PLACE IN RETURN Box to remove this checkout from your record.
To AVOID FINES return on or before date due.
MAY BE RECALLED with earlier due date if requested.

 

DATE DUE DATE DUE DATE DUE

 

 

 

 

 

 

 

 

 

 

 

 

 

 

6/01 cJClRC/DateDuep65-p. 15

 

EXPERIMENTAL AND THEORETICAL STUDIES OF THE ADSORPTION AND
PHOTOCHEMISTRY OF DIBROMODIFLUOROMETHANE (HALON-1202) ON A
MODEL CARBONACEOUS AEROSOL SURFACE
By

Michael John Dorko

A DISSERTATION
Submitted to
Michigan State University
in partial fulfillment of the requirements
for the degree of
DOCTOR OF PHILOSOPHY
Department of Chemistry

2003

ABSTRACT
EXPERIMENTAL AND THEORETICAL STUDIES OF THE ADSORPTION AND
PHOTOCHEMISTRY OF DIBROMODIFLUOROMETHANE (HALON-1202) ON A
MODEL CARBONACEOUS AEROSOL SURFACE
By
Michael John Dorko

The adsorption and photochemistry of dibromodifluoromethane, CFzBrz, on
highly ordered pyrolytic graphite (HOPG) was studied by theoretical and experimental
methods as a model for heterogeneous photochemical reactions occurring on
carbonaceous aerosols in the upper tIOposphere and lower stratosphere. The monolayer-
covered HOPG surface was characterized using temperature-programmed desorption
(TPD), X-ray photoelectron spectroscopy (XPS), and electron energy loss spectroscopy
(EELS) both before and after broadband ultraviolet (UV) irradiation from ~ 225-350 nm.
In the absence of UV radiation, molecular adsorption and desorption of CFzBrz was
observed (~ 140 K) while irradiation of the monolayer resulted in photodissociation of
the adsorbate to yield CFzBr and Br atoms. The most prevalent reaction observed in
these experiments was the recombination of a large fraction of the photolyzed molecules
(~70%) to reform CFzBrz during TPD measurements. Minor channels revealing the
formation of Brz (~6%) and C2F4Br2 (~l.4%) were present as well. Comparison of the
estimated integrated photodissociation cross section (1.9 x 10'19 cm2) to the cross section
calculated from the TPD data (4.5 x 10'19 cm2) suggested that photodissociation of
CF2BI'2 occurred by direct photoabsorption by the adsorbate and not by dissociative

electron attachment mediated by the surface. Differences in the distributions of Br; and

C2F4Br2 when compared to results of gas phase and matrix isolation experiments were

attributed to possible structural constraints or reduced surface diffusion for one or more
of the adsorbates due to the high density of photogenerated species trapped in the surface
layer. Theoretical studies of the preferred adsorption sites and orientations of CFzBrz and
CFzBr on the HOPG surface were examined using a cluster model approach to gain
additional insight into the experimental data. Second-order Moller-Plesset perturbation
theory (MP2) calculations were performed using 6-31G* and 6-31G(2d) basis sets and a
single layer slab approximation consisting of either a four (pyrene) or seven (coronene)
ring polycyclic aromatic hydrocarbon to model the HOPG surface. The preferred
adsorption site for CFzBrz was found to be located at the hollow site of a six-membered
ring for each combination of surfaces and basis sets examined. The interaction between
the molecule and the surface was dipolar in nature with a small amount of charge transfer
occurring from the surface to the adsorbate as determined from a Mulliken population
analysis. For the CFzBr radical, different preferred adsorption sites were found for the
four and seven ring substrates independent of the basis set used. The atop site was the
preferred location on the four ring system while the bridge site was preferred on the
seven ring system. The radical favored these sites because it contains a singly occupied
molecular orbital to which electron density can be added stabilizing it on the surface.
Rotation of both species on the seven ring surface revealed binding energies similar to
those Obtained for the starting geometries suggesting that free rotation occurs at each
adsorption site. These results were used to rationalize the product distributions Obtained
from the TPD experiments and the decrease in background intensity of the EELS data

after UV irradiation.

To the one who brought me into the world and the one who takes care of me now.

ACKNOWLEDGMENTS

As just about everyone knows the completion of a thesis involves more than just
the person whose name appears under the title. It is for these people that I wish to give
thanks.

First and foremost, I have to thank God. There were numerous times along this
journey when I thought I would not make it. Through His strength, guidance, and love, I
found a way to persevere and see this process to completion.

I would like to thank my adviser, Dr. Simon J. Garrett, for allowing me the
opportunity to join the group and giving me the freedom to delve into the world of
theory. His door was always open for helpful discussions, advice, and chats about things
other than research. I can now proudly say that I use “fiddly bits” and “bloody” as a
regular part of my vocabulary.

The members of my committee also deserve a round of thanks as well. Drs.
Marcos Dantus, Lynmarie Posey, and Piotr Piecuch often provided helpful suggestions in
all areas of my graduate career. I am grateful for the time and, at times, overwhelming
amounts of knowledge and advice that they freely gave to me.

I would also like to acknowledge the members of the Garrett group during my
tenure (Todd, Lili, Heather, and Jason) for the numerous conversations that we had about
work, politics, and anything else that was of current interest. Their support and
friendship were valuable during both difficult as well as good times.

All the members of both the machine (Sam, Glenn, and Tom) and electronics

(Dave, Scott, and Ron) shops need a lot of thanking. From laser emergencies to

meltdowns of the liquid nitrogen reservoir, they handled all that I could throw at them
and more in the timeliest of fashions.

Along the way, you meet many people but there are only a few that you can call
truly good friends. I would be extremely remiss if I didn’t thank John and Mike for their
friendship for the last six years. From our exploits in various fantasy leagues, MSU
hockey games with the crew, and discussions during lunches at the International Center, I
feel like I have gained a couple of brothers who I know I can always count on. GO
GREEN! GO WHITE!

My family (both biological and extended) has had an indirect role in the
completion of this thesis as well. Even though they usually had no clue as to what I was
doing and think that graduate students work forty hours a week, they would always
express much needed love and support whenever I would talk to them. I greatly
appreciate all that they have done for me.

Last, but most certainly not least, I must thank my wife Holly for every single
thing she has ever done for me. I know that I’ll never be able to truly express my thanks
enough but I’m happy that she gave me the privilege of having a lifetime with her to try.

With her love and support, I know that anything is possible.

vi

TABLE OF CONTENTS

LIST OF TABLES ............................................................................................................. ix

LIST OF FIGURES ............................................................................................................. x

KEY TO SYMBOLS AND ABBREVIATIONS ............................................................. xiv
CHAPTER 1

INTRODUCTION ............................................................................................................... l

1.1 Heterogeneous Atmospheric Chemistry ............................................................ 1

1.2 Aerosol Properties ............................................................................................. 1

1.3 Role of Carbonaceous Aerosols in Heterogeneous Atmospheric Chernistry.... 5

1.4 Halons ............................................................................................................... 9

1.5 Proposed System of Study .............................................................................. 11

1.6 Heterogeneous Reactions of Br-Containing Molecules .................................. 12

1.7 Theoretical Studies of Heterogeneous Atmospheric Chemistry ..................... 13

References ............................................................................................................. 16
CHAPTER 2

EXPERIMENTAL ............................................................................................................ 22

2.1 Ultrahigh Vacuum (UHV) Apparatus ............................................................. 22

2.2 Sample Preparation and Dosing ...................................................................... 23

2.3 X-ray Photoelectron Spectroscopy (XPS) ....................................................... 23

2.4 Temperature-Programmed Desorption (TPD) ................................................ 26

2.5 Electron Energy Loss Spectroscopy (EELS) .................................................. 30

2.6 Photochemistry ................................................................................................ 31

References ............................................................................................................. 32
CHAPTER 3

THEORETICAL METHODS ........................................................................................... 33

3.1 Introduction ..................................................................................................... 33

3.2 Perturbation Theory ........................................................................................ 33

References ............................................................................................................. 37
CHAPTER 4

ADSORPTION AND PHOTOCHEMISTRY OF CFzBrz (HALON 1202) ON HIGHLY-

ORDERED PYROLYTIC GRAPHITE (HOPG) ............................................................. 38

Abstract ................................................................................................................. 38

4.1 Introduction ..................................................................................................... 39

4.2 Results and Discussion .................................................................................... 40

4.2.1 Temperature-Programmed Desorption (TPD) ................................. 40

4.2.2 X-ray Photoelectron Spectroscopy (XPS) ........................................ 47

4.2.3 Electron Energy Loss Spectroscopy (EELS) ................................... 50

4.2.4 Adsorption and Desorption of CFzBr2(ad)/HOPG ........................... 53

vii

4.2.5 Photolysis of CFzBrz on HOPG ....................................................... 55

4.2.6 Surface Recombination Reactions ................................................... 57

4.2.7 Loss of Fluorine Species .................................................................. 60

4.2.8 Quantification of Surface Reactions ................................................ 62

4.3 Conclusions ..................................................................................................... 65

References ............................................................................................................. 67
CHAPTER 5

THEORETICAL INVESTIGATION OF THE ADSORPTION SITES AND

ORIENTATION OF CFzBrz AND CFzBr ON MODEL GRAPHITE SURFACES ........ 71

Abstract ................................................................................................................. 71

5.1 Introduction ..................................................................................................... 72

5.2 Computational Details ..................................................................................... 74

5.3 Results and Discussion .................................................................................... 80

5.3.1 Adsorption Sites and Geometries of CFzBrz .................................... 82

5.3.2 Adsorption Sites and Geometries of CFzBI' ..................................... 96

5.4 Conclusions ................................................................................................... 110

References ........................................................................................................... 1 11
CHAPTER 6

CONCLUSIONS AND FUTURE WORK ..................................................................... 116

6.1 Experimental ................................................................................................. 116

6.2 Theoretical ..................................................................................................... 123

References ........................................................................................................... 127

APPENDIX ..................................................................................................................... 130

APPENDIX A ................................................................................................................. 131

viii

LIST OF TABLES

Table 5.1 Comparison of experimental and calculated frequencies with various basis sets
(absolute and relative differences between calculation and experimental data shown in
brackets) ............................................................................................................................ 81
Table 5.2 Atomic charges, dipole moments and charge differences for CFzBrz on each

surface ............................................................................................................................... 93

Table 5.3 CFzBr atomic charges, dipole moments, and total charges on each surface. 106

ix

LIST OF FIGURES

Images in this dissertation are presented in color.
Figure 1.1 Types and sources of some common atmospheric aerosols .............................. 2
Figure 1.2 Schematic showing the growth modes, sources and size distribution of
atmospheric aerosols ........................................................................................................... 3
Figure 1.3 Formation and composition of carbonaceous aerosols ..................................... 7
Figure 1.4 Schematic of gas-phase and heterogeneous reactions of brominated
compounds in the atmosphere. Thicker lines indicate heterogeneous processes ............. 10
Figure 2.1 Surface sensitivity enhancement by variation of the electron take-off angle. 25
Figure 2.2 Surface coverage, desorption rate, and temperature dependence on time
during TPD ........................................................................................................................ 28
Figure 2.3 TPD simulations of zero (a), first (b), and second (c) order desorption
processes ............................................................................................................................ 29
Figure 4.1 Temperature-programmed desorption data (m/z = 129, CFzBr+) for increasing
exposures of CFzBrz on HOPG. Exposures were (a) 0.4 langmuir, (b) 1.5 langmuir, (c)
2.5 langmuir, (d) 3.6 langmuir, (e) 4.5 langmuir, (f) 5.1 langmuir, (g) 6.1 langmuir, (h)
8.1 langmuir, and (i) 10.1 langmuir. Inset shows integrated area Of the feature at ~ 140 K
(monolayer) as a function of exposure .............................................................................. 41
Figure 4.2 Temperature-programmed desorption data for 4.8 langmuir CFzBrz on HOPG.
Monitored masses were m/z = 31 (CF), 50 (CFf), 79 (Br+), 100 (C2F4+), 129 (CFzBr+),

160 (Brf) and 179 (C2F4Br+) ............................................................................................ 43

Figure 4.3 Temperature-programmed desorption data for 4.8 langmuir CFzBrz on HOPG
after 5 minutes of UV irradiation (filtered Hg-arc, 225-350 nm, incident power 8.6
mW/cmz). Monitored masses same as in Figure 4.2 ........................................................ 45
Figure 4.4 Temperature-programmed desorption data for 4.8 langmuir CFzBrz on HOPG
after 15 minutes of UV irradiation. Conditions same as in Figure 4.3 ............................ 46
Figure 4.5 X-ray photoelectron spectra of F ls region (a) and Br 3d region (b) of CFzBrz
monolayer as a function of UV irradiation. Vertical lines are drawn at 688.4 eV and 71.6
eV BE ................................................................................................................................ 48
Figure 4.6 (a) X-ray photoelectron spectra of the C Is region of bare HOPG (solid line),
4.8 langmuir CFzBrz-covered HOPG (solid circles) and the normalized difference
spectrum (crosses). (b) X-ray photoelectron spectra of the C ls region of 4.8 langmuir
CFzBrz-covered HOPG (solid line), 4.8 langmuir CFzBrz-covered HOPG after 30 min
UV irradiation (solid circles) and the normalized difference spectrum (crosses) ............. 49
Figure 4.7 Integrated area ratios (F ls/C ls and Br 3d/C ls) from X-ray photoelectron
spectroscopy of 4.8 langmuir CFzBrz adsorbed on HOPG, as a function of UV irradiation
time (conditions as in Figure 4.3). The solid curve in the F ls/C ls ratio is a single-
exponential fit to the data. The solid curve in the Br 3d/C ls ratio is a linear fit to the
data .................................................................................................................................... 51
Figure 4.8 High-resolution electron energy loss spectra of (a) clean HOPG, (b) 4.8
langmuir CFzBrz-covered HOPG and (c) 4.8 langmuir CFzBrz-covered HOPG after 30
minute UV irradiation ....................................................................................................... 52
Figure 4.9 Relative percentages of the different reaction channels present on the surface

as a function of UV irradiation. The lines through the data are guides for the eye ......... 63

xi

Figure 5.1 Four (a) and seven ring (b) model graphite surfaces with adsorption positions
indicated. The atop site occurs over a substrate atom, the bridge site is between to
substrate atoms and the hollow site is at the center of a six-membered substrate ring ..... 75
Figure 5.2 CFzBrz on the four ring model surface at the (a) atop, (b) hollow, and (0)
bridge sites. Carbon atoms are green, hydrogen atoms are white, bromine atoms are
purple, and fluorine atoms are gray in all figures ............................................................. 78
Figure 5.3 CFzBr on the four ring model surface at the (a) atop, (b) hollow, and (0)
bridge sites ......................................................................................................................... 79
Figure 5.4 Potential energy curves for CFzBrz approaching the four ring substrate at the
MP2/6-31G* level of theory ............................................................................................. 83
Figure 5.5 Potential energy curves for CFzBrz approaching the four ring substrate at the
MP2/6-31G(2d) level of theory ......................................................................................... 85
Figure 5.6 CFzBrz adsorbed at the (a) atop, (b) hollow, and (c) bridge positions on the
model seven ring surface ................................................................................................... 87
Figure 5.7 Rotated CFzBl‘z adsorbed at the (a) atop, (b) hollow, and (c) bridge positions
on the model seven ring surface ........................................................................................ 88
Figure 5.8 Potential energy curves for CFzBrz approaching the seven ring substrate at the
MP2/6-31G* level of theory ............................................................................................. 89
Figure 5.9 Comparison of the effects of basis sets on the binding energies and adsorption
sites for CFzBrz on the model seven ring surface. The squares denote the atop site, circles
denote the hollow site, and the triangles denote the bridge site ........................................ 91
Figure 5.10 Potential energy curves for CF2Br approaching the four ring substrate at the

MP2/6-31G* level of theory ............................................................................................. 97

xii

Figure 5.11 Potential energy curves for CF2Br approaching the four ring substrate at the
MP2/6-3 1G(2d) level of theory ......................................................................................... 99
Figure 5.12 CFzBr adsorbed at the (a) atop, (b) hollow, and (c) bridge positions on the
model seven ring surface ................................................................................................. 100
Figure 5.13 Rotated CF2Br adsorbed at the (a) atop, (b) hollow, and (c) bridge positions
on the model seven ring surface ...................................................................................... 101
Figure 5.14 Potential energy curves for CFzBr approaching the seven ring substrate at the
MP2/6-31G* level of theory ........................................................................................... 103
Figure 5.15 Comparison of the effects of basis sets on the binding energies and
adsorption sites for CFzBr on the model seven ring surface. The squares denote the atop
site, circles denote the hollow site, and triangles denote the bridge site ......................... 104
Figure 6.1 Band structures of an insulator, metal, semimetal, and semiconductor. Shaded

areas represent filled bands ............................................................................................. 117

xiii

KEY TO SYMBOLS OR ABBREVIATIONS

EC = Elemental Carbon

OC = Organic Carbon

CFCs = Chlorofluorocarbons

HOPG = Highly Ordered Pyrolytic Graphite
UV = Ultraviolet

PSC = Polar Stratospheric Cloud

PAH = Polycyclic Aromatic Hydrocarbon
UHV = Ultrahigh Vacuum

QMS = Quadrupole Mass Spectrometer

LN; = Liquid Nitrogen

XPS = X-ray Photoelectron Spectroscopy
TPD = Temperature-Programmed Desorption
EELS = Electron Energy Loss Spectroscopy
FWHM = Full Width at Half Maximum

CSF = Configuration State Function

HF = Hartree Fock

SCF = Self-Consistent Field

LCAO = Linear Combination of Atomic Orbitals
MO = Molecular Orbital

STO = Slater-Type Orbital

GTO = Gaussian-Type Orbital

xiv

CGTO = Contracted Gaussain-Type Orbital

CI = Configuration Interaction

CISD = Configuration Interaction Singles and Doubles
MPn = M ller-Plesset Perturbation Theory of Order n
CC = Couple Cluster

DFT = Density Functional Theory

LDA = Local Density Approximation

LSDA = Local Spin Density Approximation

GGA = Generalized Gradient Approximation

VDW = Van Der Waals

ODP = Ozone Depletion Potential

BE = Binding Energy

MD = Molecular Dynamics

IRAS = Infrared Absorption Spectroscopy

DEA = Dissociative Electron Attachment

UPS = Ultraviolet Photoelectron Spectroscopy

LDOS = Local Density of States

CASPT2 = Complete Active Space Second-order Perturbation
Theory

MRCISD = Multi-Reference Configuration Interaction Singles and
Doubles

XV

Chapter 1 Introduction

1.1 Heterogeneous Atmospheric Chemistry. The Earth’s atmosphere is a
complex mixture containing all three phases of matter. While the atmosphere is largely
gaseous in nature, a small portion contains liquid and solid particulate material. These
particulates can form stable suspensions in the atmosphere and are known as aerosols.l It
has been determined that reactions occurring in or on aerosols are responsible for such
processes as acid rain2 and the destruction of stratospheric ozone above Antarctica.3 (see
Figure 1.1)4 Also, these particulates have been shown to contribute to such processes as
the scattering or absorption of solar radiation, the formation of cloud condensation nuclei
and can serve as possible sites for heterogeneous chemical reactions."4

A majority of the reactions that occur in the Earth’s atmosphere are between gas-
phase molecules. However, gas-phase molecules can adsorb on the surfaces of liquid
droplets or solid particles where they can undergo reactions with each other or with the
substrate. Reactions of this type are termed heterogeneous because they are limited to
surface reactions as opposed to those that take place in the bulk.5 The mechanisms by
which heterogeneous chemical reactions take place are not as well understood as those in
the gas phase, and a need exists to be able to quantify their role in such processes as the
photochemical production of air pollution, climate changes and catalytic ozone
destruction.l

1.2 Aerosol Properties. Aerosols encompass a wide variety of sizes and
chemical compositions depending upon their source and history in the atmosphere (see

Figures 1.1 and 1.2).1'4 Typical aerosol diameters range in size from ~ 0.002 to ~ 100 um

Combustion Process Emissions
Primary OC-EC

so2
Emissions

HO
2 Gas-Phase

 

Photochemistry
Primary
HCl 3 (IV) H280. H280,
Emissions \
\
HCl 1180;, sto,
- H’ so '2
Cl+ Primary OC—EC ' 3 NH ‘ NH:
Na NH‘K, OH' 3 Emissions
Sea-Salt
Emission
Dust,Fly Ash
Condensible Secondary Metals
. oc
Organics 1 NO{,H*
T Ca+2’ Mg+2
Gas-Phase FefiletC- HNOL‘LQas—Phase
Photochemistry Photochemistry
“\\‘Gaseous Or anics //
. . g Dust,F1y Ash NO? .
Em1881ons Emissions EmlSSlonS

o , 4
Figurel.l Types and sources of some common atmospheric aerosols.

Chemical conversion

 

 

 

 

 

 

 

 

 

 

   

 

 

 

 

 

 

 

of gases to low
WW We
Condeiisation
[Primary vparticles]
Coagulation
Condensation growth
[Chain aggregates duet
+
Chemical conversion Emissions
of gases to low
volatility vapors Sea Spray
_ . Coagulation Volcanoa
I Low volatility I Coagulation Plant p+artlcles
Homogeneous
nucleation /\
I
I
/ Sedimentation
// L L
0.001 0.01 0.1 1 .2 10 100
Particle diameter (um)
Ultra fine 3%.? w: Accumulation Mechanicaly generated
particles | range ' range t aerosol range *
4 Fine particles t 1 Coarse particles —"

 

 

 

Figure 1.2 Schematic showing the growth modes, sources and size distribution of atmospheric
1
aerosols.

in the troposphere (10-15 km altitude) and stratosphere (~ 20-50 km altitude) with the
largest fraction occurring between 0.002 and 10 urn.l The lower bound of this range is
approximate due to the lack of a set definition that distinguishes between a cluster of
molecules and an aerosol6 while the upper bound corresponds to aerosols whose size is
comparable to fine sand or pollen.l

Classification of aerosols according to their size distributions can reveal some
insight into the process by which they were formed and their chemical composition. A
typical classification scheme based on mass or volume divides particles into three
different ranges: nuclei (d < 0.1 um), accumulation (0.1 < d < 1 pm) and coarse (d > 1
um).7 Aerosols that are classified as belonging to the nuclei mode are usually formed by
condensation of hot vapors in high temperature combustion processes and are emitted
directly into the atmosphere (primary aerosol) or are formed from gas-to-particle
conversion (secondary aerosol) depending on atmospheric conditions. They are typically
present in large concentrations (~10-1000 cm’3)4 but account for only a small amount of
the total aerosol mass. These aerosols do not have long atmospheric lifetimes because
they are often involved in processes that form larger aerosols. Hence, their role in
heterogeneous atmospheric processes is likely to be restricted, at best.

Accumulation range aerosols are formed during combustion processes by the
condensation of hot vapors onto nuclei mode aerosols and by coagulation of small nuclei
mode aerosols with themselves or with larger particles."‘7 While high temperature
combustion processes can form both nuclei and accumulation aerosols, the predominant
type that is formed depends upon the type of fuel, combustion method (spark, explosion,

etc.), and the amount of atmospheric mixing the particles undergo after they are formed.1

Aerosols in this size range have been found to contain a significant fraction of organic
matter as well as some soluble inorganics such as ammonium, nitrate, and sulfate ions."4
Accumulation mode aerosols are usually present in smaller concentrations compared to
nuclei mode aerosols; however, accumulation mode aerosols account for a larger fraction
of the total aerosol mass (~50%).l These aerosols tend to have the longest atmospheric
lifetimes because of inefficient methods of particle removal,7 such as washout during
precipitation events or by dry deposition due to eddy diffusion and advection."4 Due to
their long atmospheric lifetimes, these aerosols provide the greatest contribution to
heterogeneous atmospheric chemistry, cloud formation, air pollution, etc.

Aerosols in the coarse particle range are produced by a variety of mechanical
processes such as grinding, erosion, or wind. These aerosols usually contain inorganic
material (metal oxides, heavy metals, sea salt, etc.) due to their method of production.
Coarse aerosols do not have long atmospheric lifetimes because sedimentation and
washout by precipitation occurs rapidly for aerosols of this size. As a result, their
atmospheric concentration is on the order of ~ 1 cm'3 or less and their role in atmospheric
heterogeneous reactions is limited.4

1.3 Role of Carbonaceous Aerosols in Heterogeneous Atmospheric
Chemistry. While a number of chemically diverse aerosols are present in the
atmosphere, the most abundant particulate type over continental landmasses is
carbonaceous aerosolf”9 Two different types of carbonaceous aerosols are present in the
atmosphere: primary (or elemental) and secondary. Primary carbonaceous aerosols are
mainly produced during incomplete combustion of biomass and fossil fuels. Other

sources of primary carbonaceous aerosols come from chemical, geological and biogenic

sources. They are composed of tiny spherules of impure graphite that are emitted directly
into the atmosphere.7 These spherules eventually combine to form the visible clusters
termed soot and smoke (See Figure 1.3).1 Secondary carbonaceous aerosols are formed
by the coagulation of the photo-oxidation products of low vapor pressure hydrocarbons or
from emissions by vegetative matter. As a newly formed primary carbonaceous aerosol
ages in the atmosphere, it will become coated with secondary carbon increasing the
complexity of the surface. The resultant aerosol can then be viewed as consisting of a
small core of elemental carbon covered by a liquid-like layer of secondary carbon.lo A
variety of different functional groups (quinones, carboxylic acids, ethers, etc.)”'12 will be
present on the surface due to both the oxidizing nature of the atmosphere and as a result
of the process by which it was formed.9

The role that carbonaceous aerosols play in heterogeneous atmospheric reactions

”'33 Because

has been the subject of numerous studies over the past two decades.
carbonaceous aerosols are often quite diverse in the atmosphere, graphite and amorphous
carbon have been used as model surfaces in addition to natural aerosol samples. A large
portion of the work done in this area has involved characterizing heterogeneous reactions
of atmospherically important gases (802, 03, HNO3, N02, and H2804) and assessing
what role the surface played in the reaction. For example, studies of the interaction
between 802 and various carbon surfaces revealed that little or no reaction took place
implying that carbonaceous aerosols are probably not a major sink for 802.”29 The

addition of water to this system revealed no change in SO; oxidation but an increase in

the uptake coefficient for H20 was observed. ”'30 Studies of the reaction of ozone with

Solid elemental
carbon spheres

   

': a .:. .0}
Gas phase 4,513"
organics '

Surface-absorbed /
adsorbed organics

. . . . 1
Figure 1.3 Formation and composrtron of carbonaceous aerosols.

carbonaceous surfaces have shown that rapid uptake of O3 occurred and that reaction

with the surface resulted in 02, CO, and C02 being produced as seen in equations 1.1—

1.3.31'33
O3+Csoo.—>02+co(g) (1.1)
203 -i- C500; —) 302 + C500! (1.2)
Csoot oxidized + 03 —> Csoot reduced + 2 C02 (13)

These investigations showed that increasing ozone concentrations poison the surface to
further reaction and a number of surface bound carbonyl-containing species were
produced. Reactions of N02 with various graphite surfaces have shown that uptake of
N02 occurred rapidly with several surface species (C-NOz, C-ONO, and C-N-NOz) being
formed.15'“5'21 Reduction of N02 to HONO has been found to occur on most graphite
surfaces in the presence of water (see equations 1.4—1.5).
NO; + CSOOJHZOWS) —> HONO + CSoot oxidized (1.4)
2N02 + HzOmS) ——> HONO + HNO3 (1.5)
The amount of HONO that was produced varied but was not enough to account for the
high levels of HONO observed in urban air parcelsm’lg‘z"26 Comparison between the
various studies conducted to date is difficult due to the wide variety of model or collected
surfaces that have been used.

While most studies involving heterogeneous reactions on carbonaceous aerosols
have focused on using common atmospheric pollutants, fully halogenated molecules have
been studied less intensively. Chlorofluorocarbons (CFCs) are species that contain only
carbon, chlorine and fluorine and are non-toxic and chemically inert. These species were

used for such applications as propellants, refrigerants, and blowing and cleaning agents

before being restricted by the Montreal Protocol in 1987.34 Chlorofluorocarbons have
been shown to be involved in heterogeneous atmospheric reactions that result in
stratospheric ozone depletion as well as being greenhouse gases with long tropospheric
lifetimes."2 Knorr and co-workers have studied Chlorofluorocarbons (mainly CFzClz and
its analogs) adsorbed on graphite surfaces utilizing X-ray diffraction and infrared
absorption spectroscopy techniques.”39 Their studies were focused mainly on phase
transitions and structural features of CFC monolayers as model two-dimensional
systems. The relevant literature from this group is discussed below.

1.4 Halons. Halons are an additional series Of halogenated molecules with
properties similar to CFCs. These molecules also contain carbon, chlorine, and fluorine
as well as one or more bromine atoms. Halons have been used for fire and explosion
suppression, in gas and oil production, and in military aircraft applications and play a
similar role in the stratosphere as CFCs in terms of catalytic ozone destruction.l Their
removal from the atmosphere occurs by short wavelength photolysis in the upper
troposphere and lower stratosphere. Halons have higher ozone depletion potentials than
CFCs because Br atoms do not react as rapidly with organic molecules as Cl atoms
(these reactions provide for temporary removal of halogen atoms from the ozone
destruction cycle) and have larger photodissociation cross sections in the visible region of
the solar spectrum. A diagram of both the gas-phase and heterogeneous reactions of
brominated compounds in the atmosphere is shown in Figure 1.4.1

Studies Of the interaction of Halon molecules with surfaces have not received as
much attention as those of CFCs. A small number of investigations have focused on the

adsorbate structure and phase transitions Of CF3BI‘.35’40 Knorr and co-workers have

 

 

?? hv
OPP)

 

  

 

HOCI (het)

    
   

 

CIN02 (i161) *
BION02

H20 (het)
HCI (het)
HCI (bet)

BrCI

Figure 1.4 Schematic of gas-phase and heterogeneous reactions of brominated compounds in the
atmosphere. Thicker lines indicate heterogeneous processes.

10

 

 

 

hi

dd
02.;

C0”

performed experiments similar to those for CFCs adsorbed on graphite. It was
determined that CF3Br probably resides with the F3 tripod in contact with the surface and
the monolayer formed a structure that was commensurate with the graphite lattice.35’40
Robinson, et. al. have examined the decomposition of CFzBrz (and CFzClz) on alumina
surfaces at stratospheric temperatures.“ The alumina surfaces were used as a model for
solid-propellant rocket motor exhaust particles. They found that dissociative
chemisorption of both species occurred resulting in a surface-bound halomethyl fragment
and ejection of halogen-containing species during decomposition. Pre-dosing the
alumina surface with water was found to deactivate the surface only moderately with
respect to halogen uptake.

1.5 Proposed System of Study. The system presented in this thesis is the
adsorption and photochemistry of dibromodifluoromethane, CFzBrz, on the surface of
highly ordered pyrolytic graphite (HOPG). The HOPG surface will serve as a model for
primary carbonaceous aerosol as well as a source of sub-vacuum electrons or
photoelectrons upon exposure to ultraviolet (UV) radiation. It is well known that UV
irradiation of metal and semiconductor surfaces produces sub-vacuum or photoelectrons
that may initiate chemical reactions through dissociative electron attachment.42
Conversely, direct absorption of UV radiation by the adsorbate can occur and the
resulting photochemistry will be the similar to that observed in the gas phase. A
determination of which mechanism is operative will help elucidate the possible role of
carbonaceous aerosols in heterogeneous atmospheric chemistry of bromine-containing

compounds. Also, an examination of the nascent photoproducts will reveal if the HOPG

ll

surface traps or releases Br atoms and whether additional reaction products are generated
on the surface.

Dibromodifluoromethane was chosen as the adsorbate in this study for a variety
of reasons. This molecule currently is not regulated by the Montreal Protocol of 1987,34
and its concentration in the atmosphere has been increasing at a rate of 17% a year.43
Additionally, it has a short tropospheric lifetime of ~ 1-3 years, which allows for the
possibility of significant amounts of Br atoms to be introduced into the atmosphere
quickly. The initial step in the photodissociation of this molecule between 200-300 nm is
the cleavage of a C-Br bond by a single photon."""59

CF2Br2 + hv —> CFzBr + Br (1.6)
Additional photodissociation pathways are available that result in the second C-Br bond
breaking and the production of C2F4Br2 and Br2.4648' 50‘55’59 Each of these products can
be involved in the destruction of stratospheric ozone as well.

1.6 Heterogeneous Reactions of Br-Containing Molecules. Numerous studies
have been performed examining the interaction between Br-containing molecules and
such atmospherically important surfaces as polar stratospheric clouds (PSCs) and sulfuric
acid-water surfaces.“67 The main effect Of heterogeneous reactions involving bromine
on these surfaces is to increase the conversion of Cl reservoir species to produce free C1
atoms that participate in catalytic ozone destruction. The Br atoms facilitate this by
forming products that are easily photolyzed by visible solar radiation after reaction occurs

on a surface that contains a chlorine reservoir species or by generating OH radicals that

can react with HCl to produce free Cl atoms.

12

The most common brominated reservoir species in the upper troposphere and
lower stratosphere are HBr, BrO, HOBr, and BrONOz.60 Of these species, HOBr and
BrONOz are the ones most responsible for catalytic destruction of stratospheric ozone.
The formation of BrONOz occurs in the gas phase from the reaction of BrO and N02.
When it becomes adsorbed on a PSC or sulfuric acid-water surface, destruction occurs in
the following manner:

BrONOz + H20 —9 HOBr + HN03 (1.7)
This reaction produces HOBr, which is easily photodissociated in the visible region of the
solar spectrum, providing a source Of OH radicals and Br atoms. The HOBr that is
formed can be heterogeneously converted into BrCl if HCl is present on the PSC.

1.7 Theoretical Studies of Heterogeneous Atmospheric Chemistry. A
majority of theoretical calculations investigating heterogeneous atmospheric reactions
have studied interactions between chlorine—containing species and water clusters as
model PSC surfaces to obtain mechanistic information about processes that are involved
in stratospheric ozone destruction.68’74 McNamara and Hillier examined the hydrolysis

reactions of XON02 (X=Cl, Br) and N205 on bare water clusters and ClONOz and N205

with HCl-containing water clusters as shown in reactions 1.8-1.11.“72
XON02 + H20 —> HOX + HNO3 (1.8)
N205 + H20 -—> 2 HNO; (1.9)
ClONOz + HCl ——> C12 + HN03 (1.10)
N205 + HCl -—> ClNOz + HNO3 (1.11)

They observed that the mechanism for each reaction occurred by coupled proton

transfer/8N2 nucleophilic attack on ClONOz or N205 by a water molecule contained

13

within the cluster. As the size of the water cluster increased, most reactions were
observed to change from a mechanism where molecular products were formed to an ionic
mechanism where solvation of ionic products occurred. It also was observed that the
barrier heights for the reactions decreased with an increasing number of waters in the
cluster and most reactions proceeded spontaneously at stratospherically relevant
temperatures. Solvation of the ions that are formed in the larger water clusters appeared
to facilitate the reactions. Bianco and Hynes studied reactions 1.7 and 1.8 and arrived at
similar conclusions.73 Studies of other chlorine-containing molecules (HOCl, HCl, and
C12) interacting with model PSC surfaces have been performed as well.74 These studies
have shown that HOCl and HCl interact more strongly with the water surface than C12
due to the dipole-dipole interaction present between these species.

Most theoretical studies on graphite surfaces have been limited to determining
binding energies and preferred adsorption sites for atoms and diatomic molecules.”88
Cluster models have been used for the majority of these studies. Graphite surfaces are
typically approximated by the use of a single graphene layer or by a polycyclic aromatic
hydrocarbon (PAH) due to the weak interaction between graphite sheets.75"76‘73'79'8”33'85
Properties calculated using these cluster models often agreed well with experimental
values. To date, there have been no theoretical studies investigating interactions between
brominated molecules and atmospherically significant surfaces.

A variety of studies have been performed using molecular simulations to model
the adsorption of numerous small molecules on graphite surfaces.”101 These adsorption

studies have used extended graphite surfaces, graphite slit pores as well as carbon

nanotubes to address such phenomena as structural and phase transitions of mono- and

14

multi-layers and diffusion on the surface. Pair-wise potentials, usually of the Lennard-
J ones form, were used to describe the interaction between the molecule and the surface as
well as between the carbon atoms within the surface. Results of these simulations often
provided a clearer physical picture of the processes (growth mode, orientational ordering,
etc.) occurring on the surface and compared well with experimental data. Thus far, no
simulations have been carried out on the structure, growth mode, or reactions of Halons

on graphite surfaces.

15

8.

9.

References

. Finlayson-Pitts, B. J.; Pitts, J. N., Jr. Chemistry of the Upper and Lower

Atmosphere; Acdemic Press: San Diego, 2000.

Finlayson-Pitts, B. J.; Pitts, J. N., Jr. Atmospheric Chemistry: Fundamental and
Experimental Techniques; John Wiley and Sons: New York, 1986.

Solomon, S. Nature 1990, 347, 347.

Pandis, S. N.; Wexler, A. S.; Seinfeld, J. H. J. Phys. Chem. 1995, 99, 9646.

. Ravishankara, A. R.; Longfellow, C. A. Phys. Chem. Chem. Phys. 1999, I, 5433.

Wayne, R. P. Chemistry of Atmospheres; Oxford University Press: Oxford, 1991.

Seinfeld, J. H.; Pandis, S. N. Atmospheric Chemistry and Physics; John Wiley and
Sons: New York, 1998.

Andreae, M. O.; Crutzen, P. J. Science 1997, 276, 1052.

Lary, D. J .; Shallcross, D. E.; Toumi, R. J. Geophys. Res. 1999, 104, 15929.

10. McDow, S. R.; Jang, M.; Hong, Y.; Kamens J. Geophys. Res. 1996, 101, 19593.

11. Ahkter, M. S.; Chughtai, A. R.; Smith. D.M. Applied Spec. 1985, 39, 143.

12. Ahkter, M. s.; Chughtai, A. R.; Smith. D.M. Applied Spec. 1985,39, 154.

13. Rogaski, C. A.; Golden, D. M.; Williams, L. R. Geophys. Res. Lett. 1997, 24, 381.

14. Lary, D. J.; Lee, A. M.; Toumi, R.; Newchurch, M. J.; Pirre, M.; Renard, J. B. J.

Geophys. Res. 1997, 102, 3671.

15. Smith, D. M.; Akhter, M. S.; Jassin, J. A.; Sergides, C. A.; Welch, W. F.;

Chughtai, A. R. Aerosol Sci. Tech. 1989, 10, 311.

16. Kirchner, U.; Scheer, V.; Vogt, R. J. Phys. Chem. A 2000, 104, 8908.

17. Saathoff, H.; Naumann, K.-H.; Riemer, N .; Kamm, S.; MOhler, O.; Schurath, U.;

Vogel, H.; Vogel, B. Geophys. Res. Lett. 2001, 28, 1957.

16

18.

19.

20.

21.

22.

23.

24.

25.

26.

27.

28.

29.

30.

31.

32.

33.

34.

35.

Chughtai, A. R.; Jassim, J. A.; Peterson, J. H.; Stedman, D. H.; Smith, D. M.
Aerosol Sci. Tech. 1991, 15, 112.

Schurath, U.; Naumann, K.-H. Pure and Appl. Chem. 1998, 70, 1353.
Al-Abadleh, H. A.; Grassian, V. H. J. Phys. Chem. A 2000, 104, 11926.
Akhter, M. S.; Chughtai, A. R.; Smith, D. M. J. Phys. Chem. 1984, 88, 5334.

Gerecke, A.; Thielmann, A.; Gutzwiller, L.; Rossi, M. J. Geophys. Res. Lett.
1998, 25, 2453.

Kalberer, M.; Tabor, K.; Ammann, M.; Parrat, Y.; Weingartner, E.; Piguet, D.;
ROssler, E.; Jost, D. T.; Tiirler, A.; Gaggeler, H. W.; Baltensperger, U. J. Phys.
Chem. 1996, 100, 15487.

Kleffmann, J .; Becker, K. H.; Lackhoff, M.; Wiesen, P. Phys. Chem. Chem. Phys.
1999, 1, 5443.

Longfellow, C. A.; Ravishankara, A. R.; Hanson, D. R. J. Geophys. Res. 1999,
104, 13833.

Stadler, D.; Rossi, M. J. Phys. Chem. Chem. Phys. 2000, 2, 5420.
Grassian, V. H. J. Phys. Chem. A 2002, 106, 860.

Choi, W.; Leu, M.-T. J. Phys. Chem. A 1998, 102, 7618.
Baldwin, A. C. Int. J. Chem. Kinet. 1982, I4, 269.

Britton, L. G.; Clarke, A. G. Arm. Environ. 1980, 14, 829.

Sergides, C. A.; Jassim, J. A.; Chughtai, A. R.; Smith, D. M. Applied Spec. 1987,
41, 482.

Disselkamp, R. S.; Carpenter, M. A.; Cowin, J. P.; Berkowitz, C. M.; Chapman,
E. G.; Zaveri, R. A.; Laulainen, N. S. J. Geophys. Res. 2000, 105, 9767.

POschl, U.; Letzel, T.; Schauer, C.; Niessner, R. J. Phys. Chem. A 2001, 105,
4029.

United Nations Environmental Program. Montreal Protocol on Substances That
Deplete the Ozone Layer; Montreal, 1987.

Knorr, K. Phys. Rep. 1992, 214, 113.

17

36.

37

38

39.

40.

41.

42.

43.

44.

45.

46.

47.

48.

49.

50.

51.

52.

53

54.

Knorr, K.; Civera-Garcia, E. Surf Sci. 1990, 232, 203.

Nalezinski, R.; Bradshaw, A. M.; Knorr, K. Surf. Sci. 1995, 333, 255.
Volkmann, U. G.; Knorr, K. Phys. Rev. B 1993, 47, 4011.

Warken, A.; Enderle, M.; Knorr, K. Phys. Rev. B 2000, 61, 3028.

FaBbender, S.; Enderle, M.; Knorr, K.; Noh, J. D.; Rieger, H. Phys. Rev. B 2002,
65, 165411.

Robinson, G. N .; Freedman, A.; Kolb, C. E.; Worsnop, D. R.; Geophys. Res. Lett.
1994, 21, 377.

Zhou, X.-L.; Zhu, X.-Y.; White, J. M. Surf. Sci. Rep. 1991, I3, 73.

Fraser, P.J.; Oram, D. E.; Reeves, C. E.; Penkett, S. A.; McCulloch, A. M. J.
Geophys. Res. 1999, 104, 15985.

Jacox, M. E. Chem. Phys. Lett. 1977, 53, 192.
Johnson, C. A. F.; Ross, H. J . J. Chem. Soc., Faraday Trans. 1 1978, 74, 2930.
Sam, C. L.; Yardley, J. T. Chem. Phys. Lett. 1978, 61 , 509.

Wampler, F. B.; Tiee, J. J.; Rice, W. W.; Oldenborg, R. C. J. Chem. Phys. 1979,
71, 3926.

Molina, L. T.; Molina, M. J. J. Phys. Cem. 1983, 87, 1306.
Krajnovich, D.; Zhang, Z.; Butler, L.; Lee, Y. T. J. Phys. Chem. 1984, 88, 4561.
Gosnell, T. R.; Taylor, A .J.; Lyman, J. L. Chem. Phys. 1991, 94, 5949.

Talukdar, R. K.; Vaghijani, G. L.; Ravishankara, A. R. J. Chem. Phys. 1992, 96,
8194.

Vatsa, R. K.; Kumar, A.; Naik, P.D.; Rama Rao, K. V. S.; Mittal, J. P Chem.
Phys. Lett. 1993, 207, 75.

Van Hoeymissen, J .; Uten, W.; Peeters, J. Chem. Phys. Lett. 1994, 226, 159.

Vatsa, R. K.; Kumar, A.; Naik, P.D.; Rama Rao, K. V. S.; Mittal, J. P. Bull.
Chem. Soc. Jpn. 1995, 68, 2817.

18

55.Ta1ukdar, R. K.; Hunter, M.; Warren, R. F.; Burkholder, J. B.; Ravishankara, A.
R. Chem. Phys. Lett. 1996, 262, 669.

56. Dhanya, S.; Saini, R. D.; Naik, P.D.; Vatsa, R. K. Chem. Phys. Lett. 2000, 318,
125.

57. Park, M. S.; Kim, T. K.; Lee, S.-H.; Jung, K.-H.; Volpp, H.-R.; Wolfrum, J. J.
Phys. Chem. A 2001, 105, 5606.

58. Felder, P.; Yang, X.; Baum, G.; Huber, J. R. Israel J. Chem. 1993, 34, 33.

59. Cameron, M. R.; Johns, S. A.; Metha, G. F.; Kable, S. H. Phys. Chem. Chem.
Phys. 2000, 2, 2539.

60. Chaix, L.; Allanic, A.; Rossi, M. J. J. Phys. Chem. A 2000, 104, 7268.

61. Hudson, P.K.; Foster, K. L.; Tolbert, M. A.; George, S. M.; Carlo, S. R.; Grassian,
V. H. J. Phys. Chem. A 2001, 105, 694.

62. Barone, S. B.; Zondlo, M. A.; Tolbert, M. A. J. Phys. Chem. A 1999, 103, 9717.
63. Chu, L.; Chu, L. T. J. Phys. Chem. A 1999, 103, 8640.
64. Chu, L. T.; Chu, L. J. Phys. Chem. A 1999, 103, 384.

65. Lary, D. J.; Chipperfield, M. P.; Toumi, R.; Lenton, T. J. Geophys. Res. 1996,
101, 1489.

66. Gane, M. P.; Williams, N. A.; Sodeau, J. R. J. Phys. Chem. A 2001, 105, 4002.
67. Fan, S.-M; Jacob, D. 1. Nature 1992, 359, 522.

68. McNamara, J. P.; Hillier, I. H. J. Phys. Chem. A 1999, 103, 7310.

69. McNamara, J. P.; Hillier, I. H. J. Phys. Chem. A 2001, 105, 7011.

70. McNamara, J. P.; Hillier, I. H. Phys. Chem. Chem. Phys. 2000, 2, 2503.

71. McNamara, J. P.; Hillier, I. H. J. Phys. Chem. A 2000, 104, 5307.

72. McNamara, J .P.; Tresadem, G.; Hillier, I. H. J. Phys. Chem. A 2000, 104, 4030.
73. Bianco, R.; Hynes, J. T. Int. J. Quant. Chem. 1999, 75, 683.

74. Geiger, F. M.; Hicks, J. M.; de Dios, A. C. J. Phys. Chem. A 1998, 102, 1514.

19

75

76.

77

78

79.

Ishida, H.; Palmer, R. E. Phys. Rev. B 1992, 46, 15484.

Ancilotto, F.; Toigo, F. Phys. Rev. B 1993, 4 7, 13713.

Lamoen, D.; Persson, B. N. J. J. Chem. Phys. 1998, 108, 3332.

Lou, L.; Osterlund, L.; Hellsing, B. J. Chem. Phys. 2000, 112, 4788.

Jeloaica, L.; Sidis, V. Chem. Phys. Lett. 1999, 300, 157.

80. Arellano, J. S.; Molina, L. M.; Rubio, A.; Alonso, J. A. J. Chem. Phys. 2000, 112,

81.

82.

83.

84.

85.

86.

87.

88.

89.

90.

91

92

93

94

95

96.

8114.

Feller, D.; Jordan, K. D. J. Phys. Chem. A 2000, 104, 9971.

Ishikawa, S.; Madjarova, G.; Yamabe, T. J. Phys. Chem. B 2001, 105, 11986.
Sorescu, D. C.; Jordan, K. D.; Avouris, P. J. Phys. Chem. B 2001, 105, 11227.
Lushington, G. H.; Chabalowski, C. F. J. Molec. Struc. 2001, 544, 221.

Tran, F.; Weber, J.; Wesolowski, T. A.; Cheikh, F.; Ellinger, Y.; Pauzat, F. J.
Phys. Chem. B 2002, 106, 8689.

Ferro, Y.; Marinelli, F.; Allouche, A. J. Chem. Phys. 2002, 116, 8124.
Letardi, S.; Celino, M.; Cleri, F.; Rosato, V. Surf Sci. 2002, 496, 33.

Tanaka, H.; El-Merraoui, M.; Steele, W. A.; Kaneko, K. Chem.Phys. Lett. 2002,
352, 334.

Nyman, G.; Holmlid, L. J. Chem. Phys. 1986, 85, 6163.

Hansen, F. Y.; T aub, H. J. Chem. Phys. 1987, 87, 3232.

Hammond, W. R.; Mahanti, S. D. Surf. Sci. 1990, 234, 308.

Hentschke, R.; Schilrmann, B. L.; Rabe, J. P. J. Chem. Phys. 1992, 96, 6213.
Hentschke, R.; Winkler, R. .G. J. Chem.Phys. 1993, 99, 5528.

Pinches, M. R. S.; Tildesley, D. J. Surf Sci. 1996, 367, 177.

Kim, Y. H.; Rec, 1.; Shin, H. K. Chem. Phys. Lett. 1999, 314, 1.

Jackson, B.; Lemoine, D. J. Chem. Phys. 2001, 114, 474.

20

97. Meijer, A. J. H. M.; Farebrother, A. J.; Clary, D. C.; Fisher, A. J. J. Phys. Chem.
A 2001, 105, 2173.

98. Sha, X. Jackson, B; Lemoine, D. J. Chem. Phys. 2002, 116, 7158.
99. Sha, X.; Jackson, B. Surf. Sci. 2002, 496, 318.

100. Kwon, 8.; Russell, J.; Zhao, X.; Vidic, R. D.; Johnson, J. K.; Borguet, E.
Langmuir 2002, I8, 2595.

101. Zhao, X.; Kwon, S.; Vidic, R. D.; Borguet, E.; Johnson, J. K. J. Chem. Phys.
2002, 117, 7719.

21

Chapter 2 Experimental

2.1 Ultrahigh Vacuum (UHV) Apparatus. All experiments described here were
carried out in two connected stainless steel ultrahigh vacuum chambers. Each chamber
was separately pumped by a 270 US titanium ion sublimation pump (Perkin Elmer Ultek
D — 1). The first of the two chambers contained a molecular leak valve (Varian
9515106), a 1 — 200 amu quadrupole mass spectrometer (QMS) (VG Masstorr 200 DX) a
dual anode (Mg/Al K013) X—ray source (VG 3EXR2) and a hemispherical electron
energy analyzer (VG CLAM2). A nude ion gauge was used to monitor the chamber
pressure. The second chamber, which was double u-metal shielded to decrease stray
magnetic fields inside the chamber, contained a high-resolution electron energy loss
Spectrometer (EELS) (L K Technologies LK3000). Baking the entire apparatus at 100 °C
for ~ 48 hours allowed for a base pressure Of ~ 2 x 10’10 Torr to be obtained.

A long-travel manipulator (500 mm) capable of x, y, z translation and 0 rotation
(Thermionics 910438NW) was mated to the first chamber and allowed for mounting of
the sample. The sample mount consisted of a molybdenum block that was secured to a
liquid nitrogen (LNz) cooled copper reservoir by a threaded molybdenum rod and nut.
The copper reservoir was connected to a Macor spacer that was used to provide for both
thermal and electrical isolation of the sample. The sample was secured to the
molybdenum mount by two tungsten clips, under one of which was placed an E-type
thermocouple directly in contact with the sample face. This arrangement insured good
thermal contact between the mount and the sample. The sample could be indirectly

heated to > 800 K by a tungsten filament (Alfa Aesar, 0.25 mm diameter) embedded in

22

the molybdenum mount. Cooling the sample to < 85 K was accomplished by drawing
liquid N2 through the copper reservoir with a diaphragm pump.l

2.2 Sample Preparation and Dosing. A 10 x 10 x 1 mm highly ordered
pyrolytic graphite (HOPG) sample (Grade SPI-2, SPI Supplies) was repeatedly cleaved
using adhesive tape to expose a fresh, visibly flat C(OOOI) surface and then outgassed at
720 K in UHV for 5—6 hours. Such a procedure is known to produce clean, ordered
C(OOOI) surfaces.2 Prior to each day’s experiments, the sample was flashed to ~ 800 K
for 2—3 minutes to remove any accumulated adsorbates. X-ray photoelectron
spectroscopy measurements showed up to 1% oxygen-containing species were present on
the surface after flashing. These species could influence the adsorption and monolayer
formation of CF2Br2 on the HOPG surface. However, reproducible results were obtained
between experiments on subsequent days.

The clean HOPG surface was cooled to 85 i l K and exposed to CF2Br2 (~ 98%,
PCR Research Chemicals, Inc.) by backfilling the UHV apparatus for a predetermined
time using the molecular leak valve. The CF2Br2 was used without further purification.
All CF2Br2 exposures are given in langmuirs (1 langmuir = 1L = 10'6 Torr - s) and are
uncorrected for ion gauge gas sensitivity.

2.3 X-ray Photoelectron spectroscopy (XPS). X-ray photoelectron
spectroscopy was used to determine the chemical composition of the clean and adsorbate-
covered surface. This technique works by irradiating the surface with X-rays of
sufficient energy such that photoelectrons are produced according to equation 2.13

EK=hV—EB-¢ (21)

23

where Ex is the electron kinetic energy, hv is the energy of the X-ray photon, E3 is the
binding energy of the electron in the sample and (l) is the sample work function.
Collection and energy analysis of the photoelectrons reveals the binding energies of the
core levels of the atoms contained within the sample. Because each binding energy can
be uniquely associated with a core level in an atom, XPS can be used to determine both
the number and type of atoms in the sample. Changes in the chemical environment of an
element will produce a concomitant Shift in the core level binding energy relative to that
of a pure sample. This effect is commonly called the core-level shift. Information about
oxidation states and changes in bonding can be obtained by observing these shifts.

X-ray photoelectron spectroscopy is a surface sensitive technique because most of
the photoelectrons are limited to escape from sample depths of a few tens of A due to
inelastic scattering processes within the sample.4 The surface sensitivity can be increased
by changing the surface/sample geometry. This can be appreciated by looking at Figure
2.1.3 The vertical sample depth d, is described by equation 2.2

d = 3 A sinor (2.2)
where A is the electron attenuation length and or is the angle of electron ejection with
respect to the surface plane (take—off angle). The maximum vertical sampling depth will
be 3}. when or = 90°. This configuration yields 95% of the signal intensity. If, however,
the take-off angle is decreased, the vertical sampling depth becomes smaller and only
those atoms contained within this depth will be able to emit photoelectrons. The
reduction in sampling depth provides for the further increase in surface sensitivity. This
angular dependence on the XPS signal is useful for structural determination as well as

“depth profiling” in a sample.

24

+0.

01.

3.1 A

 

hv

Figure 2.1 Surface sensitivity enhancement by variation Of the electron take-off angle.3

25

All XP spectra were collected using the Mg Kong X-ray line (hv = 1253.6 eV)
Operated at 300 W (15kV, 20 mA) and an analyzer pass energy of 100 eV. To maximize
surface sensitivity, photoelectrons were collected at a take-off angle of 35° from the
surface plane. Spectra were referenced to the C Is peak from HOPG at 284.7 eV binding
energy.5 No changes in the photoemission data (C IS, F Is, and Br 3d) were observed
during extended periods (up to 1 hour) of X-ray irradiation and data typically were
acquired in less than 20 minutes. After each spectral acquisition, the surface was flashed
to 450 K and a fresh layer Of CF2Br2 was dosed. A temperature of 450 K was found to be
sufficient for removal of all F and Br from the surface after adsorption and UV exposure
experiments as evidenced from XPS measurements.

2.4 Temperature-Programmed Desorption (TPD). Characterization of the
adsorbate-covered surface was undertaken by TPD. In these experiments, the HOPG
surface was cooled to 85 i l K and the UHV apparatus was backfilled with CF2Br2 to a
4.8 L exposure (~ 1 monolayer). The surface was subsequently heated with a linear
heating rate of 10 K/s over a temperature range of 85 — 450 K. The desorbing molecules
were transmitted to the QMS through a 2 mm diameter aperature in a stainless steel
shroud enclosing the QMS, ionized by impact with 70 eV electrons, and separated
according to their mass—to-charge ratio. The QMS line-of-site was coincident with the
surface normal to insure that the largest number of desorbing molecules would be
collected. The data was then plotted as QMS intensity vs. temperature to determine the
desorption profile and kinetics.

Analysis of TPD data is often done in terms of the Polanyi-Wigner equation6

 

dN
rate = — dta = va NJ“ exp(—Ea /RT) (2.3)

26

where X, is the reaction order from the state a, v, is a frequency factor, Ea is the
desorption energy, N8 is the coverage of molecules in state a, R is the gas constant, T is
the temperature and t is time. A plot of the heating rate, surface coverage and the
desorption rate are shown in Figure 2.27 which simulates a typical TPD experiment. A
majority of TPD analyses focus on Obtaining the reaction order, xa and the desorption
energy, E... A variety of methods have been employed to obtain these parameters.8 The
desorption energy indicates the strength of the interaction between an adsorbate and the
surface. An adsorbate may be physisorbed on the surface, implying a weak adsorbate-
substrate interaction with a low desorption temperature. A chemisorbed species forms a
chemical bond to the surface and desorbs at a higher temperature compared to a
physisorbed molecule. The difference between the two species is not clearly defined but
a value of 40 kJ/mol is Often used. The reaction order is normally an integer such as 0,1,
or 2 but fractional values are possible as well.9 Plots showing typical zero, first, and
second order processes are shown in Figure 2.3 with an E3 of 40 kJ/mol, v0 Of 1 x 1028, v;
of 1 x 10”, v; of 1 x 10", a heating rate of 1.5 K/s, and an N2, of 1 — 6 x 1013
molecules/cmz. Each curve in Figure 2.3 shows an increase in Na starting from 1 x 1013
molecules/cm2 to 6 x 1013 molecules/cm2 in increments of l x 1013 molecules/cmz. The
zero order process Shows a common low-temperature leading edge with increasing
exposure and can be indicative of bulk evaporation of the adsorbate from the surface. In
this case, the rate is independent of coverage.9 Zero-order desorption kinetics also have
been observed for the desorption of adsorbates from 2-D islands modified by attractive

10-12

interadsorbate interactions and for desorption of multilayers.l3 A first order process

can be distinguished by a constant maximum desorption temperature with increasing

27

 

Rate

Coverage

 

 

Temperature

 

 

 

 

 

Time

Figure 2.2 Surface coverage, desorption rate, and temperature dependence on time
during TPD.7

28

 

 

a)

 

Intensity

 

 

 

 

I
100

I
120

I I I
140 160 180

Temperature (K)

I
200

220

Figure 2.3 TPD simulations of zero (a), first (b), and second (c) order desorption processes.

29

coverage, while a second order process has a common high temperature trailing edge
with additional exposure. Bond formation between two adsorbates prior to desorption
from the surface accounts for this high temperature trailing edge feature.

Two of the more common methods used for obtaining X, and E3 from TPD data
are Arrhenius plots and Redhead’s peak maximum method.14 In an Arrhenius plot, a
graph of ln(rate) vs. l/T for x = 0, 1, 2 is made and the order is determined by the value
that yields a linear plot with Ea being obtained from the slope of the line. The Redhead
method is usually only applicable for first-order desorption kinetics and has the following
form

E = RTm[ln(\/I‘m lli) — 3.46] (2.4)
where Tm is the temperature of the peak maximum, B is the heating rate, v is the pre-
exponential factor, and R is the gas constant. General values for v ranging between 108
and 1013 s'1 are chosen and errors of less than 2% are obtained with this method. Many
other methods for analysis of TPD data are available but will not be discussed here.

2.5 Electron Energy Loss Spectroscopy (EELS). Further characterization of
both the clean and adsorbate-covered surface was performed by electron energy loss
(spectroscopy. In EELS, a monoenergetic beam of electrons scatters from a surface and
the scattered electrons are energy analyzed. Most electrons are elastically scattered in the
specular direction (9m = 00."). However, some electrons are inelastically scattered
because they excite vibrational modes in the surface or an adsorbate. These vibrational
modes can only be excited if a component of the dynamic dipole moment lies along the
surface normal. If the dipole lies parallel to the surface plane then it will be canceled by

an equal but opposite image dipole in the surface (surface selection rule).15 The

30

 

IE

dc‘

SC.

fu}

U\
1311
Nif
Hm
Win
ape
mm

the:

inelastically scattered electrons will be collected close to the specular direction due to the
small change in their momentum. In this mechanism, the electron is scattered from long-
range dipole fields that are caused by the vibrations of adsorbates or surface atoms.
While other scattering mechanisms may occur, the long-range dipole scattering
mechanism is assumed to be the dominant scattering mechanism for the experiments
described here.

All EELS experiments were performed in the specular scattering geometry (em =
00... = 55°) with a primary beam energy of 6.09 eV. Typical resolution of the elastically
scattered beam from the clean and adsorbate-covered surface was 51-54 cm'1 (6-7 meV)
full width at half maximum (FWHM). Count rates from the clean HOPG surface were
>106 Hz while those from the adsorbate-covered surface were >105 Hz.

2.6 Photochemistry. The adsorbate-covered surface was exposed to unpolarized
UV radiation from a medium-pressure Hg-arc lamp (Oriel 6286) Operated at 350 W. The
lamp was equipped with a condenser lens and a visible/infrared liquid filter (~ 1 M
NiSO4 solution) that primarily transmitted wavelengths in the range between 225 and 350
nm. The UV radiation was introduced into the UHV apparatus through a fused quartz
window such that its angle of incidence was 45° with respect to the surface normal. An
aperature (12.7 mm diameter) affixed to the window minimized irradiation of the sample
mount. A power of 8.6 mW/cm2 was measured at the sample-lamp distance by a
therrnopile detector. All UV irradiation of the adsorbate-covered surface was performed
at a surface temperature of S 85 K. Upon exposure to UV irradiation, the surface
temperature increased by ~ 6 K. No evidence was Observed for thermally driven

reactions under these modest temperature increases.

31

10.

11.

12.

13.

14.

15.

References

. Bryden, T. R. Ph.D. Thesis, Michigan State University, 2001.

Musket, R. G.; McLean, W.; Colmenares, C. A.; Makowiecki, D. M.; Siekhaus,
W. J. Appl. Surf Sci. 1982, 10, 143.

Briggs, D. in Practical Surface Analysis; Briggs, D.; Seah, M., Eds.; Wiley and
Sons: New York, 1990; Vol. 1.

Woodruff, D. P.; Delchar, T. A. Modern Techniques of Surface Science;
Cambridge University Press: Cambridge, 1994.

Barr, T. L. Modern ES CA: The Principles and Practice of X -ray Photoelectron
Spectroscopy; CRC Press: Boca Raton, FL, 1994.

King, D. A. Surf Sci. 1975, 47, 384.

Falconer, J. L.; Schwarz, J. A. Catal. Rev. Sci. Eng. 1983, 25, 141.
de long, A. M.; Niemantsverdriet, J. W. Surf Sci. 1990, 233, 355.
Vollmer, M.; Trager, F. Surf Sci. 1987, 187,445.

Golze, M.; Grunze, M.; Hirschwald, W. Vacuum 1981, 31, 697.
Niemantsverdriet, J. W.; Markert, K.; Wandelt, K. Appl. Surf Sci. 1988, 31, 211.
Zhou, X.-L.; White, J. M.; Koel, B. E. Surf Sci. 1989, 218, 201.

Yates, J. T., Jr. Methods of Experimental Physics; Academic Press: San Diego,
1985; Vol. 22.

Redhead, P. A. Vacuum 1962, 12, 203.

Pemble, M. in Surface Analysis-The Principle Techniques; Vickerrnan, J. C., Ed.;
Wiley and Sons: West Sussex, 1997.

32

II)

Sr

Chapter 3 Theoretical Methods

3.1 Introduction. Ab initio electronic structure methods were used to Obtain
information about the localized interactions of CF2Br2 and its possible photofragments
with the HOPG surface. Calculations were performed examining the preferred
adsorption sites and geometries of the adsorbed molecules to aid in further explaining the
experimental data. To accomplish this, the electronic SchrOdinger equation must be
solved for the adsorbate/substrate system. The main interaction between the molecules
and the graphite surface occur through dispersion forces. Because these forces are weak
in nature compared to chemical bonds, ab initio methods that describe electron
correlation must be used. The simplest method that accounts for electron correlation is
second order Mallet-Plesset (MP2) perturbation theory.1 A description of the general
method of perturbation theory follows in the next section along with a more detailed
discussion of MP methods.

3.2 Perturbation Theory. In perturbation theory, the system to be studied is
assumed to be slightly perturbed in some manner compared to a known unperturbed
reference system.2 This assumption allows for the solutions of the unperturbed system,
typically from a Hartree-Fock calculation, to be used as a starting point to obtain
solutions for the perturbed system. If this is the case, the Hamiltonian of the perturbed

system can be written as
f1 = [10 +119, (3.1)
where [I 0 is the Hamiltonian Of the unperturbed problem, [-1, is the perturbed

Hamiltonian, and A is a parameter that determines the strength of the perturbation.2 The

solutions to the unperturbed Hamiltonian can be chosen such that they form a complete

33

orthonormal set. For the discussion here, the perturbed Hamiltonian is assumed to be
time-independent and non-degenerate. Values for A. will range between zero and one.
Since A. can vary continuously within this range, the energy and wavefunction for the
system will change in a similar fashion and can be written as Taylor series expansions in
powers of A

E = AOEO +AE, + 22E, + 23E, +

‘1’ = 29% + 2‘11, + 22‘1’2 + 23‘113 + (3.2)
where E0 and ‘P0 are the unperturbed energy and wavefunction when 2t=0 and all other
terms are corrections to the energy and wavefunction. Putting the Taylor series
expansions for the energy and wavefunction into the Schrtidinger equation and collecting

like powers of A (through second order) yields

to: now, = EO‘I’O

A‘: Help, + 119?, = EO‘I’, + 5311,,

A2: now, + Apr, = EO‘I’2 + 13911, + 15311,, (3.3)
The term in A0 is the solution of the unperturbed problem. The equation for the first order

corrections contains two unknowns (E1 and W1) that can be found if the first order

wavefunction is expanded in the complete set of functions produced by the unperturbed

wavefunction (‘1’, = Zed), ). The first and second order energy corrections can be

found by multiplying by (Do on the left of the equations for the Al and 7&2 terms in equation
3.3, respectively, and integrating. The corrections to the wavefunction will be found by

multiplying by <1),- on the left of the equations for the Al and 2.2 terms in equation 3.3,

34

respectively. If the unperturbed Hamiltonian is chosen as a sum of one-electron Fock
operators, the method is known as Moller-Plesset (MP) perturbation theory.1 By making
this choice for the unperturbed Hamiltonian, the sum of the zerO-order energy and first
order correction to the energy will be the Hartree-Fock energy. The higher energy
corrections correspond to wavefunction corrections that contain singly, doubly, etc.
excited Slater determinants relative to the Hartree-Fock reference configuration. These
excited determinants describe the many-electron correlation effects.

The correlation energy is not Obtained until the second order corrections (the
lowest order terms due to doubly excited Slater determinants) are added.2 The second

order energy correction in this case becomes

we <<I>O|I§r,|cr>;b)(c;b

£322

ab
i<j a<b E0 — Er}

 

ri,|¢0)

 

(3.4)

where the indices i and j indicate occupied orbitals and a and b indicate unoccupied or
“virtual” orbitals. The summation is restricted so that each double excitation is counted
only once. The matrix elements between the HF and the doubly excited determinants can
be expressed by two—electron integrals over the molecular orbitals and the total energy for

the system then becomes

 

E(MP2)= SC: "2" [ <¢i¢j ¢°¢°>—<¢i¢il¢b¢a> ]

i<j a<b 8I+8j-8a"8b

 

(3.5)

where 4),. is a molecular orbital and 8,, is the molecular orbital energy. It can be seen that
with this method that there are M4 (M = the number of basis functions in the calculation)
integrals required for the full energy calculation. However, a transformation of the

integrals from the A0 to the MO basis requires M5 integrals to be done and therefore, the

35

MP2 method scales as M5 for any system. MP2 can obtain about 70% of the correlation
energy of a system. Higher orders (MP3, MP4, etc.) will Obtain more correlation energy
but these methods scale as M6 or higher and make these types Of calculations prohibitive
for larger systems. Because MPn methods do not use the variational principle to
calculate the energy of a molecule, there is no upper bound to the energy that can be
calculated. However, MPn methods are size-extensive (MPn energies and properties
scale properly with the system size) and provide for a low cost way of including electron
correlation compared to Configuration Interaction methods. The biggest shortcoming of
perturbation methods is the accuracy with which the HF wavefunction approximates the
true ground state wavefunction of the system. If the HF wavefunction is not a good
starting point for the calculation then convergence may be very slow or not occur at all.
Additional correction terms (MP3, MP4, etc.) may need to be added in order to achieve a
more accurate description of the system. The MP2 method now will be used to calculate
the interaction between CF2Br2 and CF2Br on two model graphite surfaces (pyrene and
coronene) to determine their preferred adsorption sites and orientations. With this
information, a more complete picture of both the pre- and post-irradiated surface can be

obtained.

36

References

1. Simons, J. J. Phys. Chem. 1991, 95, 1017.

2. Jensen, F. Introduction to Computational Chemistry; John Wiley and Sons:
Chichester, 1999.

37

Chapter 4 Adsorption and Photochemistry of CF2Br2 (Halon-1202)
on Highly-Ordered Pyrolytic Graphite (HOPG)

Abstract

The adsorption and photochemistry of dibromodifluoromethane (Halon-1202)
adsorbed on highly-ordered pyrolytic graphite (HOPG) was studied using temperature-
programmed desorption (TPD), X-ray photoelectron spectroscopy (XPS) and electron
energy loss spectroscopy (EELS). Dibromodifluoromethane adsorbs molecularly at 85 K
and the first layer saturates at a fractional coverage of 0.14 i 0.01 ML. Molecular
desorption occurs at ~140 K (monolayer) and ~110 K (multilayers) with desorption
energies of 43.8 :i: 4.6 kJ/mol and 22.4 i 1.9 kJ/mol, respectively. Ultraviolet irradiation
Of the monolayer by a filtered Hg—arc lamp (225-350 nm) resulted in the formation of
CF2Br and Br (major products) and Br2 and C2F4Br2 (minor products). The estimated
integrated dissociation cross-section was 1.9 x 10'19 cm2, close to the calculated value for
CF2Br2(g) (4.5x10'19 cmz). The similarity of these two values implies that
photodissociation of CF2Br2(ad)/HOPG is dominated by direct photoabsorption of the
adsorbate and not electronic processes moderated by the substrate. The influence of the
surface is most clearly observed in the distribution of products measured by TPD. We
attribute the differences observed between adsorbate and gas phase/matrix isolation

experiments to the high density of photogenerated species trapped in the surface layer.

38

4.1 Introduction. During the past two decades, it has become increasingly
apparent that heterogeneous reactions on particulate surfaces play an important role in
atmospheric chemistry. For example, it has been shown that catalytic reactions on polar
stratospheric clouds and nitric acid trihydrate surfaces contribute to the destruction of
stratospheric ozone.” These surfaces provide a medium upon which photoactive
reservoir molecules (the oxidation products of atmospheric halocarbons) are produced
more readily than in the gas phase alone. Photolysis of these species in the stratosphere
provides halogen atoms that can catalyze ozone destruction.

A variety of natural and anthropogenic sources of halogenated molecules in the
atmosphere include Chlorofluorocarbons (refrigerants, foam-blowing agents, solvents)
and fully-halogenated bromomethanes and bromoethanes (Halons, fire retardants). The
presence of brominated molecules in the atmosphere is particularly noteworthy since the
ozone depletion potential (ODP) of these molecules is larger than their chlorinated
analogs.5 The most common brominated molecules are CH3Br (up to about 10 ppt) and
various Halons (total up to about 7 ppt).6 Since removal of Halons occurs exclusively by
short wavelength photolysis, many have relatively long atmospheric lifetimes (5-100
years)7 and so are efficient means of halogen mass transport to the stratosphere. Due to
their role in stratospheric ozone loss, most of these molecules have been restricted by
international agreement.8 Currently, Halon 1202 (CF2Br2) is not regulated by this
agreement. Indeed, the atmospheric concentration of Halon 1202 has recently begun to
rise by significant amounts, prompting debate over its environmental origin.9

The primary goal of the present study was to determine if the photochemistry of

CF2Br2 is altered when it is adsorbed on a model carbonaceous aerosol surface (highly

39

ordered pyrolytic graphite, HOPG). While the photochemistry of halogenated
compounds on ice, sulfuric acid and sea-salt surfaces have been the subject of many

"“0"“ reactions occurring on carbonaceous aerosols have been largely

investigations,
unexplored. These aerosols form the most abundant particulates over continents with the
global load estimated at about 270 x 106 kg.15 Their chemical composition is highly
variable, depending upon their formation process and history in the atmosphere. The so-
called primary organic aerosols (soot, graphitic carbon and carbon black) are structurally
similar to impure graphite and are usually formed as a result of combustion processes.

If enhanced photodissociation cross-sections for brominated adsorbates occur, the
importance Of aerosol chemistry as a net source for bromine atoms in the atmosphere will
increase. In contrast, the ODP for Halons may be decreased if chemisorption of halogen
atoms on the carbon particulate surface creates a net sink. In our experiments we use
HOPG as a model for the surface of primary carbon particulates. Depending upon
energetic coupling between the adsorbate and substrate, the. semi-metallic electronic
structure Of the graphite may significantly perturb the photochemistry of an adsorbed
molecule.l6
4.2 Results and Discussion.

4.2.1 Temperature-Programmed Desorption (TPD). The adsorption behavior
of CF2Br2 on HOPG was studied by TPD. Figure 4.1 shows a series of m/z = 129
(CF2Br+, base peak for CF2Br2) TPD spectra for increasing exposures of CF2Br2. Low
exposures produced a broad desorption peak centered at 135 K, suggesting weak

adsorbate-substrate interaction (physisorption). For progressively higher exposures this

40

 

CFZBrZIHOPG

Integrated Area

 

 

 

o'é'i'é'é'fo
CFzBr2 Dose (Langmuirs)

ion intensity (m/z=129)

 

 

 

JilllllllllllllllllllllIllllI111IJJJJJJllllllllJlljllllllll

100 110 120 130 140 150

 

Temperature (K)

Figure 4.1 Temperature-programmed desorption data (m/z = 129, CF2Br") for increasing
exposures of CF2Br2 on HOPG. Exposures were (a) 0.4 langmuir, (b) 1.5 langmuir,
(c) 2.5 langmuir, (d) 3.6 langmuir, (e) 4.5 langmuir, (f) 5.1 langmuir, (g) 6.1
langmuir, (h) 8.1 langmuir, and (i) 10.1 langmuir. Inset shows integrated area of
the feature at ~ 140 K (monolayer) as a function of exposure.

41

peak grew in area and the peak maximum shifted towards higher temperature, reaching
138 K after 4.8 L of CF2Br2. This feature displayed a common leading edge with
increasing exposure, indicative of zero-order desorption kinetics. A leading-edge
analysis17 of this feature indicated a desorption energy of 43.8 i 4.6 kJ/mol.

As the CF2Br2 exposure increased, saturation of the peak at 138 K occurred
indicating completion of the first adsorbed layer. The inset of Figure 4.1 Shows a plot of
integrated TPD area for this peak as a function of exposure. The area increased linearly
and then remained approximately constant with additional CF2Br2 exposure. Linear fits
to the increasing and constant regions yielded an intersection point of 4.8 L, and we use
this exposure as equivalent to saturation of the first layer in subsequent discussions.

Exposures > 4.8 L lead to the appearance of a lower temperature desorption peak
centered at 109 K. This peak did not saturate with increasing exposure and also exhibited
a common leading edge. Such zero-order desorption kinetics are characteristic of

18 A leading edge analysis of this peak, gave a desorption energy

multilayer desorption.
of 22.4 i 1.9 kJ/mol.
Figure 4.2 shows TPD spectra for 4.8 L CF2Br2 on HOPG measured at m/z = 31
(CF), 50 (CF2+), 79 (Br+), 100 (C2F4+), 129 (CF2Br+), 160 (Br2+), and 179 (C2F4Br+).
These ions were chosen because they are representative of the fragmentation pattern of

gas phase CF2Br2 and possible photolysis products. Each spectrum was obtained from a

separate CF2Br2 exposure. Except for the m/z = 100 and 179 data, each ion had a similar

42

 

4.8 L CFzBrleOPG

 

 

 

 

 

 

 

 

 

 

 

Unirradiated
"it/L *5
m/z = 50 x5
2‘
.5 _
ac) m/z - 79 X5
E
c m/z = 100 x200
2
m/z = 129
/ = 1
m z 60 x50
m/z = 179 x200

 

 

L1llllllllJlLJJllllllllllllllllllllllllllllllllll'IJlllllJl

120 130 140 150 160 170 180

 

Temperature (K)

Figure 4.2 Temperature-programmed desorption data for 4.8 langmuir CF2Br2 on HOPG.
Monitored masses were m/z = 31 (CF), 50 (CF2I), 79 (Br‘), 100 (C2F4*), 129
(CF2BrI), 160 (Br2+) and 179 (C2F4Br1).

desorption profile and peak temperature (138 K) suggesting a common origin. We
concluded that molecular desorption occurred by comparison of the relative ion
abundances for the TPD data presented here at m/z = 31, 50, 79, 129, and 160
(21:24:17:100:1.5) with the mass spectrum of gas phase CF2Br2 (18:21:172100z2) as
measured by residual gas analysis. A small amount of m/z = 100 (< 1% of the area of
m/z = 129) was observed both in TPD measurements and residual gas analysis of the
CF2Br2 admitted to the chamber indicating the presence of an unidentified contaminant.
No signal was observed at m/z = 179 in residual gas analysis of CF2Br2.

A significant change in the TPD spectra was seen after UV irradiation of the
adsorbate. Figure 4.3 shows a similar set of TPD spectra to those in Figure 4.2 after 5
minutes of UV irradiation. The monolayer peak originally present at 138 K broadened
and shifted to 141 K and a second desorption peak centered at 155 K appeared for m/z =
31, 50, 79, 129, and 160. Relative ion abundances Of 23:27:20:100:l.4 and
22:27:22:100:1.3 (m/z = 31, 50, 79, 129, and 160) for the 141 K and 155 K peaks,
respectively, indicated that each peak was associated with molecular CF2Br2. The m/z =
100 and 179 TPD spectra showed a broad desorption peak at ~165 K. These features are
associated with the formation of a photolysis product, C2F4Br2, which will be discussed
later.

Figure 4.4 shows TPD spectra of the previously monitored ions after 15 minutes
of UV irradiation. The peak originally present at 141 K for m/z = 31, 50, 79, and 129 in
Figure 4.3 shifted to a higher desorption temperature (~ 146 K) and had started to merge

with the leading edge of the peak at 155 K. The peak at 155 K was present for all UV

44

 

 

 

4.8 L CFzBrZIHOPG
Irradiation Time = 5 minutes
m/z = 31 x5
m/z = 50
x5

2:
FEB m/z = 79
3 x5
E
c m/z = 100 . ' x200
_q - _

m/z = 129

m/z = 160 x50

 

m/z = 179 ' I , X200

120 130 140 150 160 170 180
Temperature (K)

Figure 4.3 Temperature-programmed desorption data for 4.8 langmuir CF2Br2 on HOPG after 5
minutes of UV irradiation (filtered Hg-arc, 225-350 nm, incident power 8.6
mW/cmz). Monitored masses same as in Figure 4.2.

45

 

4.8 L CFzBr2/HOPG
Irradiation Time = 15 minutes

M X5

 

 

 

3‘
2 / - 79
9 m - x§_
E a
g m/z 100 , ,   - - _ x200
m/z = 129
m/z = 160 "50
lz =179 : I x 00

 

 

JIIllllllIlllllllliJlllIllIllllljlllllljlllllllllllllllllll

120 130 140 150 160 170 180

 

Temperature (K)

Figure 4.4 Temperature—programmed desorption data for 4.8 langmuir CF2Br2 on HOPG, after 15
minutes of UV irradiation. Conditions same as in Figure 4.2.

46

exposures up to 60 minutes irradiation time (data not shown). Furthermore, a comparison
of the m/z = 160 TPD spectra in Figures 4.3 and 4.4 revealed the growth of a new peak at
165 K with continued UV exposure. This new peak in the m/z = 160 TPD spectrum was
observed only for UV irradiation times greater than 5 minutes.

4.2.2 X-ray Photoelectron Spectroscopy (XPS). Figure 4.5 shows the F ls and
Br 3d regions for increasing UV exposure. For the unirradiated sample, the F ls binding
energy was observed at 688.4 eV similar to that of CF2Br2 adsorbed on sapphire.19 As
the UV irradiation time increased, the F ls peak decreased in intensity and shifted to
lower binding energy (687.5 eV after 60 minutes UV exposure) but the peak width
(FWHM) remained constant at 3.0 eV. The Br 3d peak maximum (unresolved spin-orbit
split doublet) was observed at 71.6 eV for the unirradiated surface. This value is
relatively high compared to simple metal bromide solids (68.3-69.3 eV)20 and about 1 eV
higher than CH3Br(ad) on Ag(111).21 The Br 3d peak shifted to 70.4 eV and the peak
broadened from 3.9 eV to 4.8 eV (FWHM) after 60 minutes of irradiation. The intensity
of the Br 3d peak remained nearly constant.

Figure 4.6(a) shows the C 1s region of the clean HOPG surface and that of a 4.8 L
CF2Br2-covered HOPG surface. The main photoemission peak in both cases is located at
284.7 eV. Furthermore, the clean graphite surface also shows the well-known shake-up
feature (Ti—Mt“) centered at about 292 eV. The addition of the CF2Br2 monolayer resulted
in the appearance of a feature at 292.8 eV. This feature was most clearly Observed in the
normalized difference spectrum (bare HOPG minus 4.8 L CF2Br2(ad)/HOPG), and was

shifted by +8.1 eV BE relative to the main photoemission peak. We assign this feature to

47

 

678 680 682 684 686 688 690 692 694 696 698 700 702
' . I ' I . I ' I ' I ' I .
a)

_
d
-
fl
-

= 1000 cps s

60 min

30 min

20 min

10 min

IT‘I‘U l T

0 min

fr!

 

 

Br 3d b)

60 min

['11

(II

C

O

O

K

'0

(D

 

 

 

 

E
75E
g- yd”
3: 20min
to”
CL
'93- kan
E?
E 0min
Fi.r.I.I.I.I.I.I.I.I.I.I
58 60 62 64 66 68 7O 72 74 76 78 80 82

Binding Energy (eV)

Figure 4.5 X-ray photoelectron spectra of F ls region (a) and Br 3d region (b) of CF2Br2
monolayer as a function of UV irradiation. Vertical lines are drawn at 688.4 eV and
71.6 eV BE.

48

 

 

 
   
 
 

I T r I l I I fi I I I I I I l I I I 1 I
— Bare HOPG
""""" 4.8 L CFzBrleOPG
a) .
+ Normalrzed
Difference
Tr? 5
8 .4341}- -+ I" W.-------E
v : 1 “we
> 'I l l l I I l l l l I l j 1 j J l l l I
t : 4 a L or B /HOPG
a) L . r
c C b) 2 O2 .
q, C —— min UV
E E- -------- 30 min UV
— ~_- + Normalized
E‘ Difference
Fae.
El: I l l J I l l l l I L l l I I l l I I I

 

 

 

 

280 285 290 295 300
Binding Energy (eV)

Figure 4.6 (a) X-ray photoelectron spectra of the C Is region of bare HOPG (solid line), 4.8
langmuir CF2Br2-covered HOPG (solid circles) and the normalized differmce
spectrum (crosses). (b) X-ray photoelectron spectra of the C ls region of 4.8
langmuir CFzBry-covered HOPG (solid line), 4.8 langnuir CF2Br2-covered
HOPG after 30 min UV irradiation (solid circles) and the normalized diffa'ence
spectrum (crosses).

49

photoemission from the adsorbate carbon atom bonded to four electronegative atoms (2 F
atoms and 2 Br atoms), with the majority of the observed chemical shift expected to
originate from the F atoms (according to literature, approximately 2.9 eV per F atom, 1.0
eV per Br atom).22

Figure 4.6(b) shows XPS data of the C Is region Of a 4.8 L CF2Br2 exposed
graphite surface and an identical surface after 30 minutes of UV irradiation. Each
spectrum represents a separate dose. After irradiation, the 292.8 eV feature was observed
to decrease in intensity and a new feature appeared at approximately 290.0 eV. This is
illustrated in the normalized difference spectrum shown in Figure 4.6(b).

Figure 4.7 shows the integrated raw area ratios F ls / C 1s and Br 3d / C Is as a
function Of UV exposure. Each data point corresponds to a separate 4.8 L CF2Br2 dose
(saturated first layer). The F ls/C ls ratio decreased by 25% between 0 and 60 minutes
irradiation time, suggesting a net loss of F atoms from the adlayer. In contrast, the Br
3d/C ls ratio remained unchanged indicating that no Br atoms were expelled from the
surface during UV exposure. Error bars represent 1 o for two separate experiments.

4.2.3 Electron Energy Loss Spectroscopy (EELS). Figure 4.8 shows electron
energy loss spectra of 4.8 L of CF2Br2 on the HOPG surface with and without exposure
to UV radiation. A spectrum of the bare HOPG surface is also shown for comparison.
The spectrum for bare graphite (a) exhibited the characteristic background of a semimetal
caused by excitation of low-energy electron-hole pair transitions across the Fermi level.23
NO Other loss features were present in this spectrum. Spectrum (b) shows the surface
exposed to 4.8 L of CF2Br2 without UV irradiation. A weak loss was observed at 1157

cm". We assign this feature to the CF2 asymmetric stretch in the adsorbed molecule

50

 

- 4.8 L CFzBrleOPG

 

1.1-
. I F1s/C1s
e Br3d/C1s
1.0-
0.9-
0.8-
0.7-

 

lntegrated Raw Area Ratio

0
(II
1 l
'—.—.l
H___1

 

 

 

Irradiation Time (min)

Figure 4.7 Integrated area ratios (F ls/C Is and Br 3d/C ls) from X-ray photoelectron
spectroscopy of 4.8 langmuir CF2Br2 adsorbed on HOPG, as a function of UV
irradiation time (conditions as in Figure 4.3). The solid curve in the F lle ls ratio
is a single-exponential fit to the data. The solid curve in the Br 3d/C ls ratio is a
linear fit to the data.

51

 

4o.

 

 

 

 

 

 

l
x20
4.8 L CFZBrleOPG
30 in
as b)
5,? ml
0- t
8
E:
'6
c
.03.
5 x20
10 -»
C)
a)
01.1. I'.-._:,=_I.i

 

 

 

 

0 500 1000 1500 2000 2500 3000 3500

Energy Loss (cm’1)

Figure 4.8 High-resolution electron energy loss spectra of (a) clean HOPG, (b) 4.8 langmuir
CF2Br2-covered HOPG and (c) 4.8 langmuir CF2Br2-covered HOPG after 30 minute
UV irradiation.

52

since it is close to the reported value for CF2Br2(g/s) of 1153 cm'l.24'25 Spectrum (0)
shows the 4.8 L CF2Br2/HOPG surface after 30 minutes of UV irradiation. Again, the
only loss peak present was a C-F-like stretch that had red-shifted to 1129 cm'1 and gained
intensity. Furthermore, a noticeable decrease in the overall background intensity was
Observed.

4.2.4 Adsorption and Desorption of CF2Br2(ad)/HOPG. The TPD data in
Figures 4.1 and 4.2 show that CF2Br2 undergoes molecular adsorption and desorption
from a clean HOPG surface in the absence of UV irradiation. Measured ion abundances
for m/z = 31, 50, 79 and 129 in our desorption data closely match those of CF2Br2(g).
For the adsorbed layer, apparent zero-order desorption kinetics were observed. Such
behavior can arise either through true zero-order desorption kinetics, first-order
desorption kinetics modified by attractive interadsorbate interactions, or half-order
desorption kinetics from 2-dimensional islands modified by attractive interadsorbate
interactions.”28

Knorr and coworkers”33 have proposed monolayer structural models based on X-
ray diffraction data for many halocarbons (but not CF2Br2) on graphite. Their data for
CF2C12 indicated coexistence Of a two-dimensional island (8 phase) and lattice gas phase
(a disordered sub-monolayer adsorbate coverage) at the temperatures and coverages
appropriate to our measurements. It was suggested that the presence of the permanent
dipole moment in CF2C12 favored either an antiparallel or zig-zag arrangement of dipoles
within the islands. We believe that the zero-order desorption kinetics observed in our
TPD experiments for the first layer of CF2Br2 are also consistent with the formation of 2-

D islands modified by attractive interadsorbate interactions.”28 Similar TPD behavior

53

has also been observed for H2O adsorption on HOPG surfaces, and attributed to the
creation of 2-D islands nucleated at defect sites.34 We speculate that the lower
temperature desorption feature (~110 K), characteristic of multilayer desorption, appears
after the 2-D islands in the monolayer coalesce.

To our knowledge, the structure of the CF2Br2 monolayer on graphite has not
been determined. However, some information on a possible structure can be obtained by
comparing the fractional coverage, <1>CF2Br2, (defined as the number of adsorbate
molecules per surface C atom) with the fractional coverage of CF2Cl2 on graphite. In our
case, the fractional coverage was calculated for CF2Br2 on HOPG using the XPS
integrated raw area ratios shown in Figure 4.7 and expressions developed in the
literature.22 Using sensitivity factors appropriate for our electron analyzer (C Is = 0.205,
F IS = 1.00 and Br 3d = 0.5920), a mean value of (DCF2Br2 = 0.14 i 0.01 (4.6 x 1014
molecules/cmz) was obtained (see Appendix A).

The calculated fractional coverage compares favorably to the maximum fractional
coverage of the [3 phase of CF2Cl2 on graphite (<1>CF2C12=0.137).30'33 Similar surface
structures for monolayers of CF2Cl2 and CF2Br2 may be anticipated based on the
similarity of their dipole moments (CF2C12 = 0.51 D, CF2Br2 = 0.66 D)35 and projected
surface areas (CF2Cl2 = 27.9 A2, CF2Br2 = 31.7 A2). Areas were estimated from van der
Waals’ radii assuming both molecules have one F and two X atoms (X = Cl, Br) in
contact with the surface (CFX2 “tripod”) as suggested by the theoretical studies in
Chapter 5.29'33 The EELS data shown in Figure 4.8(b) provides additional evidence for
this adsorption geometry. The appearance Of a single vibrational mode at 1157 cm'l,

corresponding to the CF2 asymmetric stretch (1153 cm", b1 symmetry in CF2Br2(g)), is

54

consistent with a CF2 plane that is not parallel to the HOPG surface (surface selection
rule). The companion vs(CF2) mode Of CF2Br2 expected at ~1090 cm'1 was not observed,
presumably due to a small component of the dynamic dipole parallel to the surface
normal. No loss features attributable to CBr2 vibrations were observed.

4.2.5 Photolysis of CF2Br2 on HOPG. Numerous studies have shown the initial
photodissociation step operative at 200—300 nm in CF2Br2(g), is C-Br bond cleavage in a
36-47

single photon process.

CF2Br2 + hv —+ CF2Br + Br (4.1)

Similarly, upon exposure to UV radiation, photolysis of the CF2Br2 adsorbed on graphite
was observed. The m/z = 129 TPD data of Figures 4.3 and 4.4 showed a decreased
intensity of the 138-141 K peak associated with molecularly adsorbed CF2Br2 for
increasing UV irradiation times. The dissociation cross-section for CF2Br2(ad)/graphite
was estimated by measuring this loss in intensity as a function of UV exposure. For the
range of wavelengths generated by our filtered arc lamp, the apparent total cross-section
(the integrated cross-section for 225-350 nm) was 1.9 x 10'l9 cm2 (using a mean
wavelength of 287.5 nm in the calculation). Similarly, we estimated the integrated cross-
section for CF2Br2(g) over the same range of wavelengths by multiplying the scaled
irradiance curve for our lamp,48 the NiSO4 filter transmission spectrum,49 and the gas
phase cross-section data.50'51 The integrated cross—section (225-350 nm) for gas-phase
CF2Br2 was 4.5 x 10'19 cm2, very similar to our CF2Br2(ad)/HOPG value. In fact, greater
than 95 % of the contribution to the cross-section occurs in the 240-250 nm region
(where the Hg lamp irradiance is high and the cross-section for CF2Br2(g) is large).

Absorption cross-sections for wavelengths greater than 250 nm decrease rapidly

55

(~ 10'19 cm2 at 250 nm to ~ 10’23 cm2 at 320 nm).50'51 Essentially, no contribution to the
photolysis of CF2Br2 is expected for wavelengths greater than 320 nm. There was no
evidence for photochemical modification of CF2Br2 monolayers by ambient (visible) light
entering the UHV chamber.

The EELS data shown in Figure 4.8(c) indicated that after UV irradiation the loss
originally present at 1157 cm", due to CF2 vas in CF2Br2(ad), became more intense and
red-shifted to 1129 cm“. This value is very close to that observed in a study Of the 254
nm photolysis of CF2Br2 in an AI matrix (1138 cm”).36 In that study, a definitive
symmetry could not be assigned to this mode but it was assigned to one of the
fundamental C—F stretches of CF2Br. The similarity of the integrated cross-sectional data
for CF2Br2(ad)/HOPG and CF2Br2(g), and the observation of a vibrational mode of
CF2Br(ad), implies that UV irradiation of dibromodifluoromethane produces surface-
bound CF2Br in a similar fashion to equation 4.1.

Further evidence for photochemistry of the adsorbed CF2Br2 is observed by a shift
in the C 1s photoemission peak (Figure 4.6), from 292.8 to 290.0 eV BE following UV
exposure. The direction of the shift suggests a net reduction in the number of
electronegative substituent atoms attached to the C atom of CF2Br2(ad). The magnitude
of the shift (-2.8 eV) is consistent with loss of an F atom to produce CFBr2(ad), but we
believe this to be misleading. The shift must correspond to loss of a single Br atom to
form CF2Br(ad), as clearly observed in EELS and TPD measurements. The large C ls
BE shift observed following photolysis is likely to be complicated by changes in the
intensity of the C Is shake-up feature at ~292 eV. This feature is sensitive to the

electronic structure of the surface. Indeed, based on analogy with bromine-intercalated

56

graphite compounds, the presence of Br on the graphite surface is expected to cause some
charge transfer from the graphite to the Br atoms.52 This would cause a decrease in the
population of the filled graphite Ir-band and an increase in electronic conductivity.53 The
decrease in the electron-hole pair background in our EELS data (Figure 4.8) also supports
the idea of a more metallic surface for UV-irradiated CF2Br2 on graphite.

It is somewhat surprising that electron-induced chemistry due to photon
absorption by graphite does not appear to significantly contribute to the observed
photochemistry of CF2Br2(ad). Graphite strongly absorbs in the UV generating free
photoelectrons with kinetic energies from 0 to about 0.7 eV (hv250 um = 5.0 eV, graphite
work-function, (I) = 4.35 eV54). These do not appear to cause dissociative electron
attachment-type reactions for the adsorbed molecule despite the fact that for gas phase

2 is observed

CF2Br2, a large dissociative electron attachment cross-section of >10’15 cm
for ~ 0 eV electrons.55 There is substantial evidence in the literature that, in general, low
energy electrons generated by photon irradiation of metallic and semiconductor surfaces
can cause dissociation of an adsorbed molecule.16 The apparent lack of electron-induced
chemistry for adsorbed CF2Br2 is likely due to either an increased work-function for
graphite upon adsorption (> 0.7 eV), poor spatial/energetic overlap between the graphite
and adsorbate orbitals or competitive quenching by the surface. The theoretical studies
presented in Chapter 5 provide an indication that a change in the work function of
graphite could occur as evidenced by transfer of charge from the surface to adsorbed
CF2Br2.

4.2.6 Surface Recombination Reactions. In static cell studies of the photolysis

Of CF2Br2(g) a variety of radicals and stable molecules are generated, including CF2Br,

57

CF2, Br and C2F4Br2 (formed by biradical reaction of CF2Br).3840’42'MJ'7 Reactions
generating Br2 and C2F4 do not appear to be major product channels at wavelengths
greater than about 248 nm and at low photon fluences.

The high temperature (155 K) TPD feature observed in our data is likely due to
recombination of photogenerated CF2Br(ad) and Br(ad) atoms during the TPD
experiment. The photodissociation event imparts kinetic energy to the nascent
photofragments (total up to about 2.0 eV at our shortest wavelength), separating them and
preventing immediate recombination. Prompt recombination would generate CF2Br2 that
would be indistinguishable in TPD measurements from the unphotolyzed molecules. We
assume CF2Br and Br are weakly chemisorbed and not significantly mobile, since they
are stable on the graphite surface up to ~155 K. Diffusion of one or both of these species
apparently occurs only at the elevated surface temperatures experienced during TPD.

Another recombination product observed on the HOPG surface was Br2(ad) as
indicated by the appearance of a desorption peak centered at 165 K for m/z = 160 (Figure
4.4). This feature is observed only after extended UV irradiation periods (210 minutes).
The expected m/z = 79 ion (Br+) at 165 K from fragmentation of Br2 in the ionizer has a
small abundance, preventing detection in our data (see Figure 4.4). It should be noted
that the appearance of coincident features in the m/z = 100 and 179 are not related to the
production of Br2(ad) but are associated with another photolysis product, C2F4Br2 (see
below).

Molecular bromine can result from concerted or sequential elimination from a
single CF2Br2 molecule

CF2Br2 + hv —> CF2 + Br2 (4.2)

58

OI'

CF2Br2 + hv —-> CF2Br + Br
CF2Br + hv —> CF2 + Br (4.3)
Br + Br —> Br2

or from the recombination of Br atoms produced from photolysis of two CF2Br2
molecules

2CFzBr2 + 2hv —> 2CF2Br + 2Br

(4.4)
Br + Br —> Br2

Even at the shortest wavelengths employed in this study, there is insufficient
energy in a single photon (hv250 m = 5 .0 eV) to simultaneously cleave two C-Br bonds (2

x Do (C-Br) E 6.0 eV).35 Therefore, we discount reaction 4.2 as the source of Br2. The
photodissociation cross-section of CF2Br(g) is (0.5—4.4) x 10‘18 cm2 at 248 nm,“'44416
about an order of magnitude greater than the value for CF2Br2(g) at the same wavelength.
As such, we anticipate significant photodissociation of the CF2Br photoproduct to occur.
We found no evidence for the formation of CF2(ad); however, the formation of CF2(g)
which is expelled from the adlayer following photolysis Of CF2Br(ad), does not preclude
the formation of adsorbed Br. We are, therefore, unable to comment on the relative
contributions of reactions 4.3 and 4.4 to the production of Br atoms.

The XPS data indicated Br-containing species formed during UV irradiation Of
CF2Br2 remained adsorbed on the surface, as shown in Figure 4.7. Consistent with the
production of several species (CF2Br, Br, Br2, C2F4Br2) during photolysis, the Br 3d peak
broadened from 3.9 to 4.8 eV FWHM between 0 and 60 minutes irradiation, respectively.

The peak center also shifted from 71.6 eV to 70.4 eV BE. The shift implies a net

reduction in the number of Br atoms in an electronegative environment as expected for Br

59

bi

pi.

fr

and Br2 compared with CF2Br2 (we assume the binding energies for Br in CF2Br2 and
C2F4Br2 are similar).

A second photoproduct formed was C2F4Br2 (Halon 2402) by dimerization of two
photogenerated CF2Br radicals. The dimer is indicated by the appearance of desorption
peaks at ~165 K for m/z = 100 (C2F4I) and 179 (C2F4Br+), as shown in Figures 4.3 and
4.4. The measured TPD intensity ratio at these two masses (47:100 for m/z = 100 and
179, respectively) agrees with our measured QMS ion abundances for C2F4Br2(g) (43: 100
for m/z = 100 and 179, respectively), confirming the presence of C2F4Br2(ad).

It should be noted that the unirradiated CF2Br2 monolayer showed a small TPD
peak at ~ 141 K for m/z = 100. This is believed to be an impurity; there is no peak at this
m/z in the mass spectrum of CF2Br2(g).56 It is unlikely that any Br (or Br2) is contributed
by the direct photolysis of C2F4Br2. Ultraviolet photodissociation cross sections for gas-
phase C2F4Br2 at the wavelengths used in our experiments (225-350 nm) are quite small
(~ 10.20 _ 10.21 cmz).50'51

4.2.7 Loss of Fluorine Species. The total loss of parent CF2Br2 as determined
from TPD measurements (~ 140 K peak), was approximately 14 i 3 % after 15 minutes
irradiation. This value is almost identical to the decrease in the F ls/C ls XPS intensity
ratio after the same UV irradiation period (13 i 5 %). However, it will be recalled that
there was essentially no change in the Br 3d/C ls XPS intensity ratio over these (or
extended) UV exposures. This immediately implies that the Observed loss of fluorine-
containing species cannot be due to photodesorption of CF2Br2 or CF2Br, both of which

would reduce the total Br concentration on the surface.

60

In addition to CF2Br, both CF2 and C2F4 products have been Observed in matrix
isolation experiments of CF2Br2 photochemistry.36 Although we see no evidence for
adsorbed CF2 in our EELS data, it is possible that the CF2 formed during photolysis of
CF2Br is expelled from the surface as CF2(g). The formation of CF2(ad) would be
signaled by the appearance of vibrational modes at ~1225 cm"(vs — a1 symmetry) and
~1106 cm'l (vas — b1 symmetry).56 Alternatively, CF2 photofragments may combine to
produce C2F4(ad) or C2F4(g). We can discount the formation of appreciable quantities of
C2F4(ad) since at no time did we observe losses in the EELS data due to C2F4(ad) (~ 1180
and 1330 cm")56 or in TPD data attributable to the cracking of QR; (notably at m/z = 31
and 100)?6

In principle, it should be possible to determine the overall stoichiometry of the
departing fluorine-containing species by examining the reduction in the ~293 eV feature
in the C Is XP spectrum (Figure 4.6) with UV exposure. However, the difficulties
associated with background subtraction and the relative weakness of the feature (the
number of C atoms possibly leaving the surface as CF2(g) is very small compared to the
number of C atoms sampled in the HOPG) made the results unreliable. A shift in binding
energy (-0.9 eV after 60 minutes UV irradiation) was observed for the F ls peak,
presumably due to net loss of the Br atom. In contrast to the Br 3d peak, the F ls peak
exhibited no discemable broadening, implying that only one type Of F chemical
environment was present on the surface after extended photolysis. The remaining
species, CF2Br and C2F4Br2, both contain similar bonding arrangements for F so different

binding energies for these adsorbates are not expected.

61

Several authors have studied the photodissociation of CF2Br2(g) and determined
that approximately 20-30% of the CF2Br formed following the initial CF2Br2 photolysis
undergoes a second C-Br scission via a vibrationally excited CF2Br* radical.42‘43’45 This
process generates a second Br atom and a CF2 radical. Such a scenario is also consistent
with our Observations for loss of fluorine from CF2Br2(ad)/graphite (see Figure 4.7) if the
vibrationally excited CF2Br* subsequently fragments to produce CF2(g) and Br(ad).
Indeed, the rate of loss of F-containing species is, on average, 25% the rate Of
dissociation of CF2Br2(ad)/graphite (Figure 4.9), which is in good agreement with the
value of Gosnell, et. al.42 Experimental geometry prevented us from directly monitoring
species desorbing during irradiation, and we cannot independently confirm desorption of
CF2(g) from the adlayer in our case.

4.2.8 Quantification of Surface Reactions. Calculations were performed based
on our TPD data in order to quantify the various products on the HOPG surface after UV
irradiation. Electron impact ionization cross sections were calculated for CF2Br2, Br2,
and C2F4Br2 using Deutsch-Mark (D-M) theory57 in order to correct for ionization
efficiency of the products at the electron energy of our QMS (70 eV). Deutsch-Mark
theory is a semi-classical approach based on quantum-mechanically optimized molecular
geometry, population weighting factors, and summed contributions arising from electron
ejection by each occupied molecular orbital. Of the various calculation schemes used to
determine absolute electron impact ionization cross-sections, Deutsch-M'ark theory
appears to give the most reliable results for molecules containing heavy atoms.58

Calculated electron impact ionization cross-sections for CF2Br2, Br2, and C2F4Br2 were

62

 

 

 

 

100 -
80 _ A A Total CF2Br2
‘ I
60 - CFZBr
E .
8 .
a) I
0- e
‘ 4o 4 .
‘ CFzBr+Br
. . 0
20 -
- Br2
0 _ ‘__ ___ 4 CZF48r2

 

Time (minutes)

Figure 4.9 Relative percentages of the different reaction channels present on the surface as a
function of UV irradiation. The lines through the data are guides for the eye.

63

11.25 A2, 7.78 A2, and 14.48 A2, respectively. 59 Corrections were also made for ion
fragmentation using published and measured mass spectra for CF2Br2(g), Br2(g) and
C2F4Br2(g) by multiplying each cross-section by the sum of the ion intensities of each
fragment divided by the intensity of the fragment of interest .56

Figure 4.9 summarizes the relative rates of the various processes observed during
UV irradiation of 4.8 L CF2Br2(ad)/HOPG, after correction for electron impact ionization
cross-sections and fragmentation. The most prominent reaction channel for all UV
exposures studied was the C-Br photodissociation event as shown in equation 4.1. For
example, after 15 minutes of UV irradiation, the monolayer is composed of
approximately 45 i 7 % unphotolyzed and 55 :l: 7 % photolyzed CF2Br2 molecules.
During temperature programmed desorption experiments, most of the photolyzed
molecules (70 i 10 % of the photolyzed fraction or 38 i 6 % of the total CF2Br2
molecules) recombine to regenerate parent molecules (155 K peak in TPD data). The
remainder of the photolyzed molecules (30 i 5% of the photolyzed fraction or 17 j: 3% of
the total CF2Br2 molecules) formed other reaction products. The fluorine-containing
component of this remaining 17 2t 3% appears to be almost completely lost from the
adlayer (as indicated by both F ls/C ls XPS and m/z = 129 TPD data). The formation of
C2F4Br2 and Br2 increased to 1.4 :l: 0.6% and 6 i 2% of the total number of molecules,
respectively, after 15 minutes of irradiation, indicating that recombination of CF2Br
radicals and Br atoms are relatively minor channels.

We observe various recombination reactions during TPD measurements due to the
formation of an adlayer with a high density of photogenerated radicals. However, we

observe reaction probabilities that are significantly different to those observed for the

photolysis of CF2Br2(g). The most prevalent reaction observed in our experiments was
the recombination of a large fraction of the photolyzed molecules to reform the parent
molecule during heating in TPD measurements, a channel not detected in static-cell gas

3340:4144“ or (low temperature) matrix isolation36 studies. Vatsa et. al.46 have

phase
suggested that the majority (90%) of the CF2Br radicals produced by 248 nm
photodissociation of CF2Br2(g) in a cell formed C2F4Br2 through dimerization reactions.
We Observed that the formation of C2F4Br2 was a minor channel on the surface reaching a
maximum after 10 minutes of photolysis. The proportions of the various products are not
simply statistical if only CF2Br and Br are produced. In this case, we would expect the
formation of equimolar amounts of Br2 and C2F4Br2. The formation of a higher amount
of Br2 than expected (by about a factor of four compared to C2F4Br2) is likely due to
efficient photolysis of CF2Br(ad), spontaneous C-Br scission in CF2Br* or other
controlling factors operative on the surface. Surface processes may include reduced
reaction probability for a particular recombination, perhaps due to structural
(orientational) constraints, or reduced surface diffusion for one of the species.

4.3 Conclusions. Dibromodifluoromethane (CF2Br2) adsorbed molecularly on an
HOPG surface at 85 K, with monolayer saturation corresponding to a fractional coverage
of 0.14 :l: 0.01 CF2Br2 per surface C atom. Molecular desorption occurred from the
monolayer at approximately 138-141 K. Photolysis of CF2Br2 was Observed upon
exposure to 225-350 nm Hg arc lamp irradiation. The major products formed during
photolysis were CF2Br and Br atoms. The estimated integral cross-section (225-350 nm)

2

for this process was ~ 1.9 x 10'19 cm , similar to the integrated UV photolysis cross-

section for CF2Br2(g) at these wavelengths. This implies that dissociative electron

65

attachment (DEA) does not significantly contribute to the adsorbate photochemistry. A
large fraction Of the CF2Br and Br recombine to produce CF2Br2 during TPD
experiments. Additionally, minor channels for Br2 and C2F4Br2 (Halon 2402) formation
were Observed. It should be noted that the adsorption/desorption temperatures Observed
for all species in our work are 30-50 K lower than the minimum temperatures of the
upper troposphere or lower stratosphere. Within the validity of our HOPG surface as an
accurate model for primary carbon aerosols, the observed reactions will be negligible in

the atmosphere.

66

References

Quinlan, M. A.; Reihs, C. M.;. Golden, D.M; Tolbert, M.A. J. Phys. Chem. 1990, 94,
3255.

Tolbert, M. A.; Rossi, M. J .; Malhotra, R.; Golden, D. M. Science 1987, 238, 1258.
Donsig; H. A.; Herridge, D.; Vickerrnan, J. C. J. Phys. Chem. A 1999, 103, 9211.
Berland, B. S.; Tolbert, M. A.; George, S. M. J. Phys. Chem. A 1997, 101, 9954.

Finlayson-Pitts, B. J .; Pitts, Jr., J. N. Chemistry of the Upper and Lower Atmosphere;

. Academic Press: San Diego, 2000.

10.

ll.

12.

13.

14.

15.

16.

17.

Schauffler, S. M.; Atlas, E. L.; Flocke, F.; Lueb, R. A.; Stroud, V.; Travnicek, W.
Geophys. Res. Lett. 1998, 25, 317.

Molina, M. J.; Molina, L. T.; Kolb, C. E. Annu. Rev. Phys. Chem. 1996, 47, 327.

United Nations Environmental Program, Montreal Protocol on Substances That
Deplete the Ozone Layer; Montreal, 1987.

Fraser, P. J.; Oram, D. E.; Reeves, C. E.; Penkett, S. A.; McCulloch, A. M. J.
Geophys. Res. 1999, 104, 15985.

Zhang, R.; Jayne, J. T.; Molina, M. J. J. Phys. Chem. 1994, 98, 867.
Roberts, J. T. Acc. Chem. Res. 1998, 31 , 415.

Oum, K. W.; Lakin, M. J.; DeHaan, D. O.; Brauers, T.; Finlayson-Pitts, B. J. Science
1998, 279, 74.

Hemminger, J. C. Int. Rev, Phys. Chem. 1999, 18, 387.

DeHaan, D. O.; Brauers, T.; Oum, K.; Stutz, J.; Nordmeyer, T.; Finlayson-Pitts, B. J.
Int. Rev. Phys. Chem. 1999, I8, 343.

Andreae, M. O.; Crutzen, P. J. Science 1997, 276, 1052.
Zhou, X.-L.; Zhu, X.—Y.; White, J. M. Surf Sci. Rep. 1991, 13, 73.

de Jong, A. M.; Niemantsverdriet, J. W. Surf Sci. 1990, 233, 355.

67

18. Yates, Jr., J. T. Methods of Experimental Physics, VOl.22.; Academic Press: San
Diego, 1985; p. 425.

19. Robinson, G. N.; Freedman, A.; Kolb, C. E.; Worsnop, D. R. Geophys. Res. Lett.
1994. 21 , 377.

20. Wagner, C. D.; Riggs, W. M.; Davis, L. E.; Moulder, J. F.; Mullenburg, G. E. (Ed.)
Handbook of X-ray photoelectron Spectroscopy; Perkin-Elmer Corporation: Eden
Prairie, 1979; p 94.

21. Zhou, X. —L.; White, J. M. Surf Sci. 1991, 241, 259.

22. Briggs, D. In Practical Surface Analysis, Briggs, D., Seah, M., Eds.; Wiley and Sons:
New York, 1990; Vol. 1, p 444.

23. Palmer, R. E.; Annett, J. F.; Willis, R. F. Phys. Rev. Lett. 1987, 58, 2490.
24. Eix, S. L.; Schlueter, S. A.; Anderson, A. J. Raman Spectrosc. 1992, 23, 495.

25. Baldacci, A.; Gambi, A;. Giorgianni, S.; Visinoni, R. and Ghersetti, S. Spectrochim.
Acta 1987, 43A, 455.

26. Golze, M; Grunze, M.; Hirschwald, W Vacuum 1981, 31 , 697.

27. Niemantsverdriet, J. W.; Markert, K.; Wandelt, K. Appl. Surf Sci. 1988, 31 , 211.
28. Zhou, X. —L.; White, J. M.; Koel, B. E. Surf Sci. 1989, 218, 201.

29. Knorr, K. Phys. Rep. 1992, 214, 113.

30. Knorr, K.; Civera-Garcia, E. Surf Sci. 1990, 232, 203.

31. Nalezinski, R;. Bradshaw, A.M.; Knorr, K. Surf Sci. 1995, 333, 255.

32. Volkmann, U. G.; Knorr, K. Phys. Rev. B 1993, 47, 4011.

33. Warken, A.; Enderle, M.; Knorr, K. Phys. Rev. B 2000, 61, 3028.

34. Chakarov, D. V.; Osterlund, L.; Kasemo, B. Vacuum 1995, 46, 1109.

35. Lide, D. R. Ed.; CRC Handbook of Chemistry and Physics, 75'” ed.; CRC Press: Boca
Raton, 1995.

36. Jacox, M. E. Chem. Phys. Lett. 1977, 53, 192.

37. Johnson, C. A. F.; Ross, H. J. J. Chem. Soc. Farad. Trans. 1 1978, 74, 2930.

68

(II

(_II

38.

39.

40.

41.

42.

43.

45.

46.

47.

48.

49.

50.

51.

52.

53.

54.

Sam, C. L.; Yardley, J. T. Chem. Phys. Lett. 1978, 61 , 509.

Wampler, F. B.; Tiee, J. J .; Rice, W. W.; Oldenborg, R. C. J. Chem. Phys. 1979, 71 ,
3926.

Molina, L. T.; Molina, M. J. J. Phys. Chem. 1983, 87, 1306.
Krajnovich, D.; Zhang, 2.; Butler, L.; Lee, Y. T. J. Phys. Chem. 1984, 88, 4561.
Gosnell, T. R;. Taylor A. J .; Lyman, J. L. J. Chem. Phys. 1991, 94, 5949.

Talukdar, R. K.; Vaghijani, G. L.; Ravishankara, A. R. J. Chem. Phys. 1992, 96,
8194.

. Vatsa, R. K.; Kumar, A.; Naik, P. D.; Rama Rao, K. V. S.; Mittal, J. P Chem. Phys.

Lett. 1993, 207, 75.
Van Hoeymissen, J .; Uten, W.; Peeters, J. Chem. Phys. Lett. 1994, 226, 159.

Vatsa, R. K.; Kumar, A.; Naik, P. D.; Rama Rao, K. V. S.; Mittal, J. P. Bull. Chem.
Soc. Japan 1995, 68, 2817.

Talukdar, R. K.; Hunter, M.; Warren, R. F.; Burkholder, J. B.; Ravishankara, A. R.
Chem. Phys. Lett. 1996, 262, 669.

Esposito, E.; Femandes, N.; FitzSimons, C.; Marino III, E. Eds.; Light Sources,
Monochromators and Spectrographs, Detectors and Detection Systems, and Fiber

Optics, Vol. 2; Oriel Corporation: Stratford, 1994; p. 1-42.

Calvert, J. G.; Pitts, Jr., J. N. Photochemistry; Wiley and Sons: New York, 1966; p.
729.

Orkin, V. L.; Kasimovskaya, E. E. J. Atmos. Chem. 1995, 21, 1.

Burkholder, J. B.; Wilson, R. R.; Gierczak, T.; Talukdar, R.; McKeen, S. A.; Orlando,
J. J.; Vaghjiani, G. L.; Ravishankara, A. R. J. Geophys. Res. 1991, 96, 5025.

Ellis, A. B.; Geselbracht, M. J.; Johnson, B. J.; Lisensky, G. C.; Robinson, W. R.
Teaching General Chemistry: A Materials Science Companion; American Chemical
Society: Washington D. C., 1993; p. 193.

Burdett, J. K. Chemical Bonding in Solids; Oxford: New York, 1995; p. 71.

Somorjai, G. A. Introduction to Sudace Chemistry and Catalysis; Wiley and Sons:
New York, 1994; p. 381.

69

55. Underwood-Lemons, T.; Gergel, T. J .; Moore, J. H. J. Chem. Phys. 1995, 102, 119.

56. N. M. S. D. Center; S. Stein, In NIST Chemistry WebBook, NIST Standard Reference
Database Number 69; W. Mallard, P. Linstrom, Eds.; National Institutes of Standards
and Technology: Gaithersburg, MD, 1998 (http://webbook.nist.gov).

57. Deutsch, H.; Becker, K.; Matt, S.; Mark, T. D. Int. J. Mass Spec. 2000, 197, 37.

58. Vallance, C.; Harris, S. A.; Hudson, J. E.; Harland, P. W. J. Phys. B 1997, 30, 2465.

59. P. W. Harland (personal communication).

70

Chapter 5 Theoretical Investigation of the Adsorption Sites and

Orientation of CF2Br2 and CF2Br on Model Graphite Surfaces

Abstract

Adsorption sites and orientations of CF2Br2 and CF2Br on model graphite
surfaces were determined using a cluster model approach. Second-order Mdller-
Plesset perturbation theory (MP2) calculations were performed using 6-3lG* and
6-3lG(2d) basis sets with four ring (pyrene) and seven ring (coronene) systems as
models for the graphite surface. The preferred adsorption site for CF2Br2 was
found to be located at the hollow site of a six-membered ring for each
combination of surfaces and basis sets examined. For the CF2Br radical,
different preferred adsorption sites were found for the four and seven ring
substrates. The atop site was the preferred location on the four ring system while
the bridge site was preferred on the seven ring system. Rotation of the molecules
on the seven ring surface revealed similar binding energies compared to those
Obtained for the initial geometries suggesting that free rotation may occur at each

adsorption site.

71

5.1 Introduction. Adsorption and reaction Of small molecules on graphite and
various carbonaceous surfaces has attracted a great deal of attention due to their role in
such fields as astrophysics,” air and water purification,5 and atmospheric chemistry.(*26
The ubiquity of carbonaceous surfaces in such diverse environments make them a likely
surface upon which reactions can take place. For example, the formation of molecular
hydrogen in the interstellar medium is thought to occur by recombinative desorption of
two H atoms on particles coated with a graphite-like layer."4 Heterogeneous reactions
occurring on carbonaceous aerosols in the troposphere have been implicated in the
reduction of such adsorbed species as HNO3, NO2, and O; as well.‘5'14’19’24’26 However,
due to the diverse nature of carbonaceous surfaces (porosity, surface functional groups,
etc.), the mechanisms by which these reactions take place are not well understood.

In an effort to determine the role of the surface in such processes, a number of
ab initio and molecular dynamics calculations have been performed examining the
binding energies, sites, orientations, and mono- and multi-layer structures of various

2742 The dominant interaction between most molecules and

adsorbates on graphite.
graphite occurs through dispersion forces which result in the molecules being
physisorbed on the surface. Numerous studies have been performed on the adsorption
- - - 27.29-3239
properties of some homonuclear dratonucs (H2, N2, and 02) as well as larger
adsorbates such halogenated methanes, ethane, and acetone.3‘3‘38'4°’42 For calculations
involving the homonuclear diatomics, it was observed that adsorption energies of less
than 10 kJ/mol were obtained in each case with a majority finding the preferred
27.29—32.39

adsorption site to occur over the center of a six-membered ring on the surface.

The preferred adsorption orientation for most of these species was observed to occur with

72

the axis of the molecule parallel to the surface due to steric and electrostatic effects.
Molecular dynamics (MD) calculations on a series of dipolar halogenated methanes
revealed that the global energy minimum occurred for a position slightly removed from
the atop site with the molecules in a “tipped” orientation (the molecular axis forms an
angle with the surface normal).36 Simulations of the formation of mono- and multi-layers
Of acetone41 revealed that islanding of the adsorbate occurred (Volmer-Weber growth
mode) in contrast to the layer-by-layer growth for non-polar species.37'38‘42
The particular molecules of interest in this study are CF2Br2 and CF2Br and their
interaction with a graphite surface. Dibromodifluoromethane belongs to a class of
molecules known as Halons (CF2Br2 is termed Halon 1202) that are commonly used in
fire suppression and gas and oil production.43 To date, only one experimental study has
been performed on the adsorption (and photochemistry) of CF2Br2 on a HOPG surface.44
It was determined that molecular adsorption occurred on the surface with a structure
similar to that of CF2Cl2 adsorbed on HOPG and that the initial photochemistry of
CF2Br2(ad/HOPG closely resembled that of gas phase CF2Br2.45’57
CF2Br2 + hv —> CF2Br + Br (5.1)
The present study aims to address the preferred adsorption sites and orientations
for both pre- and post-irradiated CF2Br2 adsorbed on HOPG. Cluster model calculations
have been performed examining the interaction between CF2Br2 and CF2Br with model
graphite surfaces as a way of obtaining information on the local environment of each
adsorbate. Comparison of the results presented here with those obtained from the

experimental study will allow for a more complete description of this system.

73

5.2 Computational Details. Calculations of the preferred adsorption sites and
orientations of CF2Br2 and CF2Br on model graphite surfaces were carried out using the
cluster model approach.58 A cluster model was adopted to obtain information on the
localized interaction between an isolated adsorbate and the substrate. A single adsorbate
interacting with the substrate corresponds to a zero-coverage limit for each species. The
graphite substrate was modeled using a single layer slab approximation consisting of
either a four-ring (pyrene) or seven-ring (coronene) polycyclic aromatic hydrocarbon
(PAH), as shown in Figure 5.1. Previous studies have shown that the interaction between
the first and second layers in graphite is weak compared to the in-plane bonding between
carbon atoms and that a single layer is often sufficient to provide a reasonable
representation of the electronic properties of the three-dimensional system.28‘3°""9'63
Furthermore, the use of small PAHs allows for reduced computational effort due to a
smaller number of basis functions involved for each calculation. Both ring systems were
stabilized by terminating the edge carbon atoms with hydrogen atoms eliminating the
presence of all carbon dangling bonds. The adsorption sites used for calculating all of
the potential energy curves are shown in Figure 5.1. The indicated positions will be
referred to as the atop, hollow, and bridge sites.

Two basis sets, 6-31G* and 6-31G(2d), were used to calculate the potential
energy curves between each adsorbate and substrate. These medium-sized basis sets
were chosen because they contain polarization functions that are essential in describing
weakly bound systems.“65 The 6-31G(2d) contains a second d function on all heavy

atoms that allows for added flexibility in the valence region. An attempt was made at

74

 

b)

Figure 5. 1 Four (a) and seven ring (b) model graphite surfaces with adsorption positions indicated.
The atop site occurs over a substrate atom, the bridge site is between to substrate atoms
and the hollow site is at the center of a six-membered substrate ring.

75

using the 6-31+G* basis set in these calculations because it has been suggested that
diffuse functions are important in obtaining significant portions of the interaction
energy.3°'64’65 However, convergence of the SCF equations could not be achieved with
the addition of diffuse functions. All calculations were performed with the core electrons

(C Is, F ls, Br Is, 25, 2p, 3s, 3p, and 3d) held frozen.66

Calculations of the optimized molecular geometries and potential energy curves
were performed using second-order Moller-Plesset perturbation theory (MP2). MP2
yields qualitatively similar interaction energies for weakly bound systems when

30 This theory fundamentally includes

compared to higher level correlated techniques.
contributions to the interaction energy that are due to electrostatic, induction, dispersion
and exchange effects present in weakly bound systems64‘65 and has the benefit of being
size-consistent. Also, due to the size of the systems being studied here, MP2 provides for
dynamic electron correlation at a reduced computational expense compared to multi-
reference methods. Geometry optimizations for each molecule were performed at the
MP2 level of theory using the 6-3lG* basis set. All calculations of the preferred
adsorption sites and geometries for the adsorbates on both the four and seven ring
substrates (as determined from the largest negative value for the interaction energy) were
carried out using the previously mentioned basis sets with the optimized 6-31G*
geometries.

The four and seven ring systems were chosen to investigate changes in the
binding energies, adsorption sites and orientations as the size of the substrate was

increased. The minimum height of the molecules above the surfaces was determined by

the distance between the C atom of the molecule and the plane containing the surface.

76

SL

CZ

DE

(SD;
.333

The Optimized geometries were kept rigid during the calculation of the interaction
energies and were assumed to remain unchanged during cluster formation. The
geometries of CF2Br2 and CF2Br on the four ring substrate were arranged so that the
maximum molecular surface area/electron density of the adsorbate was interacting with
the substrate as seen in Figures 5.2 and 5.3, respectively. Rotation of the adsorbed
molecules on this surface resulted in one or more atoms not being in contact with the
substrate. The use of the seven ring system allowed for rotation of the adsorbates such
that all atoms remained on the substrate and new orientations could be investigated for _
calculation of the interaction energy. New molecular orientations were obtained with the
F and Br atoms of the adsorbates placed either over the center of a ring in the substrate or
nearly on top of an atom of the substrate. The orientation of CF2Br2 interacting with the
model graphite surfaces (Br2F “tripod” approaching the surface) was determined by a
previous experimental study of the adsorption of this molecule on a highly-ordered
pyrolytic graphite surface (HOPG)44 and by comparison with the adsorption of CF2Cl2 on
HOPG.“71 Partial geometry optimization calculations confirmed this geometry. The
geometry of CF2Br with respect to the substrate was determined by partial geometry
optimization calculations as well.

Calculations of the interaction energies were performed for both basis sets by
subtracting the energy ' of each isolated fragment from the energy of the cluster
(superrnolecular approach) at every separation distance with basis set superposition error
(BSSE) being accounted for by the counterpoise method of Boys and Bemardi72 as

implemented in Gaussian 98.73

77

8a macaw Emcee»: .803 2a mfiofi 8930

3

 

.moaawm as a bum one macaw onion—w Ba .2959 2a .6.an 05.55 .223
some emote A8 v8 .Bcsoa 3V doom 3 05 a 08,35 338 met .58 05 no ~5me «d 9:5...—

§ 3

 

78

use swede so as .332 so see 3 as a 8&5 .868 we: as ea :6 Hmemo on as»...

3 so

 

79

5.3 Results and Discussion.

An initial series of calculations were undertaken to determine an appropriate basis
set for the study of these systems. Geometry optimization and frequency calculations
were performed on CF2Br2 at the MP2 level of theory with a variety of basis sets.
Comparison of bond lengths and angles between the calculated and experimental values
could not be made because the crystal structure of CF2Br2 has not been determined to
date. However, vibrational frequencies of this molecule have been experimentally
determined74 and can be compared to the calculated values. The results are summarized
in Table 5.1.

The computed frequencies obtained from calculations using each basis set were
comparable to the experimental values. The 6-31G* basis set produced frequencies that
matched well with lower frequency stretches Of CF2Br2 but yielded larger differences for
the higher frequency Stretching modes. The larger basis sets (cc-pVDZ and 6-311G*)
gave improved values for the higher frequency stretches, but comparison with the lower
frequency stretches showed increasing deviations. From these data, it was concluded that
the 6-31G* basis set would be used as the starting basis set for all calculations: this basis
set produced the smallest average error in the calculated and experimental vibrational
frequencies. Further evidence to support the use of this basis set comes from the
geometry optimization of the 4 ring substrate (pyrene). The 6-31G* basis set yielded
values for C-C bond lengths that were on average only 0.007 A larger than values
obtained from a neutron diffraction study.75 A larger basis set (6-31G(2d)) was used in
later calculations to determine the effects that additional basis functions in the valence

region would have on adsorption sites and geometries.

80

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

3.8a 3.3.. 3. m: _eoem omega

3.2+ it me: 3.3+ .NE 82 3.9? .83 22 mm: seas 6.56855 .5
3.2+ .emt 2 : 3.2+ em; 2 : 3a.? it a: 88 Ease erases do
32+ Es wee 3.3+ at new 3.3+ .3 ems E as... outside emo
3; .N+ it em... 3.0.: at E. 3.3- .3 2e 8e .88... mo
3.4.? .3 E 3.3+ at E E %m as .62 m6
3; .N+ .3 as 3.2+ .3 9a 3.2 .E E as ease assess emu
33+ it ”2 3a.? at one 3.3+ at E am we... Nmo
3.3+ 63 am 3.3+ .3 as 3%.? .3 as am as: «mo
33+ .3 E 3.»: .3 E 3.2+ .3 E we seem emu
p.63 .9 _ SE: 9.83 33-6282 pee 5:382 pee 38586 8:85

 

 

5953 80:th 3.5—2 23 03°33 88. names Botg 5.5» momoeuaaoa 88328 new 3585950 mo ecmEEoO fin 93:.

.9385 a gone 83 agatonxo 93 83328

81

5.3.1 Adsorption Sites and Geometries of CF2Br2. The results of calculations
of the interaction energies between CF2Br2 and each of the model graphite surfaces are
shown in Figures 5.4-5.9. An examination of Figure 5.4 shows broad, shallow potential
energy curves were obtained for each of the adsorption sites. For the 6-31G* basis set, it
was observed that the minimum height of CF2Br2 above the 4 ring surface for the atop
and bridge sites occurs near 4.25 A while that for the hollow site is approximately 4.1 A.
These values are larger than those obtained from molecular dynamics simulations of a
number of Chlorofluorocarbons interacting with graphite.”39 Hammond and Mahanti36
found that vertical adsorption heights ranged between 3.23 A (for CH3C1) and 3.61 A
(for CF3Cl) for dipolar molecules while Pinches and Tildesley39 observed a distance of
3.55 A for CF4. The vertical adsorption heights calculated for CFzBrz above the graphite
surface are expected to be larger due to the increase size of the van der Waals radius of a
Br atom compared to a Cl atom and the restriction that the adsorbate and surface
structures were frozen in the calculations.

Figure 5.4 shows all three adsorption sites have comparable binding energies on
the four ring substrate. The preferred adsorption site occurs over the hollow of a six-
membered ring with a binding energy of at least -1 1.79 kJ/mol. In comparison, the atop
(-10.48 kJ/mol) and bridge (-9.93 kJ/mol) sites were found to have smaller binding
energies. Errors of 10% or less are typical for conformational energies using MP2 and
double zeta basis sets so the actual preferred adsorption site cannot be determined
accurately from these calculations.76 The molecular dynamics studies of Hammond and
Mahanti (using Lennard-Jones potentials) found that dipolar molecules preferred an

adsorption site slightly removed from the atop position with the molecules in a “tipped”

82

N
01

 

 

 

 

 

: A
205
j I CFZBr2 + 4 Ring
1 MP2/6-31G*
15~
I ' I Atop
A 10: o Hollow
73 ‘ A Brid e
E : g
2 s-
>. Z
9 .
0:) J
“J 0.. A A
2 I I
-5; A
3 o I
j A
-10-1 I f a
2 o o
'15 I I I I I I I I I I I
4 5 6

C - 4 Ring Distance (Angstroms)

Figure 5.4 Potential energy curves for CF2Br2 approaching the four ring substrate at the MP2/6-
316* level of theory.

83

geometry (molecular dipole not aligned with surface normal).36 These molecules tended
to shift slightly from the atop site towards a bridge site, i.e. the carbon atom of the
adsorbate moved to a position over a CC substrate bond which was the global minimum
on the potential energy surface. As the adsorbate changed position, shifts in the binding
energy and vertical distance above the surface occurred. For example, the largest shift in
energy (0.11 kJ/mol) and distance (0.29 A) occurred for CH3C1. The energy differences
between the hollow and bridge sites calculated here are larger than those of the MD
simulations of Hammond and Mahanti.36 However, typical diffusion barriers of
adsorbates on metal and semiconductor surfaces range between 10-25% of the binding
energy77 so diffusion between all of these positions may readily occur.

The results of Figure 5.5 revealed a similar trend in the preferred adsorption site
and binding energies as the size of the basis set was increased for the system. The
potential energy curves calculated with the 6-310(2d) basis set are not as broad as those
calculated with the 6310* basis set and the minimum distance above the surface for
each site has shifted to 4.0 A compared to values of 4.1 A (hollow) or 4.25 A (atop and
bridge) for the 6-31G* basis set. Also, an increase in the binding energies as well as the
energy differences between the hollow and atop and the hollow and bridge sites occurred.
Diffusion of the adsorbate on the surface will probably occur even though the differences
between the binding energies at each site have increased. It should be noted that the goal
of this study is not to obtain the most accurate binding energy for the adsorbate to the
surface but to gain information about preferred adsorption sites. The larger basis set

provides a better description of the electronic properties of both the adsorbate and the

84

_s
U"

 

 

 

3 CFZBr2 + 4 Ring
10 l A MP2/6-31G(2d)
5 j - Atop
. ' o Hollow
3 A Bridge
0 _
9 ' A
O .-
g 1 ° -
.2. -5-
>. 1
9 - A
1% -1o-: '
4 A
: I f
-15 - o
I A i
. . I .
-20 1 D
'25 a ' I I I ' I I I I . I
4 5 6

C - 4 Ring Distance (Angstroms)

Figure 5.5 Potential energy curves for CF2Br2 approaching the four ring substrate at the MP2/6-
31G(2d) level of theory.

85

 

substrate so the interaction energy will increase and the distance above the substrate will
decrease.

Alternatively, the size of the substrate was increased while fixing the basis set (6-
316*) to examine variations in the binding energy or sites that occurred and also to
observe changes due to rotation of the molecule on the surface (Figures 5.6-5.9). The
molecule was rotated about an axis perpendicular to the surface that contained the C atom
of the adsorbate and a point directly below the molecule on the surface (see Figures 5.6
and 5.7). Comparison of the 7-ring and 4-ring surfaces utilizing the same basis set
revealed that the potential energy curves were not as broad and the binding energies at all
sites increased in the former case (Figure 5.8). The values obtained for the binding
energies on the 7-ring system using the 6-31G* basis set fall between those of the 4 ring
system using the 6-BIG* and 6-31G(2d) basis sets and follow the same trend in
adsorption site energies (hollow, atop, bridge) as seen previously. An increase in the
binding energies is expected due to the larger van der Waals attraction between the
surface and the molecule. As a first approximation, these values demonstrate that
changing the size of the substrate had less of an effect than changing the size of the basis
set.

On the 7 ring system, the minimum vertical distance of CF2Br2 above the surface
occurs at 4.0 A which is the same for the pyrene/6-31G(2d) system but slightly shorter
than the value for the pyrene/6-310* system. The energy difference between the hollow
and atop sites for the coronene/6-3lG* system becomes noticeably smaller (0.58 kJ/mol) .

compared to the other systems discussed previously (1.31 kJ/mol for 4 ring/6-31G* and

86

.83.. we: 82... _oeoa as 8 Season .33 3 e8 52.2. 3. do... E 2e 2. Been? 3.5 3. 23$

8V 3V 3

 

87

data 9E 8% 32: ea 8 2528a $63 9 2a .252 Q .33 3 2a a Bag £50 833 E 2:53

 

88

Energy (kJ/mol)

N
O

 

 

 

 

 

 

.. 10-
. ‘ I
a + . 0
. CF28r2 7 Rlng g 5. Rotated
15 _ MP2/6-31G* g 0;
j I Atop “2:34;:
a C t A
10 _ A C HC.)||OW “1.10. ; ‘
- A Bridge . . .
‘ . ~15. . a
. I 3.'5 * 4.‘o ' 45
5 - C - 7 Ring Distance (Angstroms)
: O
o .:
‘ A
‘ U
-5 _
f A
.J ‘ .
-10 _' ‘
‘ I
: o ‘ ‘ I
-15 5 I
- I
1
‘20 I I I I I I I I I I I I I
3 4 5 6

C - 7 Ring Distance (Angstroms)

Figure 5.8 Potential energy curves for CFzBrz approaching the seven ring substrate at the
MP2/6816* level of theory.

89

 

2.55 kJ/mol for 6-3lG(2d)). From this comparison, it may be concluded that the actual
adsorption site may lie between these sites.

Rotation of the molecule on the surface revealed similar magnitudes of binding
energies compared to non-rotated systems. It can be seen that the hollow and atop sites
are now equal in energy with the bridge site still having the smallest binding energy.
From these values, it appears that the molecule can freely rotate at each adsorption site
because there is almost no energy barrier to rotation. The molecular orientation does not
seem to make much difference for the binding energy to the substrate for the two
orientations tested here.

Additional calculations were performed on the 7-ring system using the 6-3lG(2d)
basis set with the molecule in a “non-rotated” orientation. A smaller number of
calculations were done on this system around the minimum of the potential energy
surface due to the large number of basis functions involved. Figure 5.9 shows a
comparison between the system using the 6-3lG* and the 6-31G(2d) basis sets. Similar
results were obtained for the energy ordering of the adsorption sites (hollow site
preferred) and the minimum vertical distance above the surface (4 A). Using a
comparatively larger double zeta basis set for the calculation yielded ~ 58% of the
binding energy as determined from temperature programmed desorption data.44 Use of
larger basis sets than those here would obtain a greater binding energy but it would
become computationally prohibitive to do so. From the data presented here, it appears
that the most probable adsorption site for CF2Br2 on graphite occurs at the hollow site

with free rotation of the molecule.

90

 

 

 

 

 

 

 

 

 

_ej onBr2 + 7 Ring -16; CFzBrz + 7 Ring
. 6-31G* . 6-31G(2d)
-7- -17-
. A i
-8- -18—
. ‘ ‘ A
-9- -19-
’3‘ -10- 320-
g - 2 4
fi '7
5, -11 d - é—Z‘l - A
> , >s ‘ I
9 -12- 9122‘ o
a) . A ‘ 8 .
‘fi -13- ° ”-234
I
-14- -24-
. . .
-15. ’ -25- .
, I I g
-16.. o -26-
-17 . . . -27 . . .
3.5 4.0 4.5 3.5 4.0 4.5
C - 7 Ring Distance (Angstroms) C - 7 Ring Distance (Angstroms)

Figure 5.9 Comparison of the effects of basis sets on the binding energies and adsorption sites for
CF2Br2 on the model seven ring surface. The squares denote the atop site, circles
denote the hollow site, and the triangles denote the bridge site.

91

An analysis of the Mulliken atomic charges and dipole moments of the complex
revealed the preference for the hollow site. Table 5.2 shows atomic charge and dipole
moment values for gas-phase CFzBrz and the 4-ring and 7-ring complexes for each basis
set. Calculated dipole moments of 0.7051 D (6-3lG*) and 0.6004 D (6-31G(2d)) were
obtained for gas phase CFzBrz. These values compare well to the experimental value of
0.66 D.78 When the molecule is adsorbed over the hollow site in each of the systems
studied, an increase in dipole moment occurred compared to the gas-phase value, while
the dipole moments for the atop and bridge sites decreased. This trend was not observed
for rotation of the molecule on the 7-ring substrate. Comparison of the dipole moments
for the atop and hollow sites for the rotated molecule on the 7-ring complex revealed
similar values and nearly identical binding energies.

Upon adsorption of the molecule on the substrate, charge is transferred from the
surface to the adsorbate. The amount of charge transferred is dependent upon the
adsorption site of the molecule. In the structures of Figure 5.6 for example, it was
observed that the largest amount of charge transfer occurred at the bridge site. At this
location, three atoms of the adsorbate (C and both Br atoms) are over sites of increased
electron density. An increase in charge on the “up” F atom (F atom farthest from the
surface) and both Br atoms compared to the gas-phase molecule can be observed in Table
5.2. The increase in charge on these atoms leads to a subsequent lowering of the dipole
moment because the charge separation within the molecule is reduced. A comparatively
small amount of charge transfer occurred at the hollow site. Here, most of the adsorbate

atoms are over sites of lower electron density except for the “down” F atom (F atom

92

 

35895 05 no mom—Eu o5 .«o 83 05 Ba macros—m 05 :o woman—0 05 mo 88 08 5933 “Canasta u a

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

«88.—0&5
«$8.? £33. 333. 383. $83- $.83. ~53- 883. 0956
«23 83° 33 E3 53 33 383 33 é a
$2.3. $33. 833. 8:3. ”53- 233. 5 S3. $2.3. .._ ago?
8283 2883 2383 $83. $33. 383. 333 :33. am
233. 88m? 8333. Read. 3E3. SE3. 2: s3. ”33. e :9...
8283 8383 2383 :53. 833. 833. 82.83 :33. am
8233 3.33 ”833 «$83 $33 $283 353 3:83 0
seem 323m 82 3pm 326: 82 939 me .9 3
32 £32 3 was. 2.23 can; so use so

 

 

6an» :03 no «ammo .8.“ Bony—obfi guano 93 85:58 38% .3930 0:52 dd 02am.

93

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

008.5me
823- 383. 833- $33. 383. $83- 283- $83- :83- ”wane
:33 ~93 323 323 833 33 «$3 383 «33 a:
8E3. 3:33. 283. 3883- 283- “53. $2.3- 383. 833. m .38..
$383 :53 :883 $83- 2 2.3- $33- a 53- S 2.3. 383- cm
8333. 2&3- $33- 2&3- 2 $3. $23. 3333. 233. 3&3. m :9:
$283 $383 3233 $33. 2:33. ”$83. 3283. :53. 333- am
383 8333 $833 2333 383 $883 2283 «883 3283 o
owvtm 323: QS< owvtm Bo=om n03. 03:5 30:03 n02

may. 2959 3 38335 $02-3 use 2.93

 

 

 

 

 

$.83 .3 2.5.

94

closest to the surface). Most of the transferred charge is localized on this F atom and the
C atom of the adsorbate and augments the already existing dipole moment. The atop site
showed the smallest amount of charge transfer between the surface and the molecule
even though the molecular orientation is similar to the bridge site. The two Br atoms of
CF2Br2 are over C atoms in the substrate, but now the C atom of the molecule is over a
surface C atom as well. In this configuration, the charge on the Br atoms and the “up” F
atom were observed to increase (become more negative) in most cases analogous to the
bridge site. A smaller amount of charge transfer occurred because the C atom in the
adsorbate is now over a region of lower electron density compared to the bridge site.
Therefore, the dipole moment decreases slightly due to a small change in the charge
separation in the molecule. This shift of charge within the molecule can best be seen by
observing the difference between the sum of the charges on the fluorines and the sum of
the charges on the bromines as seen in Table 5.2. This difference in charges shows that
the atop and bridge sites have comparable values while the hollow site has a larger
difference. However, the differences are reversed for the atop and hollow sites when the
molecule is present on the 7 ring substrate even though the dipole moment is slightly
larger at the hollow site.

A variety of studies exist that have examined the structure of polar molecules

28'36'4L67'7‘ Knorr and co-workers have investigated the

adsorbed on graphite surfaces.
structure of CFzClz on HOPG at a variety of coverages and temperatures.”71 They
determined from x-ray diffraction and IRAS data that 6 out of 9 CFzClz molecules in the

unit cell sit on the PC]; tripod at an atop site for the commensurate on phase. The other

three molecules are determined to be positioned above hollow sites and are

95

orientationally disordered with respect to the other molecules in the unit cell. The 0L
phase represents a low coverage-low temperature phase that should be comparable to the
structures calculated here. The results obtained from these calculations show that at very
low coverages the CF2Br2 molecule will sit over a hollow site which is slightly preferred
over the atop site and compares well to the structure obtained by Knorr.”71 Caution
must be exercised when comparing the preferred adsorption site from these calculations
to higher adsorbate coverages. It is believed that these dipolar molecules do not wet the
graphite surface but instead form 2-dimensional islands because of stronger inter-
molecular interactions.“

5.3.2 Adsorption Sites and Geometries of CF2Br. The results of the interaction
between CF28r and the model graphite surfaces are shown in Figures 5.10-5.15.
Comparison of the potential energy curves between CF2Br and CF2Br2 adsorbed on the
model graphite surfaces showed similar characteristics (broad shallow minima and
similar vertical adsorption heights). For example, Figure 5.10 shows the interaction
between CFzBr and the 4-ring substrate using the 6-31G* basis set. The minimum
vertical adsorption height of the radical occurred at 4.0 A, which is comparable to the
value obtained for CFzBrz interacting with the 4 ring surface (~ 4.1 A) using the 6-31G*
basis set. Also, it was observed that the binding energies of the radical to the model
surfaces were smaller than those calculated for CFzBrz on the model surfaces. The
decrease in binding energies can be explained by noting that the dominant interaction
between each species and the surface occurs through dispersion forces. The radical
“borrows” electron density from the surface to increase its stability (as discussed below),

but no covalent bond is formed due to the strong C-C bonding within the substrate.

96

1O

 

Energy (kJ/mol)

-10..

 

CF28r + 4 Ring

MP2/6-31G*
I Atop
O Hollow
A Bridge

 

Figure 5.10 Potential energy curves for CFzBr approaching the four ring substrate at the MP2/6-
310* level of theory.

1 l

5

C - 4 Ring Distance (Angstroms)

97

 

 

A change in the preferred adsorption site was observed to occur between CFzBr
and CF2Br2. The radical favors either the bridge or atop site (depending on the size of the
substrate) while the molecule preferred the hollow site. The radical contains a singly
occupied molecular orbital that is directed perpendicular to the plane containing the
fluorine and bromine atoms of the molecule. Addition of electron density to this orbital
stabilizes the interaction between the radical and the surface. Therefore, positions that
contain higher amounts of electron density, such as the atop and bridge sites, are favored
by the radical. Density functional theory calculations of the bare seven ring surface by
Arellano, et. a1. show that the atop and bridge sites contain higher amounts of electron
density compared to the hollow site.27 At present, no experimental evidence or
theoretical calculations have been performed on this or any similar system so the true
geometry and orientation of the molecule with respect to the surface are unknown.

Expanding the size of the basis set used for CF2Br adsorbed on the 4-ring system
to the 6-31G(2d) basis set increased the binding energy at each adsorption site. A more
accurate description of the system is obtained with this basis set when compared to the
calculations using the 6-31G* basis set. It can be seen from Figure 5.11 that a similar
trend in the preferred adsorption site occurred when compared to the results obtained
with the 6-31G* basis set. Now, the atop and bridge sites are separated by only 0.29
kJ/mol with the atop site having a slightly greater binding energy. A similar energy
difference was observed for the 6-31G* basis set (0.33 kJ/mol). Also, the minimum
vertical adsorption height above the surface for each site has shifted to a smaller value of
3.75 A and an increased curvature of the potential energy curve near the equilibrium

bond distance occurred when compared to the 6-31G* calculation.

98

 

 

 

 

 

CFZBr+4Ring
MP2/6-31G(2d)
I Atop
0' o Hollow
A Bridge
I
‘5‘
E -5-
i 3
V O
>~.
9’
(D O
C A
UJ I
-10.
A
o O I
I
A
A
I
~15- ‘
3' "21' ' 'é' ' '63

C - 4 Ring Distance (Angstroms)

Figure 5.11 Potential energy curves for CF2Br approaching the four ring substrate at the MP2/6-
31G(2d) level of theory.

99

.m 2:5
Hammo «—
3&8?

. E a. a

3 2a .323 Q 8a

a

350; 0&3.—

2: no me . .

>8 $608

6an wat do

3
3
3

 

data we. :28 38a 2: 8 Sousa owes 3 Ea 98:2 5 den 3 2c a 89.83 ammo 850m 2.... an»...

A3 3 3

 

101

Additional calculations were performed using the 6-31G* basis set to investigate
changes in the binding energies and adsorption sites as the size of the substrate was
increased. A further series of calculations were carried out on this system with rotation
of the radical on the surface analogous to the CF2Br2/7 ring system (see Figures 5.12 and
5.13). It can be seen from Figure 5.14 that the values for the binding energies of the
radical fell between those of the 6-31G* and 6-31G(2d) basis sets on the 4 ring surface.
Similar behavior was observed in the molecular system as explained previously. It is
interesting to note that the preferred adsorption site switched from the atop to the bridge
site and the minimum vertical adsorption decreased to ~ 3.75 A compared to 4.0 A. The
other adsorption sites remain unchanged with minimum vertical distances above the
surface of 4.0 A. A closer approach of the radical to the surface at a bridge site occurs
because more electron density is present at this position and the atoms of the adsorbate
are not directly over those of the substrate (reduced steric hindrance).

Rotation of the radical on the surface showed slight changes in the adsorption
energies. A similar trend was observed for the energy ordering of the adsorption sites
compared to the non-rotated system, but now the minimum vertical distance for the
bridge site returned to 4.0 A. This effect is probably due to the atoms in the radical
taking positions over top of the atoms in the surface. It was observed that the binding
energy at the atop site decreased indicating an unfavorable orientation of the radical on
the surface possibly due to a smaller amount of electron density transferred at this site.
Because of the small energy differences between the rotated and non-rotated orientations,

the radical may rotate freely at the preferred adsorption site.

102

 

10

 

 

 

 

 

 

-6 o Rotated
CFZBr + 7 Ring .5 -7* '
x
MP2/6-31G* 5 -8‘ _
jig-9‘ O
0 . . A
54 I Atop ““10
J O Hollow 1“ ‘ ‘
‘ Bridge '12 3:5 ' 420 ' 415
C - 7 Ring Distance (Angstroms)
3 0-
E
.3 i
as I
>.
0)
L
8
LIJ _5_ - ‘
A
‘ A
-10- '
A D
3 4 5 6

C - 7 Ring Distance (Angstroms)

Figure 5.14 Potential energy curves for CFzBr approaching the seven ring substrate at the
MP2/6616* level of theory.

103

 

4“

 

 

 

 

 

 

 

 

 

-3i CFzBr + 7 Ring ‘ crzar + 7 Ring
, 6-31G* 42* 6-31G(2d)
-4. ‘ A
. -13- u
-5- a .
A 6‘ A44-
5 - _ '5
g . g I
2 -7. ‘ 3.15.
{=3 ‘ §
.05) '8‘ ljag-w- .
i
-9. .
-17~ ‘
401 : J l
-11- ‘ . '18- ‘
‘ 4
-12 I . r n ‘19 l ' l '
3.5 4.0 4.5 3.5 4.0 4.5
C - 7 Ring Distance (Angstroms) C - 7 Ring Distance (Angstroms)

Figure 5.15 Comparison of the effects of basis sets on the binding energies and adsorption sites
for CF2Br on the model seven ring surface. The squares denote the atop site, circles
denote the hollow site, and triangles denote the bridge site.

104

Figure 5.15 shows a comparison of the binding energies of the radical on the 7-
ring surface using both the 6—3IG* and 6-3lG(2d) basis sets. It was observed that the
binding energies obtained using the 6-31(3(2d) basis set were larger than those with the 6-
31G* basis set. Similar values were obtained for the binding energies (differing by a
maximum of 0.55 kJ/mol) with the bridge site preferred over the hollow and atop sites.
The minimum height above the surface remained at 3.75 A which is the same value for
the bridge site when compared to the 6-3 16* basis set on 7-ring system. It appeared that
the bridge site is the preferred adsorption position for the radical on HOPG because it is
lowest energy site on 7-ring system using either basis set. Depending on the magnitude
of the diffusion barriers between sites, the radical may be able to easily diffuse across the
surface because of the similarity in binding energy of each site. It may be recalled that
these calculations represent the zero-coverage limit and the preferred adsorption sites and
orientation may change with increased coverage.

In comparison to the adsorption of the molecule on the surface, the radical
preferred adsorption sites where the dipole moment of the complex has its lowest value as
seen in Table 5.3. This table shows the atomic and total charges and dipole moments of
the radical in the gas phase as well as on the 4- and 7-ring substrates calculated using
both basis sets. The interaction between surface and radical is not as dipolar in nature
when compared to the molecule-surface interaction. It will be recalled that the radical
favors the at0p and bridge sites because it has a singly occupied molecular orbital to
which it can add this extra electron density. The decrease in dipole moment is the result
of the radical obtaining charge from the surface and spreading it evenly over all the atoms

in the radical.

105

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

£83. «:83. E33. 883. 383. 883.3. 3 o 3&5 Bob
33 «33 883 53 :33 £93 33 ~83 1
$83. 863. SS3. 233. 883. $33. $83. 333. m
~33. “$23. :83. :33. $83. $33. $83. 883. m
3:3 3:33 3.23 32.3. 32.3. 383. 38.3 853. am
$33 833 E83 :33... 2833 83.33 8383 «333 o
085 30%: 82. 3.5 323m .5... £503 .9 3
8.2m 9.23
max 352 3 max 3.9 3 so no

 

 

68.35. :98 no Banana 38 v5 £8808 28.6 .8920 283“ umumo .n.m 033—.

106

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

owaao
~33. 283. $83. ~88? 383. :83- 9883 8883 :83. :89
E3 083 ~83 :33 ~33 33 323 333 333 1
$33- 883- :33- 333. 833. $33- 233. $33. 833. m
833. 583. :33- :33. 833- 833. 3383- 833. $33. .._
83:3 853 S383 333- ~33- R83- 3823- :33- 353. .m
333 :33 333 8333 333 8833 £833 253 2:33 o
3cm 323m 82 3cm 328m 8% 08cm 3%: 82

max .2892 3 333 max 8.9 3 max 2.9 3

 

 

€383 .3 2.3.

107

Comparison of the gas phase atomic charges to those of the radical on both
surfaces provided further insight into the interaction of the radical with the graphite
surface. The C and Br atoms of the adsorbed radical (with the 6-31G* basis set) had
slightly more negative values than in the gas phase while the F atoms became slightly
more positive. As a result, the charge separation in the radical is reduced and more
electron density goes into the unfilled orbital. Use of the larger basis set showed that the
F atoms gained charge from the surface but the amount that they acquired was nearly
offset by the amount of positive charge on the C and Br atoms. Therefore, the radical
gained more total charge but the dipole moment was reduced to half of its gas-phase
value.

Adsorption of the radical on the 7-ring surface with the 6-31G* basis set showed
that the preferred binding site was over the bridge position as seen in Figure 5.14. A
decrease in the dipole moment was observed due to a slightly larger negative charge on
the Br atom and a slight positive charge on the C atom (compared to the gas phase
values) when adsorbed at this site. It appeared that the radical donated charge to the
surface as evidenced by the decreased total charge on CF2Br. Rotation of the radical on
the surface caused a change in the charges on the Br and C atoms to occur. The Br atom
was slightly more positive in this orientation and the C atom, as well as one of the F
atoms, became more negative compared to the non-rotated case. In this position, the F
atom that is part of the OP bond that is over a C atom in the substrate is closer to a site of
higher electron density which allowed both the F and C atoms to become more negative
and charge shifted away from the Br atom. As a result, a return of charge transfer from

the surface to the radical was observed.

108

 

Calculations of the radical on the 7-ring substrate with the 6-316(2d) basis set
showed the same preferred binding site as the 6-31G* basis set and now the signs and
magnitude of the atomic charges are comparable to those of the gas phase and adsorption
on the 4-ring surface. The C atom became more positive, and the Br and F atoms were
more negative than in the gas phase so charge became more evenly distributed. A return
of charge from the surface to the radical was observed to occur helping to stabilize the
interaction between the radical and the surface.

The data presented here can be compared to that obtained from our study of the
photochemistry of CF2Br2 on a highly ordered pyrolytic graphite (HOPG) surface.44 In
this photodissociation study, a slight peak shift and change in intensity as well as a
change in background in EELS data were observed after irradiation of a monolayer of the
molecule. This was attributed to Br atoms being produced on the surface that were
involved in charge transfer from the surface to the Br atoms. From the data presented
here, it appears that the molecule and radical are involved in this charge transfer process
as well. The combination of two open shell systems (CF2Br and Br) on the surface can
account for the large change in EELS background that is characteristic of a metal surface.
In addition, the increase in intensity and shift of the EELS peak also can be explained.
For longer irradiation times, more radicals will be present on the surface. The dipole
moment CF2Br is nearly aligned with the surface normal in the adsorption geometry of
this study. The asymmetric stretch of the radical will have a larger component of its
dipole moment perpendicular to the surface than the molecule resulting in an increase in
the intensity of the peak. The red shift in the peak occurred because the radical is not as

strongly bound to the surface compared to CF2Br2. A comparison of the results presented

109

 

here with the values obtained from the TPD data is limited. Approximately 58% of the
experimental binding energy of CF2Br2 on HOPG was obtained from these calculations
utilizing the largest surface and basis set. With the knowledge of the preferred adsorption
sites for each species, future simulations of both the pre- and post-irradiated monolayer
will allow comparison to the TPD data.

5.4 Conclusions. It was found that the preferred adsorption site for the CF2Br2
molecule occurred over the hollow position of a six-membered ring on the surface for
each basis set and model graphite surface studied. The dominant interaction between the
molecule and the surface was dipolar in nature with charge transfer occurring from the
surface to the molecule. It was observed that the site with the largest dipole moment for
the complex was the preferred one. Rotation of the molecule on the surface showed that
the molecule probably undergoes free rotation on the surface. The preferred adsorption
site for the CFzBr radical occurred at the atop site on the 4 ring substrate and the bridge
site on the 7 ring substrate for each of the basis sets used. The interaction between the
radical and the surface was not as dipolar in nature compared to CF2Br2 even though
charge was transferred from the surface to the radical. Sites of higher electron density
(bridge and atop) were found to be preferred due to the presence of a singly occupied
orbital in the radical. Rotation of the radical on the surface showed no significant
changes in binding energy at the preferred site so free rotation of the species probably

OCCUI'S .

llO

 

10.

11.

12.
13.
14.

15.

16.

17.

References

Gould, R. J .; Salpeter, E. E. Astrophys. J. 1963, I38, 393.
Hollenbach, D.; Salpeter, E. E. Astrophys. J. 1971, I63, 155.
van der Hulst, H. C. Rec. Astron. Obs. 1949, XI(II).
Williams, D. A.; Herbst, E. Surf. Sci. 2002, 500, 823.

Ruthven, D. M. Principles of Adsorption and Adsorption Processes; Wiley:
New York, 1984.

Rogaski, C. A.; Golden, D. M.; Williams, L. R. Geophys. Res. Lett. 1997, 24,
381.

Lary, D. J .; Lee, A. M.; Toumi, R.; Newchurch, M. J .; Pirre, M.; Renard, J. B.
J. Geophys. Res. 1997, 102, 3671.

Smith, D. M.; Akhter, M. 8.; Jassin, J. A.; Sergides, C. A.; Welch, W. F.;
Chughtai, A. R. Aerosol Sci. Tech. 1989, 10, 311.

Kirchner, U.; Scheer, V.; Vogt, R. J. Phys. Chem. A 2000, 104, 8908.

Saathoff, H.; Naumann, K.-H.; Riemer, N.; Kamm, S.; M6hler, 0.; Schurath,
U.; Vogel, H.; Vogel, B. Geophys. Res. Lett. 2001, 28, 1957.

Chughtai, A. R.; Jassim, J. A.;.Peterson, J. H.; Stedman, D. H.; Smith, D. M.
Aerosol Sci. Tech. 1991, 15, 112.

Schurath, U.; Naumann, K.-H. Pure and Appl. Chem. 1998, 70, 1353.
Al-Abadleh, H. A.; Grassian, V. H. J. Phys. Chem. A 2000, 104, 11926.
Akhter, M. S.; Chughtai, A. R.; Smith, D. M. J. Phys. Chem. 1984, 88, 5334.

Gerecke, A.; Thielmann, A.; Gutzwiller, L.; Rossi, M. J. Geophys. Res. Lett.
1998, 25, 2453.

Kalberer, M.; Tabor, K.; Ammann, M.; Parrat, Y.; Weingartner, E.; Piguet,
D.; Rossler, E.; Jost, D. T.; Tfirler, A.; Gaggeler, H. W.; Baltensperger, U. J.
Phys. Chem. 1996, 100, 15487.

Kleffmann, J.; Becker, K. H.; Lackhoff, M.; Wiesen, P. Phys. Chem. Chem.
Phys. 1999, I, 5443.

111

 

18.

19.

20.

21.

22.

23.

24.

25.

26.

27.

28.

29.

30.

31.

32.

33.

34.

35.

36.

Longfellow, C. A.; Ravishankara, A. R.; Hanson, D. R. J. Geophys. Res.
1999, 104, 13833.

Stadler, D.; Rossi, M. J. Phys. Chem. Chem. Phys. 2000, 2, 5420.
Grassian, V. H. J. Phys. Chem. A 2002, 106, 860.

Choi, W.; Leu, M.-T. J. Phys. Chem. A 1998, 102, 7618.
Baldwin, A. C. Int. J. Chem. Kinet. 1982, 14, 269.

Britton, L. 6.; Clarke, A. G. Am. Environ. 1980, I4, 829.

Sergides, C. A.; Jassim, J. A.; Chughtai, A. R.; Smith, D. M. Applied Spec.
1987, 41 , 482.

Disselkamp, R. S.; Carpenter, M. A.; Cowin, J. F.; Berkowitz, C. M.;
Chapman, E. G.; Zaveri, R. A.; Laulainen, N. S. J. Geophys. Res. 2000, 105,
9767.

Poschl, U.; Letzel, T.; Schauer, C.; Niessner, R. J. Phys. Chem. A 2001, 105,
4029.

Arellano, J. S.; Molina, L. M.; Rubio, A.; Alonso, J. A. J. Chem. Phys. 2000,
112, 8114.

Feller, D.; Jordan, K. D. J. Phys. Chem. A 2000, 104, 9971.
Sorescu, D. C.; Jordan, K. D.; Avouris, P. J. Phys. Chem. B 2001, 105, 11227.
Lushington, G. H.; Chabalowski, C. F. J. Molec. Struc. 2001, 544, 221.

Tran, F.; Weber, J .; Wesolowski, T. A.; Cheikh, F.; Ellinger, Y.; Pauzat, F. J.
Phys. Chem. B 2002, 106, 8689.

Ferro, Y.; Marinelli, F.; Allouche, A. J. Chem. Phys. 2002, 116, 8124.

Tanaka, H.; El-Merraoui, M.; Steele, W. A.; Kaneko, K. Chem.Phys. Lett.
2002, 352, 334.

Nyman, G.; Holmlid, L. J. Chem. Phys. 1986, 85, 6163.
Hansen, F. Y.; Taub, H. J. Chem. Phys. 1987, 87, 3232.

Hammond, W. R.; Mahanti, S. D. Surf. Sci. 1990, 234, 308.

112

37.

38.

39.

40.

41.

42.

43.

44.

45.

46.

47.

48.

49.

50.

51.

52.

53.

54.

55.

Hentschke, R.; Schilrmann, B. L.; Rabe, J. P. J. Chem. Phys. 1992, 96, 6213.
Hentschke, R.; Winkler, R. G. J. Chem.Phys. 1993, 99, 5528.

Pinches, M. R. S.; Tildesley, D. J. Surf Sci. 1996, 367, 177-

Kim, Y. H.; Ree, J.; Shin, H. K. Chem. Phys. Lett. 1999, 314, 1.

Kwon, 8.; Russell, J.; Zhao, X.; Vidic, R. D.; Johnson, J. K.; Borguet, E.
Langmuir 2002, 18, 2595.

Zhao, X.; Kwon, S.; Vidic, R. D.; Borguet, E.; Johnson, J. K. J. Chem. Phys.
2002, 117, 7719.

Finlayson-Pitts, B. J.; Pitts, Jr., J. N. Chemistry of the Upper and Lower
Atmosphere; Academic Press: San Diego, 2000.

Dorko, M. J .; Bryden, T. R.; Garrett, S. J. J. Phys. Chem. B 2000, 104, 11695.
Jacox, M. E. Chem. Phys. Ben. 1977, 53, 192.

Johnson, C. A. F.; Ross, H. J. J. Chem. Soc. Farad. Trans. 1 1978, 74, 2930.
Sam, C. L.; Yardley, J. T. Chem. Phys. Lett. 1978, 61 , 509.

Wampler, F. B.; Tiee, J. J.; Rice, W. W.; Oldenborg, R. C. J. Chem. Phys.
1979, 71, 3926.

Molina, L. T.; Molina, M. J. J. Phys. Chem. 1983, 87, 1306.

Krajnovich, D.; Zhang, 2.; Butler, L.; Lee, Y. T. J. Phys. Chem. 1984, 88,
4561.

Gosnell, T. R;. Taylor A. J .; Lyman, J. L. J. Chem. Phys. 1991, 94, 5949.

Talukdar, R. K.; Vaghijani, G. L.; Ravishankara, A. R. J. Chem. Phys. 1992,
96, 8194.

Vatsa, R. K.; Kumar, A.; Naik, P. D.; Rama Rao, K. V. S.; Mittal, J. P Chem.
Phys. Lett. 1993, 207, 75.

Van Hoeymissen, J.; Uten, W.; Peeters, J. Chem. Phys. Lett. 1994, 226, 159.

Vatsa, R. K.; Kumar, A.; Naik, P. D.; Rama Rao, K. V. S.; Mittal, J. P. Bull.
Chem. Soc. Japan 1995, 68, 2817.

113

 

56. Talukdar, R. K.; Hunter, M.; Warren, R. F.; Burkholder, J. B.; Ravishankara,
A. R. Chem. Phys. Lett. 1996, 262, 669.

57. Felder, P.; Yang, X.; Baum, G.; Huber, J. R. Isr. J. Chem. 1993, 34, 33.

58. Surjan, P. R.; Csaszar, P.; Poirier, R. A.; van Lenthe, J.-H. Acta Phys.
Hungarica 1990, 67, 387.

59. Ishida, H.; Palmer, R. E. Phys. Rev. B 1992, 46, 15484.

60. Ancilotto, F.; Toigo, F. Phys. Rev. B 1993, 47, 13713.

61. Lou, L.; Osterlund, L.; Hellsing, B. J. Chem. Phys. 2000, 112, 4788.

62. Jeloaica, L.; Sidis, V. Chem. Phys. Lett. 1999, 300, 157.

63. Ishikawa, S.; Madjarova, G.; Yamabe, T. J. Phys. Chem. B 2001, 105, 11986.
64. Chalasinski, G.; Szczesniak, M. M. Chem Rev. 1994, 94, 1723.

65. Hobza, P.; Selzle, H. L.; Schlag, E. W. Chem Rev. 1994, 94, 1767.

66. Rassolov, V. A.; Pople, J. A.; Redfem, P. C.; Curtiss, L. A. Chem. Phys. Lett.
2001, 350, 573.

67. Knorr, K. Phys. Rep. 1992, 214, 113.

68. Knorr, K.; Civera-Garcia, E. Surf. Sci. 1990, 232, 203.

69. Nalezinski, R.; Bradshaw, A. M.; Knorr, K. Surf. Sci. 1995, 333, 255.

70. Volkmann, U. G.; Knorr, K. Phys. Rev. B 1993, 47, 4011.

71. Warken, A.; Enderle, M.; Knorr, K. Phys. Rev. B 2000, 61 , 3028.

72. Boys, S. F.; Bemardi, F. Mol. Phys. 1970, 19, 553.

73. Frisch, M. J.; Trucks, G. W.; Schlegel, H. B.; Scuseria, G. E.; Robb, M. A.;
Cheeseman, J. R.; Zakrzewski, V. G.; Montgomery Jr., J. A.; Stratmann, R.
E.; Burant, J. C.; Dapprich, S.; Millam, J. M.; Daniels, A. D.; Kudin, K. N.;
Strain, M. C.; Farkas, 0.; Tomasi, J.; Barone, V.; Cossi, M.; Cammi, R.;
Mennucci, B.; Pomelli, C.; Adamo, C.; Clifford, S.; Ochterski, J.; Petersson,
G. A.; Ayala, P. Y.; Cui, Q.; Morokuma, K.; Salvador, P.; Dannenberg, J. J .;

Malick, D. K.; Rabuck, A. D.; Raghavachari, K.; Foresman, J. B.; Cioslowski,
J.; Ortiz, J. V.; Baboul, A. G.; Stefanov, B. B.; Liu, 0.; Liashenko, A.;

114

 

74.

75.

Piskorz, P.; Komaromi, I.; Gomperts, R.; Martin, R. L.; Fox, D. J .; Keith, T.;
Al-Laham, M. A.; Peng, C. Y.; Nanayakkara, A.; Challacombe, M.; Gill, P.
M. W.; Johnson, 8.; Chen, W.; Wong, M. W.; Andres, J. L.; Gonzalez, C.;
Head-Gordon, M.; Replogle, E. S.; Pople, J. A. Gaussian 98 (Revision A”);
Gaussian Inc.: Pittsburgh, PA, 2001.

N. M. S. D. Center; S. Stein, In NIST Chemistry WebBook, NIST Standard
Reference Database Number 69; Mallard, W., Linstrom, P., Eds.; National
Institutes of Standards and Technology: Gaithersburg, MD, 1998
(http://webbook.nist.gov).

Hazel], A. G; Larsen, F. K.; Lehmann, M. S. Acta Cryst. 1972, 828, 2977.

76. Jensen, F. Introduction to Computational Chemistry; John Wiley and Sons:

77.

78.

Chicester, 1999.
Seebauer, E. G.; Allen, C. E. Prog. Surf. Sci. 1995, 49, 265.

Lide, D. R., Ed.; CRC Handbook of Chemistry and Physics, 75th ed.; CRC
Press: Boca Raton, FL, 1995.

115

Chapter 6 Conclusions and Future Work

6.1 Experimental. In the experimental investigation of the photochemistry of
CF2Br2 adsorbed on HOPG it was revealed that direct excitation of the adsorbate
occurred to produce photoproducts similar to those observed from the gas-phase
molecule. It appeared that the HOPG surface was not directly involved in the
photodissociation event however; the surface did provide a medium upon which other
products (most notably C2F4Br2 and Br2) were formed. Comparison of the cross sections
of CF2Br2(g) at the photolysis wavelengths used in this study to those calculated from
TPD and other data showed that dissociation of the adsorbate by attachment of low-
energy photoelectrons produced by ultraviolet irradiation of the surface did not occur to
any measurable degree.

This finding is significant because it is well known that metal surfaces produce
bou nd sub-vacuum or free photoelectrons when irradiated with ultraviolet light.1 These
electrons can be captured by the anti-bonding orbitals of an adsorbate, and dissociation of
the molecule may occur. This process is known as dissociative electron attachment
(DEA). It would be expected that a similar DEA-type process might occur for adsorbates
on an HOPG surface because it has a band structure similar to that of a metal. Graphite is
a Setnimetal and its band structure, as well as those of a typical metal, semiconductor, and
insulator is illustrated in Figure 6.1.2 In a semimetal, the valence band overlaps slightly
in energy with the conduction band to produce a small number of holes in the valence

band and a corresponding number of electrons in the conduction band.2 Theoretically,

116

 

 

 

 

 

 

 

 

 

 

 

 

Energy

 

 

 

 

 

 

 

 

 

 

 

 

Figure 6.1 Band structures of an insulator, metal, semimetal, and semiconductor.2 Shaded areas
represent filled bands.

H7

 

electrons present in the graphite conduction band, which terminates at the Fermi level,
can be excited with ultraviolet radiation and captured by the anti-bonding orbitals of
CF2Br2(ad) resulting in dissociation of the molecule. The apparent lack of a DEA-type
process could be the result of a change in the surface properties or band structure of the
graphite.

Results obtained from ab initio calculations performed on this system suggest a
small amount of charge transfer occurred from the surface to the molecule upon
adsorption. This charge transfer process would create a graphite surface with a net
positive charge that would appear as an increase in the surface work function.
Experiments performed on bromine intercalated graphite compounds have shown that
charge is transferred from graphite to the bromine with a concurrent change in the
graphite work function between 0.5 and 1.5 eV.?"4 As a consequence, the number of
photoelectrons with sufficient energy to cause dissociation in the range of wavelengths
used here would decrease. Therefore, an accurate measurement of the work function of
graphite as a function of CF2Br2 coverage should be determined. An increase in work
function is evident as an increase in the separation of the adsorbate and substrate bands
and so charge-transfer leading to work function changes also affect processes driven by
sub-vacuum level electrons.

Poor energetic or spatial overlap of the anti-bonding orbitals of CF2Br2 with the
1:. band of graphite could generate a low probability for excitation of the adsorbate by
substrate electrons. Alternatively, favorable overlap may produce rapid quenching of any
adsorbate excited state formed through intra-adsorbate excitation. The rate of quenching

is dependent upon the lifetime of the hole generated by irradiation of the surface.5 If the

118

 

hole lifetime is long compared to bond breaking in the adsorbate, then dissociation of the
molecule will occur. The hole lifetime is determined by the resonance between the
adsorbate and substrate orbitals that are involved in the excitation process. Charge
transfer can occur between energetically resonant orbitals more easily than between non-
resonant orbitals. If a resonant charge transfer process cannot occur between the
adsorbate and substrate, then the likelihood of dissociative electron attachment occurring
will be drastically reduced. Dissociative electron attachment cross sections for Halons
and CFCs are largest for electrons with near zero electron kinetic energy.6 Large
numbers of these electrons should be available at all the wavelengths used in this
experiment. The relative positions of the graphite valence bands with respect to the
occupied orbitals of the adsorbate can be deduced from ultraviolet photoelectron
spectroscopy (UPS).

Further studies of other Halons, CFCs, and their possible replacements also
should be attempted to compare possible photodissociation pathways for this family of
molecules while adsorbed on HOPG. Dichlorodifluoromethane (CFC-11) has been
extensively studied by Knorr and co-workers and, they have proposed a structure for a
monolayer of CFzClz on graphite.7'll Ultraviolet irradiation of this monolayer may or
may not cause a similar photodissociation event to occur when compared to that obtained
for CF2Br2. Similar results would imply that direct excitation of CFzClz occurred, while
the appearance of dissimilar photoproducts could indicate that a DEA-type process may
have occurred. Photodissociation studies of other molecules in this family such as CF3Br
(Halon 1301), CHzBrz and C2F4Br2 (Halon 2402) should be investigated to determine

their photodissociation pathways as well.

119

Although crude measurements of the total photodissociation cross section for the
adsorbed and gaseous molecule were found to be comparable, excitation within the
substrate should not be dismissed. Additional experiments should be conducted to
confirm that direct adsorbate excitation predominates in this system. If the direct
excitation mechanism is the prevailing one, then the photodissociation cross section will
depend on lp. - E|2 where E is the electric field vector and p. is the transition dipole
moment.5 Experiments could be performed as a function of radiation intensity or
polarization to measure if an increase in photodissociation occurs. Similar experiments
have been used by Zhu et. al. to separate excitation in the substrate, where the E field
intensity is determined by the Fresnel equations, from excitation in the adsorbate,
exhibiting a simple c0326 dependence on polarization angle 9.12

An additional series of experiments could be performed where the excitation
wavelength is varied and the degree of photodissociation or product yield probed. If
direct intra-adsorbate excitation occurs, then the wavelength of maximum absorption for
a physisorbed molecule should be comparable to that of the gas phase species. The
wavelength of maximum absorption for CF2Br2(g) occurs at 226 nm.”"4 This
wavelength was available from our excitation source but the lamp fluence in this
wavelength region was very low, and no filters were available to isolate this wavelength.
An ideal radiation source for these studies would be a continuously tunable source such
as a Xe arc lamp and monochromator or a dye laser system.

Although graphite has been used as a simple model for a carbon aerosol particle
surface, the model can be improved in a number of ways. Hence, future experiments

should be performed by utilizing modified graphite surfaces. While carbonaceous

120

IvVL'll'llr

 

aerosols may have small domains that resemble HOPG, the majority of the surface
contains a large number of defects.15 One way to introduce defects on the surface is by
ion sputtering. Ion sputtering of the graphite surface will disrupt the pure sp2 conjugation
present in the unsputtered surface and sites of increased electron density will be created.
Previous studies of graphite surfaces sputtered with noble gas ions have shown that both
vacancy (removal of surface C atoms) and interstitial defects (trapping of the noble gas
within the graphite) were formed.”20 Molecules may preferentially adsorb at these sites,
and changes in the photochemistry may occur as a result of the change in the local
density of states (LDOS) of HOPG.

Experiments where the HOPG surface is sputtered with a reactive gas such as
molecular oxygen should be undertaken as well. Ion sputtering studies performed with
molecular oxygen revealed defects similar to those obtained with the noble gases, and
incorporation of oxygen at defects occurred to produce carbonyls and other oxygen-
containing functional groups.”22 Further oxidation of these sputtered surfaces revealed
new functional groups, such as quinones and ethers, were present which could alter both
the adsorption properties as well as photochemistry occurring on the surface.
Alternatively, etch pits of controlled sizes and depths also can be prepared on an HOPG

surface by heating the sample in air at 650 °C.23'24

It is expected that aged carbonaceous
aerosols will have oxidized defects that will influence any photochemistry that occurs on
the surface. Carbonaceous particles become oxidized as they age in the atmosphere and

heterogeneous reactions that once occurred on a newly formed particle may no longer

take place. By introducing defects in each of these ways, the photochemistry can be

121

 

observed as a function of the number of defects and type of functionalities present on the
surface.

Modification of the HOPG surface also can be accomplished by the adsorption of
straight chain alkanes and polycyclic aromatic hydrocarbons (PAHs). A large fraction of
aged carbonaceous aerosols have been found to contain varying amounts of both
polycyclic aromatic hydrocarbons and long-chain alkanes present on their surface.15
Addition of these adsorbates will alter the electronic properties of the surface. Possible
photo-oxidation of the PAHs may lead to an enhanced photolysis rate for CF2Br2 or
possible reaction between them. Bond cleavage in a PAH will lead to sites of higher
electron density (“dangling bonds”) that can be easily donated to the adsorbate. To date,
there have been no studies investigating changes in the adsorption or the photochemistry
of adsorbates on graphite using hydrocarbon spacer layers.

Additional studies using atmospherically important small molecule co-adsorbates
on the photochemistry of CF2Br2 adsorbed on HOPG could be performed. One molecule
that is present in large atmospheric quantities is water. Adsorption of water ice on HOPG
has been studied previously,”27 but its influence on the photochemistry of co-adsorbates
has not been examined. It is well known that polar molecules, such as water ice, can act
as low energy electron traps.”31 Madey and co-workers reported a large increase in the
amount of F" and Cl' ions produced when they electron irradiated water and ammonia ice
surfaces covered with CF2C12.32'36 Secondary electrons produced from the irradiation of
the surface were shown to increase dissociative electron attachment to CFzClz adsorbed

on the condensed H20 and NH3 layers by transfer of the secondary electrons in precursor

7

States to the CF2C12 molecules.3 If sub-vacuum or photoelectrons are produced by

122

ultraviolet irradiation of the HOPG surface, then an increase in the photodissociation of
CF2Br2 should be observed. The thickness of the water layer, as well as its crystal
structure, also can be varied. Roberts and co-workers have shown different adsorption
behaviors for polar H-bonding molecules on amorphous and crystalline ice surfaces.”41
It also has been observed that the degree to which DEA occurs in ice films varies with
film thickness, crystallinity, and temperature.30

Lastly, soot samples obtained from a variety of natural sources could be used as
the adsorption substrate. Carbonaceous aerosols that are produced from the combustion
of different types of fossil fuels will have differing chemical compositions and particle
size distributions depending on the combustion conditions.15 These factors may influence
the number and type of adsorption sites on the surface and any possible photochemistry
that may occur. Newly formed carbonaceous aerosols would be expected to have a larger
number of adsorption sites compared to an aged aerosol. However, an aged aerosol may
be coated with adsorbates that provide routes for alternative photochemical processes to
occur.

6.2 Theoretical. It was shown that the preferred adsorption site for the CF2Br2
molecule occurred over the hollow site of a six-membered ring on each surface
independent of the basis set used. The interaction between the molecule and the surface
was observed to be dipolar in nature with the largest dipole occurring over the hollow
site. Examination of the Mulliken atomic charges showed that (negative) charge was
transferred from the surface to the molecule. The amount of charge transferred depended

upon the position of the atoms in the adsorbate with respect to the atoms of the substrate.

At the hollow site, a larger amount of charge was transferred to the F atom closest to the

123

 

surface than at the other positions due to its proximity to larger amounts of electron
density.

It was determined that the interaction between CF2Br and the model HOPG
surfaces involved borrowing of electron density from the surface by the singly occupied
orbital of the radical. For the four ring substrate, the preferred adsorption site on HOPG
was located at the atop position. However, the preferred adsorption site shifted to the
bridge position when CF2Br was adsorbed on the seven ring substrate. Both sites contain
larger amounts of electron density than at the hollow site.

To obtain the most accurate description of both the molecule and the radical on
the HOPG surface, complete geometry optimization calculations should be done using
larger basis sets and the seven ring substrate to represent the surface. Bigger PAHs could
be used to model the graphite surface however, this would result in an increase in the
number of basis functions used in the calculations and a large increase in the amount of
computational resources as well. Larger basis sets allow for better correlation of the
valence electrons in the system and would yield binding energies comparable to those
obtained from TPD experiments. Vibrational frequencies obtained from the complete
optimization calculations could be compared to both pre- and post-irradiated EELS data
to assess the results of the calculations.

To further investigate the role of charge transfer in this system, a series of
calculations should be performed on the interaction of Br adsorbed on a model HOPG
surface. Considerable amounts of charge transfer are expected for this system as
evidenced by experimental studies of intercalation compounds involving bromine.3‘4 As

a result, single-reference wave function methods are inappropriate for this system.42

124

 

Initial MP2 calculations performed on Br interacting with both the four and seven ring
surfaces using the 6-3lG* and 6-31G(2d) basis sets revealed that the Br atom was not
bound to the substrate. Use of multi-reference wave function methods (CASSCF,
CASPT2, MRCISD) and larger basis sets will allow for a bound, charge-transfer
interaction to occur between Br and the surface. By examining the preferred adsorption
site for Br, additional information about the post-irradiated surface can be obtained.

Placing two CF2Br2 molecules on a sufficiently large surface slab and performing
geometry optimizations would allow for additional information about the structure of the
monolayer to be obtained. Experimentally, the growth of the monolayer is thought to
occur by two-dimensional islanding of CF2Br2 on the surface due to stronger
intermolecular interactions compared to the adsorbate-substrate interaction.““15 The
strength of each of these interactions as well as changes in preferred adsorption sites and
geometries should be studied. Similar information could be obtained for the post-

irradiated surface by performing calculations on CF2Br2 and CF2Br co-adsorbed on a
model HOPG surface. Information concerning variations in the adsorption enthalpy with
changes in the molecular orientation and coverage would be obtained as well.

Molecular dynamics calculations may lend further insight into the structure of the
monolayer both before and after irradiation, and would provide information on the
processes leading to formation of specific product molecules. In order to study the
structure of the monolayer, accurate intermolecular potential energy surfaces need to be
developed for various orientations between species using the same methods as described
previously. These potentials would be fit to a functional form that will be used in

modeling both mono- and multilayer growth of CF2Br2 as well as the structure of the

125

monolayer after each period of irradiation with UV light. The structure of the un-
irradiated monolayer obtained from these calculations can be compared to that proposed
for CFzClz by Knorr and co-workers.7'll The results obtained from the post-irradiated

surface can be used to rationalize the experimental TPD data.

126

10.

ll.

12.

13.

14.

15.

16.

References

. Zhou, X.-L.; Zhu, X.-Y.; White, J. M. Surf Sci. Rep. 1991, 13, 73.

Kittel, C. Introduction to Solid State Physics( 7th Edition); John Wiley and Sons,
Inc.: New York, 1996.

Cox, P. A. The Electronic Structure and Chemistry of Solids; Oxford University
Press: Oxford, 1987.

Ellis, A. B.; Geselbracht, M. J .; Johnson, B. J .; Lisensky, G. C.; Robinson, W. R.
Teaching General Chemistry: A Materials Science Companion; American
Chemical Society: Washington, DC, 1993.

Zhou, X.-L.; White, J. M. In Laser Spectroscopy and Photochemistry on Metal
Surfaces, Part II; Dai, H.-L.; Ho, W., Eds.; World Scientific: New Jersey, 1995;
Advanced Series in Physical Chemistry, Vol. 5.

Underwood-Lemons, T.; Gergel, T. J.; Moore, J. H. J. Chem. Phys. 1995, 102,
119.

Knorr, K. Phys. Rep. 1992, 214, 113.

Knorr, K.; Civera-Garcia, E. Surf Sci. 1990, 232, 203.

Nalezinski, R;. Bradshaw, A.M.; Knorr, K. Surf Sci. 1995, 333, 255.
Volkmann, U. G.; Knorr, K. Phys. Rev. B 1993, 47, 4011.

Warken, A.; Enderle, M.; Knorr, K. Phys. Rev. B 2000, 61 , 3028.

Zhu, X.-Y.; White, J. M.; Wolf, M.; Hasselbrink, E.; Ertl, G. Chem. Phys. Lett.
1991, 176, 459.

Orkin, V. L.; Kasimovskaya, E. E. J. Atmos. Chem. 1995, 21 , l.

Burkholder, J. B.; Wilson, R. R.; Gierczak, T.; Talukdar, R.; McKeen, S. A.;
Orlando, J. J.; Vaghjiani, G. L.; Ravishankara, A. R. J. Geophys. Res. 1991, 96,
5025.

Seinfeld, J. H.; Pandis, S. N. Atmospheric Chemistry and Physics; John Wiley and
Sons: New York, 1998.

Li, T.; King, B. V.; MacDonald, R. J .; Cotterill, G. F.; O’Conner, D. J .; Yang, Q.
Surf Sci. 1994, 312, 399.

127

17. An, B.; Fukuyama, S.; Yokogawa, K.; Yoshimura, M. J. Vac. Sci. Technol. B
1999, 1 7, 2439.

18. Kappel, M.; Ktippers, J. Surf Sci. 1999, 440, 387.

19. Hahn, J. R.; Kang, H.; Lee, S. M.; Lee, Y. H. J. Phys. Chem. B 1999, 103, 9944.
20. Hahn, J. R.; Kang, H. Surf Sci. 2000, 446, L77.

21. Nowakowski, M. J.; Vohs, J. M.; Bonnell, D. A. Surf Sci. 1992, 271, L351.

22. Nowakowski, M. J .; Vohs, J. M.; Bonnell, D. A. J. Am. Ceram. Soc. 1993, 76,
279.

23. Chang, H.; Bard, A. J. J. Am. Chem. Soc. 1990, 112, 4598.

24. Stevens, F.; Kolodny, L. A.; Beebe, Jr., T. P. J. Phys. Chem. B 1998, 102, 10799.
25. Chakarov, D. V.; Osterlund, L.; Kasemo, B. Langmuir 1995, 11, 1201.

26. Chakarov, D. V.; Osterlund, L.; Kasemo, B. Vacuum 1995, 46, 1109.

27. Chakarov, D. V.; Kasemo, B. Phys. Rev. Lett. 1998, 81, 5181.

28. Gilton, T. L.; Dehnbostel, C. P.; Cowin, J. P. J. Chem. Phys. 1989, 91, 1937.

29. Bass, A. D.; Sanche, L. J. Chem. Phys. 1991, 95, 2910.

30. Simpson, W. C.; Orlando, T. M.; Parenteau, L.; Nagesha, K.; Sanche, L. J. Chem.
Phys. 1998, 108, 5027.

31. Lu, Q.-B.; Sanche, L. J. Chem. Phys. 2001, 115, 5711.

32. Lu, Q.-B.; Madey, T. E. J. Chem. Phys. 1999, 111, 2861.
33. Lu, Q.-B.; Madey, T. E. Phys. Rev. Lett. 1999, 82, 4122.
34. Lu, Q.-B.; Madey, T. E. Surf Sci. 2000, 451, 238.

35. Lu, Q.-B.; Madey, T. E. J. Phys. Chem. B 2001, 105, 2779.

36. Lu, Q.-B.; Madey, T. E.; Parenteau, L.; Weik, F.; Sanche, L. Chem. Phys. Lett.
2001, 342, l.

37. Lu, Q.-B.; Sanche, L. Phys. Rev. B 2001, 63, 153403.

128

38. Schaff, J. E.; Roberts, J. T. J. Phys. Chem. 1994, 98, 6900.

39. Graham, J. D.; Roberts, J. T. Geophys. Res. Lett. 1995, 22, 251.
40. Schaff, J. E.; Roberts, J. T. Langmuir 1999, 15, 7232.

41. Schaff, J. E.; Roberts, J. T. Surf Sci. 1999, 426, 384.

42. Vila, F. D.; Jordan, K. D. J. Phys. Chem. A 2002, 106, 1391.
43. Golze, M; Grunze, M.; Hirschwald, W Vacuum 1981, 31 , 697.
44. Niemantsverdriet, J. W.; Markert, K.; Wandelt, K. Appl. Surf Sci. 1988, 31 , 211.

45. Zhou, X. —L.; White, J. M.; Koel, B. E. Surf. Sci. 1989, 218, 201.

129

Appendix

130

Appendix A

Calculation of CF2Br2 Fractional Coverage on HOPG

The fractional coverage of CF2Br2 was calculated as the number of adsorbate

molecules per surface C atom. The fraction of surface-adsorbed CF2Br2 was calculated

—a
(I) 1_ex _i
LAX§A= ABS[ {’1‘ C089]]
13 SA —a
1—<I> +ex A

where ass is the effective “size” or height of the adsorbate (CF2Br2) and is determined

as follows:

(A.1)

by dividing the molecular mass of the adsOrbate by its density multiplied by Avagadro’s
number and then taking the cube root, AA is the inelastic mean free path of photoelectrons
ejected from atom A of the adsorbate, AB is the inelastic mean free path of photoelectrons
ejected from atom B of the adsorbate, (DABS is the fractional coverage of the adsorbate, IA
is the area as measured from XPS data of peak A, 13 is the area as measured from XPS
data of peak B, SA and 8;; are atomic sensitivity factors for atoms A and B, and 9 is the
angle between the surface normal and the ejected photoelectrons. The following values
were used in calculating the fractional coverage given atom B is carbon and atom A is
bromine: aABs = 5.33 A, ltc = 14 A, 21.3, = 15 A, 0 = 65°, 13, = 63044, [C = 109456, 83, =
0.83, and Sc = 0.25. A value of 0.24877 is obtained that must be divided by 2 (there are
two bromine atoms in CF2Br2) to obtain a fractional coverage of 0.124. A similar

calculation for the fractional coverage using the fluorine (atom A) and carbon atoms

131

 

(atom B) with the following values (aABs = 5.33 A, Ac 2 14 A, A}: = 11 A, 0 = 65°, Ic =
109456, 1;: = 116761, Sc = 0.25, 8;: = 1.00) yielded a value of 0.317. Dividing by 2 (two
F atoms in CF2Br2) gives 0.159. An average was taken for the two values obtained from

the calculations and gave a fractional coverage of 0.14.

132

IIIIIIIIIIIIIIIIIIIIIIIIIIIIII

WillllWWW.l‘l‘llllillllllllll'l1|

3 1293 02493 8551 I