The adsorption of simple amines on platinum surfaces

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

Copyright statement

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.

Liability statement

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.

Access Agreement

By using a Digitised Thesis from Trinity College Library you are bound by the following Terms & Conditions. Please read them carefully.

THE ADSORPTION OF SIMPLE AMINES ON

PLATINUM SURFACES

A thesis submitted to

THE UNIVERSITY OF DUBLIN

for the degree of

DOCTOR IN PHILOSOPHY

by

A N TH O N Y IG O E

Department of Chemistry

University of Dublin

Trinity College

Dublin 2

Ireland.

DECLARATION

T h is thesis is su b m itte d by the und ersig n ed to the U n iv ersity o f D u b lin for exam ination for the degree o f D o c to r in P h ilo so p h y . T his thesis has not been su b m itte d fo r a degree in this or any o th er university. E x c e p t w here ackno w led g em en t is given, the w ork described herein is original and carried out by the au th o r alone.

DECLARATION

I agree that the Library may lend or copy this thesis upon request.

SUMMARY

T h is th esis presents the results of an investigation of the adsorption of m ethylam ine,

d im ethylam ine, trim ethylam ine and ethylenediam ine on P t ( l l l ) and Pt(331). The chem ical

co m p o sitio n and geom etric structure o f the clean crystals were characterised by A uger

electron

spectroscopy,

low energy electron diffraction,

and X-ray photoelectron

spectroscopy.

T h erm al desorption experim ents have show n that, on heating the P t ( l l l ) crystal,

m eth y lam in e undergoes decom position into H

2

, H C N and C

2

N

2

. A small am ount o f

m eth y lam in e decom poses com pletely to leave carbon and nitrogen residues on the surface.

T h e decom position o f m ethylam ine on Pt(331) produces the sam e decom position products

as are produced on P t ( l l l ) . H ow ever, H C N desorption occurs at significantly low er

tem peratures. O n both P t ( l l l ) and Pt(331) a high tem perature C

2N2 peak is observed

co n sisten t w ith p rev io u s studies on m ethylam ine, cyanogen and hydrogen cyanide adsorption

on platinum .

D im eth y lam in e decom poses on both P t ( l l l ) and Pt(331) to form H

2

, H C N and CEU-

No C2

N

2 d eso rp tio n is observed, in contrast to m ethylam ine decom position. M ethane

production is in d irect co m p etitio n to H

2 production. The decom position pathw ay involves

cleavage o f th e C -N b o n d , at w hich point either decom position o f the methyl group occurs to

give H

2

, o r the m eth y l group picks up a hydrogen to form m ethane. T rim ethylam ine

produces the sam e d esorp tio n products as dim ethylam ine on platinum and undergoes

decom position in a sim ila r manner. The decom position processes fo r both chem icals are

essentially th e sam e on both P t ( l l l ) and Pt(331). H ow ever, H C N desorbs at low er

tem peratures on P t( 3 3 1) than on P t(l 11), as was noted fo r m ethylam ine decom position.

E th y len ed iam in e decom position on platinum is a dehydrogenation process, form ing H

2

and C

2N2

, with C -C bon d cleavage to form H C N providing a com petetiv e process. C

2N2

desorbs at low tem p eratu res, the result o f dehydrogenation. N o high tem perature C

2N2

desorption event o ccu rs. X PS results indicate th at there are no CNads. species on the surface.

A dsorption and d eco m p o sitio n o f ethylenediam ine on P t(331) occurs at step sites, producing

ACKNOWLEDGEMENTS

I w ould first like to thank my supervisor Dr. M .E. B ridge for his guidance over the years. I w ould also like to thank Prof. D.R. Lloyd for taking an interest in the work and for his useful suggestions and help. I am particularly grateful to Dr. T o m M c C a b e , w hose help, guidance and friendship over the years has been invaluable.

I w ould like to th an k all o f the technical staff for their help.

T h a n k you to M a rio for his help on the m ethylam ine w ork and Jen n y for her help on dim ethylam ine. T h a n k s to Dorothy for providing the trim ethylam ine.

T o all m y friends, both in college and outside, for their c o n tin u e d friendship and support. T o Lionel, Janice, M arika, Keelin, Fiona, Conor, Elise, Evelyn, D am ien, Richard,

U na, G uilla u m e , Cecile, Gillian, Phil, B rendan, Andreas, Eileen, C hristophe , M o, Jose, Dan and D enise. A special thanks to A nne for helping m e print m y thesis.

I w ould especially like to thank D em etri, w ho has been a great friend and colleague.

T o m y friends outside college, Stephen, Pat, Neil, D ave, T a d h g , Hilda, Bren, Ita, R onan, John and A ideen, Peter and Linda, Fin and M arge, D e rm o t and T oni, Padraic, Finbar M urp h y and T im Golden.

T h a n k s to Lily O ’Brien for treating m e like a m em b e r o f the family.

I w ould also like to thank Prof. J. Kelly and the Kriebel fu n d for the financial assistance they p ro v id e d during the last year.

ABBREVIATIONS

AES A uger Electron Spectroscopy

A R U PS A ngle-resolved U ltraviolet Photoelectron Spectroscopy

DMA D im ethylam ine

EDA Ethylenedi amine

E E L S Electron Energy Loss Spectroscopy

ESC A Electron Spectroscopy for Chem ical Analysis

E E C Fast Entry C ham ber

H R E E L S High R esolution Electron Energy Loss Spectroscopy

L L angm uir

L E E D Low E nergy Electron Diffraction

N EX A ES N ear - Edge X-ray Absorption Fine Structure

R A IR S R eflection-absorption Infra-red Spectroscopy

SA C Sam ple Analysis Cham ber

SPC Sam ple Preparation Cham ber

TD S Therm al D esorption Spectroscopy

TM A Trim ethylam ine

T P D Tem perature Program m ed Desorption

U H V U ltra High Vacuum

UPS U ltra-violet Photoelectron Spectroscopy

VG V acuum Generators

V SW V acuum Science W orkshop

CONTENTS

D eclaration... j

Sum m ary... ii

A cknow ledgem ents... iii

A bbreviations... iv

C ontents... v

CHAPTER 1 INTR O D U CTIO N 1 1.1 GENERAL INTRODUCTION... 2

1.2 LOW ENERGY ELECTRON DIFFRACTION... 5

1.3 AUGER ELECTRON SPECTROSCOPY...10

1.4 THERMAL DESORPTION SPECTROSCOPY... 12

1.5 X-RAY PHOTOELECTRON SPECTROSCOPY... 15

CHA PTER 2 EXPERIM ENTAL TECHNIOUES 18 2.1 THE EXPERIMENTAL SYSTEM S...19

2.1.1 The LEED/Auger System... 19

2.1.2 The ESCA System...22

2.2 EXPERIMENTAL TECHNIQUES... 24

2.2.1 Low Energy Electron Diffraction...24

2.2.2 Auger Electron Spectroscopy... 26

2.2.3 Thermal Desorption Spectroscopy... 27

2.2.4 X-ray Photoelectron Spectroscopy... 29

2.3 CRYSTAL PREPARATION...31

2.3.1 The Platinum Crystals...31

2.3.2 Crystal M ounting...31

2.3.3 Crystal Cleaning and Characterization in Vacuo... 34

2.4 THE ADSORB ATES... 38

41 42 42 49 55 ,55 61 63 64 64 64 65 68 69 69 75 81 87 93 93 93 96 96 97 97 98 100 100 101

METHYLAMINE ON P t ( l l l ) AND Pt(33D

RESULTS... 3.1.1 Methylamine on P t ( l l l ) 3.1.2 Methylamine on Pt(331) DISCUSSION... 3.2.1 HCN desorption... 3.2.2 C2N2 desorption... 3.2.3 H2 desorption... 3.2.4 28 amu desorption... 3.2.5 Minor Products... 3.2.6 XPS Experiments... CONCLUSIONS...

DIMETHYLAMINE AND TRIMETHYLAMINE

ON P t f l l l ) AND Pt(33I)

RESULTS... 4.1.1 Dimethylamine on P t(l 11)... 4.1.2 Dimethylamine on Pt(331)... 4.1.3 Trimethylamine on P t(l 11)... 4.1.4 Trimethylamine on Pt(331)... DISCUSSION... 4.2.1 Trimethylamine on platinum...

CHAPTER 5

ETHYLENEDIAM INE ON P td 11) AND Pt(331)

103

5.1 R ESU LTS...104

5.1.1 Ethylenediamine on P t(l 11)... 104

5.1.2 Ethylenediamine on Pt(331)... I l l 5.2 DISCU SSIO N ...118

5.2.1 C2N2 desorption...118

5.2.2 H2 desorption...119

5.2.3 HCN desorption... 120

5.2.4 XPS experim ents...121

5.3 CONCLUSIONS... 123

CHAPTER 6 CONCLUSIONS AND FURTHER WORK

124 6.1 M E T H Y L A M IN E O N P t(lll) A N D Pt(331)... 125

6.2 DIM ETHYLAM INE AND TRIMETHYLAMINE ON P t(l 11) AND Pt(331) 125 6.3 ETHYLENEDIAM INE ON P t(l 11) AND P t(3 3 1)...126

REFERENCES

127

CHAPTER ONE

CHAPTER 1 INTRODUCTION

1.1 GENERAL INTRODUCTION

In order to develop a clear understanding of the fundam ental processes involved in catalysis and corrosion it is necessary to study the chem ical com position and atomic arrangem ent of solid surfaces and to investigate the m echanism s of adsorption and reaction on these surfaces. In the case o f catalysis, for exam ple, more active and selective catalysts may be designed once a detailed knowledge of the interaction o f the reacting species with the surface of the metal catalyst can be ascertained.

The developm ent o f ultra-high vacuum technology made possible the preparation and m aintenance of clean surfaces on a tim e scale sufficient to carry out surface characterisation and surface-adsorbate interaction experim ents. The adaptation of A uger electron

spectroscopy [1-4] and X-ray photoelectron spectroscopy [5] to surface studies in the late 1960’s provided a reliable means o f determ ining the surface com position of clean and adsorbate-covered surfaces. O ver the intervening years num erous techniques have been developed or adapted to probe the geom etry and electronic structure o f surface systems, including low energy electron diffraction [6-9], ultraviolet photoelectron spectroscopy [5, 10, 11], high resolution electron energy loss spectroscopy [12], reflection absorption infra­ red spectroscopy [13-15], m olecular beam studies [16, 17], secondary ion mass spectrom etry [18], photoelectron diffraction [19] and normal incidence X -ray standing waves [20]. Therm al desoiption spectroscopy [21-23], or temperature program m ed desorption as it is also known, provides inform ation on the num ber and nature of adsorbed species as well as on the thermal stability of the species and the kinetics of their evolution.

Surface chem istry experim ents are generally perform ed on single crystal surfaces, usually of low M iller indices, in order to m inim ise surface heterogeneity and allow the effect

of a lim ited num ber of adsorbate binding sites to be determined.

T he surface chem istry of platinum has been the subject o f extensive research, due to its catalytic activity [24]. It is one of the most versatile heterogeneous catalysts, used for hydrocarbon conversion reactions and for hydrogenation in the chem ical and petroleum

refining industries [2 5 , 26 ], as w ell as in the o x id a tio n o f a m m o n ia during fertiliser production, and C O and unbum ed hydrocarbons in the con trol o f car e m is s io n s [27 , 2 8], P t(l 11) is o n e o f the m ost studied crystal surfaces. It is an a to m ic a lly flat, face-cen tred cubic surface in w h ich each atom is surrounded by six near n eigh b o u rs. Its lack o f steps and kinks gives it m in im u m h eterog en eity and is therefore ideal as a starting p oin t in surface ch em istry studies. F ew in v estig a tio n s have been m ade o f the (3 3 1 ) orientated su rface [2 9 -3 1 ]. T he (33 1) surface co n sists o f three atom w id e (1 1 1 ) terraces w ith o n e-a to m high (1 1 1 ) steps. A s such the (3 3 1 ) p rov id es the opportunity to exa m in e the e ffe c t o f in trod u cin g a system atic ‘d efe ct’ to the (1 1 1 ) plane. It is w ell esta b lish ed that step s, kinks and d efe cts play an important part in surface reactions [3 2 -3 7 ]. T h e change in the electro n ic d en sity at step sites results in d ifferen t adsorption sites, different lateral interactions b etw een adsorbed sp e c ie s and different reaction p rob abilities. A d sorb ed atom s and m o le c u le s g en era lly have higher heats o f adsorption at d efect sites [38].

In this w ork the thermal d eco m p o sitio n o f sim p le organic a m in es - m eth y lam in e, dim eth ylam in e, trim ethylam ine and eth ylen ed iam in e - on P t ( l l l ) and P t( 3 3 I ) is studied. O rganic a m in es are know n for their p o iso n in g e ffe c t in ca ta lysis and are u sed to control the selec tiv ity in hydrogenation reactions [39],

T he adsorption and d eco m p o sitio n o f eth ylen ed iam in e on P t ( l l l ) have been in v estig a ted u sin g thermal d esorp tion sp ectro scop y . X -ray p h otoelectron sp ectro sco p y and high resolu tion electron en ergy lo ss sp ectro sco p y [4 0 -4 2 ]. K in g sle y e t al. p rop osed that eth y len ed ia m in e undergoes deh ydrogen ation o f the am in e grou p s to form an eth ylen ed in itren e surface sp e c ie s at or a b ove the adsorption tem perature o f 2 8 5 K . T h is is fo llo w e d by seq u en tial rem oval o f hydrogens from the m ethyl grou p s at h igh er tem peratures to produce a m ixtu re o f partially hydrogenated sp ecies on the su rface, resultin g in the

desorption o f cya n og en and hydrogen at 4 3 0 K fo llo w e d by h ydrogen cya n id e. A n alternative m odel w as p rop osed by L loyd and H em m in g er [41] fo llo w in g H R E E L S in v estig a tio n s and L indquist et al. [4 2] from X P S stu d ies. T h ey su gg ested that a d i-im in e rather than a dinitrene

sp ecies is form ed b etw een 2 7 3 K and 37 3K .

dehydrogenation an d that the N -C b ond is not broken. D is c u s s io n n o w centres on the nature o f the interm ediate or interm ediates form ed during the d eco m p o sitio n p rocess.

Few in v e stig a tio n s have been carried out on d i- and trim eth ylam ine adsorption on sin gle crystal su rfaces. Pearlstine and Friend [47] ex a m in ed trim eth ylam ine adsorption on clean, carbide and o x id e cov ered W(IOO). T he m ain d esorp tion products from th ese surfaces w ere trim ethylam ine, m ethane, hydrogen and nitrogen. N u n n ey et al. [48] found that trim ethylam ine adsorbs m olecu larly through the nitrogen lo n e pair on N i ( l l l ) at llO K .

E rley et al. [49] fou n d that it d eco m p o se s co m p letely into atom ic sp e c ie s on P t(l 11).

W ork carried out by W alker and Stair [50, 51] on M o(lO O ) resulted in the form ation o f H C N and CO (on o x id ise d M o(lO O ) surfaces on ly) as w e ll as m ethane, hydrogen and n itrogen. N o m o lecu lar trim ethylam ine w as d etected as a d esorp tion product from m olyb d en u m . T h ey co n clu d ed from their research that trim eth ylam ine u n d ergoes co m p lete d isso c ia tio n into atom ic surface sp e c ie s upon adsorption at lo w c o v e ra g es at 3 3 3 K . T h ese products ‘partially p a ssiv a te’ the surface a llo w in g m olecu lar adsorption o f trim ethylam ine at h igh er co v erag es w h ich d eco m p o ses to form H C N and C H4 d esorp tion products.

T he tech n iq u es u sed in this work w ere A u ger electron sp ectro scop y, lo w energy electro n diffraction, X -ray p hotoelectron sp ectro scop y and thermal d esorp tion sp ectroscop y. T he rem ain d er o f this chapter is d ev o ted to a b rief d escrip tion o f the theory beh in d each o f

th ese techniques. T heir application is describ ed in chapter 2.

1.2 LOW ENERGY ELECTRON DIFFRACTION

In the LEED experim ent, a beam o f energetically w ell-defined electrons is directed onto the crystal surface, and the angular distribution of elastically back-scattered electrons is m onitored to yield inform ation on the periodicity of the surface.

The theory of low energy electron diffraction has been discussed and reviewed in detail [6-8, 52], Only a brief description of LEED and the inform ation attainable from it will be given here.

An incident beam of electrons o f energy between 10 and 200eV produces electron w avelengths o f

3.9

to

0.93A,

given by:

arises at these electron energies because the corresponding w avelengths are sim ilar to the interatom ic distances in the crystal lattice. The electrons have a high elastic back-scattering cross-section resulting in a high probability of being back-scattered before penetrating any great distance into the bulk.

A

= V

(150.4/E) (1.2.1)

where

X,

the wavelength, is in A and E, the energy, is in eV. The possibility o f diffraction

<t> - I

Square | a,. | = | a,.

R ectangular C en tred R ectan g u lar I a , , k I I <t> = 9 0 " <t) = 9 0 ‘’

H exagonal (t)= 120°

J O blique

The period ic tw o-dim ensional arrangements o f ato m s at the surface can be grouped

into five types o f surface nets, whichi d escrib e all p o s s ib le d ip eriod ic surface structures. The

unit areas or unit m esh es o f these five; p o ssib le nets are sh o w n in figu re 1.1.

T*ie period icity o f the surface i s describ ed by the tw o vectors ai and a2. In reciprocal

sp a ce, the corresp on din g vectors are a i* and aj* , d efin ed by:

ai*.aj = 27r5ij (1.2.2)

where 6ij is the K ronecker delta (8ij = 1 for i = j; Sy = 0 for i j ). T h e 2n in the equation replaces the value o f unity con ven tio n a lly used in X -ray d iffraction, a llo w in g the w ave

vectors k to be drawn on the reciprocal crystal lattice diagram . T h e en ergy E o f the incident

ek ctro n is giv en by:

E = (h^/2m).k^ (1 .2 .3 )

w h ere h = h/27i (h is P la n ck ’s constant) and m is the m ass o f the electro n . T he w av e vector is

related to the w avelength o f the electron by:

k = 2ti/A. (1-2.4)

D iffraction occurs w hen the con servation o f energy and m o m en tu m selection rules are

satisfied . T h ese rules are giv en by:

l k ' | = l k o l (1-2.5)

k'li = k o | | + g , x - ( 1 . 2 . 6 )

w here k' and ko are the w ave vectors o f the diffracted and prim ary electron s resp ectively.

S in ce there is a lo ss o f period icity in o n e d im en sion , on ly the co m p o n e n t o f the m om entum

parallel to the surface is con serv ed , h en ce the n subscript. g,,k is a gen eral reciprocal lattice

vector, w h ich lies in the plane o f the surface, and is given by

g^t = h .a\* + k.a2* (1-2-7)

where h and k are integers.

It is p o ssib le to derive the sym m etry and d im en sio n s o f the clea n surface unit m esh and

also the sym m etry and d im en sion s o f ordered adlayers relative to the clean surface from a

know ledge of the direction of the d iffracted beam s a n d the p rim a ry beam energy. T he relationship b etw een the surface unit m esh and that o f a m o n o la y e r o f som e ad sorbed species can be described by usin g a transform ation m atrix M:

M =

(1.2 .8)

such that:

ais - miiaib + m^aab

(1.2.9)

a2s - rn2iaib + n\22SHb (1.2.10)

or, in m atrix notation:

3s = Mab (1.2.11)

w here as represents the translation vectors o f the adsorbed layer u n it m esh, and ab represents those o f the substrate unit m esh. The areas o f the tw o unit m esh es are given by:

T he type o f superposition o f adlayer upon substrate can be d e fin e d by reference to the values o f detM :

(a) if detM is an integer, the structure is referred to as a sim p le lattice;

(b) if d etM is a rational fraction, the structure is a coincidence lattice;

(c) if d etM is an irrational fraction, the structure is incoherent.

(a) and (b) are illustrated in figure 1.2.

A — I a j s X a2s ( 1.2 . 12)

B = I aib X a 2 b (1.2.13)

It can be show n that:

(a) SIMPLE det Ml integer

a 2 X 1 overlayer

(b) COINCIDENCE det M a rational fraction

M =

3/2 0 0 1/2

2 a , s - S Bi j,

F ig. 1.2 R elationship betw een surface and bulk m eshes. T he sim p le and co in cid en ce m eshes ure illustrated by the ca ses o f d ep osit atom s (filled circles) on the bulk exp osed (1 1 0 ) plane o f an f.c.c. material (open circles).

A sim ple notation has been proposed by W ood [53] w here the relationship betw een the

unit m eshes is d escribed by the ratio o f the lengths o f the translation vectors and by a

rotation R expressed in degrees. T his notation is illustrated in figure 1.3. U nfortunately,

W o o d ’s notation is inadequate in describing certain surface structures, fo r exam ple

incoherent structures. H ow ever, its sim plicity and suitability fo r very m any structures m eans

[image:19.502.47.492.16.529.2]

Deposit

Substrate Substrate Centred Surfact'substrate surface mesh mesh ritios normal

(a)

a,. M =

®ooo®o

o®ooo®

oomo0^

o .Q 'o 'm

©<D‘’

®’

O,0'O®O

®'®ooo®

e.g. Pt(lOO) (V2x2V2) R 45° - O

(b)

Fig. 1.3 Tw o notional exam ples o f W o o d ’s notation for structures com pared with the matrix notation. In exam ple (a) for a N i(l 10) face exposed to oxygen, the notation can be shortened to N i(l 10 )c2 -0 because the deposit m esh is rectangular.

In this work, LEED has been used for the determination o f surface and overlayer

[image:20.506.50.490.24.508.2]

1 .3 A U G E R E L E C T R O N S P E C T R O S C O P Y

A u g e r e le c tro n sp e c tro sc o p y in v o lv e s an a ly sis o f th e e n e rg y o f th e se c o n d a ry e lec tro n s

e m itte d b y th e A u g e r p ro c e ss [1] fro m a su rfa ce.

In a d ia tio n o f th e su rfa ce w ith a 2 -3 k e V e le c tro n b e a m o r X -ra y p h o to n s in a sim ila r

e n e rg y ra n g e re su lts in the io n iz a tio n o f an a to m ic c o re le v e l (E i). T h is th en re la x e s by an

e le c tro n ic tra n sitio n su ch as E2 to E i. T h e e n e rg y re le a s e d b y th is d e e x c ita tio n p ro c e ss is

tra n s f e rre d by a ra d ia tio n le s s m e c h a n is m to a th ird e le c tro n in E3 w h ic h is e m itte d fro m the

a to m (see fig u re 1.4). T h is e m itte d e le c tro n is th e A u g e r e le c tro n a n d h a s a c h a ra c te ristic

k in e tic en e rg y Ek g iv e n by;

Ek = E , - E / - E3' (1 .3 .1 )

w h e r e E j ' (~ E2) a n d E3' (~ E3) ta k e a c c o u n t o f th e sh ifts in e n e rg y o f th e s e lev els d u e to

io n iz a tio n o f th e ato m . T h e re fo re , th e k in e tic e n e rg y im p a rte d to th e e le c tro n is in d e p e n d e n t

o f th e p rim a ry ra d ia tio n . T h e io n iz a tio n c ro ss-se c tio n o f th e c o re le v e l, h o w e v e r, is

d e p e n d e n t on th e en e rg y o f the p rim a ry ra d ia tio n (Ep), ris in g fro m 0 fo r Ep< E i to a

m a x im u m fo r Ep/Ei ~ 3, fo llo w e d b y a slo w d e c re a se at h ig h e r v a lu e s , b u t w ith little

v a ria tio n in the ra n g e 2 < Ep/Ej < 6 [54],

A lth o u g h th e p rim a ry e le c tro n c a n p e n e tra te d ee p in to th e b u lk o f th e sp e c im e n and

o p ro d u c e se c o n d a ry e le c tro n s w ith in th e b u lk (d u e to an in e la s tic m e a n fre e p ath o f ~ 2 0 0 A ,

fo r a ty p ical b e a m en e rg y o f 2 - 5 k e V ), th e sh o rt e sc a p e d e p th (4 -

lOA)

fo r se c o n d a ry

e le c tro n s m e a n s th a t o n ly th o se e le c tro n s at o r n e a r th e s u rfa c e e s c a p e w ith o u t s c a tte rin g an d

a re d e te c te d . F u rth e rm o re , by d e c re a s in g th e an g le o f in c id e n c e o f th e p rim a ry e le c tro n b ea m

to n e a r g ra z in g in c id e n c e , a g re a te r n u m b e r o f s u rfa c e a to m s are s a m p le d , th u s in c re a sin g the

su rfa c e se n sitiv ity . E x p e rim e n ta lly , th e o p tim u m a n g le lies b e tw e e n 10 a n d 15° fro m g ra z in g

in c id e n c e .

E le c tro n ic d iffe re n tia tio n o f th e sig n a l fro m th e re ta rd in g fie ld a n a ly s e r d u rin g A E S

a n a ly sis y ield s the se c o n d a ry e le c tro n d is trib u tio n (N (E )) c u rv e . A u g e r e le c tro n s are

o b se rv e d as w e ak p e rtu rb a tio n s o n th is to tal s e c o n d a ry e le c tro n d is trib u tio n ; h o w e v e r.

fu rth e r d iffe n e n tia tio n o f the N (E) c-urve to g iv e d N ( E ) /d E re s u lts in g re a te r s e n s itiv ity an d

re so lu tio n , su c h th a t th e lo w e r limit o f th e d etec tio n is in th e ra n g e o f 1 - 5

%

o f a m o n o la y e r.

It is im p o rta n t to n o te that d e -e x c ita tio n o f th e io n iz e d a to m c a n a lso re su lt in X -ra y

e m is s io n , a c o m p e titiv e re la x a tio n p ro c e s s to A u g e r e m is s io n (a g a in , see fig u re 1.4).

F ollow in g th e io n iz a tio n o f a K -sh e ll, fo r ato m ic n u m b e rs (Z ) < 19 m o re th a n 9 0 % o f th e

re la x a tio n is b y A u g e r e m issio n ; at Z = 3 2 (G e ) th e re is an e q u a l p ro b a b ility o f re la x a tio n v ia

these tw o c o m p e tin g p ro c e sse s; a n d fo r Z > 33 X -ra y e m is s io n b e c o m e s th e d o m in a n t

p ro c ess [55]. T h u s , th e A u g e r yield is m a x im is e d fo r th e firs t a n d s e c o n d ro w e le m e n ts ,

m a k in g A E S m o s t u sefu l fo r the stu d y o f th e se e le m e n ts a d s o r b e d o n su rfa c e s.

A u g e r e le c tro n s p e c tro sc o p y is u se d p rim a rily as a m e a n s o f m o n ito rin g th e c h e m ic a l

c o m p o s itio n at th e su rfa ce. H o w e v e r, it m a y also be u se d to m e a s u re re la tiv e c o v e ra g e s,

d iffu s io n c o e ffic ie n ts [56], ra te s o f a d so rp tio n [57] an d d e s o rp tio n [5 8 ], a n d d e p th p ro file s

[5 9 ], as w ell as a m o n ito r in N E X A F S e x p e rim e n ts [60].

Ep o r X -ray

V

v////'//nirnmmniin\

V acuum

K e v :

• E le c tro n

O H o le

V V a le n c e L ev el

E 1 . 3 C o re E n e rg y L e v e ls Ep E le c tro n B e a m

D e-E xcitation

X -ray

V acuum

V7-,y7//77mi/l!lini!l!U\

A uger

1.4 THERMAL DESORPTION S PECTROSCOPY

T his technique involves heating the sam ple at a c o n tro lle d h eating rate and m onitoring

the partial pressure o f a species desorbing fro m the surface as a fu n ctio n o f tem perature.

A nalysis o f desorption spectra yields inform ation on the n atu re and nu m b er of

adsorbed species, the num ber of various desorbing phases fo r each species and th eir relative

pop u latio n s; the activation energies o f desorption (Ed) o f the in d iv id u al phases and the order

o f the desorption process.

T herm al desorption has been the su b ject of a n u m b er o f a rticles [21-23] and only a

b n e f d isc u ssio n o f the technique will be presen ted here.

'W hen the pum ping speed to volum e ratio is large, the partial p ressu re at any tim e is

p ro p o itio n a l to the rate o f desorption. The rate of desorption from u nit surface area is given

w h ere 9 is the surface coverage, n is the o rd er of the d e so rp tio n reactio n , v„ is the rate

c o n sta n t, R the gas constant and T the tem perature. W ith lin ear h e a tin g rates, the integral of

the partial pressure change over the tem perature range is p ro p o rtio n a l to the coverage.

R e d h e a d show ed [21] that, in the absence o f any coverage d e p e n d e n c e on Ed, the peak

tem p e ra tu re (Tp) fo r a first o rder desorption process is in d ep en d en t o f coverage:

w h ere (3 is the heating rate and vi is the pre-exponential (frequency) factor. T h u s Ed can be

d e te rm in e d directly from a m easurem ent o f Tp, provided a value o f vj is assum ed, using the

R ed h ead form ula: by:

-de/dt = V n.0".exp(-Ed/R T) (1.4.1)

E d/R T p' = (v,/|3)exp(-E d/R T p) (1.4.2)

Ed/RTp = /«(viT p/p) - 3.64 (1.4.3)

T he relation betw een Ed and Tp is very nearly linear and for lO '^ > vi/(3 > 10*, the error

in duced is sm all (approx. 2% ).

/n(Tp^/p) = Ed/RTp + Z/i(Ed/R) (1.4.4)

A plot of /n(Tp^/p) versus 1/Tp yields a straight line with a slope o f Ej/R . N o value o f vi need be assumed using this equation, but can be calculated subsequently by substituting E<j in the rate equation (1.4.1).

A more general expression, valid for all desorption orders, n, greater than zero has been derived by Falconer and Madix [23]:

/n(P/Tp^) = In (R,v„.0"-'/Ed) -Ej/RTp (1.4.5)

Plotting //2(P/Tp^) versus 1/Tp for a number o f heating rates yields a straight line o f slope

-Ed/R.

In general, Tp increases as the heating rate is increased, but since the dependence o f Tp

on (3 iis small, changes in (3 o f several orders o f magnitude are necessary to obtain reasonable accuracy in the determination of Ej from heating rate variation m ethods.

The desorption activation energy can be determined for any n w ithout recourse to h^atin'g rate variation methods, by direct analysis o f the desorption profile. From (1.4.1) it Ciin be' shown that:

/7!(-d0/dt /e") = -Ed/RT (1.4.6)

B y using the partial pressure change as a measure o f the rate and the area under the remaining higher temperature portion o f the curve as a m easure o f the coverage, a plot o f

//z(-d0/Mt /0") versus 1/T should yield a straight line o f slope -Ed/R for the appropriate order.

When analysing thermal desorption spectra, consideration m ust be given to the possibility o f lateral interactions between adsorbed species, changes o f bond strength with coverage and the interconversion o f states in the adlayer. T h ese can com plicate spectral interpretation. In the case o f lateral interactions, the activation energy and peak temperature becom e dependent on coverage. Repulsive interactions can result in a first order desorption

process displaying a second order peak shape and Tp versus 0 behaviour [61]; and a second

2 amu

SOL

0.5L

u

c«

H

HH

H

Z.

NH

l.OL

Pt(557)

P t(lll)

soo

200

300

400

600

700

TEMPERATURE (K)

F ig . 1.5 T h e rm a l d e so rp tio n s p e c tra o f H2 fo llo w in g e x p o su re s o f H2 to v a rio u s p la tin u m su rfaces.

[image:25.504.38.497.11.547.2]

1.5 X-RAY PHOTOELECTRON SPECTROSCOPY

X-ray photoelectron spectroscopy (XPS) provides inform ation about the elemental

ccomposition and chem ical environm ent of a surface. The process involves the excitation of

eelectrons in an atom or molecule into the vacuum by m eans of X-rays.

A photon with sufficient energy

hv

(where

h

is Planck’s constant and v is the

ffrequency of the electrom agnetic radiation) will ionize an electronic shell, im parting to an

eelectron enough energy for it to escape into the vacuum. The kinetic energy of this

Fph)toielectron is given by:

Ek = / z v - E b (1.5.1)

vwhere Eb is the calculated binding energy of the electron. The kinetic energy may be

rmodified by several atomic parameters associated with the electron em ission process, such

ais relaxation of orbitals and core level screening. Therefore, equation (1.5.1) represents a

hiighly sim plified relationship between E^ and Eb.

Incident photons with energies in the ultra-violet and X -ray regions of the

e;lectromagnetic spectrum provide sufficient energies to produce photoelectrons. In ultra­

v io le t photoelectron spectroscopy (UPS) the energy of the incident radiation (21.2 eV using

a helium discharge lamp) is insufficient to eject deep core electrons. UPS is therefore

g'enerally used to study electrons in the valence band of the solid. X-rays on the other hand

hiave energies capable of ionizing core electron levels. Since core level energies are elem ent

sjpecific XPS provides a powerful tool for identifying surface species.

In XPS the sources of radiation are usually alum inium o r m agnesium anodes which

giive Ka lines at 1486.6 eV and 1253.6 eV respectively. C om bining these two sources in a

swvitchable twin-anode can enable the separation of those photoelectrons w hose kinetic

energies are directly dependent on the incident photon energy from A uger electrons, which

ar e also produced during core level ionization, and whose energies are fixed, regardless of

thie energy of the incident beam (see section 1.3).

feature is the step Hke background, th e steps occurring at the lo w k in etic energy (high binding energy) sid e o f the p h otoelectron lines. T h ese steps are gen erated b y ca scad es o f inelastic c o llis io n s undergone by th o se p h otoelectron s generated at dep th s b e lo w the surface m uch greater tlhan the m ean-free-path fo r in ela stic scattering.

As stated earlier, equation (1 .5 .1 ) represents a sim p listic picture in w h ich the electron s rem aining after ion ization are assu m ed to be in the sam e initial state as b efore the ionization event. H o w ever, as an electron is ejected from a core le v e l, the other electro n s relax in energy to lo w er en ergy states to screen this h ole and so m ake m ore en ergy a vailab le to the ou tg o in g p hotoelectron . T his is ca lled a relaxation shift. Furtherm ore, an ou tg oin g photoelectron m ay interact with a v a len ce electron , either prom otin g it to a h igh er energy le v e l (a process referred to as a shake-u p), or ev en prom otin g it a b o v e the vacuum level (k n o w n as a sh a k e-o ff). In d o in g so the o u tg o in g p hotoelectron lo s e s k in etic energy. T his p r o c e ss results in satellite peaks at the high bin d ing en ergy sid e o f the p h otoelectron peak. A n oth er potential source o f spectral features are p la sm on s, w h ich are c o lle c tiv e d en sity flu ctu ation s in the electron clo u d in a solid.

Finally, the X -ray source its e lf contributes features to the X P S spectrum . B oth A l-K « and M g -K a have K« 3 ,4 satellites so m e 10 e V b elo w the m ain Ka 1,2 lin e s, w ith around 10% o f the;ir intensity.

O verall, th ese features esse n tia lly reflect the d ifferen ce in en ergy b etw een the initial (u n ex cited ) state and the final state.

In addition to id en tifyin g surface sp ecies X P S a llo w s inform ation about the ch em ical environm ent o f an elem en t to be d educed. T his is b eca u se the b in d in g en ergy o f a core electron is in flu en ced by electrostatic interaction w ith the v a le n c e electron s. T he contribution o f the valen ce electro n s to the en ergy o f the core le v e ls d ep en d s on the type o f bonding in w h ich the valen ce electron s are in v o lv ed . In sim p le term s, the greater the electron egativity o f the surrounding atom s that are in v o lv e d in ch em ic a l b on d in g, the m ore the electronic ch arge is redistributed aw ay from the atom and th e h igh er th e ob served binding en ergies o f the core electron s. T h e se ‘ch em ical s h ifts’ m ay be as large as 10 e V and provide inform ation on the v a len ce states o f the surface elem en ts and co n seq u en tly their bonding.

In XPS th e energy resolutiom is lim ited by the line w idth o f the p h oton source.

Therefore, if an elem ent exists in t\wo different chem ical e n v iro n m en ts on the surface, the

difference b etw een their respective c h e m ica l shift values m u st e x c ee d the line w idth o f the

source to be fully resolvable in th e spectrum . The line w idth o f the A l-K a and M g-K a

sources are about 0.85 eV and 0.7 eV respectively.

S om etim es, the peaks in X -ray spectra are labelled by c o n v e n tio n al X -ray notation

w here the principal quantum num bers (n = 1, 2, 3, ...e tc .) o f the o rb ita ls are assigned the

letters K , L, M , N, O, ...e tc ., with the orbital and total angular m o m e n tu m q u an tu m num bers

1 and j 0)f the hole left after photoem ission assigned respective su ffix es. H o w ever, it is not

tncom m ion fo r atom ic notation - is, 2pi/2, 2p3/2, etc. - to be used. It is this notation that is

CHAPTER TWO

EXPERIMENTAL TECHNIQUES

CHAPTER 2 E XPERIMENTAL TECHNIQUES

2.1 THE EXPERIMENTAL SYSTEMS

Z. 1.1 The LEED/A uger System

This is a bakeable, 12 inch diameter, stainless steel cham ber constructed by Vacuum

G enerators Ltd.. A num ber of ports were later added to the cham ber by Vacuum Science

Wo)rkshop Ltd.. These ports were positioned so as to give line of sight to the centre of

curvature of the LEED optics.

The ports and flanges on the chamber are sealed using high purity copper gaskets; the

mai n cham ber is sealed using gold wire as a gasket.

A section through the experimental level of the cham ber is shown in figure 2.1.

K ey:

A G Argon Ion Gun

E G l A uger E lectron Gun

E G 2 L E E D /A u ger Gun

G D G as D oser

E G 2

R F A

S Sam ple

V V iew port

Q M S Q uadrupole M ass Spectrom eter

R F A Retarding Field

A nalyser L V Leak V alve

A (

E G l

L V V

Fig. 2.1 Schem atic diagram through the experim ental level o f the L E E D -A u ger system .

The chamiber w as fitted withi a quadrupole m a s s sp ectro m eter (V G Supavac) for

residual gas analysis and therm al des.orption spectroscopy; a display type three grid retarding

field analyser (R F A ) (V acuum Gemerators Ltd.) w hich c o u ld be used fo r L E E D or AES

experim ents; a n o rm a l incidence electron gun (L E G 2) fo r L E E D and a g lancing incidence

electron gun (V acuum Science W orkshop L td.) for A E S; a n d tw o ionization gauges (Bayard-

A lp ert type), one fo r pressure m easurem ent in the torr pressure range, located

below the experim ental level o f the cham ber, the o th er fo r argon ion b o m b ard m en t (see

sectio n 2.3.3), located in the line of sight o f the sam ple.

T h e m ain ch a m b e r attained pressures o f below 2 x l 0 ‘^° torr. T his w as ach iev ed by use

of a 5'0 litre per second diode ion pum p in conjunction w ith a titan iu m su b lim atio n pum p.

The diiode ion pum p w orks by ionizing the gas in the system by m eans o f an electrical

discharge w hich, u n d e r the influence o f a m agnetic field o f a few th o u san d G auss, follow s a

flat helical path, thus m axim ising the path length and resu ltin g in a high efficiency o f ion

form ation. The gaseous ions are then c ap tu red or c h e m iso rb e d at the electrodes. T he

titanium sublim ation pum p is located below the m ain ch a m b e r and is separated from the

ch am b er by a steel plate. E lectrical resistan ce heating o f titan iu m /m o ly b d e n u m alloy

filam ents results in the sublim ation of titanium . T his co n d enses on the ch a m b e r w alls and

reacts chem ically w ith the active gases to fo rm low vapour p ressu re solid com pounds.

An auxiliary pum ping system was also em ployed to p u m p the ch a m b e r to less than

10'^ torr, at w hich pressure the ion pum p co u ld be started. T h is system consists o f a liquid

nitrogen trapped oil diffusion pum p and a rotary pum p w ith a liq u id nitrogen co o le d foreline

trap. T he rotary p u m p is an oil-sealed m echanical pum p w h ich traps and com p resses a

sam ple o f gas to slightly above atm ospheric pressure w ith each ro tatio n o f the vane. T he gas

is then expelled via an exhaust valve. O nce the rotary pum p has re a c h e d a pressure o f about

10'^ torr, the oil diffusion pum p can be started. A je t o f oil v a p o u r fro m a heated well of

liquid is directed dow nw ard and radially across a space w ithin th e pum p. G as p articles are

carried dow nw ards with the je t and c o m p re sse d until they can be rem o v ed by the rotary

pum p. M eanw hile, the oil vapour is co o le d back to liquid as it d e sce n d s, to be reh e a te d again

in the well at the base o f the pum p. T h is au x iliary system w as also used to p u m p the gas

handling lines and it can be iso lated fro m the m ain c h a m b e r by m e a n s o f an iso lation valve.

The system contains two gas handling lines both o f w hich can be pumped to less than

10"'’ torr by the auxiliary pumping system. One line is constructed m ainly o f stainless steel.

Gas dosing from this line is achieved by fillin g the chamber to the required pressure. The

other line is constructed o f glass, for handling gases which react in the steel line. Gas dosing

iis achieved by means o f a doser attached to the leak valve. A block diagram o f the entire

pum ping system is shown in figure 2.2.

Key:

TSP ODP P T A

M

o

T itanium Sublim ation Pump Oil Diffusion Pump Pirani Gauge Liquid Nitrogen Trap A ir Inlet

Leak Valve

Isolation Valve Gas Bottle Inlet

Ion Pump

Rotary Pum p ODP Cham ber

Main

TSP

2.1.2 The E S C A System

The ESCA (Electron Spectnoscopy for C hem ical Analysis) system is a three-

chambered system constructed by VSW Scientific Instrum ents Ltd.. The three cham bers are:

a sample preparation cham ber (SPC) which was used to clean and dose the sample; a sample

analysis cham ber (SAC), used to carrj out the XPS experim ents; and a fast entry cham ber

(EEC) which was not used in this work.

The SPC and EEC are each separated from the SA C by means o f a gate valve. The

SPC and SAC are both pum ped by their own liquid nitrogen trapped, w ater cooled 2001/s oil

diffusion pumps. Each pum p is backed by a tw o-stage rotary pum p (Edwards). Additionally,

the SPC is pum ped by a standard T SP whilst the SAC is pum ped by a w ater-cooled TSP.

A sketch of the experim ental level of the SAC cham ber is shown in figure 2.3.

Key:

FEC Fast Entry Chamber

I G Ion Gauge

S Sample

To SPC

S P C Sam ple Preparation Chamber

U PS U ltraviolet Photon Source

V Viewport

X PS X-ray Photon Source

UPS

XPS

Probe _{> T o F E C}

IG

V

Fig. 2.3 Sketch o f the experim ental level o f the S A C on the E S C A system (analyser not show n),

(not to scale)

Gas dosing (i.e. ex p osing it to th.e adsorbate) is lachieved by fillin g the ch a m b e r from a

:gas handling m ianifold. A dm ission of the gas is via a leak valve (V G M D 6 series). T he

m an ifo ld is p u m p e d by a rotary-backed diffusion piurnp. A block diagram o f the entire

p u m p in g system is show n in figure 2.4.

T he SA C is fitted w ith an X -ray photoelectron so u rce (V SW T A IO T w in A node), an

u ltra -v io le t p h o to electro n source, an ion gauge (B ayard-A lpert type) and a hem ispherical

ele c tro n analyser (V S W H A 100) (not show n).

C ry stal cleaning and dosing w as carried out in the SPC . T h e X P S exp erim en ts w ere

carried o u t in the SA C.

V20

ins M in i f il d V23 Y 9 y V24

Gm M a n ifo ld ] V25

V21

IG lUVI

VIO V15i

G a.s S a m p l e s

x(;

Kasi K iitry

C 'h am b er - F 4 b

S am p le Analy.si.s ( 'h a m b e r V II CD— V14 V3 V7 V6, V12 ( 'o l d T r a p

T S P

- I»2

D P I

R otary P u m p V13

V8

V17 V18

K e y :

C T C o l d T r a p T S P T i t a n i u m S u b l i m a t io n P u m p

D P D if fu s io n P u m p U V L U V - l a m p

I G Ion G a u g e V V a lv e

2.2 E X P E R IM E N T A L T E C H N IQ U E S

2.2.1 Low Energy Electron Diffraction

The apparatus consists o f a display-type 3-grid L E E D optics and an axial LEG2 electron gun, shown schematically in figure 2.5.

/ / /

Electron Gun

Sample

G1 A l

A2 A3 A4

Screen Key:

G2 G3 G4 _{F Filament}

A1 - A4 Anodes G1 - G4 Grids

Fig. 2.5 Schematic o f the LEED optics.

Electrons from the directly heated filam ent emerge as a divergent beam into the anode A l as a result o f fie ld penetration from A l at lOOOV through the grid G l. The beam is focused by the electrodes A l , A2 and A3. The beam current is trim m ed down to a few microamps from an in itia l current o f approximately 0.5 m A by means o f an aperture at the end o f A l.

The incidtent beam strikes thie sample, producing elastic, inelastic and secondary electrons. The imelastic and secondairy electrons are filtered out by a retard potential on grid G3. The anode A4 and grids G2 and G4 are held at earth; A4 and G4 to ensure that the electrons of the incident beam and the back-scattered electrons travel in a field free zone; and

02

to prevent field penetration from the screen to 0 3 . This set-up is known as a retarding field analyser.

2.2.2 A u g er Electron Spectroscopy

A uger spectra are recorded by energy analysis of the secondary electrons em itted from

the crystal upon irradiation with an electron beam of energy between 3 - 5keV. The source of

the electron beam is an electron gun with focusing and deflection electrodes. A djusting the

intensity of the prim ary electron beam alters the intensity o f the secondary electrons.

In the LEED /A uger cham ber A uger electron spectroscopy was perform ed using a VG

LEG! electron gun (E G l in fig. 2.1) and the VG LEED optics as a retarding field analyser

(RFA). A field free zone was m aintained around the sam ple by holding the sample, the

outerm ost grid (G4) of the RFA and the final anode of the electron gun at earth. The

collector of the RFA was held at +250V to ensure high collection efficiency by preventing

losses due to secondary electron em ission. Grid G3 (see fig. 2.5) was used as a high pass

filter by placing a retarding potential (-E) upon it, so that only electrons with an energy

greater than E were collected. Sw eeping this retarding potential from E = 0 to E = Ep ( the

energy of the prim ary electron beam ) produced a retarding potential curve. By differentiating

this curve, a secondary electron energy distribution curve (N(E) versus E) was obtained.

Further differentiation o f the N(E) curve to give the dN (E)/dE curve provided greater

sensitivity. The spectra were recorded on com puter. A detailed discussion of the electronic

instrum ents used for data analysis is given in reference [63].

2.2.3 Thermal Desorption Spectroscopy

In the thermal desorption experim ent the partial pressure o f a particular species is m onitored as a function of crystal tem perature. A block diagram o f the experimental arrangem ent is shown in figure 2.6.

Sample

Thermocouple

Multimeter

Chart

Recorder

Quadrupole Mass

Spectrometer

Mass Spectrometer

Control Unit

F ig. 2 .6 B lo ck diagram o f the therm al desorption experim ent.

degree of exposure is m easured in Langm uirs, L, w here IL = lO'*" torr s exposure. The sam ple was then rotated to the line o f sight of the m ass spectrom eter.

The m ass spectrom eter itself was set to the required e/m value by tuning the first mass control on the control unit. The partial pressure output was fed directly into the y-axis o f an X -Y recorder, w hilst the crystal tem perature was m easured using a chrom el/alum el therm ocouple which drove the x-axis of the X-Y recorder. For calibration purposes the therm ocouple voltage was also displayed on a digital voltmeter.

By passing a current through the crystal support wires at a constant applied a.c. voltage, a linear heating rate of 13Ks‘' in the 300 - 800K region was obtained.

2.2.4 X-ray Photoelectron Spectroscopy

X -rays are p ro d u ce d by the acceleration o f th erm ally e m itte d ele c tro n s from a hot filam ent at ground potential tow ards an anode. T his an ode is ty p ically held at a potential betw een 12kV and 20kV . T he anode used w as a tw in anode co n fig u ratio n o f alum inium and m agnesium (V S W T A 10). T he a lu m in iu m anode provides a K« rad ia tio n energy o f 1486.6 eV w ith a line w idth o f 0.85 eV. T h e m agnesium generates K« rad ia tio n en erg y o f 1253.6 eV w ith a line w idth o f 0.7 eV.

T he X -ray sp e ctro m e te r and the sam ple under investigation are c o n n e cte d electrically, en su rin g that their Ferm i energy levels are at the sam e energy.

T he analyser is a concentric hem ispherical sector analyser. T h is is a dispersion analyser, and is n o rm a lly operated in a m ode in w hich a c o n sta n t p o ten tia l difference, called the pass energy (Eq), is applied across the in n er and outer h em isp h eres so that electrons with e n e rg ies clo se to Eo (w ithin a range AE) are focused on the ex it slit o f th e analyser by m eans o f electrostatic d eflection betw een these hem ispheres.

In o rder to scan through a range o f photoelectron en erg ies, the e le c tro n s are retarded b efore they e n te r the analyser. T h is is achieved by m eans o f a reta rd in g lens system . By ram p in g the retarding field potential, electro n s of different en erg ies can b e analysed.

P re-retardation o f the electro n s also has the im p ortant a d v a n ta g e o f im proving the reso lu tio n . T his is because the reso lv in g pow er, p, o f the analyser, given by:

p = Eo/AE (2.2.1)

is fix e d by the geom etry o f the system . C onsequently, a d ecrease in Eo resu lts in a decrease in AE, i.e. a decrease in the spread o f e n ergies that arrive at the e x it slit.

am plified and then counted by a ratemeter. The output is displayed and stored on computer.

A schematic o f the photoelectron detector/analyser set-up is show n in figure 2.7.

CHSA

RP

CHSA Concentric Hemispherical Sector Analyser

RP Retarding Plate

Computer DAC

Pre-Amp

High Voltage

Supply Main Amp

and Ratemeter Hac 500

and Multiplier Power

Supply

L Lens

C Channeltron

S Sample

Fig. 2.7 Schem atic o f the X -ray source and photoelectron detector/analyser set-up.

2.3 CRYSTAL PREPARATION

2.3.1 The Platinum Crystals

The Pt(331) sam ple was cut from a 99.99% pure single crystal rod at the Material

Science Centre of the U niversity of Birm ingham . This rod was m ounted on an X-ray

goniom eter, oriented to w ithin 0.25° o f the (331) direction and a disc, 9m m in diam eter and

1mm thick, was cut in a spark cutting machine. The crystal was then polished with 0.25)i

diam ond paste. The Pt(331) surface consists of three atom wide (111) oriented terraces

divided by one atom high (111) oriented steps (see figure 2.8).

The P t(l 11) sam ple was supplied by M aTech (Jiilich). It was cut on both sides to the

above specifications and polished on one side to 0.03|im .

fcc(100)

fcc(111)

fcc(331)

Fig. 2.8 R epresentations o f the (1 0 0 ), (1 1 1 ) and (3 3 1 ) surfaces o f platinum . T h e (1 0 0 ) diagram represents

a bulk truncated surface. In reality, reconstruction occurs.

2.3.2 Crystal Mounting

Each platinum crystal was spot-w elded betw een tw o tantalum w ires of diam eter 0.25

[image:42.502.25.482.9.695.2]

To the edge o f each crystal was spot-w elded a chrom el/alum el therm ocouple of

diam eter 0.05 mm. This assembly was treated in an ultrasonic acetone bath to remove any

residues o f grease that might be a source o f contam ination.

In the case o f the LEED /A uger system the arms of the copper holder were screwed to a

ceramic block to provide the connections for the heating wires and a channel for the

electrical feed-throughs to the therm ocouple wires (see figure 2.9).

A ttached to the top of this ceram ic block was a steel driving rod which in turn was

attached to the sam ple m anipulator, allow ing the crystal to be m anipulated through the x, y,

z plane. The m anipulator used was a standard Vacuum G enerators sam ple m anipulator

(UMDI).

H eating the crystal was achieved by passing an a.c. current through the copper holders

and tantalum wires.

Chromel/Alumel Thermocouple Drive Rod

Ceramic Blocks

Copper Cooling Braid

Copper Plates Mica Sheets

XL

C l

Copper Holder

Platinum Crystal

Fig. 2.9 D iagram o f crystal m ounting for the L E E D -A uger system .

In the ESC A system the copper holder was attached to a copper ‘L ’ piece. This in turn

was screw ed into the horizontally positioned stainless steel probe (see figure 2.10).

Positioning the sam ple relative to the X-ray source was achieved by means of an x, y, z

manipulator.

A liquid nitrogen well running through the core o f the probe enabled cooling o f the

crystal. The end o f the stainless steel probe is thick to allow the sam ple to be screwed into

place. This, however, results in low therm al conductivity. To overcom e this a copper sleeve

was fitted tightly around the probe and the copper ‘L ’ piece attached to the probe through the

sleeve [32]. W ith this arrangem ent the crystal could be cooled to 83K.

Probe and Liquid Chromel/Alum el Platinum

Nitrogen Well Copper Rod Thermocouple Crystal

Teflon Insulated Thermocouple Tantalum Heating Copper Holder

Heating Wire Clamp Wires

2.3.3 Crystal Cleaning and Characterization in Vacuo

M ere e x p o su re to the atm osphere alone is su fficien t to p ro d u ce co n sid erab le surface

co n tam in atio n o f a crystal sam ple. O nce such a sam p le has been m o u n te d in the U H V

ch a m b e r this surface co n tam in atio n m u st be rem oved.

V o latile surface co n tam in an ts such as C O , C O2 an d S O2 w ere deso rb ed from the

surface sim ply by heatin g . T h e m ain n o n -v o latile c o n ta m in a n ts o f the p latin u m sam ples

d e te c te d by A E S w ere carbon, calciu m and sulphur. R em oval o f these substances w as

ach iev ed by heating the crystal at lOOOK in 5x10"’ to rr o f oxygen, p ro d u cin g volatile oxides.

F o llo w in g oxygen treatm en t, b o m b ard m en t o f the crystal surface w ith argon ions w as

necessary to rem o v e the oxygen in co rp o rate d in to the specim en as w ell as any stubborn

c o n ta m in a n ts left behind. T his pro ced u re w as rep e a te d until a clean surface w as established,

as ascertain ed by A E S and by co m p ariso n o f the C O d e so rp tio n sp ectrum to that from a

prev io u sly e sta b lish e d clean surface.

Leak

Valve Sample

500V dc

Ion Current Meter

Ion Gauge

(producing Ar+ ions)

Fig. 2.11 Set-up for argon bombardment.

Argon ion bombardment was performed using an ion gauge as the ionizing agent. The

crystal surface was set facing the ion gauge and the ion gauge em ission controller was

sw itched to 10mA. The pumps were then switched o ff and the cham ber back-filled with

argon to a pressure o f -5x10"'^ torr. In the case o f the LEED /A uger system , a bias o f about

-500V was placed on the crystal sample, resulting in an argon ion current o f 6|0.A recorded at

the sample (see figure 2.11). Argon bombardment leaves the surface in a heavily damaged

condition and it is therefore necessary to anneal the sample to restore the order.

(b)

a

3

o

u

600

0

100

200

300

400

500

Binding Energy

Fig. 2 .1 2 X -ray photoelectron spectra o f the P t(3 3 1 ) crystal; spectrum (a): initial surface condition upon

placem ent in the E S C A system ; spectra (b) - (d): surface co n d itio n fo llo w in g progressive

c lean in g cycles; spectrum (e): clean P t(331) crystal.

F ollow in g the cleaning procedure, the P t ( l l l ) crystal displayed a sharp (1x1) LEED

pseudo-LE ED pattern was not attained in this work, but rather some streaking o f the spots was

observed (see figure 2.15). This effect was also observed fo r Pt(331) by Cong et al. [29],

who suggested that this was due to m ultiple step heights on some parts o f the (331) surface.

Y. Seimiya et al. [30, 31] reported a sharp (1x1) pattern from Pt(331), although they did not

mention observing double spots. Davies and Lambert [64] observed the expected LE ED

pattern fo r Pd(331) but w ith diffuse spots indicating poor long range order.

Im m ediately p rio r to adsorption o f the specimen gases the crystal was flashed to 600K,

a temperature sufficient to drive o ff any CO and H2 that may have adsorbed from the

background. As soon as the crystal cooled to the required temperature it was exposed to the

adsorbate under study.

A il o f the adsorbates studied in this w ork left carbon and nitrogen residues on the

crystal surfaces. Consequently, it was necessary to argon bombard the surfaces between each

exposure.

50000

40000

30000

20000

10000

0

600 500

0 100 200 300 400

Binding Energy

Fig. 2.13 XPS o f clean Pt(331) surface (M g source).