A FASTER approach to air analysis – Using GC×GC with BenchTOF to analyse exhaust particulates
The health and environmental impact of particles released from vehicle exhaust have been reckoned to result in an annual cost to the European Union of over 600 billion euros. We talk to Dr Salim Alam at the University of Birmingham, to find out about how his work using Markes’ TOF MS equipment will feed into predictive models of pollution and so tackle this major threat to health.
You’re currently involved in the ‘FASTER’ project. What’s it all about?
‘FASTER’ is funded by the EU, and has the full title ‘Fundamental studies of the sources, properties and environmental behaviour of exhaust nanoparticles from road vehicles’. It’s all being conducted at the University of Birmingham, and is being led by Professor Roy Harrison from the School of Geography, Earth and Environmental Sciences.
What does the project aim to do?
It’s looking at exhaust-derived particles in the atmosphere, and in particular the semi-volatile organic compounds (SVOCs) they contain. Until now, there has been insufficient knowledge of the properties of these components, meaning that atmospheric models ignore or over-simplify their contribution. We hope to generate better-quality information about these compounds, which will feed into more accurate models of urban pollution.
What’s so important about SVOCs?
The particles present in diesel exhaust contain a lot of SVOCs, resulting from the incomplete burning of fuel. Although these start off being bound to the particles, they soon evaporate, and under ambient conditions undergo chemical changes due to the action of sunlight and atmospheric oxidants, such as ozone. These reactions drive the formation of so-called ‘secondary organic aerosol’ or SOA – poorly-understood particulate matter that’s been implicated in a wide range of health and environmental impacts.
Why have these aerosols remained poorly understood?
Basically, it’s because the tools we’ve been using to study them haven’t been up to the job. Traditional gas chromatography (GC) methods fail to resolve the SVOC components of these aerosols, giving what we analysts euphemistically term an ‘unresolved complex mixture’ – an ugly hump in the chromatogram. Such lack of resolution means that key components cannot be identified, which in turn means that the chemical and physical process cannot be reliably modelled. So we were on the look-out for something new.
How did you come to choose Markes’ equipment for the FASTER project?
We had used GC with quadrupole mass spectrometry (MS) in the past, but found that with it, we could only monitor a restricted number of target species – such as polycyclic aromatic hydrocarbons (PAHs) and n-alkanes. However, we’d heard about Steve Rowland’s groundbreaking work on petrochemicals at Plymouth University, and so asked him to run some samples for us. Using his two-dimensional (GC×GC) setup with Markes’ BenchTOF time-of-flight mass spectrometer, we found that we could detect over 10 times as many compounds as before – so we decided to purchase a BenchTOF for ourselves!,/p>
What samples have you been running?
We look at three sample types – exhaust fumes generated in our engine testing facility here at Birmingham, ‘real’ air samples collected in urban locations, and a range of fuels and engine oils. The exhaust samples are first diluted to reduce the concentration, and then passed through a 13-stage filter setup, which collects particles ranging in size from 10 nm to 10 µm in a single run. Each filter is extracted with solvent and injected directly into the GC×GC–BenchTOF system. The atmospheric samples are treated in exactly the same way, except they’re not diluted.
And I hear you keep your analytical system busy…
We certainly do! When running the exhaust samples, we can carry out 18 tests per week, consisting of six different conditions run in triplicate. Considering that there are 13 filters to be analysed in each run, our BenchTOF could typically be running up to 90 samples per week!
How do your team find that workload?
We’ve definitely been kept very busy since the system was installed in late 2013, but things have been running very smoothly. It’s a big plus that both of our Ph.D. students have found the BenchTOF and the software very easy to use, despite neither of them having a background in analytical chemistry.
How important is analytical sensitivity for you?
Sensitivity is a fundamental feature of the system, and a big reason why we’ve been so pleased with it. This point was driven home when analysing a supposedly ‘blank’ sample, which had previously showed a very small number of compounds when run on a quadrupole system. We found a huge number of compounds in it when using our Bench TOF!
And you’ve added Select-eV to your toolkit too?
That’s right. GC×GC with BenchTOF was great for separating and identifying many of the components of the ‘unresolved complex mixture’, but even this approach struggled with some of the alkane isomers. These compounds have such similar spectra that we needed to upgrade to Markes’ Select-eV variable-energy ionisation technology to improve confidence in identification and get the most out of each analytical run.
What specifically does Select-eV allow you to do?
We’re able to identify individual alkane components rather than adopt the previous ‘group’-type approach to identification. In particular, at low ionisation energies, we’ve been getting increased intensity from the molecular ion, and from the ions that indicate the branching position of the alkanes. These ions are crucial to speciating some of the isomeric alkanes.
Overall, what has the combination of GC×GC with BenchTOF and Select-eV allowed you to achieve?
Our biggest achievement to date is being able to shine a light on the composition of the ‘unresolved complex mixture’, which so many analysts looking at SOA are unable to investigate. We’ve been finding a vast diversity of chemicals in these aerosols, with everything from phthalate plasticisers to fire retardants, as well as the usual suspects such as PAHs and alkanes. It seems that there are plenty of other chemicals in exhaust aerosol, which although individually may not give cause for concern, collectively may pose a significant hazard to health.
These sound like important findings. What plans do you have for publishing this work?
We’ve had one paper out already – in the journal Atmospheric Environment – which was published after analysing just five samples, in collaboration with Steve Rowland’s group. We’ve also got four more papers in the pipeline, and believe that this information will prove really valuable when it comes to understanding how exhaust emissions affect the atmosphere, and to develop better numerical models for what’s happening.
And what’s next for your research?
Using our GC×GC–BenchTOF set up with our newly acquired UNITY thermal desorber, we’re hoping to shed more light on the SVOC composition of engine emissions. We’re also going to be taking our sampling further afield, to and plan to take air samples from London and locations overseas. We’re particularly interested in seeing how the aerosol composition varies between places that use different types of fuel – developing countries in particular tend to use fuel with a higher sulphur content. We hope that the information we’ll uncover will ultimately make it easier for urban planners and regulators to reduce the negative health impacts of living in our cities.
Dr Salim Alam, Birmingham University
Salim Alam completed his Ph.D. in Atmospheric Science at the
University of Birmingham in 2011. His research looked at the reactions of
alkenes and ozone, and involved some time at the European Photoreactor
(EUPHORE) smog simulation chamber in Valencia, Spain. Now, as a Research Fellow
at the University of Birmingham’s School of Geography, Earth and Environmental
Sciences, he’s continuing his interest in understanding the sources, properties
and mechanisms of key atmospheric pollutants, their degradation products, and
their interactions in the troposphere.