PHOTON STIMULATED FIELD DESORPTION OF HYDROGEN FROM RHODIUM

HAL Id: jpa-00226842

https://hal.archives-ouvertes.fr/jpa-00226842

Submitted on 1 Jan 1987

HAL

is a multi-disciplinary open access archive for the deposit and dissemination of sci- entific research documents, whether they are pub- lished or not. The documents may come from teaching and research institutions in France or abroad, or from public or private research centers.

L’archive ouverte pluridisciplinaire

HAL, est

destinée au dépôt et à la diffusion de documents scientifiques de niveau recherche, publiés ou non, émanant des établissements d’enseignement et de recherche français ou étrangers, des laboratoires publics ou privés.

PHOTON STIMULATED FIELD DESORPTION OF HYDROGEN FROM RHODIUM

W. Drachsel, S. Jaenicke, A. Ciszewski, J. Dösselmann, J. Block

To cite this version:

W. Drachsel, S. Jaenicke, A. Ciszewski, J. Dösselmann, J. Block. PHOTON STIMULATED FIELD

DESORPTION OF HYDROGEN FROM RHODIUM. Journal de Physique Colloques, 1987, 48 (C6),

pp.C6-227-C6-232. �10.1051/jphyscol:1987637�. �jpa-00226842�

PHOTON STIMULATED FIELD DESORPTION OF HYDROGEN FROM RHODIUM

W. Drachsel, S. Jaenicke, A. ~iszewski*, J. Dosselmann and J.H. Block Fritz-Haber-Institut der Max-Planck-Gesellschaft, Faradayweg 4-6, D-1000 Berlin 33, F.R.C.

A b s t r a c t

Using synchrotron radiation (40-120 e V ) from the BESSY storage ring, photon stimulated field desorption of hydrogen from a rhodium emitter was observed. The yield of photon-stimulated H 2 ' and H ' ,discriminated by a time-of-flight technique, displayed a strong photon energy dependence: a broad maximum a t 70 eV was found;

the onset correlated with 4p3n and 4 p l n core level excitations of rhodium, whereas the 4s-excitation caused no increase.

T h e excitation transfer from the substrate Lo the adsorbate most probably takes place via metal Auger electrons (XF,SI)-mechanism). From the knlperature dependence of the desorption yield i t is concluded t h a t only field-adsorbed t i 2 can be desorbed following phobn-excitation of the rhodium. T h e occurence of H' a s a desorbing species is attributed to field-dissociation and especially a t high photon flux, also to photo-dissociation o f t 1 2 * .

1. I n t r o d u c t i o n

The observed enhancement of field desorption by irradiation with photons of suitable energy can be described by two concepts (11 : a. a s photon-assisted field desorption or b. a s field- assisted photon-stimulated desorption. To clarify the underlying mechanism i t is essential to measure the dependence of the enhancement on photon energy and field strength. For the excitation of valence and core levels i n the substrate a s well a s in the adsorbate, synchrotron radiation provides a n ideal tunable light source. Laser excitation so far was only successful in special cases. Ethylene was desorbed from a silver surface in the presence of a field i n a one- photon process (21, and enhanced field evaporation of silicon was observed in what appeared to be a multi-photon excitation (31.

Here, the observation of a threshold in the photon-stimulated field desorption of hydrogen from rhodium is reported and arguments for a desorption model are given.

- -

*

o n leave from Institute of Experimental Physics, University o f Wroclaw. PL-50 205 Wroclaw

Article published online by EDP Sciences and available at http://dx.doi.org/10.1051/jphyscol:1987637

JOURNAL DE PHYSIQUE

2. E x p e r i m e n t a l

The experiments were performed a t the BESSY storage ring using a toroidal grating monochromator (40-120eV). The apparatus has been described elsewhere 141. Briefly, i t consists of a field ion microscope, where the tip to screen distance (- 10 cm) also serves a s the fli h t path for time of flight (ToF) mass analysis (fig. 1). Monochromatized synchrotron raiiation (SR) was focused onto the emitter tip. The flight times of the desorbed ions were recorded a t the repetition rate of the SR-light pulses of 5 MHz a t BESSY.

The relative light intensity was monitored via the photo current from a gold coated glass disc.

The etched rhodium tip was cleaned in vacuum by repeated annealing and field evaporation using neon a s image gas.

For a good signal to noise ratio i t is mandatory to work a t pressures less than 2.10-9 mbar.

Hydrogen from a pressure can (Messer Griesheim, purity better than 99.999%) was introduced with open main valve to the pumps.

marker 140 ps

Fig. 1 - Experimental arrangement of the 10 cm ToF-field ion microscope 3. Results

Previous ex eriments with "white" SR light [5] had shown that i t is possible to obtain photon assisted fieyd desorption from field emitters des ite the small target area of 10-10 em2.

For measurements with monochromatie light wit{ a 10-3 fold lower intensity, the optimum signal-to-noise ratio must be used, which is achieved a t a hydrogen pressure of 2.10-10 mbar ( a t 80 K). The hydrogen desorption signal from the rhodium emitter surface depends under these conditions linearly on the pressure, whereas a t higher pressure the hydrogen coverage and therefore also the desorption signal saturate. The competing field-ionization rate, which is responsible for the random background in the ToF spectra grows always linearly with the pressure.

Photon I lime of flight [ channel I

- ^-

^I

Pulse

r = O 1 channei a 0.8 ns 3

-

d Fig. 2 a-c Time-of-flight spectra of hydrogen from

A u

Rhodium at p=2.10-10 mb, F=20 Vlnm,

.-

^a

T = 80 K at 3 different photon energies

=-.

C

- I

5

-

0 2 a

.-

_ZI

C

-

M

% white light

Fig. 3 a-c Spectral dependence of the hydrogen yield a t 3 different field strengths, p = 2 - 1 0 - l o m b , T = 80K

o d - Q ^{al cn}

t

^{I"' I I I I 0 L}

I

^{b 1 0 h}

C1 _M X=25nm

-

^.K

-.

_V)

&

t

=

^0.5

-

S -

d

0 ) 4-

E _- l

9

0.0

-

⁰

I

^{"O ' I I I}

H

: , I

I 2

-

_{cn h=18nm}

_- ^.-

^al _h

3 1.5

-

H t

^C

4- C ^QI

3 ⁰¹

0 0

u 1.0

- -

^L

d u

h

a .c

4-

100 150 200 250

t

M 2 -

+ C 3

U 0

d

al 1 -

+ d

L.

photon energy IeV1-

-

photo- ion H2-

JOURNAL DE PHYSIQUE

0 0 0.5 1.0

light intensity [ arbitrary units

1- H;

from Rh

T=

80

K

- 2.2=v/A

hvz82.7 eV

Fig. 4 Light intensity dependence

In fig. 2a, a n original ToF-spectrum for the "white" SR light illumination is presented with the Rh tip held a t 80 K. Under the same conditions, the spectral measurements a t 25 n m and 1 8 n m a r e shown i n fig. 2b and 2c. Over the energy range of 40-120 eV, desorption spectra were taken a t three field strengths, which a r e shown in fig. 3a-c. The hydrogen intensity in these plots re resents the sum of the H2+ and H + signals above the background, taken from ToF-spectra. !he desorption rate is corrected for changes of gas pressure and light intensity To obtain the photon flux, the current from the gold diode was converted using the d a t a of Saloman [ 6 ] .

By varying the width of the entrance slit to the monochromator, the photon flux a t the tip could be changed. A linear relationship between photon flux and desorption signal, indicative for a single photon process, was observed (fig. 4).

4. Discussion

In fig. 2 some characteristic differences between ToF spectra taken with "white" a n d monochromatic illumination become evident. Gas-phase photo-ionization, which appears a s a broad band around 210 nm with white light is virtually absent under monochromatic irradiation. The H+M[2+ ratio changes from 0.6. with monochromatized light to 1.1 with white light, indicating additional photo-dissociation of desorbing H2+ under intense

The pressure was adjusted so t h a t the desorption signal did not yet saturate. Under these conditions the normalized hydrogen yield reflects the true desorption cross section a s a function of hoton ener

.

As is seen from fig. 3, varying the field strength between 30 a n d 36 Vlnm i n i u e o c e s n e i t g r the threshold a t appx. 50 eV nor the shape of the desorption band.

Steps in the desorption yield correlate with the rhodium 4~312 and 4pl/2 core excitations, whereas t h e 4 s excitation does not show up in the desorption spectrum.

core hole a t a metal atom. If this hole is filled by a valence electron of the adsorbate, t h e KF mechanism applies. Additional ejection of Auger electrons can increase the oxidation state of t h e adsorbed molecule by two or three units, so t h a t ionic desorption becomes possible. In XESD, the core hole is filled by a metal electron. Thereby Auger electrons of sufficient energy a r e generated which can lead to the desorption of adsorbed molecules.

The K F mechanism will only be active when all paths for intramolecular Auger processes a r e blocked. This is not the case, when desorption from a metal surface i s to be considered, where a large number of mobile electrons is available in the conduction band. In addition, there i s little orbital overla between the metal atom and the adsorbed hydro en molecule. Let u s

%

now consider the iniuence of the hi h electrostatic field. It was found t a t desorption starts a t kink sites [91, and the electric f e l d will certainly reduce the electron density a t such protruding atoms [lo]. Also, the valence electrons of the field-adsorbed H2 will be pulled towards the metal surface. Nevertheless this field effect should not be sufficient to alter the desorption mechanism to such degree t h a t the KF-mechanism becomes dominant.

E r n s t [ I l l determined t h a t an electron energy above 40 eV is necessary for electron- stimulated field-desorption. The Rh 4p Auger electrons have just about this energy. They can therefore lead to an efficient desorption. Since the absorption coefficient of the metal increases steeply a t the 4p edge, the number of photo-electrons which a r e created near the surface will increase sharply. The XESD mechanism therefore explains the onset behaviour of the desorption yield.

Discussing PSD, Menzel [12] pointed out t h a t structures in the desor tion yield sometimes coincide incidentially with inner core transitions in the metal. He coulc?explain the observed spectrum for CO desorption from Ru with the assumption of multiple valence-excitations in the adsorbate, which lead to long living, highly correlated excited electronic states. The fact t h a t the H2+/H+ ratio is governed only by the field strength does not favour this model for our case, because following a multiple valence excitation i n H2 one expects a high degree of dissociation.

To decide which model describes correctly the photon-stimulated field desor tion, further experiments with d i r e r e n t substrates a s well a s adsorbates are needed. So &r i t could be demonstrated t h a t the primary step is a one-photon excitation. This primary excitation takes place in the bulk of the metal substrate. We prefer the XESD-model to explain the excitation transfer to the adsorbate. However, there is a t least one other desorption mechanism active.

This i s apparent in fig. 3, where a low, but distinct desorption yield is seen below the threshold a t -50 eV. The threshold for this low-energy process has not yet been determined. I t lies a t energies below 10 eV, because the desorption signal was still observed after inserting a LiF window into the light beam. This low-energy desorption mechanism is tentatively attributed to a direct valence excitation of the field-adsorbed hydrogen.

A c k n o w l e d g e m e n t

This work was funded by the German Federal Minister for Research and Technology (BMIT) under the contract number 05 390 FX B2 TP3. A. C. thanks the Max-Planck-Society for a research scholarship.

C6-232 JOURNAL DE PHYSIQUE

References

S. Jaenicke, A. Ciszewski, W. Drachsel, U. Weigmann, T.T. Tsung, J.R. Pitts, J.H. Block, D. Menzel, J. Phys. (Paris) 47 (1986) C7-3243.

S. Nishigaki, W. Drachsel, J.H. Block, Surf. S c i . B ( l 9 7 9 ) 389.

T.T. Tsong, Phys. Rev. =(I9841 4946.

W. Drachsel, U. Weigmann, S. Jaenicke, and J.H. Block, in: Desorption Induced by Electronic Transitions DIET 11, W. Brenig and I). Menzel (eds.), Springer Series in Surface Science, Vol. &

Springer, Heidelberg 1985, p. 245.

S. Jaenicke, U. Weigmann, J.R. Pitts, W. Drachsel, J.H. Block, D. Menzel, Chem. Phys. in press.

E.B. Saloman, Nucl. Instr. Meth. m ( 1 9 8 0 ) 79.

M.1,. Knotek and P.J. Feibelman, Phys. Rev. l3ett,*(1978) 964.

R. J a e g e r , J . Stdhr, T. Kendelewicz, Phys. Rev. 828 (1983) 1145.

T.T. Tsong, T.J. Kinkus, C.F. Ai, J . C h e m . P h y s . 2 (1983) 4763.

R. Smoluchowski, Phys. Rev. m ( 1 9 4 1 ) 661.

N . Ernst, J.14. Block, Surf. S c i . N ( l 9 8 2 ) 561.

S. Auer, P. Feulner, D. Menzel, Phys. Rev.&, (1987) 7752.