v.5.

z. .3. 51:3.
2. I‘Vnufi... . .
.1...|..th.. 5:4 4. .
.Méizati : ..
x x a, 11.11..

:3: .‘ a. E

u ‘
31.2...th .
3:11 :42

. I L“

:43 x. .

‘ ‘20.

,. .... 53.5.. As
{1.13.5 XI

.25.. i...
.3.

. I:
«.31. 11

I: Ova-A523.
n a.
(a «3.1,: . ?
Yul». .5“. x...

. filtrluful l
. .w. 9.154 .
, . 7
z) ‘ y 3 . ‘3 :14}:
.5 2
l ht.— ..sin .§.
3. v k.-
..\...V1 i0... v1. 5‘ m H

.5. \pl. 9.: a.
in“! I." ‘
. . z . 51 .931. A!

(I: A. {35, ‘1... . ‘

I :29! s. ‘ ‘

$2.... .. . 1 ‘

.4 vtvux.
27...).3:

~11... I: I).

I}; (3.4..

(J . V
y‘ 'i‘ l-
3136...“.
7 (I.

 

5L... .

1.:

. .5:
in”: ants “ .

.3.

v.33

fisfigwfi

 

40°7/

This is to certify that the

thesis entitled

Adsorption and Photochemistry of 1,3-Butadiene
on HOPG(OOO1)

presented by

Jason K. Oman

has been accepted towards fulfillment
of the requirements for

 

 

M.S . degree in Chemistry

3%

WI professor

0-7639 MS U is an Affirmative Action/Equal Opportunity Institution

 

 

LIBRARY
Michigan State
University

 

 

 

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 c:/CIRC/DateOue.p65-p. 15

ADSORPTION AND PHOTOCHEMISTRY OF 1,3-BUTADIENE ON HOPG (0001)
By

Jason Kenneth Oman

A THESIS

Submitted to
Michigan State University
in partial fulfillment of the requirements
for the degree of

MASTER OF SCIENCE
Department of Chemistry

2002

ABSTRACT

ADSORPTION AND PHOTOCHEMISTRY OF 1,3-BUTADIENE ON
HOPG(0001)

By
JASON KENNETH OMAN

The adsorption and photochemistry of 1,3-butadiene (C4H6) on HOPG(0001) at < 85
K was studied using temperature-programmed desorption (TPD) and electron energy loss
spectroscopy (EELS). Butadiene weakly adsorbs on HOPG, the monolayer (ML)
exhibiting first order desorption with a maximum desorption rate at 129 K and a
corresponding desorption energy of 32 kJ/mol. Multilayer butadiene showed zero order
desorption with a maximum desorption rate at 105 K and a desorption energy of 28.5
kJ/mol. Irradiation Of the C4H6/HOPG(0001) surface (1 and 3 monolayers Of C4116) with
UV photons (193 nm, 248 nm, and 351 nm) at 100 mJ/pulse resulted in molecular
photodesorption Of the adsorbed butadiene layers. The cross section for photodesorption
was calculated at each wavelength for both coverages. For 1 monolayer coverage, the
cross-sections were 3.3x10'20 cmz, 2.4x10'20 cm2, and 1.2x 10'20 cm2 for 193 nm, 248 nm,
and 351 nm, respectively. The 3 monolayer cross-sections were 1.1x10'l7 cmz. 6.3x 10'Ix
cm2, and 1.8x10"8 cm2 for 193 nm, 248 nm, and 351 nm, respectively. It was found the
photodesorption mechanism could be described by a simple thermal process. The photo-
processes occurring in both 1 and 3 ML C4H6(ad)/HOPG were also investigated at lower
photon fluences Of S 10 mJ/pulse. Under these conditions, there was no detectable

photodesorption or condensed phase photochemistry.

ACKNOWLEDGMENTS

I would like tO thank everyone who was involved with my work at MSU. Especially
Dr. Simon Garrett for his guidance, support and I’m sure at times patience. I would also
like to thank all the group members (Todd, Lilli, Mike and Heather), and I can’t forget
the new group members Anne and Steve. They have all been helpful with my research
and provided a break from work every now and again.

I would also like to thank my family and friends for without their support during the
course Of this work it would not have been accomplished, (mom, dad, Debi, Andrew,
Matt, Dave and Eric). I would also like to give Melissa a special thanks for having to

put-up with me during the most stressful times here at MSU.

iii

TABLE OF CONTENTS

LIST OF TABLES

LIST OF FIGURES

INTRODUCTION

1.1

Heterogeneous Atmospheric Chemistry

1.2 Atmospheric Photochemistry
1.3 Carbonaceous Particles in the Atmosphere
1.4 Unsaturated Hydrocarbons in the Atmosphere
1.5 Proposed System
1.6 References
EXPERIMENTAL
2.1 Survey Of Analytical Techniques
2.2 Experimental Apparatus
2.3 References

ADSORPTION AND PHOTOCHEMISTRY OF 1,3-BUTADIENE ON HOPG

3.1 Adsorption of Butadiene on HOPG determined by TPD

3.2 Adsorption of Butadiene on HOPG determined by EELS

3.3 Photodesorption of Butadiene monitored by TPD

3.4 Photodesorption of Butadiene monitored by EELS
DISCUSSION

3.5 Mechanisms for Photodesorption

3.6 Adsorbate-mediated Processes

3.7 Substrate-mediated Processes

3.8 Laser-induced Thermal Desorption

3.9 Conclusions

CONCLUSIONS AND FUTURE WORK

4.]

Adsorption and Photochemistry Of Butadiene on HOPG

Page

vi

vii

#

\OOLII

10

20

20
25
30

31

31
33
38
43

45

45
45
46
47

56

56

4.2 Future work

4.3 References
APPENDICES

Appendix A Mass Spectra
Appendix B Calculation Of Desorption Energies for Butadiene on HOPG
Appendix C Calculation for energy absorbed by HOPG during laser irradiation

References

57
63

64
65
67
68
69

LIST OF TABLES

Page
Table 3.1 Vibrational assignments Of 1,3-Butadiene adsorbed on
HOPG(0001). 37
Table 3.2 Butadiene photodesorption cross sections and gas phase

cross sections from the literature. 41

vi

Figure 1.1

Figure 1.2

Figure 1.3

Figure 1.4

Figure 1.5

Figure 1.6

Figure 2.1

Figure 2.2

Figure 2.3

Figure 2.4

LIST OF FIGURES

Variation Of temperature and pressure from ground level to
100 km above the Earth’s surface.

Surface catalyzed heterogeneous reaction between chlorine
nitrate and a polar stratospheric cloud particle (PSC).

Comparison Of the solar spectrum outside the atmosphere, a
5800 K blackbody radiator and the solar spectrum at sea
level.

General oxidation sequence for hydrocarbons and organic
molecules in the atmosphere.

Single-bond rotation isomerization of 1,3-butadiene.

Graphite crystal structure showing basal plane dimension
and bond lengths.

Schematic diagram Of X-ray photoelectron generation.

Representation of a dipole perpendicular to a surface and
its image dipole in the surface (left) and a dipole parallel to
a surface plane with its mirror image in the surface (right).

X—ray photoelectron spectrum of the HOPG surface before
experiments.

Experimental apparatus showing excimer laser and UHV
chamber orientation.

vii

Page

21

24

27

28

Figure 3.1

Figure 3.2

Figure 3.3

Figure 3.4

Figure 3.5

Figure 3.6

Figure 3.7

Figure 3.8

Mass (C3H3+) thermal desorption spectra of C4H6 on HOPG
with increasing coverage. At 2.0 ML coverage the lst
monolayer saturates.

Thermal desorption spectra for masses 39 (C3H3+), 27
(C2H3+) and 54 (C4H6+) Of C4H6 on HOPG for S 3 ML
coverage to ensure the lst monolayer is saturated.

Electron energy loss spectra of (a) 1 ML and (b) 3 ML
C4H6 on HOPG.

Thermal desorption spectra (mass 39, C3H3+) Of 1 ML C4H6

on HOPG irradiated with increasing photon number (A =
193 nm).

Thermal desorption spectra (mass 39, C3H3+) Of 3 ML C4H6

on HOPG irradiated with increasing photon number (A =
193 nm).

Semi-logarithmic plot of TPD peak area vs. total number Of
photons/cm2 for calculation Of the photodesorption cross
section.

Electron energy loss spectra Of 3 ML C4H6 on HOPG
exposed to 1x10'7 photons/cmz.

Result for calculating the surface temperature rise using
equation 1 when HOPG is exposed to a 16 ns 100 m] pulse
Of 193 nm light.

viii

Page

32

34

36

39

40

42

44

50

Figure 3.9

Figure 4.1

Figure 4.2

Figure A]

Percent of C4H6 monolayer desorbed from the surface with
increasing pulse power. The total number of photons was
kept constant at each pulse power by varying the number Of
pulses.

Temperature programmed desorption spectra (mass 94) for
CH3Br on HOPG with increasing exposure time; a) 1 sec,
b) 5 sec, c) 10 sec, (1) 20 sec, e) 40 sec, 1) 60 sec.

X-ray photoelectron spectra of 1 ML CH3Br exposed tO 5
mJ/pulses at 10 Hz of 193 nm light for increasing times; a)
30 min, b) 10 min, c) 5 min, d) 2.5 min, e) 0.5 min, f) 0 min
showing binding energy changes.

Mass spectra Of (a) 1,3-butadiene at a pressure Of 5x10'9
Torr and (b) the background vacuum of the UHV chamber
at a pressure Of 2 x 10-10 Torr.

Page

51

59

60

66

INTRODUCTION

1.1 Heterogeneous Atmospheric Chemistry

The atmosphere is a layer Of gas that surrounds the Earth and contains molecules
necessary for respiration, helps regulate temperature and climate on the Earth’s surface and
acts as a sink for gases and particles produced from biological activity, combustion processes
and anthropogenic sources. Temperature and pressure both vary throughout the atmosphere;
Figure 1.1 shows a vertical profile from 0 to 100 km above the Earth’s surface.1 Notice, the
atmosphere is divided into several separate layers which are described based on temperature
gradients. The troposphere, closest to the Earth’s surface, contains 80% Of all atmospheric
mass and the majority of all atmospheric species, both molecular and particulate. Above the
troposphere is the stratosphere; this layer Of the atmosphere contains a high concentration of
ozone (up to 10 ppm) which filters UV radiation preventing it from reaching the Earth’s
surface.1 The mesosphere and thermosphere are high-altitude layers Of the atmosphere
containing few molecular species and are characterized by low pressures.

Over the past few decades, heterogeneous reactions on solid particle or aerosol surfaces
have been found to play an important role in atmospheric chemistry. Examples Of aerosol
particles include liquid water, ice, sea-salt, metal oxides, aluminosilicates and carbonaceous
matter or mixtures of some Of these species. 192 A majority of the research conducted to date

has focused on heterogeneous reactions on model polar stratospheric cloud surfaces (PSCs)

that contribute to ozone destruction.3'5 For example, the reaction between ice or nitric acid-

 

 

 

 

100 (1001
90
--—- 001
80
70
0.1
60
E 3’
as 50 ------------------ 1 g
0 s
"g (D
E 40 3"
< 10 v
30
20 100
10
0 w1000

 

l60 180 200 220 240 260 280 300

Temperature (K)

Figure 1.1 Variation of temperature and pressure from ground level to 100 km above the
Earth’s surface.l

containing ice and chlorine nitrate molecules produces molecular chlorine that is liberated
from the ice surface.5 Molecular chlorine is photolyzed by sunlight (k < 380 nm) to produce
free chlorine atoms that subsequently participate in reactions destroying ozone (see Figure
1.2). The formation of the ozone hole is a yearly event; surfaces that catalyze reactions
producing photo-destructive molecules are only present during the winter months. During
spring/summer the ozone-depleting species react in the gas phase to reform photoinactive
molecules that do not contribute to ozone destruction.

The impact heterogeneous reactions have on atmospheric chemistry may be much greater
than previously realized. One reason for this misconception is a relatively small portion of
atmospheric mass is condensed matter (< 2x 10‘6 %).l However, surfaces contain active sites
which catalyze reactions that would never be considered to be important in the gas phase.
Unfortunately, the composition and concentration of particulate matter varies greatly in both
time and space in the atmosphere making heterogeneous reactions in the atmosphere difficult
to predict and model. Understanding how particulate matter impacts chemistry in the
atmosphere and the reactions that these surfaces catalyze is crucial for understanding the
dynamics of the atmosphere.

Heterogeneous chemistry of N02 and H20 on soot or carbon black particles has been
shown to play an important role in tropospheric chemistry.6 Carbonaceous aerosols
reduce N02 in the presence of H20 to produce HONO. It has been shown that the gas
phase reaction between N02 and H30 is too slow to produce significant amounts of
HONO under atmospheric conditions.7 The heterogeneous production of HONO on

carbonaceous particles is fast enough to account for the high night-time levels of

C10N02(g) C12 + hv —> 2 C1-

 

 

 

 

 

 

A
2 Cl (ad)
2 ClON02(ad) ———>
2 NO3(ad)
PSC surface

Figure 1.2 Surface catalyzed heterogeneous reaction between chlorine nitrate and a polar
stratospheric cloud particle (PSC).5

HONO.7 This reaction is important because HONO is readily photolyzed in the
troposphere by sunlight at wavelengths between 380 and 330 nm8 to produce -OH and
NO-. The DH and NO- produced then participate in a series of important gas phase
reactions such as the production of 03 and the initial oxidation steps of organic molecules

in the troposphere.

1.2 Atmospheric Photochemistry

Much of the chemistry that takes place in the atmosphere is known to be the result of
photochemical processes. The sun behaves similar to a blackbody radiator emitting a broad
range of wavelengths (A = 100 nm to 1000 nm) that penetrate the Earth’s atmosphere.1 This
range of wavelengths enters the atmosphere and interacts with molecules inducing a variety
of photochemical processes. Ultraviolet and visible light is of sufficient energy to excite
electronic transitions in many molecular or atomic species. When a molecule absorbs a
photon of light it undergoes an electronic transition to an excited state. The energy between
the excited state and the ground state is equal to the energy of the photon that was absorbed.
Different atomic and molecular species absorb different wavelengths based upon energy
differences between their initial and final electronic states. Once in the excited state, a
molecule can undergo several different processes to lose the absorbed energy, including re-
emission of a photon via fluorescence, transfer of energy through collisions with other
molecules (called quenching), chemical reaction with another molecule to produce a new
species, or dissociation into two (or more) species (called photolysis).

Absorption of radiation resulting in photodissociation is very common in the atmosphere

and an important mechanism for preventing hi gh—energy radiation from reaching the Earth’s

surface where it can harm organisms. Figure 1.3 compares the differences in the solar
spectrum outside the atmosphere and at sea level. The reduction of total flux at sea level is
due to the absorption of radiation by atmospheric species in the ultraviolet, visible and
infrared regions of the solar spectrum.9 In the upper atmosphere (175 km-80 km) molecules
such as oxygen and nitrogen are the main species that absorb and filter out high-energy
radiation (A < 175 nm).10 Another species that filters out UV radiation 0» = 210 nm-300
nm)10 is ozone found in the stratosphere (15 km-50 km). Wavelengths > 300 nm can reach
both the troposphere and the Earth’s surface. Solar flux of wavelengths > 300 nm into the

troposphere is on the order of 1x10'4 photons cm'2 sec". 1

1.3 Carbonaceous Particles in the Atmosphere

Recent interest has focused on the atmospheric chemistry of carbonaceous or organic
aerosolsz’7’“ since a significant fraction (up to 30% in urban areas)1 of all atmospheric
particulate matter is carbonaceous.12 The composition of organic aerosol particles within
the atmosphere shows large spatial and temporal variations. The formation and exact
composition of these particles is complex, difficult to measure and not well
understood,13~l4 but they can include condensed biogenic organic molecules, soot, and
polycyclic aromatic hydrocarbons (PAHs). Incomplete combustion processes, such as
the burning of fossil fuels or biomass, are the major anthropogenic sources of the latter
two types of carbonaceous particles in the atmosphere.15 Penner and Novakov define
soot as consisting of a soluble organic fraction and an insoluble component that is
resistant to oxidation.15 The insoluble component has been variously termed elemental,

graphitic or black carbon. There is some evidence that this component contains a

 

 

 

 

4 Solar spectrum outside atmosphere
Blackbody radiator
I
I, \ Solar spectum at sea level
>< I \
:3
t: I
B \
‘5 I \
E" I \
I \
I
I
I ' \
I \\ \
I \ \ \
N —————
l l j l l l

 

 

200 600 1000 1400 1800 2200 2600 3000

Wavelength (nm)

Figure 1.3 Comparison of the solar spectrum outside the atmosphere, a 5800 K
blackbody radiator and the solar spectrum at sea level.10

graphite-like microcrystalline structure.l6 Black carbon particles can act as condensation
nuclei and as sites for heterogeneous reactions in the atmosphere. Indeed, it has been
speculated that fast reactions on black carbon surfaces could strongly affect atmospheric
chemistry. 15

The residence time of carbonaceous aerosols in the atmosphere is highly size-
dependant.1 Particles with a diameter of about 1 pm have a long residence time, about
1x106 sec at 1.5 km.17 Residence time decreases for particles that are either larger or
smaller than 1 um. For example, particles E 0.00] pm or E 10 um have residence times
of about 100 sec at an altitude of 1.5 km. Small particles, diameter _<_ 0.001 um are
typically removed through nucleation processes by sticking to larger particles in the
troposphere. Large aerosols, with diameters 210 um, undergo gravitational settling.
Intermediate size particles, about 1 um in diameter, are not significantly affected by
either deposition route and therefore have long residence times in the atmosphere.

Atmospheric aerosol surfaces can be labeled as being either rigid or dynamic.7 A
surface is considered to be dynamic when the vapor phase of the solid condenses on to
the surface at a high rate. This provides a consistently refreshed surface for any
heterogeneous reactions that are catalyzed by the surface. The surface of a rigid solid is
not refreshed regularly by a high condensation flux. A catalytically active rigid surface
can be permanently changed or altered by a reacting species; (this process is analogous to
the poisoning of an electrode surface). Of course, poisoning of the surface from one
reaction may allow the modified surface to catalyze another reaction. Graphitic or carbon

black aerosols are considered to be rigid solids.7

A typical young carbon black particle has a C/H ratio of ~3-413 and the particle
surface at this point is considered to be hydrophobic, preferentially adsorbing
hydrocarbon vapor over water vapor.7 The preference for hydrocarbon adsorption on
graphitic surfaces may play an important role in the heterogeneous chemistry of organic
molecules. Carbonaceous particles are susceptible to the formation of functional groups
on the surface.7 Typical functional groups are oxidation products such as mono or
dicarboxylic acids or C=O.14 The addition of oxygen-containing functional groups to the

carbon surface makes the surface more polar.

1.4 Unsaturated Hydrocarbons in the Atmosphere

Unsaturated organic molecules enter the atmosphere through a variety of pathways
including combustion of fossil fuels and biomass, industrial processes and biogenic
releases.l Biogenic sources (mainly plants) release significant amounts of unsaturated
organics including isoprene (2-methyl-1,3-butadiene) and monoterpenes. The yield of
these releases is estimated to be 1x1012 kg/yr.18 Forest fires are also known to release a
broad range of unsaturated molecular species to the atmosphere. One particular study has
shown an extensive list (about 20 species) of unsaturated molecules present in North
American forest fire plumes, including ethene, propene, 1,3-butadiene, pentene, isoprene
and octene.19 1.3-Butadiene is an example of an unsaturated organic molecule that is
commonly found in the atmosphere from several major sources including automobile
exhaust, forest fires and industrial processes. It is estimated that, in California, local

automobile exhaust contributes to about 96% of the annual emissions of 1.3-butadiene to

the troposphere. The average daytime concentration of butadiene measured in Los
Angeles is 2.2 ppb.20

Once in the atmosphere, unsaturated molecules undergo numerous oxidation reactions,

many ultimately leading to the final products C02 and H201 Strong atmospheric
oxidizers such as hydroxyl radicals, nitrate radicals, ozone and molecular chlorine initiate
the oxidation of unsaturated molecules. A wide range of reactive intermediates,
including organic peroxy radicals, organic nitrates, alkoxy radicals, aldehydes, ketones,
alcohols and organic acids may be formed. Figure 1.4 shows a general outline of
oxidation processes organic molecules undergo in the atmosphere. For example, in one

study, gas-phase butadiene was exposed to hydroxyl radicals and/or nitrate radicals and

nitric oxide,20 and the major products were found to be acrolein, HCHO, furan and
numerous organic nitrate isomers. A broad range of gas phase chemical reactions are
known to take place in the atmosphere, but it is still unclear what role the surfaces of

carbonaceous and other aerosols may play in catalyzing these reactions.

1.5 Proposed System

Our laboratory is interested in investigating adsorption and photochemistry of small
atmospherically relevant organic molecules on carbonaceous surfaces, using graphite as a
model for carbonaceous particulate matter. In this thesis, an initial study of the
adsorption and photochemistry of 1,3-butadiene on graphite is presented. Experiments
were conducted under ultrahigh vacuum (UHV) conditions, (pressure ~l.7x10'l0 torr).
The photochemistry was initiated with ultraviolet (193, 248 and 351 nm) photons

generated by an excimer laser. Analytical techniques used to study this system included

10

 

 

 

 

 

 

 

 

 

 

 

|RNo,|_£v_,[ RO-l hv ROOHI

 

 

 

 

 

 

 

 

 

 

 

Figure 1.4 General oxidation sequence for hydrocarbons and organic molecules in the

temperature programmed desorption (TPD) and high resolution electron energy loss
spectroscopy (HREELS). The goal of this work was to identify the adsorption structure
of butadiene on the graphite surface, and characterize any photochemistry that may take
place at the butadiene/graphite interface.

1,3-Butadiene was chosen both because it is environmentally relevant and its gas
phase chemistry has been well studied. It is the simplest conjugated molecule, consisting
of 4 carbon atoms that are spz hybridized resulting in a short enough bond length between
C2 and C3 for their p orbitals to somewhat overlap. Overlap between C2 and C3 gives
the central bond partial double bond character to allow for delocalization of all 4 1t
electrons over all of the 4 carbon atoms. Delocalized bonding of this nature increases the
stability of the molecule.”22 In theory, butadiene can form three different structural
isomers by rotation around C2 and C3, these are labeled s-trans, s-cis and gauche. Note,
the s preceding trans and cis designations refers to isomers formed by rotation around a
single bond.” The trans isomer, with Cgh symmetry, is the thermodynamically most
stable and at room temperature > 99 % of the molecules will take on this conformation.23
The cis conformer, with C3,. symmetry, is thought to be the second most stable
conformation of butadiene, and is formed by a 180° rotation around the C2-C3 bond as
shown in Figure 1.5. The cis (planar or almost planar) isomer is also stabilized by
delocalized 1t bonding across all four carbon atoms but unfavorable van der Walls
interactions between hydrogens on C l and C4 destabilizes the molecule relative to the
trans isomer by about 12 kJ/mol.21 The intermediate between the trans and cis isomers
is a gauche structure. In the gauche conformation, p orbitals between C2 and C3 will be

perpendicular to each other, preventing delocalized TC bonding throughout the system.

12

'
9 6—632! gauche

 

 

 

0
A
cis
:16 kJ/mo ./%_0.
o O

12 kJ/mol

 

 

trans
! V
1.1/3‘5“
0

Figure 1.5 Single-bond rotation isomerization of 1,3-butadiene.21

l3

This decreases the stability of the gauche isomer by about 28 kJ/mol relative to the trans
conformer.21

Gas-phase butadiene absorbs UV radiation from about 170 nm to 225 nm.24 The
wavelength of maximum absorption is 210 nm, where the absorption cross section is
1x10"l6 cmz.25 In this wavelength range, absorption results in a 1t——>1t* transition to two
closely spaced states (the 1Bu and 2Ag states). These states have short lifetimes of <100
fs, and decay from these states eventually leads to single-bond isomerization in about 270
fs.26 Haller et al. studied the photolysis of butadiene both in the gas phase and in a
cyclohexane solution.27 At low pressure, (4 mm Hg) the gas phase photolysis of
butadiene using wavelengths between 200 nm and 280 nm produced a series of products
such as hydrogen, acetylene, ethene, ethane, 1-butyne and 1,2-butadiene.27 It is thought
a vibrationally excited state is the precursor to the products. In the same experiment
butadiene was also found to polymerize on the walls of the vacuum chamber. In solution
the major products were found to be cyclobutene and bicyclobutane.27

Several ultrahigh vacuum studies have been published identifying adsorption
structures of butadiene on single metal crystals.28'32 The catalytic behavior of metal
surfaces toward selective hydrogenation of dienes and other unsaturated organic
molecules has spurred much of this work. Some of the most interesting work by
Bertolini et al. compared the chemistry of butadiene on Pt(l l 1) and Pd(l 1 1).28 On
Pt(l 11), the 1! bonds of butadiene are activated to produce a di-O' metallacycle species
where half the carbon atoms in the butadiene molecule rehybridize from spz to sp3.
Butadiene maintains its hybridization on the Pd(l 11) surface and forms a di-n bonded

species. Butadiene adsorption has also been studied on more inert surfaces, such as

Au(11 1) and Ag(11 1), where adsorbate/surface interactions are presumed to be weak.32
Based on inferred reflection-absorption spectroscopy, Osaka et al. suggests butadiene
adsorbs on Au(l l l) with its 1: bonds parallel to the surface in the trans conformation.

Highly ordered pyrolytic graphite (HOPG or graphite) will be used as a model for
carbonaceous aerosol surfaces. Highly ordered pyrolytic graphite is produced when
pyrolytic graphite is annealed at temperatures of 3000 °C. 33 The crystal structure of
graphite, shown in Figure 1.6, is composed of layered basal planes,33 the distance
between planes being 3.35 A. 34 Each basal plane of graphite is composed of fused
hexagons containing spz hybridized carbon atoms with C-C bond lengths of 1.4 A34
Graphite is a semimetal with a workfunction of 4.4 eV.35

As previously mentioned, the structure of the basal plane of graphite is similar to the
microcrystalline component of soot or carbon black. This component is resistant to
oxidation and insoluble in most organic solvents.15 Carbon atoms in the microcrystalline
component of carbonaceous aerosols are sp‘2 hybridized with carbon-carbon bond lengths
similar to that of graphite.22 The basis for using graphite as a model for carbonaceous
aerosols stems from the structural similarities between the two species. As such, the
graphite surface is a starting point for investigating heterogeneous photochemistry that is

relevant to atmospheric reactions on carbonaceous aerosols.

15

     

 

‘
j

Figure 1.6 Graphite crystal structure showing basal plane dimension and bond lengths.33

1.6 References

(1) Brasseur, G. P.; Orlando, J. J .; Tyndall, G. S. Atmospheric Chemistry and Global
Change; Oxford University Press, Inc: New York, 1999.

(2) Grassian, V. H. Int. Rev. Phys. Chem. 2001, 20, 467.

(3) Zondlo, M. A.; Hudson, P. K.; Prenni, A. J.; Tolbert, M. A. Annu. Rev. Phys.
Chem. 2000, 5], 473.

(4) Peter, T. Annu. Rev. Phys. Chem. 1997,48, 785.

(5) Zondlo, M. A.; Barone, S. B.; Tolbert, M. A. J. Phys. Chem. A 1998, 102. 5735.
(6) Ammann, M.; Kalberer, M.; Jost, D. T.; Tobler, L.; Rossler, E.; Piguet, D.;
Gaggeler, H. W.; Baltensperger, U. Nature 1998, 395, 157.

(7) Ravishankara, A. R.; Longfellow, C. A. Phys. Chem. Chem. Phys. 1999, l, 5433.
(8) Grassian, V. H. J. Phys. Chem. A 2002, 106, 860.

(9) Parmon, V. N.; Zakharenko, V. S. Cattech 2001, 5, 96.

(10) Finlayson-Pitts, B. J.; James N. Pitts, J. Chemistry of the Upper and Lower
Atmosphere; Academic Press: New York, 2000.

(1 1) Pandis, S. N.; Wexler, A. S.; Seinfield, J. H. J. Phys. Chem. 1995, 99, 9646.
(12) Andreae, M. O.; Crutzen, P. J. Science 1997, 276, 1052.

(13) Homann, K.-H. Angew. Chem. Int. Ed. 1998, 37, 2434.

(14) Lary, D. J.; Shallcross, D. E.; Toumi, R. J. Geophys. Res. 1999, 104, 15929.
(15) Penner, J. E.; Novakov, T. J. Geophys. Res. 1996, 10], 19373.

(16) Rosen, H. Nature T. Novakov, 266, 708.

(17) Graedel, T. E.; Crutzen, P. J. Atmospheric Change An Earth System Perspective;

W. H. Freeman and Company: New York, 1993.

(18) Atkinson, R.; Arey, J. Acc. Chem. Res. 1998, 31, 574.

(19) Friedli, H. R.; Atlas, E.; Stroud, V. R.; Giovanni, L.; Campos, T.; Radke, L. F.
Global Biogeochemical Cycles 200], 15, 435.

(20) Tuazon, E. C.; Alvardo, A.; Aschmann, S. M.; Atkinson, R.; Arey, F. Environ.
Sci. Technol. 1999, 33, 3586.

(21) Carey, F. A. Organic Chemistry, 2nd ed.; McGraw-Hill, Inc: New York, 1992.
(22) Traven, V. F. Frontier Orbitals and Properties of Organic Molecules; Ellis
Horwood: New York, 1992.

(23) Wiberg, K. B.; Rosenberg, R. E. J. Am. Chem. Soc. 1990, 112, 1509.

(24) Calvert, J. G.; Pitts Jr., J. N. Photochemistry; John Wiley and Sons: New York,
1966.

(25) Fahr, A.; Nayak, A. K. Chem. Phys. 1994, 189, 725.

(26) Fub, W.; Schmid, W. E.; Trushin, S. A. Chem. Phys. Lett. 2001, 342, 91.

(27) Haller, I.; Srinivasan, R. J. Chem. Phys. 1964, 40, 1992.

(28) Bertolini, J. C.; Cassuto, A.; Jugnet, Y.; Massardier, J.; Tardy, B.; Tourillon, G.
Surf. Sci. 1996, 349, 88.

(29) Bhattacharya, A. K.; Pyke, D. R. J. Mol Cat. A 1998, 129, 279.

(30) Chen, J. G. J. Catal. A 1995, 154, 80.

(31) Weiss, M. J.; Hagedom, C. J.; Weinberg, W. H. J. Vac. Sci. Technol. A 2000, 18.
1443.

(32) Osaka, N.; Akita, M.; Itoh, K. J. Phys. Chem. B 1998, 102, 6817.

(33) Borghesi, A.; Guizzetti, G. Handbook of Optical Constants of Solids 2; Academic

Press: New York, 1991.

(34) Reynolds, W. N. Physical Properties of Graphite; Elsevier Publishing Co.: New
York, 1968.
(35) Somorjai, G. A. Introduction to Surface Chemistry and Catalysis; John Wiley and

Sons: New York, 1994.

EXPERIMENTAL

2.1 Survey of Analytical Techniques
X-ray photoelectron spectroscopy (XPS) is frequently used to determine the chemical

composition of a surface. 1 ,2 The technique is based on the photoelectric effect. When a
surface is exposed to photons of sufficient energy, electrons will be ejected from that
surface with a specific kinetic energy. These are commonly called photoelectrons. The
kinetic energy a photoelectron possesses depends on the energy of the incident photon,
the core-level energy-state from which the electron was removed and the workfunction of
the material. The photoelectrons are then energy analyzed and their binding energies
determined using the following expression.

KE = hv — Eb - ¢ (1)
Here hv is the energy of the incident photon (usually KOL lines of Al or Mg at hv = 1486.6

and 1253.6 eV, respectively), Eb is the binding energy of the electron in the sample and

(I) the workfunction of the material.3 This process is outlined in Figure 2.1. Note that for
a solid surface, the photon energy must exceed the sum of the binding energy of the
electron at its core-energy level and the workfunction of the surface. The chemical
composition of a surface is determined from the fact that each element displays a unique

set of binding energies.

20

0 Free Photoelectron

 

 

 

 

 

A
Evacuum
Workfunction ((1))
A EFermi
hv

Binding energy (Eb)
A v
t

 

 

Core energy levels

Figure 2.1 Schematic diagram of X-ray photoelectron generation.

X-ray photoelectron spectroscopy has high surface sensitivity, probing no more then

100 A into the surface.4 This is due to the short inelastic mean free path a photoelectron
will travel in a solid material. Only photoelectrons that have been generated within the
local surface region can escape the surface to be energy analyzed. Therefore, XPS gives
an accurate description of the near-surface composition.

Temperature programmed desorption (TPD) is commonly employed to gain

information about the energetics and coverage of molecules adsorbed on a solid surface.5
It can be envisaged that a molecule adsorbed at a surface lies in a potential well of depth
Ed (Ed = desorption energy). The temperature of the sample is then raised linearly as a
function of time and at some temperature the adsorbed molecules will obtain enough
energy to break the adsorbate-surface bond and desorb from the surface. A quadrupole
mass spectrometer is used to measure the rate that the molecules desorb from the surface.
The rate is expressed as an Arrhenius type expression called the Polanyi-Wigner equation
(2).

— dn / dt = n"vx exp(-Ed/RT) (2)
In this expression x represents the “order” of desorption, n is the number of molecules, v
is a frequency factor and Ed is the desorption energy. This expression can then be fit to
the experimental data to give information about the magnitude of the adsorbate-surface
bond (Ed). If Ed is less than 40 kJ/mol the adsorbate is considered to be physisorbed on
the surface; in other words no true chemical bond has formed between the adsorbate the

surface. Desorption energies greater than 40 kJ/mol suggest the adsorbate forms a true

chemical bond with the surface, and this process is termed chemisorption.6

IO
10

The peak shape of TPD spectra can be used to describe both coverage and
desorption mechanisms.7 Because the chemical environment of monolayer and
multilayer adsorbates is usually different, often two separate desorption peaks will appear
in a TPD spectrum at coverages greater than 1 monolayer. The first monolayer typically
desorbs with first order kinetics, implying that the desorption rate depends linearly on
coverage. In this case the temperature at the maximum rate of desorption remains
constant with respect to coverage, and the peak is typically asymmetric in shape. If
interaction between the first monolayer and subsequent monolayers is weaker than the
first monolayer-substrate interaction then multilayer adsorbates will desorb at a lower
temperature than the monolayer. It is common for multilayer desorption to follow zero
order desorption kinetics. Here, the desorption rate doesn’t depend on coverage but
increases exponentially with temperature. Second and fraction order desorption
mechanisms are known but are rare and will not enter into this study of butadiene on
HOPG(0001).

High-resolution electron energy loss spectroscopy (HREELS) is used to probe

molecular vibrations of adsorbates on a surface. I ’8 High sensitivity (able to detect 1% of
a monolayer) and broad applicability have made HREELS an invaluable surface sensitive
analytical technique. The technique works by scattering a monoenergetic beam of
electrons (1-15 eV) from a surface at a specific angle. The scattered electrons are then
energy analyzed. Most of the electrons will scatter elastically, but some will lose energy

and scatter inelastically. These inelastically scattered electrons undergo energy losses

corresponding to the energy of molecular vibrations on the surface.9 There are three

scattering mechanisms: dipole scattering, impact scattering, and negative ion resonance

23

scattering. Dipole scattering is the most common and is the mechanism from which the
data will be obtained in this study.

Dipole scattering results when an electron approaching the surface is influenced

by an oscillating dipole.9 Only dipoles perpendicular to the surface affect the electron,
since a dipole that is perpendicular to the surface is enhanced by its image dipole within
the surface. A dipole parallel to the surface is cancelled by its image in the surface. This
provides the “surface selection rule” that states that only vibrations perpendicular to the
surface will been seen in an HREELS spectrum (see Figure 2.2). Vibrations parallel to
the surface will not be seen by dipole scattering. Using this information, one can

qualitatively determine the orientation of an adsorbate on a surface.

® .
0 13113016 G) 9 No net dipole

/////// ///////

069

 

 

Image dipole

Figure 2.2 Representation of a dipole perpendicular to a surface and its image dipole in
the surface (left) and a dipole parallel to a surface plane with its mirror image in the
surface (right).

2.2 Experimental Apparatus

All experiments were carried out in two connected stainless steel ultrahigh vacuum
(UHV) chambers, each one containing a 270 US ion pump to maintain ultrahigh vacuum.
The first chamber also contained a molecular leak valve, a 0-200 amu quadrupole mass
spectrometer, and an X-ray source and hemispherical electron energy analyzer for X-ray
photoelectron spectroscopy (XPS). The second chamber contained a high-resolution
electron energy loss spectrometer. The base pressure of the chamber was 1.6x 10’'0 torr.

Highly ordered pyrolytic graphite (HOPG) (grade SPI-2, SPI supplies) with
approximate dimensions of 10x10x1 mm was fastened to a molybdenum sample mount
using molybdenum clips. Before introduction to the UHV chamber, the HOPG sample
was peeled using adhesive tape. An E-type thermocouple was placed between the HOPG
sample front face and a molybdenum clip. The sample could be cooled to 77 K using
liquid nitrogen or heated to ~800 K via a tungsten filament embedded in the mount. The
sample holder was also capable of x, y, and z translation and 0 rotation. Following
bakeout, the sample was annealed to 700 K for several hours in order to remove any
contaminates present on the surface.

After annealing the sample, the surface composition was monitored using XPS. All
XPS data were collected using the Al Koz X-ray line (hv=1486.6 eV) operated at 300 W
(15 kV and 20 mA). The pass energy of the spectrometer was set at 100 eV, and XPS
data were measured collected at a take-off angle of 45° from the surface normal in order
to increase surface sensitivity. The surface of the HOPG sample was found to be free of
all species other than carbon except for ~1.0% of an oxygen species that could not be

removed. A survey scan of the annealed sample surface is shown in Figure 2.3. The

25

large peak in the spectrum at 286 eV is due to photoelectron emission from the carbon ls
orbital. The only other identifiable feature in the spectrum is due to the emission from
the oxygen ls orbital centered at 533 eV.

An excimer laser (Questek 2440) operating at 193 nm (ArF), 248 nm (KrF) or 351 nm
(XeF) was used to provide UV photons. The laser was operated at an average power of
0.5 W at all wavelengths and a repetition rate of 5 Hz. The laser pulse duration
(FWHM), quoted by the laser manufacturer, was 8-16 ns. The laser light was passed
through a quartz window on the UHV chamber and impinged on the sample at 45° from
surface normal, (see Figure 2.4). All irradiation experiments were performed at < 85 K.

1,3-Butadiene gas (Aldrich 99%) was used without further preparation. The gas was
introduced to the UHV chamber via a stainless steel manifold coupled to a molecular leak
valve. The gas was closed onto the sample through a 1/8” inside diameter stainless steel
tube (the “directed closer”). In order to expose the sample to butadiene consistently, the
leak valve was opened and mass 39 (C3H3+), the most abundant ion in the gas phase
cracking pattern of butadiene, was measured with the mass spectrometer. Once a
predetermined constant signal was obtained, the sample was placed in front of the
directed doser (~5 mm from the sample surface) for a specified amount of time.

Temperature-programmed desorption (TPD) experiments were performed by
positioning the surface in light-of—sight of the 2 mm entrance aperture of a shrouded mass
spectrometer. The electron impact ionization source was operated at 70 eV and a bias of
> -70 V was placed on the sample in order to repel any electrons from the ion source of

the mass spectrometer. During TPD experiments, the sample was linearly heated at a rate

26

 

_ C15

018

Arbitrary Intensity

 

 

 

 

. I . I . I ' I '
1000 800 600 400 200 0

Binding Energy (eV)

Figure 2.3 X-ray photoelectron spectrum of the HOPG surface before experiments.

27

Sample UHV Chamber

’Quartz window

Pressure
~ 1x10‘10torr N/firror

 

 

Excimer laser

7» = 193,
248, 351

 

 

 

Figure 2.4 Experimental apparatus showing Excimer laser and UHV chamber
orientation.

of 5 K/s up to a maximum temperature of 450 K. A fresh layer of butadiene was
prepared for each experiment.

The adsorption and photochemistry of butadiene was also characterized using EELS.
The scattering angle in our spectrometer was set to 0, = 55°, and all data presented in this

study was collected 1° off-specular (0S 2 56°) to slightly reduce the intense inelastic

scattering background from the semi-metallic graphite surface.10 The intensity of this
inelastic background decreases quickly as the scattering geometry moves away from the
specular direction. The primary beam energy was 6.1 eV and resolution from the
C4H6/HOPG surface was between 36 and 44 cm'1 as determined from the elastic peak

FW HM for all data shown here.

29

2.3 References

(1) Perry, S. S.; Somorjai, G. A. Analytical Chemistry 1994, 66, 403 A.

(2) Woodruff, D. P.; Delchar, T. A. Modern Techniques ofSurface Science, 2nd ed.;
Cambridge University Press: Cambridge, 1994.

(3) Handbook of X -ray Photoelectron Spectroscopy; Wagner, C. D.; Riggs, W. M.;
Davis, L. E.; Moulder, J. F., Eds.; Perkin-Elmer Corporation: Eden Prairie, 1979.

(4) Briggs, D.; Seah, M. P. Practical Surface Analysis, 2nd ed.; Wiley: New York,
1990; Vol. l-Auger and X-ray Photoelectron Spectroscopy.

(5) King, D. A. Surf. Sci. 1975, 47, 384.

(6) Somorjai, G. A. Introduction to Surface Chemistry and Catalysis; John Wiley and
Sons: New York, 1994.

(7) long, A. M.; Niemantsverdriet, J. M. Surf. Sci. 1990, 223, 355.

(8) Sexton, B. A. Appl. Phys. A 1981, 26, l.

(9) lbach, H.; Mills, D. L. Electron Energy Loss Spectroscopy and Surface
Vibrations; Academic Press: London, 1982.

(10) Palmer, R. E.; Annett, J. F.; Willis, R. F. Surf. Sci. 1987, 189/190, 1009.

30

ADSORPT ION AND PHOTOCHEMISTRY OF 1,3-BUTADIENE ON

HOPG

3.1 Adsorption of Butadiene on HOPG determined by TPD

The adsorption of butadiene on HOPG(0001) was studied using temperature
programmed desorption. Figure 3.1 shows TPD (m = 39, C3H3+) profiles for various
exposures of butadiene adsorbed on HOPG at < 85 K. The first monolayer of butadiene
on HOPG displays apparent first-order desorption kinetics and desorbs with a maximum
rate at about 129 K. Using Redhead analysis,1 the desorption energy was determined to
be 32 kJ/mol, suggesting that butadiene is physisorbed on the HOPG surface. With
increased butadiene exposure, a second feature appears, initially located with a maximum
rate of desorption at 105 K. This feature is ascribed to second and multilayer desorption
since it does not saturate at large exposures. Multilayer desorption shows zero-order
desorption kinetics with an activation energy determined by leading edge analysis2 to be
28.5 kJ/mol.

Interestingly, the monolayer peak does not saturate as the multilayer peak begins to
grow in. When the monolayer peak becomes saturated, approximately 16% of the total
number of adsorbed molecules are associated with the multilayer peak. This indicates
that butadiene is probably forming 3-D islands on the HOPG (0001) surface before the
first monolayer is complete. In this work one monolayer (ML) is defined as the
maximum coverage of butadiene on HOPG where no multilayer desorption is seen by

TPD. This coverage represents approximately 83% of the saturated monolayer coverage.

31

 

 

 

 

 

2.0 ML

1.5 ML

 

1.0 ML
0.75 ML

 

 

0.5 ML__

 

0.25 ML

Ion Intensity (m = 39)

 

 

 

 

l

90 I105 150' 1:15;"‘150 165
Temperature (K)

ITT—r Tfirij—U

Figure 3.1 Mass (C3H3+) thermal desorption spectra of C4H6 on HOPG with increasing
coverage. At 2.0 ML coverage the lst monolayer is saturated.

32

To confirm butadiene desorbs molecularly, further TPD experiments were
performed. Figure 3.2 shows TPD spectra for several butadiene fragment masses (m = 39
C3H3+, m = 27 C2H3+, and m = 54 C4H6+) at a constant coverage of 1.0 ML. The area
under each desorption peak was calculated and compared to the relative intensity for each
mass in the gas phase cracking pattern of butadiene (see appendix). In the gas phase
mass spectrum of butadiene, mass 39 is the most abundant ion and the relative intensities
of mass 27 and 54 are 80% and 55% respectively. The TPD data corresponds closely
with the gas phase cracking pattern with mass 27 and 54 at 75% and 51%, respectively,
of the mass 39 area. The similarity of the gas phase and desorbed molecule

fragmentation patterns indicates that no measurable irreversible adsorption occurred.

3.2 Adsorption of Butadiene on HOPG determined by EELS

Electron energy loss spectroscopy (EELS) was also used to study the adsorption of
butadiene on HOPG. Figure 3.3a shows an EELS spectrum for 1 ML of butadiene
adsorbed on HOPG at < 85 K. The spectrum shows several distinct loss features and
assignment of these peaks by comparison with IR literature data 3 suggests that the planar
butadiene molecule adsorbs, with its mirror plane containing all atoms, parallel to the
HOPG surface plane without significant C atom rehybrization. Based upon both the IR
frequency assignments between the cis and trans isomers,4 and knowing that at room
temperature less than 1% of gaseous butadiene molecules are in the cis conformation,5
the adsorption isomer of butadiene on HOPG is proposed to be the trans conformation.

Frequency shifts in IR spectra associated with the trans and cis isomers will be discussed

33

 

 

Ion Intensity

 

Mass 39

 

Massfi27

Mass 54

 

 

I T T

 

I T T T T

.,.
90 105

T T T T l T T T

120 135 160 I ' 165
Temperature (K)

Figure 3.2 Thermal desorption spectra for masses 39 (C3H3+), 27 (C2H3+) and 54 (C4Hr,+)
of C4H6 on HOPG for S 3 ML coverage to ensure the lst monolayer is saturated.

34

further below. The features visible in the EEL spectra of 1,3-butadiene(ad)/HOPG are
the p, (torsion between C: and C3) at ~186 cm], the pt (=CH2 twist) at 532 cm", the 0)
(CH2 wag) at 909cm", the 0) (CH bend) at 1014 cm", and the v (CH stretch) at 3025 cm’
'. All of the observed losses in the EEL spectrum belong to out-of—plane vibrations
except for the v (CH) mode at ~3025 cm". The appearance of the v (CH) mode is
believed to be due to the low frequency torsion mode at ~186 cm'l that lowers the
symmetry of the molecule away from Cgh, moves the C-H motion away from parallel to
the surface plane and allows it to be seen by EELS. The butadiene adsorption geometry
is also supported by the absence of several in-plane IR active molecular vibrations (for
example, the v (C=C stretch)at 1596 cm'1 and the 5 (=CH2 in plane bend) at 1381 cm")
that would be screened in the parallel geometry. The p, (=CH2 twist) at 532 cm'1
indicates that butadiene adsorbs onto HOPG in the trans conformation. In the cis
conformation the pl (=CH2 twist) shifts to about 730 cm".4 Similar conclusions about the
adsorption geometry and conformation were reached in studies of butadiene on Au(l l 1).6
There is no evidence for the formation of cycle-type adsorbates on graphite as have been
proposed on some metal surfaces where interaction between the butadiene and the surface
leads to significant double bond activation.7

Electron energy loss spectroscopy was also used to investigate multilayer (3 ML)
adsorption of butadiene on HOPG (see Figure 3.3b). The multilayer data is very
similar to the monolayer data; all of the same peaks appear at approximately the same
frequency, but with an increased signal to noise ratio. No new peaks appear in the
multilayer spectrum and the ratio of intensities between the various features is very

similar to that of the monolayer. I believe that these data indicate that butadiene forms

35

 

Normalized Intensity

 

* l (b)

M (X 8)
““ ‘51:“;m

q
'_ . L A A
j ' v ""‘T—
T T

l '

l I
0 1000 2000

 

 

 

 

 

1
3000

Energy Loss (cm'1)

Figure 3.3 Electron energy loss spectroscopy of (a) 1 ML and (b) 3 ML C4H6 on
HOPG.

36

 

 

Normal vibrational Gas phase EELS 1 ML EELS 3 ML

modes frequencies (cm'l) C4H6/HOPG C4H6/HOPG
(cm'l) (cm'l)

C-C torsion (an) 162.3 186 186

CCC deform (ag, bu) 5123, 301

CH2 twist (au, bg) 522, 770“ 532 550

CH; wag (an, bg) 908, 9123 909 919,

CH bend (an, bg) 1013, 976a 1014 1021

CC str (ag) 1 196a

CH bend (ag, bu) 1280a, 1294

CH2 scis (ag, bu) 1438a, 1294

C=C str (ag, bu) 16303, 1596

CH2 s-str (ag, b,) 2992*, 2984 3025b 3043b

CH str (3,, bu) 3003a, 3055 3025b 3043b

CH3 a-str (ag, bu) 3087a, 3101 3025b 3043b

 

‘1 IR inactive mode

b Not able to deconvolute based on the resolution of EELS

Table 3.1 Vibrational assignments of 1,3-butadiene adsorbed on HOPG(0001).

ordered multilayers on top of the first monolayer, with a similar parallel adsorption
geometry to that of the monolayer. The peak maxima for all EELS data are listed in

Table 3.1 along with IR values for C4H6(g).

3.3 Photodesorption of Butadiene monitored by TPD

Figures 3.4 and 3.5 show TPD results (mass 39) for 1 ML (Figure 3.4) and 3 ML
(Figure 3.5) of butadiene adsorbed onto HOPG at < 85 K following irradiation with 193
nm photons at incident energies of 100 mJ/pulse. Irradiating the adlayer with 3x10l7
photons/cm2 resulted in a decrease in molecularly adsorbed butadiene as indicated by the
loss of intensity for mass 39 in the TPD data shown. Continued irradiation further
decreases the butadiene coverage as clearly seen in Figure 3.4, where after irradiating the
butadiene monolayer with 5x1019 photons/cmz, only 18 % of the initial monolayer
remained on the surface. The temperature of the maximum rate of desorption and peak
shape remained constant for all TPD data collected. Furthermore, mass spectra from 0-
100 amu collected during TPD showed no desorbing fragments other than those
belonging to butadiene.

Figure 3.5 shows the same data as for Figure 3.4, but with 3 ML coverage of
butadiene on HOPG. Irradiation of the adsorbate layers produces a decrease in the mass
39 signal for both monolayer and multilayer peaks. Measurements at other masses
confirmed that, as for the monolayer, the only significant specie remaining on the surface
following irradiation of the multilayer was butadiene. There was no evidence for reaction

products remaining adsorbed, implying that butadiene desorbs molecularly (without

38

 

1 ML C4H6/HOPG

 

= 39)

 

# photons/cm:z

 

Ion Intensity (m

 

 

 

 

 

5.0 x1019

 

 

 

IfiTT

I I T I

90 "165'”‘1éo' 135"'1éd"'1es
Temperature (K)

Figure 3.4 Thermal desorption spectra (mass 39, C3H3+) of 1 ML C4H6 on HOPG
irradiated with increasing photon number (it = 193 nm).

39

 

3 ML C4H6/HOPG

= 39)

# photons/cm2

 

 

 

Ion Intensity (m

 

 

 

 

 

I'ufifi'II'H'I """ I """ T ''''' 1"
90 105 120 135 150 165
Temperature (K)

Figure 3.5 Thermal desorption spectra (mass 39 C3H3+) of 3 ML C4H6 on HOPG
irradiated with increasing photon number (1» = 193 nm).

40

fragmentation) from the surface. The multilayer peak intensity decreases at a faster rate
than the monolayer peak intensity, as will be discussed further below.

In order to quantify the photodesorption process, the area under each desorption peak
was calculated. Assuming that the process follows first order kinetics, a plot of 1n(C4H6
m/C4H6 (0,) vs. total incident photon number (nph photons/cmz), where C4H6 (t, is the area
under a TPD peak after a specific total incident photon number and C4H6 (0) is the area
under the TPD peak for an unirradiated adlayer, should give a straight line. The slope of
the line is equal to the cross section 0' (ch/photon) for photodesorption. Figure 3.6
shows a semi-logarithmic plot of TPD area vs. total incident photon number for the data
shown in Figure 3.4. The total cross section for photodesorption was calculated to be
3.3x10'20 cm2 at 193 nm and 100 mJ/pulse. The cross section for 3 ML coverage was
calculated to be 1x10'l7 cmz, about three orders of magnitude greater than the 1 ML cross
section. The experiment was repeated at 248 nm and 351 nm for both 1 ML and 3 ML
coverages. Table 3.2 lists all of the cross sections calculated here along with the gas

phase cross sections of absorption at the various wavelengths.

 

 

MWavelength (nm) 1 1 ML cross section 3 ML cross section - ‘Gas-phasemw
(cmz) (cmz) cross section (cmz)ll
248 2.4x10'20 6.3x10"8 2x1019
351 1.211103" 1.8x10'18 ~0

 

 

~'.w¢'v.-o Q'uldu M9 sv---.-.-.u.n-. .-.-.-.- .wwu...- --ru~.- «an-u sat-vs.- . -~--.~.~ '«rv-m . .-.~ - o . .u- ‘ . .-. r» m.

Table 3.2. Butadiene photodesorption cross sections and gas phase cross sections from
the literature.

41

 

0.0 4 1 ML C4H6/HOPG

  
   

-0.3 -

-°-6- e: 3.25x10'20 em2

”\o

it“ -0.9 4

<

v

C

_ -1.2 -
-1.5 -
-1.8 -

 

 

 

I ' l ' I ' I I I I j

0 1 2 3 4 5
Number of Photons (x1019)

Figure 3.6 Semi-logarithmic plot of TPD peak area vs. total number of photons/cm2 for
calculation of the photodesorption cross section.

3.4 Photodesorption of Butadiene monitored by EELS

Electron energy loss spectroscopy was also used to examine the 193 nm
photochemistry of 1 ML butadiene on HOPG. For all total incident photon numbers up
to 5x1019 photons/cmz, identical features to those observed for the unirradiated
monolayer were observed. The general loss in peak intensities and the decrease in signal-
to-noise ratio with prolonged irradiation, coupled with the pronounced inelastic
background of graphite made it difficult to determine the energy loss peak positions with
confidence. No new peaks appeared in the EELS data with prolonged irradiation,
suggesting that no adsorbed state reaction products were created in measurable quantities,
in accord with the TPD data discussed above.

The increased signal-to-noise ratio of the EELS data for multilayer butadiene
exposures allowed us to more reliably investigate changes in the EEL spectra with
irradiation. Figure 3.7 shows the vibrational spectrum obtained by irradiating 3 ML of
butadiene adsorbed on HOPG with 1x10l7 photons/cm2 at 193 nm and 100 mJ/pulse.
Clearly, the presence of loss features indicates that a substantial amount of butadiene
remained adsorbed on the surface (TPD data suggested that approximately 2 ML are
present on the surface following this photon exposure). Once again, no new features

were observed in the EELS data due to the accumulation of adsorbed products.

43

 

.. 3 ML C4H6/HOPG
~1x10'7 photon/cm2

 

 

 

 

 

 

A

3

O T

v

>5

3:

CD _.

C J

(D

q—I

E —1
d l " (X8)
_ I I I I I I I

0 1000 2000 3000

Energy loss (cm’1)

Figure 3.7 Electron energy loss spectra of 3 ML C4H6 on HOPG exposed to 1x10'7
photons/cmz.

44

DISCUSSION

3.5 Mechanisms for Photodesorption

Over the past 15 years, several photodesorption mechanisms have been proposed in
the literature. Each will be briefly discussed in relation to the experimental data
presented here. The three mechanisms considered are: (1) an adsorbate-mediated
process, commonly called photodesorption or photoejection, involving the direct
absorption of a photon by an adsorbate molecule or adsorbate-substrate complex; (2) a
substrate-mediated process involving the absorption of a photon by the surface and
electronic energy transfer to the adsorbate-substrate bond; and (3) laser-induced thermal
desorption (LITD) which occurs when energy absorbed by the surface is quickly

converted to surface phonons, and the local surface temperature rise causes desorption.

For a more in-depth discussion of these processes, the reader is referred elsewhereg‘10

3.6 Adsorbate-mediated Processes

Gas-phase butadiene absorbs UV radiation from about 170 nm to 225 nm.11 The
wavelength of maximum absorption is 210 nm where the cross section is 1x10'l6 cmz.12
In this wavelength range, absorption results in a rt——)It* transition to two closely spaced
states (the 1Bu and 2Ag states). These states have lifetimes of < 100 fs, and decay from
these states eventually leads to single—bond isomerization in about 270 fs.13

In general, adsorbate molecule-mediated surface photochemical processes exhibit a
wavelength dependence that is qualitatively similar to absorption by the analogous gas
phase molecule. As such, if the photodesorption of butadiene on graphite is initiated by

direct adsorbate absorption, we might expect it to be a maximum near 193 nm, to

45

decrease significantly at 248 and be negligible at 351 nm. The gas phase absorption cross
sections for butadiene are ~4x10'l7 cmz, < 2x10'l9 cm2 and essentially zero at 193 nm,
248 nm and 351 nm, respectively.12 If we compare these values with the cross sections
for photodesorption for 1 ML of butadiene on HOPG, 3.3x 10'20 cm2, 2.4x 10'20 cm2 and
1.2x10'20 cm2 at 193 nm, 248 nm and 351 nm, respectively, we find a poor correlation
between the photodesorption and gas-phase cross sections. It is not uncommon for the
photochemical response of adsorbed molecules to be red shifted by up to 1 eV from the
gas-phase spectrum,8 but the absorption spectrum remains qualitatively similar.
Although the cross sections for photodesorption from 3 ML of C4H6(ad)/HOPG are
about two orders of magnitude higher than those for 1 ML of C4H6(ad)/HOPG, the
variation with wavelength is very similar to the 1 ML case. For both mono and
multilayer coverages, it appears that photodesorption is not dominated by adsorbate

absorption.

3.7 Substrate-mediated Processes

Photochemical processes driven by initial photon absorption by the surface have been
commonly reported for metal and semiconductor substrates. They originate when
electrons generated in the near-surface region migrate and attach to molecular orbitals
within the adsorbed molecule or adsorbate-substrates complex. Two cases are often
described. In the first, the incident photon energy exceeds the workfunction of the
material and true photoelectrons are produced. In the second process, the incident photon

energy is less than the workfunction of the material and so-called ‘hot’ or sub-vacuum

46

level electrons are generated. These can tunnel into the adsorbate depending upon the
energetic and spatial arrangement of the surface and adsorbate energy levels.

We discount the production of free photoelectrons as being responsible for the
desorption of butadiene based upon the workfunction of graphite of 4.4 eV. ‘4 Assuming
that the workfunction does not change considerably upon adsorption (sensible for a
physisorbed system), photoelectrons cannot be generated for incident wavelengths greater
than about 280 nm. However, we observe appreciable photodesorption at 351 nm in our
experiments, and so discount photoelectron effects as dominating the photodesorption of
butadiene from HOPG.

Subvacuum-level or ‘hot’ electrons are undoubtedly created at all wavelengths used
in our work. However, we do not believe that they are responsible for the observed
photodesorption of butadiene. Other studies have indicated that the mean free path of
these hot electrons is limited to only a few layers of an adsorbate. Regardless, in the
absence of additional quenching routes operative in the first adsorbed layer, a hot electron
driven desorption process should produce a much greater rate of desorption for the
adsorbate layer in contact with the surface than for subsequent multilayers. Since we
observe a multilayer photodesorption rate that is some two orders of magnitude greater
than the monolayer photodesorption rate at all wavelengths, we conclude that hot

electrons are not contributing significantly to butadiene desorption.

3.8 Laser-induced Thermal Desorption

Laser-induced thermal desorption (LITD) is a well-known consequence of exposing

an adsorbate-covered surface to a large photon flux via pulsed laser irradiation. Excited

47

electrons generated within the optical penetration depth of the surface decay rapidly (10'
'2 s) by creation of phonons (thermal energy). The thermal energy diffuses away from
the irradiated region both during and following cessation of the laser pulse. Depending
upon the thermal diffusivity properties of the solid, it is possible to locally heat the
surface at rates corresponding to more than 10'1 K/s.15

The temperature rise of a surface can be estimated using equation (1) when the optical
and physical constants of the surface are known.16

AT(t)=(F/K)(rrt/rr)”2 (1)

Here, F is the energy absorbed by the surface, K is the thermal conductivity of graphite

(0.176 W/ch)”, K is the thermal diffusivity (0.56 curls) and t is the FWHM of the laser
pulse (simplified to a triangular temporal profile). The energy absorbed, F, can be
calculated from the optical penetration depth and the reflective properties of the surface.
The optical penetration depth is given by 1/0t, where 0t is the absorption coefficient for
graphite. The absorption coefficients, calculated from the imaginary part of the refractive
index of graphite, were 1.7x105 cm'1 at 193 nm, 3.6x105 cm'l at 248 nm and 2.5x105 cm'l
at 351 nm.18 The Fresnel equations were used to determine the amount of energy
absorbed and the amount of energy reflected from the surface.19

The results of the calculation give an estimate of the change in surface temperature
versus time and are shown in Figure 3.8 for a 193 nm, 16 ns FWHM laser pulse of 100
mJ/cmz. The calculations for 248 nm and 351 nm light (not shown) were within 50 K of
the data shown, largely because of the relatively uniform absorption coefficient of

graphite over this wavelength range. At all three wavelengths, the surface temperature

48

rises well above the desorption temperature of butadiene, up to about 380 K in 16 ns from
an initial sample temperature of 80 K. It should be noted that this isotropic model does
not take into account the highly anisotropic thermal diffusivity (K) of graphite. The value
used (0.56 cm/s) is the quoted thermal diffusivity between the graphite sheets. Thermal
diffusivity between the sheets is much slower and so our values for the transient
temperature rises represent low estimates. Similarly, the use of the longest laser pulse
FWHM in our calculations (16 ns) produces a low estimate of the maximum surface
temperature rise.

To further investigate this process, the calculation was repeated for an incident power
of 10 mJ/pulse (i.e. 10% of the original intensity). Under these conditions, it was found
that the surface temperature increase is only about 30 K, or to about 110—1 15 K. This
temperature increase corresponds to less than the temperature for onset of monolayer
desorption for butadiene (about 120 K in TPD). In Figure 3.9 the amount of 1 ML
desorbed from the surface was measured as pulse power was changed. The total number
of photons striking the surface was constant for all the data shown and was ~3x1019
photons/cmz. At low pulse power, ~0.064 W, the desorption was minimal and only about
5% of the monolayer was removed. At this power the maximum local surface
temperature was calculated to be 121 K, close to the onset for desorption. At pulse
powers above 0.1 W the amount of monolayer left on the surface greatly decreases. At
these higher powers the surface temperature rise becomes significant. As a result, the
desorption behavior of 1,3-butadiene adsorbed on graphite can be successfully described
by a simple laser-induced thermal desorption mechanism. No other photochemistry was

found to take place in the adsorbed monolayer or multilayer. This is surprising since

49

350

 

 

 

 

3004
S2,
9250—
3
4...!
<0
213
0.200—
E
a)
'—
<150-
100-
5OIuIIIIIIIIIIIII
0 2 4 6 810121416
Time(ns)

Figure 3.8 Result for calculating the surface temperature rise using equation 1 when
HOPG is exposed to a 16 us 100 m1 pulse of 193 nm light.

50

Surface Temperature (K)
82 113 144 175 206 237 268

 

70 I l f T I I I
'0 a
(D 60
.Q
L—
8
(D 50-
D
co
I 40.
v
0
_l
E 30—
‘—
L.—
O
4—1 20-
C
§
0) 10—
D.
040

 

 

 

I I I I I I I I I T j

I I
0.00 0.05 0.10 0.15 0.20 0.25 0.30

Pulse Power (W)

Figure 3.9 Percent of C4H6 monolayer desorbed from the surface with increasing pulse
power. The total number of photons was kept constant at each pulse power by varying
the number of pulses.

51

butadiene has been observed to polymerize from the gas-phase onto solid surfaces under
UV light.20 We speculate that the very short lifetime of the photoexcited state of
adsorbed 1,3-butadiene (<100 fs for C4H6(g)),I3 prevents bimolecular reaction with

neighboring species through diffusion in the monolayer.

3.9 Conclusions

The adsorption and photochemistry of butadiene (C4H6) on HOPG(0001) has been
investigated. At < 85 K butadiene adsorbs on HOPG, probably forming 3-D islands on
the surface as indicated by the appearance of multilayer TPD features before saturation of
the first adsorbed layer. Temperature programmed desorption suggests that butadiene is
physisorbed on the HOPG surface with a monolayer desorption energy of 32 kJ/mol.
Electron energy loss spectroscopy suggests that the mirror axis of the molecule
containing all atoms is parallel to the surface plane when adsorbed on HOPG, and that
multilayers adsorb with the same geometry as the first layer.

Irradiation of both 1 and 3 ML coverages of C4H6/HOPG(0001) with high UV photon
fluences (up to 100 mJ/pulse) over a range of wavelengths induces molecular
photodesorption. Three possible desorption mechanisms were considered for the
desorption process; (1) an adsorbate-mediated mechanism, (2) a substrate-mediated
mechanism, and (3) a thermal mechanism. A simple thermal mechanism proved to offer
the most reasonable explanation of the observed behavior, and calculations confirmed
that at 100 mJ/pulse, the local surface temperature will rise to well above the desorption

temperatures for butadiene. At lower energies (<10 mJ/pulse), no photodesorption or

52

other photochemistry was measured by TPD. Calculations suggest that the local surface
temperature rise at these low energies is too small to thermally desorb butadiene from the

monolayer.

53

3.10 References

(1)
(2)
(3)
1972.
(4)
3998.
(5)
(6)
(7)
1443.
(8)
(9)
(10)
(11)
1966.
(12)
(13)

(14)

Redhead, P. A. Vacuum 1962, 12, 203.
Jong, A. M.; Niemantsverdriet, J. M. Surf. Sci. 1990, 223, 355.

Shimanouchi, T. Tables of Molecular Vibrational Frequencies; NSRDS-NBS:

Mare, G. R. D.; Panchenko, Y. N.; Auwera, J. V. J. Phys. Chem. A 1997, 101,

Wiberg, K. B.; Rosenberg, R. E. J. Am. Chem. Soc. 1990, 112, 1509.
Osaka, N.; Akita, M.; Itoh, K. J. Phys. Chem. B 1998, 102, 6817.

Weiss, M. J.; Hagedom, C. J.; Weinberg, W. H. J. Vac. Sci. Technol. A 2000, 18,

Zhou, X. L.; Zhu, X. Y.; White, J. M. Surf. Sci. Rep. 1991, 13, 73-220.
Ho, W. Surf Sci. 1994, 299/300, 996.
Ho, W. J. Phys. Chem. 1996, 100, 13050.

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

Fahr, A.; Nayak, A. K. Chem. Phys. 1994, I89, 725.
Fub, W.; Schmid, W. E.; Trushin, S. A. Chem. Phys. Lett. 2001, 342, 91.

Somorjai, G. A. Introduction to Surface Chemistry and Catalysis; John Wiley and

Sons: New York, 1994.

(15)

(16)

Brand, J. L.; George, S. M. Surf. Sci. 1986, 167, 341.

Burgess, J.; Stair, P. C.; Weitz, E. J. Vac. Sci. Techno]. A 1986, 4, 1362.

54

(l7) Lide, D. R., Ed CRC Handbook of Chemistry and Physics, 75th ed.; CRC Press:
Boca Raton, FL, 1995.

(18) Borghesi, A.; Guizzetti, G. Handbook of Optical Constants of Solids 2; Academic
Press: 1991.

(19) Ingle, J. D. J.; Crouch, S. R. Spectrochernical Analysis; Prentice-Hall: Upper
Saddle River, New Jersey 1988.

(20) Wright, A. N. Nature 1967, 215, 953.

55

CONCLUSIONS AND FUTURE WORK

4.1 Adsorption and Photochemistry of Butadiene on HOPG

The work presented in chapter 3 shows that butadiene adsorbs on HOPG at 85 K to
form either mono or multilayer adsorption regimes depending on exposure. Temperature
programmed desorption spectra indicate that butadiene desorbs molecularly from the
HOPG surface at 105 K and 129 K for multi and monolayer coverages, respectively. The
molecule adsorbs on HOPG with its mirror axis parallel to the plane of the surface in the
trans conformation for all coverages as shown by EELS. Exposing the butadiene/HOPG
surface to UV photons (A = 193, 248, and 351nm) at high fluence (100 mJ/pulse) induces
photodesorption of butadiene for both 1 and 3 monolayer coverages. Several well-known
photodesorption mechanisms were considered including an adsorbate-induced process, a
substrate-induced process and a laser-induced thermal desorption process. Based upon
the wavelengths used, calculations for the surface temperature rise when exposed to high
photon flux over a short period in time, and the reduced desorption of butadiene at lower
pulse power (10 mJ/pulse) indicates that laser-induced thermal desorption is initiating the
photodesorption of butadiene from HOPG.

An interesting finding of this work is that at low photon fluences, (~10 mJ/pulse)
where l» = 193, 248, and 351 nm, no condensed-phase photochemistry of butadiene on
HOPG takes place. Photochemistry of butadiene in either the gas or liquid phase leads to

the formation of a variety of products (see chapter 1). The lack of condensed phase

56

chemistry on HOPG may result from the ultra—short excited state life-time of butadiene,
on the order of 100 fs.1 Also, surfaces are known to shorten excited state life-times of
adsorbates through a process called quenching.2 Quenching processes are fast and have
been know to compete with photo-driven chemistry of molecules adsorbed at metal

interfaces.2

4.2 Future work

The fact that butadiene undergoes no photon-initiated chemistry in the condensed
phase at low photon fluence leads to the possibility of investigating photon-initiated
bimolecular reactions between butadiene and other photochemically active
atmospherically relevant molecules, such as CH3Br and N02, on the HOPG surface. To
date, it is unclear what role the surface will play in reactions between butadiene and the
photolysis products of CH3Br and N02 in the condensed phase. The analytical
techniques discussed previously including X-ray photoelectron spectroscopy, temperature
programmed desorption and electron energy loss spectroscopy can be used to identify
adsorbed state photochemical products, however, the adsorption structures and
photochemistry of CH3Br and N02, on HOPG must first be investigated.

Methyl bromide (CH3Br) is prevalent in the atmosphere (10-15 ppt) and major
sources include biomass combustion and biological activity in the ocean.3 The
photochemistry of CH3Br in the gas phase has been studied4 and absorption of UV
radiation (7mm = 200 nm, 0' = 1x10''8 cmz) leads to n—>0'* from the Br atom to the C-Br
antibondin g orbital to form the photolysis products Br and CH3. Free halogen atoms are

known to add to unsaturated molecules across the double bond, and specific studies can

57

be found in the literature including reactions of chlorine atoms with butadiene and
isoprene under atmospheric conditions.5I6

Methyl bromide photochemistry has been studied on several different surfaces
including metalszj,8 and nonmetals9vlo under UHV conditions. On metal substrates the
Br-C bond breaking occurs in the adsorbed state at photon wavelengths between 250 and
200 nm. Zhou and White8 investigated the photodissociation of CH3Br on Ag(] 1 1) using
XPS to determine changes in the chemical environment of the Br atom as a function of
irradiation time. They measured the bromine 3d orbital binding energy in methyl
bromide/Ag(] 1 1) to be 70.7 eV. Upon irradiation of the surface with a xenon arc lamp
the two separate Br 3d peaks could be seen in the X-ray photoelectron spectra, one at
70.7 eV and the other at 68.7 eV. Further irradiation decreased the 70.7 eV peak area
relative to the 68.7 eV peak area. The authors attribute the growth of the second peak at
68.7 eV to photodissociated Br atoms on the Ag( 1 1 1) surface.

Some preliminary experiments to characterize the adsorption and photochemistry of
CH3Br on HOPG have already been preformed. Introduction of CH3Br to the UHV
chamber and HOPG surface was identical to that of butadiene as discussed in chapter 2.
Figure 4.1 shows TPD data for the parent mass of the molecule, 94 amu, with increasing
exposure times (1 sec to 60 sec) on the HOPG surface at 85 K. The broad background
present in the TPD spectra, ranging from 100 K to 220 K, is attributed to desorption of
CH3Br from the molybdenum sample mount. The sharper peaks growing into the spectra
at increased exposure are due to monolayer and multilayer desorption of CH3Br from the

HOPG surface at l 15 K and 108 K respectively. Multilayer adsorption regimes begin to

58

 

 

94)

 

 

 

Intensity (m

 

 

 

 

 

 

 

 

.J C
a.
I I I I I I I I I I I
90 120 150 180 210 240

Temperature (K)

Figure 4.1 Temperature programmed desorption spectra (mass 94) for CH3Br on HOPG
with increasing exposure time; a) 1 sec, b) 5 sec, c) 10 sec, (1) 20 sec, e) 40 sec, f) 60 sec.

59

form on the HOPG surface before the first monolayer is saturated as indicated by growth
of the monolayer peak in the TPD spectra simultaneously with growth of the multilayer
peak.

X-ray photoelectron spectroscopy was used to monitor the result of exposing 1 ML
CH3Br on HOPG to UV photons (2. = 193 nm) at low flux (~10 mJ/pulse). Figure 4.2
shows the Br (3d) XPS peak before and after various irradiation times. Before irradiation
the binding energy was found to be 71.69 eV. After irradiation for 30 sec, two new peaks
appeared in the X-ray photoelectron spectra, one at 70.92 eV and one at 68.48 eV. These
two new peaks are attributed to surface bound Br atoms at 70.92 eV and possibly Br; at
68.48 eV, although further experiments will have to be performed to confirm this.
Irradiation for longer time produces qualitatively similar results but an overall loss in
XPS intensity indicates a decrease in total surface bound bromine. If either Br or Brg are
formed on the surface as a result of irradiation, then either species could add to a double
bond on butadiene.

Nitrogen dioxide (N02) is a photoactive molecule with similar release sources as both
butadiene and carbonaceous aerosols. N02 photodissociates to form NO and O at

wavelengths between 250 nm and 400 nm; the wavelength for maximum absorption is
about 390 nm where the cross section is 6x10‘19cm2.3 N02 is known to play a significant

role in the chemistry of both organics and ozone in the troposphere.3 In fact, oxidation
reactions of organic molecules by various radicals in the presence of NO leads to the

photodissociation of NO; in the troposphere. The adsorption of N02 has already been

60

 

'7169
'fl166

 

681MB

 

M
v
\we

\wa
\LJz-e
”MA. 8.

I I I I I I I I I I I I I I I

80 78 76 74 72 70 68 66 64

Binding energy (eV)

 

 

 

 

Figure 4.2 X—ray photoelectron spectra of 1 ML CH3Br exposed to 5 mJ/pulses at 10 Hz
of 193 nm light for increasing times; a) 30 min, b) 10 min, 0) 5 min, d) 2.5 min, e) 0.5
min, f) 0 min showing binding energy changes.

61

studied on the HOPG surface by TPD and HREELS.11 The molecule forms both mono
and multilayers on graphite at 90 K which desorb at about 150 K and 140 K respectively.
At all coverages, N02 reacts on the HOPG surface to form N 204. Characterization of the
condensed phase photochemistry of N02 on HOPG would have to be performed to
determine if the products of the photochemical reaction, NO and O, remain adsorbed on
the surface at 85 K.

In this thesis, a through study on the adsorption and photochemistry of butadiene on
HOPG is presented, using HOPG as a model for carbonaceous aerosols. At high photon
fluence (100 mJ/pulse) butadiene molecularly desorbs from the HOPG surface by way of
a localized surface heating mechanism. At low photon fluences (10 mJ/pulse), butadiene
shows no condensed phase photochemistry, even at wavelengths where the molecule has
a large cross section for absorption. To further investigate any role carbonaceous
aerosols may play in heterogeneous reactions in the atmosphere, further experiments need

to be performed involving co-adsorbates with butadiene on the HOPG surface.

62

4.3 References

(1)
(2)

(3)

Pub, W.; Schmid, W. E.; Trushin, S. A. Chem. Phys. Lett. 2001, 342, 91.
Zhou, X. L.; Zhu, X. Y.; White, J. M. Surf. Sci.Rep. 1991, 13, 73-220.

Brasseur, G. P.; Orlando, J. J .; Tyndall, G. S. Atmospheric Chemistry and Global

Change; Oxford University Press, Inc: New York, 1999.

(4) Calvert, J. G.; Pitts Jr., J. N. Photochemistry; John Wiley and Sons: New York,
1966.

(5) Ragains, M. L.; Finlayson-Pitts, B. J. J. Phys. Chem. A 1997, 101, 1509.

(6) Wang, W.; Finlayson-Pitts, B. J. J. Geophys. Res. 2001, 106, 4939.

(7) Lamont, C. L. A.; Conrad, H.; Bradshaw, A. M. Surf. Sci. 1993, 280, 79-90.

(8) Zhou, X.-L.; White, J. M. Surf. Sci. 1991, 241, 259.

(9) Heyd, D. V.; Jensen, E. T.; Polanyi, J. C. J. Chem. Phys. 1995, 103, 461.

(10) Kim, S. H.; Stair, P. C.; Weitz, E. J. Chem. Phys. 1998, 108, 5080.

(1 l) Sjovall, P.; So, S. K.; Kasemo, B.; Franchy, R.; Ho, W. Chem. Phys. Lett. 1990,
171, 125.

63

APPENDICES

64

Appendix A Mass Spectra

Before experiments, the gas phase cracking pattern of 1,3-butadiene was obtained to
ensure purity. The molecule was introduced to the UHV chamber via the molecular leak
valve. Mass spectra were collected from 0-50 amu for the background and 6-56 amu for
butadiene, at 10 amu/sec. The electron multiplier voltage was set at 1.7 kV. The gas
phase cracking pattern of butadiene collected with our mass spectrometer was compared
with that reported by NIST. Qualitatively the spectra were identical expect for a mass 28
which is larger in our spectrum. The mass 28 signal is due to the oxidation of butadiene
on the ion gauge filament to form carbon monoxide. This was tested by monitoring mass
28 signal with the ion filament on and off. During experiments the ion filament was kept
on in order to monitor pressure in the chamber. Carbon monoxide does not stick to

graphite at 77 K and so did not affect butadiene adsorption on HOPG during experiments.

65

10.0

 

 

 

 

 

 

 

 

8.0- (a)
6.0-
53 .
'o
‘_ 4.0- 1
X 1
C207 MM
5 .
V II I I I I II IIII III
a) 10 20 30 40 50
L
3
$100
9 ‘ (b)
n. 8.04
.72 6 '
t .0-
c0
0- 4.0-

 

.0 .N
O O
I A I n

 

O

 

1'0 ZIOT 730 40 T50
m/z (amu)

Figure A.l Mass spectra of (a) 1,3—Butadiene at a pressure of 5x10”9 Torr and (b) the
background vacuum of the chamber at a pressure of 2x10'10 Torr.

66

Appendix B Calculation of desorption energies for Butadiene on HOPG

The rate of desorption from a surface during a TPD experiment can be described
using the Polanyi-Wigner equation.

-dn/dt = n"vx exp(-Ed/RT) (1)

When desorption follows first-order kinetics the desorption energy can be determined
using a modification of the Polanyi-Wigner equation developed by Redhead.I In order to
apply Redheads method Ed is assumed to be constant with both coverage and
temperature. By setting x = l, the temperature for the maximum rate of desorption can be

determined by differentiating (1) to get:
dr/dT = r -( Ed/RTp2 — v/B exp(-Ed/RTp)) (2)
where B is the heating rate (K/s) and r the rate of desorption (dn/dt), then the temperature
for maximum desorption occurs when.
Ed/RTPZ - v/B exp(-Ed/RTp) = 0 (3)
Rearranging (3) gives
Ed = RTp ((ln(vpr/[3) — 3.46) (4)
where plots of Ed vs Tp produces a linear line when v is between 10'2 and 10l4 Hz.
Multilayer desorption energies (zero order desorption processes) were determined by
leading-edge analysis? Equation (1) can be written as (take the In of both sides)
ln(-dn/dt) = ln(v/B) - (Ed/R- VT) (5)
Plots of ln(-dn/dt) vs l/T gives a linear line where the slope is equal to (-Ed/R) and the
intercept equal to ln(v/B). Both the frequency factor (v) and Ed can then be determined by

leading-edge analysis.

67

Appendix C Calculation for energy absorbed by

HOPG during laser irradiation

To determine the surface temperature rise of HOPG as a result of laser irradiation
using equation 1 in chapter 3 the amount of energy absorbed by the surface must be

calculated using the Fresnel equation for p polarized 1i ght.3

EP : COSG—J(T]R2 —T]I2 —sin2 0) +I(211R1y

cose + (finaz — n12 — sin2 0) + {(211m)

 

 

Where (Ep)2 is the amount of energy reflected by the surface (R), 11R and mare the real
and imaginary parts of the index of refraction for HOPG at 193 nm, 248 nm and 351 nm
and are 1.67, 0.26, 1.51, 0.71, 2.02 and 0.694 respectively and 0 the angle of incidence.
The complex quantity Ep can be reduced to a ratio of complex numbers.

R = (a2 + b2)/(c2 + d2)

Where the final result to calculate reflectivity at a incident angle is:

R = (11R —1/cos€))2 +1112
(1111 +l/c030)2 +1112

The reflectivity at wavelengths of 193 nm, 248 nm and 351 nm and an angle of incidence
= 45° are calculated to be 1.3%, 5% and 6.8% respectively. Therefore the amount of
energy absorbed by the surface is the total amount of energy exposed to the surface

minus the percent reflected.

68

References

(l) Redhead, P. A. Vacuum 1962, 12, 203.

(2) Jong, A. M.; Niemantsverdriet, J. M. Surf Sci. 1990, 223, 355.

(3) Pedrotti, F. L.; Pedrotti, L. S. Introduction to Optics; Prentice Hall: Upper
Saddle River, New Jersey, 1993

(4) Borghesi, A.; Guizzetti, G. Handbook of Optical Constants of Solids 2;

Academic Press: 1991

69

l.‘l'llllllll I...
F I"...

    

         
    

1111111111111

320 6992

1111111

l

1111