FLUORESCENCE YIELD NEAR EDGE STRUCTURE (FYNES) : A NOVEL TECHNIQUE FOR IN SITU SURFACE CHEMISTRY

HAL Id: jpa-00227315

https://hal.archives-ouvertes.fr/jpa-00227315

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.

FLUORESCENCE YIELD NEAR EDGE STRUCTURE (FYNES) : A NOVEL TECHNIQUE FOR IN SITU

SURFACE CHEMISTRY

D. Fischer, F. Zaera, J. Gland

To cite this version:

D. Fischer, F. Zaera, J. Gland. FLUORESCENCE YIELD NEAR EDGE STRUCTURE (FYNES) :

A NOVEL TECHNIQUE FOR IN SITU SURFACE CHEMISTRY. Journal de Physique Colloques,

1987, 48 (C9), pp.C9-1097-C9-1100. �10.1051/jphyscol:19879198�. �jpa-00227315�

JOURNAL DE PHYSIQUE

Colloque C9, suppl6ment au n 0 1 2 , Tome 48, dgcembre 1987

FLUORESCENCE YIELD NEAR EDGE STRUCTURE (FYNES) : A NOVEL TECHNIQUE FOR IN SITU SURFACE CHEMISTRY

D.A. FISCHER, F. ZAERA and J. L. GLAND*

National Synchrotron Light Source, Brookhaven National Laboratory, Upton, NY 11973, U.S.A.

" ~ x x o n Research and Engineering Co., Annandale, NJ 08801, U.S.A.

~6sum6. La technique appel&e en anglais Fluorescence Yield Near Edge Structure (FYNES) qui est dgcrite ici represente une progression majeure de notre capacit6 de determiner l'hybridation et l'orientation des produits intermddiaires de rhaction organique en presence d'une atmosphsre r4active. Cette technique emploie le rayonnement synchrotron dans le domaine des rayons X mous pour crder des trous au coeur des mollcules adsorbdes; la dgcroissance de ces trous est alors d6tect6e par leur fluorescence rayons X caractgristique pendant la d'esexcitation.

Puisque la technique FYNES est une spectroscopie de photon entr;, photon sorti, elle est tout B fait compatible avec un milieu gazeux re'actif. Des experiences

in

mol6cules organiques contenant le carbone et le soufre sont decrites. La technique FYNES fournit une occasion exceptionnelle d'utiliser les synchrotron pour

caracteriser les produits intermgdiaires de surface et les vitesses de r6action sur des surfaces de catalyste mode'les dans un milieu rdactif.

Abstract. The Fluorescence Yield Near Edge Structure (FYNES) technique described represents a major advance in our ability to determine the hybridization and orientation of organic reaction intermediates in the presence of a reactive atmosphere. This technique uses synchrotron radiation in the soft x-ray region to create core holes in adsorbed molecules; the decay of these core holes are then detected by their characteristic x-ray fluorescence during the deexcitation process. Since FYNES is a photon-in, photon-out spectroscopy, it is fully compatible with reactive gaseous environment. In situ experiments are described involving chemisorption, displacement, and dehydrogenation of organic molecules containing Carbon and Sulfur. The FYNES technique provides a unique opportunity for utilizing synchrotron facilities in characterizing surface intermediates and reaction rates on model catalyst surfaces in reactive environments.

Introduction

It is well known that the detection of characteristic fluorescence radiation resulting from core hole deexcitation of atoms diluted in condensed matter is a useful way to obtain local structure by means of EXAFS [I]. Although fluorescence EXAFS is a powerful technique, the fluorescence yield (FY) strongly decreases with decreasing atomic number [2]. Thus the FY technique was thought to be ill suited for Surface EXAFS studies of low-z adsorbates on high z substrates. However this class of experiments is of great interest in the study of catalysis. In fact, we have demonstrated that such a FY technique is feasible, and we have reported results on studies for thiophene adsorbed on Ni(100) [sulfur K edge] [3] and ethylene on Cu(100) [carbon K edge] [4]. We have also performed prototype synchrotron experiments coupling a soft x-ray proportional counter [5], reaction cell, and suitable soft x-ray entrance window to form a system allowing the experimenter to perform sample characterization experiments under real reaction conditions encountered in catalysis 1 6 1 . The initial series of experiments

described briefly in this paper were focussed on coadsorption of CO and hydrogen on

Article published online by EDP Sciences and available at http://dx.doi.org/10.1051/jphyscol:19879198

C9-1098

JOURNAL DE

the Ni(100) surface. We found that chemisorbed CO an be rapidly displaced from the Ni(100) surface by hydrogen pressures above losE torr in the 290 to 330 X temperature range. This unexpected displacement occurs despite the fact that CO is adsorbed with a heat of 30 kcal/mole [ 7 , 8 ] , while hydrogen is adsorbed with a heat of 23 kcal/mole [9].

Apparatus

The experimental apparatus consists of a multiple level vacuum chamber with a small high pressure reaction chamber on top where the FYNES experiments are performed. The primary vacuum chamber was equipped with standard surface science instrumentation including facilities for Auger electron spectroscopy, thermal desorption, low energy electron diffraction, and sputtering. A long travel manipulator was used to transfer the sample to the reaction chamber at the upper level. The reaction chamber could be isolated from the main chamber by a gate valve and pumped independently using a turbomolecular pump as indicated in Figure 1. The reaction chamber could also be isolated from the synchrotron ring by a combination of two thin windows and a ballast region. The windows were 1,000 boron and tin films, supported on a high transmission nickel grid mounted on the center of two 2-3/4" gate valves. Thus each window could be inserted or removed independently. The windows transmit over 50% of the radiation in the 300 eV energy range. The tin window has an absorption edge around 490 eV which reduces second order radiation coming from the monochromator for primary energies above 245 e V in the carbon edge region. Each window could withstand a hydrogen pressure

differential over 10 torr without any significant leakage to the vacuum side.

, , , E W P O R T 1 ~ V;LVE 0.1 urn BORON SAMPLE MANIPULATOR

WINDOW V A L E 0.1 u AND FEEDTHRU

PHOSPHOR SCRE

SAMPLE PREP. CHAMBER

Ficr. 1. Apparatus for soft x-ray absorption measurements under reactive atmosphere.

Fluorescence radiation from the sample was collected by a differentially pumped ultra-high vacuum compatible proportional counter mounted below the plane of incidence, perpendicular to the incident light as shown in the figure. The

detector assembly is mounted via a welded bellows and for these experiments was positioned -1 cm from the sample, providing a solid angle of collection of 10% of 47rSr. Radiation from the sample passes into the detector through two 1 gm thick polypropylene windows (each 83% transmitting for C-K radiation) [lo] with a differentially pumped region in between. The polypr~pylene windows are mounted on a 90% transmitting stainless steel grid and are capable of withstanding atomspheric pressure.

The energy discrimination characteristics of the proportional counter detector have proven to be crucial for our application, since excitation of the Ni-L edge by third order light occurs at about the same monochromator position neededafor C - K electron excitation. In order to perform carbon-edge EYNES experiments we had tg reject the signal coming from the nickel sample substrate by using a pulse height discriminator and setting a window around the adsorbate carbon peak of interest.

We believe the elastically scattered contribution from the substrate to be small because of the choice of polypropylene as window material, which strongly attenuates light above the C K edge (284.2 eV). The transmission of 2 pm of polypropylene for C - K fluorescence (277 eV) is 69% but at 290 eV it is less than 1% highlighting the cffoice of a carbon containing polymer window transparent to its own characteristic radiation (C-Ka).

Results and Discussion

Spectra at glancing (30") and normal incident photon geometries for a CO saturated Ni(100) crystal in vacuum at 100 K are shown as insets in Fig. 2. These spectra were taken with 300 mA in the storage ring (100 points, 4 sec/pt.). They have been normalized by spectra from the clean surface, but a count rate of about 4000 c/s was observed at the r* resonance peak for normal incidence (hv = 287.6 eV). These spectra illustrate the high quality of our J?Y detection technique and are qualitatively identical to those obtained using electron detection techniques Ill]. However, the signal-to-background (STB), which is the ultimate parameter in determining the sensitivity of the technique, is close to two for the edge jump of a C O saturated nickel surface, and goes as high as 10 for the r* resonance peak at normal incidence. These STB are better than those obtained by any electron yield technique by more than an order of magnitude 141. If we establish a criteria where structure determinations are possible for STB larger than .l, this means that our fluorescence detection scheme should be able to detect coverages as low as 1% of a chemisorbed C O monolayer.

The lower panel in Fig. 2-illustrates that chemisorbed CO is rapidly displaced by hydrogen pressures about 10 torr. Initially the Ni(100) surface was saturated with C O at 300 K and a FYNES spectrum was acquired. The monodhromator was then tuned to the peak of the r* resonance and the entrance windows inserted. After monitoring the intensity of the r* resonance for several hundred seconG2, hydrogen was introduced into the reaction chamber to the desired pressure (9x10 torr in this case). As indicated in Fig. 2 the C O was rapidly displaced from the surface.

FYNES spectra for both polarizations were taken after the displacement reaction to insure that the C O was actually displaced from the surface. In addition, Auger spectra taken following a desorption cycle verified the cleanliness of the surface.

Mi (100) Hydrogen Displecernent of CO Saturated Surface Norinallud F l u o m c a m YIeld

260 so0 320 280 300 320

Photon EMQy (*V) Photon E n q y (.V) Fie. 2. Displacement of a monoljiyer of chemisorbed C O by 9x10- torr of hydrogen occurs in less than 600 sec. at 330 K . The inset spectra illustrate

normalized C O spectra with normal and glancing incidence light for a C O saturated surface at 100 K .

9.0 Q =.

t

t

0.0

g

0 300 #u, r"

Tlnw (a)

A series of transient ??YNES experiments of this type were performed for several temperatures and hydrogen pressure. An example is shown in Fig. 3 for a hydrogen pressure of 0.012 torr. The logarithm of the C O coverage is plotted as a function of time. The rate of C O displacement is simply the slope of a linear coverage plot. These transient FYNES results clearly allow us to make detailed kinetic

C9-1100 JOURNAL DE PHYSIQUE

measurements in the presence of hydrogen. Reaction rates in the to

monolayer/sec region are easily accessible using current technology. The increase in rate with increasing temperature indicates that the displacement reaction is thermally activated. The results indicate that the displacement is first order in CO coverage since there is a linear decrease in the logarithm of the CO coverage.

1 .O

k

4

transient FYNES

experiments, as a function of displacement time for several temperatures.

Reaction rates in the monolayer/sec to 10 monolayer/sec can be characterized using this in-situ technique.

5 0 0

Time (s)

Conclusion

Fluorescence detection is an exciting development for soft X-ray absorption studies on surfaces. With current light sources and detector systems adsorbed monolayers can easily be detected and characterized using the fluorescence detection method. This detection method allows adsorbed nlonolayers to be characterized in the presences of reactive gases. The application of FYNES as a transient method has been demonstrated in our recent study of CO displacement from the Ni(100) surface described in this paper. The observation that chemisorbed CO is displaced by more we'akly adsorbed hydrogen was quite unexpected; this discovery highlights the Iyportancs of in-situ surface spectroscopies. Surface reaction rates in the 10 to 10 monolayer per second range can be measured using current detector systems and light sources. Pressures up to 10 torr and sample

temperatures up to 500 K can be accommodated with the current system design. A wide range of important surface reactions are currently accessible and can be characterized using this method in the near future.

References

Jaklevic, J., Kirby, J. A., Klein, M. P., Robertson, A. S., Brown, G. S., and Eisenberger, P., Solid State Comm. 3, 679 (1977).

Krause, M. O., J. Phys. Chem. Ref. Data 8 , 307 (1979).

Stohr, J . , Kollin, E. B., Fischer, D. A., Hastings, J. B., Zaera, F., and Sette, F., Phys. Rev. Lett.

2,

Fischer, D. A . , Dobler, U., Arvanitus, D., Wenzel, L., Stohr, J., Surf. Sci.

m,

Fischer, D. A., and Gland, J , L., Int. Conf. Soft X-Ray Optics Technology (Berlin 1986) to be published.

Zaera, F., Fischer, D. A., Shen, S., Gland, J. L., to be published, Surf.

Sci.

Tracy, J . C., J. Chem. Phys.

56,

Tracy, J. C. and Burkstrand, J. M., CRS Crit, Rev. in Solid St. Sci. 4 , (1974) Issue 3, 381.

Christmann, K., Schober, O., Ertle, G., Neumann, M., J. Chem. Phys.

60,

[lo] Henke, B. and Tester, M., in Advances in X-rav Analvsis, Vol. 18, Plenum Press (1975)

.

[Ill Stohr, J. and Jaeger, R., Phys. Rev. B z , 4111 (1982).