|
National Physical Laboratories - Wednesday 15th
July 1998
| |
 Figure 1. Delegates to the UK Surface Analysis Forum summer meeting at NPL |
 |
| |
|
| |
Figure 2. Bob Wild presenting Dr.
M. P. Seah with a prize to mark the release of the first ISO standard in Surface Analysis,
"The Data Transfer File Format" |
| |
|
The NPL meeting was unique in the history of the UK Surface Analysis Forum,
in that the afternoon session involved a hands on workshop on the use of
the NPL spreadsheet developed by Dr Peter Cumpson. The unusual nature of the meeting
resulted in a larger than usual number of delegates attending the meeting.
The meeting started promptly at 10 am, with a short presentation by Dr
Lea on the work of the NPL group. Dr Wild then stopped the meeting to present Dr Martin
Seah (NPL) with a prize from the group, to acknowledge his outstanding efforts in
developing the first ISO standard in the field of Surface Analysis, the Data Transfer
Format.
"Composition Depth Profile Data by
ARXPS"
Dr P Cumpson, NPL
|
|
| Figure 3. Dr. Peter Cumpson (NPL) giving the first paper at
the UK Surface Analysis Forum Meeting at NPL |
|
| |
|
Dr Peter Cumpson (NPL) presented the first Paper on the use of ARXPS,
its application and its limitations. Dr Cumpson, defined the research programme on ARXPS
at NPL and charted the development of the spreadsheet programme which was evaluated in the
afternoon workshop.
Peter went through the advantages and disadvantages of ARXPS
measurements, namely that for most instruments it is possible to perform the experiment,
but difficult to do the subsequent analysis. Over the 20 years in which the technique has
been used, 20 different models have been developed in order to construct surface profiles.
Most of these models only really work on selected systems. Peter has tried to take the
best models and insert them into a user-friendly interface in order to make the technique
more accessible to non-experts. Dr Cumpson then showed some basic examples, and
demonstrated how the software would handle these types of system. From their studies, NPL
believe that they have developed a basic “toolkit” for the surface analyst which
will make the data obtained from ARXPS measurements, more accurate and reliable.
Some Industrial Analytical Case Studies: A Menu
of Approaches Including Angle-Resolved Methods
Dr G Smith, Shell
Dr Smith described the industrial application of ARXPS and the method
of analysis currently adopted by his Research group at Shell, Thornton Laboratories. Dr
Smith initially described the basic overlayer model developed by Fadely in 1984 (CS
Fadely, Progr.Surf.Sci., 16(1984),3), and pointed out some of the basic limitations in
this approach, specifically what value of l to use? Where:
DA(q ) = l
A.cosq .ln[1+ (IA.oIB)/(IB.oIA)]
For a two layer model system
|
 |
|
Figure 4. Dr.
Graham Smith from Shell |
| |
|
At Shell they use the attenuation depth rather than the Inelastic Mean
Free Path (IMFP), as defined in the Seah and Cumpson paper (SIA 1997). Dr Smith showed
some basic examples initially SiOx on Si and subsequently the absorption of lubricant
layers on iron oxide (i.e. Lubricant/FeOx/Fe System). This particular system which is of
significant importance to Shell, has a background problem making the analysis more
difficult than usual. Most commercial packages would fit a Shirley background to the data,
but using the QUASES software it was possible to fit a Tougaard background to the data.
The different backgrounds gave different lubricant film thicknesses, of 3.74nm (Shirley)
and 3.9nm (Touggard), the latter corresponding to the values of film thickness determined
using other methods of analysis.
For further information on the data given in Dr Smith’s talk
please see the original overheads
(860k) or read the relevant sections in “Quantitative Surface Analysis for
Materials Science", GC Smith, Institute of Materials, 1991. ISBN 0-901462-79-9.
Application of QUASES for the non-destructive depth
profiling of implanted SiO2 films in Si
Prof. M Prutton, University of York
|
|
| Figure 5 Prof M Prutton presenting his
ARAES paper at the NPL Meeting. |
|
| |
|
Professor Prutton showed the analysis of a specially developed
reference sample generated by Professor Mark Dowsett’s group at Warick University,
using ion beam implantation. Using an SiO2 reference sample and the QUASES software, they
were able to predict the film thickness using the OKLL Auger line and compare the results
with measurements made using EDX, plus the thickness determined using implantation data
supplied by Professor Dowsett. The results were all in good agreement, but the data
analysis using the QUASES software took 4 hours, which was considerably longer than the
acquisition time!
Prof. JE Castle, Surrey University
|
|
|
Figure 6
Prof. Castle at the start of his presentation on Curve Fitting. |
| |
|
The work looked at the intrinsic tail in the XPS spectrum, which is
normally ignored by standard commercial packages since they tend to use a pre-determined
Shirley background. The group at Surrey University have looked at this addition to the
background in the spectrum, for a wide number of different elements as thin overlayers and
linked the tail effect to their Tougaard background subtraction method. Professor Castle
showed the analysis of Ag thin films using the Surrey University peak-resolving package,
which uses Voigt functions. A real life example was then studied, namely the presence of a
passive film on the surface of a stainless steel. Using the Surrey peak fitting software,
an accurate assessment of the Fe2p peaks could be made, even given the steeply rising
background. The entire Fe2p region could be realistically accounted for in the peak
fitting model, which is not possible with commercial packages. A further advantage of the
software package is that only a 20 eV window either side of the main lines is required for
the Tougaard background subtraction routine, most packages require a spectral window of
~100eV away from the main lines.
Dr. L Hazell, CSMA
 |
|
Figure 7. Dr Len Hazell at the
beginning of his battle with the NPL visual "aids" facilities |
|
| |
|
Len Hazell went through the standard method of data analysis done
by most analysts. He then pointed out that nearly all of us “assume” that the
surface is homogeneous within the sample volume analysed, however, we all know that in
most cases this just is not true. So why do we do it? The answer is that it is just too
difficult to construct a more accurate description of the surface even if we have the
data. In order to describe the composition of the surface in more detail, we need to vary
either q or l (see Dr Smith’s paper above), and in general we alter q, if the sample
substrate is flat. Dr Hazell then described the interactive modeling programme developed
by BP/CSMA, which is used at CSMA to analyse ARXPS data. The software is interactive and
allows the rapid assessment of the data, which is inputted via an EXCEL spreadsheet. The
model does require some prior knowledge of the likely structure, including estimates of
the possible stoichiometry of the layers and their density. The most important input
parameters included:
 |
Common Sense
|
|
Prior Knowledge
|
|
Patience
|
|
Inspiration, &
|
|
Persperation!
|
The Manufacturers Latest Innovations making ARXPS a
Practicable Method of Analysis
VG have recently launched the SProbe,
which can be used to perform ARXPS measurements simply by altering the input lens
voltages. This means that no stage/sample movement is required and that no refocusing of
the X-ray spot is needed. Also the charge compensation only has to be done once, not for
each take-off angle as with older monochromator systems.
PHI Presentation (Dr Uli Rolli)
Dr Rolli showed the basic geometry of the Quantum 2000 spectrometer.
The instrument uses a fully compensated automatic stage, which can be programmed to do
ARXPS measurements, with the sample always focussed in the X-ray spot. The collection
angle is set via a mechanical aperture on the analyser input lens.
Kratos Presentation (Dr C Blomfield)
The latest magnetic lens systems produced by Kratos, has an in-built
charge neutraliser which produces a constant flux of electrons focussed on the sample
surface. No adjustment is needed when the sample take-off angle is altered since the
system automatically compensates. Also running the instrument in the electrostatic mode,
the instrument can adjust its input lens voltages automatically and collect ARXPS data
without moving the specimen.
Workshops
XPS Datasystems: "What we analysts really need is ..."
Workshops run by:-
Dr A Carley (University of Wales (Cardiff))
&
Dr J Day (Bristol University)
Leader : AF Carley
|
 |
|
Figure 8 Dr
Carley hard at work in the first workshop. |
| |
|
This workshop focussed on the acquisition and processing aspects of
datasystems, particularly in the context of quantification and precision. Although data
acquisition is an apparently straightforward aspect of a datasystem which should have been
sorted out by now, in fact this topic occupied most of the discussion. Turnaround times
can be minimised by means of the Harrison-Hazell acquisition strategy (1992) which
estimates the optimum number of scans required for a particular precision, but no
commercial datasystem incorporates it. Other suggestions were for real-time trial
quantification of survey scans, interleaved scanning and more flexible interaction with
the process during acquisition. The ISO Standard Data Transfer Format, implemented on all
commercial datasystems, meant that users were not limited to one system for processing,
and indeed a show of hands indicated that most users did not confine themselves to a
single system for the complete acquisition to quantification sequence. Statistical errors
arising from intensity measurement and curve-fitting, which can be estimated using
published algorithms, were generally ignored, and were apparently not demanded by
customers. Systematic errors in the quantification process were much more difficult to
address. A consensus developed in the workshop that the only way to make manufacturers
aware of the failings of their datasystems, and take our suggestions seriously, was for
the UK Surface Analysis Forum to produce a synthesis of the member’s views and submit this
to the manufacturers. Members will be canvassed for their suggestions in due course.
Leaders : John Day and Keith Hallam
The intention of this workshop was to identify the needs of users and
to build on the experience of previous systems to identify future requirements. The
workshops were introduced by describing what currently existed and addressed the hardware,
data acquisition, manipulation, quantification and presentation. There was some discussion
regarding the platform used, from PDP11 to PC but the main complaint here was that
manufacturers had changed their platforms and their software but not necessarily improved
it. Data acquisition must be able to acquire spectra, depth profile, angle resolve and
map. The hope was expressed that acquisition could become more intelligent. Within data
manipulation, the various forms of smoothing, background subtraction and peak synthesis
were discussed. Within presentation and storage it was felt that software systems could
more easily export data to spreadsheets and word processing packages and that interchange
of data had still some way to go. The adoption of the VAMAS data transfer format
recognised the need for this but work must now start on producing a better transfer
format. Quantification was also discussed. Most people now prefer to quantify using
absolute data from fitted peaks. It was concluded that there was still a long way to go
before we have an ideal data system.
Hands-on computer-based ARXPS workshop
Workshop run by Dr P Cumpson
 |
|
 |
| Figure
9 Several "Senior" surface analysts who are discovering that they know less than
they thought they did! |
|
Figure
10 Even experienced users benefited from the new software! |
The practical workshops were very well attended and were
a major success. The software was shown to be very robust and easy to use, although this
varied depending upon the package used. The general view was that the product was
excellent and when will it be available to the members of the user group?
Dr RK Wild thanked the members for attending the meeting, and thanked
all the speakers in the last year for their excellent presentations. He noted that numbers
attending the meeting were increasing, with the last two meetings having over 100
delegates in both cases. This was a record for the group and justified the slight change
in structure of the meetings, where practical workshops were included.
Dr Kathy England the UK ESCA User groups treasurer then gave a brief report on the
financial position of the group, which from the current balance sheet was shown to be very
good. Both of the last two meetings should have made substantial profits, which enabled
the group to keep the meeting fees to a minimum.
The secretary’s report was given by Dr Steven Harris and praised
the continuing efforts of Dr Simon Morton in taking the standard report and converting the
document into a glamorous Web page. He also pointed out that the official membership list
had now risen to ~390, of which ~65 were retired members. The last two meetings had shown
a dramatic increase in new younger members who were essential if the group is to continue
to flourish in the year 2000 and beyond!
Elections
|
 |
|
Figure 11.
Steve Evans thanking the Group for awarding him the John Riviere Prize at the
Bristol UK ESCA User Group meeting |
| |
|
|
|
The following were elected to the main committeee:-
|
Dr. RK Wild, Chairman |
|
Dr. SJ Harris, Secretary |
|
Dr. K England, Treasurer |
Dr Wild announced at the end of the meeting that this would be the last
time that he would stand as Chairman and that a new Chairman would be appointed in two
years time. The final part of the meeting, was a brief thank you from Dr Steven Evans who
was awarded the John Riviere Prize at the Bristol Meeting, he concluded his speech by
asking the Chairman if he would be kind enough to sign the certificate this time!
Acknowledgements
The committee would like to thank Dr Alan Carrick, of Acolyte Science for supplying the photographs of the
delegates to the NPL meeting. However, the committee are not offering a prize to the first
member who correctly identifies the two distinguished members who insisted on being in
most of the shots! |
