05th August 2022 | Author: Matt Hiscock
I’ve been working with particle analysis at Oxford Instruments for a good few years now and, one of the things that has become clear to me is that there are certain commonalities in the analysis that people perform. From almost all of the customers that I have spoken to, the same themes emerge – we need a particle approach that has three key attributes:
We need the system to be capable – it must be able to do the job that we need it to do. That means it must be able to detect particles accurately and repeatably, it must be able to analyse them and give consistent results, and it must be usable by a range of different operators. Ideally, it will also have a range of functionalities that will help the user to work efficiently and to focus their efforts and analytical resources where it is needed most. All in all, it needs to be Smart.
I have yet to find a case where throughput is not important for particle analysis. Particle analysis is an approach where we can potentially be analysing hundreds of thousands of particles in a single analytical run over large areas (well, large for a microscope!). This means that we have to employ high degrees of automation and be able to optimise all of the parameters that are relevant to us as a small inefficiency in one step of the analysis can become significant when multiplied up over 1000s of fields of view and 100s of thousands of particles. So ultimately, our system needs to be Fast.
Of course, there is no point in doing any analysis unless the results that we get are representative of what is actually going on in our sample. In particle analysis, we need to be particularly trusting of the algorithms that take care of things like element identification and composition quantification as, given the huge numbers of particles that we analyse, it is impractical for us to manually review each one and check that we agree with the automatic processing that has been applied. As such, particle analysis needs to be accurate too.
Ease of use - We have tried to bear all these requirements in mind as we created our particle analysis platform, AZtecFeature. We took all the learnings from previous generations of particle analysis and then re-worked them to create a workflow that can be easily learnt by new operators of the system. We also added in-interface instructions to help explain what functions do what, and why – these can be incorporated into SOPs to ensure repeatable analyses are performed.
Smart functionalities and speed - We then developed a series of smart functionalities that make particle analysis more efficient and accurate than before – things like particle reconstruction that will put back together particles broken by field boundaries, and the ability to detect particles even when they cannot be clearly distinguished via electron imaging. We also make use of the sequence in which information is recorded when making decisions on what to analyse based on either morphology or an initial compositional screening. More recently we introduced a major speed boost called FeatureExpress – I spoke about speed in AZtecFeature in detail in my blog in May 2022.
Examples of smart functionalities in AZtecFeature – Left: Morphology Filtering to only analyse particles of particular interest – here, only the suspected asbestos fibres shown in red will be analysed – saving time by only analysing what is of interest. Middle: Reconstructing grains broken by field boundaries to ensure correct morphology and compositional measurements. Right: Detecting features by composition – the white particles are saturated in the BSE image meaning that we can’t tell if there are grains within it. With FeaturePhase, we map those particles and detect the grains within them via their composition – solving the problem.
Video showing very fast automated morphological measurement, EDS analysis, quantification and classification of additive manufacturing metal powders – making use of the FeatureExpress speed improvement.
Accuracy - Perhaps one of the most important things we do is to treat every EDS spectrum that we acquire with AZtecFeature in exactly the same way as we would if we were acquiring it manually. This means doing all of the corrections that are part of our TruQ collection of algorithms – to make sure that the elements that we detect are correct and that we are able to determine composition correctly.
Of course, I can’t get into the detail on all of the different functionalities here – have a look back through the blog history for numerous tips and tricks blogs on specific aspects of the software and please do have a look at our product page for AZtecFeature and our application note library which includes plenty of examples of AZtecFeature in use in multiple applications.
Dr Matt Hiscock,
MAG Head of Product Science, Oxford Instruments
Dr Matt Hiscock is the MAG Head of Product Science at Oxford Instruments and holds an MSci in geology from the University of Bristol and a PhD in geochemistry from the University of Edinburgh. He has also worked in the mining industry in Australia and been involved in the running of an academic SEM facility. He joined Oxford Instruments NanoAnalysis in 2013 and works with customers across a wide range of applications to understand their needs and how Oxford Instruments products can address them, as well as overseeing the development of products designed for specific applications and industries, particularly in the field of feature analysis.
We send out monthly newsletters keeping you up to date with our latest developments such as webinars, new application notes and product updates.
© Oxford Instruments 2024