; Denecke, R.; Streber, R.; Lorenz, M. Integrated X-ray photoelectron

spectroscopy and DFT characterization of benzene adsorption on Pt (111), Pt (355)

and Pt (322) surfaces. Physical Chemistry Chemical Physics 2013, 15 (47), 20662-

20671.

36. Inukai, J.; Osawa, Y.; Itaya, K. Adlayer structures of chlorine, bromine, and

iodine on Cu (111) electrode in solution: In-situ STM and ex-situ LEED studies. The

Journal of Physical Chemistry B 1998, 102 (49), 10034-10040.

37. Gartland, P.; Berge, S.; Slagsvold, B. Photoelectric work function of a copper

single crystal for the (100),(110),(111), and (112) faces. Physical Review Letters 1972,

28 (12), 738.



6.8.1 Fitting algorithm for carbon 1s core level

It was assumed that the lineshape for the C 1s core level spectrum in PPP/Cu

surfaces could be modeled by a single sp2-like antisymmetric peak. It was further

assumed that sp3-type defects induced by prolonged hydrogen exposure could be

modeled by the addition of a single additional peak at 300-800 meV higher binding

energy with a symmetric lineshape, as has been previously published for pure sp2 and

sp3 forms of carbon (graphite and diamond).1 To determine the position/width for the

sp3 peak we fit the spectrum from the sample corresponding to the maximum hydrogen

exposure time (960s) allowing the intensity, width, and position of the new peak to

vary as free parameters, while fixing all parameters save the amplitude of the original

sp2 peak. The resultant peak was then used along with the original sp2 peak to fit the

entire dataset, keeping all parameters save amplitude fixed.

6.8.2 Estimated flux atomic hydrogen source and etching rates

An estimate of the flux of the atomic hydrogen source was made by etching an

amorphous C:H film, following the procedure described previously2. A commercial

sputter coater <Leica EMACE600> was used to produce a nominal 10 nm carbon film

on a 10x10 mm silicon chip. The film thickness was measured by ellipsometry to be

t=11.9 nm in thickness.

The film was loaded into our UHV system and exposed to cycles of atomic

hydrogen etching at a substrate temperature of 325°C, followed by photoelectron

spectroscopy to determine the removal rate of the carbon film by the change in

intensity of the C 1s and Si 2p core levels. The time required to remove the entire film

(ie – to reduce the C 1s signal to zero) was linearly extrapolated from these XPS

measurements, and the flux rate was computed by 𝜙 =𝑡𝜌𝑛

𝜏𝑌, where n is the number

density of the aC:H film (9.2±0.9)E28 atoms/m3, and Y=(2.0±0.7)E-2 is the

empirically determined etch yield for the carbon film on silicon at this temperature.

The results of this procedure are shown below in Figure S6-10.

Our analysis produces an estimated flux of =(6±3)E17 /s/m2, a figure

substantially lower than the one described in reference 1, but entirely reasonable given

the longer working distance for our source. With this empirically determined measure


of the flux we are able to calculate an etching yield of (4±2)E-3 for bromine atoms

interdigitated with dense polymers on Cu(110), assuming a filled polymer surface of

the type described in prior work,3 which contains one bromine atom per unit cell in a

(4 1

−1 1) reconstruction.

We note that the assumption of a linear etch rate for Br from polymer-coated

Cu(110) is inconsistent with our observations in the manuscript, and as such the

uncertainty on the yield may well be larger than reported above. The substantially

faster etch rate on Cu(111) could be explained by the lack of regular interdigitation of

polymers with bromine on this surface, and that the additional energy due to the

observed halogen-polymer bonding on Cu(110) decreases the effectiveness of the

atomic hydrogen beam.

Figure S6-10: Summary of XPS experiments for etched aC:H film. Presented are survey spectra (top),

high resolution of C 1s (left inset, top) and Si 2p (left inset, bottom). At the bottom right, the intensity

vs time trend is shown for both these core levels, along with the extrapolated linear fit.


6.8.3 Long-range domain-orientation stability of monolayer H-etched polymers

on Cu(110)

Figure S6-11: Long-range image (100×100 nm2) of Cu(110) surface following hydrogen etching

treatment and annealing. The surface presents a near 100% orientation of the polymers along the <1-

10> direction. I=50 pA, U=-340 mV

6.8.4 Low surface coverages of dBB on Cu(110)

On surfaces where the starting coverage is substantially lower than 1 monolayer

the hydrogen etching treatment still removes the bromine, but this is accompanied with

a substantial loss of order. The lines maintain a preference for alignment along the [1-

10] direction, but are substantially less coherent than when the halogen is present. This

lends credence to the hypothesis that the halogens play an important part in the

cooperative assembly of the ordered structures previously observed on this facet.

Figure S6-12: (a) Wide and (b) close range STM images of PPP chains on Cu(110) after completely

removing bromine from a submonolayer surface, U=-1V, I=0.005nA, 73.7×73.7 nm2, U=-1.8V,

I=0.002nA, 25.3×25.3 nm2.


6.8.5 Additional images of dBB-derived polymer on Cu(111)

The figures at left show the details of the polymer surface grown on the 111 facet

of copper. There is substantial disorder present, but none further distinguishable from

the appearance of the surface prior to hydrogen etching. Because the required exposure

time to bromine removal is minimal, it is presumed that very little damage (if any) to

the polymers is introduced.

Figure S6-13: STM images of atomic hydrogen-etched the PPP on Cu(111). I= 0. 1 nA, U= -1.5 V,

100×100 nm2, and I=0.01 nA, U=-0.05V, 11×11 nm2.

6.8.6 Diffusion of PPP domains on Cu(110)

On Cu(110), partial removal of bromine from submonolayer surfaces by H

etching followed by light annealing results in surfaces like those shown in manuscript

Figure 6-7a. Islands of PPP chains oriented along the [1 -1 0] direction are dominant

on this surface, but their appearance is ragged and their morphology is variable scan-

to-scan. The islands of PPP undergo a collective motion back and forth along the [1-1

1 1 0

1 0 1

1 1 0

1 0 1

a

b


0] axis, either as a result of random thermal motion at room temperature, or by effect

of the tip scanning. To illustrate this effect, we have included a movie comprising 23

sequentially acquired STM images of the same area of the surface (subject to

instrumental drift).

The imaging parameters for each frame are U=-1.94V, I= 5 pA, at a nominal 50

x 50 nm2 field of view, and each image is acquired over a period of 450 seconds. Two

frames from the movie are presented below, with indications of the apparent motion

of features.

Figure S6-14: Two STM images of PPP on Cu(110) after partial bromine removal, collected

sequentially over a period of 450 seconds (each). The white circles are meant as a guide to the eye to

indicate changes in the PPP island morphology scan-to-scan.

References

1. Poirier, D.; Weaver, J. Carbon (as graphite, buckminsterfullerene, and

diamond) by XPS. Surface Science Spectra 1993, 2 (3), 232-241.

2. Schwarz-Selinger, T.; von Keudell, A.; Jacob, W. Novel method for absolute

quantification of the flux and angular distribution of a radical source for atomic

hydrogen. Journal of Vacuum Science & Technology a-Vacuum Surfaces and Films

2000, 18 (3), 995-1001.

3. Vasseur, G.; Fagot-Revurat, Y.; Sicot, M.; Kierren, B.; Moreau, L.; Malterre,

D.; Cardenas, L.; Galeotti, G.; Lipton-Duffin, J.; Rosei, F.; Di Giovannantonio, M.;

Contini, G.; Le Fevre, P.; Bertran, F.; Liang, L. B.; Meunier, V.; Perepichka, D. F.

Quasi one-dimensional band dispersion and surface metallization in long-range

ordered polymeric wires. Nat Commun 2016, 7, 10235.

Chapter 7: Conclusions 143

Chapter 7: Conclusions

7.1 CONCLUSIONS

In this thesis, I presented two approaches for the formation of surface-confined

nanostructures which are devoid of by-products. The first approach was the

investigating the promising “clean” reaction called decarboxylative reaction of two

monomers -with nitrogen and without nitrogen- as a which produces volatile by-

products, and the second approach focused on removing the chemisorbed halogen by-

products produced during Ullmann coupling using atomic hydrogen etching. In this

regard, three molecules has been selected, namely 3,5 pyridine dicarboxylic acid,

isophthalic acid and 1,4-dibrombenzene. The adsorption configuration and reaction for

PDC and IPA were characterized by synchrotron-based photoemission spectroscopy

and near edge x-ray adsorption fine structure techniques, while scanning tunnelling

microscopy and x-ray photoelectron spectroscopy has been employed to study both

dBB and IPA.

PDC and IPA were particularly interesting due to the same relative geometry of

two carboxylic acids in both molecule. The only difference between them was the

presence of nitrogen in the aromatic ring of PDC. The introduction of nitrogen into the

aromatic ring is an elegant way of tuning the electronic properties of the resultant

structure, although the results suggest that retaining the nitrogen in a polymeric

product may be difficult.

In the case of PDC, it was not possible to study the resultant product. This was

because of decarboxylation and molecular fragmentation, through C−N bond scission,

are competitive processes. PDC is partially deprotonated upon adsorption and adsorbs

in two different configurations on Cu(111): (1) via the nitrogen lone pair and (2) via a

carboxylate group. Following annealing at 100°C, which causes further deprotonation

of the molecules, N 1s SRPES indicates that the molecules shift away from a lone-

pair-driven adsorption geometry. With full deprotonation, the molecules assume a

lower inclination with respect to the surface, as expected for the bipodal adsorption

that orients both carboxylate groups into the surface. The activation energy of the

decarboxylation reaction has been studied by means of continuous observation of the

C 1s core level during annealing around decarboxylation temperature. The extracted

144 Chapter 7: Conclusions

activation energy of (1.93 ± 0.17) eV was higher than calculated activation energy for

decarboxylation of a planar aromatic molecule. The difference may be an indication

of the influence of adsorption geometry on the reaction barrier. This finding

demonstrates the importance of molecule-surface interactions in determining the

adsorption geometry of a molecule on the surface. The induced tilted adsorption

geometry accordingly dictates that higher temperature (and maybe for longer time) as

a prerequisite to trigger the desired reaction. Here, both the nitrogen lone pair and the

carboxylate groups influence the adsorption geometry of the molecule, suggesting for

fabrication of polymeric product containing nitrogen a careful design of building block

is imperative.

The synthesis of poly-metaphenylene polymer has been studied using IPA

precursor. IPA partially deprotonated upon adsorption on Cu(111). Data from

collecting PES at different angles indicate that molecule anchored to the surface via

deprotonated carboxylic group. These results have further been supported by using

different beam energy for the collection of PES. Following fully deprotonation of the

molecule at higher temperature the molecule reoriented and anchored to the surface

via both deprotonated carboxyl groups in a bipodal configuration. NEXAFS confirmed

the absolute inclination of the molecule for mono-deprotonated and fully deprotonated

species. Additionally, STM illustrated a poly-meta-phenylene polymer formed after

annealing at 215°C. Although STM reveals that decarboxylative coupling has

occurred, the spaghetti-like products suffer from lack of order and the coverage was

very poor. Considering the final products, decarboxylative coupling may not be the

decisive approach for fabrication of high-quality structures. It would be interesting to

employ various different building blocks -including larger molecules- since larger

molecules may adsorb flat and consequently influence the different steps of the

reaction e.g. the molecule may deprotonate in one step. Additionally, number and

relative geometry of the carboxylic acid groups in a molecule may affect both

adsorption and reactivity on surfaces and may be the focus of a future study.

In the case of dBB, a beam of H+ atoms has been used to achieve halogen-free

poly-para-phenylene from dBB on both Cu(111) and Cu(110) after Ullmann coupling.

STM showed that halogens align in interstitial rows between the PPP lines prior to

hydrogen etching. Further support of presence of halogens in the surface provided by

XPS of Br 3p core level. Dosing H+ to the samples effectively removed halogens from

Chapter 7: Conclusions 145

surface and STM showed no trace of Br on the surface. This claim was further

substantiated by the results from XPS measurements of Br 3p core level spectra that

showed Br peak completely vanished. However, STM revealed that the hydrogen

etching changed the highly ordered PPP lines to into disordered clumps. There are two

possible explanation for this: (a) Hydrogen etching not only removed the halogen but

might have had a negative effect on the polymer as well (for example, led to defective

structure in the aromatic ring), (b) the high ordering and epitaxy of PPP lines has been

dictated by halogens on the surface The platform developed here is promising to tackle

the oft-cited halide by-product challenge in surface-confined polymers fabricated by

the Ullmann reaction and helps to provide a fundamental understanding of the halogen

effect on the ordering of the products.

7.2 OUTLOOK

The main part of the presented thesis focused on decarboxylation reaction.

Although it successfully illustrated the decarboxylation of PDC and IPA with special

focus on the influence of adsorption geometry of these molecules on the reaction

pathway, of course, there are still important open questions. What would be the

molecular adsorption configuration, reaction barrier etc. for different polyfunctional

carboxylic acid precursors? For example, would terephthalic acid, which has two

carboxylic acids on the para position, adsorb flat on the surface due to the relative

geometry of the carboxylic acid groups? Would this improve the structure of the

resultant product, e.g. increase the regularity of the product? How about molecules

with two or three nitrogen in the aromatic ring instead of just one nitrogen, would for

example three nitrogens in 1,3,5-triazine molecule drives the molecule to adsorb flat?

Additionally, molecules with bigger central group -e.g. 1,3,5-triphenylbenzene or

phenelene- can be a good candidate for this reaction since the larger molecules may be

more likely to adsorb flat on the surface. This can be the focus of future work.

In the case of the Ullman reaction, it would be interesting to examine the

universality of removing halogens by atomic hydrogen etching for molecules with

three halogens in order to synthesise by-product-free 2D polymers. Furthermore,

precursors with different halogens and substrate may be the focus for a future study.

Appendices 147

Appendices

Appendix A

Comparison of dehalogenation temperature for different molecule on different

surfaces

The following table provides a simple compendium of dehalogenation

temperatures for molecules with different halogen terminations on different metal

surfaces

Table A-1: Some results of Ullmann coupling reported in the literature, classified as a function of

surface and halogen

C-I C-Br C-Cl

Au(111) CHP1, BIB2,

TIB, TIPB3,

DIB@ RT4,

DITP@ 805

BIB @185 C, TBQ @175 2, TBB did not

happened6,

TDBB starts @185 and complete @2607,

DBBA @1008, Br4Py @2009,

DBTP@ 1705

DBCTP@ 300 10

Ag(111) CHP1, 11, DITP

@RT12,

BIB @starts RT and completed 12513,

TBB not@ RT

TBTTA partially @RT 14, DMTP @60 15

DCTP@12010

Cu(111) CHP@ RT1 TBB@ RT16-17, DBBA @RT18, Br4Py @RT9,

DMTP @RT19, DBTP and TBB (sub-ML)@

between 170 and 240K, ML @473 K 20

CTP, DCTP@

8010

Au(110) DBBA @1008

Ag(110) DITP @RT12 TBB@ RT16,

Cu(110) DMTP @RT19, BCB, dCB@

15021

It should be noted that a number of additional factors may influence the reaction

temperature, including adsorption geometry (as shown in this thesis), and, for example,

the density of metal adatoms present on the surface.22

148 Appendices

Table A-2 Abbreviation definition of molecules

1,3-bis(p-bromophenyl)-5-(p-iodophenyl) benzene BIB

cyclohexa-m-phenylene CHP

1,3,5-tris(4-bromophenyl) benzene TBB/TBPB

1,3,5-tris(4′-iodophenyl) benzene TIPB

1,3,5-triiodobenzene TIB

1,3,5-tris(4-bromophenyl)-benzene TBB

1,4-diiodobenzene DIB

2,5,9,12-tetrabromoanthra(1,2-b:4,3-b′:5,6-b′′:8,7-b′′′) tetrathiophene TB2TTA

1,3,5-tris(3,5-dibromophenyl) benzene (C24H12Br6) TDBB

10,10’-dibromo-9,9’bianthryl DBBA

4,4″-diiodo-m-terphenyl, 3′,6′-diiodo-1,1′:2′,1″-terphenyl DITP

4,4′′-dibromo-1,1′:3′,1′′-terphenyl (4,4′′-dibromo-m-terphenyl) DMTP

4,4″-dibromopara-terphenyl, 3′,6′-dibromo-1,1′:2′,1″-terphenyl DBTP

4,4″-dibromo-5′-(4-chlorophenyl)-1,1′:3′,1″-terphenyl DBCTP

4-chloro-1,1′:4′,1″-terphenyl CTP

4,4″-dichloro-1,1′:4′,1″-terphenyl DCTP

1,3,6,8-tetrabromopyrene Br4Py

3,3ʹʹʹ,5,5ʹʹʹ -tetra(p-bromophenyl)1,1ʹ:4ʹ,1ʹʹ:4ʹʹ,1ʹʹʹ -quaterphenyl TBQ

1-bromo-4-chlorobenzene BCB

1,4-dichlorobenzene dCB

Appendices 149

REFERENCES





2. Eichhorn, J.; Nieckarz, D.; Ochs, O.; Samanta, D.; Schmittel, M.; Szabelski, P.

J.; Lackinger, M. On-surface ullmann coupling: the influence of kinetic reaction

parameters on the morphology and quality of covalent networks. ACS Nano 2014, 8

(8), 7880-7889.






(5), 592-597.




Nano 2017, 12 (1), 74-81.

6. Blunt, M. O.; Russell, J. C.; Champness, N. R.; Beton, P. H. Templating

molecular adsorption using a covalent organic framework. Chemical Communications

2010, 46 (38), 7157-7159.

7. Peyrot, D.; Silly, M. G.; Silly, F. Temperature-Triggered Sequential On-

Surface Synthesis of One and Two Covalently Bonded Porous Organic

Nanoarchitectures on Au(111). The Journal of Physical Chemistry C 2017.








Journal 2016, 22 (17), 5937-5944.

10. Shi, K. J.; Yuan, D. W.; Wang, C. X.; Shu, C. H.; Li, D. Y.; Shi, Z. L.; Wu, X.

Y.; Liu, P. N. Ullmann Reaction of Aryl Chlorides on Various Surfaces and the

Application in Stepwise Growth of 2D Covalent Organic Frameworks. Organic

Letters 2016, 18 (6), 1282-1285.




6919-6921.

12. Judd, C. J.; Haddow, S. L.; Champness, N. R.; Saywell, A. Ullmann Coupling

Reactions on Ag(111) and Ag(110); Substrate Influence on the Formation of

Covalently Coupled Products and Intermediate Metal-Organic Structures. Scientific

Reports 2017, 7 (1), 14541.

13. Eichhorn, J.; Strunskus, T.; Rastgoo-Lahrood, A.; Samanta, D.; Schmittel, M.;

Lackinger, M. On-surface Ullmann polymerization via intermediate organometallic

networks on Ag (111). Chemical Communications 2014, 50 (57), 7680-7682.

150 Appendices

14. Gutzler, R.; Cardenas, L.; Lipton-Duffin, J.; El Garah, M.; Dinca, L. E.;

Szakacs, C. E.; Fu, C.; Gallagher, M.; Vondracek, M.; Rybachuk, M. Ullmann-type

coupling of brominated tetrathienoanthracene on copper and silver. Nanoscale 2014,

6 (5), 2660-2668.

15. Fan, Q.; Liu, L.; Dai, J.; Wang, T.; Ju, H.; Zhao, J.; Kuttner, J.; Hilt, G.;

Gottfried, J. M.; Zhu, J. Surface Adatom Mediated Structural Transformation in

Bromoarene Monolayers: Precursor Phases in Surface Ullmann Reaction. ACS Nano

2018, 12 (3), 2267-2274.

16. Walch, H.; Gutzler, R.; Sirtl, T.; Eder, G.; Lackinger, M. Material- and

Orientation-Dependent Reactivity for Heterogeneously Catalyzed Carbon−Bromine

Bond Homolysis. The Journal of Physical Chemistry C 2010, 114 (29), 12604-12609.

17. Gutzler, R.; Walch, H.; Eder, G.; Kloft, S.; Heckl, W. M.; Lackinger, M.

Surface mediated synthesis of 2D covalent organic frameworks: 1, 3, 5-tris (4-

bromophenyl) benzene on graphite (001), Cu (111), and Ag (110). Chemical

Communications 2009, (29), 4456-4458.


Zagrebina, E. M.; Mårtensson, N.; Cafolla, A. A.; Carpy, T.; Cunniffe, J. P.;

Preobrajenski, A. B. Effect of Substrate Chemistry on the Bottom-Up Fabrication of

Graphene Nanoribbons: Combined Core-Level Spectroscopy and STM Study. The


19. Dai, J.; Fan, Q.; Wang, T.; Kuttner, J.; Hilt, G.; Gottfried, J. M.; Zhu, J. The

role of the substrate structure in the on-surface synthesis of organometallic and

covalent oligophenylene chains. Physical Chemistry Chemical Physics 2016, 18 (30),

20627-20634.

20. Chen, M.; Xiao, J.; Steinruck, H.-P.; Wang, S.; Wang, W.; Lin, N.; Hieringer,

W.; Gottfried, J. M. Combined photoemission and scanning tunneling microscopy

study of the surface-assisted Ullmann coupling reaction. The Journal of Physical

Chemistry C 2014, 118 (13), 6820-6830.




22. Dong, L.; Liu, P. N.; Lin, N. Surface-Activated Coupling Reactions Confined

on a Surface. Accounts of Chemical Research 2015, 48 (10), 2765-2774.

b.sc. and m.sc. - eprints.qut.edu.au · b.sc. and m.sc. submitted in fulfilment of the requirements...

Documents