Guía de prácticas para el estudiante

F ig . 2.6 This fig u re shows th e h yp o th etica l poten tial experienced by a molecule Y-Y approaching a surface along a reaction co-ordin ate which is normal to the surface. X represents th e surface, Y-Y is a diatomic molecule and 0 is the dissociation energy of Y-Y. The energies Ep and Ec are the energies fo r ph ysisorption and chem isorption resp ectively, where curve P represents the physisorption potential and curve C represents th e chemisorption potential. Ea is the a ctiva tio n en ergy b a r r ie r associated w ith the chemisorbed state, and (Ea-E p) is the energy b a rrie r between th e physisorbed and chemisorbed states.

AES on cylin d rica l singla c ry s ta ls Chapter 2

(2.13) T h e i n f l u e n c e o f s u r f a c e c r y s t a l l o g r a p h y o n p h y s i s o r p t i o n a n d c h a m is o r p t io n .

The above discussions o f physisorption and chemisorption are simplistic in many ways, p a rtic u la rly in t h a t th e y ign ore the surface crystallography and assume perfectly planar surfaces. For real crystallographies, there are many energetically distinguishable adsorption sites and so th e adsorbate may reside on the surface in many d ifferen t states. These sta tes v i l l be separated in energy by energy barriers determined by the crossing points o f complex p h ysisorption and chemisorption poten tia ls and it is these potential b a rriers which co n trol th e ra te o f tra n sition from one state to another. Because of the re la tiv e ly s h o rt chem isorption bond length, the local su rface c ry s ta llo g ra p h y can be expected to have a profou nd influence on the chemisorption potential. The physisorbed state, however, can be expected to be much less aware o f the local surface crystallography as the bond length is much greater.

<2.14) T h e e f f e c t o f a d s o r b a t a / a d s o r h a t e i n t e r a c t i o n s .

When two adsorbate molecules come into close proximity on a surface, interactions o f a repulsive o r a t t r a c t iv e nature w ill oome in to play between them. These adsorbate/ adsorbate in te ra ctio n s can have a profound e ffe c t on th e chemisorption potential well. These in teractions may cause an increase or decrease in (a ) the depth of the chemisorption poten tia l wall, (b ) th e a c tiv a tio n energy fo r adsorption o r (a ) the equilibrium distance o f the adsorbates from the surface.

<2.15) M e c h a n ism s f o r c h w m is o r p t io n a n d m d a o jrp tlo n k in m tic s.

Xing*? (see also r e f.11) has proposed two mechanisms fo r chemisorption; the tra ppin g Sr*# mechanism and the tra p p in g dominated mechanism.

AES on cylin d rica l sin gle crysta ls Chapter 2

In the tra p p in g fr e e mechanism th ere is no accommodation in th e physisorption v e il p r io r to chem isorption. This is ch a ra cterized fo r chemisorption on adjacent sites w ith a random d istrib u tio n o f fille d sites by a sticking coefficient which is proportional to the square of the number o f vacant sites. Thus fo r saturation defined as 8 * i :

Where sD is th e s tic k in g fa c t o r a t zero coverage. Because th ere is no energy b a rr ie r to adsorption, the temperature dependence o f s(6) will tend to be small.

In the tra p p in g dominated mechanism (the precu rsor sta te model), chemisorption is preceded by accommodation o f the adsorbate into the physisorption well. These precursor s ta te species may then undergo su rface diffu sion p r io r to e ith e r desorption or chem isorption. The transition from the physisorbed to the chemisorbed state is controlled by th e en ergy b a r r ie r between th e two states (Ea-Ep) and also by the degree o f thermal accommodation o f th e physisorbed molecules to th e surface. This la t te r temperature dependence is expressed in a pre-expon en tial fa c t o r which contains th e ra tio o f the tra n sla tio n a l, r o ta tio n a l and vibration al partition functions fo r the activated complexes fo r desorption and adsorption, qd/qa, and the in itial sticking coefficient may be written as :

Where P D is th e p ro b a b ility o f form ing a precursor s ta te species a t zero coverage. If, as may be expected, qq is g r e a te r than q a and th e temperature dependence of the pre- exponential term is substantially less than that o f the exponential, then i f P088s0, equation

s(6) * So<i-e)2 (2.16)

Sq ■ Poll ♦ 3d exp (E*~Ep) J“ 1

q> Kt (2.17)

AES on cylin drical singla crysta ls Chapiar 2

2.17 may be vritta n as :

Sq > AexpC-(Ea-Ep)/kTD (2.18)

This aquation doas, however, make the assumption o f a fully thermally accommodated precu rsor state, i t can be seen from th is aquation th a t an Arrhenius plot o f loges0 vs. 1/T « i l l have a slope equal to (Ea-E p) and an in tercep t equal to loggA. If the precursor is not fully accommodated however, then the slope « i l l be somewhat less than (Ea-E p).

(2.16) T h e k i n e t i c s o f r e a c t i o n s b e t w e e n t w o g a s e s o v e t* m e ta l s u r f a c e s .

Consider th e c a ta ly tic oxidation o f CO by oxygen atoms chemisorbed on a copper su rface. Langmuir1^ has pointed out th a t th e r e are essentially two extreme mechanisms fo r such a reaction :

(a ) Both molecules are accommodated on th e surface prior to reaction.

(b ) Only one molecule is accommodated on the surface, and reaction occurs when the o th er species, im pinging from th e gas phase, makes a co llisio n w ith an accommodated molecule.

The two possible mechanisms w ill obviou sly be characterized by rath er differen t kin etics. Mechanism (a ), known as th e Langmuir-Hinshelwood mechanism, w ill be ch a ra cterised by a low in itia l ra ta o f reaction as the surface is initially saturated with oxygen and th ere are no a va ila b le sites fo r t h e accommodation o f CO. The reaction rate w ill go through a maximum as oxygen is removed by CO as CO2 (which is desorbed), and th e number o f vacancies fo r the CO molecule increases. Finally, the reaction ra te becomes lim ited by the low concentration o f chem isorbed oxygen and begins to f a ll to zero. In

AES on cylin d rical s in g le crysta ls Chapter 2

con trast, mechanism (b ), th e E ley-R ideal mechanism, predicts a h igh in itial ra te which fa lls o ff as the number o f chemisorbed oxygen atoms decreases. This is because the chances o f a suitable c o llisio n w ith a gas phase species depends on the number of oxygen atoms on th e surface. Both these reactio n s w ill be discussed in much more d eta il in chapter

6.

(2.17) S e t r e i a t i o n .

The d riv in g fo r c e fo r a ll adsorption phenomena is the lowering of the excess free en ergy o f the su rface. This fo rc e may also cause diffu sion o f impurities from the bulk to occupy su rface s ite s and so low er the su rface energy. This phenomenon is known as segrega tion , and it s equivalence to adsorption was f i r s t h ig h lig h te d by G ibbs14. In segregation , the chem ical potential o f the system is controlled by the bulk concentration ra th e r than by th e ga s pressure as in adsorption, and a t a given chemical potential, th e same s ta te would be reached by either experiment once equilibrium had been attained. Whilst therm odynam ically the two processes may be considered equivalent, it is obvious th a t th e k in etics o f seg reg a tio n and adsorption w ill be very d iffe r e n t as in the case o f segrega tion th e seg reg a n t d iffu s es from th e bulk, w h ilst in th e case o f adsorption, th e adsorbates im pinge from th e gas phase. There are many discussions o f the thermodynamics o f seg reg a tio n eg. Oudar10 and B lakely1*, and the readers attention is directed to a recent discussion o f the kinetics o f segregation by Rowlands and Woodruff1*. As well as k in etic d iffe re n c e s between the two processes, th ere is another v e r y reel experimental d ifferen ce in that chemical potentials accessible by one experiment are often not read ily accessible by the oth er. It is o fte n th e case th a t th e low coverages often found in segrega tio n experiments would requ ire impossibly low gas partial pressures to realise in an adsorption experiment.

AES on cylin d rica l single crysta ls Chapter 2

C h a p t e r 2 R e f e r e n c e s

(1) Auger P., t/.Phys. Radium 6 (1925) p.205

(2) Chattarji D., "The Theory o f Auger Transitions", Academic Press, London, (1976) (3) Lander, Phys. Rev. 91 (1953) p.1382

(4) Chung M.F., Jenkins L.H., Surface Science 22 (1970) p.479

(5) Sevier K.D., "Low energy electron specrometry", Wiley-Interscience (1972) New York (6) Weber R.E., Peria W.T., J.A pplPh ys. 38 (1967) p.4355

(7) Blauth E., Z. Phys. 160 (1957) p.228.

(8) Palmberg P.W., Bohn G.K., Tracey J.C., A ppl. Phys. L e tt. 15 (1969) p.254.

(9) Prut ton M., "Surface Physics", Oxford Physics Series, Clarendon Press, Oxford (1975). (10) Oudar J., "Physics and Chemistry o f Surfaces", Blackie and Son Ltd. London (1975). (11) Trapnell B.M.W., Hayward D.O., "Chemisorption”, Butterworths, London

(12) Xing O.A., P ro c . 7 th In te rn a tio n a l Vacuum C ongress & 3rd In tern a tion a l Conference on Solid Surfaces v o l.ii p.769, Vienna (1977).

(13) Langmuir I., Trans. Faraday Sac. 17 (1921) p.621.

(14) Oibbs J.W., see "J.W.Gibbs, The scientific papers", Dover, New York (1961).

(15) Blakely J.M., "In trodu ction to th e properties o f c ry s ta l su rfaces", Pergamon Press, Oxford, New York, Toronto, Sydney, Braunshweig.

(16) Rowlands G., Woodruff D.P., Philosophical Magazine A, (1979) vol.40 no.4 p.459-476.

(17) Burhop E.S.H., The A uger E ffe ct and O ther Radiationless Transitions, University Press, Cambridge, 1952.

AES on cylin d rica l single crystals Chapter 3

Chapter* 3 C o n t e n t s

(3.1) Introduction to ch apter 3.

(3.2) The u lt ra high vacuum (UHV) chamber. (3.3) The HCMA and ramp generator. (3.4) The detection system.

(3.3) The ele c tro n gun and power supply. (3.6) The physical imaging system. (3.7) The a rg o n ion gun.

(3.8) The cylin drical single c ry stals. (3.9) Saaiple heating.

(3.10) The sample manipulator.

(3.11) Sample holders f o r cylin drical crystals. (3.12) The computer interface.

(3.13) The softw are.

(3.14) D igital smoothing filte rs.

AES on cylin d rical singla c ry s ta ls Chapter 3

(3.1) I n t r o d u c t i o n t o c h a p t e r 3.

This chapter describes the equipment used in the present investigation. The Vacuum Generators A u ger electron spectrom eter v h ich was used throughout the course o f this work will be described in some detail, with particular reference to some major enhancements made by the A u th or v h ic h a llo v e d the device to be computer controlled. The cylindrical sin gle crysta ls (prepared by A rm ita ge1) v i l l be described alon g with a novel method o f mounting cy lin d rica l samples v h ic h a llo v s fo r sample h ea tin g as well as accurate axial ro ta tio n . The op era tin g program s and related s o ft v a r e f o r the computer con trolled spectrom eter v i l l be b r ie fly described, along v it h a b r ie f discussion o f the application of digital smoothing filters to Auger electron spectroscopy.

(3.2) T h e u l t r a h ig h v a c u u m <VHV) c h a m b e r .

The apparatus consisted o f a conventional stainless steel UHV chamber vh ich was fit t e d v it h titanium sublim ation and ion pumps, and v h ic h had a base pressure a fte r bakeout o f b e tter than 5x10"*0 t 0rr. The chamber was equipped v ith a sample manipulator, a Vacuum Generators (V.G.) HCMA, a V.G. high intensity argon ion gun fo r sample cleaning, and a nude B a yard -A lp h ert typ e io n ization gauge head f o r pressure measurement. Gas from th e gas handling lin e could be admitted in to th e main chamber via the argon ion gun v h ic h va s equipped v it h a V.G. leak valve, and th e chamber could be isolated from the main pumping lin e by using a b a ffle valve. The main pumping lin e consisted o f a polyphenyl eth er diffu sion pump v h ich v a s separated from the experimental chamber by a liquid n itrogen cold tra p to preven t back streaming o f oil vapour. The diffusion pump v a s backed by an Edvards ED100 type ro ta ry pump v h ic h v a s fit t e d w ith a molecular sieve fo relin e tra p to p reven t any back streaming o f r o t a r y pump oil vapours into the diffu sion pump. Pressures in th e backing line were measured with a Pirani gauge head

AES on cylin drical cingla crystals Chapter 3

situ ated between the foreline trap and the diffusion pump. The main chamber with pumping and gas handling lines is shown schematically in fig . 3.1.

<3.3) T h e HCMA a n d r a m p g e n e r a t o r .

The electron energy analysis system comprised of a V.G. HCMA (described in section 2.8) and associated electron ics. The voltages fo r the HCMA outer plate and align trace were provided by a V.G. fa s t scan ramp generator which had been modified by the Authbr to accept a 0 to 10V programming vo lta g e from an extern a l source. Thus, as well as manual operation, th e ramp gen era to r could be computer controlled. The slew rate of the in te rn a l high tension supplies could be as high as lV/millisecond, and so i f a large enough s ig n a l was available, Auger spectra could be observed in re a l time on an oscilloscope. The fa s t scan r a te also allowed re a l time observation o f the elastic peak in both E.N(E) and d e riv a tiv e modes and th is fa c ilita te d the settin g up procedure w hich involved maximizing th e ela stic peak by a d ju stin g the HCMA to sample distance. An external modulating vo lta g e from a Brookdeal lo c k -in am plifier was in tern ally amplified by the ramp gen era to r electron ics and impressed on th e outer plate ramp via an iso la tin g transform er. The fast scan mode o f the spectrometer necessitated a rather high modulation freq u en cy o f about 40kHz, and care had to be taken to ensure th a t ca p a citive loss of the modulation voltage in the connecting cables was not significant. The modulation voltage could be manually varied from 0 to 25V peak-to-peak. A d ig ita l v o lt meter was added to th e ramp gen era to r and was ca lib ra ted to display the pass energy o f th e analyser. The ca lib ra tio n point taken was the carbon peak at 272eV, and the ramp was found to be reasonably linear up to about 800eV.

AES on cylin d rical sin gle crystals Chapter 3

Fig. 3.1 This figure shows the main pumping line and the gas handling line of the apparatus used throughout th is in vestiga tio n . Gas from th e gas handling line was admitted to the main chamber via a leak valve attached to the argon ion gun.

AES on cylindrical sin gle crystals C hapter 3

(3.4) T h e d e t e c t i o n sy ste m .

As described in section 2.8, the detection system o f the V.G. KCMA consisted of