ering, UCLA, Los Angeles, CA
[c] Princeton Consultants, Princeton, NJ.
As dimensions in microelectronics shrink, new materials become important, and their interfaces with other
materials play an important role in the performance of the device. One example of this is the dielectric materials
used in capacitors for random access memories and as gate dielectrics for field effect transistors. In these
applications, materials with higher dielectric constant permit physical dimensions to shrink and circuit speeds to
increase. We have used x-ray photoelectron spectroscopy to study chemical defects in two of these materials,
silicon oxynitride and tantalum pentoxide. These chemical defects are related to electrical defects. Angular
dependence of photoelectron spectra permits determination of the distribution of nitrogen as a function of depth
through the oxynitride film. In addition, the N 1s photoelectron spectrum has four components. By studying the
adsorption of nitromethane on silicon, first principal calculations, near edge x-ray absorption spectroscopy, and
electron-spin resonance experiments these components can be assigned to nitrogen bonded to two oxygen atoms
and one silicon, nitrogen bonded to one oxygen and two silicon atoms, nitrogen bonded to three silicon atoms, and
nitrogen bonded to two silicon atoms with one dangling bond. It is the presence of this dangling bond that acts as
an electron trap in the oxynitride. In tantalum pentoxide it has been difficult to directly observe the chemical
defects that act as electrical defects. By apply an electrical potential to tantalum oxide film and collecting the
photoelectrons emitted through a thin Pt film, the potential distribution through the film can be determined. This
measurement shows that the electrical defects are largely localized near the electrode interface. Through examples
like these, we will show how chemical moieties can be related to the electrical properties of dielectric films.
8th International Conference on
Electronic Spectroscopy and Structure
Abstract #
Program#
105 1 2 0 2
Soft X-ray Absorption Spectroscopy at 25 nm Spatial Resolution
GREGORY DENBEAUX [a], ERIK ANDERSON [a], WEILUN CHAO [a], THOMAS EIMUELLER [b],
PETER FISCHER [b], LEWIS JOHNSON [c], MATTHIAS KOEHLER [d], MARK LE GROS [a], ANGELIC
LUCERO [a], DEIRDRE OLYNICK [a], DAVID ATTWOOD [a]
[a] Center for X-ray Optics, Lawrence Berkeley National Lab, Berkeley, California
[b] University of Wuerzberg, Wuerzberg, Germany
[c] Florida A&M University, Tallahassee, Florida
[d] University of Regensburg, Regensburg, Germany
Recent technical advances allow imaging with the XM-1 soft x-ray microscope with a spatial resolution of 25 nm
and a spectral resolution of E/dE = 700. We report on both the technical advances which make this possible and
some of the applications which can use these capabilities. The XM-1 is a full field imaging microscope which uses
zone plates for both the condenser and objective elements and is located at the Advanced Light Source. The
illumination energy can be tuned between 250 and 900 eV. The high resolution imaging capabilities are used in
biology, materials science, and environmental science. Magnetic materials are imaged at high spatial resolution by
x-ray magnetic circular dichroism. Element specific contributions to the magnetization can be obtained within a
sample even in the presence of a capping layer. High contrast, high spatial resolution images at the L-edges of Fe
and Co will be shown for magnetic storage material.
8th International Conference on
Electronic Spectroscopy and Structure
Abstract #
Program#
106 1 2 0 4
Photoelectron diffraction and holography: Present status and future directions
D.Phillip Woodruff
Physics Department, University of Warwick, Coventry CV4 7AL, UK
The development and exploitation of photoelectron diffraction as a means of obtaining structural information
about surfaces and ultra-thin films has grown considerably in the two decades or so since its original 'discovery',
and the pace of this expansion appears to be growing. Three factors have contributed this, namely the availability
of increasingly more and better synchrotron radiation beamlines, the appreciation of what can be achieved by the
technique through a series of successful applications, and the novel attraction of structure determination through
direct inversion of the experimental data, commonly referred to as 'holographic inversion'. In this brief review some
of the strengths and dangers to be identified in work so far, and foreseeable future developments and applications
will be presented.
Although the view of the process as photoelectron holography has attracted a great deal of interest, and there now
appear to be a number of alternative theoretical approaches for inverting the data in a fashion which removes the
most serious artefacts of the earliest schemes, it is important to recognise that such methods were only ever
intended as a means of obtaining first-order approximate structures which can be tested and refined using more
exact modelling. Such a first step is important in easing the problems of uniqueness which exist in pure
trial-and-error methods, but should never be a substitute for proper quantitative structure analysis, now
well-established using photoelectron diffraction. Nevertheless, the need for large data sets (in emission angle and
energy) in order to remove many of the artefacts of 'holographic inversion' is a valuable improvement relative to
quite a number of previously under-constrained structure 'determinations'. The holographic view of the process is
of limited relevance to photoelectron diffraction conducted at high electron kinetic energies where zero-order
(forward) scattering dominates, and simple first-order structural models can often be obtained by direct inspection
of the raw data.
A key strength of photoelectron diffraction is the potential to obtain structural information in a fashion which is
both element-specific and chemical-state-specific through the spectral fingerprint of the photoelectron core level
binding energy. The increasing availability of undulator beamlines on third-generation synchrotron radiation
sources offers the necessary combination of high photon flux and high spectral resolution needed to exploit this
potential fully, and this seems likely to be am major growth area in the near future.
8th International Conference on
Electronic Spectroscopy and Structure
Abstract #
Program#
107 1 2 1 1
Suppression of the low-spin multiplet components in the 3p photoelectron spectra of atomic and solid 3d
metals
J. E. Hansen[a], A. v. dem Borne[b], R. L. Johnson[b], B. Sonntag[b], M. Talkenberg[b], A. Werveyen[b], Ph.
Wernet[b], J. Schulz[b], Ch. Gerth[c], B. Obst[c], K. Tiedke[c], P. Zimmermann[c]
[a] Department of Physics and Astronomy, University of Amsterdam, Valckenierstraat 65, NL-1018 XE
Amsterdam, The Netherlands
[b] II Institut fur Experimentalphysik, Universitat Hamburg, Luruperchaussee 149, D-22761 Hamburg, Germany
[c] Institut fur Atomare and Analytische Physik, Technische Universitat Berlin, Hardenbergstr. 36, D-10623
Berlin, Germany
In recent years the 3p-photoelectron spectra of Cr, Mn, Fe and Co have been widely used to obtain information
about the electronic and magnetic structuire of metals, compounds, surfaces, thin films and multilayer systems.
Especially spin-resolved photoemission and linear and circular dichroism in angle-resolved photoemission have
proved to be powerful atom-specific probes. The interpretation of the experimental spectra has in most cases
relied on atomic models, usually some variant of the independent electron model, while different schemes, e.g.
cluster or ligand field models, have been used to take solid state effects into account. Due to the experimental
difficulties there are only a few 3p-photoelectron spectra known of free atoms and ions compared to the situation
for solids. The spectra of atomic Cr and Mn, metals which sublimate a relatively low temperature, were the first
to be studied. The 3p-3d interaction in the final ionic state gives rise to an LS multiplet structure spanning a
binding energy range of around 20 eV. Theoretically high-spin components at low binding energy contrasts
low-spin components at high binding energies. Experimentally strong lines at low binding energy contrasts some
weak structure at high energy but the experimental spectra published so far, in particular for metals, either suffer
from insufficient energy resolution or they do not cover the full energy range. In order to determine the multiplet
splitting as well as the strength and the width of the different components we studied the 3p-photoelectron
spectra of Cr, Mn, Fe and Co atoms experimentally and theoretically. On this basis we hope to be able to provide
a reliable basis for assessing the atomic effects in the solid state spectra. For comparison the 3p-photoelectron
spectra of Cr and Mn metals were also measured.
Term dependent lifetime broadening by super-Coster-Kronig decays turns out to be the main cause of the almost
complete suppression of the low-spin lines at high binding energy. The corresponding 3p-photoelectron spectra of
the metals only display a single broad asymmetric line which we associate with the high-spin components in the
atom. We notice the significance of this result for the interpretation of, for example, experiments involving
dichroism of 3d-metals.
We emphasize the similarities but in particular the significant differences in the interpretation of the 3p- compared
to the more well known 3s-photoelectron spectrum in these