Probing the Distribution and Mobility of Aminopolymers after Multiple Sorption-Regeneration Cycles: Neutron Scattering Studies

Abstract

Solid-supported amines are effective CO₂ adsorbents capable of capturing CO₂ from flue gas streams (10-15 vol % CO₂) and from ultradilute streams, such as ambient air (∼400 ppm CO₂). Amine sorbents have demonstrated promising performance (e.g., high CO₂ uptake and uptake rates) with stable characteristics under repeated, idealized thermal swing conditions, enabling multicycle application. Literature studies suggest that solid-supported amines such as PEI/SBA-15 generally exhibit slowly reducing CO₂ uptake rates or capacities over repeated thermal swing capture-regeneration cycles under simulated DAC conditions. While there are experimental reports describing changes in supported amine mass, degradation of amine sites, and changes in support structures over cycling, there is limited knowledge about the structure and mobility of the amine domains in the support pores over extended use. Furthermore, little is known about the effects of H₂O on cyclic applications of PEI/SBA-15 despite the inevitable presence of H₂O in ambient air. Here, we present a series of neutron scattering studies exploring the distribution and mobility of PEI in mesoporous silica SBA-15 as a function of thermal cycling and cyclic conditions. Small-angle neutron scattering (SANS) and quasielastic neutron scattering (QENS) are used to study the amine and H₂O distributions and amine mobility, respectively. Applying repeated thermal swings under dry conditions leads to the thorough removal of water from the sorbent, causing thinner and more rigid wall-coating PEI layers that eventually lead to slower CO₂ uptake rates. On the other hand, wet cyclic conditions led to the sorption of atmospheric water at the wall-PEI interfaces. When PEI remains hydrated, the amine distribution (i.e., wall-coating PEI layer thickness) is retained over cycling, while lubrication effects of water yield improved PEI mobility, in turn leading to faster CO₂ uptake rates.

Figure 1

Trend of sorbent mass and working capacities through cycles. (A) Sorbent mass under dry, medium humidity, and high humidity conditions based on the sorbent mass right after regeneration. (B) CO₂ uptake under cyclic conditions. For wet cycles, relative CO₂ uptake values were taken based on the detected gas composition at the gas analyzer.

Figure 2

SANS spectra for samples treated under different cyclic conditions. Bragg peaks [10], [11], [20], [21], and [22] are annotated.

Figure 3

(A) Representative SANS spectra fitting results for the as-synthesized sample and samples after 60 cycles under varied streams (Dry, Wet MH, and Wet HH). (B) Schematic illustrations of structural contributors. (B) Was reproduced with permission from ref (58). Copyright 2023 American Chemical Society.

Figure 4

Hypothesized structures at pore wall-PEI interfaces. (A) Corona SLD as a function of cyclic conditions. Error bars represent standard deviations from three best fit results. (B) Hypothetical distribution of PEI and H₂O around pore walls.

Figure 5

(A) QENS spectra for samples treated under different cyclic conditions. (B) Highlighting QENS broadening at the lower energy window representing slower, center-of-mass diffusion of the PEI mass. Spectra were recorded at 360 K.

Figure 6

(A) EISF plots for the fresh sample and samples treated under different cyclic conditions, measured at 360 K. Dotted lines denote EISF fits to the theoretical model (eq 3). (B) Hypothesized structures and motions of PEI under different extents of intercalated H₂O (upper: as-synthesized, middle: dry-treated, and lower: wet-treated).

Figure 7

(A) Schematic of the cross-section of a mesopore in PEI/SBA-15 composites and hypothesized CO₂ diffusion path in PEI/SBA-15 sorbents. (B–D) Fractional uptake vs time for Dry, Wet MH, and Wet HH cycles, respectively (inset graphs denote 0–5 min).