Direct Air Capture of CO2 Using Amine/Alumina Sorbents at Cold Temperature

Rising CO2 emissions are responsible for increasing global temperatures causing climate change. Significant efforts are underway to develop amine-based sorbents to directly capture CO2 from air (called direct air capture (DAC)) to combat the effects of climate change. However, the sorbents’ performances have usually been evaluated at ambient temperatures (25 °C) or higher, most often under dry conditions. A significant portion of the natural environment where DAC plants can be deployed experiences temperatures below 25 °C, and ambient air always contains some humidity. In this study, we assess the CO2 adsorption behavior of amine (poly(ethyleneimine) (PEI) and tetraethylenepentamine (TEPA)) impregnated into porous alumina at ambient (25 °C) and cold temperatures (−20 °C) under dry and humid conditions. CO2 adsorption capacities at 25 °C and 400 ppm CO2 are highest for 40 wt% TEPA-incorporated γ-Al2O3 samples (1.8 mmol CO2/g sorbent), while 40 wt % PEI-impregnated γ-Al2O3 samples exhibit moderate uptakes (0.9 mmol g–1). CO2 capacities for both PEI- and TEPA-incorporated γ-Al2O3 samples decrease with decreasing amine content and temperatures. The 40 and 20 wt % TEPA sorbents show the best performance at −20 °C under dry conditions (1.6 and 1.1 mmol g–1, respectively). Both the TEPA samples also exhibit stable and high working capacities (0.9 and 1.2 mmol g–1) across 10 cycles of adsorption–desorption (adsorption at −20 °C and desorption conducted at 60 °C). Introducing moisture (70% RH at −20 and 25 °C) improves the CO2 capacity of the amine-impregnated sorbents at both temperatures. The 40 wt% PEI, 40 wt % TEPA, and 20 wt% TEPA samples show good CO2 uptakes at both temperatures. The results presented here indicate that γ-Al2O3 impregnated with PEI and TEPA are potential materials for DAC at ambient and cold conditions, with further opportunities to optimize these materials for the scalable deployment of DAC plants at different environmental conditions.

as a mixture of compounds Table S1 shows that the impregnation procedure of PEI and TEPA decreases the total pore volume and Brunauer-Emmett-Teller (BET) surface area of the γ-Al 2 O 3 support.This indicates the successful incorporation of the amines into the pores of γ-Al 2 O 3 .However, some deposition of the amines on the outer surface of the support cannot be ruled out, especially at the higher amine loadings.As expected, the BET surface areas decrease with an increase in the amine content while the pore filling increases.Apart from PEI and TEPA, spermine and spermidine (Scheme S1) were also impregnated into the pores of alumina (40 wt.%) and their CO 2 capacities were measured at 25 °C and -20 °C to understand the effect of varying chain lengths on the CO 2 uptakes.At 25 °C, the spermineimpregnated γ-Al 2 O 3 sample showed the highest CO 2 capacity among all the samples (~ 2 mmol/g), while spermidine showed the lowest capacity (~0.2 mmol/g) as seen in Figure S1.However, the CO 2 capacity for the spermine sample drastically reduced to 0.5 mmol/g while the capacity for the spermidine sample remained the same at -20 °C.Burnout experiments to determine the organic content in the γ-Al 2 O 3 support showed that much less spermidine was present in the pores of alumina (< 10 wt.%).2] Spermidine being highly volatile, escaped from the pores to a large extent during the drying of the sample in the rotary evaporator during synthesis.There was also significant loss of spermine during the activation of the sample before measurement, which may be the reason for the drastic decrease in the CO 2 uptakes at -20 °C.Since both the spermineand spermidine-incorporated γ-Al 2 O 3 sorbents displayed poor CO 2 capacities at -20 °C and are more volatile than TEPA, they were not considered as promising candidates for DAC applications at ambient or sub-ambient conditions and further exploration of their behavior is not reported.Figure S2a shows that trends for 20 wt.% TEPA samples are similar to the trends observed in Figure 1 where the CO 2 capacities decrease with temperature, however to a lesser extent as compared to the 40 wt.% sample.The 20 wt.% sample reaches a pseudoequilibrium capacity unlike the 40 wt.% sample (Figure 2a) due to lower inhibition to CO 2 diffusion since the pore volume of the 20 wt.% sample is greater than the 40 wt.% sorbent.Figure S2b shows the desorption profile for CO 2 adsorption at 25 °C and -20 °C for the 20 wt.% TEPA sample.The trends are very similar to that of the 40 wt.%TEPA sample where the CO 2 desorption profile shows a small physisorbed CO 2 peak for the adsorption at -20 C below 0 °C and a peak at ~60 °C for both adsorption temperatures.Apart from this difference, the desorption peaks for the 20 wt.% sample look similar irrespective of the CO 2 adsorption temperature.Figure S3a shows that as the amine content in the sample decreases from 40 wt.% to 10 wt.%, the time taken to reach a pseudoequilibrium decreases.The quicker approach to pseudoequilibrium with lower amine loading happens for similar reasons as described for the TEPA samples.The pore volume occupied by the PEI decreases as the amine loading decreases, as seen from Table S1, which decreases the barriers to diffusion of CO 2 , hence allowing a quicker approach to pseudoequilibrium.The trend for the desorption profiles mimics the trends for the TEPA samples.The desorption peak moves to a lower temperature as the amine content decreases, indicating that interaction of CO 2 with PEI weakens, which in turn implies that the interaction of the PEI chains with the walls has increased, on average.The differences in desorption profiles are especially obvious for the 10 wt.% PEI sample, which shows very broad peaks of CO 2 desorption.The peaks occurring at lower desorption temperatures indicate that the CO 2 interacts weakly with the amines.Reminiscent of TEPA samples, PEI would also mostly coat the walls of γ-Al 2 O 3 at lower amine loadings, thus interacting more with the walls and less with CO 2 , which manifests as lower desorption peak temperatures.On the other hand, after initially coating the walls of the support, PEI likely forms multilayers of aggregates at higher amine loadings, 3 which contain a greater fraction of free amines, thereby reducing the average degree of interaction with the walls of the support.The TPD results (Section 3.1 and S3) show that, unlike MIL-101(Cr) materials incorporated with PEI and TEPA, 4 the desorption profiles for amine-based γ-Al 2 O 3 sorbents are not significantly affected by the adsorption temperatures for the higher amine loading samples.Hence, for CO 2 adsorption at 25 °C, the same desorption conditions as those for adsorption at -20 °C were used for both PEI and TEPA.Figures S7-S10 show the adsorption-desorption cycles for the 20 wt.% and 40 wt.%TEPA and PEI sorbents.All the samples were stable across the 5 cycles at 25 °C.The average working capacity of the 40 wt.% TEPA and 20 wt.% TEPA samples were 1.3 and 0.9 mmol/g, respectively.The 40 wt.%TEPA sorbent showed lower working capacity than the pseudoequilibrium capacity obtained from Figure 2, while the 20 wt.% TEPA showed a very similar working capacity to the pseudoequilibrium capacity.This is also due to the 40 wt.%TEPA sample suffering from slower adsorption rates after the initial rapid uptake (Figure 2) as compared to the faster adsorption rates for the 20 wt.% TEPA.The 40 wt.%PEI sample showed a working capacity of 1 mmol/g sorbent while the 20 wt.% PEI sorbent showed a working capacity of 0.5 mmol/g.Both values are close to the pseudoequilibrium capacities obtained from long timescale adsorption measurements.

Figure S3 .
Figure S3.(a) CO 2 uptake profile of 40 wt.%, 20 wt.% and 10 wt.% PEI-impregnated γ-Al 2 O 3 at -20 °C and 400 ppm CO 2 (balance He).(b) CO 2 -TPD profiles after CO 2 adsorption at -20 °C.The CO 2 adsorption profiles at -20 °C are often wavy due to the injection of pulses of liquid nitrogen intermittently to maintain the temperature in the TGA/DSC.

S5.
Figure S4.Sample mass change of (a) 20 wt.% TEPA and (b) 20 wt.% PEI impregnated γ-Al 2 O 3 over 10 CO 2 adsorption and desorption cycles.CO 2 adsorption measurement was performed at -20 °C, 400 ppm CO 2 (balance He) for 2 h, and desorption at 60 °C for TEPA and 50 °C for PEI for 2 h.The CO 2 adsorption profiles at -20 °C are often wavy due to the injection of pulses of liquid nitrogen intermittently to maintain the temperature in the TGA/DSC

Table S1 :
Physical characteristics of the sorbents

Table S2 :
Summary of adsorption rates for different sorbents at 25 and -20 °C