They say that a picture speaks a thousand words. Experimental work published earlier this year by researchers from the Department of Materials Science and Engineering at the University of Sheffield and AWE indeed show this to be the case.
Digital images are composed of hundreds of thousands of pixels, each made up of red, green and blue sub-pixels of different intensities. The intensity of each sub-pixel is set to a value between zero and 255. This gives a total of more than 16.7 million colour combinations.
Researchers from the Nuclear Materials research group in the Department of Materials Science and Engineering proposed that, given the vast number of colour combinations, and the availability of high resolution imaging equipment, it should be possible to identify constituent parts of a powder, in particular, a sample of powdered nuclear material.
Successful demonstration of this proposed technique would prove useful in the area of nuclear forensics, where there is a need for consistent, objective analysis.
Nuclear forensics is a discipline dedicated to the investigation of nuclear materials to determine their processing history and provenance in the context of law enforcement and national security [1]. The International Atomic Energy Agency (IAEA) have established the IAEA Incident and Trafficking Database (ITDB), for collecting information on incidents of nuclear and other radioactive materials found out of regulatory control; between 1993 and 2019, a total of 3686 incidents had been reported to the ITDB by participating nations, justifying the need for an international forensic capability [2].
The currently accepted approach to nuclear forensics is to use spectroscopic and radioisotopic analysis, but there is still scope for analytical techniques for other bulk characteristics. One possible solution is the use of colour analysis, where there is currently no clear, objective colour analysis method.
Subjective, by-eye colour analysis is not useful to nuclear forensics, as the method is open to bias or could result in inconsistencies between laboratories.
With an objective technique, it would be possible to analyse samples quantitatively based on colour determination.
The research, performed by scientists at the University of Sheffield and published in Forensic Science International in February (https://www.sciencedirect.com/science/article/pii/S0379073820305405), details the methods of sample analysis, proposing that the technique isn't necessarily limited to laboratory work, but could be applied in the field, in situations where rapid results are required. It also highlighted the limitations and inconsistencies of by-eye subjective inspection of the powder samples.
In the work, the focus has been on uranium oxide powders which have been subjected to a particular heat treatment process. However, it is strongly believed that the process, having successfully been demonstrated to work on uranium oxide powders, could also be applied to other types of powder, such as plutonium oxides and mixed oxide nuclear fuels.
Photographs of samples from one repeat of sample, with the mean RGB values and their corresponding ‘best match’ Pantone colour code. Cumulative RGB histograms are also displayed.
While this method of analysis is showing some promise, further work is planned to investigate a wider range of materials and heat treatment regimes and provide further confirmation that this technique could be used in the field of nuclear forensics.
The research team, led by Professor Neil Hyatt, includes Nathan Thompson, Sarah O’Sullivan, Robert Howell and Dr Daniel Bailey from the University of Sheffield and Dr Matthew Gilbert from AWE.
UK Ministry of Defence © Crown Owned Copyright 2021/AWE
____
[1] International Atomic Energy Agency, "Development of a National Nuclear Forensics Library: A System for the Identification of Nuclear or Other Radioactive Material Out of Regulatory Control", IAEA, Vienna (2018)
[2] International Atomic Energy Agency, "ITDB 2020 Factsheet", IAEA (2020)