Atlanta University CenterDigitalCommons@Robert W. Woodruff Library, AtlantaUniversity CenterElectronic Theses & Dissertations Collection forAtlanta University & Clark Atlanta University Clark Atlanta University

Summer 7-31-2015

Tunable Electronic Properties of ChemicallyFunctionalized Graphene and Atomic-ScaleCatalyticsKelvin L. SuggsClark Atlanta University, kelvin.suggs@students.cau.edu

Follow this and additional works at: http://digitalcommons.auctr.edu/cauetds

Part of the Biological and Chemical Physics Commons

This Dissertation is brought to you for free and open access by the Clark Atlanta University at DigitalCommons@Robert W. Woodruff Library, AtlantaUniversity Center. It has been accepted for inclusion in Electronic Theses & Dissertations Collection for Atlanta University & Clark Atlanta Universityby an authorized administrator of DigitalCommons@Robert W. Woodruff Library, Atlanta University Center. For more information, please contactemcclenney@auctr.edu.

Recommended CitationSuggs, Kelvin L., "Tunable Electronic Properties of Chemically Functionalized Graphene and Atomic-Scale Catalytics" (2015).Electronic Theses & Dissertations Collection for Atlanta University & Clark Atlanta University. Paper 17.

NOTICE TO USERS ACCESSING THIS WORK All dissertations deposited in the Robert W. Woodruff Library must be used only in accordance with stipulations prescribed by the author in the preceding statement. The author of this dissertation is:

Name: Kelvin L. Suggs Street Address: 800 Peachtree Street, Suite 1320 City, State and Zip: Atlanta, GA 30308

The director of this dissertation is:

Professor: Xiao-Qian Wang, Ph.D. Department: Physics School: Arts and Sciences, Clark Atlanta University Office Telephone: 404-880-8649

Users of this dissertation not regularly enrolled as students of the Atlanta University Center are required to attest acceptance of the preceding stipulations by signing below. Libraries borrowing this thesis for use of patrons are required to see that each user records here the information requested. NAME OF USER ADDRESS DATE TYPE OF USE _______________ ____________ ______

____________ ______________

_______________ ____________ ______

____________ ______________

STATEMENT OF UNDERSTANDING

Clark Atlanta University Thesis or Dissertation Deposited in the Robert W. Woodruff Library of the Atlanta University Center, Inc.

Document Submitted: Thesis________ Dissertation____X___ Document Title:

Tunable Electronic Properties of Chemically Functionalized Graphene and Atomic-Scale Metallic Catalytics

Robert W. Woodruff Library of the Atlanta University Center, Inc. is organized exclusively to operate an academic library for the benefit of Clark Atlanta University, The Interdenominational Theological Center, Morehouse College, and Spelman College. As such the Library is granted the non-exclusive right to archive, reproduce, and distribute my thesis or dissertation in whole or in part in all formats, available now or in future. I acknowledge and grant permission for distribution and use of my thesis or dissertation for scholarly and research purpose only. Distribution and use of my thesis or dissertation in whole or in part for commercial purposes requires my written permission. I understand that I retain ownership of copyright of thesis or dissertation. I also retain the right to use in future works (such as articles or books) all part of this thesis or dissertation. I have obtained and attached, as appropriate, written permission statement(s) from the owner(s) of each third party copyrighted matter to be included in my thesis or dissertation, allowing distribution and use as specified above. I agree that permission to quote from, to copy from, or to publish this thesis/dissertation may be granted by the author or, in his/her absence, the Dean of the School of Arts and Sciences at Clark Atlanta University. I certify that the version submitted is the same as that officially approved by my thesis or dissertation committee and Department Chair and submitted to the office of the Dean of the School and Office of Graduate Studies.

________________________ _______________________ Signature of Author Date

TUNABLE ELECTRONIC PROPERTIES OF CHEMICALLY FUNCTIONALIZED

GRAPHENE AND ATOMIC-SCALE METALLIC CATALYTICS

A DISSERTATION

SUBMITTED TO THE FACULTY OF CLARK ATLANTA UNIVERSITY

IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR

THE DEGREE DOCTOR OF SCIENCE

BY

KELVIN L. SUGGS

DEPARTMENT OF CHEMISTRY

ATLANTA, GEORGIA

JULY 2015

© 2015

KELVIN L. SUGGS

All Rights Reserved

ABSTRACT

DEPARTMENT OF CHEMISTRY SUGGS, KELVIN L. B.S. MOREHOUSE COLLEGE, 2000

M.S. CLARK ATLANTA UNIVERSITY, 2010

TUNABLE ELECTRONIC PROPERTIES OF CHEMICALLY FUNCTIONALIZED GRAPHENE AND ATOMIC-SCALE METALLIC

CATALYTICS

Committee Chair: Xiao-Qian Wang, Ph.D. Dissertation dated July 2015

In this dissertation, we discuss the electronic properties, structural

configurations, and reaction mechanisms of chemically functionalized graphene and

charged atomic metals. In general, we analyze fundamental atomic scale and

nanoscale systems with density functional theory in order to investigate chemical

reaction energetics for peroxide synthesis as well as methanol production without

carbon emission. These systems were found to be tunable via the addition of cationic

and anionic charges, change in transition metal type, and modification through

chemical functionalization. Furthermore, transition state theory was used to predict an

optimal configuration for chemically functionalized graphene, efficient use of anionic

atomic gold and palladium for synthesis of water to peroxide, and clean conversion of

methane to methanol without carbon dioxide emission utilizing anionic gold.

ii

ACKNOWLEDGMENTS

I am greatly indebted to my parents, Ernest and Arbedella Suggs, for their

inspiring presence. Moreover, I extend great appreciation to my brothers, Tirrell and

Michael Suggs, for motivating me to endeavor in the realm science. I further gratitude

to Mr. and Mrs. Fred L. Suggs, Mr. and Mrs. Roland Suggs, Morris and Mary

Williams, Drs. Ernestine and Albert Suggs, Bryant and Loretta Suggs, Mariya

Dickens, E.S. and Doris Suggs, Denmark Suggs. I thank my colleagues Dr. Olayinka

Ogunro, Cherno Baba Kah, Dr. Duminda K. Samarakoon, Rosi N. Gunasinghe, Dr.

Darkeyah Reuvan, Dr. Zineb Felfli, Dr. Praphat Xavier Fernandes, Joyce Lockhart,

Debra Heard, and Ms. Ware for their invaluable advice and support. Finally, I extend

gratitude to Ana Maria Jauregui, Maria “Esther” del Cacho Suggs, Jose Antonio del

Cacho and family, and Maria Carmen del Cacho. I extend gratitude to Grandparents

Lucille Belcher, James Earl Jones, Eddie and Minnie Jones, Fred and Hattie Suggs,

and the Lee family and for their support and humor. A special thank you goes to Drs.

Xiao-Qian Wang and Alfred Msezane for their steadfast support throughout my

matriculation. I further appreciate financial support from the NSF (Grant No. DMR-

0934142), Title III, the Center for Functional Nanoscale Materials (CFNM) at Clark

Atlanta University, and the PRISM (Problems and Research to Integrate Science and

Mathematics) program under the auspices of Dr. P. Marsteller at Emory University.

iii

TABLE OF CONTENTS

ACKNOWLEDGEMENTS ................................................................................. ii

LIST OF FIGURES ................................................................................................................ vi

LIST OF TABLES ............................................................................................... viii

LIST OF ABBREVIATIONS ............................................................................. ix CHAPTER 1 INTRODUCTION ........................................................................ 1

1.1 Overview of Chemically Functionalized Graphene ............... 2

1.2 Methods and Calculations ...................................................... 5

1.3 Results and Discussions ......................................................... 7

1.4 Closing Remarks .................................................................... 14

1.5 Conclusions ............................................................................ 16

2 SELF-ASSEMBLY OF BIOFUNCTIONAL POLYMER ON GRAPHENE NANORIBBONS ................................................... 18

2.1 Introduction ............................................................................ 18 2.2 Results and Discussions ......................................................... 19

2.3 Methods .................................................................................. 30

2.4 Conclusions ............................................................................ 31 3 THEORETICAL INVESTIGATION OF PEROXIDE SYN-

THESIS USING ATOMIC NEGATIVE IONS ........................... 32

3.1 Introduction ........................................................................... 32

3.2 Theoretical overview ............................................................. 35

3.3 Reaction Dynamics ................................................................ 37

iv

3.4 Thermodynamics of Reactions .............................................. 38

3.5 Calculations Utilizing CAM theory, and Transition State

Theory .................................................................................... 39

3.6 Rate of Reaction Calculation ................................................. 42

3.7 Results and Data .................................................................... 42

3.8 Discussions ............................................................................ 50

3.8.1 The Atomic Physics Analysis ....................................... 50

3.8.2 Thermodynamics Calculation ....................................... 51

3.8.3 Hydrogen Bonding Calculation .................................... 54

3.8.4 Rate of Reaction Calculation ........................................ 54

3.8.5 Relativistic Effects ........................................................ 56

3.9 Summary and Conclusion ...................................................... 57 4 GOLD ANION CATALYSIS OF METHANE TO METHANOL

WITHOUT CO2 EMISSION ........................................................ 59

4.1 Introduction ............................................................................ 59

4.2 Reaction and Calculation Method .......................................... 62

4.3 Results and Discussion ........................................................... 65

4.4 Understanding the Results ...................................................... 69

4.5 Resonance Scattering Approach ............................................. 71

4.6 Thermodynamics of Reactions ............................................... 73

4.7 Transition-State Calculations ................................................ 74

4.8 Remarks on the Results ......................................................... 77

4.9 Discussion of Results ............................................................ 78

v

4.10 Conclusions .......................................................................... 80 REFERENCES ..................................................................................................... 81

vi

LIST OF FIGURES

Figure 1. Top view of the molecular structures of perfluorphenylazide (PFPA). 5

Figure 2. Calculated transition-state (TS) structure between the non-interacting

PFPA/graphene .................................................................................... 9

Figure 3. Ball-and-stick representation of optimized PFPA/graphene ............... 11

Figure 4. Calculated band structures for pristine graphene ................................ 12

Figure 5. Isosurface plot of charge densities of the hybridized valence ............ 14

Figure 6. Calculated band structures for (a) PFPA-functionalized graphene ..... 16

Figure 7. m-P2MS chemical scheme and optimized geometry .......................... 21

Figure 8. Supramolecular organization of polymer features on GNR ................ 23

Figure 9. m-P2MS polymer ordering on GNR and IgE binding ........................ 27

Figure 10. Multi-micron polymer ordering on GNR ............................................ 29

Figure 11. Catalytic cycle of the oxidation of heavy water to heavy peroxide .... 37

Figure 12. Electron Affinity calculations ............................................................. 44

Figure 13. Optimized structures of (a) intermediate water (HDO) ...................... 48

vii

Figure 14. Optimized structures of (a) heavy water oxidation ........................... 49

Figure 15. Change in entropy, ∆S (cal/mol•K), vs temperature, T (K) .............. 53

Figure 16. Complete oxidation of methane to carbon dioxide and water .......... 66

Figure 17. Oxidation of methane to carbon monoxide and hydrogen gas .......... 66

Figure 18. Oxidation of methane to methanol .................................................... 67

Figure 19. Oxidation of methane to formaldehyde and water ............................ 67

Figure 20. Oxidation of methane to formic acid and hydrogen gas ................... 68

Figure 21. Change in the Gibbs free energy ....................................................... 71

viii

LIST OF TABLES

Table 1. Calculated electron affinities (EAs) and R–T minima ......................... 45

Table 2. TS and EP represent, respectively, the calculated transition state and

energy ................................................................................................... 47

Table 3. Calculated energy barrier, (EB), and hydrogen bonding, (HB), in eV. 54

Table 4. TS, EP, and T represent, respectively, the calculated transition state,

energy of the products and temperature of the reaction ....................... 68

ix

LIST OF ABBREVIATIONS

AFM Atomic Force Microscopy

BE Binding Energy

CAM Complex Angular Momentum

CB Charge Band

CBM Charge Band Minimum

COSMO Conductor-like Screening Model

DCACPs Dispersion-Corrected Atom-Centered Potentials

EA Electron Affinities

D Deuterium

DFT Dispersion-Corrected Density Functional Theory

DLS Dynamic Light Scattering

DNP Double Numerical Approximation

EP Energy of Product

GGA Generalized Gradient Approximation

GNRs Graphene Nanoribbons

H Enthalpy

HeLa Henrietta Lax cells

HOMO Highest Occupied Molecular Orbital

x

IgE Immunoglobulen E

LST Linear Synchronous Transit

LUMO Lowest Unoccupied Molecular Orbital

MWCNTs Multi-Walled Carbon Nanotubes

P2MS Poly-2 methoxy-styrene

PBE Perdew-Burke-Ernzerhof

PBS Phosphate Buffer solution

PEO Poly-ethylene oxide

PFPA Pefluorphenylazide

QST Quadratic Synchronous Transit

R-T Ramsauer-Townsend

S Entropy

SPO Selective Partial Oxidation

T Tritium

TCSs Total Cross Sections

THF TetraHydrofuran

TS Transition state

T-S Tkatchenko–Scheffler

VB Valence Band

VBM Valence Band Maximum

VDE Vertical Detachment Energy

1

CHAPTER 1

INTRODUCTION

In this dissertation, graphene functionalized by perfluorazide (PFPA) and

helical poly 2-methoxystyrene (m-P2MS), and the utilization of anionic metallic

atomic systems as catalysts are studied utilizing density functional theory.

Furthermore, we predict novel graphene-based functional molecules, and gain insight

into the catalytic mechanism in the oxidation of water to peroxide and methane to

methanol. Our calculations are contrasted with those from other theoretical models.

Chapters 1 and 2 investigate graphene-based structures that have been chemically

functionalized. Chapters 3 and 4 explore anionic atomic Au, Ag, and Pd metals for

use as effective catalysts for the oxidation of water to peroxide, and methane to

methanol to without CO2 emission. Moreover, we have performed calculations at the

atomic level and nanoscale that predict plausible applications in nanotechnology,

industrial catalysis, and green energy fields. By probing at the atomic-scale further

insight can be obtained into the larger scale systems that include bulk metals as well as

large-scale allotropic carbon. Various properties are calculated in this dissertation

including transition states, thermodynamics, bandstructures, and molecular

geometries. In general, this dissertation concludes that the tunability of a given

system, depending on its scale, can be achieved electronically, structurally, and via

chemical functionalization.

2

1.1 Overview of Chemically Functionalized Graphene

Graphene is a one-layer sheet of carbon atoms arranged in a honeycomb

lattice. It has attracted a great deal of attention due to its remarkable properties and

promising potential applications.1-5 These applications include transistors, integrated

circuits, and biosensors. Moreover, future development of these applications requires

an improved understanding of how to control the associated structural and electronic

properties. Because of the gapless character of the graphene band structure, the future

of graphene electronics depends on developing effective methods for band gap

engineering. A gap can be formed in epitaxial graphene grown on a lattice matched

substrate.6,7 Although the approach involving lattice matched substrates is

straightforward, combining it with electronic transport remains a challenging task.

Another promising method for gap engineering relies on spatial confinement,

such as patterning graphene into nanoribbons.8, 9 The gap obtained by such a method

can be tuned by varying the spatial width of graphene ribbons. However, the

approaches relying on spatial confinement are prone to rough edges and defects.

Moreover, although graphene nanoribbon field-effect transistors have been shown to

exhibit excellent properties,8,10 mass production of graphene nanoribbon-based devices

is beyond the capability of current lithography technology.6

Recently, there has also been a number of studies on generating a band gap in

the gapless bilayer graphene with a perpendicularly applied electric field.11-14 In

bilayer graphene, the Bernal stacking can be lifted by asymmetric chemical doping or

3

electrical gating,4 leading to a gap opening. On the other hand, a wealth of approaches

has been developed for noncovalently and covalently functionalized graphene.10,15-23

Graphene contains a paucity of functional moieties and limited dispersibility in

solvents, seriously hindering the realization of its great potential.16, 21-23 As a result,

developing chemical methods in order to tune the materials properties has become one

of the most critical issues in exploring graphene technologies. Various chemical

modification techniques have been shown to not only enhance its solubility and

processability but also produce suitable properties for graphene-based nanoelectronic

and nanophotonic devices. Modification of graphene's electronic properties has been

carried out by well-established chemical functionalization techniques, in wherein

groups, such as H, OH, or F, bind covalently to carbon atoms, transforming the

trigonal sp2 orbital to the tetragonal sp3 state.15,24-28 Such transformations drastically

modify the local electronic properties.

Recent experimental studies have demonstrated an efficient method to

covalently functionalize pristine graphene with the use of nitrene chemistry, in which

a perfluorophenylazide (PFPA) undergoes cycloaddition with C-C double bonds,

forming an aziridine-ring linkage (see Figure 1).23 A wide range of aryl azide

derivatives are available and can be further functionalized with an array of polymeric

functional groups. The aziridino-ring reaction can be carried out by thermal and

photochemical activation, which results in graphene being soluble in organic solvents

and water. The advancement of graphene-aryl-aziridine adduct nanocomposites

4

brings with it the need to understand their impact on the electrical properties of

graphene. In lieu of the increasing amount of experimental and theoretical studies of

chemically functionalized graphene, a better understanding of how covalent

functionalization impacts the morphology and electron/hole transport in graphene

becomes pivotal for its future application in nanoelectronics. Experimental advances

have motivated our study of the electronic structure characteristics of PFPA

functionalized graphene. Herein, we report on comprehensive results based on first-

principles density functional theory calculations. PFPA functionalized graphene

perturbs the π conjugation of graphene, and the corresponding electronic properties

change from metallic to semiconducting. We show that, with the increase of aziridine

adducts, the resultant energy gap can be tuned. Our work thus asserts the unique

opportunity of tailoring the band gap of graphene with varying chemisorption

compositions.

5

Figure 1. Top view of the molecular structures of perfluorphenylazide (PFPA)-functionalized graphene with PFPA carrying alkyl, ethylene oxide and perfluoroalyl groups. Carbon, fluorine, nitrogen, oxygen, and hydrogen atoms are colored in gray (green for graphene), light blue, blue, red, and white, respectively.

6

1.2 Methods of Calculations

The structural and electronic properties PFPA functionalized graphene were in

vestigated using first-principles density functional theory calculations as implemented

in the DMol3 package.36 The Perdew-Burke-Erzernhof (PBE) parametrization37 of the

generalized gradient approximation (GGA), with a supercell of a vacuum space of 16

Å normal to the graphene plane was used. A kinetic energy change of 3x10-4 eV in the

orbital basis and appropriate Monchorst-Pack k-point grids of 6 x 6 x 1 were sufficient

to converge the integration of the charge density. The optimization of atomic positions

proceeds until the change in energy was less than 1 x 10-6 eV per cell. Although the

GGA approach systematically underestimates the band gaps, we are primarily

interested in the mechanism of gap opening. The GGA approach is expected to

provide qualitatively correct information and remains the popular choice for

investigations of covalent functionalizations.14

To investigate the effect of addend concentration on the electronic structures,

we have considered two configurations by adding one or two PFPA polymers onto a 7

x 7 rhombus cell. The cell constitutes 98 carbon atoms for graphene, 7 carbon, 4

fluorine, 1 nitrogen, and 3 hydrogen atoms for each PFPA molecule. A transition-state

search employing a combination of LST/QST algorithms36 facilitates the evaluation of

energy barriers. For transition-state calculations, we used a graphene flake to model

the graphene layer and found that the distortion generated in the transition-state search

is not crucial for the extracted energy barrier (error less than 0.2 eV).

7

1.3 Results and Discussions

Covalent functionalization of graphene with polymers is advantageous in that

long polymer chains facilitate solubilizing graphene into a wide range of solvents,

even at a low degree of functionalization.16,21-23 Soluble graphene can further undergo

in situ polymerizations with the immobilized functional groups. Although important

for solubility, the side chains of PFPA are not crucial to the electronic properties of

this nano- composite.29 As such, we replaced the side chains of PFPA with methyl (-

CH3) groups in order to simplify the electronic structure calculations. One of the

important chemical reactions is the [2+1] cycloaddition of nitrenes, which has been

successfully used to functionalize carbon nanomaterials with high efficiency. Shown

in Figure 2 is the transition path along with the relative energies of the corresponding

[2+1] cycloaddition reaction for PFPA functionalized graphene.

The reactant constitutes the non-interacting PFPA and graphene, whereas the

product is the PFPA functionalized graphene in which the addition of a PFPA

saturates a double bond between two graphene carbon atoms, forming a cyclopropane-

like three-membered ring. Although the energy differences between the starting and

ending configurations is fairly small (about 0.1 eV), the transition barrier is 1.92 eV, in

good agreement with the experimental estimate of ∼2-3 eV.23 As the predominant

contribution to the transition barrier is attributed to the breaking of a N-N double bond

and the associated loss of N2, our results are in conformity with the experimental

observation that functionalization occurs on the surface of graphene after the [2+1]

8

cycloaddition of PFPA. We illustrate in Figure 3 the optimized conformation of PFPA

functionalized graphene. The PFPA molecule increases the bond lengths linking to

atoms on graphene. The corresponding bond length between the C atom on graphene

and the N atom of the PFPA molecule is around 1.43 Å, whereas that of the C atom

and its nearest neighbors on graphene is around 2.21 Å. The latter C-C distance is

notably larger than the C-C bond length of 1.42 Å of graphene with sp2 hybridization

and indicates bond breaking.

The C-C bond lengths in graphene beyond the nearest neighbors are found to

be little affected by the functionalization. The graphene-PFPA molecule interaction in

the covalent functionalization has direct consequences on the electronic properties of

graphene. Previous theoretical work investigated the addition of functional groups as

free radicals to graphene.24,25,29 These functional groups drastically disrupt the

geometries and electronic structures of graphene by introducing local sp3 hybridization

defects, which induce an sp3-type “impurity” state near the Fermi level.14,30,31 In the

cases of divalent functionalization, two sp3 states induced by two neighboring

functional sites are shifted away from the Fermi level due to the rehybridization into

bonding and anti-bonding states.31 Therefore, the local bonding configuration can

significantly affect the electronic structure of functionalized graphene.

9

Figure 2. Calculated transition-state (TS) structure between the noninteracting PFPA/graphene and the PFPA functionalized graphene plus a N2 molecule. PFPA adsorbs onto the graphene surface via a nitrene radical. After losing N2, PFPA reacts with graphene via an electrophilic [2 + 1] cycloaddition reaction. Carbon, fluorine, nitrogen, and hydrogen are colored in gray (green on graphene), light blue, blue, and white, respectively.

To further pursue this point, it is instructive to recall that, for nitrene

functionalized carbon nanotubes, the cyclopropane ring structure introduced by [2+1]

cycloadditions can either remain intact or lead to cleavage of the sidewall bonds with

the increase of the nanotube curvature, resulting in two valence tautomeric forms that

display distinct electronic characteristics and markedly different transport properties in

metallic tubes.30 In those cases, the nitrene chemistry introduces cyclopropane

10

functionality in place of the partial double bonds initially present in the π-conjugated

electronic structure. Each addition saturates a conjugated bond and causes the valence

of a pair of carbon atoms to revert from sp3 to sp2 hybridization.30,31

We depict in Figure 4 the calculated band structures for PFPA functionalized

graphene, along with the pristine graphene for comparison. It is readily observed that,

after the covalent functionalization, the π and π* linear dispersion of pristine graphene

in the proximity largely preserves the Dirac point (K). Therefore, a gap is created

between the π and π* states. These electronic properties of PFPA-functionalized

products are in sharp contrast to the sp3 rehybridization and loss of π electrons found

upon the addition of monovalent chemical groups in other functionalization

schemes.14,31 The absence of sp3-type “impurity” states in the vicinity of the Dirac

point is also consistent with the rationale that the C-C bond between the two

bridgehead atoms is either broken or substantially weakened, leading to partial

recovery of the π-electron system.

On the other hand, our present results are clearly distinct from those of the

noncovalent functionalization.14,32,33 For noncovalent functionalization, there is little

modification of the band structures close to the Fermi level, and the corresponding

bandstructure constitutes flat and dispersed bands that can be readily classified as

arising from functional group and pristine graphene contributions.34 By contrast, the

PFPA-functionalized graphene displays profound level hybridizations. In particular,

the bandgap opening at the Dirac point implies important perturbations generated by

11

the functionalization. All of the band gaps of the PFPA functionalized graphene

appear at the Dirac point. It is worth noting that, although the C atoms on graphene

connecting to PFPA retain their sp2 hybridization, the sp2 hybridization angle is

changed. As a result, the electronic structure of graphene is inevitably affected by

PFPA functionalization. An important ramification of the [2 + 1] cycloaddition

induced perturbation is that the alteration in the electronic structure of graphene

increases with incrementing PFPA functionalization concentration. We have

investigated the functionalization of graphene at a higher PFPA concentration by

including another PFPA functional group in the unit cell (see Figure 3).

Figure 3. Ball-and-stick representation of optimized structures of PFPA functionalized graphene with one and two PFPA addends in the left and right panels, respectively. d and d0 are two characteristic bond lengths of 1.56 and 1.42 Å, respectively.

12

Figure 4. Calculated band structures for pristine graphene (left panel), one-PFPA functionalized graphene (middle panel), and two-PFPA functionalized graphene (right panel). Γ = (0,0), K = (π/3a,2π/3a), M = (0,π/2a), where a = 17.22 Å for a 7 × 7 rhombus unit cell. The Fermi level is shifted to 0 eV (dashed blue line).

The results from geometry optimizations indicate that bridgehead C-C bond

breaking persists at higher concentrations. The extracted energy gap is 0.16 and

0.29 eV for one and two PFPA molecules on a graphene unit cell consisting of 98

carbon atoms, respectively. Closer scrutiny of the band alignments32 and dispersions

near the Dirac point reveals that the gap opening is primarily attributed to the

functionalization-induced modifications of the π conjugation. The disruption of the

original π conjugation is manifested in the level hybridization, as seen in the band

13

structure (Figure 4). Specifically, the highest occupied molecular orbital (HOMO) and

the lowest unoccupied molecular level (LUMO) of PFPA line up with the π and π*

bands of graphene at about -1 and 1 eV, respectively. The band alignment is such that

the interaction between flat and dispersed bands leads to hybridization induced level

avoided-crossing, which leads to the split of π and π* bands of graphene into two

hybridized bands each.

We show in Figure 5 charge densities of the corresponding hybridized bands at

the band center (the Γ point). For those states, the charge density distributions display

predominant charge confinements on PFPA molecules for hybridized conduction and

valence bands. This is to be contrasted to the conjugated π and π* pattern on graphene.

As can be seen in Figure 5, the increase of the addend concentration leads to a

proportional increase of the change of the π conjugation. This correlates with the

associated increase of the band gap and thus provides support of the suggested

scenario of the functionalization-induced band-gap opening. Careful examination of

the charge density distributions also indicates the existence of σ and σ* bonds in the

hybridized states that contribute to the gap formation as well.

14

Figure 5. Isosurface plot of charge densities of the hybridized valence band maximum (VBM), conduction band minimum (CBM), and the next near-gap states at the band center. The isovalue is 0.025 au.

1.4 Closing Remarks

A few remarks are in order. (i) The semi-metallic graphene is more sensitive to

the π-conjugation changes than the metallic single-walled carbon nanotubes. For the

latter to open a gap, it is necessary to have a higher functionalization

concentration.30,31 This appears to be attributed to the curvature of the nanotube.30 (ii)

The formation of a band gap in PFPA functionalized graphene is analogous to the

epitaxial graphene in that Stone-Wales defects and the graphene-substrate interaction

generate band gaps due to the disruption of π conjugation. (iii) In this work, we focus

mainly on the electronic structure characteristics, specifically, the mechanism of band-

gap formation for PFPA functionalized graphene. The issue of solubility of alkyl,

ethylene oxide, and perfluoroalkyl groups can be addressed by alternative theoretical

approaches, such as density functional tight-binding calculations. (iv) In addition to

15

the absence of midgap impurity states, it is worth noting that the gap formation

mechanism of PFPA functionalized graphene is qualitatively distinct from that of NH

functionalized graphene.35

We illustrate in Figure 6 the calculated band structure. As is readily observable

from Figure 6 that, although both schema lead to a gap at the Dirac point (K) that is

attributed to the functionalization-induced symmetry breaking,35 NH functionalized

graphene generates a crossing in the vicinity of the Dirac point. By contrast, the

PFPA-functionalized graphene sustains the gap formation. This clearly demonstrates

the crucial difference between NH-radical and aziridine-ring linkages. (v) The

concentration dependence of the [2+1] cycloaddition is investigated with additional

PFPA adsorption on the same side of the graphene, in accordance with an

experimental study.23 If the absorption is on two different sides of graphene, our

results indicate that the gap is still opened, but the value of the gap is almost identical

(slightly smaller) than the single adsorption. This shows that the distortion of the π-

conjugation network depends sensitively on the adsorption configurations as well.

16

Figure 6. Calculated band structures for (a) PFPA functionalized graphene and (b) NH-functionalized graphene, along with that for the pristine graphene (blue dashed lines).

1.5 Conclusions

In summary, we have studied the electronic characteristics of PFPA

functionalized graphene. We have shown that the [2+1] cycloaddition preserves the

sp2 hybridization network of the carbons on graphene. However, the π conjugation of

graphene near the Fermi level is greatly disturbed by functionalization, which leads to

the opening of a band gap dependent upon the PFPA concentration. This contrasts

with the free-radical functionalization case where the sp3-type band is induced close to

the Fermi level. Such dependence of the electronic properties on the degree of

functionalization of graphene suggests a novel and controllable method for the “band

engineering” of graphene. Our findings on the nature of a PFPA functionalization-

17

induced band gap provide useful guidelines for enabling the flexibility and

optimization of graphene-based nanodevices.

18

CHAPTER 2

SELF-ASSEMBLY OF BIOFUNCTIONAL POLYMER ON GRAPHENE NANORIBBONS

2.1 Introduction

The planar structure of graphene has potential applications in electronics,

sensor devices, spintronics, nanoelectronics, and biodiagnostics.38-42 Laterally

constraining the carriers in a quasi-one-dimensional system, graphene nanoribbons

(GNRs) can be fabricated using lithographic methods43-45 and by metal particle-

assisted46-48 or oxidative49,50 longitudinal unzipping of multi-walled carbon nanotubes

(MWCNTs). GNRs can be processed for specific applications by modification of the

basal plane and edge functional groups composed of carboxylic acid, hydroxyl,

epoxide, and carbonyl. Hydrazine significantly reduces the amount of oxygen

functional groups on GNRs, resulting in improved conductivity.49,50

Recently, significant efforts have been devised to create self-assembled

hierarchical graphene-based materials.51-53 Examples include the co-assembly of

graphene and organic monolayer,54, 55 titania nanosheets,50 and proteins.53 These

ordered structures are assembled by intermolecular forces arising from electrostatic,

π– π stacking, dipolar, van der Waals, hydrogen bonding, or metal–ligand interactions.

Various types of polymers have been shown to interact with graphene56 or graphene

oxide, forming stable hybrids.52,54 For instance, modified poly(phenylene vinylene)

19

conductive polymer can specifically attach to GNRs and tune the corresponding

electronic properties.56 In view of the rapid progress made in preparing controlled

polymer self-assembly, a better understanding of the interfacial interactions between

the helical polymer and GNRs is clearly desirable.

2.2 Results and Discussions

We have investigated the self-assembly of biocompatible polymer, R, ω-bi [2,

4-dinitrophenyl caproic] [poly(ethylene oxide)-b-poly-(2-methoxystyrene)-b-

poly(ethylene oxide)] (DNP-PEO-P2MS-PEO-DNP, hereafter referred to as m-

P2MS), onto chemically prepared GNRs. P2MS polymer with more than 20 2-

methoxystyrene monomers forms a helix rod structure.58 The m-P2MS assembles into

secondary structures in solution, resulting in associated optical activity.59 Furthermore,

the chiral initiated polymer surfaces are better supports for HeLa, mouse osteoblast,

and human osteoblast cell as compared to non-chiral initiated counterparts, owing to

the moderately periodic topography.60 Dinitrophenol groups attached at the ends of m-

P2MS are suitable for bio-sensing applications due to their high affinity to anti-DNP

IgE protein in solution and IgE on the surface of mast cells.61 The dangling glycol

end-segment is hydrophilic, aiding in the extension of the DNP groups away from the

P2MS backbone in aqueous environments, thereby assisting DNP's availability for

protein interaction. However, m-P2MS typically forms non-uniform surfaces upon

deposition on substrates. The lack of control in the formation of uniform surfaces

20

hinders the application of this type of versatile polymer. In this regard, the hierarchal

self-assembly of m-P2MS into anisotropic ordered patterns on chemically prepared

GNRs is timely and of considerable interest. This is particularly the case because

surface-adsorbed m-P2MS GNR specifically binds with anti-DNP IgE protein.

Illustrated in Figure 7 is chemical scheme ofm-P2MS, along with optimized

geometries of m-P2MS and parallel aligned P2MS chains on graphene.

While the electronic structure for graphene and GNRs is distinctive from each

other, the ensuing changes arising from the GNR edges become dormant for ribbons

with a width larger than 100 nm. Monomers of 2-methoxystyrene in Figure 7(a) were

optimized using dispersion-corrected density functional theory (DFT) method. First-

principles calculation results show that helical P2MS aligns parallel on graphene

owing to van der Waals interactions (Figure 7(b)). The calculated electronic band

structure of m-P2MS-functionalized graphene shows substantial molecular orbital

hybridization, which indicates component charge transfers. Specifically, 2-

methoxystyrene monomers in the polymer backbone serve as charge donors to

graphene.

Consequently, the helical m-P2MS chains and graphene form donor-acceptor

complex with enhanced van der Waals interactions. The diameter of m-P2MS

backbone is about 5 nm. As a result, GNRs generated from unzipping MWCNTs, with

typical width of 300-500 nm and length of a few micrometers, provide desired planar

“flatbed” for m-P2MS self-assembly. Furthermore, the oxygen groups at the GNR

21

edge and the basal plane contribute to m-P2MS self-assembly onto GNRs via

hydrogen bonding.

Figure 7. m-P2MS chemical scheme and optimized geometry. (a) Schematic representation of α, ω-bi[2,4-dinitrophenyl caproic] [poly(ethylene oxide)-b-poly(2-methoxystyrene)-b-poly(ethylene oxide)] (m-P2MS), and side views of optimized helical m-P2MS section. Helical poly (2-methoxystryene), glycol segment, and pendant 2, 4-dinitrophenyl are highlighted in red, yellow, and blue, respectively. (b) Top view of the optimized structure of attached helically P2MS on graphene (green color).

S O P

22

The m-P2MS of 1015K molecular weight (74112 repeat units) forms

aggregates of 265(35 nm diameters in tetrahydrofuran (THF) solvent, which was

confirmed by dynamic light scattering (DLS). The strong van der Waals interactions

between the m-P2MS helical rigid rods lead to the formation of polymer-specific sized

aggregates in solution. The optical rotation of polarized light by the m-P2MS

secondary polymer structures in solution, yet the formation of nonuniform films on

silicon or silicon oxide surfaces, indicates that aggregates form weakly associated

superstructures in solution. As such, GNR's adhesive properties, due to its inherent

van der Waals forces,62 make it suitable for the controlled attachment of m-P2MS

secondary polymer structures, which is of interest from the perspective of better

understanding graphene/polymer nanoscale film topography. To this effect, we have

prepared GNRs with a typical height of approximately 0.50-0.75 nm using oxidative

chemistry technique developed by Tour and co-workers.49,50 We show in Figure 8

typical atomic force microcopy (AFM) images of m-P2MS spin-cast onto single-layer

GNR adsorbed on a silicon oxide (SiO2) surface.

23

Figure 8. Supramolecular organization of polymer features on GNR showing three-dimensional AFM topography data. (a) Lamella ordered m-P2MS polymer drop-cast on the GNR. Bottom inset: structured polymer periodicity of 200 nm and height of 75 nm along the m-P2MS–GNR hybrid. (b) Aligned herringbone-shaped lamella polymer features after 24 h exposure to solvent-rich environment. Bottom inset: structured polymer periodicity of 195 nm and height of 75 nm along the m-P2MS–GNR hybrid. (c) Aligned herringbone-shaped polymer features in polymer deposited on chemically reduced GNR (r-GNR). Bottom inset: structured polymer periodicity of 150 nm with heights of 4.5–20 nm extracted from the height profile along the m-P2MS–GNR hybrid.

The m-P2MS polymer spontaneously self-assembles along the entire ribbon in

a platelet pattern (Figure 8(a)). The platelet nanopattern has a periodicity of 200 (20

nm and an overall height of 59-74 nm. The mean corrugation height along the cross

section of the hybrid ribbon is 23 (5 nm (Figure 8(a)). The gradient processed image

(inset of Figure 8(a)) displays partially overlapping polymer lamella in “scale like”

fashion along the GNR. The size of apparent platelet pattern is consistent with DLS

data that reveal similarly sized m-P2MS aggregates in THF solvent. This demonstrates

that the ∼265 nm m-P2MS secondary polymer structures attach to the graphene

surface via electrostatic and van der Waals interactions, forming supramolecular layers

and presumably further mediated by the graphene edges. In contrast, there are no

24

distinguishable ordered m-P2MS polymer structures on the amorphous SiO2 surface

(Figure 8(a)).

The nanocomposites were processed to form complex networks upon further

exposure to a solvent vapor-rich atmosphere for a 12 h period (Figure 8(b)). The m-

P2MS on the GNR spontaneously forms into aligned herringbone features on the

GNR. The herringbone pattern has a periodicity of 200 (20 nm, along with an overall

height and width of 5974 and 550 nm, respectively. The gradient processed image

(inset of Figure 2(b)) reveals a distinct herringbone pattern on the ribbon. The average

corrugation height along the cross section of the hybrid ribbon is 14 (5 nm (inset of

Figure 8(b)), a moderate reduction of film topography after solvent exposure. These

herringbone-shaped polymer features are attributed to the incorporation of solvent

vapor into the polymer matrix, allowing for increased m-P2MS chain mobility. The m-

P2MS chains are subsequently able to undergo further van der Waals mediated

adhesion to the GNR.

The chemical unzipping of MWCNTs invariably leads to the presence of

oxygen species on the basal plane and ketone groups along edges of the GNR, which

can interact with the ether groups available on the m-P2MS aggregate. The

combination of the hydrophobic P2MS chain and hydrophilic glycol terminal groups

is essential for supramolecular attachment to the GNRs. A circular dichroism

spectroscopy study of m-P2MS and the hybrid with GNRs reveals that, while the

feature of the helical rod of m-P2MS is modified slightly with the inclusion of GNRs,

the addition of water plays an important role in the enhanced interactions between m-

25

P2MS and GNRs. We demonstrate in Figure 8(c) the height tuning of the m-P2MS

polymer nanopattern on chemically reduced GNR (r-GNR). A herringbone polymer

pattern is clearly evident on r-GNR (Figure 8(c)).

The height data (inset of Figure 8(c)) profile along the P2MSr-GNR hybrid

shows the structured polymer periodicity of 150 (10 nm and an overall corrugation

height ranging from 4.5 to 20 nm. Thus, the reduction in m-P2MS surface topology is

likely related to the decrease of the oxygen functional groups, while the preservation

of the herringbone pattern can be attributed to van der Waals interaction with the

GNR. Consequently, r-GNR facilitates flexibility in controlling polymer height, while

retaining the characteristic polymer pattern on the r-GNR surface. The polymer

interaction with the GNR nanofiller was investigated using differential scanning

calorimetry (DSC) over the temperature range of 40 to 200 C·.The GNR thermogram

exhibits an endothermal transition approximately at 130 C·, which is attributed to the

release of adsorbed water. Upon further heating, the GNRs undergo an exothermal

transition at 170 C· that is associated with the thermal decomposition. The DSC

thermogram of 1% GNR loaded m-P2MS shows a melting temperature at 102 C·, an

increase of 7 C· in comparison to m-P2MS and m-P2MS loaded with 1% r-GNR,

which have melting temperatures of ∼95 C·.

The addition of low concentrations of r-GNRs has little effect on the polymer

thermal performance. The prominent endothermic peak of the GNR thermogram is

notably absent from the GNR and r-GNR polymer composite thermograms. The

composite thermograms indicate that the m-P2MS polymer hinders the release of

26

water from the GNR surface. The composites undergo exothermic deflagration

between 150 and 165 C·. The thermal performance of the composite implies that the

incorporation of the chemically prepared GNRs at low concentrations into the m-

P2MS matrix moderately inhibits the polymer chain movement, consistent with the

effect of other carbon-based fillers on polymer thermal properties.63-67

To demonstrate the potential of these polymer nanostructures as a biological

interface, we performed protein interaction studies using confocal laser scanning

microscopy and AFM. IgE is a well-studied protein known to be involved in the

body's immune response. The m-P2MS adsorbed on the GNRs retains the bioactivity

of the divalent DNP functional groups. Three-dimensional laser scanning microscope

and AFM images (Figure 9(a), (c)) show polymer ordering on the basal plane and

edges of the GNR. As seen in Figure 9(a), the polymeric structures are ordered along

the GNR. The darker region along the axis is of higher thickness in comparison to the

edges. Discernable polymeric structures yield ovate and overlapping configurations

(Figure 9(c)), forming hierarchal polymer structures at GNR basal and edge interfaces.

The ovate polymer structures have an average height, width, and length of 7.8, 336,

and 900 nm, respectively. The polymer lamella patterns appear to follow along the

smooth edges and basal plane of the macromolecular ribbon. These polymeric

structures, attached at the edges, undergo further conformational change into

secondary periodic structures on the GNR basal plane.

27

Figure 9. m-P2MS polymer ordering on GNR and IgE binding. (a) Laser confocal microscope and 3D rendering of AFM topography image of m-P2MS polymer ordering on GNR before ((a),(c)) and after ((b),(d)) IgE exposure, respectively. Inset of b: Fluorescence confocal microscope images.

The bioactivity was investigated by incubating them-P2MS GNR

nanocomposite in phosphate buffer solution (PBS) containing fluorescently labeled

Alexa 488-IgE and a blocking protein for 15 min. Fluorescent microscope images

(inset of Figure 9(b)) demonstrate lengthwise fluorescence on the nanocomposite

ribbon. Dyes attached to graphene undergo fluorescence resonance energy transfer,

which typically results in dye fluorescence quenching.62, 69 By contrast, fluorescence is

not quenched in them-P2MS GNR nanocomposite, as the polymer attached to the

GNR surface and edges acts as a spacer that keeps the fluorescently labeled IgE from

contacting the GNR surface. AFM topography images in Figure 3(d) reveal that the

IgE protein clusters at the edges of the polymer structures on the GNR. The DNP

groups being accessible to anti-DNP IgE in the solid state is relevant for biosensing

28

applications. The features showing IgE protein edge clustering indicate that the DNP

groups are spatially confined at the edges of the polymer superstructures and available

for protein interactions. It is worth noting that protein fibrils propagate from the edges

of the polymer structures. The height profile (inset of Figure 9(c)) of the P2MS GNR

hybrid shows a periodicity of 270 (50 nm and an average corrugation height of 25 (5

nm. Subsequent to the IgE exposure, the composite ribbon has a periodicity of 100 (35

nm and an average corrugation height of 15 (5 nm (inset of Figure 9(d)). The decrease

in surface roughness is attributed to the addition of IgE protein.

A typical Raman spectroscopy (433 nm laser excitation) graph for the GNR

polymer hybrid composes a graphene (G) band (1603 cm-1) and a graphene defect (D)

band (1354 cm-1) of roughly equal size (Figure 10(b)). Four bands observed from 2600

to 3200 cm-1 are typical of chemically prepared GNRs.50 Adsorption of polymer

retains the characteristic Raman spectroscopy G band and D band peaks. Shown in

Figure 10 is a microscope image of m-P2MS/GNR composites, on a silicon oxide

substrate, which have been exposed to a solvent atmosphere over a period of a week.

Remarkably, the polymer features are spatially confined the approximate width of a

GNR (light blue) and anisotropically extend beyond the footprint of the GNRs by

several micrometers. Micro-Raman mapping with a spatial resolution of 1 µm was

used to investigate the polymer surface features.

29

Figure 10. Multi-micron polymer ordering on GNR. Long-range polymer ordering of P2MS centered on GNR, propagating on SiO2. The composite feature is typically 10 µm in length and 1 µm in width. Inset: Micro-Raman map spectroscopy of the peak width of the G (1603 cm–1) line intensity acquired on a 12 × 8 µm scan window highlighted by the red box.

The Raman mapping of the G line (1603 cm-1) peak (inset of Figure 10) shows

that the center point of the polymer feature (yellow region) is unambiguously assigned

to GNRs. These results confirm that the GNRs are capable of controlled propagation

of m-P2MS over several micrometers on a SiO2 surface, and the polymer structure

retains the approximate width of the nucleating GNR. Understanding charge transfer

at the polymer/graphene interface and the spatial distribution of the resultant charge

carriers is important to the development of graphene-based devices. Interestingly,

density functional calculations show that charge accumulates at the m-P2MS chain

ends for the valence band states of the m-P2MS GNR nanocomposite. Graphene

preferentially maintains a charge-neutral molecular orbital level.70-73 Consequently,

the energy level alignment relaxation of doped graphene induces optimal charge

30

redistribution from the length of the polymer chain to the m-P2MS chain ends, leading

to the electrostatic joining of the charged m-P2MS chain ends. The induced charge

transfer between m-P2MS and GNRs promotes long-range attractions, thus allowing

the preferred alignment of m-P2MS chains. It is worth mentioning that the planar

topography and the smooth edges of GNRs play an important role in the self-assembly

process. This is in contrast to the deposition of m-P2MS on highly ordered pyrolytic

graphite (HOPG), in which no regular m-P2MS polymer self-assembly pattern is

observed. These observations strongly suggest that GNR facilitates the controlled

assembly of m-P2MS, owing to enhanced van der Waals interactions and its unique

planar conformation with regular edges.

2.3 Methods

The MWCNTs were unzipped and reduced using an optimized method

developed by Tour and coworkers.49 A modified GNR purification procedure was

utilized, in which the product was isolated by repeated mixing and centrifugation

steps, for the separation of exfoliated graphene ribbons from unzipped MWCNTs. The

oxidation reaction product was poured into 5 mL of liquid nitrogen cooled 30%

hydrogen peroxide, which prevented the precipitation of potassium permanganate. The

resultant single and bilayer GNRs were spin-cast from an ethanol/water (50:50)

solution onto a silicon wafer with 300 nm SiO2 wafers. The GNRs were reduced by

their adsorption on the SiO2 surface and subsequent treatment with 1 vol% hydrazine

monohydrate and 1 vol% concentrated ammonium hydroxide. The reduced GNRs

31

were washed with de-ionized water and dried under nitrogen gas flow. Protein Binding

was initiated via the m-P2MS/GNR composites adsorbed on a silicon oxide wafer

were sensitized with Alexa488-IgE containing 1 mg/mL of BSA for about 15 min,

which were subsequently washed in phosphate buffer solution (PBS) containing BSA.

2.4 Conclusions

We have described a method for controlled nanopatterning of GNRs by the

self-assembly of m-P2MS onto the GNR surface. The m-P2MS self-assembly on the

GNRs is attributed to van der Waals interaction between the GNR basal plane and the

m-P2MS backbone. The present investigation provides a basis for studying polymer

surface topology on GNRs and the associated effect on protein binding.

32

CHAPTER 3

THEORETICAL INVESTIGATION OF PEROXIDE SYNTHESIS USING ATOMIC NEGATIVE IONS

3.1 Introduction

The exploration of the role of atomic particles and nanoparticles in catalysis

has attracted a wide range of fundamental and industrial investigations.75-89 In

particular, the direct synthesis of hydrogen peroxide from H2 and O2 using supported

Au, Pd, and Au–Pd nanoparticle catalysts has been reported.81,89 The experiments

found that the addition of Pd to the Au catalyst increased the rate of H2O2 synthesis

significantly as well as the concentration of the H2O2 formed. These findings have

motivated us to study at the fundamental atomic physics level the mechanism

underlying the Au and Pd nanoparticles’ excellent catalytic properties.80,81 including

the substantial enhancement of the Au–Pd nanocatalyst over the individual Pd and Au.

Recently, we investigated the transition state of the oxidation of water to peroxide by

performing dispersion-corrected density-functional theory calculations on the catalytic

properties of atomic Au– and atomic Pd– negative ions.92 From the results, we

concluded that the atomic Au– negative ion catalyst will most likely catalyze a reaction

whenever water is the medium, but the atomic Pd– negative ion catalyst acts as the

33

most efficient and economical catalyst when compared to the atomic Au– negative ion

catalyst.

In this study, we want to obtain definitive answers to the following questions:

(1) Can these catalysts, namely, atomic Au– and atomic Pd– negative ions, efficiently

catalyze heavy water to heavy peroxide? (2) How does the performance of these

negative ion catalysts compare when catalyzing H2O2, HDO2, and D2O2 from H2O,

HDO, and D2O, respectively?

Answers to these questions could provide valuable insight into the bond

strengths in H–O–H, H–O–D, and D–O–D. Toward this end, we have investigated and

compared the catalysis of H2O, HDO, and D2O conversion to H2O2, HDO2, and D2O2,

respectively. Deuterium (D) is obtained by combining two nuclei of hydrogen via

nuclear fusion at very high temperature. Deuterium is unique among heavy stable

isotopes in being twice as heavy as the lightest isotope. This difference increases the

strength of water’s H–O bond, and in turn, this is sufficient to cause differences that

are important to some biochemical reactions.98 Deuterium can replace the light

hydrogen in water molecules to form heavy water (D2O), which is about 11% denser

than normal water (this is enough that ice made from it sinks in ordinary water).99

Biologically, this difference means that large amounts of heavy water can have

harmful effects on animals, although it would entail approximately 2 weeks of

consuming only D2O and no H2O to be terminal to humans. It has been found that H2O

has a longer intramolecular O–H bond length than D2O’s corresponding O–D bond

34

length. Specifically, the O–H bond is longer by about 0.03 Å, or 3%. Also, the

intermolecular hydrogen bond that connects two separate molecules is shorter in H2O

than in D2O. Here, the difference is about 0.07 Å, or 4%. Geometrical differences

between the structures of light and heavy water also exist.

Previous research predicted an overall broadening of the H2O structure

compared to the D2O structure. The hydrogen molecule and its isotopomers, HD and

D2, are of interest because of their presence in tokomak edge plasmas, planetary

atmospheres, and different astrophysical environments.95 Recent interest in deuterated

hydrogen includes the isotope effect,96 electron-impact cross sections,95 the question

of the ability of fruit flies to sniff out heavy hydrogen,98 and the sensitivity of the

endohedral translation–rotation dynamics to the differences in the interaction

potentials, including to the large variations in the masses and the rotational constants

of H2, HD, and D2 inside C60.97 The strength of hydrogen bonds per water molecule is

less in H2O than in D2O (3.62 eV vs 3.76 eV). Together, these structural differences

give light water a broader structure and heavy water a narrower, tetrahedral shape.98 In

the formation of H2O2, HDO2, and D2O2, we have also calculated the corresponding

transition states with and without the presence of the atomic negative ion catalysts. We

found that the entropy decreases when moving from H2O → HDO → D2O because of

the higher average number of hydrogen bonds per molecule and bond length

contribution to the ordering of the molecules. Additionally, in this study, we consider

the catalytic effect of both the Au– and Pd– ions on the oxidation of the light,

35

intermediate, and heavy water to the corresponding peroxides from a fundamental

approach through the combined theoretical atomic physics and the quantum chemistry

perspectives.90-92

3.2 Theoretical Overview

The Regge-pole methodology 99 has been employed to explore, through the

elastic total cross sections (TCSs), the near–Threshold scattering of slow electrons

from both the ground and excited states of simple and complex atoms.100-103

Embedded in the Regge-pole methodology are the crucial electron–electron

correlations and the vital core polarization interactions. These physical effects are

responsible for the existence and stability of typical negative ions. It has been found

that Ramsauer–Townsend (R–T) minima, shape resonances, and the dramatically

sharp long-lived resonances characterize the low energy electron elastic scattering

TCSs. From the electron energy positions of the very sharp long-lived resonances, the

important binding energies (BEs) of the resultant negative ions formed during the

collisions as Regge resonances have been extracted. These binding energies are

identified with the EAs of the relevant atoms when the binding energies correspond to

the negative ions formed in the ground state.

Very recently, the same fundamental mechanism that underlies the well-

investigated muon-catalyzed nuclear fusion using deuterium (D) and tritium (T) has

been proposed to drive nanoscale catalysis.90,91 The fundamental atomic mechanism

36

responsible for the oxidation of water to peroxide has been attributed to the interplay

between Regge resonances and R–T minima in the electron elastic TCSs for Au and

Pd atoms, along with their large electron affinities.90 Furthermore, dispersion-

corrected density-functional theory (DFT) transition state calculations have been

performed on the atomic Au– and Pd– ion catalysis of water conversion to H2O2,

revealing the important role played by the formation of the Au–(H2O)2 and Pd–(H2O)2

anion molecular complexes.105

The formation of these anion complexes in the transition state, with the

interaction of the Au– and Pd– ions with H2O being comparable to the strong hydrogen

bond, has been identified as the fundamental mechanism, as shown in Figure 11, for

breaking the hydrogen bonding strength in the catalysis of H2O2 using the atomic Au–

and the atomic Pd– ions. Thus, the crucial link between low-energy electron elastic

scattering resonances and low-energy chemical reaction dynamics has now been fully

established.92 In the H2O2 catalysis, the anion Au–(H2O)2 molecular complex formed

during the transition state weakens (breaks) the H–O bonds, thereby promoting the

formation of the H2O2 in the presence of O2. This important mechanism can also be

used to understand the experiments with ozone gas that have demonstrated that

bacteria and viruses were torn apart, with the Ag acting as an extremely efficient

oxidative catalyst.106

37

Figure 11. Catalytic cycle of the oxidation of heavy water to heavy peroxide through the atomic Au– negative ion catalyst.

3.3 Reaction Dynamics

Following reference 90, in this study, we first consider the slow oxidation of heavy

water to heavy peroxide without the atomic negative ion catalyst

(1) 2D2O + O2 → 2D2O2

Then, we apply the atomic Au– negative ion to speed up reaction 1 and obtain

(2) Au- (D2O) 2 + O2 →Au- + 2D2O2

(3) Au- + 4D2O + O2 →Au- (D2O)2 + 2D2O2

Then, we add reactions (2) and (3) and obtain

(4) 4D2O + 2O2 → 4D2O2

Reactions (2) and (3) are captured in the self-explanatory Figure 11 for clarity.

A similar result as in equation 4 is obtained when the atomic Au– negative ion catalyst

38

is replaced by the atomic Pd– negative ion catalyst and when the H2O and HDO

molecules are used. Note, it is important to introduce the Au- ion into the (D2O2)2

before adding the O2 as in Equation (2). The question we want to address here is

which of the two negative ion catalysts is more effective in the catalysis of the heavy

water to peroxide. The anionic complexes Au–(H2O)2 and Pd–(H2O)2 have been

characterized as atomic Au– and atomic Pd– negative ions interacting with two water

molecules, respectively, i.e., as anion–molecule complexes.105 The large electron

affinities of atomic Au and atomic Pd played the essential roles; they are important in

the dissociation of the Au–(H2O)2 and Pd–(H2O)2 complexes breaking up into atomic

Au– or atomic Pd– negative ion and (H2O)2, respectively.105 Very important here, the

experiment84 found a stronger interaction between the atomic Au– negative ion and

H2O and that the atomic Au– negative ion does not react with O2. Similar reactions as

in Equations (1)–(4) can be generated when the heavy water, D2O, is replaced by the

intermediate water, HDO, and a figure similar to Figure 11 can be generated as well.

3.4 Thermodynamics of Reactions

In the H2O2 catalysis from water using the atomic Au– negative ion, the

hydrogen-bond-breaking mechanism has been attributed to the formation of the

complexes involving the atomic Au– negative ion and two water molecules. Water

possesses the unique properties that are rare in other materials and are of biological

importance. These properties are evident in hydrogen bonded environments,

39

particularly in liquid water. In liquid water, the hydrogen bond’s enthalpy is

approximately 0.24 eV and the total dissociation energy is 5.09 eV. Hydrogen bonding

has a direct impact on the change in the Gibbs free energy, G (∆G = ∆H – T∆S),

where H, T, and S represent enthalpy, temperature, and entropy, respectively. When

the atomic negative ion catalyst is introduced into the oxidation of water, there is

breaking of hydrogen bonding. Therefore, the system changes from relative order to

disorder. Hence, the entropy of the system increases, whereas the enthalpy of the

system decreases. The overall result leads to the Gibbs free energy being negative, and

the process results in the spontaneous formation of peroxide. To gain a deeper

understanding of the process of atomic negative ion catalysis, we have also calculated

the rate of a reaction using the Arrhenius equation.107

3.5 Calculations Utilizing CAM Theory, and Transition State Theory

For the calculation of the elastic scattering TCSs for the atoms of interest here

(Au and Pd), we have used the Mulholland formula, in atomic units, implemented

through the complex angular momentum (CAM) description of electron-atom

scattering.99

We have furthermore employed the first principles calculations based on

density functional theory (DFT) and dispersion-corrected DFT approaches. For

structural molecular confirmation, we further utilized the generalized gradient-

corrected approximation (GGA) Perdew–Burke–Ernzernof (PBE) parametrizations of

40

the exchange-correlation functional, along with the double numerical plus polarization

basis set as implemented in the DMol3 package.36,37 The calculations used a tolerance

of 1.0 × 10–3 eV and a basis set cutoff of 0.4 nm for the energy convergence. To assess

the relativistic effects, we have also recalculated the transition barriers with the use of

an all-electron relativistic potential. The resulting energy barriers were found to be

2.63, 3.02, and 3.37 eV for H2O, HDO, and D2O, respectively. However, the use of the

nonrelativistic potential yields 2.69, 3.07, and 3.39 eV, respectively.

The relativistic effects contribution is seen to be small, less than 2.3%. The

dispersion correction is based on the Tkatchenko–Scheffler (T–S) scheme,108 which

exploits the relationship between polarizability and volume. The T–S dispersion

correction takes into account the relative modification in dispersion coefficients of a

variety of atomic bonding by weighting values extracted from the high-quality ab

initio database with atomic volumes derived from the partitioning of the self-

consistent electronic densities. The T–S scheme has been successfully applied to a

variety of systems for a much improved accuracy. The physical effects of electron–

electron correlations and the core polarization interactions are vital for the existence

and stability of typical negative ions.

Consequently, this justifies the adoption of the dispersion-corrected atom-

centered potentials (DCACPs) for use in the present calculations of the atomic Au–

and atomic Pd– negative ion catalysis of light, intermediate, and heavy water to the

corresponding peroxides. The Kohn–Sham formalism of density functional theory

41

(DFT) combined with many popular approximated exchange-correlation functionals

inadequately accounts for the vital London dispersion forces that are of crucial

importance in chemical and biological systems.109

When an effective atom-centered nonlocal term was added to the exchange-

correlation potential in order to cure the lack of London dispersion forces in standard

DFT,110 the corrected generalized gradient approximation DFT calculations yielded

correct equilibrium geometries and dissociation energies of argon–argon, benzene–

benzene, graphite–graphite, and argon–benzene complexes. The DCACPs have been

tested by evaluating the interaction energy of different weakly bound molecular

systems (P2, PH3, and PN dimers) and applied successfully to phosphorus as well.107

The ability of DCACPs to improve the GGA treatment of hydrogen-bonded systems

has been confirmed for the hydrogen bond lengths and binding energies of 20

complexes containing the elements C, H, N, O, and S.112 It was concluded that

DCACPs improve significantly the BLYP description of hydrogen-bonded systems.

The application of the DCACPs to the study of liquid water leads to a softening of

liquid water’s structure, resulting in higher mobility.

This demonstrates that van der Waals interactions are essential in fine-tuning

both structural and dynamical properties of liquid water.113 Also, the convergence to

the correct long-range asymptotic behavior of the multicenter expansion for DCACPs

has been demonstrated in the case of the H2 van der Waals dimer.114 A transition-state

search employing nudged elastic bands facilitates the evaluation of energy

42

barriers.108,115-119 After the initial construction of the reaction path, a transition-state

search followed by a transition state optimization was performed using the linear

synchronous transit (LST) method and the quadratic synchronous transit (QST). In

addition, we have performed a vibrational analysis of the transition state and

confirmed that that the transition state has only one imaginary frequency associated

with it.

3.6 Rate of Reaction Calculation

We have used the Arrhenius equation107 to calculate the rate of a reaction and

compared the number of molecules that can react in the absence and presence of the

atomic negative ion catalyst at constant (room) temperature using the expression (5)

where K is the rate constant, T is the temperature in kelvin, R is the gas constant

8.31 J•mol/K, Ea is the activation energy in J•mol, and A is the frequency factor

which includes factors such as the frequency of collisions and their orientation. It

varies slightly with temperature, although not much. It is often taken as constant

across small temperature changes.

3.7 Results and Data

In our recent study, we investigated the catalytic properties of both atomic Au–

and atomic Pd– ions when catalyzing H2O2 from H2O from the atomic physics

perspective. The low energy electron elastic TCSs for both Au and Pd atoms are

43

characterized by two R–T minima as shown in Figure 12;90,102 the most important one

in this context being the second local R–T minimum. Table 1, taken from reference

85, shows the first R–T minima for Au at 0.692 eV and for Pd at 0.930 eV to be much

higher than the corresponding second minima at 2.057 eV for Au and 3.134 eV for Pd.

As shown in Figure 12, we note that the excited-state TCSs for both atomic Au and

atomic Pd, characterized by huge deep R–T minima, followed by shape resonances,

typify the ground-state electron elastic collision cross sections for many atoms.98, 99

Sitting at 2.262 and 1.948 eV in very close proximity to the second local R–T

minimum of atomic Au are the bound states of the atomic Au– and atomic Pd–

negative ions, respectively (see reference 90). The observed81,89 exceptional catalytic

property of both nano Au and nano Pd has been attributed to the unique positioning of

the second local R–T minimum of atomic Au together with the resonance at around

the same position. The importance of the R–T minimum as already indicated in

reference 85 is to facilitate the creation of new molecules. At this minimum and

appropriate environment, the attachment of the atomic Au– negative ion to (H2O)2 can

result in the formation of the Au–(H2O)2 anion complex at a BE of about 3.20 eV and

characterized as a gold anion interacting with two water molecules.105 The BE value

corresponds to the vertical detachment energy (VDE) of the Au–(H2O)2 anionic

complex.

Figure 12. Electron Affinity calculations in eV within the crossY,Ru, Ag represented in purple, green, pink, and light blue, respectiv

. Electron Affinity calculations in eV within the cross-Y,Ru, Ag represented in purple, green, pink, and light blue, respectiv

44

-sections of Au, Y,Ru, Ag represented in purple, green, pink, and light blue, respectively.

45

Table 1.

Calculated Electron Affinities (EAs) and R–T Minima, in eV, for the Atomic Pd and Au Ground States.

The proposed mechanism of catalysis, using the atomic Au– negative ion

catalyst as an example, is as follows. When an electron collides elastically with a

ground-state neutral gold atom, attachment can result, leading to the formation of a

negative ion resonance due to the formation of compound atomic states. The energy

position of this negative ion resonance corresponds to the stable bound state of the

atomic Au– negative ion formed during the collision as a resonance. The BE of the

atomic Au– negative ion defines the EA of atomic Au. The energy position of the

atomic Au– negative ion is roughly at the second local R–T minimum of the TCS for

Au, where two H2O molecules can attach to the atomic Au– negative ion through

strong resonances with large rates forming the Au–(H2O)2 anion molecular complex.

The atomic Au– negative ion breaks the hydrogen bond and is released after the

chemical reaction. We note that the dissociative energy of the Au–(H2O)2 molecule is

about 3.20 eV. This energy is within the effective range of the second local R–T

46

minimum of the electron elastic TCS of the Au atom. From the studies in references

90 and 91, we understood the exceptional catalytic nature of the gold negative ion

from the atomic physics perspective. What is important and revealing in those studies

90, 91 is the appearance of the ground state of the atomic Au– negative ion at the second

local R–T minimum of the atomic Au TCS. Clearly, this demonstrates the importance

of identifying and delineating the resonances as well as the minima in low-energy

electron elastic collisions with neutral atoms. The same analysis applies to the atomic

Pd– negative ion catalyst. The study 90 cannot be deemed complete if we cannot

understand the problem of anion catalysis of H2O2 from H2O using the atomic Au– and

atomic Pd– negative ions from the theoretical chemistry perspective.

Thus, to address this problem, we have performed dispersion-corrected

density-functional theory for transition state (TS) calculations of the catalytic

properties of the atomic negative ions Au– and Pd– for the oxidation of light,

intermediate, and heavy water to corresponding peroxides. Figures 13 and 14 present

the optimized structures of the reactants, transition states, and products (EP) of

oxidation of intermediate and heavy water, respectively to corresponding peroxides

when the atomic Au– and atomic Pd– negative ion catalysts are absent and when they

are present. The red, white, dark blue, gold, and green spheres represent, respectively,

oxygen, hydrogen, deuterium, gold, and palladium atoms; the transition states and

products of oxidation are in eV. Table 2 summarizes the results of Figures 13 and 14.

47

Table 2.

TS and EP represent, respectively, the calculated transition state and energy of the products, all in eV.

48

Figure 13. Optimized structures of (a) intermediate water (HDO), (b) intermediate water catalyzed by atomic gold negative ion to peroxide, and (c) intermediate water catalyzed by atomic palladium negative ion to peroxide, along with corresponding energies of reactants, transition states (TS), and products (EP). O, H, D, Au–, and Pd– are represented by red, white, dark blue, gold, and green, respectively.

49

Figure 14. Optimized structures of (a) heavy water oxidation, (b) heavy water catalyzed by atomic gold negative ion to peroxide, and (c) heavy water catalyzed by atomic palladium negative ion to peroxide, along with corresponding energies of reactants, transition states (TS), and products (EP). O, D, Au–, and Pd– are represented by red, dark blue, gold, and green, respectively.

50

To determine the effect of the solvent on the calculations, we have used the

COSMO solvation model for the transition state calculations. For example, when the

atomic Au– negative ion catalyst is used, the resultant energy barrier for HDO is 3.02

eV (all-electron relativistic), 3.05 eV (with COSMO), and 3.07 eV (non-relativistic

potential). This demonstrates that the effect of the solvent is negligible.

3.8 Discussions

3.8.1 The Atomic Physics Analysis

We first consider the slow oxidation of light water to peroxide in the absence

of the atomic negative ion catalyst. Upon the addition of the atomic Au– or atomic Pd–

negative ion to H2O, an anionic complex is formed at the second local R–T minimum

of the TCS with a transition state Au–(H2O)2 or Pd–(H2O)2, respectively. Then, the

large EAs of the Au and Pd atoms play the important roles in the breaking up of the

complex to form the atomic Au– or atomic Pd– negative ion and H2O2 products. These

atomic Au– and atomic Pd– negative ions weaken/break the hydrogen bonding in the

H2O, thereby allowing the additional O2, usually provided by the support to attach to

form the desired H2O2. The same argument applies to the oxidation of HDO and D2O

to HDO2 and D2O2, respectively. Although the first and second local R–T minimum in

the TCSs for both atomic Au and atomic Pd are qualitatively the same, including their

EAs as indicated in Figure 12,90 there is little understanding why the atomic Pd–

negative ion has a higher catalytic activity than the atomic Au– negative ion.

51

3.8.2 Thermodynamics Calculation

The entropy of a system increases as it becomes more disordered. For

convenience, we use the case of H2O as an illustration. In the presence of the atomic

negative ion catalysts such as the atomic Au– negative ion, the complex Au–(H2O)2 is

formed in the transition state. In this anionic complex, the two water molecules are

attached to an atomic Au– negative ion. By definition, a catalyst speeds up the rate of a

reaction by lowering the activation energy without changing the energy of the

reactants and the products. Also, the activation energy is the minimum energy required

for a reaction to take place. We followed the steps below to calculate the percentage of

H-bond strength broken when we applied the Au– (Pd–) anion catalyst:

(1) In liquid water, for example, the energy of attraction between water molecules

(hydrogen bond enthalpy) is approximately 0.24 eV.

(2) Our results indicate that the energy of a product changes from 2.21 to 2.13 eV

when we apply the Au– anion catalyst and to 2.04 eV when we apply the Pd– anion

catalyst.

(3) These results can be explained through the percentage of H-bond strength broken

upon the application of a catalyst to the system.

For example, when the atomic Au– negative ion catalyst is used, we obtain:

The percentage of H-bond strength broken is approximately 30%. A similar result is

obtained when the atomic Pd– negative ion catalyst is used; it is equal to 50%.

Therefore, at least 30% of the H-bond strength is broken when the atomic Au–

52

negative ion catalyst is applied, while the percentage increases to 50% with the

replacement of the atomic Au– negative ion by the atomic Pd– negative ion catalyst in

the oxidation of H2O to H2O2. Indeed, these results are consistent with the findings in

the measurements using nanoAu and nanoPd catalysts.81, 89

With the replacement of H by D in the water molecule, the contribution of H-

bonding and the bond length change. Upon the introduction of the atomic negative

ions into the reactions, the entropy of the system increases and its enthalpy decreases.

This leaves the change in Gibbs free energy more negative, resulting in

thermodynamically favorable formation of peroxide.

Figure 15 contrasts the change in entropy (cal/mol•K) versus temperature, T

(K), for the oxidation of D2O to D2O2. The black, pink, and green curves represent the

oxidation of heavy water to peroxide in the absence of a catalyst, in the presence of

atomic Au–, and in the presence of atomic Pd– negative ion catalysts, respectively.

Clearly, the addition of the atomic Au– negative ion catalyst increases the ∆S

significantly throughout the range of temperatures. This is indicative that the atomic

Au– negative ion catalyst disrupts the D-bond and this disruption increases ∆S with the

temperature as expected. The addition of the atomic Pd– negative ion catalyst to the

D2O increases the change of enthalpy even higher compared to the case when the

atomic Au– negative ion catalyst was added, green curve in Figure 15.

Figure 15. Change in entropy, of D2O to D2O2. The black, red, and green curves represent the oxidation of heavy water to peroxide in the absence of a catalyst and in the presence of atomic Auatomic Pd– negative ion catalysts, respectively.

Figure 15. Change in entropy, ∆S (cal/mol•K), vs temperature, T (K), in the oxidation . The black, red, and green curves represent the oxidation of heavy

water to peroxide in the absence of a catalyst and in the presence of atomic Aunegative ion catalysts, respectively.

53

S (cal/mol•K), vs temperature, T (K), in the oxidation . The black, red, and green curves represent the oxidation of heavy

water to peroxide in the absence of a catalyst and in the presence of atomic Au– and

54

3.8.3 Hydrogen Bonding Calculation

Using the EP differences from Table 2, the hydrogen bonding (HB) of H2O,

HDO, and D2O is calculated utilizing density functional theory; the results are

presented in Table 3. The results indicate that regardless, in the absence and presence

of the atomic negative ion catalysts, the HBs of HDO and D2O are about 10 and

15.6% greater than that for H2O, respectively. The results are consistent with the

previous finding.

Table 3.

Calculated energy barrier, (EB), and hydrogen bonding, (HB), in eV

3.8.4 Rate of Reaction Calculation

In this study, we have also interpreted the energy barrier difference in terms of

the rate of a reaction by using the Arrhenius equation.107 In the equation, the rate of a

reaction depends on the number of collisions, orientation, activation energy, and

temperature. To understand the catalytic effect of the atomic Au– and atomic Pd–

negative ions, we assume the temperature to be 298 K and the activation energy to be

55

in eV. For example, the activation energy and the rate constant upon the action of the

atomic Au– negative ion catalyst is Ea1 = 0.54 eV and 7.58 × 10–10, respectively. Upon

the action of the atomic Pd– negative ion catalyst, these quantities become Ea2 = 0.18

eV and 9.3 × 10–4, respectively, for the oxidation of ordinary water. On the basis of the

extracted energy barriers, one can estimate that there are 1.22 × 106 more molecules

that react in the presence of the atomic Pd– negative ion catalyst compared to the case

when only the atomic Au– negative ion catalyst is present.

To find the energy barrier (EB), the EP was subtracted from the TS (refer to

Table 2), and the results are presented in Table 3. Our results show that, regardless of

whether light or heavy water molecules are being catalyzed, the atomic Au– and

atomic Pd– negative ions catalyze the reactions differently.

For example, if we compare the ratio of EB (no catalyst/ (atomic Au– negative

ion catalyst)), it is equal to 3.2, 2.4, and 1.8 for the oxidation of H2O, HDO, and D2O,

respectively. Also, the ratio of EB (atomic Au– negative ion catalyst/ (atomic Pd–

negative ion catalyst)) is equal to 3.0, 2.6, and 2.1 for the oxidation of H2O, HDO, and

D2O, respectively. It can then be concluded that, although the atomic Au– negative ion

catalyst speeds up the oxidation of water to peroxide, the atomic Pd– negative ion

possesses a higher catalytic activity than the atomic Au– negative ion when catalyzing

light, intermediate, and heavy peroxide consistent with the recent experimental

findings.81,89 Both the negative ion catalysts increase the EB of H2O such that EB

(H2O) < EB(HDO) < EB(D2O).

56

3.8.5 Relativistic Effects

A comment on relativistic effects in the calculations of the structure and the

dynamics of atomic Au and its negative ion Au– is appropriate. In calculating the

electronic structure of atomic Au and atomic Au– and their interactions, relativistic

effects are known to be important; see Gorin and Toste79 and Hakkinen et al.120 and

references therein. However, accounting for relativistic effects does not necessarily

guarantee reliable results, if the crucial electron–electron correlation effects and the

core polarization interaction, vital for the existence and stability of most atomic

negative ions, are not adequately accounted for. Most existing theoretical methods

used for calculating the binding energies of the atomic negative ions, including the

atomic Au– negative ion, are structure-based. Generally, for negative ions, the diffuse

nature of the orbitals translates into the need for an extensive partial wave expansion;

this makes calculations intractable in most cases, particularly for heavy and complex

systems. Thus, one is forced to truncate the expansion and/or selectively include what

are estimated to be the most relevant contributions. This approach is indeed adopted

by nearly all structure-based calculations, and the results obtained through these

methods are often riddled with uncertainties and lack definitiveness for complex

systems, such as the Au and Pt atoms.

57

3.9 Summary and Conclusion

We have performed dispersion-corrected density-functional theory calculations

for transition states to investigate the effect of the atomic Au– and atomic Pd– negative

ion catalysis on the formation of peroxides from H2O, HDO, and D2O. We found that

in all three cases both the atomic Au– and atomic Pd– negative ions exhibit excellent

catalytic properties in the formation of H2O2, HDO2, and D2O2. These atomic negative

ions speed up the reactions by lowering the activation energy, with the atomic Pd–

negative ion accomplishing the catalysis by a factor of about 3 times faster than that

by the atomic Au– negative ion. We have also used the Arrhenius equation to calculate

the rate of reactions and compared the number of molecules that can react in the

absence (presence) of an atomic negative ion catalyst at constant room temperature

and found that about 1.22 × 106 times more molecules react in the presence of the

atomic Pd- negative ion compared to when the atomic Au– negative ion catalyst is

present. Using the EP differences from Table 3, the hydrogen bonding (HB) of H2O,

HDO, and D2O has been calculated.

The results indicate that, regardless of the presence (absence) of the atomic

negative ion catalysts, the HBs of HDO and D2O are about 10 and 15.6% greater than

the HBs of H2O, respectively. Furthermore, upon the introduction of the atomic

negative ion catalysts to the reaction, we find that as we go from D2O → HDO → H2O

the entropy of the system increases while the enthalpy of the system decreases. This

58

leaves the ∆G more negative, resulting in thermodynamically favorable formation of

peroxide.

Previously, it was demonstrated that slow electron-Au and electron-Pd

collisions form the atomic Au– and atomic Pd– negative ions.90 Then, through the

transition state, the anionic molecular complexes Au–(H2O)2 and Pd–(H2O)2 are

formed. In the anionic molecular complex formation, the Au– and Pd– ions break up

the hydrogen bond strength in the two water molecules, permitting the formation of

H2O2, HDO2, and D2O2 in the presence of O2 usually provided by the support. These

results, together with those of references 90−92, now complete the fundamental

understanding of negative ion catalysis, at the atomic physics and chemical reaction

dynamics levels.

Namely, the negative ion resonances formed in the electron elastic scattering

TCSs with neutral atoms and transition state chemistry provide the mechanism for

negative ion catalysis. In conclusion, this important mechanism of negative ion

catalysis can now be used to understand the experiments with ozone gas in the

demonstration that bacteria and viruses were torn apart, with the Ag acting as an

extremely efficient oxidative catalyst.106 The atomic Ag– negative ion binds two water

molecules, breaking the strong hydrogen bonding; the oxygen from the ozone can then

attach to the water molecules to form hydrogen peroxide, the desired oxidation

product for bacteria or viruses destruction.

59

CHAPTER 4

GOLD ANION CATALYSIS OF METHANE TO METHANOL WITHOUT CO2 EMISSION

4.1 Introduction

Considerable efforts continue to be devoted to finding ways to reduce CO2

emissions and atmospheric concentrations. Carbon sequestration, improving the

efficiency of energy use, and reducing the carbon content of fuels are three major

pathways that are currently being pursued to address the stabilization of greenhouse

gas concentrations.125 Carbon sequestration uses various approaches for CO2 capture,

storage, and reuse.125,126 One such process, CO2 mineralization, uses carbonic

anhydrase enzyme to convert dilute, unseparated CO2 to HCO3 and finally to

everlasting calcium and magnesium carbonates. Biogenic methane is another of the

carbon sequestrations; it involves geologic storage of CO2 in depleting and depleted

oil and gas reservoirs, with subsequent conversion of the CO2 to CH4 via designer

microbes or biometric systems that operate above or below ground.125 Common among

many of these concepts is the enhancement of naturally occurring biochemical and

geochemical processes through the identification and replication of natural processes

for the purposes of carbon sequestration.

60

The catalytic partial oxidation of methane into valuable products is of great

scientific importance and considerable industrial, economic, and environmental

interest. However, a great challenge is that in the absence of an appropriate catalyst,

methane undergoes complete combustion yielding carbon dioxide and water at

approximately 340 K with minimal competition with the formation of useful products

that can occur at elevated temperatures. The fundamental ideas of muon-catalyzed

nuclear fusion utilizing a negative muon, a deuteron, and a triton127 are used in the

proposed oxidation of CH4 to methanol for which we have selected the atomic gold

anion as the catalyst. Here we propose the use of the atomic Au− ion catalyst to control

the temperature of the oxidation of methane to methanol around 325 K. This has the

effect of lowering the transition state (TS) by 32 % compared to the case of the

absence of the catalyst for the complete oxidation of methane to methanol without

carbon dioxide emission. We have employed the first principles density functional

theory (DFT) and dispersion-corrected DFT calculations for the transition state on the

Au− ion and analyzed the thermodynamics properties of the reactions as well.

The main motivations for the investigation are: (1) the direct synthesis of H2O2

from H2 and O2 using supported Au, Pd, and Au–Pd nanoparticle catalysts128,129

including the theory130,131 that attributed the catalytic properties of Au and Pd to the

formation of negative ion resonances in low-energy electron elastic total cross sections

(TCSs) for Au and Pd atoms, along with their large electron affinities (EAs); (2) the

61

recent dispersion-corrected density functional theory transition-state calculations

performed on the atomic Au− ion catalysis of water conversion to H2O2, revealing that

the formation of the Au−(H2O)2 anion molecular complex in the transition state

provides the fundamental mechanism for breaking up the hydrogen bonding strength

in the catalysis of H2O2 using the Au− ion.132 It is important to note that the Au− ion is

employed here as a prototype for negatively charged gold clusters or surfaces. The

relatively large binding energy associated with the Au− ion is of fundamental

significance as compared to that of the Au+ ion or the neutral Au atom. Contrary to

bulk gold, nanogold exhibits surprisingly high activity and/or selectivity in the

combustion as well as partial oxidation of various molecules and compounds.133 Since

the publication of the paper,133 there have been considerable research activities on

nanogold, particularly on its catalytic properties.133–143 The mechanisms of charge

transfer135,136 and relativity137 have been advanced as possible explanations for the

excellent catalytic properties of gold nanoparticles.

Recently, the negative ion resonances that characterize the electron elastic

scattering TCSs for atomic Au have been proposed as the fundamental mechanism

driving nanoscale catalysis.130,131 The catalytic combustion of methane, the main

component of natural gas, including its conversion to useful products, has recently

received extensive experimental and theoretical attention because of the potential to

reduce pollutant emissions and synthesize useful chemicals.155–160 A recent

investigation demonstrated the selective conversion of a mixture of methane and

62

oxygen to formaldehyde at temperatures below 250 K through temperature-controlled

Au2+ nanocatalysis.160 Experimentally, it has been established that the Au− anion

interacts with water molecules to form the Au− (H2O)1,2 complexes, causing bond

breaking and with methane to form the Au−(CH4) complex,161 thereby weakening the

C–H bond. Furthermore, the strong interaction between the Au− anion and H2O is

comparable to the hydrogen bonding in H2O and the Au− anion interaction with CH4 is

significant as well, but the Au− ion does not interact with O2.154 These findings154,169

are vital to the fundamental understanding of nanocatalysis using Au nanoparticles. To

our knowledge, our proposed approach is the first to use the Au− negative ion in the

catalytic combustion of methane to useful products without the emission of CO2.

4.2 Reactions and Calculation Method

The complete combustion of methane leads to the formation of carbon dioxide and

water:

CH4 + 2O2 → CO2 + 2H2O (1)

Possible by-products of the partial oxidation of methane are:

CH4 + �

� O2 → CO + 2H2 (2)

CH4 + �

� O2 → CH3OH (3)

CH4 + O2 → H2CO + H2O (4)

CH4 + O2→HCO2H + H2 (5)

63

Generally, there is little competition between the complete oxidation, reaction

(1) and the selective partial oxidation (SPO), reactions (2)-(5), of methane. There are

two reasons why the overall reaction leads to the formation of carbon dioxide and

water: (1) Complete combustion of methane occurs at the lowest temperature

compared to its SPO and (2) the corresponding transition state for reaction (1) is

lowest compared to that of any SPO of methane to the desired products. However, the

atomic Au− negative ion activates molecular oxygen in CH4 and increases the level of

the SPO of methane to produce useful compounds. Here the atomic Au− catalyst is

used to control the oxidation temperature of methane around 325 K to lower the

transition state by about 32 % compared to the case of the absence of the catalyst for

the complete oxidation of methane to methanol and further oxidize methanol to

formaldehyde and formic acid without CO2 emission. We follow exactly the same

procedure as in130,131 when applying the atomic Au− ion catalyst to each of the

reactions (1)-(5).

The proposed mechanism of catalysis using the negative Au− ion catalyst is as

follows. When a slow electron collides elastically with a ground-state neutral gold

atom, attachment can result, leading to the formation of a negative ion resonance due

to the formation of compound atomic states. The energy position of this negative ion

resonance corresponds to the stable bound state of the Au− negative ion formed during

the collision as a resonance. The binding energy of the Au− ion defines the EA of

atomic Au. Theoretically, it has been demonstrated that the EA of Au is right at the

64

absolute minimum or the second R-T minimum (absolute) of the elastic TCS of Au.130,

131,162,163 At this minimum and within the appropriate environment, the attachment of

the Au− negative ion to the CH4 molecule results in the formation of the Au−(CH4)

anionic molecular complex. This complex formation results in the disruption of the

stable C–H bonds in the methane molecule. The attendant change in the Gibbs energy

of the system becomes negative, thereby thermodynamically favoring the formation of

methanol. The Au− ion is released after the chemical reaction. We note that the

dissociative energy of the Au−(CH4) molecular complex is within the second R-T

minimum of the Au elastic TCS.

We have also employed the first principles calculations based on DFT and

dispersion-corrected DFT approaches for the investigation. For geometry optimization

of structural molecular confirmation, we utilized the gradient-corrected Perdew–

Burke–Ernzerfof parameterizations164 of the exchange correlation rectified with the

dispersion corrections.161 The double numerical plus polarization basis set was

employed as implemented in the DMol3 package.166 The dispersion correction method,

coupled to suitable density functional, has been demonstrated to account for the long-

range dispersion forces with remarkable accuracy. We used a tolerance of

1.0 × 10−3 eV for energy convergence. A transition-state search employing nudged

elastic bands facilitates the evaluation of energy barriers.167–169 Finally, the energy of

the transition state and thermodynamic curves of the reactions were calculated from

the DMol3 package.166 As the calculation of the transition barrier depends crucially on

65

the exchange correlation scheme employed, the use of reliable dispersion-corrected

approach is essential. The error in extracting the transition barrier associated with the

transition pathway was estimated to be less than 0.001 eV.167–169

4.3 Results and Discussion

Figures 16-20 present the optimized structures of the reactants, transition

states, and products of oxidation of methane leading to the formation of CO2, CO,

CH3OH, H2CO, and HCO2H, respectively. The data in (a) correspond to the absence

of the Au− ion catalyst while those in (b) are data when the Au− ion catalyst is present.

The red, white, gray, and gold spheres represent respectively oxygen, hydrogen,

carbon, and gold atoms. The TS and EP, both in electron volts, represent respectively

the calculated transition-state energy and the energy of the products. The breaking of

the stable C–H bonds in the methane molecule in the transition state resulting in the

formation of methanol in the presence of O2 is attributed to the formation of the

anionic Au− (CH4) complex. The role of the Au− ion is to disrupt the stable C–H bonds

in the methane molecule, allowing the formation of methanol in the presence of O2. It

is noted that the optimized structure corresponding to the reaction (3), namely the

production of methanol, has the lowest transition-state energy (see Figures. 16(b)-

20(b)). These results are also summarized in Table 4.

Figure 16. Complete oxidation of methane to carbon dioxide and water in the absence (a) and presence (b) of the Auspheres represent respectively oxygen, hydrogen, carbon, and gold atoms.

Figure 17. Oxidation of methane to carbon monoxide and hydrogen gas in the absence (a) and presence (b) of the Auspheres represent respectively oxygen, hydrogen, carbon, and gold atoms.

( )

Complete oxidation of methane to carbon dioxide and water in the absence and presence (b) of the Au− negative ion catalyst. The red, white, gray, and gold

spheres represent respectively oxygen, hydrogen, carbon, and gold atoms.

Oxidation of methane to carbon monoxide and hydrogen gas in the absence and presence (b) of the Au− negative ion catalyst. The red, white, gray, and gold

spheres represent respectively oxygen, hydrogen, carbon, and gold atoms.

( )

66

Complete oxidation of methane to carbon dioxide and water in the absence negative ion catalyst. The red, white, gray, and gold

spheres represent respectively oxygen, hydrogen, carbon, and gold atoms.

Oxidation of methane to carbon monoxide and hydrogen gas in the absence negative ion catalyst. The red, white, gray, and gold

spheres represent respectively oxygen, hydrogen, carbon, and gold atoms.

Figure 18. Oxidation of methane to methanol in the absence (a) and presence (b) of the Au− negative ion catalyst. The red, white, gray, and gold spheres represent respectively oxygen, hydrogen, carbon, and gold atoms.

Figure 19. Oxidation of methane to formaldehyde and water in the absence (a) and presence (b) of the Au− represent respectively oxygen, hydrogen, carbon, and gold atoms.

Oxidation of methane to methanol in the absence (a) and presence (b) of the negative ion catalyst. The red, white, gray, and gold spheres represent

respectively oxygen, hydrogen, carbon, and gold atoms.

Oxidation of methane to formaldehyde and water in the absence (a) and negative ion catalyst. The red, white, gray, and gold spheres

represent respectively oxygen, hydrogen, carbon, and gold atoms.

67

Oxidation of methane to methanol in the absence (a) and presence (b) of the negative ion catalyst. The red, white, gray, and gold spheres represent

Oxidation of methane to formaldehyde and water in the absence (a) and negative ion catalyst. The red, white, gray, and gold spheres

Figure 20. Oxidation of methaand presence (b) of the Auspheres represent respectively oxygen, hydrogen, carbon, and gold atoms.

TS, EP, and T represent, respectively,products and temperature of the reaction.

CH4 + 2O2→CO2 + 2H2O

CH4 + O2→CO + 2H2

CH4 + O2→CH3OH

CH4 + O2→H2CO + H2O

CH4 + O2→HCO2H + H2

Oxidation of methane to formic acid and hydrogen gas in the absence (a) Au− negative ion catalyst. The red, white, gray, and gold

spheres represent respectively oxygen, hydrogen, carbon, and gold atoms.

Table 4

TS, EP, and T represent, respectively, the calculated transition state, energy of the products and temperature of the reaction.

TS (eV) EP (eV) T(K) TS (eV) EP (eV)

No

catalyst

No

catalyst G�=�0

Catalyst

Au−

Catalyst

Au

3.21 −1.23 340 3.22 −1.21

4.47 −1.61 500 3.51 −1.60

4.41 −1.56 475 3.01 −1.56

4.24 −1.42 450 3.29 −1.43 3.98 −1.33 425 3.71 −1.34

68

ne to formic acid and hydrogen gas in the absence (a) negative ion catalyst. The red, white, gray, and gold

spheres represent respectively oxygen, hydrogen, carbon, and gold atoms.

the calculated transition state, energy of the

EP (eV) T(K)

Catalyst

Au−

G�=�0

−1.21 340

−1.60 375

−1.56 325

−1.43 350 −1.34 400

69

4.4 Understanding the Results

Here we discuss the results of the complete oxidation of CH4, reaction (1), and

of the SPO of CH4, reaction (3) as illustrations; the latter analysis also applies to the

remaining reactions. In130,131 we explained the catalytic production of H2O2 from H2O,

using the atomic Au− ion catalyst, in the presence of O2. Similarly, here we first apply

the atomic Au− ion catalyst to the complete oxidation of CH4, reaction (1), and obtain:

Au− (CH4) + 2O2→Au−+2H2O2 + CO2 (6) Au− + 2CH4 + 2O2→Au− (CH4) + 2H2O2 + CO2 (7) Adding the reactions (6) and (7), we get: CH4 + 2O2→2H2O + CO2 (8)

The Au− ion catalyst has changed nothing in the reaction, demonstrating

complete combustion. The results of Table 4 (same TS values for the absence and

presence of the catalyst) and Figures.16 and 20 are illustrations of the complete

combustion process. We note that the purpose of a catalyst is to decrease the reaction

temperature to ambient temperature.170 So, the Au− catalyst cannot be effective since

the 340 K temperature (Table 1) is the ambient temperature for CO2 production.

Next we apply the Au− ion catalyst to the reaction (3) and obtain:

Au− (CH4) + �

�O2→Au− + CH3OH (9)

70

Au− + 2CH4 + �

�O2→Au− (CH4) + CH3OH (10)

Adding the reactions (9) and (10), we have: CH4 +

�

�O2→CH3OH (11)

Contrary to the complete oxidation of methane, reaction (1), the Au− ion

catalyzes the SPO of CH4 to a new product, namely CH3OH without CO2 emission,

reaction (11). As seen from comparing the TSs in column 2 and column 5 of Table 4,

the complete oxidation leaves the TS virtually unchanged when the Au− ion catalyst is

introduced. However, for the case of the SPO of CH4, reaction (3), the TSs are 4.41

and 3.01 eV in the absence and presence of the Au− ion catalyst, respectively. So, no

barrier reduction is a manifestation of the complete oxidation of CH4. For this case the

catalyst has no effect on reaction (1). The obtained results in Table 4 and Figures 16-

20 can be understood from three perspectives: resonance scattering theory,

thermodynamics consideration, and transition-state calculations.

Specifically in Figures 21 (a) and 21 (b) we establish that the reactions

involving the production of methanol, represented by the purple curves, experience a

dramatic reduction in temperature from about 475 K to approximate 350 K due to the

addition of Au-. This a very promising result as we observed a minimization of

temperature needed for the reaction as well as methanol production without CO2

emission as represented by the black curve. There is a temperature threshold of about

50 K between methanol and CO

establishes insight for further research and development of efficie

green energy applications utilizing anionic metal systems.

Figure 21. (a) We show the region Gibbs free energy, ∆G (in electron volts) versus temperature, T (in Kelvin), in the presence of the Au− ion cat(green), and fifth (blackproduction of CH3OH, COtemperatures.(b) Change in the Gibbs free energy (in electron volts) versus temperature, T (in Kelvin), in the absence of the Ausecond (black), third (red), fourth (purplerespectively to the reactions leading to the production of COCH3OH, and CO beyond the optimum temperatures.

4.5 Resonance Scattering A

Most importantly, when a slow electron

stable negative Au− ion is formed almost exactly at the second deep R

the electron elastic scattering TCS of atomic Au.

50 K between methanol and CO2 emission when we add Au-. This is a key result that

insight for further research and development of efficient fuels and other

applications utilizing anionic metal systems.

(a) We show the region −0.50 eV ≤ ∆G ≤ 0.50 eV. Change in the ∆G (in electron volts) versus temperature, T (in Kelvin), in the ion catalyst. The first (purple), second (blue), third (red), fourth

(green), and fifth (black) curves correspond respectively to the reactions leading toOH, CO2, H2CO, CO, and HCO2H beyond the optimum

b) Change in the Gibbs free energy (in electron volts) versus temperature, T (in Kelvin), in the absence of the Au− ion catalyst. The first (blue),

ed), fourth (purple), and fifth (green) curves correspond respectively to the reactions leading to the production of CO2, HCO2H, H

OH, and CO beyond the optimum temperatures.

Scattering Approach

Most importantly, when a slow electron collides elastically with atomic Au, a

ion is formed almost exactly at the second deep R-T minimum of

tic scattering TCS of atomic Au.162,163 The binding energy of this

71

This is a key result that

nt fuels and other

0.50 eV. Change in the G (in electron volts) versus temperature, T (in Kelvin), in the

alyst. The first (purple), second (blue), third (red), fourth ) curves correspond respectively to the reactions leading to the

H beyond the optimum b) Change in the Gibbs free energy (in electron volts) versus

atalyst. The first (blue), ) curves correspond

H, H2CO,

collides elastically with atomic Au, a

T minimum of

The binding energy of this

72

atomic Au− ion has been determined experimentally to be 2.309 eV.161,171,172 This

value also corresponds to the EA of atomic Au. If CH4 is introduced at the second R-T

minimum of the electron elastic TCS of atomic Au, it attaches to the Au− ion forming

the anionic Au−(CH4) molecular complex,154,161 with the vertical detachment energy

(VDE) of 2.34 eV154 (incidentally, the R-T minimum is used in the creation of exotic

molecules such as RbCs170,174). Here we observe the remarkable characteristic of

atomic Au with respect to CH4, namely the EA of Au and the VDE of Au− (CH4) are

in the second R-T minimum of the Au elastic TCS. The interaction between the Au−

ion and CH4 is comparable to the C–H bond strength in CH4.154 Thus the Au− ion

weakens or disrupts the C–H bond in CH4 permitting the formation of CH3OH in the

presence of O2. We note that the interaction between the Au− ion and O2 is weak,154

showing the inertness of the Au− ion toward O2. After the reaction the Au− ion catalyst

is free to catalyze another reaction (the process is similar to the destruction of the

ozone by the Cl− ion). This was the determining factor in our selecting the Au− ion as

our catalyst. In154 it has been remarked that the binding energies of the corresponding

Au neutral complexes are significantly less than those of the anion species (for

example, the complex Au−(H2O) has a binding energy that is more than an order of

magnitude larger compared with that of the neutral Au(H2O) complex.154

73

4.6 Thermodynamics of Reactions

Low-energy chemical reaction dynamics provides the mechanism for making

and breaking bonds. In the CH4 catalysis to CH3OH using the atomic Au− ion, the C–H

bond breaking has been attributed to the formation of the anionic Au− (CH4) molecular

complex. The C–H bonding has a direct effect on the change in the Gibbs free energy,

G (∆G = ∆H − T∆S) where H, T, and S represent respectively the enthalpy,

temperature, and entropy. When the atomic Au− ion is introduced into the oxidation of

CH4, the breaking of the C–H bonding occurs. Therefore, the system changes from

relative order to less order. Hence, the entropy of the system increases, whereas the

enthalpy of the system decreases. The overall process results in the Gibbs free energy

being negative, resulting in the spontaneous formation of methanol. To gain a deeper

understanding of the process of atomic Au− ion catalysis, the rate of the reaction was

calculated using Arrhenius Equation.175 In Figure 16, the ∆G versus T for all the

reactions (1)-(5) is depicted. What is remarkable about the effect of the Au− ion

catalyst on the SPO of CH4 to CH3OH and the complete oxidation of CH4 is that

whereas in the absence of the Au− ion catalyst, the production of methanol is at a

much higher temperature (Table 4 and Figure 21(b)). However, the introduction of the

Au− ion catalyst into the reaction (3) dramatically impacts the rate of the reaction,

lowering the temperature at which ∆G = 0, from 475 to 325 K (Table 4 and Figure

21 (b)); this temperature is lower than that for the emission of CO2 (340 K). Indeed the

74

Au− catalyst is incredibly effective in catalyzing the conversion of CH4 to CH3OH

without the emission of CO2.

4.7 Transition-State Calculations

Figures 16(a), (b) present, respectively, in the absence and presence of the Au−

catalyst the TSs and EPs for the complete oxidation of CH4 to CO2. As already

indicated, it is seen from both the figures that the TSs in the absence and presence of

the Au− ionic catalyst are virtually the same. Also the EPs differ only slightly. These

results represent the signature of the complete combustion of CH4. Henceforth, they

will be used as the benchmark for assessing the SPO of the various reactions (1)-(5).

Figures 17(a), (b) displays the calculated TSs and EPs, in the absence and presence of

the Au− ionic catalyst, respectively, for the SPO of CH4 to CO  +  2H2, reaction (2).

Without the Au− ionic catalyst, the TS is 4.47 eV (Figure 17(a), while when the Au−

ionic catalysts is present the TS drops down to 3.51 eV (Figure 17(b)). This is to be

expected since the role of the catalyst is to reduce the barrier. Figures 18(a), (b)

presents respectively the data without and with the Au− ion catalyst for the SPO of

CH4 to methane, reaction (3). The introduction of the Au− ionic catalyst drops down

the TS from 4.41 eV (Figure 19(a)) to 3.01 eV (Figure 19(b)). We note that this

dramatic reduction of the TS of the reaction (3) in the presence of the Au− ionic

catalyst to a value below that of the complete oxidation of CH4 is the main result. It

75

represents a significant accomplishment in the field of catalysis using the Au− ionic

catalyst. The EPs are the same in both Figures 19(a), (b) as expected.

The results for the SPO of CH4 to H2CO + H2O without and with the Au−

ion catalyst are plotted, respectively in Figures 20(a), (b). Just as for the reactions (2)

and (3), given in Figures 18(b) and 19(b), the Au− ionic catalyst reduces the barrier

significantly. However, the TS of 3.29 eV shown in Figure 20(b) is still slightly higher

than that of the complete oxidation of CH4, reaction (1). Perhaps, another atomic

negative ion such as Pd− or Pt−162 added to the Au− ionic catalyst could reduce further

the TS of 3.29 eV to a value significantly lower than that of the complete oxidation of

CH4. We believe that with a combination of the various atomic negative ion catalysts

(see for example the various figures in reference 162), all the reactions (2)-(5) could

be catalyzed directly as in the case of the reaction (3) without CO2 emission. This calls

for further investigations. Figures 15(a), (b) contrast the results for reaction (5), in the

absence and presence of the Au− ion catalyst, respectively. Interestingly, for this

reaction, the Au− ionic catalyst reduces the TS by a small amount, 3.98 versus 3.71

eV. As expected, the EP remains unchanged in both figures.

Comparing all the results presented in Figures 16-20, it is seen that the Au− ion

catalyst has a dramatic effect on reaction (3). Namely, it reduces the TS of the reaction

to a value below that obtained for the complete oxidation of methane. Hence, our main

focus is on reaction (3). The results of these figures are summarized in Table 1.

Figures 16(a), (b) presents the results of ∆G (in electron volts) versus T (in Kelvin) for

76

the reactions (1)-(5). Figure 16(a) represents the data in the presence of the Au− ionic

catalyst, while Figure 16(b) gives the results in the absence of the catalyst. We focus

our discussion on reactions (1) and (3), namely the complete oxidation of CH4 and the

production of methanol. Note the position of the curve for the complete oxidation of

CH4, represented by the first curve in Figure 16(b), blue, and by the second curve in

Figure 16(a), blue. In Figure 16(b), without the Au− catalyst, the production of the

methanol curve occupies the position 4, purple. However, in the presence of the Au−

catalyst, curve 4 jumps dramatically to position 1(Figure 16(a)) ahead of the CO2

production curve; the temperature at ∆G = 0 is 325 K. This can be compared with

that of the CO2 production at 340 K. Important here is that the CO2 curve does not

change its position from that it occupied in Figure 16(b). This clearly demonstrates the

considerable effect the catalyst has on the methanol production. Again this represents

the main result of this calculation. These data exhibit clearly the extent to which a

reaction has been influenced by the presence of the Au− catalyst. By controlling the

temperature around 325 K, methane can be completely oxidized to methanol, rather

than to carbon dioxide (see Figure 16(a), first graph), and methanol can further oxidize

to formaldehyde and formic acid.

77

4.8 Remarks on the Results

As seen from Table 4, the thermodynamics properties agree excellently with

the transition-state calculations of the complete and selective partial oxidation of

methane. Combustion of methane to carbon dioxide and water in the presence and

absence of the Au− ionic catalyst yields almost the same transition state. However, for

the selective partial oxidation of methane, there is a significant change in the transition

states when we compare the results in the presence and absence of the Au− ionic

catalyst. The introduction of the Au− ion catalyst lowers the transition states for the

formation of CO, CH3OH, H2CO, and HCO2H by 21, 32, 22, and 7 %, respectively.

Also when we compare the transition states in the absence of a catalyst for the

formation of carbon dioxide and methanol, we clearly see that the TS for the formation

of CO2 is smaller than that for the methanol formation. This elucidates why methane

undergoes complete oxidation to carbon dioxide, resulting in the increased pollutant

emissions. However, if the Au− ion catalyst is used, the oxidation of methane favors

the formation of methanol because its TS is lower than that of carbon dioxide. This is

much like the separation of a mixture of alcohol and water through the temperature

control. In summary, this proposed catalytic process involving the use of the atomic

Au− ionic catalyst promises a first and a giant step toward finding and assembling

nanocatalysts atom by atom for various chemical reactions, including the direct partial

oxidation of methane to useful products without CO2 emission. This will certainly

78

address the problem of greenhouse gas emissions, with considerable impact on the

environment.

4.9 Discussion of Results

Nanoparticles are essentially a small cluster of atoms; here we are dealing with

a single atom (more specifically, its negative ion). The origin of the catalytic activity

of supported gold nanoparticles is still not fully understood.176 Turner et al.176

investigated the catalytic behavior of very small size (approximately 1.4 nm) gold

nanoparticles obtained from atomic gold clusters. They speculated that the remarkable

catalytic behavior of the atomic nanoparticles was due partly to the strong electronic

interaction between the gold and the titanium dioxide support. Here we use atomic

gold and atomic gold anion, such as used in the experiment of Zheng et al.161, which

are obtained from laser-ablated gold foil. This completely avoids any complication

associated with the support. In130,131 we have used a similar analysis to understand the

experiments128,129 on the catalysis of H2O2 from H2O using Au and Pd nanoparticles.

This investigation could also help toward understanding the issue of the support since

our approach uses simply atoms and atomic anions. As pointed out in references130

and 131, our approach worked for the catalysis of H2O to H2O2 using the atomic Au−

catalyst for the reasons: the large EA of atomic Au, the presence of the second deep R-

T minimum in the electron elastic scattering TCS for atomic Au, and the existence of

the VDE for the anionic Au− (H2O) complex within this R-T minimum. For CH4

79

catalysis the first two conditions still hold. However, the VDE (2.34 eV)157 of the

anionic Au− (CH4) complex is still within this second R-T minimum of the Au elastic

TCS.

To get a sense of how the proposed mechanism might be affected when small

clusters are used rather than the atoms, we recently used density functional theory to

investigate the structure and dynamics of small clusters of 2, 3, 4, and 5 Pt atoms;173

the geometric optimization was achieved using the DMol3 package under the

generalized gradient approximation with the Perdew–Wang exchange correlation

functional.162

The electron affinities for the clusters were evaluated and compared with

measurement and other theoretical calculations. Our calculated EAs were found to be

closer to the measurement, demonstrating the importance of careful geometric

optimization of the structures. Furthermore, the EAs for the clusters did not deviate

significantly from that of the atom. This implies that the proposed mechanism would

still be applicable to small clusters. However, we do not know yet how far this would

hold as the cluster size is increased beyond 5 atoms. Importantly, Hakkinen et al.174

investigated the VDE for Au7 ; they found that the calculated VDE varied between

2.75 and 3.57 eV, with their value being 3.46 eV which agrees well with the

experimental value of 3.5 eV cited in reference 174. Even for a cluster of this size, our

analysis would work because the VDE of Au7 is still within the effective range of the

second R-T minimum of the Au elastic TCS.126,127To firm this, we would need to

80

calculate the electron elastic TCS for the Au7 cluster and identify the R-T minimum

and the various resonances.162

This will certainly be one of our future research projects. Finally, the present

paper could also lead to a better understanding of the role of the noble metal particle

(Au) size and the TiO2 polymorph in the catalytic production of H2 from ethanol.178

Notably, Au nanoparticles of size in the range 3–12 nm were found to be particularly

photo-reactive.

4.10 Conclusion

The atomic Au− ionic catalyst is found to reduce the optimum temperature for

the SPO of methane to about 325 K for CH3OH production. Consequently, in the

presence of the atomic Au− ion catalyst, by controlling the temperature around 325 K,

methane can be completely oxidized to methanol without the emission of the CO2,

thereby broadening considerably the scope of gold’s applications. Using the Au− ion

as the catalyst essentially disrupts the C–H bonding in CH4 oxidation through the ionic

Au− (CH4) molecular formation, thereby eliminating the competition from the carbon

dioxide formation. We conclude by recommending that the negative ions of the atoms

such as those in162 be investigated individually or in combinations for possible

catalytic activities in the selective partial oxidation of methane; the Pt− negative ion

will accomplish similar results as the Au− ionic catalyst.

81

REFERENCES

1. Geim, A. K.; Novoselov, K. S. Nat. Mater. 2007, 6, 183. 2. Rao, C. N. R.; Sood, A. K.; Subrahmanyam, K. S.; Govindaraj, A. Angew.

Chem., Int. Ed. 2009, 48, 7752. 3. Neto, A. H. C.; Guinea, F.; Peres, N. M. R.; Novoselov, K. S.,Geim, A.

K.;Rev. Mod. Phys. 2009, 81, 109. 4. Allen, M. J.; Tung, V. C.; Kaner, R. B. Chem. Rev. 2010, 110, 132. 5. Oostinga, J. B.; Heersche, H. B.; Liu, X.; Morpurgo, A. F.; Vandersypen, L.

M. Nat. Mater. 2008, 7, 151. 6. Berger, C.; Song, Z.; Li, X.; Wu, X.; Brown, N.; Naud, C.; Mayou,, D.; Li, T.;

Hass, J.; Marchenkov, A. N.; Conrad, E. H.; First, P. N.; de Heer, W. A. Science 2006, 312, 1191.

7. Leenaerts, O.; Partoens, B.; Peeters, F. M. Phys. Rev. B 2009, 80, 245422. 8. Li, X.; Wang, X.; Zhang, L.; Lee, S.; Dai, H. Science 2008, 319, 1229. 9. Xia,F.; Farmer,D.B.; Lin, Y.-M.;Avouris, P. Nano Lett. 2010, 10, 715. 10. Wang, X.; Ouyang, Y.; Li, X.; Wang, H.; Guo, J.; Dai, H. Phys. Rev. Lett.

2008, 100, 206803. 11. Castro, E. V.; Novoselov, K. S.; Morozov, S. V.; Peres, N. M. R.; dos Santos,

J. M. B. L.; Nilsson, J.; Guinea, F.; Geim, A. K.; Neto, A. H. C. Phys. Rev.

Lett. 2007, 99, 216802. 12. Zhang, Y.; Tang, T.-T.; Girit, C.; Hao, Z.; Martin, M. C.; Zettl, A.; Crommie,

M. F.; Shen, Y. R.; Wang, F. Nature 2009, 459, 820. 13. Mak, K. F.; Lui, C. H.; Shan, J.; Heinz, T. F. Phys. Rev. Lett. 2009,102,

256405. 14. Samarakoon, D. K.; Wang, X.-Q. ACS Nano 2010, 4, 4126. 15. Elias, D. C.;

Nair, R. R.; Mohiuddin, T. M. G.; Morozov, S. V.; Blake, P.; Halsall, M. P.; Ferrari, A. C.; Boukhvalov, D. W.; Katsnelson, M. I.; Geim, A. K.; Novoselov, K. S. Science 2009, 323, 610.

16. Liu, L.-H.; Yan, M.-D. Nano Lett. 2009, 9, 3375. 17. Chen, Z.; Nagase, S.; Hirsch, A.; Haddon, R. C.; Thiel, W.; Schleyer, P. v. R.

Angew. Chem., Int. Ed. 2004, 4, 1552. 18. Zhao, J.; Chen, Z.; Zhou, Z.; Park, H.; Schleyer, P. v. R.; Lu, J. P.

ChemPhysChem 2005, 6, 598. 19. Nduwimana, A.; Wang, X.-Q. ACS Nano 2009, 3, 1995. 20. He, H.; Gao, C. Chem. Mater. 2010, 22, 5054. 21. Quintana, M.; Spyrou, K.; Marek Grzelczak, M.; Browne, W. R.;

Rudolf, P.; Prato, M. ACS Nano 2010, 4, 3527.

82

22. Choi, J.; Kim, K.-J.; Kim, B.; Lee, H.; Kim, S. J. Phys. Chem. C 2009, 113, 9433.

23. Liu, L.-H.; Lerner, M. M.; Yan, M. Nano Lett. 2010, 10, 3754. 24. Partoens, B.; Peeters, F. M. Phys. Rev. B 2006, 74, 075404. 25. Flores, M. Z. S.; Autreto, P. A. S.; Legoas, S. B.; Galvao, D. S.

Nanotechnology 2009, 20, 465704. 26. Li, Y.; Zhou, Z.; Shen, P.; Chen, Z. Chem. Commun. 2010, 46, 3672. 27. Li, Y.; Zhou, Z.; Shen, P.; Chen, Z. J. Phys. Chem. C 2009, 113, 15043. 28. Li, Y.; Zhou, Z.; Shen, P.; Chen, Z. ACS Nano 2009, 3, 1952. 29. Samarakoon, D. K.; Wang, X.-Q. ACS Nano 2009, 3, 4017. 30. Lee, Y.-S.; Marzari, N. Phys. Rev. Lett. 2006, 97, 116801. 31. Park, H.; Zhao, J.; Lu, J.-P. Nano Lett. 2006, 6, 916. 32. Ogunro, O. O.; Wang, X.-Q. Nano Lett. 2009, 9, 1034. 33. Ogunro, O. O.; Karunwi, K.; Khan, I. M.; Wang, X.-Q. J. Phys. Chem. Lett.

2010, 1, 704. 34. Suggs, K.; Wang, X.-Q. Nanoscale 2010, 2, 385. 35. Denis, P. A.; Iribarne, F. J. Phys. Chem. C 2010, 115, 195. 36. DMol3; Accelrys Software Inc.: San Diego, CA, 2010. 37. Perdew, Perdew, J. P.; Burke, K.; Ernzerhof, M. Phys. Rev. Lett. 1996, 77,

3865. 38. Chen, Z. H.; Lin, Y. M.; Rooks, M. J.; Avouris, P. Graphene Nano-Ribbon

Electronics. Physica E 2007, 40, 228–232. 39. Wang, X.; Ouyang, Y.; Li, X.; Wang, H.; Guo, J.; Dai, H. Room-Temperature

All-Semiconducting Sub-10-nm Graphene Nanoribbon Field-Effect Transistors. Phys. Rev. Lett. 2008,100, 206803.

40. Koenig, S. P.; Boddeti, N. G.; Dunn, M. L.; Bunch, J. S. Ultrastrong Adhesion of Graphene Membranes. Nat. Nanotechnol. 2010, 6, 543–546.

41. Collins, W. R.; Lewandowski, W.; Schmois, E.; Walish, J.; Swager, T. M. Functionalized Graphenes and Thermoplastic Nanocomposites Based upon Expanded Graphite Oxide. Angew. Chem., Int. Ed. 2011, 50, 8848–8852.

42. Sprinkle, M.; Ruan, M.; Hu, Y.; Hankinson, J.; Rubio-Roy, M.; Zhang, B.; Wu, X.; Berger, C.; de Heer, W. A. Scalable Templated Growth of Graphene Nanoribbons on SiC. Nat. Nanotechnol. 2010, 5, 727–731.

43. Wang, X.; Dai, H. Etching and Narrowing of Graphene from the Edges. Nat.

Chem. 2010, 2, 661–665. 44. Han, M. Y.; Ozyilmaz, B.; Zhang, Y. B.; Kim, P. Energy Band- Gap Engineering of Graphene Nanoribbons. Phys. Rev. Lett. 2007, 98, 206805.

45. Tapaszto, L.; Dobrik, G.; Lambin, P.; Biro, L. P. Tailoring the Atomic Structure of Graphene Nanoribbons by ScanningTunneling Microscope Lithography. Nat. Nanotechnol.2008, 3, 397–401.

83

46. Datta, S. S.; Strachan, D. R.; Khamis, S. M.; Johnson, A. T. C.Crystallographic Etching of Few-Layer Graphene. NanoLett. 2008, 8, 1912–1915.

47. Kim, K.; Sussman, A.; Zettl, A. Graphene Nanoribbons Obtained by Electrically Unwrapping Carbon Nanotubes. ACS Nano 2010, 4, 1362–1366.

48. Elias, A. L.; Botello-Mendez, A. R.; Meneses-Rodriguez, D.; Jehova-Gonzalez, V.; Ramirez-Gonzalez, D.; Ci, L.; Munoz-Sandoval, E.; Ajayan, P. M.; Terrones, H.; Terrones, M. Longitudinal Cutting of Pure and Doped Carbon Nanotubes To Form Graphitic Nanoribbons Using Metal Clusters as Nanoscalpels. Nano Lett. 2010, 10, 366–372.

49. Kosynkin, D. V.; Higginbotham, A. L.; Sinitskii, A.; Lomeda, J. R.; Dimiev, A.; Price, B. K.; Tour, J. M. Longitudinal Unzipping of Carbon Nanotubes To Form Graphene Nanoribbons. Nature 2009, 458, 872–876.

50. Higginbotham, A. L.; Kosynkin, D. V.; Sinitskii, A.; Sun, Z.; Tour, J. M. Lower-Defect Graphene Oxide Nanoribbons from Multiwalled Carbon Nanotubes. ACS Nano 2010, 4,2059–2069.

51. Xu, J.; Wang, K.; Zu, S.-Z.; Han, B.-H.; Wei, Z. Hierarchical Nanocomposites of Polyaniline Nanowire Arrays on Graphene Oxide Sheets with Synergistic Effect for Energy Storage. ACS Nano 2010, 4, 5019–5026.

52 Wang, Q. H.; Hersam, M. C. Room-Temperature Molecular-Resolution Characterization of Self-Assembled Organic Monolayers on Epitaxial Graphene. Nat. Chem. 2009, 1,206–211.

53. Zeng, Q.; Cheng, J.; Tang, L.; Liu, X.; Liu, Y.; Li, J.; Jiang, J. Self-Assembled Graphene_Enzyme Hierarchical Nanostructures for Electrochemical Biosensing. Adv. Funct. Mater. 2010, 20, 3366–3372.

54. Zhang, T.; Cheng, Z.; Wang, Y.; Li, Z.; Wang, C.; Li, Y.; Fang, Y. Self-Assembled 1-Octadecanethiol Monolayers on Graphene for Mercury Detection. Nano Lett. 2010, 10, 4738–741.

55. Wang, Q. H.; Hersam, M. C. Characterization and Nanopatterning of Organically Functionalized Graphene with Ultrahigh Vacuum Scanning Tunneling Microscopy. MRS Bull. 2011, 36, 532–542.

56. An, X.; Simmons, T.; Shah, R.; Wolfe, C.; Lewis, K. M.; Washington, M.; Nayak, S. K.; Talapatra, S.; Kar, S. StableAqueous Dispersions of Noncovalently Functionalized Graphene from Graphite and Their Multifunctional High-Performance Applications. Nano Lett. 2010, 10, 4295–4301.

57. Nduwimana, A.; Wang, X. Q. Energy Gaps in Supramolecular Functionalized Graphene Nanoribbons. ACS Nano 2009, 3, 1995–1999.

58. Ogunro, O. O.; Karunwi, K.; Khan, I. M.; Wang, X. Q. Chiral Asymmetry of Helical Polymer Nanowires. J. Phys. Chem. Lett. 2010, 1, 704–707.

59. Gordon, K.; Sannigrahi, B.; McGeady, P.; Wang, X. Q.; Mendenhall, J.; Khan, I. M. J. Biomater. Sci., Polym. Ed. 2009, 20, 2055–2072.

84

60. Sannigrahi, B.; Sil, D.; Baird, B.; Wang, X. Q.; Khan, I. M. Synthesis and Characterization of Bi[2,4-dinitrophenyl (DNP)] Poly(2-methoxystyrene) Functional Polymers.J. Macromol. Sci. A 2008, 45, 664–671.

61. Baird, E. J.; Holowka, D.; Coates, G. W.; Baird, B. Biochemistry 2003, 42, 12739–12748.

62. Koenig, S. P.; Boddeti, N. G.; Dunn, M. L.; Bunch, J. S. Nat. Nanotechnol. 2011, 6, 543–546.

63. Peters, J. E.; Papavassiliou, D. V.; Grady, B. P. UniqueThermal Conductivity Behavior of Single-Walled Nanotube-Polystyrene Composites. Macromolecules 2008, 41, 7274–7277.

64. Manchado, M. A. L.; Valentini, L.; Biagiotti, J.; Kenny, J. M.Thermal and Mechanical Properties of Single-Walled Carbon Nanotubes-Polypropylene Composites Prepared by Melt Processing. Carbon 2005, 43, 1499–1505.

65. Fang, M.; Wang, K.; Lu, H.; Yang, Y.; Nutt, S. Single-Layer Graphene Nanosheets with Controlled Grafting of Polymer Chains. J. Mater. Chem. 2010, 20, 1982–1992.

66. Fang, M.; Wang, K.; Lu, H.; Yang, Y.; Nutt, S. Covalent Polymer Functionalization of Graphene Nanosheets and Mechanical Properties of Composites. J. Mater. Chem. 2009, 19, 7098–7105.

67. Meador, M. A. B.; Capadona, L. A.; McCorkle, L.; Papadopoulos, D. S.; Leventis, N. Chem. Mater. 2007, 19, 2247–2260.

68. Kim, J.; Cote, L. J.; Kim, F.; Huang, J. Visualizing Graphene Based Sheets by Fluorescence Quenching Microscopy. J. Am. Chem. Soc. 2009, 132, 260–267.

69. Chang, H.; Tang, L.; Wang, Y.; Jiang, J.; Li, J. Graphene Fluorescence Resonance Energy Transfer Aptasensor for the Thrombin Detection. Anal.

Chem. 2010, 82, 2341–2346. 70. Topsakal, M.; Ciraci, S. Static Charging of Graphene and Graphite Slabs. Appl.

Phys. Lett. 2011, 98, 131908. 71. Coletti, C.; Riedl, C.; Lee, D. S.; Krauss, B.; Patthey, L.; von Klitzing, K.;

Smet, J. H.; Starke, U. Charge Neutrality and Band-Gap Tuning of Epitaxial Graphene on SiC by Molecular Doping. Phys. Rev. B, 2010, 81, 235401.

72. Bostwick, A.; Speck, F.; Seyller, T.; Horn, K.; Polini, M.; Asgari, R.; MacDonald, A. H.; Rotenberg, E. Observation of Plasmarons in Quasi-Freestanding Doped Graphene. Science, 2010, 328, 999–1002.

73. Koenig, S. P.; Boddeti, N. G.; Dunn, M. L.; Bunch, J. S. Ultrastrong Adhesion of Graphene Membranes. Nat. Nanotechnol 2011, 6, 543–546.

74. Bernhardt, T. M.; Heiz, U.; Landman, U. Chemical and catalytic properties of size-selected free and supported clusters. In Nanocatalysis (Nanoscience and

Technology); Heiz, U., Landman, U., Eds.; Springer: Berlin, 2007; pp 1−244.

85

75. Wong, M. S.; Alvarez, P. J. J.; Fang, Y. L.; Akcin, N.; Nutt, M. O.; Miller, J. T.; Heck, K. N. J. Chem. Technol. Biotechnol. 2009, 84, 158−166. Beltrán, M.

R.; Suárez Raspopov, R.; González, G. Eur. Phys. J. D 2011, 65, 411−420. 76. Thompson, D. T.; Nano Today 2007, 2, 40−43. Bond, G. C.;Louis, C.;

Thompson, D. T.; In Catalysis by Gold Catalytic Science Series; Hutchings, G. J., Ed.; Imperial College Press: London, 2006. von Gynz-Rekowski, F.; Gantefor, G.; Kim, Y. D. Eur. Phys. J. D., 2007, 43, 81−84. van Bokhoven, J.

A. Chimia, 2009, 63, 257−260. Daniel, M. C.; Astruc, D. Chem. Rev. 2004, 104, 293−346.

77. Lang, S. M.; Bernhardt, T. M.; Barnett, R. N.; Landman, U. J. Phys. Chem. C , 115, 6788−6795. Moshfegh, A. Z. J. Phys. B 2009, 42, 233001−233032.

78. Hurst, J. K. Science 2010, 328, 315−316. Gorin, D. J.; Toste, F. D. Nature 2007, 446, 395−40.

79. Kimble, M. L.; Castleman, A. W., Jr.; Mitrić, R.; Burgel, C.; Bonacić-Koutecký, V. J.; Am. Chem. Soc. 2004, 126, 2526−2535.

80. Edwards, J. K.; Carley, A. F.; Herzing, A. A.; Kiely, C. J.; Hutchings, G. J. J.

Chem. Soc., Faraday Discuss. 2008, 138, 225−239. 81. Maljusch, A.; Nagaiah, T. C.; Schwamborn, S.; Bron, M.; Schuhmann, W.

Anal. Chem., 2010, 82, 1890−1896. 82. Zhao, Z.-F.; Wu, Z. J.; Zhou, L.-X.; Zhange, M. H.; Li, W.; Tao, K.-Y. Catal.

Commun,. 2008, 9, 2191−2194. 83. Gao, Y.; Huang, W.; Woodford, J.; Wang, L. S.; Zeng, X. C. J. Am. Chem.

Soc. 2009, 131, 9484−9485. 84. Jurgens, B.; Kubel, C.; Schulz, C.; Nowitzki, T.; Zielasek, V.; Bienrt, J.;

Biener, M. M.; Hamza, A. V.; Baumer, M. Gold Bull. 2007, 40/2, 142−149. 85. Tsai, S.-H.; Liu, Y.-H.; Wu, P.-L.; Yeh, C.-S. J. Mater. Chem. 2003, 13,

978−980. 86. Sun, C.-L.; Hsu, Y.-K.; Lin, Y.-G.; Chen, K.-H.; Bock, C.; MacDougall, B.;

Wu, X.; Chen, L.-C. J. Electrochem. Soc. 2009, 156, B1249−B1252. 87. Haruta, C. M. Catal. Today 1997, 36, 153−166. 88. Edwards, J. K.; Solsona, B.; Landon, P.; Carley, A. F.; Herzing, A.; Watanabe,

M.; Kiely, C. J.; Hutchings, G. J. J. Mater. Chem. 2005, 15, 4595−4600. 89. Msezane, A. Z.; Felfli, Z.; Sokolovski, D. J. Phys. B 2010, 43, 201001. 90. Msezane, A. Z.; Felfli, Z.; Sokolovski, D. Europhys. News 2010, 41, 11. 91. Tesfamichael, A.; Suggs, K.; Felfli, Z.; Wang, X.-Q.; Msezane, A. Z.; 2012,

arXiv:1201.2191v1. 92. Ball, P.; Nature (London) 2008, 452, 291−292. Franco, M. I.; Turin, L;

Mershin, A.;Skoulakis, E. M. C.; Proc. Natl. Acad. Sci. U.S.A. 2011, 108, 3797−3802. Ball, P. Nature. DOI: 10.1038/news.2011.39. Published Online: Feb 14, 2011.

93. Debenedetti, P. G.; Stanley, H. E. Phys. Today 2003, 56, 40−46.

86

94. Yoon, J.-S.; Kim, Y.-W.; Kwon, D.-C.; Song, M.-Y.; Chang, W.- S.; Kim, C.-G.; Kumar, V.; Lee, B. J. Rep. Prog. Phys. 2010, 73, 116401−116421.

95. Machida, S.-I.; Hirai, H.; Kawamura, T.; Yamamoto, Y.; Yagi; T.Phys. Rev. B 2011, 83, 144105.

96. Xu, M.; Sebastianelli, F.; Bac ić, Z.; Lawler, R.; Turro, N. J. J. Chem. Phys. 2008, 129, 064313−064321.

97. Gordon, T. H.; Hura, G.; Water structure from scattering experiments and simulation. Chem. Rev. 2002, 102, 2651−2670. Ball, P. Life’s matrix: A biography of water; Farrar, Straus and Giroux: NewYork, 2000, Franks, F., Ed. Water: A Comprehensive Treatise; Plenum Press: New York, 1972−1979; Vols. 1−6.

98. Sokolovski, D.; Felfli, Z.; Ovchinnikov, S. Y.; Macek, J. H.; Msezane, A. Z.; Phys. Rev. A 2007, 76, 012705.

99. Felfli, Z.; Msezane, A. Z.; Sokolovski, D. J. Phys. B 2008, 41, 041001. Msezane, A. Z.; Felfli, Z.; Sokolovski, D. Chem. Phys. Lett. 2008, 456, 96−100.

100. Felfli, Z.; Msezane, A. Z.; Sokolovski, D. Nucl. Instrum. Methods Phys. Res.,

Sect. B 2011, 269, 1046−1052. 101. Felfli, Z.; Msezane, A. Z.; Sokolovski, D. Phys. Rev. A 2009, 79, 012714;

2011, 83, 052705. 102. Felfli, Z.; Msezane, A. Z.; Sokolovski, D. J. Phys. B 2011, 44, 135204; 2012,

45, 045201. 103. Amour, E. A. G. J. Phys.: Conf. Ser. 2010, 225, 012002. 104. Zheng, W.; Li, X.; Eustis, S.; Grubisic, A.; Thomas, O.; De Clercq, H.; Bowen,

K. Chem. Phys. Lett. 2007, 444, 232−236. 105. Davies, R. L.; Etris, S. F. Catal. Today 1997, 36, 107−114.

106. Levine, R. D.; Molecular Reaction Dynamics; Cambridge University Press: United Kingdom, 2005.

107. Tkatchenko, A.; Scheffler, M. Phys. Rev. Lett. 2009, 102, 073005. 108. Lin, I. C.; Rothlisberger U. Chimia 2008, 62, 231−234. 109. von Lilienfeld, O. A.; Tavernelli, I.; Rothlisberger, U.; Sebastiani, D. Phys.

Rev. Lett. 2004, 93, 153004−153007. 110. Cascella, M.; Lin, I. C.; Tavernelli, I.; Rothlisberger, U. J. Chem. Theory

Comput. 2009, 5, 2930−2934. 111. Arey, J. S.; Aeberhard, P.; Lin, I. C.; Rothlisberger, U. J. Phys. Chem. B 2009,

113, 4726−4732. 112. Lin, I.-C.; Seitsonen, A. P.; Coutinho-Neto, M. D.; Tavernelli, I.;

Rothlisberger, U. J. Phys. Chem. B 2009, 13, 1127−1131. 113. Tavernelli, I.; Lin, I.-C.; Rothlisberger, U. Phys. Rev. B 2009, 79, 045106. 114. Perdew, J. P.; Burke, K.; Ernzerhof, M. Phys. Rev. Lett. 1996, 77, 3865−3868. 115. DMol3

; Accelrys Software Inc.: San Diego, CA, 2011.

87

116. Suggs, K.; Reuven, D.; Wang, X.-Q. J. Phys. Chem. C 2011, 115, 3313−3317. 117. Suggs, K.; Person, V.; Wang, X.-Q. Nanoscale 2011, 3, 2465−2468. 118. Samarakoon, D.; Chen, Z.; Nicolas, C.; Wang, X.-Q. Small 2011, 7, 965−969. 119. Hakkinen, H.; Moseler, M.; Landman, U. Phys. Rev. Lett. 2002, 89, 033401. 120. Wu, Z. J.; Kawazoe, Y. Chem. Phys. Lett. 2006, 423, 81−86. 121. Wu, J.; Yuan, J. J. Phys. B 2006, 39, 2493−2503. 122. van der Hart, H.; Laughlin, C.; Hansen, J. E. Phys. Rev. Lett. 1993, 71,

1506−1509. 123. Walter, C. W.; Gibson, N. D.; Li, Y. G.; Matyas, D. J.; Alton, R. M.; Lou, S.

E.; Field, R. L., III; Hanstorp, D.; Lin, P. L.; Beck, D. R.; Phys. Rev. A 2011, 84, 032514.

124. Beecy D.J., Ferrell F.M., and Carey J.K.; In Proceedings of 1st National

Conference on Carbon Sequestration, May 14–17, 2001. 125. Song, C.S.; Global challenges and strategies for control, conversion and

utilization of CO2 for sustainable development involving energy, catalysis, adsorption and chemical processing. Catal Today 2006, 115:2–32.

126. Armour, E.A.G. J Phys Conference Series, 2010, 225:012002 and references therein.

127. Edwards ,J.K.; Carley, A.F.; Herzing, A.A.; Kiely, C.J.; Hutchings, G.J. J

Chem Soc Faraday Discuss 2008, 138:225. 128. Edwards, J.K.; Solsona, B.l Landon,P., Carley; A.F., Herzing; A.,Watanabe;

M., Kiely; C.J., Hutchings, G.J. J Mater Chem, 2005, 15:4595. 129. Msezane, A.Z.; Felfli, Z.; Sokolovski, D. J Phys B 2010, 43:201001. 130. Msezane, A.Z.; Felfli Z.; Sokolovski D. Europhys News 2010, 41:11. 131. Tesfamichael,A.; Suggs, K.; Felfli, Z.; Wang X.-Q.; Msezane, A.Z.; 2012,

arXiv:1201.2191v1. 132. Haruta M. Catal Today 1997, 36:153. 133. Dumur, F.; Guerlin, A.; Dumas, E.; Bertin, D.; Gigmes, D.; Mayer, C.R.Gold

Bull, 2011, 44:119, and references therein. 134. Sanchez A.; Abbet S.; Heiz U.; Schneider, W.D.; Hakkinen, H.; Barnett R.N.;

Landman U. J Phys Chem A 1999, 103:9573. 135. Bernhardt,T.M.; Heiz U.; Landman, U.; Chemical and catalytic properties of

size-selected free and supported clusters. In: Heiz U.; Landman U.; (eds) Nanocatalysis (nanoscience and technology) 2007, Springer, Berlin, pp 1–244.

136. Gorin, D.J.; Toste,F.D. Nature 2007, 446:395. 137. Moshfegh, A.Z. J Phys D 2009, 42:233001. 138. Beltrán, M.R.; Suárez Raspopov, R.; González, G.; Eur Phys J D 2011,

65:411. 139. Thompson, D.T. Nano Today 2007, 2:40. 140. Bond, G.C.; Louis,C.; Thompson, D.T.; In: Hutchings J.; (ed) Catalysis by

gold catalytic science series. 2006, Imperial College Press,London.

88

141. Lim, D.-C.; Hwang, C.-C.; Ganteför G.; Kim Y.D.; Model catalysts of supported Au nanoparticles and mass-selected clusters. Phys Chem Chem Phys

2010, 12:15172. 142. van Bokhoven, J.A. Chimia 2009, 63:25. 143. Daniel, M.-C.; Astruc D. Chem Rev 2004,104:293. 144. Hashmi, A.S.K.; Hutchings G.J.; Gold catalysis. Angew Chem Int Ed 2006,

45:7896. 145. Hashmi, A.S.K.; Gold-catalyzed organic reactions Chem Rev 2007, 107:3180. 146. Jurgens, B.; Kubel, C.; Schulz, C.; Nowitzki, T.; Zielasek, V.; Bienrt, J.,

Biener, M.M.; Hamza A.V.; Baumer M. Gold Bull 2007, 40(2):142. 147. Kimble, M.L.; Castleman, A.W.; Jr, Mitric R.; Burgel, C.; Bonacic-Koutecky

V. J Am Chem Soc, 2004, 126:2526. 148. Liu Y.-C.; Lin L.-H.; Chiu W.-H. J Phys Chem B 2004,108:19237. 149 González ,Orive A.; et. al.;Nanoscale 2011, 3:1708. 150. Wong M.S.; Alvarez P.J.J.; Fang Y.L.; Akcin N.; Nutt M.O.; Miller J.T.;

Heck,K. N.; J Chem Tech Biotech 2009, 84:158. 151. Pretzer, L.A.; Nguyen, Q.X.; Wong,,M.S. J. Phys Chem C 2010, 114:21226. 152. Gao,Y.; Huang,W.; Woodford, J.; Wang, L.-S.; Zeng, X.C. J Am Chem Soc

2009, 131:9484. 153. Sorokin B.; Kudrik E.V.; Bouchu D. Chem Technol 2008, 5:T43. 154. Vafajoo, L.; Sohrabi, M.; Fattahi M. World Academy of Science, Engineering

and Technology, 2011 73:797. 155. Mohr F (ed) Gold chemistry, applications and future directions in the life

sciences, 2009 Wiley, New York. 156. Yuan, J.; Wang, L.; Wang Y. Ind Eng Chem Re,2011 50(10):6513. 157. Chen, W.; et. al.;Catal Today 2009,140:157. 158. Zhang, Q.; He, D.; Zhu, Q. J Nat Gas Chem 2008,17:24. 159. Lang S.M.; Bernhardt T.M.; Barnett R.N.; Landman U. J Phys Chem C

2011,115:6788. 160. Zheng W.; Li X.; Eustis S.; Grubisic A.; Thomas O.; De Clercq H.; Bowen K.

Chem Phys Lett. 2007, 444:232. 161. Felfli Z.; Msezane A.Z.; Sokolovski D. J Phys B 2011, 44:135204. 162. Felfli Z.; Eure A.R.; Msezane A.Z.; Sokolovski D. NIMB 2010,268:1370. 163. Tkatchenko A.; Scheffler M. Phys Rev Lett. 2009, 102:073005. 164. Perdew J.P.; Burke K.; Ernzerhof M. Phys Rev Lett. 1996, 77:3865. 165. DMol3 Accelrys Software Inc., San Diego, 2011. 166. Suggs K.; Reuven D.; Wang X.-Q. J Phys Chem C 2011, 115:3313. 167. Suggs K.; Person V.; Wang X.-Q. Nanoscale 2011, 3:2465. 168. Samarakoon D.; Chen Z.; Nicolas C.; Wang X.-Q. Small 2011,7:965. 169. Nam,T.H.; Dat V.T.; Loan N.T.T.; Radnik J., Roduner E. J Chem 2010,48:149. 170. Hotop H.; Lineberger W.C. J Phys Chem 1985, Ref Data 14:731.

89

171 Andersen T.; Haugen H.K.; Hotop H. J Phys Chem 1999, Ref Data 28:1511. 172. Simoni A.; Launay J.M.; Soldan P. 2009, arXiv:0901.3129v1. 173. Balakrishnan N.; Quéméner G.; Dalgarno A.; Inelastic collisions and chemical . reactions of molecules at ultracold temperatures. In: Stwalley, W.C.; Krems R.V.; Friedrich B.; (eds) Cold molecules: theory, experiment, applications.

CRC Press, BocaRaton, Florida, 2009. 174. Levine, R.D. Molecular reaction dynamics. Cambridge University Press, Cambridge, 2005.

175. Turner, M.; Golovko,V.B.; Vaughan, O.P.H.; Abdulkin P.; Berenguer- Murcia A.; Tikhov M.S.; Johnson, B.F.G.; Lambert R.M. Nature 2008, 454:981. 176. Chen, Z.; Msezane A.Z. Density functional theory investigation of small Pt clusters. Bull Am Phys Soc 2010, 55(57). 177. Hakkinen, H.; Moseler, M.; Landman ,U. Phys Rev 2002, 89:033401. 178. Murdoch, M.; Waterhouse, G.I.N.; Nadeem, M.A.; Metson, J.B.; Keane, M.A.; Howe, R.F.; Llorca, J.; Idriss, H. Nat Chem 2011, 3:489.