Application of high-resolution positron-emission-tomography for quantitative spatiotemporal process monitoring in dense material

Methods for sensitive quantitative recognition of transport processes in opaque media without retroaction on the process itself are desirable in many scientific and technical fields. Tomographic methods based on the detection of substances labeled with radioisotopes are both most sensitive and without impact on physical or chemical properties.
During the last decade, we developed the “GeoPET-method”, applying positron-emission-tomography (PET) as laboratory method for observing flow and diffusion of reactive and non-reactive chemical species and particles in geomaterials. The substances are labeled with positron-emitting radionuclides, like 18F (decay time T1/2=110 min.), 64Cu (T1/2=12.7 h), 124I (T1/2=4.18 d), 58Co (T1/2=70.9 d), 22Na (T1/2=2.60 y), which are chosen with regard to their chemical properties and the required observation time. We use a pre-clinical PET-scanner (ClearPET by Raytest, Straubenhardt/Germany), taking advantage of its higher resolution (nearly 1 mm in dense material) and higher sensitivity, compared to clinical PET-scanners. The FOV (Field Of View) has a maximum diameter of 160 mm and a length of 100 mm. Process monitoring is accomplished by sequential recording of 3D-images with a minimum frame rate of 1 min and a maximum observation period of about 8*T1/2.
PET clearly outclasses any other tomographic modality with respect to sensitivity and selectivity. However, its application on process observation in dense material is more intricate than the habitual medical application, because substantial adverse effects have to be considered, which are due to attenuation and scattering of the annihilation photons and unfavorable other decay radiation. Applying Monte-Carlo simulations of these effects, we aim at their elimination based on fundamental physical principles.
We successfully applied GeoPET for monitoring of transport of conservative and reactive tracers, particles and humic substances in soil columns and porous or fractured rock cores. Frequently, we observe more strongly localized preferential pathways than anticipated in common transport models, and retention of – even supposedly conservative – tracers during their passage through the material. Recent studies with aerosols in flow loops aim at the verification of CFD-computations of particle deposition and remobilization. As further PET-applications in industrial process tomography we suggest e.g. the characterization of filter processes in catalysts and reactors.
Based on our quantitative images of the spatiotemporal tracer propagation in our column experiments, we parameterize 2D and 3D-numerical reactive transport simulations by inverse modeling. Examples will be demonstrated.