The main body of this work explores ionic liquids (ILs) to be used as a magnetic
lubricant and sealant in a vacuum pump. The problem has arisen due to
upcoming legislation governed by the Stockholm Convention with respect to
persistent organic pollutants; where the current lubricant falls into this category
due to it’s fluorinated composition and it’s inability to decompose naturally.
Other factors include the current lubricant’s high viscosity and difficulty in
removing and cleaning the parts of the pump where the lubricant has come into
contact. Further to this, due to a gap in the current market it was hypothesized
that a magnetic lubricant could enhance performance by using a magnetic field
to hold the lubricant in the relevant parts of the pump, enabling less lubricant to
be used, whilst also being used to create a seal and in turn enhance
performance of the vacuum pump. In addition to this, a contribution to the IL
modelling field with respect to XPS is completed, where novel dicationic ILs are
modelled with respect to the C1s region using models derived and developed
from existing methods.

12 core samples based on ionic liquids with paramagnetic anions derived from
Fe3+ and Dy3+ and imidazolium or pyridinium-based cations are explored as
lubricants. Key tests were performed to begin collecting data to allow for
informed decisions on what samples may be more suitable for purpose.
Viscosity measurements were performed using a Brookfield DV-III digital
rheometer equipped with a small sample adapter at room temperature initially,
followed by ramped measurements up to 80 °C and then followed by viscosity
index (VI) calculations. VI calculations require a known density of the sample,
this was completed by weighing 1 mL of the sample at room temperature, this
was a fixed density where it is known density will decrease with a temperature
increase. Thermal stability was measured using thermogravimetric analysis on
a ramped rate moving from room temperature up to 450 °C at a rate of 10 °C
per minute. Contact angles were measured on three different substrates namely
steel aluminium and a fluoroelastomer comprised of hexafluoropropylene and
vinylidene fluoride. The method used a KrĂĽss DSA 25E contact angle
goniometer with a standard automated syringe dosing unit, applying the touch
on method where an average timed dynamic measurement was recorded.

Paramagnetism was measured using an Evans balance and friction behaviour.
Specifically, the coefficient of friction was measured using a Bruker UMT
Tribolab. This test made used of a ball of specified material applied at a
specified force onto a steel plate which reciprocated at a specified Hz with a
small amount of sample applied between the surfaces. Further to this, a surface
profilometer was employed to analyse the wear scar left on the steel plate after
the test had been completed using a Bruker DektakXT stylus profilometer.
Friction behaviour also entails analysis of the sample after being subjected to
friction testing to determine any chemical composition changes utilising X-Ray
Photoelectron Spectroscopy (XPS) and infrared (IR) spectroscopy.

At this point, trends relating chemical composition to physical properties have
began to be deduced; whereby in it’s simplest form extending the alkyl chain
component of the cation within the ionic liquid generally results in an increase in
viscosity as well as a decrease in density. With regards to VI, increasing the
alkyl chain length results in a decrease in VI. The same effect of alkyl chain
length on contact angle can be seen whereby a longer alkyl component results
in a decrease in contact angle for all surfaces investigated. The 12 core
samples all displayed a paramagnetic response to the Evans balance with the
Dysprosium samples showing a higher magnetic moment than the Iron-based
samples, an expected result.

Work involving coefficient of friction experiments also found that the core
samples gave lower COF values than the control lubricants, but XPS
discovered the short chain ionic liquid samples were potentially less stable than
their longer chain counterparts when compared to their original state based on
changes to their spectra for before and after COF studies, IR spectroscopy also
deduced subtle changes in composition.

The work finishes with one sample being scaled up, based on performances in
previous tests and tested in a live pump with mixed results. Despite performing
well as a lubricant relative to two control lubricants, compatibility issues arise
resulting in problematic changes in chemical and physical composition of the
scaled-up sample.

The remainder of the work covered troubleshooting the scaled-up sample by
utilising the same previous analytical tools such as XPS and IR but also
covered Mass Spectrometry (MS). This was used to determine potential
impurities present in the used sample and inform on potential degradation
pathways. This troubleshooting phase concluded that residual Fomblin (the
control lubricant used previously) was causing contamination when analysing
with XPS, further to this MS found peaks relating to a brand of antioxidants,
known as Irganox which is likely to be from its use as an additive in Fomblin.

Furthermore, an unexpected result in this work was the magnetic properties of
the samples were deemed not strong enough to fulfil their original objective; this
resulted in further work being completed to remedy the issue.

Initially, inspired by ferrofluids it was hypothesised that perhaps an ionic liquid
could be used as the carrier solvent for magnetic nanoparticles as an option to
boost magnetism based on magnetite, Fe3O4 following a co-precipitation
method. Confirmation of nanoparticle synthesis was completed using XRD with
peaks matching a reference sample taken from ICSD. However, difficulties were
experienced with relation to nanoparticle stability within the ionic liquid.
Surfactants were briefly explored based on oleic acid and decyl, methylimidazolium bromide with no positive results. Novel in-situ methods using an
ionic liquid as reaction media for the preparation of magnetite were explored
with preliminary dynamic light scattering (DLS) measurements suggesting
differences in size of nanoparticle relating to ionic liquid being used as the
medium.

Further novelty was beginning to be explored with respect to a new type of
poly(Ionic Liquid) based on a dysprosium anion. There were difficulties in
characterizing the new material which leaves room for future work, however
again the paramagnetic response despite being larger than the core samples
explored was still not deemed strong enough for manipulation with a typical
magnet.

In summary, this work has demonstrated that ionic liquids can be used as a
lubricant in a vacuum pump, purely as a lubricant as opposed to a lubricant and
sealant which was the aim of this project. The scaled up sample performed well on performance of the pump however clearly it had degraded over the 2000
hours that it had run for. The inability to be manipulated with a standard magnet
raised issues quickly with respect to utilising as a sealant within the pump.
However, these issues enabled further work to be explored including
nanoparticles and polymers and despite only being in the preliminary stages
offers optimism that both objectives could perhaps be achieved.

This work offers further insight into how ionic liquids may perform in large scale
experiments as opposed to bench top experiments in the laboratory. It has been
seen that COF measurements may not give the full picture of how ionic liquids
perform as lubricants as demonstrated by the unanticipated degradation when
in the live test pump. But with respect to the test pump, this is a great
contribution to the field where it is believed this type of experiment has not been
reported in the literature to date.