Cyclic Iodine Loading and Regeneration of Silver-Loaded Engineered Zeolites for Off-Gas Treatment Applications
Loading...
Authors
Jenks, James Davidson
Issue Date
2025
Type
Thesis
Language
en_US
Keywords
Material characterization , Nuclear wasteforms , Radioiodine capture , Regeneration of silver , Silver zeolite
Alternative Title
Abstract
Aqueous reprocessing of used nuclear fuel results in the release of gaseous radioiodine. Silver-exchanged zeolites have long been considered effective sorbents for the capture of radioiodine from off-gas streams. Although regeneration and reuse of these costly, yet effective, sorbents were explored decades ago by the Department of Energy (DOE), the approach was likely abandoned due to poor performance during subsequent capture and regeneration cycles. In this study, loading and regeneration cycles were performed on two different sorbents to understand the properties that lead to regeneration longevity: silver-exchanged faujasite (Ag-400) with 37 mass% silver and silver-containing mordenite (Nex) with 10 mass% silver. The sorbents were in the form of engineered beads or extrudates; however, no binder material was identified. Pure silver wire (Ag) was also evaluated to understand the reactions between silver and iodine without the influence of an additional sorbent substrate. Cyclic testing was conducted by exposing sorbents to a saturated I2(g) environment at 150°C to form AgI(s), followed by regeneration at 500°C using H2(g) to strip iodine as HI(g). Ag-400 showed degradation after initial iodine loading, resulting in increased capture performance during the second cycle due to more exposure of previously inaccessible silver sites. However, the capture performance is unclear as full regeneration of the sample did not occur after the second loading cycle. Nex exhibited no visible structural changes, nor any decline in capture performance, after 5 cycles. Silver wire had limited iodine penetration of 60 µm after 18 h and AgI(s) formation during saturation testing with an 8 µm h-1 AgI(s) transformation rate that slowed down as exposure time increased. Regeneration of fully transformed wire with 18 h regeneration
time failed, only reducing a thin shell of the wire back to Ag0(s) with varying thickness between 10-30 µm. Although unsuccessful, these observations offered insight into how slow regeneration kinetic rates were and how this led to the failed regeneration of larger silver particles later observed in the Ag-400 cyclic testing. Ag-400 has silver ions uniformly distributed throughout the material. The structural degradation of Ag-400 is attributed to a combination of AgI(s) formation upon loading and reduction of Ag+ to Ag0 during regeneration, the effects of which are exacerbated due to the high silver loading. The X-ray diffraction (XRD) patterns showed amorphization of the faujasite (FAU) framework by the first cycle of regeneration. The removal of Ag+ from its framework-stabilizing position and subsequent precipitation as AgI(s) initiates the collapse of the FAU structure. Upon regeneration, the release of iodine as HI(g) is accompanied by the reduction of silver, which subsequently aggregates to form metallic particles of 0.5-4 µm. Further cycles lead to repeated aggregation and particle growth, resulting in silver- rich particles of 5-25 µm. Nex is comprised of 400-700 nm Ag-containing particles forming large clusters of 10-25 µm, and silver uniformly distributed throughout an unidentified mordenite (MOR) matrix. Changes in the silver distribution and particle morphology were observed with cycling- silver clusters aggregate into AgI(s) particles of a similar size as the whole clusters that, upon regeneration, separate and redistribute as smaller, 1-2 µm reduced metallic particles that do not sinter or aggregate further, which repeats upon subsequent cycles. MOR regions without the silver clusters experience small-scale aggregation of the uniformly distributed silver into sub-micron (400-700 nm) particles throughout the bulk upon initial loading and regeneration, but do not aggregate further upon additional cycles.
The absence of bulk structural changes indicates that the compliant matrix and low silver content were suitable to accommodate the volumetric stresses induced by repeated loading and regeneration. Despite the significant difference in silver concentration, the Ag-400 and Nex had similar iodine loading levels of 448 and 324 mg g-1, respectively. Based on these results, sorbents with silver loading and forms that minimize mechanical disruption of the substrate at the bulk scale are expected to perform better under regeneration conditions. Although the sorbents used in this study differ somewhat from those evaluated in the original DOE regeneration tests, similar trends were observed with MOR-based sorbents overperforming the FAU-based counterpart. It remains unclear whether FAU would undergo the same degree of structural collapse at lower silver exchange levels similar to the MOR. However, the excess negative charge associated with the high aluminum content in low Si/Al FAU (X) reduces framework stability, particularly in acidic environments and elevated temperatures. The ease of migration and growth of the silver particles through the bulk upon regeneration leads to the structural collapse of the engineered pellet. In contrast, the higher Si/Al of the MOR (Z) is less reliant on cation stabilization and exhibits improved thermal stability. Prolonged cycling would be required to determine whether embedded silver particles eventually lead to structural damage in the sorbent matrix, but even in the absence of regeneration, MOR would make a better choice than FAU for iodine capture in extreme environments.