University of Southern California Dissertations and Theses

Chemisorption on the (111) and (100) faces of platinum-tin bimetallic surfaces

(USC Thesis Other)

Chemisorption on the (111) and (100) faces of platinum-tin bimetallic surfaces

PDF

Download a page range

Download transcript

Copy asset link

Request this asset

Transcript (if available)

Content INFORMATION TO USERS
This manuscript has been reproduced from the microfilm master. U M 1 films
the text directly from the original or copy submitted. Thus, some thesis and
dissertation copies are in typewriter face, while others may be from any type of
computer printer.
The quality of this reproduction is dependent upon the quality of the
copy submitted. Broken or indistinct print, colored or poor quality illustrations
and photographs, print bleedthrough, substandard margins, and improper
alignment can adversely affect reproduction.
In the unlikely event that the author did not send UMI a complete manuscript
and there are missing pages, these w ill be noted. Also, if unauthorized
copyright material had to be removed, a note will indicate the deletion.
Oversize materials (e.g., maps, drawings, charts) are reproduced by
sectioning the original, beginning at the upper left-hand corner and continuing
from left to right in equal sections with small overlaps.
Photographs included in the original manuscript have been reproduced
xerographically in this copy. Higher quality 6" x 9” black and white
photographic prints are available for any photographs or illustrations appearing
in this copy for an additional charge. Contact UMI directly to order.
ProQuest Information and Learning
300 North Zeeb Road, Ann Arbor, M l 48106-1346 USA
800-521-0600
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
CHEMISORPTION ON THE (111) AND (100) FACES OF Pt-Sn BIMETALLIC
SURFACES
by
Chameli Panja
A Dissertation Presented to the
FACULTY OF THE GRADUATE SCHOOL
UNIVERSITY OF SOUTHERN CALIFORNIA
In Partial Fulfillment of the
Requirements for the Degree
DOCTOR OF PHILOSOPHY
(Materials Science)
April 2000
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
UMI Number: 3018024
___ ®
UMI
UMI Microform 3018024
Copyright 2001 by Bell & Howell Information and Learning Company.
All rights reserved. This microform edition is protected against
unauthorized copying under Title 17, United States Code.
Bell & Howell Information and Learning Company
300 North Zeeb Road
P.O. Box 1346
Ann Arbor, Ml 48106-1346
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
UNIVERSITY OF SOUTHERN CALIFORNIA
THE GRADUATE SCHOOL
UNIVERSITY PARK
LOS ANGELES, CALIFORNIA 90007
This dissertation, written by
under the direction o f h &. ' X T .. .. . Dissertation
Committee and approved by a U its members,
has been presented to and accepted by The
Graduate School, in partial fulfillment of re
quirements for die degree of
DOCTOR OF PHILOSOPHY
Dean of Graduate Studies
Date
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Dedicated to
my grandparents;
Dasharathi and Champa Adak
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Acknowledgments
First, I would like to express my heartiest appreciation to my research advisor Dr.
Bruce Koel, for his support and guidance without which this work would not have been
possible. His excellence, enthusiasm and creativity will always be a source of inspiration
to me in both my professional and personal endeavors. The great patience and
understanding which he showed me, especially towards the difficulties and problems that
most foreign students often meet, has made my time here at USC one of the most
memorable and rewarding periods of my life.
The continuous collaboration and encouragement provided by a long list of former
and present Koel group members was very important to the completion of this research.
While everybody contributed to this, I am specially thankful to Dr. John Peck, Dr. Jiang
Wang, Dr. Nathan Swami, Dr. Cristof Baur, Dr. Hong He, Dr. Michael Quinlan, for not
only providing technical knowledge, good ideas and moral support, but also for creating an
environment where scientific work was made more stimulating and challenging. I would
like to thank Dr. Najat Saliba and Dr. Yi-Li Tsai for initially training me in UHV
technology, and David Beck and Dmetri Jerdev for solving all the computer problems for
me. My special thanks goes to Dennis Semin, your friendship and encouragement over
the past five years have been invaluable.
My family has provided a constant and seemingly unlimited source of support and
encouragement during my studies in California. This thesis will be incomplete without
acknowledging my parents, brother and my in-laws, for their love, confidence in me and
support from thousands of miles away. I am greatful to my parents for giving me the
opportunity for an education and a fife which very few people are fortunate enough to
iii
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
have. Finally, I am very indebted to my loving husband Siddhartha. Together we shared
the ups and downs of my graduate years here at USC.
The funding by the Army Research Office and the Division of Chemical Sciences,
Office of Basic Energy Sciences, US Department of Energy is gratefully acknowledged.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Table of Contents
ACKNOWLEDGMENTS
LIST OF FIGURES
LIST OF TABLES
ABSTRACT xiii
1. Introduction 1
2. Experimental Methods 10
2.1 Apparatus 10
2.2 Preparation of Sn/Pt surface alloys 12
3. Adsorption of Methanol, Ethanol and Water on Well-Characterized
Pt-Sn Surface Alloys 23
3.1 Introduction 24
3.2 Experimental Methods 26
3.3 Results and discussion 28
3 > 5 Conclusion 35
3.6 Acknowledgments 36
3.7 References 37
4. Probing the Influence of Alloyed Sn on Pt(100) Surface Chemistry by CO
Chemisorption 53
4.1 Introduction 54
4.2 Experimental Methods 57
4.3 Results 58
4.4 Discussion 63
4.5 Conclusion 66
4.6 Acknowledgments 67
4.7 References 68
5. Influence of Alloyed Sn on Adsorption and Reaction of NO on Pt(IOO)
Surfaces 82
5.1 Introduction 83
5.2 Experimental Methods 85
5.3 Results and 87
5.4 Discussion 95
5.4 Conclusion 101
5.5 Acknowledgments 102
5.6 References 103
6. Coking Resistance of Pt-Sn Alloys Probed by Acetylene Chemisorption 123
6.1 Introduction 124
6.2 Experimental Methods 125
6.3 Results and Discussion 126
6.5 Conclusions 131
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
7. Acetylene Chemisorption on Sn/Pt(100) Surface Alloys 140
7.1 Introduction 141
7.2 Experimental Methods 143
7.3 Results 145
7.4 Discussion 151
7 5 Conclusion 155
7.6 Acknowledgments 156
7.7 References 157
8. Conclusion 175
Bibliography 178
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
List of Figures
Fig. 2.1: Ultrahigh vacuum (UHV) chamber where the experiments 16
were performed.
Fig. 2.2: Arrangement of upper (top) and lower (bottom) levels of the 17
UHV chamber.
Fig. 2.3: LEED image of the reconstructed hex-Pt(lOO) surface taken 18
at a beam energy of 60 eV.
Fig. 2.4: Schematic top-view of the structure of (2x2)Sn/Pt(l 11) (top) 19
and (V3xV3)R30° Sn/Pt(lll) (bottom) surface alloys.
Fig. 2.5: LEED image of the Sn/Pt(100) surface taken at a beam 20
energy of 60 eV. Top: c(2x2)Sn/Pt(100). Bottom:
(3V2xV2)R45° Sn/Pt(100).
Fig. 2.6: Schematic top-view of the structure of two Sn/Pt(100) 21
surfaces. Top: c(2x2)Sn/Pt(100) overlayer. Bottom:
c(2x2)Sn/Pt(100) surface alloy.
Fig. 3.1: Methanol TPD spectrum after methanol exposures on the 39
P t(lll) surface. The multilayer desorption peak at the
highest exposure has been cut off. Exposures from the
bottom to top are 10,20,40,60 and 70 s.
Fig..3.2: TPD spectrum of methanol on (2x2) Sn/Pt(l 11) surface with 40
different coverages. The exposures from the top to bottom
are 5,20,40 and 60 s.
Fig. 3.3: Methanol TPD spectrum after methanol exposures on 41
(V r 3xVf 3)R30° Sn/Pt(lll) surface with coverages bottom to
top 10,20,40 and 60 s.
Fig. 3.4: Ethanol TPD spectrum after ethanol exposures on P t(lll) 42
surface with different coverages. The exposures from top to
bottom were 5,10,20,40,60 and 80 s.
Fig. 3.5: TPD spectrum of ethanol on (2x2) Sn/P t(lll) surface with 43
different coverages. The exposures from top to bottom were
10,20,40,60 and 120 s.
Fig. 3.6: Ethanol TPD spectrum after ethanol exposures on 44
(a /3 x-/3)R30° Sn/Pt(l 11) surface with exposures top to
bottom 5, 10,20,40 and 80 s.
Fig. 3.7: Methanol and ethanol uptake curve results fron TPD 45
experiments for different exposures on Pt(l 11), (2 x2) and V3
surfaces.
vu
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Fig. 3.8:
Fig. 3.9:
Fig 3.10:
Fig. 3.11:
Fig. 3.12:
Fig. 3.13:
Fig. 4.1:
Fig. 4.2:
Fig. 4.3:
Fig. 4.4:
Fig. 4.5:
Fig. 4.6:
Fig. 4.7:
Fig. 4.8:
Comparison of methanol and ethanol desorption spectra for 46
0 s O.10sat of methanol and ethanol on Pt(l 11), (2x2) and V3
surface alloy.
TPD spectrum of water dosed at 100 K on P t(lll) surface 47
with different exposure.
Water TPD spectrum after water exposures on ^8
(2x2)Sn/Pt(l 11) surface alloy with different coverages at 100
K.
TPD spectrum of water on (V3x\/3)R30° S n/P t(lll) surface 49
with different coverages.
Comparison of desorption peak temperature of about 10% of 50
saturation monolayer coverage of watert on P t(lll) and the
(2x2) and V3 alloy surfaces.
Influence of the alloyed Sn concentration on the desorption ^1
peak temperature of methanol ethanol and water in the Sn/pt
surface alloys. The right hand axis gives an estimation of the
corresponding molecular desorption energies.
CO TPD spectra after CO exposures on (5x20)-Pt(100) at ^0
150 K.
CO TPD spectra following CO exposures on the 71
c(2x2)Sn/Pt(100) alloy surface at 150 K.
CO TPD spectra after CO exposures on the (3V2xV2)R45°
Sn/Pt(100) alloy surface at 150 K.
Comparison of CO TPD spectra from Pt(100)(top) and ^3
c(2x2)Sn/Pt(100) overlayer (bottom) to those from the
c(2x2) alloy and 3a/2 alloy.
CO uptake by hex-Pt(100) and the Sn/Pt(100) surface alloys. ^4
Relative CO coverages were obtained by using the integrated
CO TPD peak areas from Figs. 4.1-4.3, with 0s a tco on Pt(100)
set to 0.77 ML [28, 29]. Uptake curves on an expanded scale
below 0.6 L CO exposure are reshown in the inset in order to
more easily see the effects of alloyed Sn on the initial sticking
coefficient of CO.
HREELS spectra of CO on hex-Pt(100) at 150 K. ^
HREELS spectra of CO for an annealing sequence following a 76
saturation dose of CO on Pt(100) at 150 K.
HREELS spectra of CO on the c(2x2)Sn/Pt(100) alloy at 150 77
K.
vui
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Fig. 4.9:
Fig. 4.10:
Fig. 4.11:
Fig. 4.12:
Fig. 5.1:
Fig. 5.2:
Fig. 5.3:
Fig. 5.4:
Fig. 5.5:
Fig. 5.6:
Fig. 5.7:
Fig. 5.8:
Fig. 5.9:
Fig. 5.10:
HREELS spectra of CO on the (3V2xV2)R45° Sn/Pt(100) 78
alloy at 150 K.
Possible model for the 2D structure of CO adsorbed on the 79
c(2x2) Sn/Pt(100) alloy at saturation CO coverage.
Comparison of the saturation CO coverage on Pt(l 11), Pt(100), 80
(2x2)Sn/Pt(l 11) alloy, (V3x/3)R30° S n/P t(lll) alloy,
c(2x2)Sn/Pt( 100) alloy, and (3^2xV2)R45° Sn/Pt(100) alloy at
150 K.
Comparison of the initial sticking coefficient of CO on Pt(l 11), 81
Pt(100), (2x2)Sn/Pt(lll) alloy, (V3xV3)R30° S n/P t(lll) alloy,
c(2x2)Sn/Pt(100) aUoy,and (3V2xi/2)R45° Sn/Pt(100) alloy at
150 K.
TPD curves used to identify products from the reaction of a 106
NO monolayer on the hex-Pt(100) surface at 100 K.
NO, N 2 and 0 2 TPD traces following NO exposures on the 107
hex-Pt(100) surface.
Uptake curves constructed from TPD results detailing NO 108
adsorption kinetics on hex-Pt(100) at 100 K. The value for
0 gf) on Pt(100) was set to 0.73 ML [29].
109
TPD traces from a NO monolayer on the c(2x2)Sn/Pt(100)
alloy at 100 K.
NO, N ,0 and 0 2 TPD traces obtained after NO exposures on
the c(2x2) Sn/Pt(100) alloy at 100 K.
TPD curves from a NO monolayer on the (3Vr 2xV2)R45°
Sn/Pt(100) alloy at 100 K.
NO, N ,0 and 0 2 TPD traces obtained following NO ^
exposures on the (3Vr 2xV2)R45° Sn/Pt(100) alloy at 100 K.
113
Uptake curves obtained from TPD results for NO adsorption
kinetics at 100 K on (a) the c(2x2) Sn/Pt(100) alloy and (b) the
(3V f 2xV2)R45° Sn/Pt( 100) alloy.
Comparison of NO TPD results from hex-Pt(100), the two ^ 4
alloys, and a c(2x2) overlayer at two coverages of NO (a) 0N O
= 0.10 (b) ® n o = 6 NO a t 100 K o n eack surface.
HREELS spectra following NO exposures on the hex- 115
Pt(100) alloy at 100 K.
ix
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Fig. 5.11:
Fig. 5.12:
Fig. 5.13:
Fig. 5.14:
Fig. 5.15:
Fig. 5.16:
Fig. 6.1:
Fig. 6.2:
Fig. 6.3:
Fig. 6.4:
Fig. 6.5:
Fig. 7.1:
Fig. 7.2:
Fig. 7.3:
HREELS spectra of NO after sequentially heating to higher 116
temperature following a saturation dose of NO on hex-Pt(lOO)
at 100 K.
HREELS spectra following NO exposures on the 117
c(2x2)Sn/Pt(100) alloy at 100 K.
HREELS spectra obtained after stepwise annealing a 118
saturation dose of NO on the c(2x2) alloy at 100 K.
HREELS spectra after NO exposures on the (3V2xV2) R45° 119
Sn/Pt(100) alloy at 100 K.
Schematic representation of models for (a) monomeric NO 120
species adsorbed at different sites and (b) dimeric NO species
adsorbed as surface dinitrosyl and (NO)2 dimer configuration.
Comparison of (a) normalized NO saturation coverage and (b) 121
initial sticking coefficient S0 of NO on Pt surfaces and several
Pt-Sn surface alloys at 100 K. The concentration of Pt atoms
in the surface layer is shown with the dashed line and scaled on
the right hand side.
Acetylene TPD curves after saturation acetylene exposures on 135
Pt and Sn/Pt alloys at 100 K.
D, evolution in TPD after saturation exposure of acetylene 136
on Pt and Pt-Sn alloys at 100 K.
(a) C6D6 and (b) C4D6 TPD traces after acetylene saturation 137
exposures on these surfaces at 100 K.
Comparison of (a) acetylene monolayer coverage at 100 K and 138
(b) acetylene desorption activation energy versus surface Sn
concentration in the Sn/Pt(l 11) and Sn/Pt(100) surface alloys.
Fractional decomposition of acetylene in the monolayer 139
versus surface Sn concentration on in the Sn/Pt(l 11) and
Sn/Pt(100) surface alloys.
TPD spectra of products from the decomposition of a 159
saturation monolayer of acetylene on hex-Pt(100).
D, TPD spectra after C,D, exposures on hex-Pt(100) surface 160
at lOO K.
TPD spectra obtained from the reaction of acetylene on the 161
c(2x2) Sn/Pt(100) alloy after saturation exposure of acetylene
at 100 K.
x
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Fig. 7.4:
Fig. 7.5:
Fig. 7.6:
Fig. 7.7:
Fig. 7.8:
Fig. 7.9:
Fig. 7.10:
Fig. 7.11:
Fig. 7.12:
Fig. 7.13:
Fig. 7.14:
Fig. 7.15:
Fig. 7.16:
C2D 2 TPD spectra obtained after acetylene exposures on the 162
c(2x2) alloy at 100 K.
Product formation from the reaction of a saturation 163
monolayer of acetylene on the (3V2xV2)R45° Sn/Pt(100)
surface during TPD.
TPD spectra of acetylene after C2D2 exposures on the 3V2 164
alloy at 100 K.
Comparison of C,D2, D,and C6D6 TPD spectra obtained from 165
a saturation exposure of acetylene at 100 K on different
surfaces.
Plots of ln((3/Tp 2 ) versus 1/Tp for desorption of acetylene on the 166
two Sn/Pt alloyed surfaces at heating rates 1-24 K/s.
C (ls) spectra of acetylene after acetylene exposures on clean 167
hex-Pt(100) at 100 K.
(a) C (ls) spectra obtained following adsorption of a saturation 168
monolayer of acetylene on hex-Pt(100) at 100 K as a function
of annealing temperatures. (b) C(ls) intensity versus
temperature for each annealing temperatures shown in (a).
C (ls) spectra of acetylene after acetylene exposures on the (a) 169
c(2x2) alloy and (b) 3V2 alloy at 100 K.
XPS C (ls) spectra obtained following adsorption of a 170
saturation monolayer of acetylene on the 3v2 alloy at 100 K as
a function of annealing temperature. The inset shows the C (ls)
intensity versus annealing temperature for acetylene on the 3V2
alloy.
Comparison of C(ls) spectra for a saturation monolayer of 171
acetylene on several surfaces.
The uptake curve constructed from the intensity of C (ls) 172
photoelectron line versus the C2D2 exposures on clean hex
Pt(100) and the two alloyed surfaces. The C2D2 coverage was
plotted in the right hand side axis as estimated from the known
saturation coverage of CO on the clean Pt(100) surface [37].
Comparison of (a) saturation monolayer coverage and (b) initial 173
sticking coefficient of acetylene versus surface Sn
concentration on Sn/Pt alloyed surface. The squares represents
Pt atom concentration in atoms/cm2, plotted on right hand side
axis.
Fraction of decomposition of acetylene versus surface Sn on 174
several Sn/Pt surfaces.
xi
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
List of Tables
Table 2.1:
Table 3.1:
Table 5.1:
Table 52:
Table 53:
Pt-Sn alloy preparation conditions. 22
Thermal desorption peak temperatures (top entries) and 52
desorption activation energies in kJ/mol (bottom entries) for
methanol, ethanol, and water obtained at about 10% of the
monolayer saturation coverages on P t(lll) and two Pt-Sn
surface alloys. The left hand values given in parentheses are
activation energies obtained by Redhead analysis, and the right
hand values were determined by leading edge analysis.
Vibrational energies (cm'1 ) of adsorbed NO for a saturation 122
dose of NO on Pt(100) and Sn/Pt(100) at 100 K.
Vibrational energies (cm'1 ) of N2 0 for a saturation dose of NO 122
on Sn/Pt(100) alloy surfaces at 100 K.
Vibrational energies (cm'1 ) of the dinitrosyl intermediate on 122
Sn/Pt(100) alloys after a saturation exposure of NO at 100 K.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Abstract
Chemisorption and reaction of CH3 OH (methanol), C2H5 OH (ethanol), and H2 0
(water) on Pt(l 11) and Sn/Pt(l 11) alloys, and CO (carbon monoxide), NO (nitric oxide),
and C2D, (acetylene) on Pt(100) and Sn/Pt(100) have been studied under ultrahigh
vacuum conditions using temperature programmed desorption (TPD), Auger electron
spectroscopy (AES), low energy electron diffraction (LEED), X-ray photoelectron
spectroscopy (XPS) and high resolution electron energy loss spectroscopy (HREELS).
Small organic molecules like CH3 OH and C,HsOH are potential fuels for low-temperature
hydrocarbon fuel cells and it is important to understand the role of tin as a promoter in
electrooxidation of these molecules. Also, the catalytic reactions of CO, NO and C 2H 2 are
of considerable interest for improving of automotive exhaust-gas catalytic converters and
other heterogeneous catalysts. Ordered Pt-Sn alloys can be prepared by vapor deposition
of Sn on Pt surfaces. A (2x2) structure (0S n = 0.25) and a (V r 3xV3)R30° structure are
formed on P t(lll), and a c(2x2) and (3v(2xV2)R45° structures with 0S n = 0.5 and 0.67,
respectively, are formed on Pt(100). CH3OH, C2HsOH and H ,0 are all weakly bound
and reversibly adsorbed on P t( lll) and both of the S n /P t(lll) alloys under UHV
conditions. Alloying Sn into the P t(lll) surface weakens the adsorption of these
molecules from that on Pt(l 11) and leads to a lower reactivity as the surface concentration
of Sn increases. TPD measurements reveal a reduction in the saturation coverage and
chemisorption bond energy for CO, NO and C,H-? chemisorption, on the two
Sn/Pt(100)alloys compared to that on Pt(100). CO chemisorption is completely reversible
on these two Sn/Pt(100) alloys. However, NO is partially reduced to form N ,0 on these
alloys, so that N ,0 along with NO and 0 2 desorption was observed. We propose that
dinitrosyl species, i.e., two NO molecules bound to one Pt atom, are intermediates in N2 0
xiii
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
formation from adsorbed NO on these Sn/Pt(100) surface alloys. Irreversible dissociative
adsorption of acetylene was strongly suppressed (~80-90%) on both of the two
Sn/Pt(100) alloys. About 15% of the adsorbed acetylene monolayer was converted to
benzene and desorbed during TPD on the (3V2xV2)R45°Sn/Pt( 100) alloy, but no benzene
was desorbed from the c(2x2) alloy.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Chapter 1
Introduction
Bimetallic catalysts play an important role in industrial catalytic processes. Catalytic
activity or selectivity can be greatly enhanced by the addition of a second metal to the
primary active metal [1-4]. Fundamentally, there is a continuing effort to identify the
origin of the effect of a second component in bimetallic systems: electronic and geometric
(ensemble) effects [5-6]. Catalysis is a kinetic phenomenon, and it should be useful to
know how many metal atoms are required per adsorbed molecule for the most favorable
path of a catalytic reaction. This can be called the ensemble requirement [5-6]. Although
this is based on rather elementary chemical and geometric principles, it can be of
considerable help in rationalizing performance of industrial catalysis and designing
superior catalysts. The size of the ensemble available for chemisorption and catalysis can
be varied by alloying one metal with another, or by adding impurities which block certain
sites.
One reason for lack of progress in studying Pt-Sn bimetallic catalysts is that detailed
information about the composition and structure of the reactive sites of the catalyst is not
available. Surface segregation and adsorbate-induced reconstruction has limited the
understanding of fundamental chemisorption experiments on bulk Pt-Sn alloys [7].
Changes in the surface chemistry of (111) and (100) faces of platinum single crystals
1
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
following the addition of tin to form surface alloys under UHV conditions, are helpful for
modeling of electronic and ensemble effects in related industrial supported-catalyst
systems.
Ordered alloys limit the Pt ensemble sizes available for adsorption and reaction, often
to rather small numbers. In addition, specific reaction sites can be blocked and this allows
fine control of reactive sites available at the surface. The stability and reproducibility of
these surface alloys allows us to probe more fully the chemistry of Pt-Sn bimetallic
systems at a molecular level. Also the ability to control (to some extent) the geometry and
composition of these alloy surfaces provide us with additional capability to separate
geometric and electronic effects in explaining the unique chemistry of these surfaces.
The work reported in this thesis specifically, we used ordered Pt-Sn surface alloys as
model catalysts under UHV conditions contributes to the understanding of the chemistry
of small molecules (CO, NO, H2 0 , C2H2, CH3 OH and C,HsOH) on these surfaces. The
thesis is divided into two parts. In the first part (Chapter 4), results of chemisorption
studies of methanol, ethanol, and water on P t(lll) and S n /P t(lll) surface alloys is
described. The second part (Chapters 4-7) describes the effect of Sn on the chemisorption
of small molecules like CO, NO and C2H2 on Pt(100).
These Sn/Pt surface alloys can be obtained by Sn deposition on Pt single crystal
surfaces. A p(2x2) structure (0 S n = 0.25 ML) and a 0/3xV3)R3O° structure ( 0 S n = 0.33
ML) can be formed on P t(lll), and c(2x2) and (3/2x\/2)R450 structures ( 0 s n = 0.5 and
0.67 ML, respectively) can be formed on Pt(lOO) [8-11]. An advantage of using these
systems is the 2D and even 3D structure of Sn is known because the surface alloy is
ordered. A disadvantage is that the Sn coverage cannot be varied continuously over a large
range on the same face of the Pt single crystal.
The initial investigation of S n/P t(lll) surface alloys was made by Paffett and
Windham [8], and their analysis that surface alloys were formed was confirmed later by
2
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Low energy alkali ion scattering spectroscopy (ALISS) [9], low energy electron diffraction
(LEED) I-V [12] and x-ray photoelectron diffraction (XPD) [13]. These studies show
that the alloyed Sn was incorporated into the Pt(l 11) surface layer with a 0.2 A outward
buckling.
The first study of the Sn/Pt(100) system was also reported by Paffett et. al. [10].
They conjectured that the observed c(2x2) and (3Vr 2xVr 2)R45° LEED patterns formed by
this system were also due to alloy formation. Later it has been confirmed by ALISS [11].
The observed c(2x2) LEED pattern on Pt(100) following Sn deposition can be due to two
different surface structures, the c(2x2) structure was formed by an overlayer of Sn
adatoms over a wide range of temperature upto 700 K. The c(2x2) alloy formation
occurred over a relatively narrow temperature range of 700-750 K. A more thermally
stable alloy structure with LEED pattern (3V2xV2)R45° was formed at higher annealing
temperature. The c(2x2) surface alloy on Pt(100) has a geometric structure resembling a
“checker board”, with one-half of the lattice positions each occupied by Sn and Pt and
where each surface-Pt atom has only Sn nearest neighbors and vice versa [10-11]. ALISS
studies suggest that the (3v'2xV2)R45° Sn/Pt(100) surface alloy has a geometric structure
that is similar to that of the c(2x2) alloy, and may be composed of small c(2x2) alloy
domains with the same buckling distance within the domains [11]. Recent STM images of
a related, bulk Pt^SnClOO) single crystal alloy surface, after sputtering and annealing the
surface to 600 K, show pyramidal features which consist of (102)-facets [14], Flat
portions between and on top of the pyramids formed by the facets were mainly comprised
of triple rows that are probably Pt. A c(2x2) LEED pattern with streaky facet spots was
observed after heating to 600 K and annealing to 1000 K leads to an improved c(2x2)
LEED pattern with no facet spots [10]. STM studies in our lab are underway to further
probe the (3V2xV2)R45° Sn/Pt(100) alloy and the relationship that it has to the structure
3
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
of the bulk Pt^SnClOO) surface. At this time, however, the (3V f 2xV2)R45°-Sn/Pt(100) alloy
structure has not been determined.
For the p(2x2)Sn/Pt(lll) alloy, pure-Pt 3-fold hollow sites exist but no adjacent
pure-Pt 3-fold sites are present. All pure-Pt 3-fold sites are eliminated on the
(V3xV3)R30°-Sn/Pt( 111) alloy and the distance between adjacent 2-fold bridge sites is
increased. On the c(2x2) alloy on Pt(100), each surface atom of Pt is isolated, i.e. there
are no Pt-Pt nearest neighbors. The importance of pure-Pt 3-fold sites and two adjacent
pure-Pt 3-fold sites is revealed by the adsorption activation energies and reactivities of
small molecules and hydrocarbons on these surfaces [15-18].
Another interest in bimetallic Pt-Sn surface chemistry comes from the possible use
of Pt-Sn alloys as an electrode material in advanced fuel cells [20-22]. Pt-Sn alloys have
generated interest as promising electrocatalysts for methanol oxidation, even though there
are conflicting results about whether these alloys causes an enhancement or inhibition of
the rate of methanol oxidation [20-22]. In order to aid discussion and resolution of the
controversies about the role of Sn on Pt-Sn alloy surfaces and to provide a firm
foundation for understanding chemistry relevant to electrooxidation of alcohol over Sn-
promoted Pt electrodes, it is important to investigate the interactions of simple alcohols
with well-defined Pt-Sn alloy surfaces.
In Chapter 3 we report on the interaction of methanol, ethanol, and water with Pt-Sn
surfaces under UHV conditions by temperature programmed desorption (TPD). The
objective of this study was to determine the adsorption energies of methanol, ethanol and
water on P t(lll) and two different bimetallic surfaces, the (2x2) and C/3xV3)R30°
Sn/Pt(l 11) surface alloys. Specifically, we wanted to evaluate whether the presence of Sn
in the surface layer leads to an increase in adsorption energy or thermally activates these
molecules for reaction due to the thermodynamic driving force provided by the Sn-O
interaction.
4
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Alloying with Sn reduces the adsorption energy of these molecules compared to that
on Pt(l 11). This is consistent with recent results that show that Pt-Sn alloys are not more
active than pure Pt for electrooxidation of methanol [21, 22]. Our data provide
benchmarks for discussing the surface chemistry of alcohols on Pt-Sn alloys, opening the
road for an improved understanding of electrooxidation and catalysis of alcohols and other
oxygenated molecules on bimetallic Pt-Sn catalysts.
The structure and composition of Sn/Pt(100) alloys formed on Pt(100) single
crystals are quite different than the two Sn/Pt(l 11) alloys that were previously
investigated. The c(2x2) alloy removes all pure-Pt two-fold bridge and four-fold hollow
sites. Hence, chemisorption of small molecules provide essential information on the roles
that surface structure and composition play in tailoring the surface chemistry of Pt/Sn and
other bimetallic alloys. The second part of the thesis comprised of Chapter 4-7 examines
the differing reactivities between Pt(100) and Sn/Pt(100) for CO, NO and C2H2
adsorption.
Fundamental aspects of the catalytic reaction of CO and NO to form C 0 2 and N 2
have been given considerable attention for design of automotive exhaust-gas converters
[23, 24]. In Chapters 4 and 5, we discuss the role of Sn in altering CO and NO
chemisorption on Pt(100). NO is isoelectronic with CO', and this additional electron in the
anti-bonding 2 k * orbital makes NO a weaker Jt-acceptor ligand and a much more
versatile ligand than CO. The adsorption energy of CO on the Sn/Pt(100) alloy surfaces
is only decreased by 12-16 kJ/mol from that on Pt(100). In contrast, NO is chemisorbed
more weakly by 60-70 kJ/mol on these two alloy surfaces than on Pt(100). Also, while
the adsorption of CO on Pt(100) and the two Sn/Pt(100) alloys is completely reversible, a
substantial change occurs in the NO chemistry on Sn/Pt(100) surfaces because of the
presence of Sn. On Pt(100), about 25% of the NO monolayer decomposes to eventually
desorb N0 and O,. In the presence of Sn on the two alloys, N2 0 formation occurs from
5
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
NO reaction at very low temperatures, even at 100 K. We believe that this reactivity arises
from the low temperature formation of a NO dinitrosyl complex which facilitates N ,0
formation. Our data suggest that the absence of adjacent strong bonding sites that exist
for NO on pure Pt surfaces leads to the formation of a dinitrosyl complex which
subsequently decomposes to form N ,0 .
Chapters 6 and 7 represents another aspect of heterogeneous catalysis, that of
catalytic reforming of hydrocarbons. P t( lll) and Pt(100) surfaces are known to be
highly reactive towards acetylene decomposition to form hydrogen and adsorbed carbon.
This was the only reaction pathway observed in UHV studies [25-27]. Such a high
reactivity, which leads to non-specific carbon build-up, is not desired in most industrial
reactions and so commercial hydrocarbon conversion catalyst often utilize bimetallic Pt-
based catalysts containing a second metal component, to modify (reduce) the reactivity of
Pt. In addition to being a prototype “coke precursor” molecule with a C:H stoichiometry
of unity, C2H, can undergo a rather unique C-C bond coupling reaction that can be studied
in UHV, i.e. benzene formation via cyclotrimerization. This reaction was found previously
to occur on S n /P t(lll) by our group [28]. Still, 35% of the adsorbed acetylene
monolayer decomposed on the (V3x- '/3)R30° S n /P t(lll) alloy with 0S n = 0.33 and only
about 10% of the adsorbed acetylene underwent cyclization to form benzene, and so it is
of interest to investigate further Sn addition.
Alloying Sn to form the c(2x2) and (3V2xV2)R45° Sn/Pt(100) surface alloys
removes all pure-Pt four-fold hollow sites and pure-Pt two-fold bridge sites, leaving only
isolated Pt surface atoms. There are many changes in acetylene adsorption and reaction
caused by this modification. Decomposition of acetylene was greatly suppressed (more
than 90%) on both alloys. The monolayer coverage of C2D, decreases from 0.54 ML on
Pt(100) to 0.32 ML on the c(2x2) alloy and 0.25 ML on the (3V2xV2)R45° Sn/Pt(100)
alloy also reduced the initial sticking coefficient (S0 )of acetylene at 100 K by a factor of
6
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
two. Cyclotrimerization of acetylene to benzene was observed only on the (3Vr 2xVr 2)R45°
Sn/Pt(100) alloy. This may be because of (lll)-like steps or facets present on the
surface.
7
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
REFERENCES
[1]. J. H. Sinfelt, Bimetallic Catalysts: Discoveries, Concepts and applications, Wiley,
New York, 1983.
[2]. H. Lieske and J. Volter, J. Catal. 90 (1984) 96.
[3]. R. Srinivisan, R. J. DeAngelis and B. H. Davis, J. Catal. 106 (1987) 449.
[4]. Y. Zhou and S. M. Davis, Catal. Lett. 15 (1992) 51.
[5]. J. W. A. Sachtler and G. A. Somoqai, J. Catal. 81 (1983) 77.
[6]. W. M. H. Sachtler, Handbook of Catalysis, to be published.
[7]. W. Unger and D. Marton, Surf. Sci. 218 (1989) L467.
[8]. M. T. Paffett and R. G. Windham, Surf. Sci. 208 (1989) 34.
[9]. S. H. Overbury. D. R. Mullins, M. T. Paffett and B. E. Koel, Surf. Sci. 254 (1991)
45.
[10]. M. T. Paffett, A. D. Logan, R. J. Simonson and B. E. Koel, Surf. Sci. 250 (1991)
123.
[11]. Y. Li and B .E . Koel, Surf. Sci. 330 (1995) 193.
[12]. A. Atrei, U. Bardi, G. Rovida, M. Torrini, E. Zanazzi an P. N. Ross, Phys. Rev. B 46
(1992) 1649.
[13]. M. Galeotti, A. Atrei, U. Bardi, G. Rovida and M. Torrini, Surf. Sci. 313 (1994)
349.
[14]. M. Hoheisel, J. Kuntze, S. Speller, A. Postnikov, W. Heiland, I. Spolveri and U.
Bardi, to be published.
[15], M. T. Paffett, S. C Gebhard, R. G. Windham and B. E. Koel, Surf. Sci. 223 (1989)
449.
[16], C.Xu, B.E. Koel and M .T. Paffett, Langmuir 10(1994) 166.
[17]. C. Xu, Y.-L. Tsai and B. E. Koel, J. Phys. Chem. 98 (1994) 585.
[18], C. Xu and B. E. Koel, Surf. Sci. 304 (1994) 249.
[19]. (a)M. M. P. Janssen and J. Moolhuysen, Electrochim. Acta. 21(1976) 861.
(b)J. Catal. 46 (1977) 289.
[20], K. J. Cathro, J. Electrochem. Soc. 116 (1969) 1608.
[21], A. N. Haner and P. N. Ross, J. Phys. Chem., 95 (1991) 3740.
8
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
[22]. K. Wang, H. A. Gasteiger, N. M. Markovic and P. N. Ross, Jr, Electochimica Acta.
41 (1996)2587.
[23]. H. Hirano, T. Yamada, K. Tanaka, J. Siera and B. E. Nieuwenhuys, Vacuum 41
(1990) 134.
[24]. P. Cobden and B. E. Nieuwenhuys, Surf. Sci. 262 (1992) 97.
[25]. H. Ibach and S. Lehwald, J. Vac. Sci. Technol. 15 (1978) 407.
[26]. C. E. Megiris, P. Berlowitz, J. B. Butt and J. J. Kung, Surf. Sci. 159 (1985) 184.
[27]. T. E. Fischer, S. R. Kelemen and H. P. Bonzel, Surf. Sci. 64 (1977) 157.
[28]. C.X u, J.W . Peck and B.E. Koel, J. Am.Chem. Soc. 115 (1993)751.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
9
C hapter 2
Experimental Methods
2.1. Apparatus
The work described in the subsequent sections was performed in a two-level,
stainless steel, ultra high vacuum (UHV) chamber. A schematic drawing of the overall
system, in addition to a drawing for each level, are shown in Figs. 2.1 and 2.2. The base
pressure of the chamber was below 2 x 10' 1 0 Torr. The chamber was pumped by a 220-1/s
Perkin Elmer ion pump, a Ti-sublimation pump, and a 170-1/s turbomolecular pump
backed with an Alcatel® direct-drive mechanical roughing pump. The background
pressure and gas exposures were monitored by a nude ion gauge.
The upper level of the chamber was equipped with UTI 100C quadrupole mass
spectrometer (QMS) for TPD, double-pass cylindrical mirror analyzer (CMA) for AES
and XPS, dual X-ray source for XPS, four-grid optics for LEED, and an ion-sputtering
gun. The lower level contained a spectrometer for high resolution electron energy loss
spectroscopy (HREELS) using single 127°-cylindrical sectors in the monochromator and
analyzer. Facilities for gas dosing and metal evaporation were also located at each level
and in a row of ports between two levels. A stainless steel shield, with an aperture of about
1-cm dia. covered the ionizer region of the QMS to effectively attenuate the background
gas contribution in TPD signals when the crystal was positioned at a distance of about 1
10
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
mm from the aperture opening. In order to inhibit electron emission from the QMS
ionizer onto the sample during TPD, two high transparency, stainless steel screens were
used between the ionizer region and the sample. One was placed across the end of the
ionizer grid (with a bias of -55 V) and one was placed across the entrance aperture of the
shield (at ground potential). Sample currents of 6 .6 ,0.6 and 0.06 pi A were obtained using
no screens, only one screen across the ionizer grid, and both screens, respectively, with the
QMS operating at 2-mA-emission current and the sample at ground potential and
positioned in front of the entrance aperture. XPS was carried out by using 300 W Mg K a
radiation (hv = 1253.6 eV) with the CMA operated at a pass energy of 25-50 eV (AE =
0.4-0.8 eV). The AES measurements were carried out at 4 eV peak-to-peak modulation,
10 uA sample current, and E; = 3 kV. The HREELS spectra were taken with a typical
resolution of 70-80 cm '1 , count rate of 50-100 kcps, incident beam energy of 3-5 eV, and
incident current of 10'1 0 A.
The sample holder was made of a split copper block that was separated from a solid
copper block by a ,9”x .9”x 0.10” thick sapphire plate. This solid was silver-soldered
into the end of a long stainless steel tube filled with liquid nitrogen for sample cooling to
95 K. Sapphire provides high thermal conductivity for sample cooling along with low
electrical conductivity needed for resistive heating [1]. The Pt crystal was suspended
between two 10-mil Ta wires, which were attached to two Ta posts embedded into Cu
blocks. The temperature was measured by a chromel-alumel thermocouple that was spot-
welded direcdy to the side of the crystal.
The Pt crystal surface was cleaned in the chamber by a combination of Ar+ ion
sputtering (V = lkV, I = 5pA, t = 5 min) at 300 K, followed by annealing in vacuum at
1240 K for 30 sec and oxygen treatments (1 x 10'8 Torr O,) for five minutes with the
sample at 1000 K. The clean surface gives a (lx l) LEED pattern for Pt(l 11) surface and
11
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
a reconstructed (5x20) hexagonal LEED pattern for Pt(100) surface. The LEED pattern
for the clean hex-Pt(lOO) surface is shown in fig. 2.3.
Gas dosing utilized a microcapillary array gas doser attached to the end of a 1/8-inch
stainless steel tube that was connected to a Varian® variable leak valve. The microcapillary
array (500-u. thickness and 10-u. diameter holes) collimates the gas flow, which results in
an increased gas flux at the crystal surface over that predicted by the background chamber
pressure. Gas exposures are reported in Langmuirs ( 10"6 torr.s) whenever possible. A
correction factor for the doser enhancement factor was obtained by constructing uptake
curves from both background gas exposures and doser. The ion gauge sensitivity factor
was also taken into account. This procedure is described in the appendix A.
2.2. Preparation of (111) and ( 100) Sn/Pt surface alloys
The Sn/Pt surface alloys were produced by evaporating ~0.7-1.2 monolayer of Sn
on clean Pt surfaces at 300 K followed by annealing to different temperatures to obtain
Sn/Pt(l 11) and Sn/Pt(100) surface alloys [2-7].
The Sn doser was made by using an enclosed "boat" made from 0.13 mm Ta foil
containing a Sn ingot (6N purity). This doser was outgassed before use by warming it for
at least 5 min to 300 K. This boat was resistively heated at a constant current that caused it
to turn an orange-red color (~1000 K), however, no careful temperature measurements
were made. Sn was deposited on the Pt crystal, which was held at room temperature, and
the coverage was monitored by AES. Values for the Sn(430)/Pt(237 eV) peak-to-peak
height ratio were measured at 3 kV incident beam energy and 4 Vp-p modulation voltage.
AES was always taken before and after annealing rather than simply recording the applied
current and time for operating the Sn doser. This procedure was more reliable given the
12
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
variations in Sn doser designs, doser operating temperatures and the distance between the
Sn doser and crystal.
The (2x2)Sn/Pt(lll) and (V3xV3)R30°-Sn/Pt(l 11) surface alloys were prepared
following the procedure used by Paffett and Windham [2] via evaporation of ~0.7-0.9 ML
of Sn onto a Pt(l 11) single crystal at 300 K (with the smaller amount used for the (2x2)
surface alloy) and then annealing the crystal at 1000 K for 10 s. The general conditions
used to prepare the two Sn/Pt(l 11) alloys is described in table 2.1.
Two S n /P t(lll) alloys with stoichiometry of Pt3 Sn and Pt,Sn, respectively are
shown in fig 2.4. Both of these alloys have been extensively studied by low energy ion
scattering spectroscopy [6], LEED I-V [8] calculations and X-ray forward scattering [9].
These surface alloys are quite “flat”, with an outward buckling distance for Sn (above the
Pt surface plane) of ~ 0.20 A.
On Pt(100), two different alloyed surfaces can be formed by a similar procedure.
These two alloys are denoted as the c(2x2) and the (3Vr 2x‘ /2)R45° Sn/Pt(100) surface
alloys with 0S n= 0.50 and 0.67, respectively [3,7]. These surface alloys could be prepared
in a reproducible manner. After dosing Sn on the surface the Sn(430)/Pt(237 eV) peak-
to-peak height ratio in AES was measured before and after annealing the surface.
The c(2x2) and p(3/2xV2)R45° Sn/Pt(100) surface alloys were prepared by first
evaporating a thin Sn film onto the clean hex-Pt(100) surface at 300 K and then annealing
to 750 K or 900 K, respectively, as described by Paffett et al [3]. The LEED pattern for
the c(2x2) and p(3-/2xV2)R450 Sn/Pt(100) surface alloys are shown in fig. 2.5. The
observed c(2x2) LEED pattern following Sn deposition on Pt(100) can be due to two
different surface structures [3,7]. An overlayer of Sn adatoms is formed over a relatively
wide range of temperatures below 750 K. This Sn adlayer is ordered for T = 500-750 K.
Formation of the c(2x2) alloy occurs over a narrow range of 750-800 K, depending on the
13
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
initial Sn concentration. The c(2x2)Sn/Pt(100) overlayer and surface alloy are shown in
Fig. 2.6. A more thermally stable alloy with a (3V2xV2)R45° structure is formed at higher
annealing temperatures of about 800 K and is stable up to 1050 K.
The c(2x2) alloy is comprised of 0.5 ML of Sn incorporated into the surface plane
of the Pt(100) sample to form a “checker board” pattern. Each Pt atom is isolated from
nearest neighbor Pt contacts and surrounded by all Sn neighbors, and vice versa [3, 7].
The alloy surface is quite “flat”, with an outward buckling distance for the Sn of ~ 0.20 A
above the surface Pt atom plane [7]. The (3Vr 2xV2)R45° (0S n = 0.67 ML) LEED pattern
suggests a periodic surface reconstruction with Sn atoms occupying the domain
boundaries [7]. Studies with low energy alkali ion scattering spectroscopy (ALISS)
suggested that the (3V2xV’ 2)R45° alloy has nearly the same local geometric structure as
the c(2x2) alloy and is composed of small c(2x2) alloy domains with the same buckling
distance within the domains. However, the structure of this surface is still undetermined.
Recent STM images, of a related bulk Pt3 Sn(100) alloy surface after sputtering and
annealing to 600 K showed pyramidal features consisting of (102)-facets [10]. Flat
portions between and on top of the pyramids formed by the facets are mainly comprised
of triple rows that are probably Pt. A c(2x2) LEED pattern with streaky facet spots was
observed after heating to 600 K and annealing to 1000 K leads to an improved c(2x2)
LEED pattern with no facet spots. STM studies in our lab are underway to further probe
the structure of the (3Vr 2xV2)R45° Sn/Pt(100) alloy and any relationship that it has to that
of the structure of the bulk Pt,Sn(100) surface. For brevity, the c(2x2)Sn/Pt(100) and the
p(3Vr 2x\72)R45° Sn/Pt(100) alloys will be referred as the c(2x2) and 3V2 alloys,
respectively, throughout this thesis.
14
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
REFERENCES
[1]. R. G. Windham in: The Chemistry of Ethylene on Bimetallic Pt(l 11) Surfaces, Ph.
D. thesis 1988.
[2]. M. T. Paffett and R. G. Windham, Surf. Sci., 208 (1989) 34.
[3]. M. T. Paffett, A. D. Logan, R. J. Simonson and B. E. Koel, Surf. Sci. 250 (1991)
123.
[4]. Y.-L. Tsai, in: Surface Science Studies of Thermal and Electron-Induced Reactions
on Metal and Semiconductor Surfaces, Ph. D. Thesis 1996.
[5]. N. A. Saliba, in: Oxidation of Pt and Sn/Pt Alloy Surfaces and the Thermal Stability
of the Oxides Produced, Ph. D. Theses 1999.
[6]. S. H. Overbury, D. R. Mullins. M. T. Paffett and B. E. Koel, Surf. Sci., 254 (1991)
45.
[7]. Y. Li and B. E. Koel, Surf. Sci. 330 (1995) 193.
[8]. A. Atrei, U. Bardi, G. Rovida, M. Torrini, E. Zanazzi and P. N. Ross, Phys. Rev., B
46 (1992) 1649.
[9]. M. Galeotti, A. Atrei, U. Bardi, G. Rovida and M. Torrini, Surf. Sci., 313 (1994)
349.
[10]. Hoheisel, M.; Kuntze, J.; Speller, S.; Postnikov, A.; Heiland, W.; Spolveri, I.; Bardi,
U. to be published.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
D t
O
Turbo Pump
g n . C M A
g HREELS
!cn
Ion Pump
Fig. 2. 1 Schematic drawing of the UHV chamber used in this work
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
XPS
LEED
Turbo Pump
CMA
QMS
Variable 5 °-3 0 °
,25°
HREELS
Fig. 2.2 Arrangement of the upper (top) and lower (bottom) levels of the UHV chamber.
17
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Fig. 2. 3 LEED image of the reconstructed hex-Pt(lOO) surface taken at a beam energy of
60 eV.
18
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
p(2x2)
Pt3S n (lll) 6Sn = 0.25
(V3xV3)R30° Pt2Sn ®Sn 1 1 3
Fig. 2 .4 Schematic top-view of the structure of two Sn/P t(lll) surface alloys. Top:
(2x2)Sn/Pt(l 11) alloy. Bottom: (V3xV3)R30° Sn/Pt(l 11) alloy.
19
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Fig. 2.5 LEED image of the c(2x2) (top) and (3V r 2xV2)R45° (bottom) Sn/Pt(100)
surface alloys taken at a beam energy of 60 eV.
20
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
c(2x2) Sn Overlayer on Pt(100) ©Sn = ML
c(2x2) Sn Alloy on Pt(100) 0 S n = 0-5 ML
Fig. 2 .4 Schematic top-view of the structure of two Sn/Pt(100) surfaces. Top:
c(2x2)Sn/Pt(100) overlayer. Bottom: c(2x2)Sn/Pt(100) alloy.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Table 2.1 Pt-Sn alloy preparation conditions
surface LEED pattern
Annealing
temp, tune
Sn(430)/Pt(237) AES ratio
Before annealing After annealing
Pt( 111) (2x2) 1000 K,10s 4-5 1.7-1.8
(V3xV3)R306 1000 K, 10s >8 2.4-2.6
Pt(100) c(2x2) adlayer 500 K,30s 10-11 8-9
c(2x2) 750 K,30s 10-11 7-8
(3V2xV2)R45° 850 K ,30 s 10-11 3.5-4.5
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Chapter 3
Adsorption of Methanol, Ethanol and Water on Well Characterized
Pt-Sn Surface Alloys
Abstract
Adsorption and desorption of methanol (CH3 OH), ethanol (C2 H5 OH), and water on
P t( lll) and two, ordered Pt-Sn alloys has been studied using primarily temperature
programmed desorption (TPD) mass spectroscopy. The two alloys studied were the
p(2x2)Sn/Pt(lll) and (/3xVr 3)R30° S n /P t(lll) surface alloys prepared by vapor
deposition of Sn on P t(lll), with 0S n = 0.25 and 0.33, respectively. All three molecules
are weakly bonded and reversibly adsorbed under UHV conditions on all three surfaces,
molecularly desorbing during TPD without any decomposition. The two Pt-Sn surface
alloys were found to chemisorb both methanol and ethanol slightly more weakly than on
the P t(lll) surface. The desorption activation energies measured by TPD, and hence the
adsorption energies, of both methanol and ethanol progressively decrease compared to
Pt(l 11) as the surface concentration of Sn increases. The decreased binding energy leads
one to expect a lower reactivity for these alcohols on the two alloys. The sticking
coefficients and the monolayer coverages of these alcohols on the two alloys were identical
to that on P t(lll) at 100 K, independent of the amount of Sn present in the surface layer.
23
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Alloying Sn in P t(lll) also slightly weakens the adsorption energy of water. Water
clusters are formed even at low coverages on all three surfaces, eventually forming a water
bilayer prior to the formation of a condensed ice phase. These results are relevant to a
molecular level explanation for the reactivity of Sn-promoted Pt surfaces that have been
used in the electrooxidation of simple organic molecules.
3.1. Introduction
The use of simple organic molecules such as methanol, ethanol, formic acid, and
formaldehyde as future electrochemical fuels has several advantages. In addition to their
high energy density [1] they are relatively nontoxic and easy to store and handle.
Advanced fuel cells using methanol are already being developed [2]. Crucial to the
operation of these electrochemical systems is the interaction of the fuel with the electrode
surface.
The oxidation of methanol on different metal electrodes has been studied [3-8]. The
principle limitation in using an electrode (electrocatalyst) with a viable organic fuel is the
poisoning of the electrode by some intermediate and/or product produced by the reaction.
Identified poisons are adsorbed hydrogen (Ha d s ), formyl species (CHOa d s ), and carbon
monoxide (COa d s ) [9-10]. The most active catalysts for methanol oxidation are defined as
those having a low surface concentration of all poisoning species. These are either
platinum catalysts promoted by electrodeposition of certain metals [11] or alloys of
platinum such as Pt-Ru [12-13, 19] or Pt-Sn [13-19]. For the Pt-Sn systems, there are
controversies about the state of Sn in these catalysts and the activity of such electrodes.
Most reports [13-16] are that Pt-Sn appears to be the most active catalyst in a sulfuric
acid solution at a temperature above 40° C. In contrast, others report either an inhibition or
an activity comparable to that of pure Pt [17-19]. Haner et. al [17] studied the
24
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
electrooxidation of methanol on Pt-Sn alloys and found no Pt-Sn alloy of any
composition that was more active than pure platinum. While the use of ethanol in a fuel
cell has been considered [20-21], the oxidation of ethanol on different metal electrodes is
much less studied and ethanol is still of primarily academic interest due to the addition of
one carbon atom that implies more intermediates that could possibly poison the electrode
surface.
Surface science studies in UHV of methanol and ethanol adsorption on metal
surfaces such as, Pd(l 11), Pt(111), and Pt(l 10) have been carried out [22-24], Methanol
decomposes on Pd( 111) to form CO and H2 , while ethanol undergoes C-C bond cleavage
to form CO, H, and methane (CH4 ) [22]. Methanol and ethanol weakly, reversibly adsorb
and desorb molecularly from Pt(l 11) [25-26]. Indeed, the first four (C,-C4 ) alcohols all
have small heats of adsorption (11-15 kcal/mole) on P t(lll) [25]. The small amount of
decomposition (10%) into CO, H,and G ^, in early reports concerning adsorbed alcohols
on Pt(l 11) was eventually shown by Dubois et al [26] to be due to defect sites present at
the surface. An outstanding question is whether or not the presence of Sn in the surface
layer of Pt-Sn alloys will alter the chemistry of methanol and ethanol on the surface.
The interaction of water with transition metal surfaces is also an important topic for
discussion of the electrode chemistry in electrooxidation of alcohols. Because water/metal
interactions are so ubiquitous for practical as well as fundamental considerations in many
disciplines including corrosion, electrochemistry, and catalysis, the adsorption of water on
transition metals has been extensively investigated in the past and the platinum surface has
been of special interest in heterogeneous catalysis and electrochemistry. A nice review of
water adsorption has been given by Thiel and Madey [27]. On P t(lll) at 110 K, water
adsorbs nondissociatively with unit sticking probability independent of coverage [28].
Even at the lowest coverages, water desorbs in two peaks, 196 and 178 K, and at higher
25
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
coverages there are three distinct physisorbed peaks 160-167 K, 170-171 K and 177-180
K ascribed to multilayer ice, a bilayer region and a non-bilayer region [28, 29]. Hydrogen
bonded clusters are present even at very low coverages. From photoelectron [30] and
vibrational spectra [31, 32], it was concluded that the “monolayer” consists of a two-
tiered, 3D structure, called the bilayer. The structure of the bilayer is such that the bottom
half of the water molecules are directly bonded to the surface through the oxygen atoms
and the top half of the water molecules are held in the structure by two or three hydrogen
bonds to the lower molecules. The high temperature, “chemisorbed” state of water
corresponding to the desorption peak at 185-196 K on Pt(l 11) surfaces has been assigned
to a surface recombination reaction of coadsorbed hydroxyls that are formed at defects or
from dissociation induced by preadsorbed oxygen (from the background) [28,29].
In order to aid discussion and resolution of the controversies discussed above and to
provide a firm foundation for understanding chemistry relevant to electrooxidation of
alcohols over Sn-promoted Pt electrodes, it is important to define the interactions of simple
alcohols with well-defined Pt-Sn alloy surfaces. In our present study we have investigated
the interaction of methanol, ethanol, and water with Pt-Sn surfaces under UHV conditions
by temperature programmed desorption (TPD). The objective of this study was to
determine the adsorption energies of these molecules on P t( lll) and two different
bimetallic surfaces, the (2x2) and (V r 3x1 /3)R30o Sn/P t(lll) surface alloys. Specifically,
we wanted to evaluate whether the presence of Sn in the surface layer leads to an increase
in adsorption energy or thermally activates these molecules for reaction due to the
thermodynamic driving force provided by the Sn-O interaction.
3.2. Experimental Methods
The experiments were conducted in an ion-pumped stainless steel vacuum chamber
(base pressure lxlO '1 0 torr), equipped with low energy electron diffraction (LEED), Auger
26
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
electron spectroscopy (AES) using a double-pass cylindrical mirror analyzer, a shielded
UTI 100C quadruple mass spectrometer (QMS) for temperature programmed
spectroscopy (TPD), ion gun for sputtering, and gas and metal dosing facilities.
The Pt( 111) crystal could be heated resistively to 1100 K or cooled to 100 K using
liquid nitrogen. A chromel/alumel thermocouple was spot welded to the crystal to monitor
the temperature. The crystal was cleaned by repeated cycles of Ar+ ion bombardment,
annealing in vacuum at 1100 K, and heating in 5xl0"8 torr O, at 800 K. The cleanliness
and long range order of all surfaces were checked using AES and LEED prior to each
experiment.
Methanol (Mallinckrodt chemical, 99.9 %), Ethanol (Quantum chemical corporation,
99.9 % ), and deionized water were placed in a glass reservoir attached to a stainless steel
dosing line, and used as supplied after degassing by multiple freeze-pump-thaw cycles.
These gases were exposed to the Pt crystal by a microcapillary array doser connected to
the gas line through a leak valve. For water, some background dosing was also used for
small exposures, in addition to microcapillary dosing. All of the exposures listed in this
paper are given simply in terms of the dosing time for a fixed dosing pressure (about
4xlO'1 0 torr in the background); no attempt has been made to correct for flux enhancement
of the doser or ion gauge sensitvitiy. The mass spectrometer
in the chamber was used to check the purity of the gases during dosing. For all of the
TPD experiments, the heating rate was ~ 4 K/s.
The (2x2) Sn/Pt(l 11) and (•/3xvr3)R30° Sn/Pt(l 11) surface alloys were prepared by
evaporating several monolayers of Sn onto the P t(lll) crystal surface and subsequently
annealing the sample to 1000 K for 10 sec. Depending on the initial, deposited Sn
coverage, the annealed surface exhibits either a (2x2) or (V3xV3)R30° S n/P t(lll)
structure as observed by LEED [33]. These LEED patterns for the surfaces prepared as
27
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
above are due to substitutional surface alloys with 0S n = 0.25, corresponding to the (111)
face of PtjSn, and 0S n = 033, corresponding to a Pt,Sn surface, in which the Sn atoms
protrude 0.02 nm above the surface Pt plane [34]. For the (2x2) structure, 3-fold Pt sites
are present but no adjacent 3-fold Pt sites are present. All 3-fold sites comprised of only
Pt are eliminated for the V 3 structure and the distance between adjacent 2-fold sites is
increased. For brevity throughout this paper, we will refer to the p(2x2)Sn/Pt(l 11) and
(V3xv3)R30° Sn/Pt(l 11) surface alloys as the (2x2) and V 3 alloy surfaces, respectively.
3.3. Results and discussion
33.1. Methanol and ethanol adsorption
A series of TPD spectra for methanol (CH3 OH) desorption from Pt( ill) and the
(2x2) and V 3 alloys are shown in Figs. 3.1, 3.2 and 33, respectively. All methanol
exposures were given with the surface temperature at 95-100 K. In each case, a clear
separation occurs between a high temperature peak due to a chemisorbed state and a low
temperature peak arising from desorption from a condensed, physisorbed layer. With
increasing coverage in the monolayer, a small shift to lower temperatures is seen for the
chemisorption peak. The multilayer, or condensed phase, peak formed at larger exposures
occurs at 145 K for all the three surfaces. On Pt(lll) at relatively low exposures, a
desorption peak at 194 K is observed from the monolayer, and this peak shifts only
slightly to lower temperature (183 K) at saturation coverage of the monolayer, presumably
due to lateral interactions between methanol molecules. Alloying Pt with Sn lowers the
desorption temperature for low coverages of methanol in the monolayer from 194 K on
Pt(lll) to 186 K and 179 K on the (2x2) and \3 alloy surfaces, respectively. At
saturation coverage in the monolayer, the desorption peak is reduced from 183 K on
Pt(l 11) to 177 K and 170 K on (2x2) and V3 alloy surfaces, respectively. Because of the
very small shifts in the TPD peak maxima with increasing methanol coverage for these
28
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
three surfaces, we have attributed first order kinetics to the methanol desorption from all of
these surfaces.
Evidence for CO and H, evolution from methanol decomposition was monitored
during TPD from all three surfaces following methanol exposure. No appreciable CO or
H2 desorption was detected. Estimation of the maximum amount of decomposition using
these CO and H2 peak areas yield in each case decomposition amounts < 5%, and we
attribute this to contributions from defect sites and coadsorption of impurities from the
background gases. Consistent with these TPD results, no carbon or oxygen was detected
by AES following TPD.
In Figs. 3.4, 3.5 and 3.6 we show a series of TPD spectra for ethanol (C2 H5 OH)
desorption following ethanol dosing on Pt(lll) and the (2x2) and V3 alloy surfaces,
respectively, at 95-100 E C . These TPD spectra are quite analogous to those for methanol
desorption, except that the desorption temperature of ethanol is higher than methanol on all
the three surfaces at all coverages due to higher molecular weight and larger size of ethanol
which leads to stronger adsorption and larger condensation energies on the surface. In
each case, a distinct monolayer desorption peak can be observed, and this peak shifts
slightly to lower temperatures with increasing coverage in the monolayer. The ethanol
multilayer formed for larger exposures has a desorption peak temperature of 155 E C from
all the three surfaces for nearly identical exposures. On Pt(lll), the desorption peak
maximum for a low coverage of ethanol in the monolayer occurs at 213 K, but shifts to
202 K at saturation coverage in the monolayer due to lateral interactions. The desorption
peak temperature of ethanol in the monolayer progressively decreases with increasing Sn
concentration in the surface alloy. At relatively low ethanol coverages, desorption occurs
in peaks at 213, 206, and 197 K on Pt(lll) and the (2x2) and V3 alloy surfaces,
respectively, and peaks occur at 202, 200 and 190 K for monolayer saturation coverages
29
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
on these three surfaces, respectively. The appearance of the ethanol TPD spectra on all the
three surfaces also suggests first order desorption kinetics.
The adsorption / desorption behavior of ethanol was found to be entirely reversible
with no evidence of decomposition, even though ethanol is expected to have a higher
reactivity than methanol due to the presence of (3-hydrogen atoms.
Uptake curves showing the adsorption kinetics and relative saturation coverages for
methanol and ethanol chemisorbed on the three surfaces are shown in Figs. 3.7 (a) and
(b), respectively. The slopes of these uptake curves are proportional to the sticking
coefficients, S, of these molecules on these surfaces. First, we find that S=1 for
population of the chemisorbed state on these surfaces at 100 K by comparison to the
uptake into the condensed multilayers at higher coverages (not shown) and a knowledge
that S=1 for condensation of many hydrocarbons [35] and water [29]. Secondly, we find
that the value of the initial sticking coefficient at "zero" coverage, S0, is maintained
throughout population of the monolayer, i.e., constant sticking coefficient, for adsorption
on these surfaces at 100 K. We find that the value of S0 (and S) is not affected by the
presence of Sn in the surface layer. Thus, alloyed Sn does not effectively decrease the
adsorption rate constant of these two alcohols on Pt-Sn alloys compared to the Pt(lll)
surface up to 0S n = 0.33. Even if Sn is not directly involved in the bonding of these
molecules, this would be consistent with the important influence of a precursor state
present at the Sn sites that we have previously discussed as the modifier precursor state
[36]. These curves also show that the same monolayer coverage is obtained for methanol
and ethanol chemisorbed on the two Pt-Sn surface alloys compared to that on Pt(lll).
We find that the chemisorbed methanol and ethanol monolayer coverages on Pt( 111) and
the (2x2) and V3 surface alloys are independent of the alloyed Sn concentrations in these
three surfaces. Thus, alloyed Sn does not effectively inhibit access to the surface for the
30
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
adsorption of these two alcohols on Pt-Sn alloys compared to the Pt(lll) surface up to
0S n = 0.33. This suggests that there is only a small ensemble of a few Pt atoms required
for chemisorption.
A comparison of the desorption peaks in TPD for relatively low initial surface
coverages of about 10% of a saturation monolayer coverage (to avoid lateral interactions)
of methanol and ethanol on Pt(l 11) and the (2x2) and V3 alloy surfaces is shown in Fig
3.8. These spectra show clearly the influence of alloyed, surface Sn in reducing the
desorption temperatures to progressively lower values as the surface concentration of Sn
increases in the surface alloy. A decrease in peak temperature indicates slightly weaker
adsorbate-surface bonding, and this effect can be quantified to give desorption activation
energies in several ways. The simplest method is an analysis of the desorption peak
maxima using Redhead [37] analysis, assuming a preexponential factor of 101 3 sec'1 and
first order kinetics. These results are given in Table 3.1. A method that does not require
an assumption of the preexponential factor or kinetic order, and is often more accurate, is
derived from making an Arrhenius plot from the desorption rate curves. This is also
sometimes referred to as leading edge analysis [38,39], These Arrhenius plots of ln(rate)
versus 1/T, where the desorption rate is proportional to the TPD peak intensity, gave
curves that were excellently described by straight lines for temperatures below the peak
maximum. The slopes (slope=Ea/R) of these lines give the desorption activation energies
for methanol and ethanol on the three surfaces, and these are reported in Table 3.1. The
Redhead method and the Arrhenius plots give consistent results within + 6 kJ/mole. Since
the molecular adsorption of these alcohols is not activated on these surfaces, the
desorption energy is equal to the adsorption energy. The presence of alloyed Sn in Pt-Sn
surfaces does not have a large effect on the methanol and ethanol heats of adsorption, but
31
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
does slightly decrease the adsorption energy of methanol and ethanol on these surfaces in
comparison to Pt(l 11).
To summarize methanol and ethanol interactions on Pt-Sn alloys, alloyed Sn does
not have a large influence on Pt(lll) chemistry. We note that ultraviolet photoelectron
spectroscopy (UPS) has shown that alloying causes only small changes in the Pt d-DOS
and in particular the DOS near Ef [40]. However, because of the stronger Sn-O bond
compared to the Pt-O bond and the thermodynamic driving force for forming tin oxide, it
is possible that Sn would increase the reactivity of the clean Pt(lll) surface, either
increasing the adsorption energies of alcohols or lowering the activation barrier to
dissociative adsorption. This intuitive idea does not correctly predict our results. We
observed a decrease in the adsorption energies because Sn is not directly involved in the
bonding of these molecules to the surface and the subtle localization of charge due to Pt-
Sn bonding weakens the adsorbate-surface interaction. Our values for the adsorption
energetics should be quite useful in calculating surface coverages and residence times
under a variety of conditions other than those of UHV. The intuitively expected increase
in alcohol reactivity due to Sn was also not observed. We know from our desorption
measurements that the activation barriers to dissociate methanol or ethanol on these alloy
surfaces certainly exceed 41-54 kJ/mol. It is reasonable to assume that the barriers to
dissociation exceed those on Pt(lll) since the adsorption energies are smaller. Our data
are consistent with the more recent results for electrooxidation of methanol [17-19] that
show that Pt-Sn alloys are not more active than pure Pt.
33.2. Water adsorption
Fig. 3.9. shows the TPD spectra after H2 0 was dosed on clean Pt(lll) at 100 K.
The inset shows additional spectra for very low exposures. Even for these very low
exposures, two peaks appear, one near 185 K and a second peak at 163 K, similar to that
32
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
observed previously [28, 29]. The feature at 185 K is always small and saturates at
relatively low exposures. This broad peak has been assigned to a surface recombination
reaction of coadsorbed hydroxyls that are formed at defects or from dissociation induced
by preadsorbed oxygen (from the background) [28, 29]. The formation of the 163 K
peak, which shifts to 167 K near saturation coverage, is attributed to water island or
cluster formation due to hydrogen bonding even at very low coverages, because the energy
associated with hydrogen bonding is comparable or greater than the water-surface
interaction. We did not resolve a second peak in the bilayer region like Jo et al. [29]
observed, and this is most likely due to a small amount (~ 5% ML) of preadsorbed
hydrogen from background contamination on the Pt(lll) surface [29]. Finally, another
peak arises at 161 K which shifts to higher temperature with increasing exposure and does
not saturate. One “monolayer” has been defined as the coverage corresponding to the
highest exposure which did not give a 161 K desorption peak. This monolayer saturation
coverage of water on Pt(lll) corresponds to a bilayer with 0.67 H2 0 molecules per Pt
atom [27].
On both the (2x2) and V3 surface alloys, no high temperature “chemisorption”
peak was found, as shown in Figs. 3.10 and 3.11. Immediately at the lowest exposures, a
H2 0 TPD peak arises at about 156 K for (2x2) alloy and at 152 K for V3 alloy and shifts
to higher temperatures (165-166 K) with increasing coverages. These peak shapes are
appear like those for zero order kinetics. First, alloying with Sn completely poisons the
Pt(lll) surface for the high temperature “chemisorption” peak of water. This could be
due to titration of defect sites by alloyed Sn and/or the fact that these Pt-Sn alloys do not
dissociatively chemisorb 0 2 [41]. Secondly, the peak that appears at 166-165 K for both
of the alloyed surfaces is identical to that (167 K) for the Pt(l 11) surface, and the overall
behavior of this peak with increasing coverage is also nearly the same. This desorption
33
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
peak is certainly due to island formation from hydrogen-bonded water clusters, and
furthermore we propose that this peak arises from the water bilayer on the alloy surfaces.
A comparison o f the areas of the 165-167 K peaks at saturation coverage in these states on
all three surfaces gives the same values. Hence, the “monolayer” of water that forms on
Pt(lll) with a bilayer structure also forms on the two alloys, and the kinetics and
energetics of the desorption of this bilayer and subsequent multilayers is identical on all
three surfaces. Evidently the presence of alloyed Sn in the surface layer of the Pt-Sn alloy
has a similar effect to that observed for preadsorbed hydrogen or oxygen [29] on Pt(l 11)
in that some orientation or ordering requirement can not be satisfied on the alloy surface to
cause the characteristic “two peak” desorption of the bilayer structure. (The
preadsorption of contaminant hydrogen or oxygen can no longer be an explanation for the
merging of these two peaks, like on Pt(lll) in Fig. 3.9, since the two alloys do not
dissociatively adsorb H ,or O, [41].)
In order to probe any subtle differences in the direct bonding interactions of water
with these three surfaces, we show desorption spectra in Fig. 3.12 for initial water
coverages that are about 10% of the saturation monolayer coverage. The main TPD peak
decreases from 163 K on Pt(lll) to 156 K on (2x2) alloy to 152 K on V3 alloy. The
results from the Arrhenius plots from a leading edge analysis are summarized in Table 3.1
along with those obtained by using Redhead analysis. With increasing Sn concentration
in the alloys, the bonding of water to the surface is reduced slightly. Regarding the
reactivity of these surfaces, the activation barrier for dissociative adsorption of water
exceeds 38-44 kJ/mol on these three surfaces.
Fig. 3.13 illustrates and compares the influence of increasing alloyed Sn
concentration on the adsorption energetics of methanol, ethanol and water on Pt(111) and
the (2x2) and V3 alloy surfaces. The molecular desorption peak temperatures are shown
34
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
for desorption at relatively low coverages of about 10% of the monolayer saturation
coverage in order to reduce the influence of lateral interactions in the adsorbed layer. In
addition, the right hand axis in Fig. 3.13 approximately indicates the corresponding
desorption activation energies (this is an approximate guide since there is not strictly a
linear relationship between the peak temperature and the activation energy). As we have
discussed, alloying Sn with Pt(l 11) causes in all cases a weak reduction in the adsorbate-
surface bonding, presumably due to increased localization of Pt orbitals resulting from Pt-
Sn bonding interactions. The decrease in the desorption temperature is not linear with an
increase in the surface Sn concentration for any of the three molecules studied, the
decrease in adsorption energy being relatively larger for the V3 alloy than the (2x2) alloy
for all of the molecules. We have observed a similar behavior for several alkenes [42] on
these alloys, and we interpreted this as due to the absence of pure-Pt 3-fold sites on the V3
alloy and the importance of these sites in chemisorption bonding. Because these
molecules are so weakly bonded to the surfaces, such a connection is more tenuous. The
smaller influence of alloying on the bonding of water on the (2x2) and a/3 alloys is
justifiable due to the relatively smaller role played by the substrate surface in the
adsorption energy because of the much larger importance of hydrogen bonding in the
water overlayer compared to that for methanol and ethanol.
3 5 . Conclusions
Methanol and ethanol molecularly desorb from Pt(lll) and the (2x2) and
(V3xV3)R30° Sn/Pt(lll) surface alloys without any decomposition under UHV
conditions. Both alloy surfaces chemisorb these molecules slightly more weakly than the
clean Pt(lll) surface. The adsorption rate constant, i.e., the sticking coefficient, and the
monolayer coverage of methanol and ethanol are uneffected by alloying Sn into the Pt
surface. The later observation suggests that an ensemble size of only a few Pt atoms is
35
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
required for chemisorption. Also the activation barriers for dissociation of methanol and
ethanol exceed 41-54 kJ/mol on these Pt-Sn alloys and chemisorption of these alcohols
does not oxidize the Sn in the alloy under UHV conditions.
Water is only weakly adsorbed on all three surfaces studied, and a thermally
stabilized bilayer structure is formed on the two alloy surfaces similarly to Pt(lll). The
desorption temperature and desorption activation energy of this bilayer and subsequent
condensed layers are unaffected by alloying, even though there is evidence that at very low
coverages there is a small weakening of the direct interaction of water molecules with the
surface with increasing Sn concentration.
In general, for all three molecules, a trend is observed for a small decrease in the
desorption activation energies with increasing Sn concentration in the series Pt(111), (2x2)
alloy, and V3 alloy surfaces, with a relatively larger influence upon forming the V3 alloy.
This is consistent with recent results that show that Pt-Sn alloys are not more active than
pure Pt for electrooxidation of methanol [17-19]. Our data provide benchmarks for
discussing the surface chemistry of alcohols on Pt-Sn alloys, opening the road for an
improved understanding of electrooxidation and catalysis of alcohols and other
oxygenated molecules on bimetallic Pt-Sn catalysts.
Acknowledgment
This work was supported by the Division of Chemical Sciences, Office of Basic
Energy Sciences, U.S. Department of Energy.
36
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
REFERENCES
[1]. C. Lamy, Electrochim. Acta. 29 (1984) 1581.
[2], S. Surampudi, S. R. Narayanan, E. Vamos, H. Frank, G. Halpert, A. La Conti, J.
Kosek, G. K. Surya Prakash and G. A. Olah, J. Power Sources 47 (1994) 377.
[3]. E. M. Belgsir, H. Huser, J. M. Leger and C. Lamy, J. Electroanal. Chem. 225 (1987)
281.
[4]. F. Kadigam, B. Beden, J. M. Leger and C. Lamy, J. Electroanal. Chem. 1 (1981) 89.
[5]. M. Shibata and S. Motoo, J. Electroanal. Chem. 229 (1987) 385.
[6]. K. Kunimatsu, J. Electroanal. Chem. 213 (1986) 149.
[7]. P. Olivi, B. Beden, F. Hahn, J. M. Leger and C. Lamy, J. Electroanal. Chem. 346
(1993)415.
[8]. B. Beden, S. Juanto, J. M. Leger and C. Lamy, J. Electroanal. Chem. 238 (1987) 323.
[9]. T. Iwasita, W. Vielstich and E. Santos, J. Electroanal. Chem. 229 (1987) 367.
[10]. R. Parsons and T Vandemoot, J. Electroanal. Chem. 257 (1988) 9.
[11]. R. Adzic, In Advances in Electrochemistry and electrochemistry Engineering,
Gerischer, H. Ed. Wiley-Interscience: New York, 1984, vol 13 pp. 159-260.
[12]. M. M. P. Janssen and J. Moolhuysen, Electrochim. Acta. 21 (1976) 869.
[13]. (a)M. M. P. Janssen and J. Moolhuysen, Electrochim. Acta. 21(1976) 861.
(b)J. Catal. 46 (1977) 289.
[14]. K. J. Cathro, J. Electrochem. Soc. 116(1969) 1608.
[15]. S. Szabo, J. Electroanal. Chem. 172 (1984) 359.
[16]. S. Gilman and M. W. Breiter, J. Electrochem. Soc. 109 (1962) 1099.
[17]. A. N. Haner and P. N. Ross, J. Phys. Chem. 95 (1991) 3740.
[18]. S. A. Campbell and R. Parsons, J. Chem. Soc. Faraday Trans. 88(6), (1992) 833.
[19]. K. Wang, H. A. Gasteiger, N. M. Markovic and P. N. Ross, Jr, Electochim. Acta. 41
(1996)2587.
[20]. J. Willsam and J. Heitbaum, Electrochim. Acta. 31 (1986) 943.
[21]. B. Beden, M. C. Morin, F Hahn and C. Lamy, J. Electroanal. Chem. 229 (1987) 353.
37
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
[22]. J. L. Davis and M. A. Barteau, Surf. Sci. 187 (1987) 38S.
[23]. S. Akhter and J. M. White, Surf. Sci. 167 (1986) 101.
[24]. J. Wang and R.I. Masel, Surf. Sci. 243 (1991) 199.
[25]. B. A. Sexton, K. D. Rendulic and A. E. Hughes, Surf. Sci. 121 (1982) 181.
[26]. K. D. Gibson and L. H. Dubois, Surf. Sci. 223 (1990) 59.
[27]. P. A. Thiel and T. E. Madey, Surf. Sci. Rep. 7 (1987) 211.
[28]. G. B. Fisher and J. L. Gland, Surf. Sci. 94 (1980) 446.
[29]. S. K. Jo, J. A. Polanco and J. M. White, Surf. Sci. 253 (1991) 233.
[30]. E. Langenbach, A. SpitzerandH. Luth, Surf. Sci. 147 (1984) 179.
[31]. H. Ibach and S. Lehwald, Surf. Sci. 91 (1980) 187.
[32]. F. T. Wagner and T. E. Moylan, Surf. Sci. 191 (1987) 121.
[33]. M. T. Paffett and R. G. Windham, surf. Sci. 208 (1989) 34.
[34]. S. H. Overbury, D. R. Mullins, M. T. Paffett and B. E. Koel, Surf. Sci. 313 (1994)
349.
[35]. L. Q. Jiang and B. E. Keol, J. Phy. Chem. 96 (1992) 8694.
[36]. C. Xu and B. E. Koel, J. Chem. Phy. 100 (1994) 664.
[37], P. A. Redhead, Vacuum 12 (1962) 203.
[38]. J. B. Miller, H. R. Siddiqui, S. M. Gates, J. N. Russel, J. T. Yates, Jr., J. C. Tully, J.
Ach and S. Lehwald, Surf. Sci. 91 (1980) 187.
[39]. D. H. Parker, M. E. Jones and B. E. Koel, Surf. Sci. 233 (1990) 65.
[40]. C. Xu and B. E. Koel, Surf. Sci. 304 (1994) L505.
[41]. M. T. Paffett, S. C. Gebhard, R. G. Windham and B. E. Koel, J. Phys. Chem. 94
(1990) 6831.
[42]. Y.- L. Tsai, C. Xu and B . E. Koel, Surf. Sci. Submitted.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
183 K
1 3
E
c O
C M
CO
CO
c
0 3
C
100 150 2 0 0 250 3 0 0
Temperature (K)
Fig. 3.1 Methanol TPD spectrum after methanol exposures on the Pt(l 11) surface. The
multilayer desorption peak at the highest exposure has been cut off. Exposures
from the bottom to top are 10,20,40,60 and 70 s.
39
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
C H 3O H
(2x2)Sn/Pt(111)
177
u
E
c c
C\J
C O
CO
c
3
c
X
o
c
5
C O
300 2 5 0 200 1 50 1 00
Temperature (K)
Fig. 3.2 TPD spectrum of methanol on (2x2) Sn/Pt(l 11) surface with different
coverages. The exposures from the top to bottom are 5,20,40 and 60 s.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
c h 3oh
(V’ 3x^3)Sn/Pt(111)
170 K
CM
C O
CO
250 200 3 0 0 1 00 1 50
Temperature (K)
Fig. 3.3 Methanol TPD spectrum after methanol exposures on C/3x/3)R30° Sn/Pt(l 11)
surface with coverages bottom to top 10,20,40 and 60 s.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
202
3
E
CO
C O
■ 'O '
C O
c
03
C
X
O
m
X
cf1
200 2 5 0 3 0 0 100 1 5 0
Temperature (K)
Fig. 3.4 Ethanol TPD spectrum after ethanol exposures on Pt(l 11) surface with different
coverages. The exposures from top to bottom were 5,10,20,40,60 and 80 s.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
c2h5oh
(2x2)Sn/Pt(111)
200 K
| !---------1 --------1 I [ I I I * | * * 1 1 | 1 1 1 1
100 150 200 2 5 0 3 0 0
Temperature (K)
Fig. 3.5 TPD spectrum of ethanol on (2x2) Sn/Pt( 111) surface with different coverages.
The exposures from top to bottom were 10, 20,40,60 and 120 s.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
C2H5OH
(•/3xVr 3)Sn/P t(111)
190 K
3
E
c c
C O
■ ' 3 '
C O
a
0
a
X
O
n f
c f
250 1 50 200 300 100
Temperature (K)
Fig. 3.6 Ethanol TPD spectrum after ethanol exposures on (V3xV3)R30° Sn/Pt(l 11)
surface with exposures top to bottom 5,10, 20,40 and 80 s.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
T P D A rea T P D Area
■ (2x2) alloy
▲ V3 alloy
8 0 0 40 6 0 20
Exposure (s)
■ (2x2) alloy
▲ V3 alloy
1 00 8 0 0 4 0 60 20
Exposure (s)
Fig. 3.7 Methanol and ethanol uptake curves results from TPD experiments for different
exposures on Pt(l 11), (2x2) and a /3 surfaces.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
3
E
CO
O J
C O
179 K
V3 alloy
186 K
c
3
E
(2x2) alloy
x
o
CO
X
194 K
Pt(1 11)
O
300 200 250 150 100
Temperature (K)
3
E
CO
C O
197 K
V3 alloy
206 K
CO
c
CD
(2x2) alloy
x
o
x
213 K
Pt(111)
300 200 250 100 1 50
Temperature (K)
Fig. 3.8 Comparison of methanol and ethanol desorption spectra for 0 s O.10sat of
methanol and ethanol on Pt(l 11), (2x2) and V3 surface alloy.
46
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
H g O Intensity (18 amu)
161 K
163 K
167 K
185 K
120 1 60 2 0 0 2 40
185 K
2 5 0 3 0 0 200 1 00 1 5 0
Temperature (K)
Fig. 3.9 TPD spectrum of water dosed at 100 K on P t( lll) surface with different
exposure. The inset shows additional spectra for very low exposures.
47
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
H 20/(2X2)Sn/Pt(111)
160 K
1 3
E
c c
00
66 K
C/3
c
CD
c
o
n ?
58 K
100 120 140 160 180 2 0 0 220 240
Temperature (K)
Fig. 3.10 Water TPD spectrum after water exposures on (2x2) S n/P t(lll) surface alloy
with different coverages at 100 K.
48
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
H 20/(Vr 3Xvr 3)Sn/Pt(111)
160 K
3
E
c o
00
165 K
c
0 3
C
O
c
X
cvi
152 K
100 120 140 160 180 200 220 240
Temperature (K)
Fig. 3.11 TPD spectrum of water on (V3xV3)R30° Sn/P t(lll) surface with different
coverages.
49
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
157 K
^3 alloy
3
E
c a
C O
158 K
>s
c n
c
0)
c
O
c
X
(2x2) alloy
CM
164 K
185 K
Pt(111)
1 80 1 40 200 220 120 160
Temperature (K)
Fig. 3.12 Comparison of desorption peak temperature of about 10 % of saturation
monolayer coverage of watert on Pt(l 11) and the (2x2) and V3 alloy surfaces.
50
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
220-1
- 5 5
EtOH
200 -
- 5 0
MeOH
Q .
1 8 0
- 4 5
C L
1 6 0 -
- 4 0
(2x2) V 3
1 40-1
0 .0 0 0 .0 5 0.10 0 .1 5 0 .2 0 0 .2 5 0.3 0 0.35
O
C D
C O
O
o
>
o
< ‘
0 9
m
3
C D
C Q
7T
C _
Sn Concentration (%)
Fig. 3.13 Influence of the alloyed Sn concentration on the desorption peak temperature of
methanol ethanol and water in the Sn/pt surface alloys. The right hand axis
gives an estimation of the corresponding molecular desorption energies.
51
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Table 3.1. Thermal desorption peak temperatures (top entries) and desorption activation
energies in kJ/mol (bottom entries) for methanol, ethanol, and water obtained at about 10%
of the monolayer saturation coverages on P t(lll) and two Pt-Sn surface alloys. The left
hand values given in parentheses are activation energies obtained by Redhead analysis, and
the right hand valuse were determined by leading edge analysis.
Molecule P t(lll) (2x2) V3
Methanol 194 K 186 K 179 K
(49,47) (47,44) (45,41)
Ethanol 213 K 206 K 197 K
(54,51) (52,53) (50,47)
Water 162 K 156 K 152 K
(bilayer) (41,43) (39,43) (38,44)
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Chapter 4
Probing the Influence of Alloyed Sn On Pt(100) surface Chemistry by
CO Chemisorption
Abstract
The chemisorption properties of the c(2x2) and (3V2x/2)R45° Sn/Pt(100) alloy
surfaces, along with the clean Pt(100) surface, were investigated using CO as a probe
molecule. Temperature programmed desorption (TPD) studies reveal a reduction in CO
desorption peak temperature, and thus the chemisorption bond energy, by alloying Sn into
the Pt(100) surface. A large decrease was observed in the saturation coverage of CO on
these alloyed surfaces at 150 K compared to the Pt(100) surface. The initial sticking
coefficient of CO was found however to be nearly independent of the surface Sn
concentration. High resolution electron energy loss spectroscopy (HREELS) studies
showed that CO was only chemisorbed on atop sites on both alloys. Sn incorporation
forms isolated Pt atoms at the surface of these two Sn/Pt(100) alloys and eliminates the
possibility of CO bonding to multiple Pt centers, i.e., pure-Pt two- and three-fold bridging
sites.
53
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
4.1. Introduction
Understanding and controlling chemical reactions at alloy surfaces is one of the
outstanding challenges in surface science. The ability to prepare and define ordered
surface alloys provides new opportunities for fundamental studies of the chemistry of
bimetallic surfaces. Much of this work in surface chemistry is motivated by catalysis.
Bimetallic catalysts often exhibit higher selectivity and/or improved lifetimes compared to
single-metal supported catalysts [1-3]. For example, Pt-Sn bimetallic catalysts are used
commercially in refining and petrochemistry [3, 4] and selective catalytic
(de)hydrogenation [5, 6], and Pt-Sn alloys have also been investigated as possible
electrocatalysts in fuel cell applications [7, 8]. The activity and selectivity of bimetallic
catalysts depend strongly on the surface composition and arrangement of the constituent
metals, with the observed effects commonly explained as due to electronic (ligand) or
geometric (ensemble) factors [9, 10]. Difficulties in the past in elucidating the origin of
the observed changes in reactivity due to alloying often come from lack of information at
the atomic level on the composition and structure of the surface. Herein, we report surface
science studies that utilize well-defined, ordered Pt-Sn bimetallic surfaces and the
interaction of CO as a probe of important fundamental aspects of the chemistry and
reactivity of these surfaces.
CO chemisorption and bonding on Pt(100) has been extensively studied, in
particular that associated with adsorbate induced-surface reconstruction [11-14]. The
clean Pt(100) surface exhibits a quasi-hexagonal, close-packed (5x20) reconstructed
surface normally referred to as the "hex" phase [15, 16]. The hex surface undergoes
reconstruction upon chemisorption of small molecules like CO [11-14], NO [17], and H,
or O, [18] to produce the (lxl)-Pt(lOO) square lattice. Thiel et al. [13] identified that the
higher heat of adsorption of CO on the (lx l) phase (157 kJ/mol) than on the
54
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
reconstructed phase (115 kJ/mol) is the driving force for the CO-induced surface phase
transition. They also proposed that the phase transition occurs by adsorption of CO on
the hex phase, followed by migration and trapping of CO into growing (lxl)-C O
"islands" initially formed by nucleation. Scanning tunneling microscopy (STM) later
confirmed an island growth mechanism [19]. The HREELS results of Behm et al. [12]
indicated that a critical coverage of ~0.05 ML CO at 310 K was required to induce the
hex— >(lxl) phase transition, while Jackman et al. [20] determined this critical coverage to
be 0.08+0.05 ML from Rutherford back-scattering spectroscopy (RBS). The local CO
coverage on (lxl)-C O islands is 0.5 ML when both the (lx l) and hex Pt(100) surface
shows a c(2x2)-CO Pt(100) LEED pattern, and further CO exposure gives rise to several
higher coverage ordered structures: c(5V2xV2) (0C O = 0.6 ML); c(3\/2xV2)R45° (0C O =
0.67 ML); and finally, at saturation, a c(2x4) structure together with a c(3Vr 2xV2)R45°
pattern, corresponding to a CO coverage of 0C O = 0.75 ML [13, 21], where 1 ML is
defined relative to the (lxl)-Pt(lOO) surface (1.28x1015 cm'2 ).
Martin et al. [21] investigated adsorption of CO on both hex and (Ixl)-Pt(lOO)
surfaces by infrared reflection absorption spectroscopy (IRAS) at 90 and 300 K. A single
atop-CO band was obtained on the (lx l) surface at 2065 cm-1 at 90 K and 2090 cm' 1 at
300 K, which shifted to higher frequency with increasing exposure, along with at least
three different bands in the bridging region between 1867 and 1910 cm' 1 depending on the
coverage. On the hex surface, at low CO coverages, only one atop band was observed at
2083 cm' 1 at 300 K. This was assigned to CO bound at atop sites on the hex surface
without reconstruction. With increasing CO exposures, a second atop-CO band at 2087
cm' 1 appeared along with a bridge-bonded CO band at 1870 cm'1 . It was proposed that
these bands were related to a CO-induced surface reconstruction. An earlier HREELS
study showed that the major contrast was at low coverages, with a much higher fraction of
55
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
bridging CO on the (lxl)-Pt(lOO) surface compared to only atop-CO on the hex surface
[12].
Two Sn/Pt(100) alloys can be prepared by the vapor deposition of Sn onto Pt(100)
single crystal substrates in UHV [22, 23]. These alloys have surface nets giving c(2x2)
and p(3V2x/2)R45° LEED patterns and have 0S n = 0.5 and 0.67 ML, respectively [22].
CO TPD spectra only after saturation CO exposures on these two alloy surfaces were
reported previously [22], and the activation energy for CO desorption was estimated to be
104.5 and 100 kJ/mol, respectively. This indicates a reduction of the heat of adsorption
for CO at saturation coverages by 17-21 kJ/mol compared to clean Pt(l00) [22]. Also it
was reported that the saturation CO coverage decreased from 0.77 ML on Pt(100) at 300
K to 0.28 and 0.22 ML, respectively, on the c(2x2) and (3V2xVr 2)R45° Sn/Pt(100) alloys
at 300 K.
This paper reports on a much more detailed investigation of CO adsorption on these
two Pt-Sn surface alloys and hex Pt(100) by using primarily TPD and HREELS. The
structure of the c(2x2) surface alloy is such that 0S n = 0.5 ML with the surface Pt atoms
isolated, surrounded by only Sn nearest neighbors, to form a "checkerboard" pattern. In a
related study on Sn/Pt(i 11) surface alloys, we found that alloyed Sn had little effect on the
adsorption of CO on P t(lll) [24], The CO chemisorption bond strength was reduced
only about 21 kJ/mol compared to P t(lll), and little change was observed in the
population of the thermodynamically preferred (bridging) adsorption site or saturation CO
coverage [24]. The structure and composition of the two Sn/P t(lll) alloys are quite
different from the Sn/Pt(100) alloys investigated here, and hence studies of these other two
alloys formed on Pt(100) provide essential information on the role that surface structure
and composition play in tailoring the surface chemistry of Pt-Sn alloys.
56
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
42. Experimental Methods
These experiments were performed in a stainless steel UHV chamber that has been
described previously [25] with a base pressure of 8x10'“ torr. Briefly, it was equipped
with low energy electron diffraction (LEED) optics, a cylindrical mirror analyzer (CMA)
for Auger electron spectroscopy (AES), an ion sputtering gun, a UTI 100C quadrupole
mass spectrometer (QMS) for TPD, and facilities for gas dosing and metal evaporation.
The Pt(100) crystal was cleaned by repeated cycles of Ar+ ion sputtering and oxygen
treatments (PO,=5xlO'7 torr at 1200 K). The cleanliness of the crystal was checked by
AES and LEED. The clean surface showed a reconstructed (5x20)-Pt(100) pattern [15,
16]. The crystal was resistively heated and cooled to 95 K by liquid nitrogen, with the
temperature measured by a chromel-alumel thermocouple spot-welded to the side of the
crystal.
CO exposures were carried out by using a leak valve connected to a multichannnel-
array gas doser. CO (Matheson, 99.5% purity) was used without further purification.
Exposures of CO are given in units of Langmuirs (L), corrected for ionization gauge
sensitivity (Sco/SN , =1.08) and a gas-doser flux enhancement factor (~40) over that for
background gas flux, as determined by forcing the initial slope of the uptake curve on hex
Pt(100) at 100 K to correspond to an initial sticking coefficient of 0.8 [26].
All of the TPD experiments were done with a heating rate of 4 K/s. HREELS
spectra were taken with a spectrometer containing single 127° cylindrical sectors in the
monochromator and analyzer. All spectra were taken in specular reflection with 0in = 0o u t
= 65° from the surface normal. The electron incident energy was 4.5 eV, with a typical
resolution of 80 cm"1 from clean Pt(100).
The c(2x2) and (3V2x/2)R45° Sn/Pt(100) surface alloys were prepared by
evaporating Sn on the clean, reconstructed (5x20)-Pt(100) surface at 300 K and then
57
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
annealing to 750 and 900 K respectively, as described by Paffett et al. [22], The c(2x2)
LEED pattern observed following Sn deposition on Pt(100) can be due to two different
surface structures [22,23]. Over a range of temperatures (500-750 K), a c(2x2) overlayer
of Sn adatoms is formed [23]. Alloy formation by incorporation of Sn into Pt sites in the
outermost layer into a c(2x2) structure occurs over a narrow range between 750-800 K,
depending on the initial Sn concentration. A more thermally stable (3V2xV2)R45° alloy is
formed at higher annealing temperatures of over 800 K and this structure is stable up to
1050 K. The structure and composition of the c(2x2) alloy has been determined by low
energy alkali ion scattering spectroscopy (ALISS) [23], with 0S n = 0.5 ML and the surface
quite "flat", with an outward buckling distance of 0.19+0.02 A for Sn atoms above the
surface Pt plane [23]. ALISS revealed that the alloy formed with the (3\/2x‘ /2)R450
LEED pattern had a very similar surface structure to that for the c(2x2) alloy, but the
detailed structure and composition of this surface has not been elucidated yet. The surface
Sn coverage in the (3V2xV2)R45° alloy was estimated to be 0.67 ML, and Paffett et al.
[22] have suggested that the most stable and reproducible (3V2x-/2)R45° Sn/Pt(100)
surface is a faceted surface with Sn rich domain boundaries. ALISS results suggest that
the p(3/2xV2)R45° alloy has nearly the same geometric structure as the c(2x2) alloy, and
is composed of small c(2x2) alloy domains with the same buckling distance within the
domains [23]. For brevity, the c(2x2)Sn/Pt(100) and p(3Vr 2xVr 2)R45° Sn/Pt(100) surface
alloys will be referred as the c(2x2) and 3V2 alloys, respectively, throughout this paper.
4.3. Results
43.1. TPD
TPD spectra of CO following CO adsorption on a clean hexagonally reconstructed
(5x20)-Pt(100) surface at 150 K is shown in Fig. 4.1 and can be briefly summarized as
follows. The desorption of CO at low coverage occurred in a broad peak at 465 K. With
58
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
increasing CO coverage, the peak maximum shifted to higher temperatures and other
features formed at lower temperatures. At saturation CO coverage (0C O = 0.77 ML), three
peaks appeared. The highest temperature peak at 520 K corresponds to a local CO
coverage on the (1x1) phase up to 0.25 ML, the peak at 475 K corresponds to a local
coverage on the (lx l) phase between coverages 0.3 - 0.5 ML and the lowest temperature
peak at 435 K corresponds to higher coverage structures [11]. This shift of the highest
temperature desorption peak to higher temperatures with increasing coverage is consistent
with previous results [11, 12], and was explained in terms of a “nucleation-trapping”
model of the surface phase transition from hex (5x20)-Pt(100) — > (lxl)-Pt(lOO).
A series of TPD spectra for CO desorption from the c(2x2) alloy are shown in Fig.
4.2. All of the CO exposures were made with the substrate at ~150 K. Completely
reversible adsorption and desorption of CO occurred on the alloyed surface. AES spectra
taken after the TPD experiments showed no traces of carbon or oxygen on the surface. At
low coverages, the CO desorption peak appeared at 420 K and gradually shifted to lower
temperature (402 K) at saturation coverage. The higher temperature desorption peak is
possibly associated with desorption from defects sites that arise from incomplete coverage
of the alloy on the Pt(100) crystal.
CO TPD spectra from the 3V2 alloy is shown in Fig. 4.3. The CO desorption peak
temperature shifted from 408 K at low CO coverage to 396 K at saturation CO coverage.
The CO desorption peaks obtained from the 3V2 alloy were significantly narrower than
those measured on clean Pt(l00). This reflects the reduction of possible CO bonding
sites on the alloy. This is consistent with the presence of CO bonding at only atop sites
on the alloy, as indicated by HREELS spectra reported in Section 4.3.2.
The desorption temperature of the most strongly bound CO decreased from 520 K
on Pt(100) to 402 K on the c(2x2) alloy and to 396 K on 3V2 alloy. Using Redhead
59
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
analysis [27], the desorption activation energies were estimated to be 135, 104 and 102
kJ/mol on Pt(100), c(2x2) and 3V2 alloys, respectively. This decrease in desorption
energy of ~33 kJ/mol indicates weaker adsorbate-surface bonding and thus shows an
important aspect of the influence of Sn in reducing the reactivity of the Pt(100) surface.
Similar CO adsorption behavior was also found on the Sn/Pt(l 11) alloys in our previous
studies [24].
Fig. 4.4 compares the CO TPD spectra from several surfaces after saturation
exposure of CO on these surfaces at 150 K. Also included are results from CO
adsorption on a c(2x2) Sn overlayer, prepared by evaporating 0.5 ML of Sn on the hex
Pt(100) surface and then annealing to 500 K for 10 sec. CO adsorption was completely
eliminated on this surface at 150 K. These spectra show clearly the dramatic difference in
the site blocking ability of Sn depending on its location — as either an adatom or alloyed
atom — on the Pt(100) surface. Fig. 4.4 also illustrates nicely the effect of surface Sn in
reducing CO the desorption temperature, and thus the lifetime and coverage of CO at
higher pressures, on Pt-Sn alloys.
CO adsorption kinetics are revealed in the uptake plots for CO on Pt(100), and the
c(2x2) and 3V2 alloys shown in Fig. 4.5. The CO coverage scale was calibrated using the
value of the saturation coverage of CO on the hex Pt(100) surface at 160 K defined to be
0C O =0.77 ML [28,29]. Relative CO coverages were determined by utilizing the CO TPD
peak areas. The CO coverages determined in this manner were verified independently
using X-ray photoelectron spectroscopy (XPS) measurements of the C(ls) peak areas.
The saturation coverage of CO decreased from 0.77 ML on the Pt(100) surface to 0.28
ML on the c(2x2) and to 0.25 ML on the 3V2 alloy. This decrease occurs because of the
extensive loss of Pt atoms (which are required to chemisorb CO) in the surface layer of
the alloys. The initial slopes of the uptake curves in Fig. 4.5 (inset) also indicate that the
60
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
initial sticking coefficient of CO on these substrates at 150 K does not decrease linearly
with the surface Sn concentration, as would be expected from Langmuirian adsorption
site-blocking, but rather decreases from the value of the initial sticking coefficient of CO
on Pt(100) surface as 1:0.75:0.64. This indicates CO adsorbs via a mobile, modifier
precursor state on the two alloy surfaces [30].
43.2. HREELS
Vibrational spectra of chemisorbed CO for increasing coverages on Pt(100) at 150
K are shown in Fig. 4.6 . At low coverages, one vco peak at 2078 cm' 1 was seen and this
band is assigned to CO molecules bonded at atop sites [11, 21]. With increasing
coverage, an additional, less intense band due to CO bound at bridging sites [21] was
observed at 1887-1917 cm'1 . At higher coverages two Vp,^ peaks were observed at 380
and 471 cm'1 . A shift of 10-15 cm'1 of the CO stretching mode to higher energies
occurred for both atop and bridge-bonded CO with increasing CO coverage, and this is
expected from dipole-dipole coupling between CO molecules [31].
Consistent with previous work [11, 12, 21] CO is adsorbed only at atop sites up to
some low coverage (0~O.O6 ML) without removing the hexagonal reconstruction, and then
a phase change to (Ixl)-Pt(lOO) is induced even at 150 K at some coverage of CO (higher
than that required at 300 K) where bridge-bonded CO appears -- a characteristic feature of
adsorption of CO on (lxl)-Pt(lOO).
Fig. 4.7 shows HREELS spectra after a saturation dose of CO on Pt(100) at 150 K
and subsequent annealing. Annealing reduces the CO coverage and both the peaks for
atop and bridge-bonded CO are shifted to lower energies with increasing temperature
because of reduced dipole-dipole coupling. The intensity of the atop-CO peak decreased
while the intensity of the bridge-bonded CO peak remained almost constant during
annealing, until all CO desorbs near 500 K. The ratio of bridging to atop loss peaks
61
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
initially was 0.28 at 150 K, and was reduced to 0 2 and 0.6 after annealing to 300 and 450
K, respectively. This also indicates CO-induced reconstruction of the hex-»(lxl) surface
at 300 K because this ratio is very small for low coverages of adsorbed CO on the hex
surface. After annealing to 500 K, almost all of the adsorbed CO desorbed, and the
surface reconstructed back to the hex surface. Small loss peaks near 1650 cm' 1 appeared,
assigned to a small amount of adsorbed CO on fourfold hollow sites on the (lx l) surface
remaining during reconstruction, and the intensity of the elastic peak increased by a factor
of 10 after annealing to 500 K.
Vibrational spectra of CO on the c(2x2) alloy are shown in Fig. 4.8 following
sequential CO exposures. At low CO coverage, the spectrum shows only one peak at
2037 cm' 1 and this shifts to 2057 cm' 1 at saturation CO coverage. The intensity of this
peak (relative to the elastic peak) was much lower than the loss peaks observed on Pt(100)
at 2088 cm'1 . We assign this peak to CO molecules bound at atop Pt sites on the alloy
surface (only atop Pt sites exist on this alloy). Hence, Sn alloying reduced the population
of atop sites beyond the expected 0.5 ML, and more significantly completely eliminated
the low energy band (~1890 cm'1 ) of bridge-bonded CO at high CO coverages on the
alloy.
Fig. 4.9 shows that also only one vibrational peak was observed at 2047 cm' 1 on the
3V i2 alloy. On both of these two alloyed surfaces, the loss peak shifted to higher energies
with increasing CO coverage. On both alloys, a larger frequency shift of ~20-30 cm' 1
occurred compared to ~10 cm' 1 on clean Pt(100) by increasing the CO coverage from 0.1
ML to saturation coverage. Assuming that in all cases the shift is due to dipole-dipole
coupling between adsorbed CO molecules, this may be due to shielding by bridging CO
on Pt(100) or due to the relatively small change in molecular environment on clean Pt(100)
with increasing coverage because of island formation.
62
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
For both of the c(2x2) and 3a /2 surface alloys, the CO stretching mode was at a
lower energy compared to that on Pt(100), whereas remained almost constant. This
small shift of vco can be attributed to the reduction of CO dipole-dipole coupling because
of the lower CO coverages on these Sn/Pt alloys.
4 3 3 . LEED
Chemisorption of CO on both of the Sn/Pt(100) alloys was examined by LEED. No
new spots indicative of ordered structures with different unit cell size or orientation were
obtained for CO adsorption on these two alloy surfaces. HREELS spectra shows that CO
occupies only on atop sites on these alloys. While the c(2x2) alloy structure implies an
ideal saturation CO coverage of 0.5 ML, the saturation coverage determined from the
integrated TPD area (0C o= 0-28 ML) was found to be a factor of about two less than this
coverage. A possible real space structure consistent with this coverage and our HREELS
and LEED observations is shown in Fig. 4.10. This proposed real space model of
p(2x2)CO / c(2x2)Sn/Pt(100) has an ideal CO coverage of 0.25 ML. In this structure CO
atoms occupy only atop sites in alternating rows, possibly because of CO-induced
relaxation changes in the surface Pt and Sn atoms that lift Pt atoms bonded to CO
molecules toward the surface and thus force nearest-neighbor Pt atoms away from the
surface into positions that are more effectively blocked by Sn from CO access. The
proposed p(2x2) CO adlayer and the CO-induced structural changes should be addressed
in the future by LEED-IV or similar surface crystallography probes.
4.4. Discussion
Alloying the Pt(100) surface with Sn to form the c(2x2) and 3V2 alloys causes a
reduction in the heat Discussion of adsorption of CO, i.e. Pt-CO bond strength, compared
to the clean Pt(100) surface. We estimate that the desorption activation energy decreased
63
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
from 135 kJ/mol on the clean Pt(100) surface to 104 and 102 kJ/mol on the c(2x2) and
3V2 alloys, respectively. A smaller reduction in the CO adsorption energy was found in a
related series of P t(lll) and the p(2x2) and (V3xV3)R30° S n /P t(lll) alloys: 121.5,
1043, and 100 kJ/mol, respectively. These CO chemisorption results indicate that for
strong jc-acceptors such as CO, Sn alloyed with Pt does not strongly modify the
chemisorption bond strength at a given adsorption site, i.e., Sn has only a weak electronic
effect, but does strongly modify the availability of Pt sites (e.g., bridging versus atop)
primarily via a geometric, or site blocking, effect.
Much more extensive suppression of CO chemisorption occurred on the Sn/Pt(100)
alloys, decreasing 0 ^ 0 from 0.77 ML on clean Pt(100) to 0.28 and 0.25 ML on the two
alloys with 0S n = 0.5 and 0.67 ML, respectively. For CO on S n /P t(lll) [24], O^-q
decreased from 0.68 ML on Pt(l 11) to 0.65 ML on the p(2x2) alloy (0S n = 0.25 ML) and
0.53 ML on the V3 alloy (0S n = 0.33 ML). All of these results are compared in Fig. 4.11.
This curve provides a qualitative guide to the influence of alloyed Sn on CO chemisorption
over a wide range of Sn concentrations. It certainly appears that alloyed Sn should have
little effect on the saturation coverage of CO on Pt-Sn alloy surfaces until Sn
concentrations are reached that eliminate the pure-Pt threefold hollow sites.
From the uptake curves of Fig. 4.5, one can see that the initial sticking coefficient is
reduced but not as strongly as one might guess based on the surface Sn concentration.
Similar results were obtained for CO adsorption on Sn/Pt(l 11) surface alloys [30], as well
as many other bimetallic and modified surfaces, where the sticking coefficient was nearly
independent of modifier coverage up to -0.25 ML [32, 33]. These results clearly show
that under many conditions a simple Langmuir or modified-Langmuir site-blocking
equation for the dependence of the initial sticking coefficient on modifier coverage S =
So (l-a0)b fails due to the important influence of a “modifier precursor” state on the
64
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
adsorption kinetics. This factor allows for much faster adsorption kinetics than would a
linear decrease of the initial sticking coefficient with increasing modifier (Sn) coverage.
The values for the CO initial sticking coefficient on four different ordered Pt-Sn alloys at
130-150 K are compared in Fig. 4.12.
HREELS spectra of chemisorbed CO on both Sn/Pt(100) alloys, indicating CO
chemisorption on atop sites only at 150 K, are consistent with the structural models
proposed for these alloyed surfaces [22,23]. Pure-Pt twofold and fourfold bridging sites
for chemisorption are completely eliminated. TPD spectra of the alloyed Sn/Pt(100)
surfaces are much narrower than those on Pt(100) consistent with the reduction in
possible adsorption sites. It is interesting to compare the present HREELS results with
those for adsorbed CO on the two S n/P t(lll) alloy surfaces. For both p(2x2) and /3
alloys, both the bridging and atop sites were populated at saturation CO coverage due to
the presence of pure Pt twofold bridge sites on both of these two Sn/P t(lll) alloys and
little change in population was indicated. For saturation CO adlayers, vco for CO at atop
sites appeared at 2090 and 2085 cm"1 on the (2x2) and /3 alloys, respectively, compared to
vco at 2105 cm' 1 on Pt(l 11) (a shift of vco for bridge-bonded CO towards lower energy
was also observed).
Similarly, the atop CO stretching mode was shifted to lower energy than on Pt(100)
for both the c(2x2) and 3 /2 Sn/Pt(100) surface alloys. This can be attributed to the
reduction of CO dipole-dipole coupling due to corresponding lower CO coverages on all
of these Pt-Sn alloys. We note that the shift of the atop CO mode is smaller on the 3 /2
alloy than on the c(2x2) alloy compared to clean Pt(100) even though ISS showed a
higher surface Sn concentration than on the c(2x2) alloy [23]. This could be due to a
weaker Pt-CO bond, although this is not likely because of the TPD results, or due to the
special arrangement of Pt and Sn atoms in the 3/2 alloy. Consistent with this latter
65
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
possibility, we have recently found benzene formation from the trimerization of acetylene
occurs only on the 3V2 alloy, and not on the c(2x2) alloy [34].
A final comment is to say that probing chemistry at surfaces with CO is not
particularly general. CO is one of the strongest ir-acceptor ligands often encountered.
The electron transfer involved in CO adsorption is usually discussed within the Blyholder
model [35], involving 5 a donation from CO to empty da states of the metal and back-
donation from the filled dT states of the metal to 2t c antibonding orbitals of CO. Hence
characterization of chemisorption bond strengths at these Pt-Sn alloy surfaces by CO
chemisorption might be expected to not be very general when considering other reactive
molecules, in particular molecules that act as donor ligands. But these experiments do
reveal important fundamental insight into quite general features of the behavior of
adsorption rate constants and maximum chemisorption coverages that should be useful for
understanding a wide variety of reactions on bimetallic catalysts.
4.5. Conclusions
Insight was gained into the chemisorption properties of the alloyed Sn/Pt(100)
surfaces by using CO adsorption to probe the structural and electronic differences
between the alloyed surfaces and the Pt(100) surface. Alloying with Sn causes only a
small reduction of 30-34 kJ/mol in the adsorption energy of CO on these surface alloys
compared to the Pt(100). The saturation coverage of CO on the alloyed surfaces at 150 K
decreased from 0.77 ML on clean hex Pt(100) to 0.28 and 0.25 ML on c(2x2) and
(3Vx2vr 2)R45° Sn/Pt(100) respectively, whereas the initial sticking coefficient is nearly
independent of surface Sn concentration. No CO chemisorption on the bridged Pt sites
has been found as indicated by the lack of CO stretching band in HREELS compared to
clean Pt(100) surface, because Sn has altered the ability to bind CO on bridged sites as a
66
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
site blocking agent. Hence the molecular chemisorption can be modified by altering the
available metal ensemble sizes
Acknowledgment
This work was supported by the Division of Chemical Sciences, Office of Basic
Energy Sciences, U.S. Department of Energy. We thank Dr. Hong He for his assistance
in the HREELS experiments.
67
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
REFERENCES
[1]. J. H. Sinfelt, Bimetallic Catalysts: Discoveries, Concepts and applications, Wiley,
New York, 1983.
[2]. H. Lieske and J. Volter, J. Catal. 90 (1984) 96.
[3]. R. Srinivisan, R. J. DeAngelis and B. H. Davis, J. Catalysis. 106 (1987) 449.
[4]. Y. Zhou and S. M. Davis, Catalysis Lett. 15 (1992) 51.
[5]. R. D. Cortright and J. A. Dumesic, J. Catal. 148 (1994) 771.
[6]. R. D. Cortright and J. A. Dumesic, J. Catal. 157 (1995) 576.
[7]. B. Beden, F. Kardigan, C. Lamy and J. Leger, J. Electroanal. Chem. 127 (1981) 75.
[8]. A. N. Haner and P. N. Ross, J. Phys. Chem. 95 (1991) 3750.
[9]. R. Burch, Acc. Chem. Res. 15 (1982) 24.
[10]. J. W. A. Sachtler andG. A. Somoijai, J. Catal. 81 (1983) 77.
[11]. P. A. Thiel, R. J. Behm, P. R. Norton and G. Ertl, Surf. Sci. 121 (1982) L553.
[12]. R. J. Behm, P. A. Thiel, P. R. Norton and G. Ertl, J. Chem. Phys. 78 (1983) 7437.
[13]. P. A. Thiel, R. J. Behm, P. R. Norton and G. Ertl, J. Chem. Phys. 78 (1983) 7448.
[14]. Y. Y. Yeo, L. Vattuone and D. A. King, J. Chem. Phys. 104 (1996) 3810.
[15]. R. J. Behm, W. Hosier, E. Ritter and G. Binnig, Phy. Rev. Lett; 56 (1986) 228.
[16]. X.-C. Guo, A. Hopkinson, J. M. Bradley and D. A. King, Surf. Sci. 278 (1992) 263.
[17]. P. Gardner, M. Ttishaus, R. Martin and A. M. Bradshaw, Surf. Sci. 240 (1990) 112.
[18]. M. A. Barteau, E. I. Ko and R. J. Madix, Surf. Sci. 102 (1981) 99.
[19]. E. Ritter, R. J. Behm, G. Potschke and J. Wintterlin, Surf. Sci. 181 (1987) 403.
[20]. T. E. Jackman, K. Griffiths, J. A. Davies and P. R. Norton, J. Chem. Phys. 79
(1983) 3529.
[21], R. Martin, P. Gardner and A. M. Bradshaw, Surf. Sci. 342 (1995) 69.
[22]. M. T. Paffett, A. D. Logan, R. J. Simonson and B. E. Koel, Surf. Sci. 250 (1991)
123.
[23], Y. Li and B. E. Koel, Surf. Sci. 330 (1995) 193.
68
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
[24]. M. T. Paffett, S. C. Gebhard, R. G. Windham and B. E. Koel, J. Phvs. Chem. 94
(1990)6831.
[25]. R. G. Windham, M. E. Bartram and B. E. Koel, J. Phys. Chem. 92 (1988) 2862.
[26]. A Hopkinson and D. A. King, Chem. Phys. 177 (1993) 433.
[27]. P. A. Redhead, Vacuum. 12 (1962) 203.
[28]. P. R. Norton, J. A. Davies D. K. Creber, C. W. Sitter and T. E. Jackman, Surf. Sci.
108 (1981)205.
[29]. P. R. Norton, J. A. Davies andT. E. Jackman, Surf. Sci. 122 (1982) L593.
[30]. C. Xu and B. E. Koel; J. Phys Chem. 100 (1994) 664.
[31]. A. Crossley and D. A. King, Surf. Sci. 68 (1977) 528.
[32]. S. Johnson and R. D. Madix, Surf. Sci. 83 (1979) 487.
[33]. L. Q. Jiang, B. E. Koel and J. L. Falconer, Surf. Sci. 273 (1992) 273.
[34]. C. Panja, N. Saliba and B. E. Koel; to be published.
[35]. G. Blyholder, J. Phys. Chem. 68 (1964) 2772.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
C0/Pt(100) TPD
< * )
c
B
c
O
o
oo
C M
< z >
CO
c o
0.75
0.37
0.25
0.16
0.08
5 0 0 200 300 4 0 0 60 0
Temperature (K)
Fig. 4.1 CO TPD spectra after CO exposures on (5x20)-Pt(100) at 150 K.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
CO/c(2x2)Sn/Pt(100)
TPD
CO
C M
C/3
0 3
C O —
0.28
0.20
0.12
0.06
0.03
4 0 0 600 200 5 0 0 300
Temperature (K)
Fig. 4.2 CO TPD spectra following CO exposures on the c(2x2)Sn/Pt(100) alloy surface
at 150 K.
71
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
C O /(3V 2x/2)R 45 Sn/Pt(100) TPD
O
o
C O
C M
C O
CO
cc
0.25
0.18
0.12
0.06
0.03
2 0 0 4 0 0 6 0 0 3 0 0 5 0 0
Temperature (K)
Fig. 4.3 CO TPD spectra after CO exposures on the (3\^2xV2)R45° Sn/Pt(100) alloy
surface at 150 K.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Saturation CO Coverage
TPD
C O
c
0 3
C
o
O
Pt(100)
0 0
CM
C/3
C O
CO
c(2x2)Sn/Pt alloy
3 /2 Sn/Pt alloy
c(2x2)Sn/Pt overlayer
600 500 300 400 200
Temperature (K)
Fig. 4.4 Comparison of CO TPD spectra from Pt(100)(top) and c(2x2)Sn/Pt(100)
overlayer (bottom) to those from the c(2x2) alloy and 3v2 alloy.
73
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
C O T P D Peak Area
- 0.8
0.2 0.3 0.4 0.5 0.6 0.0
Pt(100)
- 0.6
- 0 .4
c(2x2)Sn/Pt(100)
- 0.2
(3 V 2 x \^ 2 )R 4 5 Sn/Pt(100)
h 0.0
0 3 4 1 2 5
CD
o
o
CO Exposure (L)
Fig. 4.5 CO uptake by hex Pt(100) and the Sn/Pt(100) surface alloys. Relative CO
coverages were obtained by using the integrated CO TPD peak areas from Figs.
4.1-4.3, with 0 c o on Pt(100) set to 0.77 ML [28, 29]. Uptake curves on an
expanded scale below 0.6 L CO exposure are reshown in the inset in order to
more easily see the effects of alloyed Sn on the initial sticking coefficient of CO.
74
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
T = 150 K
C0/Pt(100)
x 25
2 0 8 8
3 8 0
2078
471
c
CO
CO
o
_i
0.72 ML
1 887
0.18 ML
0.06 ML
0.03 ML
2000 3 0 0 0 1000 0
Loss Energy (cm 1)
Fig. 4.6 HREELS spectra of CO on hex Pt(100) at 150 K.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Saturation CO on Pt(100)
Annealing Temp
500 K
450 K
CO
c
d >
c
CO
CO
o
400 K
300 K
150 K
3 0 0 0 1000 2000 0
Loss Energy (cm 1)
Fig. 4.7 HREELS spectra of CO for an annealing sequence following a saturation dose
of CO on Pt(100) at 150 K.
7 6
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
C O /c ( 2 x 2 ) S n /P t ( 1 0 0 ) T = 150 K
x 25
2 0 5 7
481
0.28 ML
2 0 4 5
0.15 ML
2 0 3 7
0.08 ML
3 000 2000 1000 0
Loss Energy (cm 1)
Fig. 4.8 HREELS spectra of CO on the c(2x2)Sn/Pt(100) alloy at 150 K.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
CO/(3/2xV2)R 45 Sn/Pt(100)
T = 150 K
x25
2 0 7 8
4 8 5
C /3
C
0.25 ML
C/3
C/3
o 2 0 5 8
0.12 ML
2 0 4 7
0.06 ML
0 1000 2000 3 0 0 0
Loss Energy (cm 1)
Fig. 4.9 HREELS spectra of CO on the (3V2xV2)R45° Sn/Pt(100) alloy at 150 K.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Fig. 4.10 Possible model for the 2D structure of CO adsorbed on the c(2x2) Sn/Pt(100)
alloy at saturation CO coverage.
79
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
T=150 K
1 0 -
Pt(100)
8 -
CM
4 -
o
2 -
3 / 2 c(2x2)
0-I
0 .2 0.3 0.4 0.5 0.6 0.7 0.0 0.1
Sn Concentration (ML)
Fig. 4.11 Comparison of the saturation CO coverage on Pt(l 11), Pt(100), (2x2) Sn/Pt
(111) alloy, (V3x/3)R30°Sn/Pt(l 11) alloy, c(2x2)Sn/Pt(100) alloy, and
(3/2xV2)R45°Sn/Pt(100) alloy at 150 K.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
C O Initial Sticking Coefficient (S0)
T=150 K
0 . 8 -
0 . 6 -
0 . 4 -
0 . 2 -
(2x2) yf 3
3 / 2 c(2x2)
0 . 0 4
0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7
Sn Concentration (ML)
Fig. 4.12 Comparison of the initial sticking coefficient of CO on Pt(l 11), Pt(100),
(2x2)Sn/Pt(l 11) alloy, (/3x/3)R30°Sn/Pt(l 11) alloy, c(2x2)Sn/Pt(100) alloy,
and (3V2xV2)R45°Sn/Pt(100) alloy at 150 K.
81
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Chapter 5
Influence of Alloyed Sn On Adsorption and Reaction of NO on
Pt(100) surfaces
Abstract
Adsorption and reaction of NO on the (5x20)-Pt(100) surface and two Sn/Pt(100)
surface alloys has been studied using temperature programmed desorption (TPD) and
high resolution electron energy loss spectroscopy (HREELS). On the (5x20)-Pt(100)
surface in the absence of Sn, NO is primarily reversibly adsorbed and most of the
chemisorbed NO desorbs molecularly from the surface during TPD. About 25% of the
adsorbed NO monolayer decomposes at temperatures higher than 400 K and this leads to
N, and O, desorption from the surface. Alloying Sn into the surface layer of Pt(100)
forms two ordered surface alloys having c(2x2) and (3^2xV2)R45° Sn/Pt(100) surface
structures with 0S n = 0.5 and 0.67 ML, respectively. Alloying reduced the saturation
coverage of NO in the chemisorbed monolayer from that on Pt(100) at 100 K, and also
reduced the adsorption energy of molecularly bound NO by more than a factor of two.
Alloyed Sn, which removes all pure-Pt two-fold bridge and four-fold hollow sites,
completely changed the NO reaction pathway: nitrogen in NO was partially reduced to
82
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
form N ,0 on these alloys so that N 20 , along with NO and O, desorption was observed in
TPD. NO is bonded at the same site with a similar geometry on both Pt(100) and the
Sn/Pt(100) alloyed surfaces at low NO coverages, based on the HREELS spectra. At
saturation (monolayer) coverages, however, quite different HREELS spectra were observed
on both of the Sn/Pt(100) alloyed surfaces compared to that on Pt(100). Vibrations were
observed from adsorbed N20 , along with a shift of more than 60 cm' 1 for two vN O peaks
on both of the Sn/Pt(100) alloys compared to Pt(100). The two vN O peaks can be assigned
either as: (i) two vs modes of bent (tilted) and linearly-bonded atop NO, or (ii) vs and
stretching modes of a surface dinitrosyl species, i.e., two NO molecules bound to one Pt
atom. Dinitrosyl species have been proposed as intermediates for N2 0 formation in
reactions of NO on Mo(110) [Queeney and Friend, J. Chem. Phys. 107, 6432 (1997)],
and we suggest that a similar reaction mechanism occurs on Sn/Pt(100) alloys.
5.1. Introduction
Nitric oxide (NO) adsorption and reaction on transition metal surfaces is of
fundamental and technological interest [1-4]. As one example, catalytic removal of NOx
from combustion emissions is a crucial air pollution control problem because NO favors
the photochemical formation of ozone in the lower atmosphere and is also a “green
house” gas [5]. Catalysts for NO reduction are typically supported transition metals and
so NO adsorption and chemistry on transition metal surfaces is at the heart of this
catalysis.
Pt or Rh are good metals for NO reduction [6 , 7], but Pt-Rh bimetallic catalysts
outperform either Pt or Rh catalysts to remove simultaneously NOx, CO and
hydrocarbons in catalytic converters for automotive exhaust gases [8, 9]. This is not
surprising because bimetallic catalysts often have strikingly different performance in
83
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
catalytic processes compared to the single metal component systems [10, 11]. The activity
for NO+CO reactions is enhanced on Pd-Cu [12] and Pd-Cr [13] bimetallic surfaces
compared to Pd alone, and it has been suggested that the main reason is because NO
dissociation is more facile on the bimetallic Pd surfaces compared to pure Pd surfaces [12,
13].
Surface science studies have shown that the catalytic activity of Pt surfaces in the
CO+NO reaction is related to the efficiency of dissociation of NO, which varies for
different crystal planes [14], For example, NO partially decomposes to yield N, and O ,
on the Pt(100) surface [15-18], while no dissociation occurs on P t(lll) [16, 19, 20].
HREELS and infrared reflection absorption spectroscopy (IRAS) studies show that NO
adsorbs molecularly on the reconstructed hex-Pt(lOO) surface at room temperature, but
initially adsorbs dissociatively on the (lxl)-Pt(lOO) surface prior to molecular adsorption
at higher NO coverages [15,17,18],
The clean Pt(100) surface reconstructs to expose a “quasi-hexagonal”, close-
packed outermost layer with a (5x20) unit cell, and a stable Pt(100)-hex-R0.7° layer when
the first phase is annealed above 1100 K [21]. This is generally known as the “ h e x ”
phase and has been characterized by many techniques [21-23]. Adsorption of small
molecules like H, [24], CO [25-27] or NO [27-30], induce a phase transition from the hex
structure to a “square” (lxl)-Pt(lOO) structure with an unit cell as expected from the
ideal termination of the bulk lattice. Such adsorbate-induced phase transitions play a role
in the mechanism of several oscillatory reactions on Pt(100) [25-30].
On the hex-Pt(100) surface at temperatures between 200 and 300 K, NO adsorption
immediately lifts the reconstruction to form NO-covered (lxl)-islands [27-30]. At
temperatures below 200 K, adsorption does not immediately lift the reconstruction and this
results in significant NO accumulation directly on the hex phase. A temperature-
84
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
dependent critical coverage of NO is required for the structural phase transition to occur
[29].
In order to probe adsorption and reaction site requirements on bimetallic Pt surfaces,
we have been exploring the influence of Sn on Pt surface chemistry. We have previously
studied NO adsorption on two ordered Sn/Pt(l 11) surface alloys [20]. Alloying the
Pt(l 11) surface layer with Sn reduced the binding energy of the most strongly bound NO
chemisorbed on Pt(l 11) from 92 kJ/mol to 58 and 50 kJ/mol on the (2x2)Sn/Pt(l 11) and
the (V r 3xVr 3)R30° S n/P t(lll) alloys, respectively. No dissociation of NO occurred on
P t(lll) or either of these alloyed surfaces during TPD. Alloying with Sn reduced the
saturation monolayer coverage of NO from 0.5 ML on Pt(111) to 0.4 ML on the (2x2)
alloy and to 0.3 ML on the (V r 3x/3)R30° Sn/Pt(l 11) alloy, at 100 K [20].
Two additional, ordered Sn/Pt(100) alloys can be prepared by deposition and
annealing of thin Sn films on Pt(100) single crystal substrates in ultra high vacuum
(UHV) [31,32]. These alloys have surface nets giving c(2x2) and p(3V2x-/2)R45° LEED
patterns and have 0S n = 0.5 and ~ 0.67 ML, respectively. The structure and composition
of the Sn/Pt(100) alloys formed on Pt(100) are quite different than the two S n /P t(lll)
alloys that were previously investigated [20]. The c(2x2) alloy removes all pure-Pt two
fold bridge and four-fold hollow sites. Hence chemisorption studies of NO on these two
Sn/Pt(100) alloy surfaces provide essential information on the roles that surface structure
and composition play in tailoring the surface chemistry of Pt-Sn and other bimetallic Pt
alloys.
5.2. Experimental Methods
These experiments were performed in a stainless steel UHV chamber that has been
described previously [33] with an operating base pressure of 2xlO"1 0 torr. Briefly, it was
equipped with low energy electron diffraction (LEED) optics, a cylindrical mirror analyzer
85
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
(CMA) for Auger electron spectroscopy (AES), an ion sputtering gun, an UTI 100C
quadrupole mass spectrometer (QMS) for TPD, a HREELS spectrometer containing
single-127° cylindrical sectors in the monochromator and analyzer, and facilities for gas
dosing and metal evaporation.
The Pt(100) crystal was cleaned by repeated cycles of Ar* ion-sputtering and oxygen
treatments (PO 2 =5xl0'7 torr at 1200 K for 2 min.) followed by heating in vacuum to 1240
K for 30 sec. The cleanliness of the crystal was checked by AES and LEED. The clean
surface showed a reconstructed (5x20)-Pt(100) pattern [21-23]. The crystal was
resistively heated to 1250 K and cooled to 95 K by liquid nitrogen, with the temperature
measured by a chromel-alumel thermocouple spot-welded to the side of the crystal.
NO exposures were carried out by using a leak valve connected to a multichannnel-
array gas doser. NO (Matheson, 99.5% purity) was used without further purification.
Exposures of NO are given herein in units of Langmuirs (L), corrected for ionization
gauge sensitivity (SN O /SN , = 1.3) and a gas-doser flux enhancement factor (45 over that for
the background gas flux).
All of the TPD experiments were done with a heating rate of 4 K/s. HREELS
spectra were taken in specular reflection with 0in = 0out= 65° from the surface normal. The
electron incident energy in HREELS was 4.5 eV, with a typical resolution of 70-80 cm' 1
from clean Pt(100).
The c(2x2) and (3V2xV2)R45° Sn/Pt(100) surface alloys were prepared by
evaporating Sn on the clean, reconstructed (5x20)-Pt( 100) surface at 300 K and then
annealing to 750 and 900 K, respectively [31, 32]. The c(2x2) LEED pattern observed
following Sn deposition on Pt(100) can be due to two different surface structures [31, 32].
A c(2x2)-Sn overlayer (comprised of Sn adatoms) is formed over the temperature range of
500-750 K [31]. A c(2x2) alloy is formed by incorporation of Sn substitutionally into Pt
86
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
sites in the outermost layer over a narrow range between 750-800 K, depending on the
initial Sn concentration. A more thermally stable alloy with a (3\/2xV2)R45° LEED
pattern is formed at higher annealing temperatures of over 800 K, and this structure is
stable up to 1050 K. The structure and composition of the c(2x2) alloy has been
determined by low energy alkali ion scattering spectroscopy (ALISS) [32]. This surface
has 0S n = 0_5 ML and is quite “flat”, with an outward buckling distance of 0.19+0.02 A
for Sn atoms above the surface Pt plane [32], ALISS revealed that an alloy was formed
also for the (3V2xV2)R45° LEED pattern, with a similar local geometric structure to the
c(2x2) alloy and composed of small c(2x2) alloy domains with the same Sn-buckling
distance within the domains [32]. The surface Sn coverage in the (3Vr 2xV2)R45°
Sn/Pt(100) alloy was estimated to be 0.67 ML. Paffett et. al. [31] suggested that this
surface was faceted with Sn-rich domain boundaries. Recently, in what may be a closely
related structure, STM images of the (100) surface of a bulk PtjSn alloy crystal show
pyramidal features consisting of ( 102)-facets and terraces between and on top of the
pyramidal facets after sputtering and annealing the surface to 600 K [34]. For brevity, the
c(2x2)Sn/Pt(100) and (3V2xV2)R45°Sn/Pt(100) surface alloys will be referred to simply
as the c(2x2) and 3V2 alloys, respectively, throughout this paper.
5.3. Results
53.1. TPD
The principle desorption signals from reaction of a NO monolayer during TPD on
the hex-Pt(100) surface are shown in fig. 5.1. These curves were obtained simultaneously
following adsorption of a saturation dose of NO on a clean, hex-Pt(100) surface at 100 K.
Signals for C (12 amu) and N (14 amu) spectra can be used to distinguish the N, (28
amu) and N2 0 (44 amu) desorption products of NO dissociation from CO (28 amu) and
CO, (44 amu) contamination. Most of the adsorbed NO desorbed molecularly from the
87
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
surface, but -25% (as determined below) of the NO at monolayer coverage decomposes
during TPD to give mainly N, and O, desorption products, with a small amount of N2 0
desorption also detected.
NO, N2 and O, TPD curves following increasing NO exposures on the hex-Pt(lOO)
surface at 100 K are shown in fig. 5.2. Overall, this data is consistent with that reported
previously [15-17]. At low coverages, NO desorbs in a single peak at 430 K. This shifts
to higher temperature with increasing initial NO coverage and a high temperature shoulder
occurs at 500 K. The peak at 500 K does not grow until the main peak at 470 K saturates.
The same phenomena, i.e., shifting of the desorption peak to higher temperature and
nonsequential filling of desorption states with increasing coverage also occurred for
adsorbed CO on hex-Pt(100). This behavior was attributed to reconstruction of the hex
phase to the (lxl)-Pt(lOO) surface [25, 26]. At higher initial NO coverages, a low
temperature NO desorption peak appeared at 288 K which shifted to lower temperature
with increasing NO exposure. N, desorbed in a single peak at 465 K, with a low
temperature shoulder around 390 K. The peak at 465 K also shifted to higher temperature
with increasing NO exposures. Molecular oxygen did not initially desorb from the
surface, but eventually a peak was observed at 650 K along with a shoulder at 800 K.
Fig. 5.3 provides the NO adsorption kinetics and shows the changes in the nature of
reversible adsorption and reaction that occur with increasing NO coverage. The amounts
of NO that desorbed or decomposed into other products were constructed from TPD peak
areas after NO exposures on the hex-Pt(lOO) surface at 100 K. Following small NO
exposures, most of the oxygen formed by NO dissociation reacted with small amounts of
coadsorbed CO to form CO, and little oxygen desorbed from the surface. Some
reversible adsorption, and thus desorption of NO, occurred even at low coverages. At
monolayer coverage of NO, about 75% of the adsorbed NO desorbed from the surface
88
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
without dissociation. The NO coverage scale was referenced to the saturation coverage of
NO reported to be 9 n o= 0-73 ML on hex-Pt(lOO) at 100 K. This assumes 50%
reversible adsorption for 0N O = 0.5 ML which is associated with saturation of the NO
desorption peak at 470 K [16, 29]. This calibration can be checked independently by
referencing the O, TPD peak area from NO dissociation to that for an oxygen coverage of
0O = 2.2 ML [35] that was created by a saturation exposure of ozone (O.) on the Pt(100)
surface at 300 K. The results of this comparison were consistent.
TPD curves after a saturation exposure of NO on the c(2x2) alloy at 100 K are
shown in fig. 5.4. The main desorbed products were N2 0 (44 amu), NO (30 amu) and O,
(32 amu) in contrast to that on Pt(100). Small, contaminant CO and CO, desorption
peaks were identified by comparing all of the other TPD signals. Coadsorbed CO and
carbon which were present on the surface reacted with oxygen to produce reaction rate-
limited CO and CO-, desorption. Almost no N2 production from reaction of NO on the
alloyed surface was found.
Fig. 5.5 shows TPD curves for NO, N ,0 and 0 2 desorption products for increasing
exposures of NO on the c(2x2) alloy at 100 K. At low coverages, NO desorbed in a large
peak at 210 K and a smaller peak at 320 K. With increasing coverage, NO desorption
shifted to lower temperature, probably because of NO-NO repulsive interactions. At high
coverages, the peak at 320 K became relatively larger and a new peak appeared at 420 K.
N ,0 desorbed in a single peak at 150 K (all other, higher temperature peaks were from
CO, desorption). O, desorbed in a peak at 1030-1045 K. This high temperature state has
been assigned to the reduction of oxidized Sn in previous studies of the oxidation of these
Pt-Sn alloys [35]
TPD, as shown in fig. 5.6, was used after a saturation exposure of NO on the 3\72
alloy at 100 K to identify all desorbed products. As on the c(2x2) alloy, NO, N2 0 and O,
89
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
were the main desorbed species. While the amount of molecular NO adsorption was
lower on the 3V2 alloy than on the c(2x2) alloy, the fraction of adsorbed NO that desorbed
as NO was higher on the 3V2 alloy (vide infra).
Fig. 5.7 shows a series of NO, N2 0 and 0 2 TPD traces for increasing NO
exposures on the 3\/2 alloy at 100 K. For all of the products, the peak shapes were very
similar to those on the c(2x2) alloy, but the peak temperatures were slightly lower. This
indicates a weaker interaction with the surface due to the larger amount of surface Sn on
the 3V i2 alloy. This suggests a similar NO chemistry on these two alloys.
Uptake curves characterizing NO adsorption kinetics and the extent of reversible
adsorption are shown in fig. 5.8 (a) and 5.8 (b). NO TPD peak areas provide the amount
of reversible NO adsorption and this was found to be 0.12 and 0.09 ML on the c(2x2) and
the 3V2 alloys, respectively. O, TPD peak areas indicate 0O = 0.06 and 0.03 ML on the
c(2x2) and 3\/2 alloys, respectively, desorbs as O, at saturation NO coverage. On the
c(2x2) alloy, about 70% of the adsorbed NO at saturation coverage reacted to eventually
desorb N ,0 and O, and 30% desorbed molecularly. For the 3a/2 alloy, about 40% of the
adsorbed NO desorbed molecularly and 60% reacted to desorb N2 0 and O,. Alloyed Sn
reduced the monolayer saturation coverage for NO from 0 no= ^-73 ML on Pt(100) to
0 no= 0-35 and 0.22 ML on the c(2x2) and 3V2 alloys, respectively.
The initial slopes of the uptake curves are proportional to the initial sticking
coefficient (S0 ) of NO. Alloyed Sn decreases (S0 ) of NO from 0.8 on hex-Pt(lOO) [27] at
100 K to 0.16 and 0 .1 2 on the c(2x2) and 3V2 alloys, respectively.
Fig. 5.9 compares directly NO TPD curves for low 0N O (10% of 6 n o ) ^
saturation coverage (©no) on eac^ surface at 100 K. This figure illustrates more clearly
the other important effect of Sn, i.e., reducing the NO desorption temperature due to
90
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
weaker chemisorption bonding on the surface. Assuming first order desorption kinetics
and using Redhead analysis [36] we estimate the NO desorption activation energy at low
0N O to be 118 kJ/mol on Pt(100), 55 kJ/mol on the c(2x2) alloy, and 48 kJ/mol on the 3\^2
alloy. The 60-70 kJ/mole decrease in desorption energy corresponds to 50% weaker
adsorbate-surface bonding. Reductions in the monolayer saturation coverage and
desorption activation energy (30-40 kJ/mol) for NO chemisorption on the S n/P t(lll)
alloys compared to Pt(l 11) were also found in our previous studies [20].
We also investigated NO adsorption on a c(2x2)-Sn overlayer (0S n = 0.5), prepared
by depositing a thin layer of Sn on the hex-Pt(lOO) surface at 300 K and then annealing to
500 K. The results are shown in fig. 5.9. NO adsorption was completely eliminated on
the Sn adlayer-covered surface at 100 K, even though the surface Sn coverage and Sn-Sn
distance were identical to that for the c(2x2)-Sn alloy. This clearly illustrates that the site-
blocking ability of Sn depends on whether it is present as an adatom or alloyed atom.
5.3.2. HREELS
Vibrational spectra taken using HREELS are shown in fig. 5.10 for increasing
exposures of NO on hex-Pt(100) at 100 K. NO was molecularly adsorbed on the hex
surface at 100 K without decomposition at all coverages. At low coverage, a single vN O
peak occurred at 1676 cm' 1 which shifted to 1645 cm' 1 with increasing exposures. At
higher coverages, an additional band appeared at higher energy. The inset shows a
decomposition of the broad vN O peak obtained for saturation exposures of NO on Pt(100),
into two peaks at 1645 and 1756 cm'1 . These spectra are consistent with those reported
previously [15, 29]. The shift of vN O to lower energy with increasing NO exposures is
contrary to the behavior expected on the basis of dipole-dipole interactions [37]. IRAS
experiments by Gradner et. al. [29] showed that the 1676 cm' 1 band exhibited a positive
dipole shift but it was later replaced at higher NO coverages by a new band at lower
91
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
frequency. Several components appear at different stages during the hex— >(lxl) phase
change.
The assignment of NO adsorption sites using the vN O energy is problematic. A wide
variety of bonding modes for NO are possible because of the presence of the unpaired
antibonding electron in the 2k * orbital of NO, and the vN O energies for these various
species overlap strongly. A peak at 1676 cm' 1 has been assigned to NO adsorption at atop
sites in a bent configuration because it falls in-between the energies for linear-atop 1700-
2000 cm' 1 and bridge-bonded 1480-1545 cm' 1 sites [38]. The band at 1756 cm"1 was
assigned to NO bound at linear-atop sites [29], but it is not associated with (lxl)-terraces
because this band was not observed on the unreconstructed (Ixl)-Pt(lOO) surfaces [29].
Apparendy, it is also not associated with the hex-phase because the intensity of this band
increases with increasing exposures while the area of the hex phase decreases [29]. It has
been suggested that this band is due to NO adsorption at kink sites or other defect sites
formed at domain boundaries during reconstruction [29]. The two low energy peaks at
280 and 410 cm' 1 have been assigned as Pt-NO stretching and Pt-NO bending modes,
respectively, on the hex-Pt(100) surface [15, 20] All of these mode assignments are
shown in Table 5.1.
Fig. 5.11 shows the effect of annealing on HREELS spectra following a saturation
exposure of NO on hex-Pt(100) at 100 K. Heating to 200 K caused a decrease in the
intensity of the vN O peak at 1777 cm'1 , and this peak was eliminated by heating to 450 K.
The 1656 cm' 1 peak also shifted to 1586 cm"1 by heating to 450 K. All vN 0 loss peaks
disappeared after headng to 500 K. The value for vN O of 1586 cm' 1 is lower than the value
observed following a small NO exposure to give a similar NO coverage on a hex-Pt(100)
surface at 100 K. This indicates that the hex-Pt(100) surface has been reconstructed to a
(lxl)-structure upon heating to 450 K in the presence of adsorbed NO. A new peak at
92
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
540 cm' 1 appeared after heating to 400 K. This peak broadened and was shifted to 590
cm' 1 and reduced in intensity after heating to 800 K. We assign this peak to of
oxygen adatoms formed by decomposition of NO, which occurs for T > 400 K. The
broad nature of the peak suggests the presence of two kinds of oxygen atoms on the
surface. These data are consistent with the TPD results showing that NO and N , have
desorbed from the surface by 600 K and oxygen adatoms are almost fully removed from
the surface as O, by heating to 800 K.
HREELS spectra after increasing NO exposures on the c(2x2) alloy at 100 K are
shown in fig. 5.12. An exposure of 1L NO, produced a peak at 1636 cm' 1 and two peaks
in the lower energy range at 371 and 632 cm"1 . We assign the peak at 632 cm' 1 to a NO
bending mode that is dipole active when NO is adsorbed in a bent (non-linear or tilted)
configuration and the peak at 1636 cm' 1 to vN O for NO adsorbed at atop sites in a bent
geometry. The 371 cm"1 peak can be assigned to a Pt-NO stretching mode Vp,.^. A 3L
NO exposure increased the NO coverage and this broadened the vN O peak and caused a
shift of the vN O peak to 1806 cm' 1 and the NO bending mode to 672 cm'1 . A new peak
appeared at 1023 cm' 1 but we have no assignment for this peak. A similar peak was found
for NO adsorbed on C u (lll) [39] and assigned to a combination mode of (in this case)
the 371 and 672 cm' 1 vibrations. However, a highly inclined NO species, such as that
found on Rh(100) [40], could also give a vibrational loss peak at this energy. At
saturation NO coverage (6L), the Pt-NO stretching band at 361 cm' 1 became very intense
and the higher energy vN O peak grew and became the dominant contribution in the broad
peak at 1806 cm"1 The inset shows a decomposition of this 1806 cm' 1 peak into two peaks
at 1818 and 1694 cm'1 . The electron energy loss data alone do not distinguish among the
potential origins of the two vN O peaks: i) two different monomeric species; ii) vs NO and
v e N O of a dimer or a dinitrosyl, or iii) a combination of all of these. Some insight comes
93
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
from observation of peaks at 1365 and 863 cm'1 . The peak at 1365 cm' 1 falls in the range
1350-1545 cm' 1 that is typical for NO adsorbed in two-fold bridge sites. Though pure-Pt
two-fold short bridge sites do not exist on this c(2x2) alloyed surface, mixed Pt-Sn two
fold short bridge sites exist. This peak might also arise from NO in a bent configuration
on the pure-Pt two-fold long bridge sites. The peak at 863 cm' 1 was assigned to a bending
vibration of bent, bridge-bonded NO. NO bonded to two-fold bridge sites in a bent
configuration has been reported on Cu(l 11) [39]. We attribute the small peak at 2228
cm' 1 to the N-N stretching vibration of adsorbed N ,0. The N-N-O stretching vibration of
N ,0 would be expected to cause a peak at 1250 cm'1 , but this was not observed here. All
of the mode assignments for N ,0 are shown in Table 5.2.
Fig. 5.13 shows HREELS spectra after a monolayer of NO on the c(2x2) alloy was
heated sequentially to 1000 K. The spectrum taken after heating to 200 K indicates that
molecular NO was still present, adsorbed on the surface. The elastic peak intensity
decreased sharply at 200 K, indicating a disordered adlayer and possibly NO dissociation
or reaction. Heating the surface to 300 K eliminated the Pt-NO stretching vibration and
caused broad loss features near 460 and 570 cm' 1 to appear. This aids the interpretation of
the TPD spectra by showing that all of the molecularly adsorbed NO and NzO species are
desorbed or decomposed at this temperature. Further heating to 500 K nearly eliminates
the peak at 460 cm'1 . The loss peak at 570 cm' 1 is from Pt-O stretching vibrations,
because only oxygen adatoms remain at the surface. The elastic peak count rate from this
surface was quite high, consistent with a large chemical ordering change. Heating the
surface to 1000 K, desorbed all of the oxygen as 0 2, but also destroyed the c(2x2) alloy
structure.
A set of HREELS spectra after NO exposures on the 3V2 alloy at 100 K is shown in
fig. 5.14. At low NO exposures, a vN O peak was observed at 1666 cm'1 . This is higher
94
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
than that on the c(2x2) alloy and close to that on Pt(100). Similar behavior is found for
adsorbed CO on these two alloyed surfaces41 With increasing NO exposures, the 602
cm' 1 peak grew larger and the peak near 1736 cm' 1 broadened. At saturation coverage the
spectra was dominated by peaks at 351 and 1816 cm'1 , with smaller loss peaks at 580, 863,
1234 and 2198 cm'1 . Similar to the c(2x2) alloy, the inset shows the decomposition of the
broad peak at 1816 cm' 1 to two peaks at 1821 cm' 1 and 1698 cm'1 .
On the 3 /2 alloy, formation of N ,0 from the reaction of adsorbed NO occurs at 100
K. The peaks at 580,1234 and 2238 cm' 1 are assigned to the three normal modes of linear
N ,0 molecules: N-N-O bending (S^q), N-N-O stretching (vN N ,0) and N-N stretching
(yxx) modes respectively [42]. These mode assignments are shown in Table 5.2. This is
consistent with TPD showing the evolution of N2 0 and furthermore shows that this
reaction occurs at temperature as low as 100 K.
5.4. Discussion
Alloying the Pt(100) surface with Sn to form the c(2x2) and the 3 /2 Pt-Sn alloys
strongly effects the adsorption and reaction of NO on these surfaces. On hex-Pt(100) at
100 K and 300 K, NO is molecularly adsorbed and NO decomposition only occurs at
temperatures higher than 400 K [15-18]. On the (lxl)-Pt(lOO) surface at 300 K, NO is
partly dissociatively adsorbed on the surface and heating in TPD leads to desorption of
N,, O, and NO as the main reaction products [15-18]. Hence, surface structure plays an
important role in NO reactions on Pt surfaces. Alloying the Pt(100) surface with Sn to
form the c(2x2) and 3 /2 alloys produces a square lattice structure similar to that of (lx l)-
Pt(100) and these surfaces are also more reactive than hex-Pt(lOO). However, alloying
changed the selectivity for the products formed by thermal reactions of NO on the surface:
N ,0 rather than N,, along with NO and O, are the major desorbed species. The reactivity
was also enhanced on the c(2x2) and 3/2 alloys: NO adsorption was only 30-40%
95
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
reversible on the alloys compared to 75% on the hex-Pt(lOO) surface. In a related system,
on Pt-Rh alloys used as model catalysts for NO reduction, both N, and N2 0 were formed
as reaction products [43]. In those studies, the selectivity towards N, and N20 depended
on the relative concentration of NOa d s and on the surface and the reaction temperature.
The presence of surface-bound N2 0 formed from reaction of NO can account for
the peaks at 2228 and 2238 cm' 1 on the c(2x2) and 3/2 alloys, respectively, in the
HREELS spectra of a saturated NO layer along with two other peaks at 580 and 1234
cm' 1 on 3 /2 alloy. This conclusion is based on spectra following N2 0 adsorption on
Pt(l 1 1)4 2 and N2 0 formation from NO reactions on Ag(l 11) [44, 45] and Mo(l 10) [46].
All of the mode assignments for our spectra are shown in table 5.2. TPD data showed that
N ,0 was produced and desorbed from these two Pt-Sn alloys and HREELS shows
furthermore that N2 0 is formed at 100 K.
We consider below two possible mechanisms for N ,0 formation on the two
Sn/Pt(100) alloys. The first is a dissociative mechanism:
NOw - Nw + Ow (1)
N(a ) + NOw - N2Ofa ) (2)
2N« -* N2 ( g ) (3)
2 0 w - 0 2 ( g ) (4)
the second mechanism involves dimerization:
2NO( a ) - (NO)2 (a ) (5)
(NO)2 ( a ) - N20 (a ) + 0 ( a ) (6)
2 0 « ^ °2(g , C 7)
The dissociative mechanism may also desorbs N2 ( g ) depending on the relative rates of the
reactions (2) and (3).
9 6
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Alloying Sn into the Pt(100) surface causes a large reduction in the heat of
adsorption of NO, i.e., the Pt-NO bond strength. The desorption activation energy, which
is equal to the adsorption energy for non activated adsorption, decreases from 118 kJ/mol
on hex-Pt(lOO) to 55 kJ/mol on the c(2x2) alloy and 48 kJ/mol on the 3\/2 alloy. Hence,
it is unlikely that these alloys would be more reactive for dissociation of NO at lower
temperatures. TPD also showed no N, desorption from these two alloys which might
occur for a dissociative mechanism. Thus, we favor the dimerization reaction mechanism
for N2 0 formation on these two Pt-Sn alloys.
Several possible configurations of adsorbed NO at different sites are shown in fig.
5.15(a). At low coverages, NO is adsorbed with a tilted or bent geometry at atop sites on
hex-Pt(lOO) at 100 K [15, 18]. At high coverages on the Pt(100) surface, some NO is
bonded at atop sites in a linear geometry. We found a similar adsorption geometry of NO
at lower coverages on both of the Pt-Sn alloys. However, there was a large difference for
the vibrational spectra from the two Pt-Sn alloys at saturation coverage. Two peaks appear
in the vN O region at ~1695 and ~1820 cm"1 on the two alloys. These two modes can be
assigned as atop NO in bent and linear configurations, respectively, as was done
previously on Pt(100) surfaces at high coverages [15, 17, 18]. But, the frequencies and
intensity ratios of these two peaks are quite different on the alloyed surfaces than on
Pt(iOO), and it is worthwhile to consider other possibilities.
On both Ag(l 11) and Cu(l 11) surfaces at 90 K, a surface-bound NO dimer (NO),
is formed [45, 47]. The characteristic frequencies for the NO symmetric (vs NO) and
asymmetric (v^NO) stretching modes are 1863 cm"1 and 1788 cm"1 [43], Dimeric NO
species have also been observed with IR on supported catalysts containing tungsten,
chromium and molybdenum with typical values of vs NO at 1820 cm' 1 and v.ls NO at 1710
cm"1 [48]. These latter dimeric species, have been identified as dinitrosyls, in which two
97
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
NO monomers are bound to a single metal atom without N-N bonding that would be
present in an actual dimer, by analogy to polymeric dichlorodinitrosyl-molybdenum and
tungsten complexes which have similar v s NO and v^NO frequencies [48]. Schematic
representation of adsorbed dinitrosyl and dimeric species are presented in fig. 5.15(b). If
(NO), species have C,v symmetry, i.e. the two N-O bonds are tilted at equal angles from
the surface normal, the vJMO stretch will be dipole-forbidden [48]. But if the surface
dinitrosyl species has Cs symmetry, with the two N-O bonds having different tilt angles,
the v m( N O ) mode will be dipole-allowed. (However, this may still give a weak dynamic
dipole relative to vs NO and thus may be difficult to detect.) On Mo(llO), from isotopic
shifts in IR, a dinitrosyl intermediate with vs NO at 1821 cm' 1 was identified for N ,0
formation from NO reaction at low temperatures [46]. Mode assignments for the surface
dimer and dintrosyl species are given in Table 5.3.
In our spectra on the two alloyed surfaces, NO vibrations at -1820 cm' 1 and -1695
cm' 1 are significandy perturbed from that of (NO), dimeric species vs NO (1863 cm'1 ) and
vJ^ O (1788 cm"1 ) on the weakly interacting Ag(l 11) and Cu(l 11) surfaces [45, 47]. But
1820 and 1695 cm' 1 are very similar to the vibrations of a surface dinitrosyl complex with
vs NO (1820 cm '1 ) and va s NO (1710 cm'1 ). We propose that a surface dinitrosyl complex,
in which two NO molecules bound to one Pt site is an intermediate for low temperature
N ,0 formation on these two Sn/Pt(100) surface alloys. The intensity ratio of these two
peaks indicates that the two NO molecules in the dinitrosyl species are tilted with different
tilt angles from the surface normal and thus the v^NO mode is dipole-allowed. (A tilted
monomeric species at atop sites can also contribute to the 1695 cm"1 peak).
On the c(2x2) alloy, each Pt atom is surrounded by Sn-atom nearest neighbors.
Elimination of pure Pt two-fold and four-fold hollow sites also suggests formation of the
dinitrosyl species (two NO molecules bound to one Pt atom) as an intermediate for N2 0
98
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
formation from adsorbed NO. Hence, N, formation from the dissociation of NO requires
an ensemble of at least two Pt atoms which does not exist on these alloyed surfaces.
While a detailed IR study, including isotopic shift data, is required for an unequivocal
identification of a surface-bound dinitrosyl species, our data suggests that the absence of
adjacent strong bonding sites that exist for NO on pure Pt surfaces leads to the formation
of a dinitrosyl complex which subsequently decomposes to form N20 .
A NO stretching frequency of 1666 cm'1 , at low coverage on the 3V2 alloy is close to
that on the hex-Pt(lOO) surface and higher energy than on the c(2x2) alloy (1636
cm '1 ). This is surprising because ISS and AES indicate a higher Sn concentration on the
3Vi2 alloy [31,32]. Also the relative amount of molecular NO desorption (40%) is higher
on the 3V2 surface than that (30%) on the c(2x2) alloy. This could be due to a weaker Pt-
NO bond strength or some special arrangement of Pt and Sn atoms in the 3V2 alloy. In a
similar study of CO adsorption the vco frequency for the 3V2 alloy was closer to that on
Pt(100) than the vco on the c(2x2) alloy [41]. Both of these results are likely due to a
special arrangement of Pt and Sn atoms on the 3V2 alloy because benzene formation from
the cyclotrimerization of acetylene only occurred on the 3V2 alloy and not on the c(2x2)
alloy [49], This reaction is highly structure sensitive and indicates the formation of
"(11 l)-like" sites on the 3V2 alloy.
It is informative to compare the results for NO adsorption on these two Sn/Pt(100)
alloys to those on Sn/Pt(l 11) alloys [20]. The saturation monolayer coverage of NO on
P t(lll) at 100 K is 0.5 ML and this decreased to 0.4 ML on the (2x2)Sn/Pt(l 11) alloy
and 0.3 ML on the (VW 3)R30° S n /P t(lll) alloy at 100 K. On Pt(100) at 100 K
saturation monolayer coverage of NO is 0.73 ML and this decreased to 0.35 ML on the
c(2x2)Sn/Pt(100) alloy and 0.22 ML on the (3vr 2xvr 2)R45° Sn/Pt(100) alloy at 100 K [20,
29]. These results are shown in fig. 5.16 (a). We have normalized the monolayer coverage
99
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
to 1.0 for both Pt(l 11) and Pt(100). The Pt atom concentration in each surface layer is
also plotted on the right hand side and represented by the dashed line. Fig. 5.16 (b)
shows similar curves for S0(NO) on these surfaces at 100 K. These figures provide a
qualitative guide to the influence of alloyed Sn on NO chemisorption over a wide range of
Sn concentrations. The NO monolayer coverages and initial sticking coefficients are
reduced by alloyed Sn but not necessarily in proportion to the surface Sn concentration.
Similar results were obtained for CO [41,50] and CTL, [51,49] adsorption on Sn/Pt(100)
and Sn/P t(lll) alloys. Under many conditions, a simple Langmuirian site-blocking
model for the dependence of the saturation coverage and initial sticking coefficient on
modifier coverage fails. Non-linear behavior in Offo ® so can be understood if the
molecular size exceeds the ensemble-requirement for adsorption [52]. In such cases the
saturation coverage on the clean (unalloyed) surface is not determined by the adsorption
ensemble size but rather by the molecular size. An explanation for behavior of S0 has
been given which describes the importance influence of a “modifier precursor” state on
the adsorption kinetics [53, 54]. This factor allows for faster adsorption kinetics than
expected for a linear decrease in S0 with increasing modifier coverage.
Adsorption and reaction of NO on these S n/P t(lll) and Sn/Pt(100) surface alloys
are structure sensitive. The activity for decomposition and selectivity of the NO reactions
to produce N ,0 depends on the surface geometry and the availability of Pt-Pt nearest
neighbors on the surface. NO adsorption is completely reversible on Pt(l 11) [16, 19] and
S n/P t(lll) alloys [20],but substantial NO decomposition occurs on Pt(I00) [15-18] and
Sn/Pt(100) alloy surfaces (due to the higher heat of adsorption of NO on the (100) plane).
As a final comment that may be helpful in understanding the chemistry of NO, we
note that NO is isoelectronic with CO' and this additional electron in the anti-bonding 2 k *
orbital makes NO a weaker k acceptor ligand and a much more versatile ligand than CO.
100
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
For example, the adsorption energy of CO on the Sn/Pt(100) alloy surfaces is only
decreased by 12-16 kJ/mol compared to Pt(100) [41]. In contrast, NO is chemisorbed
more weakly by 60-70 kJ/mol on these two alloy surfaces compared to Pt(100). Also,
while the adsorption of CO on Pt(100) and the two Sn/Pt(100) alloys is completely
reversible, a substantial change occurs in the NO chemistry on Sn/Pt(100) surfaces due to
the presence of Sn. On Pt(100), about 25% of the NO monolayer decomposes to
eventually desorb N 2 and 0 2. In the presence of Sn on the two alloys, N.,0 formation
occurs from NO reaction at very low temperatures, even at 100 K. We believe that this
reactivity is due to the low temperature formation of a NO disitrosyl complex which
facilitates N ,0 formation.
55. Conclusion
Adsorption and reaction of NO on hex-Pt(lOO) and two Sn/Pt(100) surface alloys
was investigated using primarily TPD and HREELS. On hex-Pt(lOO), NO molecularly
adsorbed on the surface at 100 K and partially (25%) decomposed during heating in TPD
to eventually desorb N 2 and O,. On hex-Pt(100) at 100 K, NO mostly adsorbed at atop
sites in a bent configuration with some linearly adsorbed atop NO at higher coverages.
Alloying Sn into the Pt(100) surface to form the c(2x2) Sn/Pt(100) and the
(3V2xV2)R45° Sn/Pt(100) alloys with 0S n = 0.50 and 0.67, respectively, has large effects
on the adsorption and reaction of NO. Alloying completely changed the selectivity of NO
reactions to produce mostly N2 0 and O, as the reaction products. N2 0 formation occurs
at very low temperatures, even at 100 K. About 60-70 % of the adsorbed NO in the
monolayer follows this pathway on these alloyed surfaces. At low coverages on these two
alloys, NO adopts the same bent-atop configuration as on hex-Pt(100). However, at
higher coverages a surface dinitrosyl complex is formed which we believe an intermediate
to low temperature N20 formation. Alloyed Sn significantly reduced the NO desorption
101
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
energy, equal in this case to the adsorption energy or Pt-NO bond strength D(Pt-NO),
from 118 kJ/mol on hex-Pt(lOO) to 55 kJ/mol on the c(2x2) alloy and 48 kJ/mol on the
3 /2 alloy. The saturation coverage of NO decreased from 0.73 ML on Pt(100) to 0.35
ML on the c(2x2) alloy and 0.22 ML on the 3 /2 alloy at 100 K.
Acknowledgement
This work was supported by the Division of Chemical Sciences, Office of Basic
Energy Sciences, U.S. Department of Energy. We thank Dr. Hong He for his assistance
in the HREELS experiments.
102
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
REFERENCES
[1]. T. E. Fisher, S .R . Kelemen, J. Catal. 53 (1978) 24.
[2]. B. E. Hyden, Surf. Sci. 131 (1983) 419.
[3]. J. T. Rummer, J. Phys. Chem. 90 (1986) 4747.
[4]. J. K. Brown, A. C. Luntz, Chem. Phys. Lett. 204 (1993) 451.
[5]. G. V. Biinau, T. Wolff, Photochemie (VCH, Weinheim, 1987). p. 178.
[6]. M.Shelef,Catal. Rev. Sci.Eng. 11 (1975) 1.
[7]. K. C. Taylor, Catal. Rev. Sci. Eng. 35 (1993) 457.
[8]. H. Hirano, T. Yamada, K. Tanaka, J. Siera B. E. Nieuwenhuys, Vacuum 41
(1990)134.
[9]. P. Cobden B. E. Nieuwenhuys, Surf. Sci. 262 (1992) 97.
[10]. H. Lieske, J. Volter, J. Catal. 90 (1984) 96.
[11]. R. Srinivisan, R. J. DeAngelis B.H. Davis, J. Catal. 106 (1987)449.
[12]. Y. Debauge, M. Abon, J. C. Bertolini, J. Massardier, A. Rochefort, Appl. Surf. Sci.
90 (1995) 15.
[13]. A. E. Hamdaoui, G. Bergeret, J. Massadier, M. Primet, A. Renouprez,. J. Catal. 148
(1994) 47.
[14]. W. F. Banholzer, J. M. Gohndrone, G. H. Hatzikos, J. F. Lang, R. I. Masel,
Y. O. Park, K. Stolt,. J. Vac. Sci. Technol. A 3 (1985) 1559.
[15]. G. Pirug, H. P. Bonzel, H. Hopster, H. Ibach, J. Chem. Phys. 71 (1979) 593.
[16], R. J.G orte,L.D . Schmidt, J.L. Gland, Surf. Sci. 109(1981) 367.
[17], S. B. Schwartz, G. B. Fisher, L. D. Schmidt, J. Phys. Chem. 92 (1988) 389.
[18]. D. Yu. Zemlyanov, M. Smimov, V. V. Gorodetskii,. React. Kinet. Catal. Lett. 53
(1994)97.
[19]. H. Ibach, S. Lehwald, Surf. Sci. 76 (1978) 1.
[20]. C. Xu, B. E. Koel, Surf. Sci. 310 (1994) 198.
[21]. P. Heilmann, K. Heinz, K. Muller, Surf. Sci. 83 (1979) 487.
[22]. R. J. Behm, W. Hosier, E. Ritter, G. Binnig, Phys. Rev. Lett. 56 (1986) 228.
103
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
[23]. X.-C. Guo, A-Hopkinson, J. M Bradley, D. A. King, Surf. Sci. 278 (1992) 263.
[24]. M. A. Barteau, E. I. Ko, R J. Madix, Surf. Sci. 102 (1987) 99.
[25]. P. A. Thiel, R. J. Behm, P. R. Norton, G. Ertl, J. Chem. Phys. 78 (1983) 7437.
[26]. P. A. Thiel, R. J. Behm, P. R. Norton, G. Ertl, J. Chem. Phys. 78 (1983) 7448.
[27]. Y. Y. Yeo, L. Vattuone, D. A. King, J. Chem. Phys. 104 (1996) 3810.
[28]. P. Gardner, M. Tiishaus, R. Martin, A. M. Bradshaw, Vacuum 41 (1990) 304.
[29]. P. Gardner, M. Tiishaus, R. Martin, A. M. Bradshaw, Surf. Sci. 240 (1990) 112.
[30]. J. Radnik, F. Gitmans, B. Pennemann, K. Wandelt, Surf. Sci. 287/288 (1993)330.
[31]. M. T. Paffett, A. D. Logan, R. J. Simonson, B. E. Koel, Surf. Sci. 250 (1991) 123.
[32], Y. Li, B. E. Koel, Surf. Sci. 330 (1995) 193.
[33]. R. G. Windham, M. E. Bartram, B. E. Koel, J. Phys. Chem. 92 (1988) 2862.
[34]. M. Hoheisel, J. Kuntze,S. Speller, A. Postnikov, W. Heiland, I. Spolveri, U. Bardi,.
to be published.
[35]. N. A. Saliba, D. E. Beck, C. Baur, B. E. Koel, in preparation.
[36]. P. A. Redhead, Vacuum 12 (1962) 203.
[37]. A. Crossley, D. A. King, Surf. Sci. 68 (1977) 528.
[38]. F. A. Cotton, G. Wilkinson, G. Advanced Inorganic Chemistry 3rd ed. (Wiley, New
York, 1972) p. 713.
[39], S. K. So, R. Franchy, W. Ho, J. Chem. Phys. 95 (1991) 1385.
[40], J. S. Villarrubia, W. Ho, J. Chem. Phys. 87 (1987) 750.
[41]. C. Panja, B. E. Koel, Israel J. Chem. 38 (1998) 365.
[42]. N. R. Avery, Surf. Sci. 131 (1983) 501.
[43]. H. Hirano, T. Yamada, K I. Tanaka, J. Siera, P. Cobden, B. E. Nieuwenhuys, Surf.
Sci. 262 (1992) 98.
[44], S. K. So, R. Franchy, W. Ho, J. Chem. Phys. 91 (1989) 5701.
[45], W. A. Brown, P. Gardner, D. A. King, J. Phys. Chem. 99 (1995) 7074.
[46]. K. T. Queeney, C. M. Friend, J. Chem. Phys. 107 (1997) 6432.
104
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
[47]. A. Anderson, B. Lassier-Govers, Chem. Phys. Lett. 50 (1977) 124.
[48]. W. S. Millman, W. K. Hall,. J. Phys. Chem. 83 (1979) 427.
[49]. C. Panja, B. E. Koel, in preparation.
[50]. M. T. Paffett, S. C. Gebhard, R. G. Windham, B. E. Koel. J. Phys. Chem. 94 (1990)
6831.
[51]. C. Xu, J. W. Peck, B . E. Koel, J. Am. Chem. Soc. 115 (1993) 751.
[54]. B. E. Koel, To be published.
[53]. S. Johnson, R. D. Madix, Surf. Sci. 83 (1979) 487.
[54], L. Q. Jiang,, B. E. Koel, J. L. Falconer, Surf. Sci. 273 (1992) 273.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Q M S Intensity
NO/(5x20)Pt(100)
Saturation Dose
mass 12 (C)
x 20
mass 14 (N)
CO.
mass 30 (NO)
CO.
mass 32 (0 2)
mass 44 (N20+C02)
x 20
6 0 0 4 0 0 8 0 0 200
Temperature (K)
Fig. 5.1 TPD curves used to identify products from the reaction of a NO monolayer on
the hex-Pt(100) surface at 100 K.
106
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Reproduced w ith permission o f the copyright owner. Further reproduction prohibited without permission.
C O
C
C D
C M
00
C M
(0
(0
C O
N2 TPD A
NO (L) I
3 5 I f /
1 [L I
o.so K/i
0 2 5 l| nil
0.10
I T i 11 | 11 n t i ' m i | i i 11 | i m i | i m i
200 4 0 0 600
Temperature (K)
NO TPD
NO (L)
1111M11111111111
200 400 600
Temperature (K)
C M
" ■ jTffnTiTi 11111111111111111111
400 600 800 1000
Temperature (K)
Fig. 5.2 N2, NO and 0 2 TPD traces following NO exposures on the hex-Pt(lOO) surface.
o
-j
T P D Peak A rea (arb. units)
N O /P t(1 00) T = 100 K
0.8
Total Monolayer
0.7
0.6
NO
0.4
0.3
0.2
0.0
0.5 2.0 2.5 3.0 3.5 0.0 1.0 1.5
Exposure (L)
Fig. 5.3 Uptake curves constructed from TPD results detailing NO adsorption kinetics
on hex-Pt(lOO) at 100 K. The value for Offo on Pt(100) was set to 0.73 ML
[29].
108
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
NO/c(2x2)alloy
Saturation Exposure
TPD
mass 12 (C)
x 20
mass 14 (N)
CO C O
c
03
C
CO
mass 28 (N2+CO) co
a
mass 30 (NO)j
mass 32 (0 2) CO.
CO,
CO.
mass 44 (N20+C02)
1 200 8 0 0 1000 4 0 0 6 0 0 200
Temperature (K)
Fig. 5.4 TPD traces from a NO monolayer on the c(2x2)Sn/Pt(100) alloy at 100 K.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Reproduced w ith permission o f the copyright owner. Further reproduction prohibited without permission.
c o
c
0
c
O
Z
o
C O
C O
< 0
C O
NO TPD
c o
c
0
C -
O
C M
z
c o
c o
c o
I I I I I I I I I I I I I I I I I I II I
200 400
Temperature (K)
c o
c
0
o
CM
CO
C O
C O
C O
NO (L)
0 o TPD
3.5
0.50
0.25
0.10
600 100 200 300 400 500 400 600 800 1000
Temperature (K)
Temperature (K)
Fig. 5.5 NO, N2 0 and 0 2 TPD traces obtained after NO exposures on the c(2x2) Sn/Pt(100) alloy at 100 K.
o
Q M S Intensity
Saturation Exposure
NO/3V2 alloy
TPD
mass 12 (C)
x 20
mass 14 (N)
CO
mass 30 (NO)
mass 44 (N20+C02)
1000 1200 4 0 0 6 0 0 8 0 0 200
Temperature (K)
Fig. 5.6 TPD curves from a NO monolayer on the (3'/2xV2)R45°Sn/Pt( 100) alloy at 100
K.
Ill
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Reproduced w ith permission o f the copyright owner. Further reproduction prohibited without permission.
NO TPD
>
■ * — >
D
0)
c
< D
'co
c
CD
-t-j
c -
c
c
O
Z
pj—
o
C M
z
NO (L)
NO (L)
o
C O
0)
0)
0
2
NO (L)
0.5
0.5
0.5
1000
Temperature (K)
1200 800 200 4 0 0 6 00
2 0 0 4 0 0
Temperature (K)
6 0 0
Fig. 5.7 NO, N2 0 and 0 2TPD traces obtained following NO exposures on the (3\/2x/2)R 45nSn/Pt(100) alloy at 100 K.
to
T P D Peak A rea (arb. units) T P D Peak A rea (arb. units)
(a)
b 0.4
NO/c(2x2) alloy T= 100 K
Total monolayer
- 0.3
- 0.2
NO
b o.o
5 20 0 1 0 1 5
O
o
<
CD
0 3
C Q
CD
NO Exposure (L)
(b)
NO/3V2 alloy T=100 K
Total Monolayer
NO
20 1 5 0 5 1 0
L- 0 . 2 0
o
o
<
CD
0 3
C Q
CD
0
0.00
NO Exposure (L)
Fig. 5.8 Uptake curves obtained from TPD results for NO adsorption kinetics at 100 K
on (a) the c(2x2) Sn/Pt(100) alloy and (b) the (3V2xV2)R45°Sn/Pt(100) alloy.
113
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
(a)
NO TPD
Low Coverage
3a /2 alloy
0 3
O _
c(2x2) alloy
co
c(2x2) overlayer
Pt(1 00)
400 5 0 0 600 200 3 0 0 100
Temperature (K)
(b)
NO TPD
High Coverage
O D
C
03
3v^2 alloy c
O
z
c(2x2) alloy
c(2x2) overlayer o
co
03
0 3
c o
Pt(100)
600 5 0 0 400 200 3 0 0 100
Temperature (K)
Fig. 5.9 Comparison of NO TPD results from hex-Pt(lOO), the two alloys, and a c(2x2)
overlayer at two coverages of NO (a) 0N O = 0.10 anc^ (b) 9n o = 9 NO at 109 K
on each surface.
114
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
N 0 /P t(1 00)
HREELS
T= 100 K
1645
,1756
x 40
1400 1600 1800
1676
2000
1 7 5 6
C O
c:
( D
280
c
410
0.5 L
70
0.2 L
2500 1500 2000 0 500 1000
Loss Energy (cm 1 )
Fig. 5.10 HREELS spectra following NO exposures on the hex-Pt(100) alloy at 100 K
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Intensity
HREELS
N 0/P t(100)
Saturation D ose
x 20
Annealing Temp
590
800 K
600 K
1586
500 K
540
450 K
1 7 7 7 400 K
300 K
28 0
7 0
200 K
1656
41 0
100 K
2500 1000 1500 2 0 0 0 500 0
Loss Energy (cm’1 )
Fig. 5.11 HREELS spectra of NO after sequentially heating to higher temperature
following a saturation dose of NO on hex-Pt(lOO) at 100 K.
116
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Intensity
NO/c(2x2) alloy
HREELS
T= 100 K
x 50
361
1 6 9 4
1 6 0 0 220 ( 1 4 0 0 2000 1 8 0 0
1x1 50
6 9 2
1806
863
1 0 2 3
1 1 3 6 5 /
V J J
1 6 3 6
371
2 228
6 7 2
70
6 3 2
2000
3 0 0 0 2000 1000
Loss Energy (cm"1 )
Fig. 5.12 HREELS spectra following NO exposures on the c(2x2)Sn/Pt(100) alloy at
100 K.
117
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
L o s s Intensity
HREELS
NO/c(2x2) alloy
Saturation Dose
Annealing Temp
50
1000 K
5 7 0
46 0
500 K
i m
3 6 0
300 K
6 9 2
1 8 1 7
1 3 6 5
200 K
8 6 3
70
100 K
3 0 0 0 2000 0 1 000
L oss Energy (cm'1)
Fig. 5.13 HREELS spectra obtained after stepwise annealing a saturation dose of NO on
the c(2x2) alloy at 100 K.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Intensity
N O /3-/2 alloy-
HREELS _
x 50 T = 100 K
351
1821
1698
1400 1600 1800 2000 2200
1 8 1 6
5 8 0
8 6 3
1234
61 21
2 2 3 8
1 7 3 6
7 0
6 0 2
1 666
3 0 0 0 2000 1 000
L oss Energy (cm 1)
Fig. 5.14 HREELS spectra after NO exposures on the (3Vr 2xVi2) R45° Sn/Pt(100) alloy
at 100 K.
119
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
(a)
N
■ < / 1 1 / c
r / \ / _
o o oo 00
Linear atop Bent atop Bridged Bent Bridged
(b)
D O
Surface Dinitrosyl (NO) 2 Dimer
Fig. 5.15 Schematic representation of models for (a) monomeric NO species adsorbed at
different sites and (b) dimeric NO species adsorbed as surface dinitrosyl and
(NO)2 dimer configuration.
120
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
T=100 K
1 .0 -
g 0.8-
0 . 6 -
- 0.8
0 .4 -
Pt(111)
Pt(100) - 0.4
0 . 2 -
(2x2) V 3
c(2x2) 3V2
h 0.0 0.0 — I
0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7
o
o
O
CD
2 -
o '
= 3
o
3
C /J
o '
3
Sn Concentration (ML)
T=100 K
1 . 0 -
- x1 0
0 . 8 -
O)
• Pt(100)
- 0.4
(2x2) V 3 c(2x2) 3 / 2
I- 0 .0 0.0 — I
0.0 0.1 0.2 0.3 0 .4 0.5 0.6 0.7
2
o
o
Z 3
O
CD
Z 3
3
o '
o
o
3
w.
o
3
Sn Concentration (ML)
Fig. 5.16 Comparison of (a) normalized NO saturation coverage and (b) initial sticking
coefficient S0 of NO on Pt surfaces and several Pt-Sn surface alloys at 100 K.
The concentration of Pt atoms in the surface layer is shown with the dashed line
and scaled on the right hand side.
121
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Table 5.1. Vibrational energies (cm'1 ) of adsorbed NO for a saturation dose of NO on
Pt(100) and Sn/Pt(100) at 100 K.
Mode
Assignmnt
NO/Pt(100)
bent linear
atop atop
N0/Pt(100)
bent linear
atop atop
N0/Sn/Pt(100) alloys
bent linear
bridged atop atop
v (Pt-NO) 280 361
6 (Pt-NO) 410
5 (NO) 863 612-632
v (NO ) 1645 1757 1365 1636-
1666
1818-
1821
Table 5 2 . Vibrational energies (cm'1 ) of N2 0 for a saturation dose of NO on Sn/Pt(100)
alloy surfaces at 100 K.
Mode Assignment N2 0 (a /Pt(l 11) [42] Sn/Pt(100) alloy
v (Pt-N,0) 325 351-361
5 (NN-O) 575 580
v (NN-O) 1300 1234
v (N-NO) 2310 2198-2228
Table 5 3 . Vibrational energies (cm'1 ) of the dinitrosyl intermediate on Sn/Pt(100) alloys
after a saturation exposure of NO at 100 K.
Mode Assignment (NO), Dimer (I)
Ag/(fl 1) [44]
Dinitrosyl complex
(H)
Mo(l 10) [46]
Sn/Pt(100)
va (NO) 1788 1720 1694-1698
vs (NO) 1863 1821 1818-1821
122
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Chapter 6
Coking Resistance of Pt-Sn Alloys Probed by Acetylene
Chemisorption
Abstract
Acetylene (C,H2 ) is a reactive molecule with a low C:H stoichiometry that can be
used to evaluate aspects of the resistance of metal-based catalysts to the formation of
carbonaceous residue (coking). Herein we summarize our results for C,H, chemisorption
and thermal reaction on four well-defined, ordered surface alloys of Pt-Sn prepared by Sn
vapor deposition on Pt(100) and P t(lll) single crystals under UHV conditions. While
chemisorption of C2H, under UHV conditions on Pt is completely irreversible, i.e. thermal
decomposition leads to complete conversion of the chemisorbed monolayer into surface
carbon, alloying with Sn strongly reduces the amount of carbon thus formed. In addition,
the temperature for complete dehydrogenation of the carbonaceous residue formed from
acetylene decomposition (polymerization) is increased by up to 100 K, from 760 to 860
K. Both of these phenomena are consistent with observations of increased lifetimes and
decreased coking for technical Pt-Sn bimetallic catalysts compared to Pt catalysts used for
hydrocarbon conversion reactions.
123
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
6.1. Introduction
Alumina-supported Pt-Sn catalysts have been the subject of many basic studies
because of the importance of these catalysts in refining and petroleum chemistry,
particularly in reforming reactions [1-3]. While there is strong data indicating that Pt-Sn
alloy phases are not formed on the active reforming catalysts using A120 3 supports [4, 5],
the complexity of the materials systems causes difficulties in unambiguously establishing
the activity and identity of all the phases present on the catalyst under working conditions.
Indeed, Pt-Sn alloy formation has been given as an important contributor to improved
reforming selectivity and activity maintenance [6-8]. Furthermore, there are many useful
catalysts that have been reported to have good performance for a variety of reactions where
Pt-Sn alloy phases are fairly clearly identified as the metallic phase responsible for the
observed chemistry [9,10]. Thus it is useful to probe further the influence of alloyed Sn
on inhibiting carbon build-up on Pt surfaces in order to better understand the origin of
decreased coking for these bimetallic catalysts.
C,H, is a reactive molecule with a low H:C stoichiometry: a suitable “coke
precursor”. Adsorption of acetylene on a variety of low-index Pt surfaces has been
studied previously [11-17]. On Pt(100) [11] and P t(lll) [13-18] complete thermal
decomposition to surface carbon and H2 ( g ) occurs under UHV conditions. We found
previously that decomposition of acetylene on Pt(l 11) was reduced by alloying the surface
with Sn, reducing carbon formation on the surface by promoting reversible acetylene
adsorption and C-C bond coupling reactions leading to desorption of butadiene and
benzene [18]. Still, about 35 % of the adsorbed acetylene monolayer decomposed on the
(V3xV3)R30° Sn/Pt(l 11) alloy with 0S n = 0.33. About 10 % of the adsorbed acetylene
underwent cyclization to form benzene [18].
124
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Two alloys can also formed by vapor deposition of Sn on the Pt(100) surface, and
the structure and composition of these two alloys are different from those on Pt(l 11) [19-
22]. We report here several results of a more extensive investigation [23] of the
adsorption and reaction of acetylene on two ordered Sn/Pt(100) alloyed surfaces. We
compare these results to those on S n/P t(lll) alloy surfaces in order to further elucidate
effects of alloyed Sn on inhibiting carbon deposition on Pt surfaces.
6.2. Experimental Methods
All of the experiments were performed in a stainless UHV chamber described
elsewhere [24], which had a base pressure of 8x10'*1 torr and was equipped with a four-
grid low energy electron diffraction (LEED) optics, double-pass cylindrical mirror
analyzer (CMA) for Auger electron spectroscopy (AES), ion sputtering gun, UTI-100C
quadrupole mass spectrometer (QMS) for temperature programmed desorption (TPD),
and various metal and gas dosing facilities.
The Pt(100) crystal was cleaned by repeated cycles of Ari ion sputtering and
oxidation at 1200 K with PO 2=5xl0'7 torr. The surface purity and order were checked by
AES and LEED. The crystal could be resistively heated to 1200 K and cooled by using
liquid nitrogen to 95 K. The temperature was measured by a chromel-alumel
thermocouple spot-welded to the side of the crystal. All of the TPD experiments were
done utilizing a linear heating rate of 4 K/s. d,-Acetylene (C2D2 ) (Cambridge Isotope
Laboratories, 99% purity) was used as received and dosed onto the crystal through a
Varian ® leak valve connected to a multichannel array gas doser positioned directly in
front of the crystal surface.
Two ordered surface alloys can be formed on P t(lll) following Sn deposition and
annealing as previously reported [19, 20]. These are denoted as the (2x2) and the
(1 /3xV, 3)R30° S n/P t(lll) surface alloys with 0S n = 0.25 and 0.33 ML, respectively. On
125
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Pt(lOO), two different alloyed surfaces can be formed by a similar procedure. These two
alloys are denoted as the c(2x2) and the (3-/2xV2)R45° Sn/Pt(100) surface alloys with 0S n
= 050 and 0.67, respectively [21,22]. All of these surface alloys are quite “flat”, with an
outward buckling distance for Sn (above the Pt surface plane) of ~ 0.20 A. The c(2x2)
surface alloy has a geometric structure resembling a “checker board” , with one-half of the
lattice positions each occupied by Sn and Pt and where each surface Pt atom has only Sn
nearest neighbors and vice versa [22], Low energy alkali ion scattering spectroscopy
(ALISS) initially suggested that the (3V2xV2)R45° Sn/Pt(100) surface alloy had a
geometric structure similar to that of the c(2x2) alloy, and was composed of small c(2x2 )
alloy domains with the same buckling distance within the domains [22]. STM studies in
our lab are underway to further probe the (3V2xV2)R45° Sn/Pt(100) alloy and the
relationship that it has to the structure of the bulk Pt3Sn(100) surface. At this time, the
(3V'2xVr 2)R45°-Sn/Pt(100) alloy structure has not been fully determined. For brevity in
the rest of this paper, the (2x2)-Sn/Pt(l 11) and the (i/3x1 /3)R30o -Sn/Pt(l 11) surface
alloys will be referred to as the (2x2) and the V3 alloys and the c(2x2)-Sn/Pt(100) and the
(3Vr 2xVr 2)R45°~Sn/Pt(100) alloys will be referred to as the c(2x2) and the 3V2 alloys,
respectively.
Acetylene, benzene, and butadiene TPD curves on both S n /P t(lll) and Sn/Pt(100)
surface alloys were normalized using the results reported previously for S n/P t(lll) and
coverage calibrations on Sn/Pt(100) using XPS and TPD as reported elsewhere [23].
6.3. Results and discussion
Thermal desorption of molecular acetylene in TPD from Pt and Sn/Pt surface alloys
are compared in fig. 6.1. On clean hex-Pt(100) [23], irreversible adsorption of C2D2
occurs; D2 and a small amount of ethylene (C2D4 ) desorb from the surface, and carbon is
left on the surface. This is consistent with previous studies [11]. XPS can be used to
126
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
quantify the extent of irreversible adsorption, and we find that about 90% of the
chemisorbed acetylene monolayer undergoes decomposition on Pt(100) [23].
The amount of C,H, desorption from the chemisorbed monolayer is increased by
alloying Sn into the Pt(100) surface. This phenomenon was predictable based on our
previous studies of S n /P t(lll) surface alloys [18]. Only 2% of an acetylene monolayer
desorbed molecularly from clean P t(lll), but increasing amounts of Sn in the (2x2) and
V3 alloys increased the molecular desorption of acetylene.
Acetylene is more weakly chemisorbed on Pt-Sn alloys than on Pt surfaces. In both
sets of S n /P t(lll) and Sn/Pt(100) surface alloys, a higher Sn concentration shifted the
desorption temperature of acetylene to lower temperatures. However, C2D, is apparently
more strongly bound on the c(2x2) Sn/Pt(100) alloy with 0S n = 0.5 ML than on the (2x2)
S n /P t(lll) surface with 0S n = O.33 ML.
Fig. 6.2 shows the D2 evolution in TPD after saturation doses of cf-acetylene on
several surfaces. Measurements of D2TPD peak areas were used to quantify the amount
of acetylene decomposition that occurs during TPD. Compared to that on P t(ll 1), there
was only 72% of the amount of acetylene decomposition on the (2x2) alloy and only 35%
decomposed on the V3 alloy. This effect on Sn/Pt(100) alloy is larger with higher Sn
concentration, i.e., the decomposition of acetylene is decreased to only about 7% on the
c(2x2) and 5% on the 3V2 alloy surfaces compared to that on Pt(100). It is not possible
for us to guarantee that these small amounts of D2 do not originate from defect sites. Very
small amounts of carbon was found in AES following TPD which is fully consistent with
these results. Also, the temperature for complete dehydrogenation of the carbonaceous
residue formed from acetylene decomposition (polymerization) is increased by up to 100
K, from 760 on clean Pt surface to 860 K on the 3V2 Sn/Pt(100) alloy.
127
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
TPD curves for benzene desorption resulting from a saturation dose of acetylene on
these surfaces are shown in fig. 6.3(a). No benzene was desorbed on Pt(l 11) or Pt(100)
surfaces. Acetylene is strongly rehybridized on clean Pt surfaces and, under UHV
conditions, the only reaction pathway is decomposition to carbon and H, [11-16].
However, benzene desorption is observed following acetylene adsorption on both
S n /P t(lll) surfaces, and about 10% of the adsorbed acetylene monolayer converted to
benzene on the V3 alloy. Alloyed Sn weakens Pt-C2D, bonding, decreases C2D2
rehybridization, and inhibits C2D2 decomposition. This ultimately leads to C-C bond
coupling to give C4 and C6 surface species [18].
If the only factor involved in benzene formation was weaker chemisorption, then one
would expect a higher amount of benzene formation on the c(2x2) Sn/Pt(100) alloy than
that on the (2x2) Sn/Pt(l 11) alloy. However, no benzene signal in TPD was detected from
the C2D2 monolayer on the c(2x2) Sn/Pt(100) alloy. Obviously, some other factor such as
surface geometry plays an important role in benzene formation and desorption.
On the c(2x2) Sn/Pt(100) surface alloy, only a single Pt atom ensemble exists (each
Pt atom has only Sn nearest neighbors). Hence, isolated Pt atoms with C4 v symmetry do
not have acetylene cyclization activity under these conditions and apparently an ensemble
of at least two Pt atoms is required for cyclotrimerization. However, given the low
cyclotrimerization activity of Pd(100) [20], it could be that a surface with C4 v symmetry
may not be favorable for this reaction and isolated Pt atoms with another surface
symmetry may be active. Surprisingly, about 15% of the adsorbed acetylene monolayer
undergoes cyclization and desorbs as benzene [23] on the 3V2 alloy with 0S n = 0.67. This
amount was estimated from the XPS C(ls) intensity and benzene TPD peak areas. This
implies that sites are present on this alloy in which Pt atoms are exposed at the surface in
an ensemble which has locally C3 or C6 symmetry, perhaps with facets like the Sn/Pt(l 11)
128
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
surface alloys. Support for this conjecture comes from a related study of CO on the two
Sn/Pt(100) alloys, in which it was observed that the 3 /2 alloy showed more "Pt-Iike"
chemistry than the c(2x2) alloy [25].
On the S n /P t(lll) alloys, a C4 product desorbed as shown in fig. 6.3(b). A
mechanism has been proposed for benzene formation that includes a C4 metallopentacycle
intermediate [18]. The gas evolved in TPD is presumably butadiene that results from
hydrogenation of a C4 intermediate, but we can not rule out a cyclobutene product from
our limited data. This is an interesting contrast to the results for the two Sn/Pt(100) alloys,
where we monitored all possible C,, C4 and C6 products but detected no butadiene or any
other hydrocarbons desorbing during TPD. The addition of a third acetylene molecule to
the C4 intermediates must be strongly favored over hydrogenation of the C4 intermediate
on the 3/2 Sn/Pt(100) alloy compared to that on the S n/P t(lll) alloy surfaces. This
result probably arises simply from a lack of surface deuterium because of the reduced
dehydrogenation activity on the 3 /2 Sn/Pt(100) alloy.
Fig. 6.4(a) provides a comparison of the relative coverages of acetylene in the
chemisorbed monolayer on these surfaces at 100 K. On P t(lll), the saturation coverage
of acetylene has been reported to be 0.25 ML (3.8 x 101 4 molecules/cm2 ) at 100 K [18,
26]. On this basis, the saturation CjH, coverage decreases to 0.17 ML on the (2x2) alloy
and 0.16 ML on the /3 alloy at 100 K [18]. On Pt(100), the acetylene monolayer
coverage has been reported to be 0 5 ML (6.5 x 101 4 molecules/cm2 ) at 100 K [11, 12, 23].
Using these references, the monolayer coverage decreases to 0.32 ML on the c(2x2) alloy
and 0.25 ML on the 3 /2 alloy at 100 K. The monolayer coverage is unexpectedly
different on these two Pt surfaces, and there is some controversy over the monolayer
coverage of acetylene on clean Pt(l 11) [18, 26-28]. This has been proposed to be either
0.25 ML [18,26], or0.5 ML [27, 28]. In fig. 6.4(a) we have also indicated the saturation
129
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
monolayer coverage calculated from a closest packing model (0c p ) with Van der Wall
contacts using the C,D, gas phase structure. This coverage, 0c p = 5 .7 x l0 1 4 molecules/cm2
(0.44 ML on Pt(100) and 0.37 ML on P t(lll)), sets a reasonable upper limit for the
number of molecules in the chemisorbed layer. The coverage on Pt(100) has been
reported to be slightly higher than 0c p which can be explained by uncertainty in the
experimental determination or perhaps by the fact that we used gaseous C2H2 molecular
structure for calculation of 0c p and formation of either sp3 or sp2 hybridized species upon
chemisorption can lead to slightly higher coverages. The C2H2 monolayer coverage
decreases 20-25 % upon alloying 0.25-0.33 ML Sn into the P t( lll) surface. There is a
larger decrease of 40-60 % upon alloying 0.50-0.67 ML Sn into the Pt(100) surface.
Such changes in the coverage of adsorbate due to site-blocking by surface modifiers is
often modeled by using a Langmuirian adsorption type expression. This expression can
be written as 0 C2D2= ( l - a 0 Sn )b, where a is the number of adsorption sites that are blocked
by one modifier atom and b is the number of adsorption sites required for adsorption. For
b=l, both of the dashed lines in fig. 6.4(b) give the same value of a = 0.35 for Sn/Pt alloy
surfaces.
Fig. 6.4(b) plots the desorption activation energy of acetylene on these surfaces,
calculated using Redhead analysis. Because adsorption is not appreciably activated these
energies correspond to adsorption energies. The adsorption energy of acetylene on clean
P t( lll) and Pt(100) surfaces can be estimated by extrapolating the adsorption energy
curves to zero Sn concentration. We note that these values for acetylene adsorption
energy on (111) and (100) Pt surfaces agrees well with the "sp2 -5 " adsorption energy of 54
kcal/mol predicted from sp3(61 kcal/mol) and sp2 (46 kcal/mol) C2 species as estimated
using the QVB theory by Koel and Carter [29] and a Pt-C covalent a-bond strength of
D(Pt-C) = 53 kcal/mol. Alloyed Sn weakened the chemisorption bond strength of
130
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
acetylene by 45-65% on these surfaces and the acetylene desorption activation energy
decreases concomitantly with increasing surface Sn concentration. However, the C2D,-Pt
bond strength on c(2x2) alloy on Pt(100) with 0S n = 0.5 ML is apparently higher than that
on the (2x2) alloy on Pt(l 11) with 0S n = 033 ML.
The amount of acetylene decomposition (after normalizing to the saturation
monolayer coverage) on each o f the surfaces studied is shown in fig. 6.5. The Pt atom
surface concentration is plotted as a dashed line on the right hand side as given by the
scale. On the Sn/Pt(l 11) alloys, 72 and 35% of the acetylene monolayer decomposed for
0S n= 0.25 and 0.33, respectively, alloyed in the Pt(l 11) surface. On the Sn/Pt(100) alloys,
acetylene adsorption is largely reversible and 90% of the acetylene monolayer reversibly
desorbed on the c(2x2) alloy. On the (2x2) Sn/Pt(l 11) alloy, three-fold pure-Pt sites are
present but there are no two adjacent three-fold pure Pt sites. Only two-fold bridge and
atop sites are present on the /3 alloy, with all of the pure Pt three-fold sites eliminated, but
still about 35% of acetylene monolayer decomposed on this surface. On the c(2x2) alloy,
each surface atom of Pt is isolated, i.e. there are no Pt-Pt nearest neighbors, and a
“checker board” pattern of Pt and Sn is formed. The strong suppression of
decomposition of acetylene on both Sn/Pt(100) alloys can be attributed to an ensemble
requirement of at least a pure Pt two-fold bridge site to stabilize the transition state or
reaction products of acetylene decomposition.
6.4. Conclusions
C,H, is a reactive molecule with a low H:C stoichiometry used to model coking
reactions on Pt and Pt-Sn alloy surfaces. Sn reduced the C2H2 chemisorption bond
strength and reactivity for dissociative adsorption of C2H2 compared to Sn-free Pt
surfaces. Also, the temperature for complete dehydrogenation of the carbonaceous residue
formed from C 2H 2 decomposition was increased by up to 100 K. Gaseous C4 and C 6
131
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
products were formed from C-C bond coupling reactions of chemisorbed C2H2 in UHV.
Acetylene chemisorption on these Sn/Pt alloyed surfaces is relatively structure-insensitive,
but acetylene reaction and decomposition is structure-sensitive. Our results indicate that a
single Pt atom is sufficient to chemisorb acetylene, but an ensemble of at least two Pt
atoms is required for decomposition activity. C2D2 cyclotrimerization reactions on the
3V2 alloy surface indicate that reconstruction to produce this alloy leads to the formation
of Pt sites that have a local C, symmetry.
Acknowledgment
This work was supported by the Division of Chemical Sciences, Office of Basic
Energy Sciences, U.S. Department of Energy.
132
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
REFERENCES
[1
[2
[3
[4
[5
[6
[7
[8
[9
[10
[11
[12
[13
[14
[15
[16
[17
[18
[19
[20
[21
[22
[23
. E. Marlen, P. Beccat, J. C. Bertolini, P. Delichere, N. Zanier and B. Didillon, J. Catal.
159 (1996) 178.
. Y. Zhou and S. M. Davis, Catal. Lett. 15 (1992) 51.
. Y. X. Li, K. J. Klabunde and B. H. Davis, J. Catal. 128 (1991) 1.
. G. Meitzner, G. H. Via, F. W. Lytle, S. C. Fung, and J. H. Sinfelt, J. Phys. Chem. 92
(1988) 2925.
. R. Srinivasan, L. A. Rice, and B. H. Davis, J. Catal. 129 (1991) 257.
. H. Lieske and J. Volter, J. Catal. 90 (1984) 96.
. R. Srinivasan, R. J. De Angelis, and B. H. Davis, J. Catal. 106 (1987) 449.
. K. Balakrishnan, and J. Schwank,J. Catal. 132(1991)451.
. R. D. Cortright and J. A. Dumesic, J. Catal 148 (1994) 771.
. R. D. Cortright and J. A. Dumesic, J. Catal 157 (1995) 576.
. T. E. Fischer, S. R. Kelemen, and H. P. Bonzel, Surf. Sci. 64 (1977) 157.
. T. E. Fischer and S. R. Kelemen, J. Vac. Sci. Technol. 15 (1978) 607.
. H. Ibach and S. Lehwald, J. Vac. Sci. Technol. 15 (1978) 407.
. L. L. Kesmodel, L. H. Dubois and G. A. Somoijai, J. Chem. Phys. 70 (1979) 2180.
. M. Salmeron and G. A. Somoijai, J. Phys. Chem. 86 (1982) 341.
. C. E. Megiris, P. Berlowitz, J. B. Butt, and J. J. Kung, Surf. Sci. 159 (1985) 184.
. N. A. Avery, Langmuir. 4 (1988) 445.
. Xu, C.; Peck, J. W.; Koel, B. E. J. Am. Chem. Soc. 115 (1993) 751.
. M. T. Paffett, and R. G. Windham, Surf. Sci. 208 (1989) 34.
. S. H. Overbury. D. R. Mullins, M. T. Paffeett, and B. E. Koel, Surf. Sci. 254 (1991)
45.
. M. T. Paffett, A. D. Logan, R. J. Simonson, and B. E. Koel, Surf. Sci. 250 (1991)
123.
. Y. Li, and B. E. Koel, Surf. Sci. 330 (1995) 193.
. C. Panja, N. A. Saliba and B. E. Koel, in preparation.
133
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
[24]. R. G. Windham, M .T. Bartram and B .E. Koel, J. Phys. Chem. 92 (1988) 2862.
[25]. C. Panja and B . E. Koel, Israel J. Chem. 38 (1998) 365.
[26]. M. Abon, J. Billy and J. C. Bertolini, Surf. Sci. 171 (1986) L387.
[27]. N. Freyer, G. Pirug, and H. P. Bonzel, Surf. Sci. 125 (1983) 327.
[28]. N. Freyer, G. Pirug, and H. P. Bonzel, Surf. Sci. 126 (1983) 487.
[29]. B. E. Koel and E. A. Carter, Surf. Sci. 226 (1990) 339.
134
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Saturation Exposure of C2D .
3V2 alloy
= 3
E
c o
co
CM
c(2x2) alloy
CO
c
CD
V3 alloy C
(2x2) alloy
Pt(100)
Pt(111)
x 10
700 300 400 600 1 00 200 500
Temperature (K)
Fig. 6.1 cU-Acetylene (C2D2 ) TPD curves after saturation cL,-acetylene exposures on Pt
and Pt-Sn alloys at 100 K.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Saturation Exposure of C2D .
3v^2 alloy
c(2x2) alloy
CO
^3 alloy
(2x2) alloy
Pt(100)
Pt(111)
200 300 400 500 600 700 800 900
Temperature (K)
Fig. 6.2 D2 evolution in TPD after saturation exposure of d2 -acetylene on Pt and Pt-Sn
alloys at 100 K.
136
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Intensity (60 amu) Intensity (84 amu)
Saturation Exposure of C2D ,
3^2 alloy
■w w ill
c(2x2) alloy
a /3 alloy
(2x2) alloy
Pt(100)
500 400 200 600 300 700
Temperature (K)
Saturation Exposure of C2D 2 c4D 6 tpd
3-72 alloy
c(2x2) alloy
v"3 alloy
(2x2) alloy
Pt(111)
Pt(100)
200 300 400 500 600 700
Temperature (K)
Fig. 6.3 C6D6 (d6-benzene) and (b) C4D6TPD traces after acetylene saturation exposures
on these surfaces at 100 K.
137
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
0.6
CM
C^H2 hep monolayer
0.5
CO
0.4
j c + c j
XC+CT
0 . 2 -
(100) (2 x 2) / 3 c ( 2 x 2 ) 3 /2
0.0 0.0
Sn Concentration ( atoms/cm2)
o
o
Z 3
O
CD
03.
o '
o
o
o
3
C O
o'
3.
60-3
50
o 40H
| 3 ( H
j* :
* 2 0 -3
sp
\ Sp
2 . 5 .
sp
A v
\
iff
1 O d
0 “ i 1 r
(2x2) / 3 c(2x2) 3 / 2
A , f ,C \ 4 ,
o .o 0.5 1.0 1.5
15 2
Sn Atom Concentration (x 10 atoms/cm )
Fig. 6.4 (a) Comparison of acetylene monolayer coverage at 100 K. (b) Influence of
alloyed Sn on the acetylene desorption activation energy of the most strongly
chemisorbed state, which is essentially the adsorption energy of the most
strongly bound state on each of the form ordered Pt-Sn alloys. Values given
by the brackets for clean Pt(100) and P t( lll) surfaces from predictions
made in ref. [29].
138
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
1 00
8 0 -
Q.
4 0 -
2 0 -
( 100 )
(2x2) / 3
3 / 2
1 .0 0.6 0.8 0.4 0.0 0.2
15 2
Sn Atom Concentration (x10 atom s/cm )
Fig. 6.5 Fractional decomposition of acetylene in the monolayer (normalized to the
saturation monolayer coverage on each surface) on the four S n/P t(lll) and
Sn/Pt(100) surface alloys.
139
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Chapter 7
Acetylene Chemisorption on Sn/Pt(100) Alloys
Abstract
Adsorption and reaction of acetylene on a hexagonally reconstructed (5x20)-Pt(l00)
surface and two ordered Sn/Pt(100) alloy surfaces was investigated using temperature
programmed desorption mass spectrometry (TPD), Auger electron spectroscopy (AES),
low energy electron diffraction (LEED) and X-ray photoelectron spectroscopy (XPS).
Vapor deposition of Sn on a Pt(100) single crystal was used to form two Sn/Pt alloys, the
c(2x2) and (3/2xVr 2)R45°Sn/Pt(100) structures with 0S n = 0.5 and 0.67 ML, respectively,
depending on the initial Sn concentration and annealing temperature. Acetylene nearly
completely decomposed during TPD on Pt(100) in the absence of Sn, forming hydrogen
which desorbs as R, and surface carbon. This decomposition associated with irreversible
dissociative adsorption was strongly suppressed on the two Pt-Sn alloy surfaces and a
large acetylene desorption peak in TPD was observed. Additionally, 15% of the adsorbed
acetylene monolayer was converted to gaseous benzene during TPD on the (3Vr 2xA/2)R45°
Sn/Pt(100) alloy, but no such benzene desorption occurred on the c(2x2) alloy. Alloyed
Sn in the c(2x2) alloy decreased the initial sticking coefficient of acetylene on Pt(100) at
100 K by -40%, but additional Sn in the other alloy had no additional effect. The
140
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
saturation coverage of C,D, in the chemisorbed monolayer at 100 K was decreased from
that on Pt(100) by 35% on the c(2x2) alloy and 50% on the (3\^2xV2)R45° Sn/Pt(100)
alloy. The acetylene chemisorption bond energy, estimated by the acetylene desorption
activation energy measured in TPD, also decreased as the alloyed Sn concentration
increased. Multiple C2D2 TPD peaks for desorption from both the c(2x2) and the
(3''/2xa/2)R45° Sn/Pt(100) alloy surfaces indicate either several energetically
distinguishable adsorption sites for acetylene on these surfaces or the rate-limiting
influence of more complex surface reactions.
7.1. Introduction
Acetylene (C2H2 ) chemisorption and reaction on Pt and other transition metal
surfaces has been studied many times, mainly to probe surface reactions related to
heterogeneous catalysis. P t(lll) and Pt(100) surfaces are known to be highly reactive
towards acetylene decomposition, and the formation of hydrogen and adsorbed carbon on
the surface is the only reaction pathway observed in UHV studies [1-4]. Such a high
reactivity, which leads to non-specific carbon build-up, is not desired in most industrial
reactions and so commercial hydrocarbon conversion catalysts often utilize bimetallic Pt-
based catalysts containing a second metal component to modify (reduce) the reactivity of
Pt [5, 6]. Surface science studies of Pt-Sn alloys have also shown consistent results
indicating that hydrocarbon dehydrogenation rates on Pt-Sn alloys were much slower than
on clean Pt [7-9].
As a highly reactive molecule with a C:H stoichiometry of unity acetylene can serve
as a prototype for reactions of “coke precursors” . Also, C,H 2 can undergo a C-C bond
coupling reaction (under UHV conditions) that of benzene formation from the
cyclotrimerization. Benzene desorption occurs only on Pd(l 11) [10-13] and Cu(l 10) [14,
141
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
15] pure-metal surfaces. Lambert and coworkers [10-12] elucidated that the mechanism
first involves dimerization to form a C4H4 metallopentacycle intermediate, which then
forms benzene with a third C2H2 molecule in an associative mechanism without cleavage
of any C-C or C-H bonds [10-12]. The clean P t(lll) surface [1-3] does not desorb
benzene from this reaction. On N i(lll), there is evidence of benzene formation on the
surface at high C2H2 coverages [16], but it decomposes during heating to desorb H2 and
leave carbon on the surface.
Acetylene cyclotrimerization on Pd single crystal surfaces is sensitive to both the
atomic geometry and the electronic structure of the metal surface. For example, benzene
formation is more efficient on Pd(l 11) than on Pd(l 10) [17]. More recently, this reaction
was found to be effectively carried out on reduced TiO2(001) [18] and several bimetallic
alloys [19-22]. Benzene formation from cyclotrimerization of acetylene has now been
found on Sn/Pt(111) [19], Au(l 11)/Pd [20], Pd/Au( 111) [21 ], and Pd(111)/Sn [22].
The two ordered Pt-Sn surface alloys formed on P t(lll) that were studied
previously [23,24] havec(2x2) and 0/3xV3)R45° unit cells and 0S n = 0.25 and 0.33 ML
in the outermost layer, respectively. Decomposition of acetylene is strongly suppressed
by alloying Sn in the surface layer of pt( 111) and benzene desorption was observed on
both alloys [19]. On the V3 alloy, which desorbed more benzene than the (2x2) alloy,
33% of the adsorbed acetylene monolayer decomposed to form carbon and hydrogen on
the surface and 10% of the adsorbed acetylene cyclotrimerized to form benzene (under
those TPD conditions). On P d (lll), a relatively weak interaction of C2H2 preserves the
alkyne bond character, in contrast to the strong rehybridization on Pt, Rh and Ir and this
contributes to benzene formation [10-13]. Sn in the two Pt-Sn alloys studied thus far
weakens the acetylene-surface interaction and decreases the activity for decomposition, and
thus facilitates benzene formation [19].
142
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
C2D, adsorption measurements were carried out on a hexagonally reconstructed
(5x20)-Pt(100), referred to as the “hex” phase [27]. Two additional ordered Pt-Sn alloys
can be formed by heating Sn films deposited on Pt(100) single crystals. These have
c(2x2) and p(3/2xV2)R450 unit cells and 0S n = 0.5 and 0.67 ML, respectively in the
topmost layer [25, 26]. Both of these two alloys have a higher surface Sn concentration
than the two previous surface alloys on Pt(l 11). On the c(2x2) Sn/Pt(100) alloy, each Pt
atom has only Sn atom nearest neighbors, forming a “checker-board" pattern of isolated
Pt atoms surrounded by Sn atoms. This alloy can be used to probe the reactivity of single
Pt-atom reactive ensembles. The 3 /2 alloy has a similar, but unknown structure.
In this paper, we report studies of the adsorption of acetylene on a clean hex-Pt(lOO)
surface and the two Pt-Sn ordered surface alloys formed on Pt(100) described above. We
probe the influence of alloyed Sn in (i) suppressing the decomposition of acetylene and
thereby reducing carbonaceous accumulation, and (ii) the efficiency of gaseous benzene
formation by cyclization of acetylene. XPS measurements were performed to estimate the
C2H2 coverage on the hex-Pt(lOO) and other Pt-Sn alloy surfaces. In addition to Pt-Sn
alloys, we investigated yhe chemistry of a c(2x2)-Sn/Pt(100) adlayer for comparison.
7.2. Experimental Methods
All of the experiments were performed in a stainless steel ultrahigh vacuum (UHV)
chamber with a base pressure of 8x 10'“ torr, which previously described elsewhere [28],
It was equipped with a low energy electron diffraction (LEED) optics, cylindrical mirror
analyzer (CMA) for AES and XPS, ion sputtering gun, UTI 100C quadrupole mass
spectrometer (QMS) for TPD, dual anode X-ray source for XPS, and facilities for gas
dosing and metal evaporation.
The Pt(100) crystal was cleaned by repeated cycles of Ar+ ion sputtering at 300 K
and oxygen treatments with PQ 2 = 5x1 O '7 torr while the Pt crystal was held at 1200 K. The
143
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
cleanliness of the crystal was checked by AES, and LEED showed the expected,
reconstructed (5x20)-Pt(100) pattern [27]. The manipulator allowed for the Pt crystal to
be resistively heated to 1240 K and cooled to 95 K by contact with a liquid nitrogen
reservoir. The temperature was measured by a chromel-alumel thermocouple which was
spot-welded to the side of the crystal. Acetylene (C2D,) (Cambridge Isotope Laboratories,
99.96 %) was used without further purification. Gas exposures reported herein have been
corrected for an ionization gauge sensitivity factor ( 1.9) and also a directed beam doser
enhancement factor of 60.
All of the TPD experiments were done by using a heating rate of 4 K/s. XPS was
carried out by using 300 W Mg K a radiation (hv= 1253.6 eV) with the CMA operated at
a pass energy of 50 eV (AE = 0.8 eV).
The c(2x2) and (3v'2xv^2)R45° Sn/Pt(I00) surface alloys were prepared by
evaporating a thin Sn film onto the clean hex-Pt(lOO) surface at 300 K and then annealing
the sample to 750 K or 900 K, respectively, as described by Paffett et al [25]. The
observed c(2x2) LEED pattern following Sn deposition on Pt(l00) can be due to two
different surface structures [25, 26]. An overlayer of Sn adatoms is formed over a
relatively wide range of temperatures below 750 K. This adlayer is ordered for T = 500-
750 K. Formation of the c(2x2) alloy occurs over a narrow range of 750-800 K,
depending on the initial Sn concentration. A more thermally stable alley with 0S n = 0.67
ML and (3V2xV2)R45° LEED pattern is formed at higher annealing temperatures
between 800-1050 K. The c(2x2) alloy is comprised of 0.5 ML of Sn incorporated into
the surface plane of the Pt(100) sample to form a “checker board” pattern. Each Pt atom
is isolated from nearest neighbor Pt-atom contacts, i.e., surrounded by all Sn-atom
neighbors, and vice versa [25]. The alloy surface is quite flat, with Sn atoms buckled
outwardly by ~ 0.20 A above the surface Pt-atom plane [26]. The LEED pattern may be
144
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
the results of a periodic surface reconstruction with Sn atoms occupying the domain
boundaries [25]. Low energy alkali ion scattering spectroscopy (ALISS) experiments
suggested that the (3V2xV2)R45° alloy has nearly the same local geometric structure as
the c(2x2) alloy and is at least partially composed of small c(2x2) alloy domains with the
same buckling distance within the domains [26]. However, the structure of this surface is
still undetermined. Recent STM images of a somewhat related (100) surface of a bulk
Pt^Sn alloy after sputtering and annealing to 600 K, show pyramidal features consisting of
(102) facets [29]. Flat portions between and on top of the pyramids formed by the facets
are mainly comprised of triple rows that are probably Pt. A c(2x2) LEED pattern with
streaky facet spots was observed after heating to 600 K, and an improved c(2x2) LEED
pattern with no facet spots occurred after annealing to 1000 K [29]. STM studies in our
lab are underway to further probe the structure of the (3Vr 2xVr 2)R45° Sn/Pt(100) alloy and
the relationship that its structure has to that of other surface alloys and bulk alloy surfaces.
For brevity, the c(2x2)Sn/Pt(100) and the (3V2xVr 2)R45° Sn/Pt(100) alloys will be
referred to as the c(2x2) and 3V2 alloys, respectively, for brevity throughout this paper.
7.3. Results
73.1. TPD
At low initial C2D2 coverages, completely irreversible adsorption of C2D2 occurred.
cL,-Acetylene decomposed during TPD to liberate only D2 ( g ) and form a carbon adlayer.
Thermal desorption spectra following adsorption of a chemisorbed monolayer of d2-
acetylene (C2D2 ) on the clean hex-Pt(100) surface is shown in fig. 7.1. In addition to D2,
a small amount of C2D4 and C2D2 desorbed. No desorption of other hydrocarbons such
as ethane, propane, propene, C4HX , or benzene was detected. Hence, C2D2 adsorption is
essentially irreversible, with no dimerization or cyclotrimerization products formed during
145
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
TPD on the hex-Pt(lOO) surface. This is consistent with the results of a previous study
[4].
Fig. 7.2 shows D2 evolution from the Pt(100) surface during TPD after increasing
C2D2 exposures on hex-Pt(lOO) at 100 K. The bottom curve gives the D -, TPD trace after
a saturation exposure of D2 on the hex-Pt(lOO) surface at 100 K to give 0D = 1.20 ML
[30]. By comparing this curves with those from C2D2 decomposition, we conclude that
the lowest temperature D2 TPD peak at 378 K could be D.,-desorption rate-limited, but all
of the other higher temperature peaks are clearly rate-limited by reactions forming D(a ). D2
desorption shows that at least partial dehydrogenation occurs on Pt(l00) at or below 378
K. The three high temperature peaks for low initial C2D, coverage for D, evolution from
C,D 2 are very similar to those on Pt( 111) [1-3] and so sequential dehydrogenation
reactions of acetylene and its decomposition products at higher temperatures may be
similar on these two surfaces. Acetylene adsorbs molecularly on Pt(l 11) below 330 K in
a u3-r]2 configuration, disproportionate to form ethylidyne (CCH3 ) and other species
(possibly C,H) between 330-400 K, and then undergoes further dehydrogenation to form
surface carbon and H0 (g ) [1-3]. Ethylidyne hydrogenation on P t(lll) forms a small
amount of C-,D4 (g ) during TPD, and ethylidyne decomposition causes a characteristic D,
peak at 435 K. The peak near 435 K in fig. 7.2 on Pt may also arise from decomposition
of ethylidyne that was formed as a metastable intermediate on Pt(100), but mechanistic
information about C,D, decomposition on the Pt(100) surface awaits additional studies.
Products desorbed during TPD after saturation exposures of C2D, on the c(2x2)
alloy at 100 K are shown in fig. 7.3. We monitored the 26amu signal, due to cracking
fragment of C-,D-,, in order to exclude contributions from background CO adsorption.
The principle desorbed species were C2D0 reversible adsorption and D2. The amount of
D, desorption corresponds to 0D = 0.15 ML, as determined by a comparison to the D,
146
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
TPD peak area following a saturation exposure of D2 to give 0D = 1.2 ML on this surface.
This amount is consistent with the small amount of adsorbed carbon that was found in
AES following TPD, and corresponds to = 0-022 ML or 7% of the amount of
adsorbed C,D-, (as shown later). We believe that this is accurate measure of the reactivity
of this surface, but it is difficult to exclude the possibility that defect sites lead to this small
amount of C,D, decomposition. We searched for the desorption of other hydrocarbon
products, but except for a small amount of C 2D4 after large exposures, no other C,, C4 or
C6 products were found.
Fig. 7.4 shows the C,D., TPD curves after increasing exposures of C2D2 on the
c(2x2) alloy at 100 K. A C,D, peak occurs at 430 K even at low coverages, and this peak
shifts down slightly to 423 K at saturation coverage in the chemisorbed layer. At high
exposures, peaks appear at 387 K and 310 K while the peak at 423 K decreases slightly in
size. Desorption from a weakly bonded state also occurs in a peak at 180 K. We believe
that this latter peak is due to a species in the chemisorbed monolayer that is Jt-bonded to
the surface rather than due to acetylene adsorbed in the second layer i.e. first physisorbed
layer, because the coverage associated with this peak is much less than that of the
chemisorbed layer. C,H, multilayers could be formed and caused an apparent peak
appeared at 120 K which could be increased in intensity by larger exposures.
TPD spectra from the C,D, chemisorbed monolayer on the 3v^2 alloy are shown in
fig. 1 3 . Acetylene was mostly reversibly adsorbed. By using the D2 TPD peak area, we
estimate that =0.012 ML or -5% of the adsorbed acetylene monolayer (as shown
later) decomposed during TPD. However, benzene desorption was observed following
cyclotrimerization of acetylene.
Fig. 7.6 shows TPD spectra after increasing exposures of acetylene (C2D2 ) on the
3V2 alloy at 100 K. Initially, a peak grows at 305 K and then a second peak grows larger
147
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
at 283 K. Two small, high temperature peaks were also observed at 422 and 520 K. As
on the c(2x2) alloy, peaks at 180 and 120 K were observed and are assigned to a weakly
adsorbed, jr-bonded species in the monolayer and to a C2D, multilayer peak, respectively.
In fig. 7.7, we compare C2D2, D,, and C6D6 TPD spectra following the adsorption
and reaction of a saturation coverage of acetylene on several surfaces at 100 K. C-.D, is
irreversibly adsorbed on the hex-Pt(100) surface, liberating only D2 in TPD. Sn adatoms
in an ordered c(2x2) Sn overlayer (0S n= 0.5 ML) on the Pt(100) surface allows only very
small chemisorption and completely suppressed acetylene decomposition and reaction.
(The c(2x2) Sn overlayer was prepared [25, 26] by annealing a thin layer of Sn on the
hex-Pt(100) to 500 K). On the c(2x2) alloy, with 0S n = 0.5 ML, decomposition was
strongly reduced during TPD but the alloy surface retained a large chemisorption capacity.
These spectra illustrate an important difference in the effectiveness of Sn to "block" sites
depending on its location as an adatom or alloyed atom on Pt surfaces. The most strongly
chemisorbed C,D2 on the c(2x2) alloy desorbed at 423 K, and no benzene desorption was
observed. On the 3V2 alloy with 0S n = 0.67 ML, the C 2D 2 desorption energy is decreased
further, as indicated by the shift of acetylene desorption peak to lower temperatures, and
cyclotrimerization reactions form benzene which desorbs from the surface.
In order to determine the C 2D2 desorption activation energy (Ed ) on the two alloy
surfaces, we measured C2D2 desorption kinetics and the desorption peak temperature (T )
using heating rates ((3) of 1-24 K/s. We only considered the highest temperature
desorption peak at about 10% of the monolayer coverage in order to minimize lateral
interactions. Plots of ln((3/Tp ) versus 1/Tp associated with these measurements are shown
in fig. 7.8. Such plots should yield a straight line with slope = -E^R, where R is the gas
constant, which is independent of the value of pre-exponential factor [31]. Values for Ed
of 29.8 and 19.9 kcal/mole for the c(2x2) and 3V2 alloy surfaces, respectively. We note
148
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
that these values are not too different those calculated more simply by the Redhead method
[31] 1 of 27 and 19 kcal/mole for the c(2x2) and the 3 /2 alloys, respectively, assuming
first order desorption kinetics and a pre-exponential factor of 101 3 s".
132. XPS
Fig. 7.9 shows the C(ls) core-level photoelectron spectra corresponding to
molecularly adsorbed C2D2, as a function of C,D, exposure on hex-Pt(lOO) at 100 K.
The spectra are all described well by a single peak of 2.0 eV FWHM at 283.9 eV binding
energy (BE) at low coverage which shifts 0.2 eV to higher BE at saturation coverage.
These results are consistent with a previous measurement of the C(ls) peak for molecular
acetylene on Pt(l 11) at 284.0 eV BE [32].
The C(ls) XPS spectra shown in fig. 7.10(a) were obtained after an exposure of 4L
C2D2 (to form a little more than a monolayer) on hex-Pt(100) at 100 K and stepwise
annealing to higher temperatures. The width of the C(ls) peak is 2.0 eV FWHM in each
spectrum. Heating to 600 K caused the C(ls) peak to shift to 0.4 eV higher BE, and the
final value of 284.5 eV BE is consistent with that for graphitic carbon at 284.3 eV BE on
Pt(l 11) [32]. Fig. 7.10(b) quantifies the corresponding C(ls) peak areas. A decrease in
C(ls) intensity occurs after heating to 200 K is because of the desorption of all
physisorbed acetylene. An additional decrease of ~10% in the C(ls) intensity occurs
from annealing the adlayer to 400 K. C2D4 desorption at 357 K is the main contribution
to this decrease because of the very small amount of C2D2 desorption at 375 K (as will be
seen later). The C(ls) intensity remained constant from 400 to 800 K, consistent with the
absence of any significant desorption of hydrocarbons or diffusion of carbon into the bulk
of the crystal.
C(ls) XPS spectra for increasing C,D, exposures on the c(2x2) and the 3 /2 alloys at
100 K are shown in figs. 7.11(a) and (b), respectively. The C(ls) peak for monolayer
149
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
G,D, coverage is at 284.3 eV BE on the c(2x2) alloy and 283.9 eV BE on the 3V2 alloy.
The peak width is 2.0 eV FWHM at low coverages on both surfaces, as on Pt(100), but
broadens to 2.6 eV at higher coverages. The C(ls) spectra after 8L of C2D, in fig. 7.11
can be decomposed to give a new peak at 285.9 eV BE. We assign this peak to jr-bonded
acetylene molecules that are more loosely coordinated to the surface. Similar rc-bonded
acetylene species was found on Cs/Pd(110) surfaces [33]. In related studies, Jt-bonded
ethylene species are formed on Cs/Pt(l 11) and K/Pt(l 11) surfaces [34, 35].
Annealing studies using XPS are used on the 3V2 alloy to estimate the amount of
benzene desorption. The C(ls) XPS spectra for about a monolayer of adsorbed acetylene
on the 3V2 alloy at 100 K and adlayer after stepwise annealing are shown in fig. 7.12.
The inset quantifies the C(ls) intensity changes. At 100 K, there are two peaks at 283.9
and 285.9 eV BE. The peak at 285.9 eV is eliminated by heating to 200 K and the main
C (ls) peak shifts to 0.9 eV lower BE. A new peak appears at 284.6 eV BE. A 20%
decrease in the C(ls) intensity results from desorption of weakiy adsorbed acetylene
corresponding to the peak in TPD at 180 K. Annealing to 300 K decreases the C(ls)
peak intensity corresponding to molecular desorption of 75% of the acetylene monolayer
that is reversibly adsorbed. Two C(ls) peaks indicate the presence of more than one kind
of surface species. The decrease in intensity by heating from 300 to 500 K indicates d™1 1
= 0.04 ML and so -15% of the adsorbed acetylene monolayer undergoes cyclization to
form benzene on the 3V2 surface. The C(ls) intensity after annealing to 500 K indicates a
small amount - 10% of decomposition to form carbon and hydrogen on the surface.
Adsorbed acetylene decomposed to give two forms of atomic carbon, graphitic (284.6 eV)
and carbidic (283 eV) carbon on the 3\/2 surface [36].
The C(ls) spectra for near-monolayer coverages of acetylene on the hex-Pt(100) and
the two Sn/Pt alloy surfaces at 100 K are directly compared in fig. 7.13. We estimated
150
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
that the coverage of acetylene in the chemisorbed monolayer on Pt(100) at 100 K is
0C2d 2 = consistent with a previous value of 0.5 ML that was also derived from
XPS measurements [38]. Our estimate utilized a comparison of the C(ls) peak area of a
monolayer of CO on Pt(100) at 100 K, where the CO coverage is known to be 0C O = 0.75
ML or 9.7xl0 1 4 molecules/cm2 [37]. The amount of acetylene adsorbed in the
chemisorbed layer is clearly decreased due to the presence of alloyed Sn in the surface.
The monolayer coverages of C2D2 on the c(2x2) and the 3V2 Sn/Pt(100) alloys were
determined to be 0.32 ML and 0.25 ML, respectively.
XPS data can also be used to determine information on acetylene adsorption kinetics
on the three surfaces at 100 K Fig. 7.14 shows the “uptake curves” where he initial slope
(at low exposures) is proportional to the initial sticking coefficient (S0) of acetylene. The
value of S0 on both alloyed surfaces at 100 K is one-half of that on hex-Pt(lOO).
73 3 . LEED
No extra spots in LEED were observed following acetylene adsorption on the c(2x2)
or 3V2 surfaces at 100 K. This indicates that either no ordered adsorbate structures were
formed or that these structures had the same size and orientation of their unit cells. Also a
number of annealing experiments were carried out up to 600 K following multilayer
adsorption, and no new spots were observed. Importantly, none of these experiments
caused a noticeable degradation of the initial alloy LEED pattern.
7.4. Discussion
On the hex-Pt(100) surface, ~90% of the chemisorbed acetylene monolayer
decomposes to give D, and leave adsorbed carbon on the surface during heating in TPD.
Molecular desorption of acetylene is minimal (~5%) but a small amount of ethylene
desorbed at 375 K. C ,H , adsorbs on P t(lll) in a di-cr-jr bonding fashion [1-3], and
151
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
presumably this also occurs on Pt(100) surfaces, however determination of this must await
vibrational studies.
Alloyed Sn with 0S n = 0.5-0.67 ML in the Pt(100) surface still allowed a large
amount of acetylene to chemisorb but almost completely suppressed acetylene
decomposition during heating in TPD. On the c(2x2) alloy, the monolayer coverage was
0 cf?D2 = 0-32 and only desorption of molecular acetylene and a small amount of D,
desorption occurred. The D2 TPD peak area and C(ls) XPS intensity changes give a
consistent estimate that about 93 % of the acetylene monolayer desorbed from the surface.
The small amount of decomposition (~7 %) may be due to defect sites, but most likely
not. XPS and TPD spectra also show a weakly bound species assigned to n-bonded
acetylene that desorbs at 180 K on the alloy surface. A large molecular desorption of
C2D2 (75 %) along with benzene desorption was also observed on 3V2 alloy with
saturation monolayer coverage =
Closely related studies have been carried out on acetylene adsorption and reaction on
Pt(l 11) and two Sn/Pt(l 11) surface alloys, a (2x2)Sn/Pt(l 11) alloy with 0S n = 0.25 and a
(V r 3xVr 3)R45° S n /P t(lll) alloy with 0S n = 0.33 ML [23 , 24], Molecular desorption of
acetylene was increased by alloying with Sn and decomposition was inhibited. However,
even on the V3 alloy with the highest Sn concentration, more than 35% of the chemisorbed
acetylene monolayer decomposed to liberate H,and form surface carbon. The monolayer
coverage decreased about 20-25% upon alloying Sn in to the Pt(l 11) surface [19]. This
compares to an even large decrease of 40-60% on Sn/Pt(100) surfaces. The higher Sn
concentration in the Sn/Pt(100) alloys isolates the Pt atoms and thereby reduces pure-Pt
adsorption sites to ensembles of a single Pt atom.
Given the small changes detected in the Pt valence levels and core levels by UPS and
XPS [24,25], it is important to attempt to understand the influence of Sn due to "ensemble
152
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
effect". This is useful in order to determine if predictive capability (based simply on
geometry) for other alloy surfaces can be developed. However, in doing so, it is also
important to remember that electronic effects of Sn on Pt electronic structure may be an
important factor in the observed decrease in acetylene adsorption energy and the increase
in C-H dissociation activation energy barrier.
Looking at the ensembles available on the (2x2) Sn/Pt(l 11) alloy, three-fold pure-Pt
sites are present but there are no two adjacent three-fold pure-Pt sites. Only two-fold
bridge and atop sites are present on the V3 alloy, and all of the pure Pt three-fold sites are
eliminated. On the c(2x2)Sn/Pt(100) alloy, Pt atoms are isolated, surrounded only by
nearest neighbor Sn atoms and pure-Pt two-fold bridge and four-fold hollow sites are
eliminated. (The 3V2 alloy is thought to have a similar structure, which somehow
incorporates additional surface Sn). The strong suppression of decomposition of
acetylene on both Sn/Pt(100) alloys can be attributed to an ensemble requirement of at
least a pure-Pt two-fold bridge site to stabilize the transition state or reaction products of
acetylene decomposition. Hence, acetylene decomposition at single Pt atoms on Pt-Sn
alloys is a highly activated process, especially compared to other available reaction
channels including acetylene desorption.
Benzene formation and desorption from the cyclotrimerization of acetylene on Pt-Sn
alloys is a structure sensitive reaction. Benzene desorption occurs from both of the
Sn/Pt(l 11) alloys and the highest amount of benzene (accounting for 10% of the acetylene
monolayer) is formed on the V3 surface [19]. Acetylene is strongly rehybridized on clean
Pt( 111) and from this reaction on Pt(l 11) benzene desorption does not occur. Alloyed Sn
weakens the Pt-C interaction, but if only a weaker interaction caused benzene formation
then one would expect a higher amount of benzene desorbed from the c(2x2) surface on
Pt(100) with 0S n = 0.5 ML than on the V3 Sn/Pt(ll I) surface with 0S n = 0.33 ML. No
153
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
benzene desorption was found on the c(2x2) alloy. However, it is well known that surface
geometry plays key roles in benzene formation on Pd surfaces and that is apparently the
case for Pt-Sn alloys as well. On the 3>/2 alloy with 0S n = 0.67 ML, about 15% of the
acetylene monolayer undergoes cyclization to form benzene. We propose that benzene
formation is a probe reaction for Pt ensembles that have hexagonal or (lll)-lik e
symmetry. This would imply that such facets or steps exist on the 3a / 2 alloy. Consistent
with that, CO chemisorption reveals more "Pt character" on the 3V2 alloy than on the
c(2x2) alloy [39]. This could result from a higher Pt coordination number.
Fig. 7.15 shows how the C,D , monolayer coverage and initial sticking coefficient
are affected by Sn addition to form surface alloys on the Pt(100) surface. These curves
provide a qualitative guide to the influence of alloyed Sn on acetylene chemisorption over a
wide range of Sn concentrations, and more importantly reveal whether specific site-
requirements exist. On Pt(lll), the saturation coverage of acetylene has been reported to
be 0.25 ML (3.8 x 101 4 molecules/cm2 ) at 100 K [19, 39]. On this basis, the saturation
G,H, coverage decreases to 0.17 ML on the (2x2) alloy and 0.16 ML on the V3 alloy at
100 K [19]. On Pt(100), the acetylene monolayer coverage has been reported to be 0.5
ML (6.5 x 101 4 molecules/cm2 ) at 100 K [4, 5]. Using these references, the monolayer
coverage decreases to 0.32 ML on the c(2x2) alloy and 0.25 ML on the 3a/2 alloy at 100
K. The monolayer coverage is unexpectedly different on these two Pt surfaces, and so it is
not surprising that there is some controversy over the monolayer coverage of acetylene on
clean Pt(lll) [19, 39-41], This has been proposed to be either 0.25 ML [19, 39] or 0.5
ML [40,41]. In fig. 7.11(a) we plotted the saturation monolayer coverage calculated from
a closest packing model (0c p ) on Pt surface and the saturation monolayer coverage of
Sn/Pt alloy surfaces were estimated on the basis of that. The coverage 0c p = 5.7 x 101 4
molecules/cm2 (0.44 ML on Pt(100) and 0.37 ML on Pt(l 11)) is the upper limit for the
154
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
number of molecules in the chemisorbed layer. The changes in the chemisorbed
monolayer coverage of C,D, induced by Sn are close to those expected by a simple site-
blocking model in which 0C 2D 2=( 1"a® sn)b» where a =1 (the number of adsorption sites that
are blocked by one modifier atom) and b =1 (the number of adsorption sites required for
adsorption).
The values for the C,D, initial sticking coefficient S0(C2D,) on four Pt-Sn alloys at
100 K are compared in Fig. 7.15 (b). S0 is reduced, but not as strongly as one might
guess based on Langmuirian adsorption kinetics and the surface Sn concentration.
Similar results were obtained for CO adsorption on these Pt-Sn surface alloys [42], A
simple Langmuirian site-blocking equation for the dependence of S0 given by S = S0(l-
a0 S n )b fails due to the important influence of a “modifier precursor” state on the
adsorption kinetics [43,44]. This factor allows for faster adsorption kinetics than would a
linear or greater decrease S0 with increasing Sn concentration.
The fraction of chemisorbed acetylene that decomposes to give carbon deposition on
Pt-Sn alloys is plotted in fig. 7.16. The presence of > 0.5 ML of Sn in the Pt(100) surface
completely suppressed acetylene decomposition. Acetylene was reversibly adsorbed and
desorbed on these alloyed surfaces. Elimination of pure Pt two-fold bridge and four-fold
hollow sites on Sn/Pt(100) alloy surfaces indicates an ensemble requirement of at least a
pure Pt two-fold bridge site to stabilize the transition state or reaction products of
acetylene decomposition
7.5. Conclusions
Alloying Sn to form the c(2x2) and 372 Sn/Pt(100) surface alloys removes all
pure-Pt four-fold hollow and pure-Pt two-fold bridge sites, leaving only isolated Pt surface
atoms. There are many changes in acetylene adsorption and reaction caused by this
modification. Decomposition of acetylene was greatly suppressed (more than 90%) on
155
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
the c(2x2) Sn/Pt(100) alloy. The monolayer coverage of C,D, decreases from 0.54 ML
on Pt(100) to 0.32 ML on the c(2x2) alloy and 0.25 ML on the 3V2 alloy. Alloying Sn
with Pt(100) to form the c(2x2) alloy also reduced the initial sticking coefficient of
acetylene at 100 K by a factor of two. Acetylene chemisorption on these Sn/Pt alloyed
surfaces is structure insensitive, whereas acetylene decomposition is structure sensitive and
at least an ensemble of two Pt atoms is required for decomposition activity.
Cyclotrimerization of acetylene to benzene was observed on the 3V2 Sn/Pt(100) alloy, and
this may be due to presence of (11 l)-like steps or facets present on the surface.
Acknowledgment
This work was supported by the Division of Chemical Sciences, Office of Basic
Energy Sciences, U.S. Department of Energy.
1 5 6
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
REFERENCES
[1]. M. Salmeron, and G. A. Somoijai, J. Phys. Chem. 86 (1982) 341.
[2]. C. E. Megiris, P. Berlowitz, J. B. Butt, and H. H. Kung, Surf. Sci. 159 (1985) 184.
[3]. N. R. Avery, Langmuir 4 (1988) 445.
[4]. T. E. Fisher, S. R. Kelemen, and H. P. Bonzel, Surf. Sci. 64 (1977) 157.
[5]. Y. Zhou, and S. M. Davis, Catal. Lett. 15 (1992) 51.
[6]. R. D. Cortright, and J. A. Dumesic, J. Catal. 148 (1994) 771.
[7]. M. T. Paffett, S. C. Gebhard. R. G. Windham, and B. E. Koel, Surf. Sci. 223 (1989),
448.
[8]. C. Xu, and B. E. Koel, M. T. Paffett, Langmuir 10 (1994) 166.
[9]. Y. L. Tsai, and B. E. Koel, Surf. Sci. 385 (1997) 37.
[10]. W. T. Tysoe, G. L. Nyberg, and R. M. Lambert, J. Chem. Soc. Chem. Commu.
(1983) 623.
[11], C. H. Patterson, and R. M. Lambert, J. Am. Chem. Soc. 110 (1988) 6871.
[12]. R. M. Ormerod, and R. M. Lambert, Catal. Lett. 6 (1990) 121.
[13]. A. J. Gellman, J. Am. Chem. Soc. 113 (1991) 4435.
[14], N. A. Avery, J. Am. chem. Soc. 107 1985)6711.
[15]. J. R. Lomas, C. J. Baddeley, M. S. Tikhov, and R. M. Lambert, Langmuir 11
(1995)3048.
[16]. J. C. Bertolini, J. Massardier, and G. Dalmai-Imelik, J. Chem. Soc. Faraday Trans. I
(1978) 1720.
[17], T. M. Gentle, and E. L. Muetterties, J. Phys. Chem. 87 (19830 2469.
[18]. K. J. Pierce, and M. A. Barteau, J. Phys. Chem. 98 (1994) 3882.
[19]. C. Xu, J. W. Peck, and B. E. Koel, J. Am. Chem. Soc. 115 (1993) 751.
[20]. R. M. Ormerod,C. J. Baddeley, and R. M. Lambert, Surf. Sci. 259 (1991) L709.
[21]. C. J. Baddeley, R. M. Ormerod, A. W. Stephenson, and R. M. Lambert, J. Phys.
Chem. 99 (1995) 5146.
[22]. F. A. Lee, C. J. Baddeley, C. Hardacre, G. D. Moggridge, R. M. Ormerod, R. M.
Lambert, J. P. Candy, and J. M. Basset, J. Phys. Chem. B 101 (1997) 2797.
157
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
[23]. R. G. Windham, and M. T. Paffett, Surf. Sci. 208 (1989) 34.
[24]. S. H. Overbury, D. R. Mullins, M. T. Paffett, and B. E. Koel, Surf. Sci. 254 (1991)
45.
[25]. M. T. Paffett, A. D. Logan, R. J. Simonson, and B. E. Koel, Surf. Sci. 250 (1991)
123.
[26], Y. Li, and B. E. Koel, Surf. Sci. 330 (1995) 193.
[27]. R. J. Behm, W. Hosier, E. flitter, and G. Binnig, Phys. Rev. Lett. 56 (1986) 228.
[28]. R. G. Windham, M. E. Bartram, and B. E. Koel, J. Phys. Chem. 92 (1988) 2862.
[29]. M. Hoheisel, J. Kuntze, S. Speller, A. Postnikov, W. Heiland, I. Spolveri and U Bardi,
to be published.
[30]. P. R. Norton, J. A. Davies, D. K. Creber, C. W. Sitter, and T. E. Jackman, Surf. Sci.
8 (1981) 205.
[31]. P. A. Redhead, Vacuum 12 (1962) 203.
[32]. N. Freyer, G. Pirug, and H. P. Bonzel, Surf. Sci. 126 (1983) 487.
[33]. T. Takaoka, T. Sekitani, T. Aruga, and M. Nishijima, Surf. Sci. 306 (1994) 179.
[34], A. Cassuto, M. Mane, M. Hugenschmidt, P. Dolle, and J. Jupille, Surf. Sci. 237
(1990) 63.
[35]. A. Cassuto, M. Mane, J. Jupille, G. Tourillon, and P. Parent, J. Phys. Chem. 96
(1992)5987.
[36]. J. Vaari, J. Lahtinen, and P. Hautojarvi, Catal. Lett. 44 (1997) 43.
[37]. G. Borden, G. Pirug, and H. P. Bonzel, Surf. Sci. 72 91978) 45.
[38]. T. E. Fischer, and S. R. Kelemen, J. Vac. Sc.Technol. 15 (1977) 607.
[39]. M. Abon, J. Billy and J. C. Bertolini, Surf. Sci. 171 (1986) L387.
[40]. N. Freyer, G. Pirug, and H. P. Bonzel, Surf. Sci. 125 (1983) 327.
[41]. N. Freyer, G. Pirug, and H. P. Bonzel, Surf. Sci. 126 (1983) 487.
[42]. C. Panja, and B. E. Koel, Israel J. Chem. 38 (1998) 365.
[43]. L. Q. Jiang, and B. E. Koel, J. L Falconer, Surf. Sci. 273 (1992) 273.
[44], C. Xu, and B. E. Koel, J. Chem. Phys. 100 (1994) 664.
158
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
TPD
375K
CO
s z
CD
C Z
C O
O
347K
515K
585K 680K
435K
(x 0.05)
8 0 0 200 6 0 0 4 0 0
Temperature (K)
Fig. 7.1 TPD spectra of products from the decomposition of a saturation monolayer of
acetylene on hex-Pt(100).
159
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
512 K
665 K
CO
d
CD
d
3
E
C O
'vt'
pi
Q
(x 0.5)
D2 from saturation
exposure of D2
6 0 0 8 0 0 200 4 0 0
Temperature (K)
Fig. 7.2 D2 TPD spectra after C2D2 exposures on hex-Pt(100) surface at 100 K.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Q M S Intensity
TPD C^D2/c(2 x2) alloy
10 L Exposure
C6D6 (84 amu)
C 4D4 (56 amu)
D2 (mass 4)
200 4 0 0 600 800
Temperature (K)
Fig. 7.3 TPD spectra obtained from the reaction of acetylene on the c(2x2) Sn/Pt(100)
alloy after saturation exposure of acetylene at 100 K.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
C ^ D g (2 6 amu) Intensity
C2D2/c(2x2) alloy
TPD
117 K
423 K
387 K
310 K
Exp (L) CpD
I ' i t t j t
1 00
I i i- i t y v i i i | i i i i [ i i ii | i f 11 | 1 r i i i | i i i i •
2 0 0 3 0 0 4 0 0 5 0 0 6 0 0
Temperature (K)
Fig. 7.4 G,D2 TPD spectra obtained after acetylene exposures on the c(2x2) alloy at 100
K .
1 6 2
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
T P D Peak Intensity (amu)
TPD
C2D4 (mass 32)
D2 (4 amu)
600 800 400 200
Temperature (K)
Fig. 7.5 Product formation from the reaction of a saturation monolayer of acetylene on
the (3\^2x\^2)R45° Sn/Pt(100) surface during TPD.
163
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
C,D2 Peak Intensity (2 6 amu)
C2D2/3 V2 alloy
TPD
|120
283 K
305 K
Exp (L) C,D.
25
181
422 K
519 K
4 0 0 6 0 0 3 0 0 5 0 0 200 100
Temperature (K)
Fig. 7.6 TPD spectra of acetylene after C,D2 exposures on the 3V2 alloy at 100 K.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
rmission o f the copyright owner. Further reproduction prohibited without permission.
3V2 alloy
c(2x2) alloy
c(2x2) overlayer
Pt(100)
"*|~TTT T '[l T'l I I I I II | II I I I I I I I p I
200 400 600
Temperature (K)
3V2 alloy
c(2x2) alloy
c(2x2)
overlayer
1 /1 0
Pt(100)
1111111 > i \u 11111111111111 it 111111 p T r
200 400 600 800
Temperature (K)
D
E
c d
co
w
c
Q )
3V2 alloy
c(2x2) alloy
X
c(2x2) overlayer
A IllWilW I ■ ■ I I II* M M I * l>
Pt(100)
11 i | > i'll 11 m i j 11 i i | i r r r | n r 1
200 400 600
Temperature (K)
Fig. 5.2 Comparison of C2 D2 , D2 and C6D6 TPD spectra obtained from a saturation exposure of acetylene at 100 K on
different surfaces.
o\
C^D2 Desorption
(3=1-24 K/s
= 0.1 ML
CM
Q _
I —
02.
C
— I
3\/2 alloy c(2x2) alloy
4.0x10 2.0 3.0 3.5 2.5
1/Tp(1/K)
Fig. 7.8 In((3/Tp 2 ) versus l/Tp for desorption of acetylene on the two Sn/Pt alloyed
surfaces at heating rates 1-24 K/s.
1 6 6
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
C(1s) XPS
284.1
D2 Exposure LU
284.0
x
X
m m * f 1 . ■» >
• - 1 ^ • • •
283.9
.0.5 L
0.3 L
2 90 2 8 8 286 2 8 4 28 2 280 278 276
Binding Energy (eV)
Fig. 7.9 C(ls) spectra of acetylene after C2H, exposures on hex-Pt(lOO) at 100 K.
167
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
C(1s) XPS
284.1
100 K
L U
284.3
200 K
300 K
284.4'
400 K
284.
500 K
600 K
2 9 0 2 8 8 2 86 284 2 8 2 2 8 0 278 2 7 6
Binding Energy (eV)
35x10'
3 0 -
2 5 -
S 2 0 -
<
^1 5 -
c o
Q_1 0 -
5 -
0 -|
100 2 0 0 30 0 400 5 0 0 6 00 700 800
Annealing Temperature (K)
Fig. 7.10 C(ls) spectra obtained following adsorption of a saturation monolayer of
acetylene on hex-Pt(100) at 100 K as a function of annealing temperatures.
The lower curve shows the C (ls) intensity versus temperature for each
annealing temperatures shown in upper curves.
168
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
C(1s) XPS
284.3
285.9
C^D2 Exposure
L U
2 8 4 .2 s '
2 9 0 288 286 2 84 282 2 8 0 278 276
Binding Energy (eV)
LU
C(1s) XPS 283.9 C2D2/3V 2 alloy
C^D2 Exposure
285.9
2 8 3 .7V
t . ~ • •? %
’ • • 4 l
-V^ . S r . / t j r
• . •••
... . 2 L
* * * P I
• * * v m ;. * . * *
l i | I I 1 | I »"7 | M » r f r ■ i i - \ | \ t i 1 i' n r - i —r ■ i » i i | i i ^ | i t a H |*
2 9 0 288 286 2 84 282 2 8 0 278 276
Binding Energy (eV)
Fig. 7.11 C(ls) spectra of acetylene after acetylene exposures on the c(2x2) alloy and
3V2 alloy at 100 K.
169
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
C p 2/3 ^ 2 alloy 12x1 o'
C(1s) XPS
283.9
vu o
C D 8 -
4 -
O-l
00 200 300 4 0 0 500
Annealing Temperature (K )
285.9
V
283
L U
284.6
284.2 200 K
L 2 83
■ » .
• • • •
300 K
284.8
■ ’ yxv-***-
284.6
400 K
^ V ~ 500 K
270 280 290 275 285
Binding energy (eV)
Fig. 7.12 XPS C(ls) spectra obtained following adsorption of a saturation monolayer of
acetylene on the 3a/2 alloy at 100 K as a function of annealing temperature.
The inset shows the C(ls) intensity' versus annealing temperature for acetylene
on the 3V2 alloy.
170
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
C(1s) XPS
O 2 D2
1 M L Exposure
T= 100 K
284.1
284.3
L U
Pt(100)
28 5 .9
c(2x2) alloy
283.9
I • •
2 8 5 .9
3-/2 alloy
29 0 28 8 2 86 28 4 2 8 2 2 80 278 2 7 6
Binding Energy (eV)
Fig. 7.13 Comparison of C(ls) spectra for a saturation monolayer of acetylene on several
surfaces.
171
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
C(1s) X P S P eak A re a (arb. units)
0.6
P t(100)
0.5
0.4
c(2x2) alloy
P G
0.3
3V2 alloy
0.2
0.0
10 12 14 16 4 6 0 2 8
Exposure (L)
Fig. 7.14 The uptake curve constructed from the intensity of C (ls) photoelectron line
versus the C,D, exposures on clean hex Pt(100) and the two alloyed surfaces.
The C,D, coverage was plotted in the right hand side axis as estimated from the
known saturation coverage of CO on the clean Pt(100) surface [37].
172
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
0.6
QjH2 hep monolayer
c v i
x1 0
0.8
Jt+ C T
in
0 . 2 -
0.4
(100) ( 2 x 2 ) /3 c ( 2 x 2 ) 3 /2
0.0
0.0
Sn Concentration ( atoms/cm2)
o
o
3
O
CD
P
o '
Z3
P
O
3
C O
o '
3.
1.6x10
0 . 8 -
a > 0 .6 —
0.8
E ? 0 .4 -
0.4
( 1 0 0 )
M 0 . 2 -
(2x2)V 3 c(2x2) 3 / 2
C O
0.0 0 .0 -j
N )
2
Sn Concentration ( atom s/cm )
Fig. 7.15 Comparison of acetylene monolayer saturation coverage and initial sticking
coefficient of acetylene versus surface Sn concentration in atoms/cm2 on Sn/Pt
alloyed surface. The squares represents Pt atom concentration in atoms/cm2,
plotted on right hand side axis.
173
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Reaction 1 00
8 0 -
6 0 -
2 0 -
( 100 )
0.8 0 .4 0.6 1.0 0.2 0.0
15 2
Sn Concentration (x10 atom s/cm )
Fig. 7.16 Fraction of decomposition of acetylene versus surface Sn on several Sn/Pt
surfaces.
174
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Chapter 8
Conclusion
Addition of a second component to a metal surface can greatly affect the observed
chemistry, either hindering or enhancing selected reaction pathways. Dispersed, supported
bimetallic catalysts often show enhanced reactivity or selectivity and are of central
importance commercially. In order to probe aspects of the basic chemistry of these
bimetallic systems, chemisorption and reaction of a number of small molecules was
explored on several Pt and Pt-Sn alloy surfaces. Methanol (CH3 OH), ethanol (C2H5OH),
and water (H20 ) on P t(lll) and Sn/Pt(lll), and carbon monoxide (CO), nitric oxide
(NO), and acetylene (C2H2 ) on Pt(100) and Sn/Pt(100) was studied. These experiments
were carried out under ultrahigh vacuum conditions using temperature programmed
desorption (TPD), Auger electron spectroscopy (AES), low energy electron diffraction
(LEED), X-ray photoelectron spectroscopy (XPS) and high resolution electron energy loss
spectroscopy (HREELS).
Changes of the surface chemistry of (111) and (100) faces of platinum single
crystals can be studied following the addition of tin to form different surface alloys. A
p(2x2) structure ( 0 S n = 0.25 ML) and a (V3x\(3)R30o structure (0 S n = 0.33 ML) can be
formed on P t(lll), and c(2x2) and (3V2x"/2)R45° structures (0 S n = 0.5 and 0.67 ML,
respectively) can be formed on Pt(100).
175
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Interaction of methanol, ethanol, and water with Pt-Sn surfaces was explored to
evaluate whether the presence of Sn in the surface layer leads to an increase in adsorption
energy or thermally activates these molecules for reaction due to the thermodynamic
driving force provided by relatively strong Sn-O bond. Surprisingly, alloying with Sn
reduces the adsorption energy of these molecules compared to that on Pt( 111). This is
consistent with recent results that show that Pt-Sn alloys are not more active than pure Pt
for electrooxidation of methanol. Our data provide benchmarks for discussing the surface
chemistry of alcohols on Pt-Sn alloys, opening the road for an improved understanding of
electrooxidation and catalysis of alcohols and other oxygenated molecules on bimetallic
Pt-Sn catalysts.
Additionally, the chemisorption properties of CO, NO and C,H, was investigated to
probe the structural and electronic differences between alloyed Pt-Sn surfaces and the
Pt(100) surface. Alloying with Sn causes only a small reduction of 30-34 kJ/mol in the
adsorption energy of CO on the two Sn/Pt(100) surface alloys compared to the Pt(100).
In contrast, NO is chemisorbed more weakly by 60-70 kJ/mol on these two alloy surfaces
than on Pt(100). Also, while the adsorption of CO on Pt(100) and the two Sn/Pt(100)
alloys is completely reversible, a substantial change occurs in the NO chemistry on
Sn/Pt(100) surfaces because of the presence of Sn. On Pt(100), about 25% of the NO
monolayer decomposes to eventually desorb N, and O,. In the presence of Sn for the two
alloys, N2 0 formation occurs from NO reaction at very low temperatures, even at 100 K.
We believe that this reactivity arises from the low temperature formation of a NO
dinitrosyl complex which facilitates N2 0 formation. Our data suggest that the absence of
adjacent strong bonding sites that exist for NO on pure Pt surfaces leads to the formation
of a dinitrosyl complex which subsequently decomposes to form N2 0 .
Finally, we used C2H2 as a reactive molecule with a low H:C stoichiometry in order
to model coking reactions on Pt(100) and Sn/Pt(100) alloy surfaces. Sn reduced the C ,H 2
176
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
chemisorption bond strength and reactivity for dissociative adsorption of C,H2 compared
to Sn-free Pt surfaces. In addition, the temperature for complete dehydrogenation of the
carbonaceous residue formed from acetylene decomposition (polymerization) was
increased by up to 100 K. Both of these phenomena are consistent with observations of
increased lifetimes and decreased coking for technical Pt-Sn bimetallic catalysts compared
to Pt catalysts used for hydrocarbon conversion reactions. In addition to being a prototype
“coke precursor” molecule with a C:H stoichiometry of unity, C2H, can undergo a rather
unique C-C bond coupling reaction that can be studied in UHV, i.e. benzene formation via
cyclotrimerization. Cyclotrimerization of acetylene to benzene was observed only on the
(3Vr 2xVr 2)R45° Sn/Pt(100) alloy. This may indicate that reconstruction to produce this
alloy leads to the formation of Pt sites that have a local C3-symmetry.
177
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Bibliography
-A-
[1]. M. Abon, J. Billy and J. C. Bertolini, Surf. Sci. 171 (1986) L387.
[2], R. Adzic, In Advances in Electrochemistry and electrochemistry Engineering,
Gerischer, H. Ed. Wiley-Interscience: New York, 1984, vol 13 pp. 159-260.
[3]. S. Akhter and J. M. White, Surf. Sci., 167 (1986) 101.
[4]. A. Anderson, B . Lassier-Govers, Chem. Phys. Lett. 50 (1977) 124.
[5] A. Atrei, U. Bardi, G. Rovida,M.Torrini,E.Zanazzi an P. N. Ross, Phys. Rev. B. 46
(1992) 1649.
[6]. N. R. Avery, Surf. Sci. 131 (1983) 501.
[7]. N. A. Avery, J. Am. Chem. Soc. 107(1985)6711.
[8]. N. A. Avery, Langmuir. 4 (1988) 445.
-B-
[9]. C. J. Baddeley, R. M. Ormerod, A. W. Stephenson, and R. M. Lambert, J. Phys.
Chem. 99 (1995) 5146.
[10]. K. Balakrishnan, and J. Schwank, J. Catal. 132(1991)451.
[11]. W. F. Banholzer, J. M. Gohndrone, G. H. Hatzikos, J. F. Lang, R. I. Masel,
Y. O. Park, K. Stolt,. J. Vac. Sci. Technol. A 3 (1985) 1559.
[12], M. A. Barteau, E. I. Ko and R. J. Madix, Surf. Sci. 102 (1981) 99.
[13]. B. Beden, F. Kardigan, C. Lamy and J. Leger, J. Electroanal. Chem. 127 (1981) 75.
[14], B.Beden,S. Juanto,J.M . Leger and C. Lamy, J. Electroanal. Chem., 238 (1987) 323.
[15]. B. Beden, M. C. Morin, F Hahn and C. Lamy, J. Electroanal. Chem., 229 (1987) 353.
[16]. R. J. Behm, P. A. Thiel, P. R. Norton and G. Ertl, J. Chem. Phys. 78 (1983) 7437.
[17]. R. J. Behm, W. Hosier, E. Ritter and G. Binnig, Phy. Rev. Lett; 56 (1986) 228.
[18]. E. M. Belgsir, H. Huser, J. M. Leger and C. Lamy, J. Electroanal. Chem., 225 (1987)
281.
[19]. J. C. Bertolini, J. Massardier, and G. Dalmai-Imelik, J. Chem. Soc. Faraday Trans. I
(1978) 1720.
[20]. G. Blyholder, J. Phys. Chem. 68 (1964) 2772.
178
with permission of the copyright owner. Further reproduction prohibited without permission.
[21]. G. Borden, G. Pirug, and H. P. Bonzel, Surf. Sci. 72 91978) 45.
[22]. J. K. Brown, A. C. Luntz, Chem. Phys. Lett. 204 (1993) 451.
[23]. W. A. Brown, P. Gardner, D. A. King, J. Phys. Chem. 99 (1995) 7074.
[24]. G. V. Biinau, T. Wolff, Photochemie (VCH, Weinheim, 1987). p. 178.
[25]. R. Burch, Acc. Chem. Res. 15 (1982) 24.
-C-
[26]. S. A. Campbell, and R. Parsons, J. Chem. Soc. Faraday Trans., 88(6), (1992) 833.
[27]. A. Cassuto, M. Mane, M. Husenschmidt, P. Dolle, and J. Jupille, Surf. Sci. 237
(1990) 63.
[28]. A. Cassuto, M. Mane, J. Jupille, G. Tourillon, and P. Parent, J. Phys. Chem. 96
(1992) 5987.
[29]. K. J. Cathro, J. Electrochem. Soc., 116 (1969) 1608.
[30]. P. Cobden and B. E. Nieuwenhuys, Surf. Sci. 262 (1992) 97.
[31]. R. D. Cortright and J. A. Dumesic, J. Catal. 148(1994)771.
[32]. R. D. Cortright and J. A. Dumesic, J. Catal. 157 (1995) 576.
[33]. F. A. Cotton, G. Wilkinson, G. Advanced Inorganic Chemistry 3r d ed. (Wiley, New
York, 1972) p. 713.
[34]. A. Crossley and D. A. King, Surf. Sci. 68 (1977) 528.
-D-
[35]. J. L. Davis and M. A. Barteau, Surf. Sci., 187 (1987) 38S.
[36]. Y. Debauge, M. Abon, J. C. Bertolini, J. Massardier, A. Rochefort, Appl. Surf. Sci. 90
(1995) 15.
-F-
[37]. T. E. Fischer, S. R. Kelemen, and H. P. Bonzel, Surf. Sci. 64 (1977) 157.
[38]. T. E. Fisher, S. R. Kelemen, J. Catal. 53 (1978) 24.
[39]. T. E. Fischer and S. R. Kelemen, J. Vac. Sci. Technol. 15 (1978) 607.
[40]. G. B. Fisher and J. L. Gland, Surf. Sci., 94 (1980) 446.
[41]. N. Freyer, G. Pirug, and H. P. Bonzel, Surf. Sci. 125 (1983) 327.
179
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
[42]. N. Freyer, G. Pirug, and H. P. Bonzel, Surf. Sci. 126 (1983) 487.
-G-
[43]. M. Galeotti, A. Atrei, U. Bardi, G. Rovida, and M. Torrini, Surf. Sci. 313 (1994) 349.
[44]. P. Gardner, M. Tiishaus, R. Martin, A. M. Bradshaw, Vacuum 41 (1990) 304.
[45]. P. Gardner, M. Tiishaus, R. Martin and A. M. Bradshaw, Surf. Sci. 240 (1990) 112.
[46]. A .J.G ellm an,J.A m .Chem .Soc. 113(1991)4435.
[47]. T. M. Gentle and E. L. Muettertles, J. Phys. Chem. 87 (1983) 2469.
[48]. K. D. Gibson and L. H. Dubois, Surf. Sci., 223 (1990) 59.
[49]. S. Gilman and M. W. Breiter, J. Electrochem. Soc., 109 (1962) 1099.
[50]. R. J.Gorte, L. D. Schmidt, J.L . Gland, Surf. Sci. 109(1981)367.
[51]. X.-C. Guo, A. Hopkinson, J. M. Bradley and D. A. King, Surf. Sci. 278 (1992) 263.
-H-
[52]. A. E. Hamdaoui, G. Bergeret, J. Massadier, M. Primet, A. Renouprez,. J. Catal. 148
(1994)47.
[53]. A. N. Haner and P. N. Ross, J. Phys. Chem., 95 (1991) 3740.
[54]. P. Heilmann, K. Heinz, K. Muller, Surf. Sci. 83 (1979) 487.
[55]. H. Hirano, T. Yamada, K. Tanaka, J. Siera and B. E. Nieuwenhuys, Vacuum 41
(1990) 134.
[56]. H. Hirano, T. Yamada, K.I. Tanaka, J. Siera, P. Cobden, B. E. Nieuwenhuys, Surf.
Sci. 262 (1992) 98.
[57]. M. Hoheisel, J. Kuntze, S. Speller, A. Postnikov, W. Heiland, I. Spolveri, U. Bardi, to
be published.
[58]. A Hopkinson and D. A. King, Chem. Phys. 177 (1993) 433.
[59]. B.E.Hyden, Surf. Sci. 131 (1983)419.
- I -
[60]. H. Ibach and S. Lehwald, J. Vac. Sci.Technol. 15 (1978) 407.
[61] H. Ibach, S. Lehwald, Surf. Sci. 76 (1978) 1.
[62], H. Ibach and S. Lehwald, Surf. Sci., 91 (1980) 187.
180
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
[63],
[64],
[64],
[65].
[66]
[67],
[68].
[69].
[70],
[71].
[72].
[73].
[74].
[75],
[76],
[77],
[78].
[79],
[80],
[81].
T. Iwasita, W. Vielstich and E. Santos, J. Electroanal. Chem., 229 (1987) 367.
-J-
T. E. Jackman, K. Griffiths, J. A. Davies and P. R. Norton, J. Chem. Phys. 79 (1983)
3529.
M. M. P. Janssen and J. Moolhuysen, Electrochim. Acta., 21(1976) 861.
M. M. P. Janssen and J. Moolhuysen, Electrochim. Acta., 21 (1976) 869.
M. M. P. Janssen and J. Moolhuysen, J. Catal., 46 (1977) 289.
L. Q. Jiang and B. E. Keol, J. Phy. Chem., 96 (1992) 8694.
L. Q. Jiang, B. E. Koel and J. L. Falconer, Surf. Sci. 273 (1992) 273.
S. K. Jo, J. A. Polanco and J. M. White, Surf. Sci., 253 (1991) 233.
S. Johnson and R. D. Madix, Surf. Sci. 83 (1979) 487.
-K-
F. Kadigam, B. Beden, J. M. Leger and C. Lamy, J. Electroanal. Chem., 1 (1981) 89.
L. L. Kesmodel, L. H. Dubois and G. A. Somoijai, J. Chem. Phys. 70 (1979) 2180.
J. T. Kummer, J. Phys. Chem. 90 (1986) 4747.
K. Kunimatsu, J. Electroanal. Chem., 213 (1986) 149.
-L-
C. Lamy, Electrochim. Acta., 29 (1984) 1581.
E. Langenbach, A. Spitzer and H. Luth, Surf. Sci., 147 (1984) 179.
F. A. Lee, C. J. Baddeley, C. Hardacre, G. D. Moggridge, R. M. Ormerod, R. M.
Lambert, J. P. Candy, and J. M. Basset, J. Phys. Chem. B 101 (1997) 2797.
Y. Li and B. E. Koel, Surf. Sci. 330 (1995) 193.
Y. X. Li, K.J.Klabunde and B.H. Davis, J. Catal. 128(1991) 1.
H. Lieske and J. Volter, J. Catal. 90 (1984) 96.
J. R. Lomas, C. J. Baddeley, M. S. Tikhov and R. M. Lambert, Langmuir. 11 (1995)
3048.
181
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
[82].
[83].
[84].
[85].
[86],
[87].
[88],
[89].
[90],
[91].
[92],
[93],
[94],
[95].
[96].
[98],
[99].
-M-
E. Marlen, P. Beccat, J. C. Bertolini, P. Delichere, N. Zanier and B. Didillon, J. Catal.
159 (1996) 178.
R. Martin, P. Gardner and A. M. Bradshaw, Surf. Sci. 342 (1995) 69.
C. E. Megiris, P. Berlowitz, J. B. Butt, and J. J. Kung, Surf. Sci. 159 (1985) 184.
G. Meitzner, G. H. Via, F. W. Lytle, S. C. Fung, and J. H. Sinfelt, J. Phys. Chem. 92
(1988)2925.
J. B. Miller, H. R. Siddiqui, S. M. Gates, J. N. Russel, J. T. Yates, Jr., J. C. Tully and
J.ach and S. Lehwald, Surf. Sci., 91 (1980) 187.
W. S. Millman, W. K. Hall,. J. Phys. Chem. 83 (1979) 427.
-N-
P. R. Norton, J. A. Davies D. K. Creber, C. W. Sitter and T. E. Jackman, Surf. Sci.
108 (1981)205.
P. R. Norton, J. A. Davies and T. E. Jackman, Surf. Sci. 122 (1982) L593.
- O -
P. Olivi, B. Beden, F. Hahn, J. M. Leger and C. Lamy, J. Electroanal. Chem., 346
(1993)415.
R. M. Ormerod, and R. M. Lambert, Catal. Lett. 6 (1990) 121.
R. M. Ormerod,C. J. Baddeley, and R. M. Lambert, Surf. Sci. 259 (1991) L709.
S. H. Overbury. D. R. Mullins, M. T. Paffeett, and B. E. Koel, Surf. Sci. 254 (1991)
45.
S. H. Overbury, D. R. Mullins, M. T. Paffett and B. E. Koel, Surf. Sci., 313 (1994)
349.
-P-
M. T. Paffett, and R. G. Windham, Surf. Sci. 208 (1989) 34.
M. T. Paffett, S. C Gebhard, R. G. Windham, and B. E. Koel, Surf. Sci. 223 (1989)
449.
M. T. Paffett, S. C. Gebhard, R. G. Windham and B. E. Koel, J. Phys. Chem., 94
(1990) 6831.
M. T. Paffett, A. D. Logan, R. J. Simonson and B. E. Koel, Surf. Sci. 250 (1991)
123.
182
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
100],
101],
102].
103].
104].
105].
106].
107],
108].
109].
110].
111].
112].
113].
114].
115].
116].
117].
118].
119].
120].
121].
C. Panja, B. E. Koel, Israel J. Chem. 38 (1998) 365.
C. Panja, N. Saliba and B. E. Koel; to be published.
D. H. Parker, M. E. Jones and B. E. Koel, Surf. Sci., 233 (1990) 65.
R. Parsons and T Vandemoot, J. Electroanal. Chem., 257 (1988) 9.
C. H. Patterson, and R. M. Lambert, J. Am. Chem. Soc. 110 (1988) 6871.
K. J. Pierce, and M. A. Barteau, J. Phys. Chem. 98 (1994) 3882.
G. Pirug, H. P. Bonzel, H. Hopster, H. Ibach, J. Chem. Phys. 71 (1979) 593.
-Q -
K. T. Queeney, C. M. Friend, J. Chem. Phys. 107 (1997) 6432.
-R-
J. Radnik, F. Gitmans, B. Pennemann, K. Wandelt, Surf. Sci. 287/288 (1993)330.
P. A. Redhead, Vacuum, 12 (1962) 203.
E. Ritter, R. J. Behm, G. Potschke and J. Wintterlin, Surf. Sci. 181 (1987) 403.
T. G. Rucker, M. A. Logan, T. M. Gende, E. L. Muetterties and G. A. Somorjai, J.
Phys. Chem. 90 (1986) 2703.
-S-
J. W. A. Sachtler, and G. A. Somoijai, J. Catal. 81 (1983) 77.
W. M. H. Sachtler, Handbook of Catalysis, to be published.
N. A. Saliba, D. E. Beck, C. Baur, B. E. Koel, in preparation.
M. Salmeron and G. A. Somoijai, J. Phys. Chem. 86 (1982) 341.
S. B. Schwartz, G. B. Fisher, L. D. Schmidt, J. Phys. Chem. 92 (1988) 389.
B. A. Sexton, K. D. Rendulic and A. E. Hughes, Surf. Sci., 121 (1982) 181.
M. Shelef, Catal. Rev. Sci. Eng. 11 (1975) 1.
M. Shibata and S. Motoo, J. Electroanal. Chem., 229 (1987) 385.
J. H. Sinfelt, Bimetallic Catalysts: Discoveries, Concepts and applications, Wiley, New
York, 1983.
S. K. So, R. Franchy, W. Ho, J. Chem. Phys. 91 (1989) 5701.
183
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
[122]. S. K. So, R. Franchy, W. Ho, J. Chem. Phys. 95 (1991) 1385.
[123]. R. Srinivisan, R. J. DeAngelis and B. H. Davis, J. Catalysis. 106 (1987) 449.
[124]. R. Srinivasan, L. A. Rice, and B. H. Davis, J. Catal. 129 (1991) 257.
[125]. S. Surampudi, S. R. Narayanan, E. Vamos, H. Frank, G. Halpert, A. La Conti, J.
Kosek, G. K. Surya Prakash and G. A. Olah, J. Power Sources, 47 (1994) 377.
[126]. S. Szabo, J. Electroanal. Chem., 172 (1984) 359.
-T-
[127]. T. Takaoka, T. Sekitani, T. Aruga, and M. Nishijima, Surf. Sci. 306 (1994) 179.
[128]. K. C. Taylor, Catal. Rev. Sci. Eng. 35 (1993) 457.
[129]. P. A. Thiel, R. J. Behm, P. R. Norton and G.Ertl, Surf. Sci. 121 (1982) L553.
[130]. P. A. Thiel, R. J. Behm, P. R. Norton, G. Ertl, J. Chem. Phys. 78 (1983) 7437.
[131]. P. A. Thiel, R. J. Behm, P. R. Norton and G. Ertl, J. Chem. Phys. 78 (1983) 7448.
[132]. P. A. Thiel and T. E. Madey, Surf. Sci. Rep., 7 (1987) 211.
[133], Y. L. Tsai, and B.E. Koel, Surf. Sci. 385 (1997)37.
[134]. Y.- L. Tsai, C. Xu and B. E. Koel, Surf. Sci., Submitted.
[135]. W. T. Tysoe, G. L. Nyberg and R. M. Lambert, Surf. Sci. 135 (1983) 128.
[136]. W. T. Tysoe, G. L. Nyberg, and R. M. Lambert, J. Chem. Soc. Chem. Commu.
(1983) 623.
- U -
[137]. W. Unger and D. Marton, Surf. sci. 218 (1989) L467.
-V-
[138]. J. Vaari, J. Lahtinen, and P. Hautojarvi, Catal. Lett. 44 (1997) 43.
[139]. J. S. Villarrubia, W. Ho, J. Chem. Phys. 87 (1987) 750.
[140]. M. R. Voss, H. Busse and B.E. Koel, Surf. Sci.414 (1998) 330.
-W-
[141]. J. wang and R.I. Masel, Surf. Sci., 243 (1991) 199.
184
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
[142]. K. Wang, H. A. Gasteiger, N. M. Markovic and P. N. Ross, Jr, Electochimica Acta.,
41 (1996) 2587.
[143]. K. Wang, H. A. Gasteiger, N. M. Markovic and P. N. Ross, Jr, Electochimica Acta.,
41 (1996) 2587.
[144]. F. T. Wagner and T. E. Moylan, Surf. Sci., 191 (1987) 121.
[145]. J. Willsam and J. Heitbaum, Electroctiim. Acta., 31 (1986) 943.
[146]. R. G. Windham, and M. T. Paffett, Surf. Sci. 208 (1989) 34.
[147]. R. G. Windham, M. E. Bartram and B. E. Koel, J. Phys. Chem. 92 (1988) 2862.
-X-
[148]. C. Xu, J. W. Peck, and B. E. Koel, J. Am. Chem. Soc. 115 (1993) 751.
[ 149]. C. Xu, and B . E. Koel, Surf. Sci 304 (1994) 249.
[150], C. Xu and B. E. Koel, Surf. Sci., 304 (1994) L505.
[151]. C. Xu, B.E. Koel, Surf. Sci. 310 (1994) 198.
[152]. C. Xu, Y.-L. Tsai and B. E. Koel, J. Phys. Chem. 98 (1994) 585.
[153]. C. X u,B.E. Koel,andM .T.Paffett,Langmuir 10 (1994) 166.
[154]. C. Xu and B. E. Koel, J. Chem. Phy., 100 (1994) 664.
-Y-
[155]. Q. Y. Yang, K. J. Maynard, A. D. Johnson and S. T. Ceyer, J. Chem. Phys. 102
(1995) 7734.
[156]. Y. Y. Yeo, L. Vattuone and D. A. King, J. Chem. Phys. 104 (1996) 3810.
-Z-
[157]. D. Yu. Zemlyanov, M. Smirnov, V. V. Gorodetskii,. React. Kinet. Catal. Lett. 53
(1994)97.
[158]. Y. Zhou and S. M. Davis, Catalysis Lett. 15 (1992)51.
185
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.

Linked assets

University of Southern California Dissertations and Theses

Conceptually similar

PDF

Chemisorption studies of hydrocarbons on clean and hydrogen precovered platinum(111) and tin/platinum(111) surface alloys

PDF

Characterization of ordered tin-platinum surface alloys and tin-oxide overlayers on platinum(111) using ultra-high vacuum-scanning tunneling microscopy

PDF

Crotonaldehyde hydrogenation on platinum(111)-based model catalysts

PDF

Chemisorption studies of oxygen-containing molecules on Au(211), Pt(111) and Sn-modified Pt(111) surfaces

PDF

Electrochemical test methods for advanced battery and semiconductor technology

PDF

Developments of chromium-free conversion coatings on galvanized steels

PDF

Application of electrochemical methods in corrosion and battery research

PDF

Expression and regulation of B7 on antigen presenting cells

PDF

Creep deformation in superplastic yttria -stabilized zirconia ceramics and yttria-stabilized zirconia /alumina composites

PDF

Cyclostratigraphy and chronology of the Albian stage (Piobbico core, Italy)

PDF

A syntax-based statistical translation model

PDF

Development of steel foam processing methods and characterization of metal foam

PDF

EFAB: A novel, high-aspect-ratio true three-dimensional microfabrication process for rapid, low-cost desktop micromachining of MEMS

PDF

Combining compile -time and run -time parallelization

PDF

Evaluation of the superplastic potential in commercial aluminum alloys through equal -channel angular pressing

PDF

III-V compound semiconductor nanostructures by selective area growth using block copolymer lithography

PDF

An implicit-based haptic rendering technique

PDF

Functional testing of constrained and unconstrained memory using march tests

PDF

Deformation of bulk nanocrystalline aluminum alloys

PDF

Endogenous regulatory factors in the inhibition and down-regulation of immune responses

Asset Metadata

Creator Panja, Chameli (author)

Core Title Chemisorption on the (111) and (100) faces of platinum-tin bimetallic surfaces

School Graduate School

Degree Doctor of Philosophy

Degree Program Materials Science

Publisher University of Southern California (original), University of Southern California. Libraries (digital)

Tag engineering, materials science,OAI-PMH Harvest

Language English

Contributor Digitized by ProQuest (provenance)

Advisor Koel, Bruce (committee chair), [illegible] (committee member)

Permanent Link (DOI) https://doi.org/10.25549/usctheses-c16-65326

Unique identifier UC11327958

Identifier 3018024.pdf (filename),usctheses-c16-65326 (legacy record id)

Legacy Identifier 3018024.pdf

Dmrecord 65326

Document Type Dissertation

Rights Panja, Chameli

Type texts

Source University of Southern California (contributing entity), University of Southern California Dissertations and Theses (collection)

Access Conditions The author retains rights to his/her dissertation, thesis or other graduate work according to U.S. copyright law. Electronic access is being provided by the USC Libraries in agreement with the au...

Repository Name University of Southern California Digital Library

Repository Location USC Digital Library, University of Southern California, University Park Campus, Los Angeles, California 90089, USA