X-ray vision enhanced
High-energy X-ray colour imaging holds the key to producing images of materials with an unprecedented amount of structural information. Jacqueline Hewett reports on research into novel detector materials to satisfy the demanding technical requirements of this approach.
The vision most synonymous with X-ray technology is an image of a broken bone, where the bone appears white because it has a higher density, and therefore has absorbed more X-rays than the surrounding soft tissue. But this is only half of the picture. All substances scatter, as well as absorb X-rays. This means that if you can collect and measure the different wavelengths of the scattered X-ray light, you can produce a fingerprint of what gave rise to the scatter.
The technology of 3D imaging, however, is limited by the available detector arrays, outlines Robert Cernik, Professor of Synchrotron Radiation and Materials Science at The University of Manchester, UK. Cernik is Principal Investigator of the UK-based project High-Energy X-ray Imaging Technology (HEXITEC).
He explains that, at present, it is not possible to obtain defect-free detector material of an area greater than about one square centimetre. High energy X-ray colour imaging, however, requires high quality semiconductor material. The technique would then generate a 3D density contrast image as well as provide structural information, such as of strain inside a weld.
After four years’ of R&D, the partners involved in HEXITEC claim they can fabricate suitable detector materials for the technique and are on the verge of taking high-energy colour X-ray imaging into application areas ranging from medicine, security scanning, gas and oil exploration, and aerospace and automotive engineering.
Finding the right material
The starting point was developing a detector material capable of working with high-energy X-rays on the order of 60–80keV all the way up to 400keV.
Most medical X-rays are around 30keV and are detected using silicon, germanium or standard scintillation counters. When it comes to high-energy X-rays, however, silicon atoms are too light to stop them.
‘An X-ray detector works when an X-ray photon hits the material and generates a number of electron-hole pairs. The number of pairs is directly proportional to the energy of the incident photon, so counting the number of electron hole pairs will tell you the energy of the photon,’ says Cernik.
The next material worth considering was therefore germanium, which is heavier than silicon, with 32 electrons. ‘You can make effective X-ray detectors with germanium,’ comments Cernik. ‘The problem is that germanium’s bandgap is small and its leakage current is therefore high. To overcome this, you need to cool the detector to liquid nitrogen temperatures, which is cumbersome and it is not practically scalable. It is also not very good at stopping 60-70keV rays’.
Other possibilities are indium antimonide, mercuric iodide and thalium bromide, but each has its disadvantages. The material of choice became cadmium zinc telluride (CdZnTe) as it is heavy, does not require cooling and, most importantly, can stop high energy X-rays. It also has a large bandgap, is good at generating electron pairs, and has longer charge carrier lifetime. The challenge was to produce CdZnTe on a wafer scale.
Optimising growth conditions
Although CdZnTe can be grown by a number of methods, including from the melt phase, the fundamental difficulty is the difference in the melting point and vapour pressure of the constituent atoms. At the point of crystallisation, the in-built strains tend to cause dislocations, cracks, pipes and inclusions.
The HEXITEC team grew its detector grade CdZnTe using multi-tube physical vapour transport (MTPVT), a technique patented by Professor Andy Brinkman of Durham University, UK.
It is said to be inherently immune to many of these difficulties and has been shown to work well for CdZnTe, after many trials to optimise parameters such as temperature, growth rate, and crystal seed orientation.
Under the right conditions, the researchers are able to grow two-inch diameter wafers that are two-to-six millimetres thick – the right size and quality for detector grade material.
An added advantages of MTPVT is that it inhibits the introduction of defects such as tellurium inclusions.
From material to detector
‘Having a detector material, however, is not [having] a detector,’ remarks Cernik. The next major goal was taking the detector material, pixellating it and bonding it to all of the electronics necessary to create a fully functioning detector.
Paul Sellin, at the University of Surrey determined the quality of the material for electronic contacting to evaluate properties such as resistivity, the number of trapped states and charge-carrier lifetimes. In addition, his team was responsible for research into passivation methods and contacting.
‘CdZnTe is the Teflon of the semiconductor world in that it is difficult to get a material to permanently adhere to its surface,’ comments Cernik.
This structure was handed to Paul Seller at Rutherford Appleton Laboratory who led the effort to create an 80x80 array of 250 square pixels. Seller’s group was responsible for developing a flip-chip bonding process to attach the application specific integrated circuit onto the back of the detector chip.
‘The last stage is to acquire and process the huge volume of data produced in a meaningful way,’ says Cernik. ‘This is where the project is right now. We are developing data acquisition, visualisation and evaluation software that allows you to recognise a diffraction pattern relating to a certain thing, such as diseased tissue, strained aluminium or explosives’.
The HEXITEC team has now secured funding for a further three years to push the technology into viable applications. One challenge that remains is how to scale up manufacturing to an industrial level.