LEABHARLANN CHOLAISTE NA TRIONOIDE, BAILE ATHA CLIATH TRINITY COLLEGE LIBRARY DUBLIN OUscoil Atha Cliath The University of Dublin
Terms and Conditions of Use of Digitised Theses from Trinity College Library Dublin
All material supplied by Trinity College Library is protected by copyright (under the Copyright and Related Rights Act, 2000 as amended) and other relevant Intellectual Property Rights. By accessing and using a Digitised Thesis from Trinity College Library you acknowledge that all Intellectual Property Rights in any Works supplied are the sole and exclusive property of the copyright and/or other I PR holder. Specific copyright holders may not be explicitly identified. Use of materials from other sources within a thesis should not be construed as a claim over them.
A non-exclusive, non-transferable licence is hereby granted to those using or reproducing, in whole or in part, the material for valid purposes, providing the copyright owners are acknowledged using the normal conventions. Where specific permission to use material is required, this is identified and such permission must be sought from the copyright holder or agency cited.
By using a Digitised Thesis, I accept that Trinity College Dublin bears no legal responsibility for the accuracy, legality or comprehensiveness of materials contained within the thesis, and that Trinity College Dublin accepts no liability for indirect, consequential, or incidental, damages or losses arising from use of the thesis for whatever reason. Information located in a thesis may be subject to specific use constraints, details of which may not be explicitly described. It is the responsibility of potential and actual users to be aware of such constraints and to abide by them. By making use of material from a digitised thesis, you accept these copyright and disclaimer provisions. Where it is brought to the attention of Trinity College Library that there may be a breach of copyright or other restraint, it is the policy to withdraw or take down access to a thesis while the issue is being resolved.
By using a Digitised Thesis from Trinity College Library you are bound by the following Terms & Conditions. Please read them carefully.
THE ADSORPTION OF AMINES AND
OTHER SMALL MOLECULES ON
A thesis submitted to
THE UNIVERSITY OF DUBLIN
for the degree of
DOCTOR IN PHILOSOPHY
EVELYN NOELLE TIMOTHY, B.Sc., MRSC
Department of Chemistry
University of Dublin
1 0 FEB 2004
LIBRARY D U B L IN ^
This thesis has not been submitted as an exercise for a degree at any other university. Except where stated, the work described therein was carried out by me alone.
I give permission for the Library to lend or copy this thesis upon request.
This thesis presents an investigation of the adsorption and thermal decom position of m ethylam ine, dim ethylam ine, trim ethylam ine and s-triazine on P t(llO ) and carbon m onoxide, nitric oxide and nitrous oxide on Pt(331) using low
energy electron diffraction, thermal desorption spectroscopy and angle resolved ultra violet photoelectron spectroscopy.
Therm al desorption experim ents show that m ethylam ine decom poses to H2, H CN and C2N2 following thermal treatment. A second m inor pathw ay produces CH4 and N2. H2 and D2 co-adsorption studies reveal that HCN is the product o f a decom position pathway leaving intact H -CN bonds. No C2N2 is observed during the co-adsorption experim ents proving that CN exists as individual CN fragm ents on the surface and not as a surface polymer. These fragm ents com bine with H(a) in excess hydrogen form ing a second HCN phase.
s-Triazine thermal decom position also releases HCN, C2N2 and H2. C o adsorption of s-triazine with H2 and D2 reveal that the main low tem perature HCN peak is the result of decom position leaving intact H-CN bonds while the high tem perature shoulder is the result of H(a) and CN(a> recom bination. Cyanogen form ed during C3H3N3 decom position is the result o f CN(a) recom bination. The s-triazine ring has broken down by 500K.
N o n e o f the above m o lec u le s form ordered L E E D patterns on the P t(llO )
surface or rem ove the (1 x 2 ) reconstruction. A (1 x 1 ) pattern is obtained by heating the
m eth yla m in e/s-triazin e co v ered surfaces to 6 0 0 K lea v in g o n ly CN(a).
A dsorption and desorption o f CO and N O are m olecu lar on P t(3 3 1 ). Initial
adsorption is on to the step sites. Therm al desorption occurs as tw o p hases - a low
tem perature peak representing desorption from the terraces and a high tem perature
peak representing step site desorption. A t saturation cov era g e 57% o f C O is adsorbed
on the terraces w h ereas 57% o f N O is adsorbed on the step sites. T h e surface
concentration o f C O on the steps is 1.5 greater than on the terraces. W ith N O this
valu e increases to 2.6 . A R U P S exp erim en ts reveal a 0 .4 e V sh ift in the binding
en ergies o f the 4 a , 5 a and Iti orbitals o f CO on g o in g from a saturated surface to a
surface co m p risin g o n ly step site C O con firm in g the ex iste n c e o f tw o typ es o f C O on
N2O and O2 do not adsorb on P t(3 3 1 ) at room tem perature.
C o-ad sorp tion o f step site N O w ith CO on P t(3 3 1 ) d o es not result in any ch em ical
reaction products. A transform ation reaction is o b serv ed in w h ich the p resen ce o f the
C O ca u ses a d ecrease in the N O desorption tem perature from 4 4 3 to 3 7 0 K . T his
su g g ests C O is ad sorb in g onto the step sites and cau sin g N O to m o v e to the terraces.
I w ould first like to thank my supervisor Dr. M.E. Bridge for his guidance and support. I am also thankful to Dr. T. M cCabe for his help in collecting the ARUPS data. I am particularly grateful to Dr. A.C. Igoe for his help in getting this project off the ground.
Special m ention m ust go to Dr. C.J. Barnes and Dr. E. A lSham aileh who
supervised the TDS and LEED experim ents o f chapter five which were carried out in their laboratory at DCU.
Thank you to all the technical staff especially Brendan, Fred, Ed and Theresa and to all my colleagues in the chem istry department.
To all my friends, in Galway, Los A ngeles and D ublin, Johanna, Dara, Nuala, Turlough, Bridie, Ed, Joanne, Anthony, Cecile, G uillaum e, Una, Jerom e, Gillian, Phil, D am ien, Trinny and Nigel.
To my parents, Evelyn and N oel, and my brothers, John, Stephen, Jam es and Ronan.
I am indebted to G alw ay County Council and the D epartm ent o f Social,
Com m unity and Fam ily Affairs for financial support. Thanks also to the Kriebel fund for assistance provided in m y first year.
Auger Electron Spectroscopy
atomic mass units
Angle Resolved Ultraviolet Photoelectron Spectroscopy
Electron Energy Loss Spectroscopy
H R EELS
High Resolution Electron Energy Loss Spectroscopy
Low Energy Electron Diffraction
Near-Edge X-ray Absorption Fine Structure
Reflection-Absorption Infra-Red Spectroscopy
Thermal Desorption Spectroscopy
Temperature Programmed Desorption
Ultra High Vacuum
Ultra-violet Photoelectron Spectroscopy
Vacuum Science Workshop
THE SPECIMEN GASES
H3C— N— CH3
D eclaration i
Sum m ary ii
A cknow ledgem ents iv
A bbreviations v
The specim en gases vi
1.1 General Introduction 2
1.2 The Vacuum 6
1.3 The Pt Crystals 7
1.4 Low Energy Electron D iffraction 9
1.5 Therm al D esorption Spectrscopy 15
1.6 Angle Resolved Photoelectron Spectroscopy 19
2 EXPERIMENTAL TECHNIQUES
2.1 The Experim ental Systems 24
2.1.1 The LEED/TPD System 24
2.1.2 The ARUPS System 25
2.2 The Vacuum Pumps 26
2.3 The Gas H andling Lines 29
2.4 Crystal Preparation 29
2.4.1 The Crystals 29
2.4.2 Sample M ounting 30
2.4.3 Sample Cleaning in Vacuo 32
2.5 The Specimen Gases 35
2.6 The Experim ental Techniques 38
2.6.1 Low Energy Electron D iffraction 38
2.6.2 Therm al D esorption Spectroscopy 39
METHYLAMINE AND s-TRIAZINE ON Pt(llO )
3.1 R esults 43
3.1.1 M ethylam ine on P t(llO ) 43
3.1.2 M eth y lam in e and H2/D2 C o -ad so rp tio n 50
3.1.3 s-T riazine on P t(llO ) 53
3.1.4 s-T riazine and H2/D2 C o -ad so rp tio n 55
3.2 D iscussion 58
3.2.1 M eth y lam in e on P t(l 10) 58
22.214.171.124 H2 desorption 58
126.96.36.199 H C N d esorption 59
188.8.131.52 C2N2 desorption 6 6
184.108.40.206 CH4 and N2 d esorption 67
3.2.2 M ethylam ine and H2/D2 C o -ad so rp tio n 6 8
3.2.3 s-T riazine on P t(l 10) 70
3.2.4 s-T riazine and H2/D2 C o-ad so rp tio n 71
3.3 C o n clu sio n s 72
3.3.1 M eth y lam in e on P t(l 10) 72
3.3.2 s-T riazine P t(l 10) 73
4 DIMETHYLAMINE AND TRIMETHYLAMINE ON Pt(llO )
4.1 R esults 75
4.1.1 D im ethylam ine on P t(llO ) 75
4.1 .2 T rim eth y lam in e on P t(l 10) 80
4.2 D iscussion 85
4.2.1 D im ethylam ine on Pt 85
220.127.116.11 28 amu desorption 88
18.104.22.168 LEED experim ents 88
4.2.2 Trim ethylam ine on Pt 89
22.214.171.124 Ha desorption 89
126.96.36.199 CH4 desorption 91
188.8.131.52 HCN desorption 92
184.108.40.206 28 amu desorption 92
220.127.116.11 LEED experim ents 93
4.3 Conclusions 94
4.3.1 D im ethylam ine on P t(llO ) 94
4.3.2 Trim ethylam ine on P t(l 10) 95
CO, NO AND N2
O ON Pt(331)
5.1. Results 98
5.1.1. Carbon M onoxide on Pt(331) 98
5.1.2. Nitric O xide on Pt(331) 104
5.1.3. Nitrous O xide on Pt(331) 108
5.1.4. NO and CO C o-adsorption 109
5.2. D iscussion 111
5.2.1. Carbon M onoxide on Pt(331) 111
5.2.2. Nitric O xide on Pt(331) 116
5.2.3. Nitrous O xide on Pt(331) 118
5.2.4. NO and CO Co-adsorption 119
5.3 Conclusions 120
One of the m otivations behind m odem surface studies is a better understanding of heterogeneous catalysis. The rates of certain chem ical reactions are greatly increased in the presence o f a solid catalyst. It is the surface o f the solid that is responsible for this catalytic activity. The increased reaction rates result from the m odification of at least one of the constituent chem icals to a state with enhanced ability to interact with the other constituents on the solid surface .
One aims, therefore, to understand what these m odifications are, w hether there are new interm ediate species formed, what the rate lim iting steps and activation energies are, what kind o f sites on the catalyst surface are active and how these processes depend on the catalyst material. This know ledge m ight lead to m ore active, more selective or cheaper catalysts.
The approach used to study the adsorption o f gases on metal surfaces is to first investigate highly sim plified versions o f these systems. This involves taking flat, usually low m iller index, faces o f single crystals of the metal o f interest and studying the adsorption and/or co-adsorption of small know n quantities o f the sam ple gas(es) on them in an otherw ise ultra high vacuum (UHV) environm ent. The purpose o f these m ethods is to characterize the surface and the adsorption and reaction processes in fine detail so that the conditions are very well defined.
atom s. S tu d ies o f adsorption on sin g le crystal su rfaces can therefore be d irectly related to in v estig atio n s o f the real catalyst .
In this w ork the therm al d eco m p o sitio n o f m eth yla m in e, d im eth ylam in e, trim eth ylam ine and s-triazin e on the P t( llO ) surface is in v estig a ted as w e ll as the d eso rp tion /d ecom p o sitio n o f carbon m o n o x id e, nitric o x id e , nitrous o x id e and s- triazine on P t(3 31). M eth yla m in e is the sim p lest stable m o le c u le co n ta in in g a C -N sin g le bond, and m ay thus serve as a p rototype for m ore co m p lica te d m o lec u le s.
T he am ines are in v estig a ted w ith a v ie w to com p arin g the results to those p rev io u sly obtained in this laboratory on P t ( l l l ) and P t(3 3 1 ) . M eth ylam in e adsorption is know n to be se n sitiv e to the m etal structure [4 ], V ery d ifferen t results w ere obtained by T h om as and M a sel w hen m eth ylam in e w a s therm ally treated on the (5 x 2 0 ) and (1 x 1 ) recon stru ction s o f Pt(lOO). V ery little, i f any, M A d eco m p o sitio n w a s ob served on the (5 x 2 0 ) reconstructed surface w h ereas 50% o f the M A adsorbed on the (1 x 1 ) surface d eh yd rogen ated to H C N , C2N2 and N2, w ith 30% co m p letely d eco m p o sin g to N2. T h e se authors co n clu d ed that the (1 x 1 ) Pt(lOO) surface is ten tim es more active in partially or c o m p le te ly d e c o m p o sin g m eth yla m in e than the (5 x 2 0 ) surface.
T he surface ch em istry o f platinum has b een w id e ly stu d ied d ue to its catalytic a ctivity . M eth ylam in e has been the su bject o f a n um ber o f in v estig a tio n s on platinum under U H V c o n d itio n s [3, 6 -9 ], A ll o f th ese deal w ith P t ( l l l ) , w h ile  also reports on desorption from the P t(3 3 1 ) surface. T o the k n o w le d g e o f the author n o previou s stu d ies o f m eth y lam in e on the P t ( llO ) surface h a v e b een carried out. On P t ( l l l ) and P t(3 3 1 ) the m ajor desorption products ob serv ed have b een
N2have also b een d etected but th ese are o n ly a fe w percent o f the
m ajor sp ecies. T h e authors agree that therm al d eco m p o sitio n o f m eth ylam in e on th ese su rfaces occurs via a p ath w ay o f d eh ydrogen ation, starting w ith
N-Hb ond cleava ge. Current d isc u ssio n s center on any interm ediates form ed and the exa ct nature o f HCN
Less work has been carried out on dim ethylam ine, trimethylamine and
s-triazine. HCN is thought to polym erize to s-triazine on P d ( l l l ) follow in g high
exposures , s-Triazine has also been proposed as a m odel for a paracyanogen-like
surface polym er formed on Pt(lOO)  and P t ( l l l )  fo llo w in g C2N2 adsorption.
D espite this interest very little thermal desorption data is available for this compound.
Thermal decom position on P t ( l l l ) yielded H C N , C2N2 and H2 . H el
photoelectron spectra o f the adsorbed m olecule recorded by Som ers could not be
aligned with the gas phase spectrum suggesting the adsorbed m olecule was not
com p letely intact at room temperature.
Igoe  found di- and trimethylamine decom posed on P t ( l l l ) and (331)
yielding H2, CH4 and HCN. HCN was concluded to result from decom position
leaving intact H-CN bonds and not from H(a) and CN(a) recombination. Kang and
Trenary  observed only H2 and HCN desorption follow in g low temperature
exposure o f DM A to P t ( l l l ) . These authors believe D M A dehydrogenates leaving
m ethylam inocarbyne which in turn decom poses to several intermediates, including
m ethyl isocyanide, surface CN and possibly CNH2. D M A decom position on
Si(lOO)-( 2x1) also yielded H2 and HCN as the main desorption products as observed by
M ulcahy et al. , A further product at 43 amu was detected at exposures o f 0.25L and above, m ost likely N-m ethylm ethanim ine, CH3N = C H2.
Trimethylamine decom position was investigated by Erley, Xu and
H em m inger on P t ( l l l ) using HREELS, FT-IRAS and TPD . They found
adsorption at 85K to be m olecular with bonding through the N lone pair. Com plete
decom position occurred after heating to 310K yielding H2, N2, som e parent m olecule,
and som e CH3N, a surface intermediate. Pearlstine and Friend  also found TM A
to undergo com plete and irreversible decom position on W(IOO) for exposures lower
than 0.25L producing H2 and N2. At higher exposures traces o f trimethylamine and
CH4 were also detected.
Walker and Stair  found initial TM A adsorption on Mo(lOO) to be
dissociative. Once the atomic species produced had partially passivated the surface
CO has been studied extensively on Pt surfaces using TPD [22-31], LEED
[22, 29, 31, 32-34], ARUPS [35-36], HREELS [24, 29] and IRAS [27, 30, 37],
Adsorption and desorption are molecular. D esorption from the flat (111) surface
occurs as a single main peak at ~ 450K. D esorption from stepped surfaces results in
tw o CO peaks - the low tem perature peak representing teiTace site desorption and the
high tem perature peak representing desorption from step sites. The step sites are
observed to be m ore densely populated. CO adsorbs on two different binding sites on
both steps and terraces - atop sites at low coverage with bridge sites occupying at
higher coverages. Only one TPD study of CO on the Pt(331) crystal is available .
NO [38-47] also desorbs as a single peak from P t ( l l l ) and as separate step
and terrace peaks from stepped Pt surfaces. A dsorption is bridge bonded at low
coverages and was thought to convert to atop NO at higher coverages. Recently, the
assignm ent of the saturation coverage band to atop NO has been questioned. LEED-
IV  and DFT  calculations favor the 3-fold fee hollow sites.
The techniques used in this work are therm al desorption spectroscopy (TDS),
low energy electron diffraction (LEED) and angle resolved ultra-violet photoelectron
spectroscopy (ARUPS). The rest of this chapter deals with the theory behind these
techniques, the structure of the P t(llO ) and (331) surfaces and the need for ultra high
1.2 T H E V A C U U M
In order to study atom ically clean surfaces we must w ork under ultra high vacuum (UHV) conditions. As the concentration o f atoms on the surface o f a solid is of the order of lO'^ cm"^, to keep the surface clean for 1 hour we m ust ensure the flux o f contam inant m olecules incident on the initially clean surface is less than ~ lO'^ molecules/cm^/sec.
From the kinetic theory of gases, the flux F o f m olecules striking the surface of unit area at a given am bient pressure P is
F(atoms/cm^.sec) = 3.51 x 10^^ P(torr)/M (g.mole)T'^^
where M is the average m olar weight o f the gaseous species and T is the temperature.
For exam ple, CO m olecules at room tem perature at 1 torr have an arrival time of 3.88 X 10^*^ m olecules cm'^ s ''. The m onolayer time is about 3 x 10“*^ s at this pressure, 3 s at 10“ torr or alm ost 1 hr at 10”^ torr.
1.3 T H E Pt C R Y S T A L S
The clean P t(llO ) surface is observed to be reconstructed, i.e., the surface exhibits an atomic structure that differs fundam entally from the structure one w ould expect if the bulk structure terminated abruptly at the surface. The LEED pattern obtained from clean P t(llO ) show s a (1x2) surface structure as opposed to the (1x1) pattern expected o f a ‘hill and valley’ structure o f clo se packed rows o f Pt atoms.
F ig. 1.1 R epresentations o f the (1 x 1 ) and (1 x 2 ) surfaces o f P t ( l 10) .
Experimental studies including LEED [49-50], AES  and x-ray diffraction  support the ‘m issing row ’ m odel, in w hich, alternate (110) rows are m issing from the top surface layer as an explanation for the (1x2) pattern.
Pt(331) is a stepped surface with three atom wide (111) oriented terraces interrupted by one atom high (111) oriented steps. There is no observed reconstruction of this surface.
LOW ENERGY ELECTRON DIFFRACTION
The low energy electron diffraction experim ent gives inform ation on the basic
periodicity of the ordered com ponents of both the clean crystal surface and any
overlayer subsequently adsorbed.
A finely focused beam o f m onochrom atic electrons is directed onto the metal
surface and the elastically back-scattered electron beam s are detected and displayed
on a phosphorescent screen.
For a diffraction pattern to occur, the w avelike properties o f the electrons
m ust have a wavelength sim ilar to the interatom ic distances in the crystal lattice.
Substitution o f values into the de Broglie relationship gives the electron
w avelengths in A
= V (150/E)
with E expressed in eV.
Therefore incident electron energies between 20 and 200eV are used which
correspond to w avelengths of between 0.93 and 3.9A. At these energies electrons
interact strongly with m atter with inelastic mean free paths in solids o f the order o f 4
to 20A. As LEED m onitors the angular distribution o f elastically back-scattered
electrons only most of the electrons detected are scattered from the top three or four
There are five types o f tw o-dim ensional Bravais lattices from which all
I d ) le)
F ig. 1.3 T he fiv e tw o-d im en sion al B ravais lattices .
T h e u n it cell in real sp a c e ca n be d e s c rib e d b y the b a sis v e c to rs ai a n d a 2
(fo llo w in g th e n o ta tio n o f E rtl a n d K iip p ers ). T h e re c ip ro c a l sp a c e re p re s e n ta tio n
is d e s c rib e d by a i* a n d aj*. T h e real a n d re c ip ro c a l sp a c e la ttic e s a re re p re s e n te d by
a,.aj* = 6,j
w h e re i, j = 1 o r 2 , a n d 8 ij is th e k ro n e c k e r d e lta fu n c tio n (6 y = 0 if i 7^ j a n d 5ij = 1 if i
= j) . ai* 1 3j fo r i ^ j. In tro d u c in g y an d y*, th e a n g le s b e tw e e n (ai a n d a 2) an d (a i*
a n d a 2*), re s p e c tiv e ly , w e h av e
a i* = 1/a i sin y
a 2* = l / a 2 sin y
T his in v erse relatio n sh ip betw een real and recip ro cal space m eans that a long v ecto r in real space co rresp o n d s to a short v ector in reciprocal space (fig. 1.4).
Fig. 1.4 The principal o f diffraction pattern form ation in a L E E D experim ent. T h e incident electron beam approaches alon g So and the specular beam e x its alon g Sqo .
T he relatio n sh ip b etw een the substrate lattice and any ad sorbate lattice
(represented by the basis vectors bi and b2) can be e x p ressed by the ratios o f the lengths of the basis vectors, and any angle o f rotation 0, betw een the tw o lattices. T h is notation w as p ro p o sed by W o o d  and takes the form (n x m )R 9 w here n = |bi|/|ai|
e.g. Ni (110) c ^ ( 2 x 2 ) - 0 ---S u bstrate ---S u t ^ r a ie C e n tre d ^ u rf a c e /
surface m esh substrate norm al mesh ratios
e .g . P i ( I O O ) ( , 2 x 2 V 2 ) ' * < 5 * - 0 (W
F ig. 1.5 T w o ex a m p les o f W o o d ’s notation for surface structures com pared with the matrix notation. In exam p le (a) for the N i(I lO ) face e x p o se d to o x y g e n the notation can be shortened sligh tly to N i( 1 1 0 ) c 2 - 0 b ecau se the d ep osit m esh is rectangular, aib and a2b represent aj and a2and ais and a2s represent bi and b2 ,
The substrate and adsorbate lattices can also be related by a transformation matrix M such that
b = M .a
where M =
T h e basis vectors are represen ted by
bi = miiai + mi2a2
b2 = 0121 ai + m22a2
T h e elem en ts o f M are determ in ed ex p e rim e n tally from the d iffraction pattern m ea su red on the L E E D screen.
T he areas o f the tw o unit m eshes are given by
A = |ai X a2| B = |b| X b2|
It can be show n that
A = B d et M
w here the d eterm in an t o f M is d efin ed as
det M = m i i m22 - m2i mi2
(a ) S I M P L E .d e l M in teg er
A 2 X 1 overlayer
(b) C O IN C ID E N C E , det M a ratio n al fraction
• ( b
F ig. 1.6 T h e re la tio n s h ip b e tw e e n su rfa c e an d b u lk m esh es. T h e s im p le a n d c o in c id e n c e m esh es are
illu s tra te d b y th e c a s e s o f a d s o rb a te a to m s (fille d c irc le s ) o n the b u lk e x p o se d (1 1 0 ) p la n e
THERMAL DESORPTION SPECTROSCOPY
W hen a crystal is heated resistively, any adsorbed species on the surface may desorb. This is because its surface residence tim e depends exponentially on
tem perature (x = To exp(AE/RT)). If the adsorbate is not re-supplied from the gas
phase, its surface concentration dim inishes rapidly with increasing tem perature until the surface becomes clean. This phenom enon is the basis of therm al desorption spectroscopy (TDS). In this technique the partial pressure of the desorbing species is m onitored as a function o f temperature.
A nalysis of desorption spectra yields inform ation on the nature and num ber of adsorbed species, the num ber o f various desorbing phases for each species and their relative populations, the activation energies of desorption
(Ed)of the individual phases and the order o f the desorption process.
W here the pum ping rate is very large, com pared with the desorption rate, the pressure rise is sim ply proportional to the desorption rate and a peak in the thermal desorption spectrum at some tem perature Tmax indicates a m axim um desorption rate at that temperature.
The rate o f desorption from unit surface area m ay be w ritten as
-d0/dt = Vn6" exp(-Ed/RT) (1.5.1)
where n is the order o f the desorption reaction, 0 is the surface coverage
Using a linear change of sample temperature with time (T = Tq + [3t) and
assuming that Ed is independent of 0 Redhead  solved equation (1.5.1) to find the
temperature (Tp) at which the desorption rate is a maximum
Ed/RTp^ = (vi/P) exp(-Ed/RTp) for n = 1 (1.5.2)
where (3 is the heating rate and vi is the rate constant for a first order desorption
We can see from equation (1.5.2) that Tp is independent o f coverage for a first
order reaction with constant Ej and thus Ej can be found directly from a measurement of Tp provided a value of V| is assumed. The relation between Ej and Tp is very nearly
linear and, for lO'^ > V|/(3 > 10^ (K~'), is given to ±1.5% by
Eci/RTp = / « ( v , T p / p ) - 3 . 6 4 (1.5 .3 )
The activation energy can be determined without assuming a value of the rate
constant by varying (3 and plotting log Tp vs. log (3. Ej can then be obtained from the relation
Ed/RTp+ 2 = d(log P)/d(log Tp) (1.5.4)
The rate constant can then be found by substituting Ed in eqn. (1.5.2).
For second order procedures Redhead derived the following equation
Ed/RTp^ = 0ov^/p exp(-EdZRTp) (1.5.5)
where Go is the initial surface coverage.
Tp now depends on surface coverage. 0o may be found from the area under the
Redhead thus states that the order o f the desorption reaction can be
determ ined from the behavior of the m axim um in the desoprtion rate curves with
coverage. If a peak in the desorption rate curve does not change in tem perature with
coverage, the reaction is first order with a fixed activation energy o f desorption. If the
tem perature of the peak decreases with increasing coverage, the reaction may be
second order with fixed Ed, or first order with an Ed dependant on coverage. To
distinguish between these two cases one plots log(0oTp^) V5. 1/Tp - a second order
reaction with fixed Ed will yield a straight line.
The shape of the experim ental desorption rate curve as a function of
tem perature can also be used to determ ine the order o f the reaction. Redhead found
that the half-w idth of a peak at half m axim um is independent o f the initial coverage
for first order processes. For a second order process the half width changes with
Falconer and M adix  derived a m ore general expression valid for all
desorption orders, n, greater than zero
/n(p/Tp^) = /n(Rvn0"'‘/Ed) - Ed/RTp (1 .5 .6 )
A plot of /n(P/Tp^) versus 1/Tp for a num ber of heating rates should yield a straight line o f slope - Ej/R.
In general, Tp increases as the heating rate is increased, but since the
dependence of Tp on P is small, changes in P of several orders of m agnitude are
necessary to obtain reasonable accuracy in the determ ination of Ed from heating rate
The desorption activation energy for any n can alternatively be obtained by direct analysis o f the desorption profile. From equation (1.5.1) it can be shown that
/n(-de/dt/0") = -Ej/R T (1.5.7)
By plotting Zn(-d0/dt/9") versus 1/T, using the pressure change as the rate and the area under the higher tem perature portion o f the spectrum as a m easure o f the coverage, a straight line of slope -Ed/R is obtained for the appropriate desorption order.
ANGLE RESOLVED PHOTOELECTRON SPECTROSCOPY
Photoelectron spectroscopy involves the m easurem ent o f the kinetic energies
Ek of photoelectrons ejected from a surface by m ono-energetic radiation. From the
kinetic energies one can determ ine the binding energies, intensities, and angular
distributions of these electrons and use this inform ation as a probe to elucidate the
electronic structure of m olecules and ions.
In XPS the energies of the photons used lie in the X -ray region and are high
enough (~ 1000-1500eV) to eject electrons from core level orbitals. As the deep core
electrons do not participate in bonding and are characteristic o f the atom from which
they originate this technique is particularly useful for elem ental analysis.
U ltraviolet photons (lO-lOOeV) excite em ission from valence levels. As
valence electrons are used in chem ical bonding UPS is well suited to the study of
bonding at surfaces. It can provide m easurem ents o f the w orkfunction and band
structure of the surface and adsorbed layers.
W hen a photon of energy hv interacts with a m olecule that has electrons
bound to it by energy
BEelectrons with kinetic energy
Ekare ejected provided that hv
Conservation o f energy between ion and electron requires that
Ek + BE(1.6.1)
BEis referenced to the vacuum level.
The kinetic energy o f photoelectrons em itted from the surface, however, is
w here Eb is the binding energy o f a core level with respect to the Fermi level and O is the workfunction. (The work function is the m inim um potential the most loosely bound valence electrons in the solid must overcom e to be ejected into the vacuum outside the solid with zero kinetic energy at absolute zero tem perature (fig. 1.7)).
Fig. 1.7 Energy level diagram to d efin e the work function. T he Fermi en ergy Ef refers to the
potential energy at the top o f the valen ce band .
The binding energy o f an electron is the energy difference between the initial and final states o f the atom. That is, the difference in energy betw een an atom with n electrons and the ion with n-1 electrons
BE = Ef(n-1) - Ei(n)
If there were no rearrangem ent of the rem aining electrons, the binding energy
BE w ould equal the negative o f the energy o f the orbital from which the electron was
ejected. This approxim ation is known as K oopm an’s theorem
BE = -£k
T he energies required for equation 1.6.3 are
Ei(n) = <\j/,(n) 1 H ' j \i/i(n)>
Ef(n-1) = < \|/f(n-l) I H ' | \|/f(n-l)> (1.6 .6)
w here i|/i is the initial state o f the atom , v|/f is the final state and H ' is the perturbation
c aused b y the irrad iatin g photon.
T he transition p robability p e r unit tim e betw een these initial an d final states is
given by F e rm i’s G olden R ule
w here 8(Ef - Ej - h\)) is the D irac delta function (equal to unity w hen Ef - Ej - hu = 0
and zero elsew here).
As w ith energy, m om entum m ust be co n serv ed du rin g the transition. T he
photon m om entum is insignificant in the en ergy range used in U PS therefore the
electron m o m en tu m is unchanged by photon absorption i.e. ki ~ kf. T h is m eans only
vertical tran sitio n s are ob serv ed in U PS.
T he photoelectron kinetic energy Ek is related to the parallel and
p erp en d icu lar com p o n en ts o f m om entum , k|| and k i , resp ectiv ely , by
P.f = 2kI\\ I <\j/f I H ' I ij/i> I ^5(Ef - Ei - hi)) (1.6.7)
The m om entum o f the photoelectron in vacuum is directly related to the
m om entum of the electrons in the solid. The angular distribution o f photoelectrons
m ust be m easured in order to determ ine the com ponents of the m om entum . This is the
basis of angle-resolved photoelectron spectroscopy (ARUPS). The variables in an
A RUPS experim ent are shown in figure 1.8.
THE EXPERIMENTAL SYSTEMS
The LEED/TPD System
This bakeable, 12 inch diameter, stainless steel vacuum cham ber was
constructed in the early 1970’s by Vacuum G enerators Ltd., using a method o f
internal argon-arc welding in order to avoid inclusion of crevices as trapped volumes.
A num ber of ports were later added to the cham ber by V acuum Science W orkshop
Ltd., to the design and specification o f M.E. Bridge and J.S. Somers.
The cham ber was fitted with a quadrupole mass spectrom eter (initially VG
Supavac replaced by VGQ Onix) for residual gas analysis and thermal desorption
spectroscopy; reverse view LEED optics and electron gun (O micron); an ionization
gauge (Bayard-A lpert type) for pressure m easurem ent in the torr pressure
range and an ion gun for in situ cleaning o f the sample. A standard Vacuum
G enerators sample m anipulator (U M D l) was also fitted. The platinum crystal was
m ounted on the m anipulator so that it could be turned to the line o f sight of the mass
spectrom eter, the LEED gun, the ion gun and the gas doser.
(The TPD and LEED experim ents of chapter five investigating CO and NO
were carried out at D ublin City University using a sim ilar system which is described
in detail elsew here .)
The ports and flanges on the cham ber were sealed using high purity copper
A section through the experim ental level o f the cham ber is shown
schem atically in fig. 2.1.
AG Argon Ion Gun
E G l Auger Electron Gun
EG2 LEED/Auger Gun
GD Gas Doser
LV Leak Valve
QMS Quadrupole Mass Spectrometer
RFA Retarding Field Analyser
E G l
Fig. 2.1 S chem atic diagram through the experim ental level o f the L E E D /T P D system (not to scale).
2.1.2 The ARUPS System
The ARUPS cham ber was also bakeable, stainless steel and 12 inches in
diam eter. It was fitted with a differentially pum ped noble gas discharge lamp, a
m oveable electron analyser (VSW HA45), a channeltron electron counting system, an
ion gun for surface cleaning and an ion gauge for pressure m easurem ent in the 10'^ to
5 X 1 0 '" torr pressure range (fig. 2.2).
The main cham ber was pum ped by a liquid nitrogen trapped oil diffusion
pum p and a titanium sublim ation pump. Base pressures of 2 x 10"'® torr were
A G Argon Ion Gun
D L Discharge Lamp
EG Electron Gun
Q M S Quadrupole Mass Spectrometer
SDA Spherical D eflection Analyser
V Viewport AC QMS EG V DL SDA V
Fig. 2.2 Schematic diagram through the experimental level o f the ARUPS system (not to scale).
THE VACUUM PUMPS
In order to achieve UHV conditions within the experim ental cham ber, gas
particles must be rem oved from it. Vacuum pum ps were used for this purpose. A
rotary pum p was used first to ‘ro ugh’ pum p the system from atm ospheric pressure to 'y a pressure range within which the diffusion pum p could begin to function (~ 10"
torr). Rotary pum ps are mechanical pum ps which operate by creating a periodically
increasing and decreasing cham ber volume. The diffusion pum p consists basically of
a pum p body with a liquid nitrogen cooled wall and a three or four stage nozzle
system. The low vapor density oil is in a boiler and is vaporized there by electrical
heating. The oil vapor stream s through the chim neys and em erges with supersonic
speed through the different nozzles. This je t of oil vapor widens like an um brella
carrying gas particles dow nw ards with it and com pressing them until they can be
rem oved by the rotary pump. M eanwhile, the oil vapor condenses when it reaches the
pum p w ails and flow s back in liquid form to the boiler. This auxiliary system (rotary
and diffusion pum ps) was also used to pum p the gas handling lines and could be
This system was used to bring the cham ber pressure down to approxim ately
10"^ torr at which stage the ion pum p could be used. The diode ion pum p works by
ionizing the gas in the system in a Penning gas discharge between a cathode and an
anode in the pump. The pump contains two parallel cathodes made of titanium
between which are arranged a system of cylindrical anodes m ade o f stainless steel.
The cathode is m aintained at high negative electrical potential, of the order o f a few
kV, with respect to the anode. The whole electrode assem bly is m aintained in a strong
hom ogeneous m agnetic field. Electrons near the anode are trapped in the high
m agnetic field and set up a region of high electron density in the anode cylinders.
Here the electrons collide with gas particles in the system and ionize them. Due to
their greater mass, these ions will be accelerated tow ard and im pinge on the cathode.
A titanium sublim ation pum p was used to supplem ent the ion pump. This is a
form of getter pum p in which the getter material (titanium ) is evaporated and
deposited on the inner wall o f the cham ber as a getter film. The titanium is
evaporated from a titanium /m olybdenum alloy filam ent by resistance heating.
Particles from the gas which im pinge on the getter film becom e bound to it by
chem isorption and form stable com pounds with the titanium , which have
im m easurably low vapor pressures. This pum p w as not used continuously but
switched on in short bursts of a few m inutes, controlled by a tim er, every few hours.
W ith the use o f these pum ps pressures of below 2 x 10“’^ torr were m aintained
in the main cham ber. A block diagram of the entire pum ping system is shown in fig.
T S P
T ita n iu m S u b ilm a tio n P u m p
Oli D iffusion P u m p
P ira n i G a u g e
L iquid N itro g en T ra p
Leai< V alve
A — is o la tio n V alve
G a s B o ttle Inlet Ion P u m p
R o ta ry P u m p ODP C h a m b e r
T S P
THE GAS HANDLING LINES
S p ecim en gas d o sin g w a s ach iev ed through a g lass g as h and ling line. T his
c o n siste d o f a set o f gas b ottles, each attached to a m a n ifo ld , separated from the m ain
ch am b er by a leak va lv e. G as d o sin g w as a ch iev ed b y back fillin g the ch am b er to a
set pressure in the 5 x 10'^ - 10'^ torr range. A cap illary d oser w a s attached to the leak
v a lv e to in crease the flu x o f the adsorbate gas at the surface o f the sam p le. In order to
reduce the p rob lem o f an u neven flu x distribution b ein g p rod uced at the fa ce o f the
sam p le, the d ista n ce b etw een the gas d oser and the sam p le w a s rela tiv ely lo n g (~
50m m ).
A separate stain less steel line w as u sed to introduce o x y g e n and argon to the
cham ber for clean in g.
T he gas h and ling lin es co u ld be p um ped to pressures <10'^ torr by the
auxiliary p u m p in g system before b ein g filled with the required gases.
T h e initial preparation o f the P t( llO ) and P t(3 3 1 ) sam p les w as carried out at
the M aterial S c ie n c e C entre o f the U n iv ersity o f B irm in gham . A 99 .99% pure sin g le
crystal rod (M eta ls R esearch L td.) w as m ou n ted on an x-ray g on io m eter, oriented to
w ith in 0 .2 5 ° o f the (1 1 0 ) direction and a d isc w as cut o f f in a spark cu ttin g m achine.
T his p rocedure w as repeated in the (3 3 1 ) direction. T h e d iscs are ap proxim ately 9m m
in diam eter and 1m m thick. In preparation for this w ork, the crystals w ere p o lish ed
2.4.2 Sample Mounting
In the TPD /LEED system the platinum crystal under consideration was spot-
w elded between two 0.25 mm tantalum wires, positioned opposite each other on the
edges of the crystal (fig. 2.4). A chrom el/alum el therm ocouple o f 0.05m m diam eter
was also spot-w elded to the edge of the crystal. This assem bly was treated in an
ultrasonic acetone bath to rem ove any residues o f grease that m ight be a source of
The tantalum wires were then clam ped betw een the tw o plates o f a copper
holder. The arms of the copper holder were screw ed to a ceram ic block to provide the
connections for the heating wires and a channel for the electrical feedthroughs to the
therm ocouple wires. A steel driving rod, connected to the sam ple m anipulator, was
attached to the top of this ceram ic block. The sam ple m anipulator allow ed the crystal
to be m anipulated in the x, y, z plane.
The sam ple was heated by passing an a.c. current through the copper holder
and tantalum wires and its tem perature was m onitored by the chrom el/alum el
C eram ic Blocks \
•Copper P la te s
Mica S h e e t s
C hrom el/A lum el T h erm ocou p le
Tantalum H e a tin g , , Wire
'C o p p e r H older
F ig. 2 .4 D iagram o f crystal m ounting in the L E E D /T P D system .
M ounting the crystals in the ARUPS system also involved spot-w elding
tantalum wires to the edge of the crystal (fig. 2.5). These tantalum wires were then
spot-w elded to two tungsten rods, one of which was insulated from earth via a sheet
of mica. These rods were in turn attached to the end of a probe. A chrom el/alum el
therm ocouple was spot-w elded to the edge of the crystal and channeled by a ceramic
tw in-capillary to the electrical feedthroughs. The crystal could be cooled to 85K by a
Ceramic Twin Capillary
To H eater Power
T u n gsten Rod
Fig. 2.5 D iagram o f crystal m ounting in the A R U P S system .
2.4.3 Sample Cleaning in Vacuo
Exposure of the crystals to atm osphere results in considerable surface
contam ination. Once the sam ple is m ounted in the UHV cham ber the contam ination
must be removed. The crystal was first flashed to lOOOK to rem ove any contam inants
w hich could be desorbed into the vacuum as volatile oxides, sulphides or carbides.
M ore stubborn contam inants such as carbon (which form s strongly bound com pounds
with the substrate m aterial) were rem oved by cycles of heating at lOOOK in 10'^ torr
of oxygen and bom barding the crystal with argon ions. Argon bom bardm ent was also
necessary after each amine TDS experim ent as som e C-N bond breakage occurred
leaving C on the surface. Follow ing bom bardm ent the crystal was always annealed at
1300K to restore order to the surface which was left in a heavily dam aged state, with
em bedded argon atoms.
An unm odified, open ended, Bayard-A lpert type ionization gauge (Leisk
IG 32N ) was the source of ions for cleaning in this work. The sample, turned toward
ion pum p and back filled to a partial pressure of 5 x torr with argon. W ith a 1mA
em ission from a standard ion gauge controller (IGP3) an ion current o f 6.5|xA is
obtained at the sample. The ion current could be varied by altering the argon pressure.
T he gauge is connected as a standard ion gauge, so it could be used to m easure
pressure during bom bardm ent.
Ion Gauge(producing Ar* ions)
F ig. 2 .6 D iagram o f the argon bom bardm ent set-up.
The LEED patterns obtained from the clean P t(llO ) and (331) surfaces are
Fig. 2.7 (1x2) LEED pattern o f the clean P t(l 10) surface. Beam energy 59eV.
THE SPECIMEN GASES
M ethylam ine and dim ethyl amine were prepared in this laboratory. They were
prepared by the addition of a solution of the corresponding HCl salt to N aO H pellets
(l.T ratio). The C H3NH2HCI and (CH3)2N H H C1 were supplied by A ldrich Chemical
Com pany, and were of reagent grade. Riedel-de H aen supplied the N aO H pellets
(extra pure). The resultant gases were collected and purified (by at least three freeze-
pum p-thaw cycles) in the vacuum gas-handling line. Purity was determ ined by mass
spectrom etry (figures 2,9 and 2.10). The peak at 40 am u is argon rem aining after a
cleaning cycle. Some of the 18 amu peak results from w ater present as the gases were
not dried after synthesis.
Fig. 2 .1 0 M ass spectrum for dim ethylam ine.
T rim etiiylam ine w as su p p lied by F lu k a Ltd. w ith a p u rity > 99% . A liquid at
room tem perature, it w as tran sferred into a glass gas bottle and su b jected to a nu m b er
o f freeze-p u m p -th aw cycles (fig. 2.11).
s-Triazine was supplied by Sigma with approx. 97% purity. It was subjected
to at least three freeze-pum p-thaw cycles to ensure com plete degassing (fig. 2.12).
F ig. 2 .1 2 M ass spectrum for s-triazine.
Carbon m onoxide, nitric oxide and nitrous oxide were supplied by M esser
G riesheim (purity 99%). Supplied in gas cannisters, they were attached directly to the
Low Energy Electron Diffraction
This apparatus consists of an O M ICRO N SPECTA LEED 4-grid rear view
LEED optics and an internal electron gun with LaBe filam ent (fig. 2.13).
/ / / /
/ / /
S am p le
A2 A3 A4
N s V
G2 G3 G4
Al - A4 Anodes
G1 - G4 Grids
Fig. 2 .1 3 Schem atic o f the L E E D optics.
Electrons from the directly heated filam ent em erge as a divergent beam into
A l. A l, A2 and A3 act as a three elem ent focus lens. Em ission from the filam ent is ~
0.3m A which gives a beam current o f ~ 1.3|iA. The potential on G1 controls the spot
size and intensity.
Im pingem ent of the beam on the crystal produces elastic, inelastic and
secondary electrons. The inelastic and secondary electrons are filtered out by the
incident beam and back-scattered electrons travel in a field free zone. The grid G2 is
held at earth to prevent field penetration from the screen to G3.
The screen is m aintained at 6kV to accelerate the elastically diffracted
electrons tow ard it. W hen the electrons strike the screen they excite its phosphor
coating thus displaying the LEED pattern. The pattern can be viewed and
photographed through the view port opposite.
2.6.2 Thermal Desorption Spectroscopy
In the therm al desorption experim ent the partial pressure of a particular
desorbing product is m onitored as a function of crystal tem perature. A block diagram
of the experim ental set-up is shown in figure 2.14.
Quadrupole Mass Spectrometer
Mass Spectrometer Control Unit
Prior to exposing the crystal to the adsorbate, the mass spectrom eter filam ent
was degassed for 15 mins. The crystal was then flashed to lOOOK and allow ed to cool
to the desired exposure tem perature (~ 300K). Once this tem perature was reached the
crystal was dosed by preferential front face dosing from the capillary doser. (The
degree of exposure is m easured in Langm uirs, L, where IL = 10'^ torr exposure for 1
sec.) W hen the background pressure had settled (come down to 10'^ torr) the thermal
desorption spectrum was taken with the crystal rotated to the line o f sight of the
2.6.3 Angle Resolved Ultraviolet Photoelectron Spectroscopy
The noble gas discharge lamp used in this work could generate photons of
21.2eV (H el) or 40.8eV (H ell). Follow ing irradiation of the sample by these photons
the em itted electrons were energy analyzed in the plane perpendicular to the sample
m anipulator axis. The moveable spherical deflection analyzer was a dispersive
instrum ent so the electron energy distribution could be recorded directly. The angular
resolution of the analyzer is determ ined by the angle subtended by the first aperture
METHYLAMINE AND s-TRlAZINE ON Pt(llO)
Industrially m ethylam ine is produced by the reaction o f am m onia with
m ethanol as follows:
CH3OH + NH3 -> CH3NH2 + H2O
Further reaction with methanol produces di- and trim ethylam ine
CH3NH2 + CH3OH ^ (CH3)2NH + H2O
(CH3)2NH + CH3OH ^ (CH3)3N + H2O
The product ratio can be influenced by altering the reaction conditions. The
reaction m ixture is separated and dried by distillation into three pure products - MA,
DM A and TM A . The chem ical properties of the m ethylam ines are derived from
the reactivity of the am ino group which serves as a convenient means to introduce
m onom ethyl, dim ethyl or trimethyl am ino groups into various organic com pounds.
As such they are valuable starting m aterials for organic com pounds where N-methyl
groups have to be introduced at specific positions .
M ethylam ine is used in tanning, in the m anufacture of dyestuffs and in the
treatm ent of cellulose acetate rayon . Like di- and trim ethylam ine it is also used
as a raw material in the m anufacture of m edicine, pesticides, rubber adjuvant,
explosives, chem ical fibre solvent, surfactant, ion exchange resin and in photography
Methylamine on Pt(llO )
300 400 500 600 700 800 900 1000
3.1). Sets of thermal desorption spectra for each o f these products are shown in
figures 3.2 - 3.6. A small peak in the 30 amu spectrum was also detected (fig. 3.7).
At low exposures of M A (up to 5L), hydrogen desorbs as a single peak (~
415K at 0 = 0.5L and ~ 422K at 9 = 5L). At higher exposures, this peak appears as a
high-tem perature shoulder on the new dom inant peak at 400K.
300 400 500 600 700 800 900
F ig. 3 .2 Therm al desorption spectra o f H2 fo llo w in g exp osu res o f M A to P t ( l 10).
A main peak at ~ 4 3 IK is observed for H CN desorption with a lesser peak at
A main peak at ~ 4 3 IK is observed for H CN desorption with a lesser peak at
- 5 4 IK. The temperature o f the peak maxim um is independent o f coverage.
300 400 500 600 700 800 900
Fig. 3.3 Thermal desorption spectrum o f HCN follow ing a 20L exposure o f M A to Pt( 1 10).
C2N2 desorbs as two broad peaks between 650 and 900K . The main peak
shifts downwards with increasing coverage from ~ 737K (9 = 0.5L ) to ~ 713K (9 =
CjNs/CHjNHj/PtCllO) 52 amu
300 400 500 ISDD 700 800 900 1000
T em p eratu re (K )
Fig. 3.4 Therm al desorption spectra o f C2N2 follow ing exposures o f M A to P t(l 10).
CH4 desorbs as a single peak at ~ 427K. This broad peak starts at ~ 373K and
spans around 235K.
16 am u CH4/CH3NHj/R (110)
300 400 500 600 700 800 900
Three desorption p eaks are o b served in the 2 8 amu spectrum . T h e first peak
has a m axim u m at ~ 4 2 5 K and the third peak at ~ 8 18 K . A s sim ilar peaks are
ob serv ed in the 14 am u spectrum , w e can co n clu d e that th ese peaks are due to N2
d esorption. T he sec o n d peak is b eliev ed to b e C O , fo llo w in g a com parison o f this
peak w ith on e taken o f background C O d esorb in g from the P t(l 10) surface.
28 amu Nj/CHjNH^/PtCllO)
300 400 SOO 600 700 800 900
F ig. 3 .6 Therm al desorption spectrum o f N2 fo llo w in g a 2 0 L exp osu re o f M A to P t(lIO ).
A sm all peak is o b served in the 3 0 am u spectrum at ~ 37 8 K .
b 30 amu 30 amu/CH3NH2/Pt(110)
300 400 500 600 700 800 900
No desorption o f NH
Some o f the HCN spectra recorded in this work showed a prominent peak at ~
390K which was not present in others. This peak was observed by Igoe  at 396K in
the HCN/MA/Pt(331) spectra and at 383K in the H C N /M A /P t(lll) spectra. As this
low temperature spike was not reproducible experiments were carried out to test the
effect o f varying the exposure time and pressure used to achieve a certain coverage
and o f varying the background pressure at which the spectra were recorded. Three
additional HCN spectra were recorded (fig. 3.8) following exposure o f the crystal to
20L o f MA under differing conditions.
In the first o f these, (a), the exposure time and pressure was as before (200s at
10"^ torr) but the spectrum was taken when the background pressure had reached 5 x
10'^ torr instead o f waiting until
10'^ torr as before.
For (b) the 20L coverage was achieved by exposing the crystal at 10'* torr for
2000s, and the spectrum was taken with the leak valve open (i.e. the background
pressure was not allowed to recover).
The crystal was exposed at 5 x 10'^ torr for 40G0s for the third exposure, (c).
Again the spectrum was taken with the leak valve still open.
In all three spectra the 4 3 IK peak is reproduced with the lesser peak at ~
540K. However, in this set o f spectra one can see the additional low temperature peak
which was not observed in the HCN spectrum o f fig. 3.3. This peak reaches a
maximum at ~ 388K in spectrum (a). In (b) the peak is less intense and has its
maximum at ~ 395K. In (c) the additional peak is reduced to a shoulder at ~ 405K on
300 400 500 600 700 800 900
Fig. 3.8 Thermal desorption spectra o f HCN follow ing 20L exposures o f M A under differing
E xposure o f the P t(llO ) surface to m ethylam ine resulted in a decrease in the
intensity of the (1x2) Pt spots and an increase in background intensity as evidenced
by LEED. M ethylam ine did not form any ordered overlayers on the crystal surface at
any coverage and did not rem ove the (1x2) surface reconstruction. H owever, a (1x1)
pattern was observed after heating the M A covered crystal to 600K. Further heating
3.1.2 Methylamine and Hydrogen/Deuterium Co-adsorption on Pt(llO)
M eth y lam in e and h ydrogen co-ad sorp tion on P t( llO ) y ield ed o n ly H C N w h o se spectrum sh o w ed tw o d istin ct peaks at ~ 4 3 0 K (P i) and ~ 5 1 0 K (P2) (fig- 3.9 ). N o C 2 N 2 , C H 4 , N a or C H 3 N H 2 w ere d etected. T h e exp erim en ts w ere carried out by first adsorbing the m eth y lam in e then back fillin g the cham ber w ith 5 x 10'^ torr o f H2 and taking the spectrum w ith the H2 tap open. U nder th ese co n d itio n s any H2 d esorb ing w o u ld not be d etected b eca u se o f the high background H2 pressure.
H C N /C H jN H j + H j/R (1 1 0 )
27 am u
i/i c 0) c
300 400 500 600 700 800
T e m p e ra tu re (K)
F ig. 3.9 Therm al desorption spectrum o f H C N fo llo w in g a 2 L exp osu re o f M A and H2to P t(llO ).
C o-ad sorp tion o f m eth ylam in e and deuterium resulted in three thermal d eco m p o sitio n products - H2, H C N and a peak at 28 amu representing D C N .