"Joint" studies in Surface Analysis
British Aerospace, Bristol Wednesday 5th July 1996
The Group was welcomed by Dan Kells, the Director of
Sowerby Research Centre.
Dave Dixon (BAe, Bristol) opened the morning session with a talk
describing operation of joint projects at BAe and the benefits accrued. The main reason
for BAe's involvement with joint projects is to obtain additional funding and there may be
twenty or so running at any one time. In most cases the funds come from the European
Commission via one of many schemes to encourage technological development in strategic
areas - aerospace being a major one
The EC is not just a "pot of gold", it has very strict
conditions and published guidelines for collaborative projects and those proposals
submitted must be seen to be well thought out, well planned, well managed and well
resourced. Of ~1100 proposals for Brite-Euram funding 20% were rejected due to incorrect
presentation, 60% were rejected after consideration and 20% were funded. Progress against
milestones must be reported at specified intervals and it is the project chairpersons
responsibility to ensure that the contractors know what is expected and when. The EC
audits the project and treats the final programme as a contract. Each proposal is assessed
for clear, verifiable and realistic technical aims, a significant socio-economic impact,
its financial credibility and technical merit. It is imperative that each project has a
potential for commercialisation.
Most industrial Brite-Euram projects involve a minimum of 10 man years,
spread over 2-4 years and involve at least two countries. The industry gets 50% funding
and any universities involved get 100% of their marginal costs. For basic research
projects in the 0.5-1 MECU cost range, industry gets 15% of its costs.
The results of a collaborative project are always a compromise.
Although it is essential to work within the rules of the EC, the advantages of seeing what
other industries do and how they work and the improvement in European scientific cohesion
makes these projects very rewarding. For small companies, that can't afford to devote the
time necessary to submit and manage these projects, there are specific schemes available
and there are also firms of consultants who can be engaged to manage and monitor the
operation of the project.
Dave warned of the potential for "breakdowns in
communications" as the main reason why aspects of a project fail but, in general, the
work was highly successful. The EC has carried out a cost: benefit analysis on the
projects funded and has found that the C/B ratio is ~7:1. It also expects that it may be
necessary to inject extra funds to see the development through to commercial exploitation.
Alison Crossley (AEA Technology, Harwell)
discussed some medical and electrical applications of "joints" but pointed out
that all the work done at AEA is carried out as joint projects because AEA technology has
been set up as an independent contract research company. Alison has been involved with
industrial sponsors, universities and various DTI, DOE and MOD groups. Currently she is
associated with an ESPRIT funded project on scanning probe microscopy which is aimed at
developing new electronic materials and micro- systems technology. AEA Technology also
sets up R & D club activities and will manage European Commission sponsored work
programmes on behalf of the participants.
Surface analysis and other techniques are used in collaborative studies
of coatings and interfaces for tribological applications in orthopaedic implants such as
hips and knees. There are 50k hips and 20k knee replacement operations per annum in
Britain alone and the average lifetime of the joint is 10-15 years which is exactly the
same as it was 30 years ago. The main technical reason for failure is corrosion and
erosion (at ~0.1mm/year) as the metal spike rubs in an ultra high molecular weight polymer
cup (to ensure a low friction coefficient). There are also biological problems with the
adhesive used to bond the spike to the bone. Modern methods that are being examined to
improve the lifetime encompass (1) use of carbon fibre polymer composite systems for the
spike, (2) ion implanted metals, TiN or diamond like carbon (DLC) for the wearing surface
and (3) cementless implants where compounds of hydroxyapatite and phosphates are used to
encourage the bone to grow into the implant.
Studies of electrical joints are carried out to improve stability,
adhesion, ohmic resistance behaviour and compatibility with micro-patterning techniques.
For high temperature contacts bonded to DLC, wide band gap semi-conductors are required.
Collaborative studies using XRD and RBS are used to characterise the layer structure.
To provide ohmic contact, refractory metals are used which form
carbides with the DLC. Titanium gives good carbide formation and the interfacial oxide can
be made to dissolve in the metal such that stability is within 5% after 5000 hours at
800C. However, titanium will diffuse through a gold contact layer. This can lead to
undesirable electrical characteristics and so a platinum barrier layer has to be employed.
For molybdenum, AES studies show that oxide causes weakening of the joints. Collaborative
effort and surface analysis all play a part in understanding the interfacial properties
and reactions that occur.
Steve Harris (BAe, Bristol) described the
vast array of "sticky topics" that the Sowerby Research Centre is set up to
support within BAe's operations. Steve's group works on adhesion, diffusion bonding,
welding, brazing, composites, sealants and electronic devices as part of internal
partnerships within the BAe operating companies. They also sponsor fundamental work in
universities. They have a Kratos XSAM800 and a VG ESCALAB MkII fitted with various
fracture stages (bending and peeling) and a gas cell in which they can cure composites and
paints under controlled conditions. The in-situ fracture stages are considered essential
because they reduce confusion in interpretation resulting from post fracture reaction and
contamination which occurs after ex-situ fracture.
As an example of the work on sealants, Steve described how XPS
determined that a low peel strength failure from a polysulphone - aluminium bond was a
cohesive failure in a weak boundary layer. SEM showed that this was due to the presence of
an abrupt interface whereas high strength bonds are characterised by rough interfaces
(ensured by suitable pre-etching) where a substantial amount of mechanical interlocking is
also possible.
A contrasting failure mode was found for a multi-ply polymer system
where a fluorocarbon de-bulking agent was incorporated at a very low level. Unfortunately,
even at this low level, XPS showed that it had segregated to specific areas in the
composite and, under mechanical stress, a crack propagated by following the line of least
resistance through the segregated regions.
XPS and AES analysis of porous titanium welds was used to illustrate
that the accepted wisdom is not always correct. The porosity was "known" to be
due to hydrogen incorporation but surface analysis not only detected impurities but also
found evidence that residual organic material was a more likely cause of the problem.
Steve's group have recently begun to use molecular modeling techniques
to support the experimental studies. This is with a longer term view to validating the
modeling approach and reaching the stage where it can be used to provide useful insights
into the interactions in interfacial adhesion.
John Watts (University of Surrey) discussed
collaborative studies that his group at Surrey had carried out on fundamental aspects of
adhesion in conjunction with industrial sponsors. There is a difference in approach
between engineers who like to couple the failure and degradation effects in their tests
and the surface science approach which usually tries to separate them to aid
interpretation of the failure mechanism. Commercial adhesives are multi-component
formulations and the first chemical transformation stage occurs during curing. This leads
to the initial interfacial chemistry which is further modified during environmental
testing.
Using an example of an acrylic photo-cured micro-electronics
encapsulant, John showed how a combination of XPS and SIMS was necessary to convince the
formulators that an aliphatic diluent was segregating to the resin - ceramic interface and
causing susceptibility to attack by water in humidity tests. John had used molecular
modeling techniques to indicate that a monolayer of this type of material should adhere to
the surface such that the molecules were tilted at 30 to the vertical and this would
produce the 10nm thick surface layer that was found by XPS electron take-off angle
studies. The water susceptibility was eliminated when this component was removed from the
formulation but, unfortunately, it turned out to be a necessary viscosity improver which
was essential in other applications.
John gave a second example of cooperation with industrial sponsors
which involved resin curing for coatings applications. In this case, the formulators were
using PF6 as an initiator. In SIMS studies the ratio of F- to PF6 fragments was shown to
be an indicator of how far the curing had proceeded. Thus, the surface analysis results
could be fed back to the formulators to indicate potential product modification which may
lead to faster curing.
Finally, John described his philosophy of using XPS to obtain
adsorption isotherms for the ingredients in a formulation and deriving the heats of
adsorption. This data can indicate which ingredients are most likely to segregate to
surfaces and reveal potential problems with the ultimate bond strength and susceptibility
to degradation.
Gary Critchlow (ISST, Loughborough) described
work on conversion coatings for aluminium, carried out in conjunction with industrial
sponsors. Phosphates and chromates have been used successfully for many years but
environmental considerations are forcing a search for equally effective alternatives.
Gary showed data which confirmed that the surface pre-treatment of
aluminium had a dramatic impact on bond durability. Whilst a simple degrease gave a 70
hour time to failure, a combination of grit blasting, anodising and application of an
adhesion promoter increased this to beyond 2000 hours when the test was stopped. The
treatments are designed to produce surfaces that are micro-porous, clean, wettable and
hydration resistant. AES, XPS and electron microscopies are required to characterise these
properties of the substrate.
Using an example from the automotive industry where bonded aluminium
joints are required to remain intact during a collision in order to absorb the energy of
the impact, Gary showed that the surface magnesium oxide layer was first removed and a
40nm thick layer of phosphate was deposited as a nodular layer. There was an optimum
thickness because thicker layers failed cohesively under impact using a variable mass
pendulum test.
Results of studies using a zirconium based "green"
alternative pre-treatment revealed that it is a ZrO2 film that resists hydration and the
ideal topography occurs at 50nm film thickness. In tests at low load the treatment gives
dramatic differences although at high loads these are less significant, probably due to
creep in the layer.
Gordon Tatlock (IRC, Liverpool) discussed
work on the adhesion of native oxides on Fe20Cr5Al alloys. In general, these materials
grow a porous Al2O3 oxide scale at elevated temperature but the presence of sulphur,
carbon and yttrium impurities can have a severe effect on the adherence of the scale. To
investigate this problem, Gordon used FEM AES to analyse specially prepared high purity
alloys free from these elements or with deliberate addition of them. In the presence of
sulphur and carbon the scale morphology is convoluted and there are tunnels beneath the
scale which appear to follow the grain boundaries of the alloy. This is clearly an
indication of a weakly adherent scale. With addition of yttrium, at the 0.1wt% level, the
scale is more uniform and apparently well bonded to the substrate.
Using thin samples which can be bent easily in the vacuum system using
a wobble stick, Gordon was able to see that even the yttrium doped scale detached
preferentially near to the grain boundaries. By contrast, the commercial grade alloys
broke completely. AES analysis showed that the detachment seemed to be associated with
sulphur or carbon in the large grain boundaries formed as a result of annealing. It was
possible to map the apparent amounts of these impurities at the grain boundaries. However,
this raised a question regarding quantification when applied to individual single crystal
grains where channeling of the incident beam and diffraction of the emitted beam can give
rise to preferential signal enhancement phenomena. It is known for example that a
sub-monolayer of sulphur can dramatically alter the signal from a nickel substrate
dependent upon the angles subtended in the spectrometer.
The work is continuing but Gordon warned of the difficulties encountered in very high
resolution analysis of individual grains in an alloy.
Simon Church (IAC, Bristol) discussed his
work on diffusion bonding of AlLi 8090 alloy using a novel development of a tensile
fracture stage where it is used in reverse to apply a constant compressive load at
elevated temperature, typically 66% of the melting point. The temperature was raised by
placing the sample at one focus of twin elliptical mirrors and a 2kW halogen lamp at the
other. The advantage of diffusion bonding is that it can be used to bond complex shapes
with low deformation. Titanium bonds well by this method because it is capable of
dissolving its own native surface oxide. However, for aluminium alloys, the native oxide
has to be removed chemically or mechanically, by polishing. Polishing is not practical for
a commercial operation but Simon has compared the behaviour of chemical etching and
polishing in laboratory scale experiments.
Chemical etching produces porous surfaces compared to polishing. SIMS
detects Li, Na, Al, K, O Cl with a significant amount of F on the chemically etched
surface but much less on the polished alloy. XPS shows that the oxide remaining on the
treated alloys is initially thin enough to detect the metallic aluminium in both cases.
Based on a set of 60 compressive shear test samples to expose the
interfaces, Simon used AES to show that the oxide on a properly bonded alloy had a
granular structure with a mixed Al, Mg and Li oxide. In contrast, a poorly bonded
interface had grown a ceramic scale which was almost entirely Al2O3. Contamination
elements Li, S, O, F, Cu, Mg and Al were detected beneath the scale. Simon's general
conclusions were that, although chemical etching removes the native oxide it also
introduces contamination. Joints formed at low deformation allow air to enter along the
interface and cause oxide growth. At high deformation pressures this is not such a
problem. The art is to utilise the correct pressure in a production line environment.
Frank Jones (University of Sheffield)
described the problem of obtaining convincing evidence that the chemistry of carbon fibre
surfaces ranked equally in importance alongside mechanical keying in defining the strength
of composite materials. His approach was to use plasma treatment methods to introduce
functional groups to alter the surface chemistry without changing the physical surface
roughness or to deliberately smooth the surface by using a non-functionalised coatings.
The method involves plasma processing the fibres in a mixture of functional and
non-functional monomers, e.g. acrylic acid and hexane or an ally-amine and octadiene. The
ratio of the monomers defines the surface produced which was characterised using XPS. This
showed that surface acid groups predominated at low power but hydroxyl groups at high
power.
To correlate the surface chemistry with the bond strength, a single
fibre fragmentation test was used which produces a mean length of fibre fragment that is
related to the tensile load that the fibre-resin bond can support compared to the fibre
itself. Frank found that part of the misunderstanding about the effect of surface
treatments is due to the presence of inherent surface functionality on the fibres as
manufactured. It then became possible to treat fibres so that the composite strength was
reduced compared to the "as received" fibres. Amine and acid groups were then
particularly good at improving the bond strength whereas hydroxyl groups were of limited
value.
Applying finite element analysis methods to the composite has enabled
Frank's group to propose a better method of characterising adhesion by use of the
Composite Stress Transfer Function. The work has not only confirmed the importance of
chemistry in fibre reinforced composites but has also highlighted the possibility of using
plasma treatment methods to create graded bonds at the fibre-resin interface in the
composite.
Simon Morton (University of Surrey) gave an
update on the status of the UK ESCA User Group web site which can be found at:-
http://www.surrey.ac.uk/MSE/ESCA/ESCA/home.html
(the capital letters must be entered as shown).
This site is now the world's largest and most comprehensive web site
dedicated to surface analysis. After only a very short time Simon and Chris Walker
(University of Ulster) have produced an extremely professional, informative and
technically useful facility which all members are urged to inspect in their own time.
Meeting reports and future meeting information and registration forms will be available on
the page to download directly.
Recent upgrades include elemental data from reference books by kind
permission of the publishers. There is database of Auger parameters and other element
specific information. The development of the Web site has acquired its own momentum and it
is worth logging on fairly frequently to see what's new. The membership list (names and
affiliations only) has been included so that individuals can make sure they are on the
list or notify Simon if they or a colleague are no longer in the field and do not wish to
receive our mailshots.
Simon ended his presentation with a request for members to inform him
of information that they would like to see on the site. All will be done to include items
that are generally useful and do not infringe copyright.
|
