The discovery of X-rays by Wilhelm Röntgen1 in 1895 led to many great advances in medical and industrial radiography. But, perhaps surprisingly, high-resolution X-ray imaging of 3D objects remains a daunting challenge. Writing in Nature, Ou et al.2 report a potential solution to this problem, using nanocrystals that can trap the energy of X-rays for several weeks.
Early radiography techniques involved passing X-rays through an object and capturing an image of the remnant beam on photographic film. In the early 1980s, a method known as computed radiography was developed, in which a device called an image plate replaces the photographic film. The image plate contains compounds called phosphors, which generate excited charge carriers when irradiated by X-rays; these charge carriers become trapped close to the sites at which they were generated, and thereby produce a latent image of the remnant X-ray beam. The plate is then scanned with a laser beam to convert the latent image into luminescence, which is, in turn, converted into digital signals that are processed by a computer to reconstruct the final image3. However, computed radiography has several drawbacks, such as low image resolution and high cost.
In the mid-1990s, alternative digital X-ray-imaging techniques were developed in which X-ray energy is converted, either directly or indirectly, into electrical signals straight away4, rather than by processing a latent image as a later step. Digital radiography is generally performed using flat-panel X-ray detectors consisting of a layer of scintillators — materials that convert X-rays into light emission — and a layer of highly pixellated phototransistors, which convert the emitted light into electrical current for computational image reconstruction. But, despite enormous research efforts, these flat, non-flexible X-ray detectors are unable to produce high-resolution images of curved or irregularly shaped 3D objects.
Luminescent materials are key tools for X-ray sensing, and a group known as lanthanide-containing phosphors has been widely used in various fields, including biological detection and nanothermometry5,6. Some of these phosphors can glow for several seconds, minutes or even hours after irradiation with light7–9. The origin of this persistent luminescence has been debated9, but the consensus is that defects in the crystal lattices of the phosphors play a large part by trapping excited charge carriers generated by irradiation. Defects are produced during the synthesis of conventional lanthanide-containing phosphors by heating them to high temperatures (sometimes up to 1,700 °C)7. However, this treatment produces large particles of the phosphors, which are unsuitable for the fabrication of flexible, large-area X-ray detectors.
Ou et al. now report lanthanide-containing nanocrystals that can store excited charge carriers — produced by X-ray irradiation — in defects in the crystal lattice for several weeks. To explain this behaviour, the authors propose that fluoride ions in the crystal lattice can be displaced by collisions with X-ray photons. This produces vacancies where the ions used to be, as well as interstitials — fluoride ions at sites that are normally unoccupied. The vacancies pair up with interstitials to produce irregularities in the lattice known as Frenkel defects10.
The authors’ quantum-mechanical calculations suggest that the Frenkel defects act as traps for charge carriers in the nanocrystals, and that the traps have different depths (that is, the amount of energy that a trapped charge carrier needs to escape varies). However, the energy of the charge carriers in shallow traps can slowly escape and migrate to lanthanide ions in the lattice under ambient conditions. This process occurs concurrently with the self-healing of the defects. The researchers find that this energy migration produces luminescence that persists for more than 30 days. Such persistence is potentially useful for applications, because it extends the period of time during which a latent image can be stored in a detector before being converted into electrical signals for analysis, compared with previously used phosphors.
Ou and colleagues used these persistent luminescent nanocrystals to make flexible X-ray detectors for high-resolution 3D radiography, developing a new technique that they call X-ray luminescence extension imaging (Xr-LEI). The detectors consist of a sheet of a silicone polymer into which the nanocrystals have been embedded (Fig. 1). The sheet is wrapped around the 3D object to be imaged, and then irradiated with X-rays. Charge carriers become trapped in Frenkel defects in nanocrystals that are in regions of the detector through which X-rays pass, producing a latent image of the remnant X-ray beam. The detector is then removed and heated at 80 °C, rapidly converting the latent image into luminescence as the energy of trapped charge carriers is stimulated to migrate to lanthanide ions. The resulting image can simply be recorded using a digital camera or smartphone.
The authors demonstrated the capabilities of their technique by using it to visualize the internal structures of curved circuit boards. They found that the resolution of the images can be increased if the detector is stretched when it is applied to the object under investigation. By embedding nanocrystals in a highly stretchable silicone, the authors achieved a resolution of about 25 micrometres. This is much higher than the resolution that can be achieved using conventional flat-panel detectors (typically in the order of 100 micrometres).
Several issues will need to be addressed before Xr-LEI can be translated into medical and industrial applications. For example, there is room for improvement in the X-ray sensitivity of the detectors, because only a small amount of nanoparticles (about 2% by weight) is incorporated into the silicone sheet. More fundamentally, work is also needed to understand exactly how Frenkel defects affect luminescence, but this presents challenges — for instance, identifying Frenkel defects formed by X-ray irradiation is difficult. Advanced techniques, such as time-resolved X-ray absorption spectroscopy and solid-state nuclear magnetic resonance spectroscopy, could be used to directly probe the change of positions of fluoride ions that lead to defect formation. Nevertheless, the insights presented by Ou et al. open up a promising avenue of research that might provide a new approach for non-invasive medical radiology and inspection of nanoelectronics.