Professor Spivey is the McLaurin Shivers Professor of Chemical Engineering at Louisiana State University and Director of the DOE Energy Frontier Research Center at LSU. Professor Spivey’s research interests include the application of the principles of heterogeneous catalysis to catalytic combustion, control of sulfur and nitrogen oxides from combustion processes, acid/base catalysis (e.g., for condensation reactions), hydrocarbon synthesis, and the study of catalyst deactivation.

Catalysis Volume 11

A Review of Recent Literature

By James J. Spivey, Sanjay K. Agarwal

The Royal Society of Chemistry

Copyright © 1994 The Royal Society of Chemistry
All rights reserved.
ISBN: 978-0-85186-654-3

Contents

Chapter 1 Applications of Secondary Ion Mass Spectrometry in Catalysis and Surface Chemistry By Herman J. Borg and J. W. (Hans) Niemantsverdriet, 1,
Chapter 2 Frequency Response Techniques for the Characterization of Porous Catalytic Solids By Sebastian C. Reyes and Enrique Iglesia, 51,
Chapter 3 H2 Adsorption on Supported Noble Metals and Its Use in Determining Metal Dispersion By Calvin H. Bartholemew, 93,
Chapter 4 Binding Properties of Hydroxyl Groups on Substrates in Aqueous Environments: Their Relationship to Catalyst Preparation By James A. Schwan, Cristian Contescu, and Jacek Jagiello, 127,
Chapter 5 Variation in the Mechanism of Catalytic Reactiom with Crystal Face By Eriko Yagasaki and Richard I. Masel, 165,
Chapter 6 Partial Oxidation and Ammoxidation of Propane: Catalysts and Processes By Yoshihiko Moro-oka and Wataru Ueda, 223,
Chapter 7 Selective Oxidations of C4 Paraffins By Fabrizio Cavani and Ferrucio Trifiro, 246,
Chapter 8 Bimetallic Catalysts: Structure and Reactivity By László Guczi and Antal Sárkány, 318,
Chapter 9 Catalytic Dehydrogenation of Lower Alkanes By Daniel E. Resasco and Gary L. Haller, 379,
Chapter 10 Catalytic Partial Oxidation of Methane to Synthesis Gas (Syngas) By Gary A. Foulds and Jack A. Lapszewicz, 412,
Chapter 11 Characterization of Catalysts with Microcalorimetry By Paul J. Andersen and Harold H. Kung, 441,

CHAPTER 1

Applications of Secondary Ion Mass Spectrometry in Catalysis and Surface Chemistry

BY HERMAN J. BORG AND J.W. (HANS) NIEMANTSVERDRIET

1 Introduction

Secondary ion mass spectrometry (SIMS) is an extremely sensitive ultra-high vacuum (UHV) technique for surface analysis, which has found general recognition in catalysis and surface chemistry. However, SIMS is not as widely applied as, for example, X-ray photoelectron spectroscopy (XPS), electron microscopy, or the vibrational spectroscopies. Unique features of SIMS for studying catalysts are its unsurpassed sensitivity, enabling one to establish the presence of promoters and poisons with concentrations down to the parts-per-million level, and its capability to demonstrate through the analysis of molecular secondary ions that elements are in contact with each other. In addition, SIMS can give its information as a function of depth, in sputter profiles, or as a function of position, in chemical maps. In surface science, SIMS offers the attractive possibility to monitor the concentrations of surf ace species in real time, making the technique a promising tool for kinetic studies.

SIMS has its weaknesses also, which relate to the fact that the physical principles behind the formation of secondary ions are not yet fully understood. In particular, the variation of SIMS yields of an element in different bonding geometries, commonly referred to as the matrix effect, forms a serious drawback with respect to quantification.

SIMS as a characterization technique reveals its information in several ways:

• Single ions straightforwardly reveal the presence of certain elements in the sample;

• Molecular ions such as MoO+ or RhCi- indicate which elements may be in contact in the catalyst;

• Characteristic fragmentation patterns sometimes give evidence for the type of compound present (for example, MoO3 gives a highly specific pattern of Mo-, MoO-, MoO-2, MoO-3, and MoO-4 ions);

• Suitably chosen intensity ratios can — after careful calibration — be used to obtain (semi) quantitative information (for example, for hydrogen adsorbed on nickel, the ratio Ni2H+/Ni+ correlates well with the hydrogen coverage).

In this paper we describe briefly the physical phenomena such as sputtering, ion emission, ionization, and neutralization that are involved in SIMS. Next we review applications of SIMS in catalysis. We first concentrate on the catalyst itself and the chemistry involved in catalyst preparation. Then we describe how SIMS has been used to reveal information on the interaction of catalysts with reacting gases, promoters, and poisons. Finally we discuss SIMS in reactivity studies on model surfaces, where the technique has been used to address questions as adsorption and surface reactions. Thorough introductions to SIMS have been written by Benninghoven et al., and Vickerman et al., while many applications have been described by Briggs et al. Earlier reviews on SIMS in catalysis and surf ace chemistry, with the focus on surface reactivity studies, have been given by Delgass et al. and Greenlief and White. More recently, a number of successful SIMS applications in catalysis were reviewed by Niemantsverdriet.

2 Theory of SIMS

2.1 The Principle of SIMS. – The principle of SIMS is conceptually simple: a primary ion beam (Ar+, 0.5 to 5 keV) is used to remove atoms, ions, and molecular fragments from the surface which are consequently analyzed with a mass spectrometer. It is as if one scratches some material from the surf ace and puts it in a mass spectrometer to see what elements are present. However, the theory behind SIMS is quite complicated and far from being understood.

When a surface is exposed to a beam of ions, energy is transferred to the surface region of the sample by a collisional cascade. Some of the energy will return to the surface and stimulate the ejection of atoms, ions, and multiatomic clusters (Figure 1). In SIMS, secondary ions (positive or negative) are detected directly with a mass spectrometer. In the more recently developed secondary neutral mass spectrometry (SNMS), the secondary neutral species are ionized, for example by electron impact ionization, before their mass is analyzed. Most SIMS instruments use a quadrupole mass spectrometer, which has the advantage that it is relatively inexpensive. However, for higher sensitivity and higher mass resolution, magnetic sector or time-of-flight (TOF) mass spectrometers are used. With such instruments one can, for example, distinguish between the masses of Al+ (26.98) and C2H3+ (27.024), which is not possible with a quadrupole-based instrument.

Strictly speaking, SIMS is a destructive technique, but not necessarily a damaging one. In the dynamic mode, used for making concentration depth profiles, several dozens of monolayers are removed per minute. In static SIMS, however, the rate of removal corresponds to one monolayer per several hours, implying that the surface structure does not change during the measurement (between seconds and minutes). In this case one can be sure that the molecular ion fragments are truly indicative of the chemical structure on the surface. Static SIMS is a very gentle, nondamaging technique, which causes less damage to a surface than Auger electron spectroscopy (AES), for example, or even standard XPS does. For example, Benninghoven and coworkers have been able to observe the intact emission of very large biomolecules such as vitamin B12 (m/e=l,356) adsorbed on silver.

For a good understanding of SIMS spectra, it is important to have at least a qualititative understanding of phenomena such as sputtering, ionization, and neutralization; ion-induced electron and light emission; and the energy distribution of sputtered particles.

2.2 Secondary Ion Yields. – The signal intensity of an elemental positive or negative secondary ion is given by

I ±s = Ip Y R± csurf T (1)

where

I ±s = Measured flux of positive or negative secondary ions [s-1];

IP = Flux of primary ions [s-1];

Y = Sputter yield of the element, equal to the number of atoms ejected per incident ion;

R± = Probability that the particle leaves the surface as a positive or negative ion;

csurf = Fractional concentration of the element in the surface layer (a number between 0 and 1); and

T = Transmission of the mass spectrometer, typically 10-3 for a quadrupole and 10-1 for a TOF instrument.

The essential quantities that determine the yield of secondary ions in a SIMS spectrum are thus the sputter yield Y and the ionization probability R±.

Sputtering is a reasonably well-understood phenomenon. Sputter yields depend on the properties of the sample as well as on those of the incident ions. Sputter yields of the elements vary roughly between 1 and 10 (see Table 1), with a few exceptions to the low side, such as bismuth with a sputter yield around 0.1 under SIMS conditions, and to the high side such as zinc, which has a sputter yield of around 15 under 5 keV argon bombardment.

The sputter yield depends on the mass of the primary ion, its energy, and the angle of incidence. This is illustrated in Figure 2. The effects of these three variables on the sputter yield are

• Mass: a heavy ion such as xenon transfers more energy to a surface and sputters more efficiently than a lighter ion such as argon.

• Energy: increasing the energy of the primary ions initially increases the sputter yield. At high energy, however, ions penetrate deeper into the solid and their energy is dissipated further away from the surface. The result is that fewer collision cascades reach the surface and the sputter yield decreases at high energy. Thus the sputter yield goes through a maximum (Figure 2[a]).

• Angle: the penetration depth of primary ions with a certain energy into the sample will be at maximum when the angle of incidence is perpendicular to the surface (if we exclude channeling phenomena). Hence, the sputter yield will increase when the angle of incidence, 0, measured with respect to the surface normal, increases. We expect that the sputter yield varies with l/cos0 and this is indeed approximately the case. At glancing angles, however, the ions may scatter back from the surface into the vacuum. As a result, the sputter yield maximizes at angles around 70° (Figure 2[b]).

Sputtering can be done with ions and with atoms. For metal samples, the sputter yields are the same. Actually, an incident ion neutralizes rapidly when it enters the metal, and hence there is no difference in the sputtering effect of argon atoms and argon ions. For insulators or semiconductors, sputter yields are up to a factor of 2.5 higher for ions than for atom beams, which has consequences for fast atom bombardment SIMS (FABSIMS).

2.3 Electron and Photon Emission Under Ion Bombardment. – Inherently connected to the interaction of ions with a solid is the emission of electrons from the sample. Typical yields are 0.1 to 0.2 electrons per incident argon ion of 2 keV, and 0.2 to 0.5 at 5 keV for all noble gases.

In addition to emitting electrons, a solid bombarded with ions in the keV range emits electromagnetic radiation from the near infrared (IR) to the near ultraviolet (UV), with a photon yield of typically 10-4 per incident ion for a metal, and 10-2 to 10-1 for insulators. If the primary beam is intense, as in the dynamic SIMS range, and the sample is an insulator, one observes a bright glow at the point where the beam hits the sample. With conductors, the effect is either small or insignificant.

2.4 Energy Distribution of Secondary Ions. – It is important to know the energy distribution of secondary ions because it has consequences for their detection, especially in the case of insulators. As Figure 3 shows, the energy distribution of elemental secondary ions usually has a peak between 15 and 30 eV and falls off rapidly at higher energy but exhibits a low level tail to a few hundred eV. The maximum in the energy distribution is hardly affected by the energy of the primary ions. The high energy tail, however, is the result of short collision cascades close to the point of impact of the primary ion. The intensity of this high energy part of the distribution increases when the primary ion energy goes up.

The energy distributions of molecular ions peak at significantly lower energies and do not tail to high energies. Figure 3 illustrates this for the emission of secondary Cu+2 and Cu+3 ions from a copper target. As quadrupole mass spectrometers operate with relatively narrow energy windows of typically 10 eV, it is evident that the setting of the window is essential for the detection of molecular ions.

The angular dependence of the secondary ion intensity is expected to follow a simple cosine law, in particular for randomly oriented polycrystalline surfaces. The explanation for this is that upon impact the collision cascade produces an isotropic distribution of the energy through the sample. Hence the intensity of collision cascades that arrive at the surface under an angle Θ with the surface normal varies as cos Θ. For single crystals, however, anisotropic emission may occur due to the channeling of the beam and the focusing of collisions along close-packed directions in the crystal and may lead to anisotropies in the secondary ion emission.

2.5 The Ionization Probability. – The formation of secondary ions is the most difficult feature in SIMS. While sputtering is relatively well understood, the process of sputter ionization is not, and a theory that describes the process of secondary ion formation satisfactorily does not yet exist. A number of trends can be rationalized, though.

For the formation of positive ions from initially neutral elements, there is a clear relation between the ionization probability, R±, and the ionization potential, I. Elements such as Na, K, Mg, Ca, which all have a low ionization potential, give high yields in positive SIMS, whereas elements with a high ionization potential, such as N, Pt, and Au, give low yields. A similar relation exists between the probability for the formation of negative ions and the electron affinity, εA: electronegative elements such as 0, F, and Cl give high intensities in negative SIMS. Also the noble metals Pt and Au usually appear much more intense in negative than in positive SIMS.

The ionization probabilities R± vary over some five orders across the elements in the periodic table. In addition, they vary also with the chemical environment of the element. This matrix effect makes quantification of SIMS spectra extremely difficult. As illustrated in Table 1, positive secondary ion yields from metal oxides are typically two orders of magnitude higher than those of the corresponding metals. A similar increase in yields from metals is observed after adsorption of gases such as oxygen or carbon monoxide.

There have been several attempts to develop models for secondary ion formation. The interested reader may consult the literature for reviews. Here we will briefly describe one model which accounts quantitatively for a number of observations on metals: the perturbation model of NØrskov and Lundqvist. It assumes that the formation of a secondary ion occurs just above the surface, immediately after emission. Then

• The probability for the formation of positive secondary ions increases when the ionization potential becomes smaller.

• When the secondary ion is close to the surface, there is a chance that it becomes neutralized by an electron from the surface of the metal. This process becomes more likely if the work function of the metal is lower. In other words, a high work function prevents neutralization and is favorable for positive ion emission.

• Both ion formation and neutralization just above the surface are more likely if (the normal component of) the velocity v of the departing particle is small, or, in other words, when its residence time in the interaction zone just above the surface is long.

These features are recognized in the following expression for the ionization probability:

[MATHEMATICAL EXPRESSION NOT REPRODUCIBLE IN ASCII] (2)

where

R+ = Ionization probability,

φ = Work function of the sample,

I = Ionization potential of the sputtered particle, and

v = Velocity of the sputtered particle.

Expression (2) accounts qualitatively for the observed variations of secondary ion yields with ionization potential. It also describes correctly that the yields of positive secondary ions from metals increase when the surface is covered by molecules which increase the work function, such as CO or oxygen. Although the model elegantly predicts a number of trends correctly and is conceptually useful, it is not detailed enough to be a basis for quantitative analysis of technical samples.

2.6 Emission of Molecular Clusters. – Molecular cluster ions are highly useful because they reveal which elements are in contact in the sample. Of course, this presupposes that such clusters are emitted intact and are not the result of recombination processes above the surface. The mechanism of molecular cluster formation has been described by two different models. The direct emission model assumes that the appearing clusters are emitted as a whole, i.e., atoms in a cluster were adjacent in the surface. In the recombination model, on the other hand, clusters are formed in the near-surf ace region by combination of sputtered atoms. In this model, then, atoms in a cluster were not necessarily neighbors in the surface. A number of experimental and theoretical studies has been devoted to this question.

Winograd et al. and Garrison used classical dynamics calculations to simulate the effect of primary ions impacting on adsorbate-covered single crystal surfaces, e.g., c(2×2)CO/Ni(001) and c(4×4)C6H6/Ni(OOl). Their simulations indicate that adsorbate molecules are to a large extent emitted intact. A small part of the adsorbed molecules may be fragmented, due to direct collisions with the primary ion or energetic substrate atoms. Clusters containing both adsorbate and substrate atoms, often more intense in SIMS spectra, are believed to be formed in the near-surface region by combination after emission of intact adsorbate molecules and substrate atoms. The probability that molecular fragments recombine is almost negligible.

(Continues…)Excerpted from Catalysis Volume 11 by James J. Spivey, Sanjay K. Agarwal. Copyright © 1994 The Royal Society of Chemistry. Excerpted by permission of The Royal Society of Chemistry.
All rights reserved. No part of this excerpt may be reproduced or reprinted without permission in writing from the publisher.
Excerpts are provided by Dial-A-Book Inc. solely for the personal use of visitors to this web site.