Periodic Mesoporous Organosilica Nanoparticles for CO2 Adsorption at Standard Temperature and Pressure

(1) Background: Due to human activities, greenhouse gas (GHG) concentrations in the atmosphere are constantly rising, causing the greenhouse effect. Among GHGs, carbon dioxide (CO2) is responsible for about two-thirds of the total energy imbalance which is the origin of the increase in the Earth’s temperature. (2) Methods: In this field, we describe the development of periodic mesoporous organosilica nanoparticles (PMO NPs) used to capture and store CO2 present in the atmosphere. Several types of PMO NP (bis(triethoxysilyl)ethane (BTEE) as matrix, co-condensed with trialkoxysilylated aminopyridine (py) and trialkoxysilylated bipyridine (Etbipy and iPrbipy)) were synthesized by means of the sol-gel procedure, then characterized with different techniques (DLS, TEM, FTIR, BET). A systematic evaluation of CO2 adsorption was carried out at 298 K and 273 K, at low pressure. (3) Results: The best values of CO2 adsorption were obtained with 6% bipyridine: 1.045 mmol·g−1 at 298 K and 2.26 mmol·g−1 at 273 K. (4) Conclusions: The synthetized BTEE/aminopyridine or bipyridine PMO NPs showed significant results and could be promising for carbon capture and storage (CCS) application.


Introduction
In recent years, a significant increase in the concentration of CO 2 into the atmosphere, one of the main greenhouse gases, has given rise to serious concerns, and there is a clear need for greater efforts to reduce CO 2 emissions. Indeed, CO 2 emissions between 1974 and 2004 were greater than those during the 600,000 previous years [1]. Moreover, the current atmospheric concentration of CO 2 reached a value of 414 ppm in 2020, the highest value on Earth for 3 million years [2]. At the same time, CO 2 is an ecological and almost unlimited source of carbon and can be used for many applications such as the production of value-added chemicals and energy products including carbon monoxide [3], methane [4], methanol [5,6], formate [5], methoxide [7], cyclic carbonate or polycarbonate [8], formic acid [9], and urea [10]. Thus, the reduction of CO 2 in the atmosphere has become a very important research topic.
Initially, liquid amine solutions or aqueous alkanolamines such as ethanolamine (MEA), di-ethanolamine (DEA), di-glycolamine (DGA) and N-methyldiethanolamine (MDEA) were widely used in the chemical industry to trap acid gases such as carbon dioxide (CO 2 ) from gas mixtures [11]. These solutions are able to react and reversibly form other chemical species such as carbamate, carbonate, bicarbonate, etc. [11]. However, this approach to chemical absorption has many disadvantages, such as huge energy consumption for the absorption and regeneration of amines, the high cost of CO 2 evaporation, and other problems related to amines including poor stability and oxidative degradation of amines, and interference from impure gases such as SO 2 , NO 2 , and NO [12]. Thus, the use of solid adsorbents has gradually replaced this technique.
To overcome the problems of conventional liquid amines, new studies have gradually focused on the synthesis of solid adsorbents modified by amines. Recently, porous silica and periodic mesoporous organosilicas (PMOs) have become popular solid adsorbents for CO 2 capture [22,23]. An increase in CO 2 adsorption capacity was observed for PMOs modified with compounds containing amines [24], amines having good affinity with CO 2 , allowing templating diamino-alkyltrimethoxysilanes and leading to lamellar materials after the sol-gel procedure [25]. The impregnation of porous silica adsorbents with amine groups has several advantages such as simple manufacture, low cost, and the ability to load a large number of amines [26]. However, the disadvantage of this technique concerns the amines, which are located at the surface of the porous solids and not inside the pores, which may affect the optimal conditions for CO 2 adsorption. Other scientists have observed a significant improvement in the CO 2 adsorption by using mesoporous silica modified with various grafted amines such as 3-(aminopropyl)triethoxysilane and N- [3-(trimethoxysilyl)propyl]ethylenediamine [27,28]. Furthermore, the use of aminefunctionalized organosilica for CO 2 capture and valorization is growing significantly [29]. The use of silica-based nanoparticles for the storage of CO 2 has been less described than the use of bulk materials.
FTIR molecular spectroscopy analysis was carried out ( Figure S1). The obtained spectra allowed the identification of the BTEE precursor with characteristic absorption bands at 1020 cm −1 and 1160 cm −1 , respectively, corresponding to Si-O-Si and Si-OH bonds. The absorption bands corresponding to the aromatic cycle of the pyridine precursor were also observed around 690-750 cm −1 (C=C and C=N bonds), 1600 cm −1 (amide II band), and 1690 cm −1 (amide I band). The presence of these bands proved the success of the co-condensation between the BTEE and the aminopyridine precursors. Moreover, an increase in the nitrogen content was confirmed by elementary analysis (Table S1).
PMO 1 NPs showed monodisperse nanoparticles of about 80 nm diameter by TEM ( Figure 3a and Table 1) and mesoporosity was clearly visible. Dynamic Light Scattering (DLS) in EtOH ( Figure S2) confirmed monodispersity with a hydrodynamic diameter of 200 nm. Compared to reference NPs, the addition of aminopyridine led to more polydisperse systems. The pyPMO 6 NPs showed mean diameters of 120 nm but with a standard deviation of 115 nm by TEM (Figure 4a, Table 1). The nanoparticle hydrodynamic diameters distribution was between 100 and 200 nm but aggregates were observed by DLS ( Figure S3) at 900 nm. The pyPMO 15 NPs led to smaller nanoparticles of 100 nm diameter but more polydispersed than PMO 1 NPs, (Figure 5a, Table 1) for which DLS showed polydisperse NPs holding 200-500 nm hydrodynamic diameters with large aggregates in the range of 1-2 µm ( Figure S4). increasing aminopyridine molar fraction: 6% (pyPMO 6 NPs) and 15% (pyPMO 15 NPs). Reference PMO NPs were synthesized with 100% BTEE (PMO 1 NPs). FTIR molecular spectroscopy analysis was carried out ( Figure S1). The obtained spectra allowed the identification of the BTEE precursor with characteristic absorption bands at 1020 cm −1 and 1160 cm −1 , respectively, corresponding to Si-O-Si and Si-OH bonds. The absorption bands corresponding to the aromatic cycle of the pyridine precursor were also observed around 690-750 cm −1 (C=C and C=N bonds), 1600 cm −1 (amide II band), and 1690 cm −1 (amide I band). The presence of these bands proved the success of the co-condensation between the BTEE and the aminopyridine precursors. Moreover, an increase in the nitrogen content was confirmed by elementary analysis (Table S1).
PMO 1 NPs showed monodisperse nanoparticles of about 80 nm diameter by TEM ( Figure 3a and Table 1) and mesoporosity was clearly visible. Dynamic Light Scattering (DLS) in EtOH ( Figure S2) confirmed monodispersity with a hydrodynamic diameter of 200 nm. Compared to reference NPs, the addition of aminopyridine led to more polydisperse systems. The pyPMO 6 NPs showed mean diameters of 120 nm but with a standard deviation of 115 nm by TEM ( Figure 4a, Table 1). The nanoparticle hydrodynamic diameters distribution was between 100 and 200 nm but aggregates were observed by DLS (Figure S3) at 900 nm. The pyPMO 15 NPs led to smaller nanoparticles of 100 nm diameter but more polydispersed than PMO 1 NPs, (Figure 5a, Table 1) for which DLS showed polydisperse NPs holding 200-500 nm hydrodynamic diameters with large aggregates in the range of 1-2 µm ( Figure S4).  increasing aminopyridine molar fraction: 6% (pyPMO 6 NPs) and 15% (pyPMO 15 NPs). Reference PMO NPs were synthesized with 100% BTEE (PMO 1 NPs). FTIR molecular spectroscopy analysis was carried out ( Figure S1). The obtained spectra allowed the identification of the BTEE precursor with characteristic absorption bands at 1020 cm −1 and 1160 cm −1 , respectively, corresponding to Si-O-Si and Si-OH bonds. The absorption bands corresponding to the aromatic cycle of the pyridine precursor were also observed around 690-750 cm −1 (C=C and C=N bonds), 1600 cm −1 (amide II band), and 1690 cm −1 (amide I band). The presence of these bands proved the success of the co-condensation between the BTEE and the aminopyridine precursors. Moreover, an increase in the nitrogen content was confirmed by elementary analysis (Table S1).
PMO 1 NPs showed monodisperse nanoparticles of about 80 nm diameter by TEM ( Figure 3a and Table 1) and mesoporosity was clearly visible. Dynamic Light Scattering (DLS) in EtOH ( Figure S2) confirmed monodispersity with a hydrodynamic diameter of 200 nm. Compared to reference NPs, the addition of aminopyridine led to more polydisperse systems. The pyPMO 6 NPs showed mean diameters of 120 nm but with a standard deviation of 115 nm by TEM ( Figure 4a, Table 1). The nanoparticle hydrodynamic diameters distribution was between 100 and 200 nm but aggregates were observed by DLS (Figure S3) at 900 nm. The pyPMO 15 NPs led to smaller nanoparticles of 100 nm diameter but more polydispersed than PMO 1 NPs, (Figure 5a, Table 1) for which DLS showed polydisperse NPs holding 200-500 nm hydrodynamic diameters with large aggregates in the range of 1-2 µm ( Figure S4).         In the case of PMO 1, the shape of the sorption isotherm revealed a type IV isotherm, with a narrow pore size located at 2.7 nm. Its specific surface area was found to be 626 m 2 ·g −1 (Figure 3b). Higher values of specific surface areas were obtained with pyPMO 6, (918 m 2 ·g −1 , Figure 4b) and interestingly, the pore size was similar, compared to PMO 1 (2.7 nm, Figure 4b). The specific surface area seemed to decrease with the increase in the amino pyridine molar fraction to 15% (around 800 m 2 ·g −1 , Figures 4b and 5b). Note that the contribution of mesopores to the sorption capacity was higher compared to PMO 1, which showed a better structuration of the mesoporosity when using pyridine moieties, even though larger distributed pore sizes from 2.5 to 2.8 nm were observed for the larger aminopyridine fraction materials (Table 1, Figures 4b and 5b).  In the case of PMO 1, the shape of the sorption isotherm revealed a type IV isotherm, with a narrow pore size located at 2.7 nm. Its specific surface area was found to be 626 m 2 ·g −1 (Figure 3b). Higher values of specific surface areas were obtained with pyPMO 6, (918 m 2 ·g −1 , Figure 4b) and interestingly, the pore size was similar, compared to PMO 1 (2.7 nm, Figure 4b). The specific surface area seemed to decrease with the increase in the amino pyridine molar fraction to 15% (around 800 m 2 ·g −1 , Figures 4b and 5b). Note that the contribution of mesopores to the sorption capacity was higher compared to PMO 1, which showed a better structuration of the mesoporosity when using pyridine moieties, even though larger distributed pore sizes from 2.5 to 2.8 nm were observed for the larger aminopyridine fraction materials (Table 1, Figures 4b and 5b).
CO2 adsorption measurements were performed at 298 K and 273 K ( Figure 6, Table  2). A slight increase in the CO2 adsorption capacity with the aminopyridine molar fraction was observed at 298 K. Moreover, an increase of 14% in the CO2 adsorption was noticed for the pyPMO 15 (0.92 mmol·g −1 ) having the highest aminopyridine content (15%) compared to the 100% BTEE (0.81 mmol·g −1 ). This shows the potential of adding aminopyridine to BTEE for CO2 capture. However, different results were observed at 273 K, where the CO2 adsorption for the 100% BTEE was higher than with others PMOs, showing in this case the potential of PMO 1 for CO2 capture. CO2 adsorption increases with decreas-  Table 2). A slight increase in the CO 2 adsorption capacity with the aminopyridine molar fraction was observed at 298 K. Moreover, an increase of 14% in the CO 2 adsorption was noticed for the pyPMO 15 (0.92 mmol·g −1 ) having the highest aminopyridine content (15%) compared to the 100% BTEE (0.81 mmol·g −1 ). This shows the potential of adding aminopyridine to BTEE for CO 2 capture. However, different results were observed at 273 K, where the CO 2 adsorption for the 100% BTEE was higher than with others PMOs, showing in this case the potential of PMO 1 for CO 2 capture. CO 2 adsorption increases with decreasing temperature. Indeed, adsorption being an exothermic process, increasing the temperature of the system decreases the adsorption capacity [39]. In terms of affinity, it is worth looking at low CO 2 pressure, where the highest interaction sites can be found. The shape of the sorption isotherms is almost linear up to 266 hPa, which confirms the rather poor affinity of these materials for carbon dioxide, regardless of temperature. In these cases, similar Henry's constants can be derived (~1.5·10 −3 mmol·g −1 ·Pa −1 at 298 K). ing temperature. Indeed, adsorption being an exothermic process, increasing the temperature of the system decreases the adsorption capacity [39]. In terms of affinity, it is worth looking at low CO2 pressure, where the highest interaction sites can be found. The shape of the sorption isotherms is almost linear up to 266 hPa, which confirms the rather poor affinity of these materials for carbon dioxide, regardless of temperature. In these cases, similar Henry's constants can be derived (~1.5·10 −3 mmol·g −1 ·Pa −1 at 298 K).    Three BTEE/iPrbipyridine PMO NPs were synthetized with BTEE and 5,5 -bis(triisopropoxysilyl)-2,2 -bipyridine (iPrbipy) (Figure 7) by increasing the iPrbipyridine molar fraction from 6 to 15%: 6% (iPrbipyPMO 6), 10% (iPrbipyPMO 10) and 15% (iPrbipyPMO 15).
FTIR molecular spectroscopy analysis ( Figure S5) revealed the success of the cocondensation between the BTEE and the iPrbipyridine precursors, as the pyridine precursors exhibited the same characteristic absorption bands. Indeed, a new absorption band, centered at 750 cm −1 appeared. This band can be attributed to the C=N bonds from the two aromatic cycles. An increase in the intensity of this band with the bipyridine molar fraction was observed, due to the four-fold increase in nitrogen content (Table S2). This offers convincing proof of the success of the polycondensation between the BTEE and the bipyridine precursors.
TEM images (Figures 8a, 9a and 10a) showed highly porous (hollow systems) but polydispersed nanoparticles with a mean diameter of 92 nm (iPrbipyPMO 6) or 111 nm (iPrbipyPMO 10). DLS in EtOH confirmed polydispersity of iPrbipyPMO 6 with two populations centered at 200 nm and 900 nm (nanoparticles and aggregates, respectively, Figure S6). For iPrbipyPMO 10 and iPrbipyPMO 15 the populations were centered between 500-600 nm, which is characteristic of dispersed aggregates of nanoparticles ( Figures S7 and S8).        The three iPrbiPyPMO materials were mesoporous, as indicated by a clear uptake at intermediate relative pressures followed by saturation (Figures 8b, 9b and 10b). However, this saturation became less clear as the bipyridine moieties fraction increased. Concomitantly, the hysteresis loop increased. The shape of the hysteresis loop (H3 type) indicates the presence of hollow materials, in which desorption occurs through small mesopores compared to the size of the cavities. The specific surface area with the addition of 6% and 10% bipyridine was higher compared to that of PMO 1 (958 m 2 ·g −1 and 796 m 2 ·g −1 ). The derivation of the pore size distribution using the desorption branches of the sorption isotherms gave two populations of pores, the most important located at ~2.5 nm and a second one located at 3.7 nm. The latter is reminiscent of cavitation effects when desorption occurs in small mesopores connected to large cavities, which is precisely the case with hollow materials.
In terms of CO2 capture, the bipyridine-based PMOs behaved differently compared to the aminopyridine-based PMOs ( Figure 11, Table 3). At low pressure, the four traces were different, and they were almost superposed in the case of aminopyridine-based The three iPrbiPyPMO materials were mesoporous, as indicated by a clear uptake at intermediate relative pressures followed by saturation (Figures 8b, 9b and 10b, Table 3). However, this saturation became less clear as the bipyridine moieties fraction increased. Concomitantly, the hysteresis loop increased. The shape of the hysteresis loop (H3 type) indicates the presence of hollow materials, in which desorption occurs through small mesopores compared to the size of the cavities. The specific surface area with the addition of 6% and 10% bipyridine was higher compared to that of PMO 1 (958 m 2 ·g −1 and 796 m 2 ·g −1 ). The derivation of the pore size distribution using the desorption branches of the sorption isotherms gave two populations of pores, the most important located at~2.5 nm and a second one located at 3.7 nm. The latter is reminiscent of cavitation effects when desorption occurs in small mesopores connected to large cavities, which is precisely the case with hollow materials.
In terms of CO 2 capture, the bipyridine-based PMOs behaved differently compared to the aminopyridine-based PMOs ( Figure 11, Table 4). At low pressure, the four traces were different, and they were almost superposed in the case of aminopyridine-based PMOs. In the latter case, the influence of aminopyridine moieties was found to be moderate at low pressure, as the sorption capacity was close to that of PMO 1. Using bipyridine moieties, PMO 1 had the lowest CO 2 sorption capacity, which indicates a clear interaction between bipyridine moieties and carbon dioxide. Quantitatively, the highest CO 2 adsorption capacity was obtained with iPrbipyPMO 6 with 6% iPrbipyridine (1.045 mmol·g −1 at 298 K and 2.26 mmol·g −1 at 273 K). A straightforward explanation is that the greatest specific surface area is found with this material, compared to the other PMOs of the series (Figure 8b, Table 3). In terms of sorption efficiency, it is worth differentiating the sorption capacity per square meter. This calculation yields different sorption capacities: 1.09 µmol·m −2 , 0.992 µmol·m −2 , and 1.97 µmol·m −2 for iPrbipyPMO 6, iPrbipyPMO 10, and iPrbipyPMO 15, respectively. This trend shows that bipyridine moieties interact with CO 2 , even though other effects may come into play, such as structuration, pore blocking, and so on. At 273 K, it can be noted that the sorption capacities of the bipyridine-based PMOs reach those of PMO 1, which indicates that even if the CO 2 /bipyridine-based PMOs' affinity is greater compared to PMO 1, their sorption capacities are similar (apart from 6% bipyridine).
On the IR spectra side, ( Figure S9) as for the synthetized PMO NPs with the iPrbipyridine precursor, the characteristic absorption band of the C=N aromatic bond located   Figure 11. Evolution of the CO 2 adsorption of BTEE/iPrbipyridine PMO NPs with different bipyridine molar fractions versus absolute pressure at 298 K (left) and 273 K (right).
On the IR spectra side, ( Figure S9) as for the synthetized PMO NPs with the iPrbipyridine precursor, the characteristic absorption band of the C=N aromatic bond located around 750 cm −1 increased with the Etbipyridine molar fraction, a feature of the co-condensation between the BTEE and the Etbipyridine precursors. This band did not appear for the aminopyridine precursor, probably due to the difference in chemical structure with the presence of two aromatic cycles with nitrogen for the iPrbipyridine and the Etbipyridine, against only one aromatic cycle for the aminopyridine precursor. An increase in the percentage of nitrogen with the increase in the molar fraction of Etbipyridine was observed. However, these percentages were lower than with the other precursors (Table S3).
TEM images (Figures 13a, 14a and 15a) showed the formation of larger particles than pyPMO NPs and iPr PMO NPs, with diameters ranging from 300 to 588 nm. A significant standard deviation was also observed for the three PMO NPs (Table 5). Hollow structure was not observed. DLS analysis revealed a single population of nanoparticles with mean hydrodynamic diameter around 1100 nm, due to aggregation of the NPs in EtOH (Figures S10-S12).
Molecules 2022, 27, x FOR PEER REVIEW 10 hydrodynamic diameter around 1100 nm, due to aggregation of the NPs in EtOH ( Fig  S10-S12).   hydrodynamic diameter around 1100 nm, due to aggregation of the NPs in EtOH (Figures S10-S12).       In terms of textural properties (Figures 13b, 14b and 15b), the sorption isotherms obtained with the EtbipyPMO series belong to type IV, as found for the other series studied in this work. In this particular case, the saturation plateaus were very flat, which shows that these materials are purely mesoporous, without side adsorption in voids. This is confirmed by the absence of hysteresis loops for these PMOs, which suggests that they are  In terms of textural properties (Figures 13b, 14b and 15b), the sorption isotherms obtained with the EtbipyPMO series belong to type IV, as found for the other series studied in this work. In this particular case, the saturation plateaus were very flat, which shows that these materials are purely mesoporous, without side adsorption in voids. This is confirmed by the absence of hysteresis loops for these PMOs, which suggests that they are not aggregated. The mesopore volume decreased from 41% down to 27% for increasing fractions of ethoxybipyridine moieties. At the same time, the surface areas obtained were lower compared to PMO 1 NPs, but also compared to other PMOs studied in this work. This could be due to the greater thickness of the pores' walls as a consequence of the condensation of the ethoxybipyridine moieties. Additionally, the average mesopore size was found to slightly increase for growing fractions of ethoxybipyridine moieties (from 2.8 nm up to 3.4 nm).
The carbon dioxide sorption isotherms obtained with this series of PMOs strongly depended on the fraction of Etbipyridine fraction ( Figure 16, Table 6). This trend was even more pronounced compared to the iPrbipyridine series. Indeed, at low pressure, the slopes were completely different, regardless of temperature, and the uptake at 10 5 Pa was also different. Intriguingly, iPrbipyridine PMO 10 showed a greater affinity compared to the other PMOs, however this feature was not present at 273 K. Two materials showed a poor CO 2 capture capacity compared to PMO 1, namely the Etbipyridine 6 and Etbipyridine 10 PMOs. On the other hand, Etbipyridine PMO 15 exhibited better sorption capacities than PMO 1. We further derived the adsorbed amount per square meter to evidence the efficiency of these materials. The adsorbed amounts were found to be 1.2 µmol·m −2 , 2 µmol·m −2 , 4 µmol·m −2 , and 2.5 µmol·m −2 for PMO 1 NPs, EtbipyPMO 6 NPs, EtbipyPMO 10 NPs, and EtbipyPMO 15 NPs, respectively, which shows that carbon dioxide interacts with Etbipyridine moieties more than with PMO 1 NPs.  Despite the similarity of the chemical structures between the iPrbipyridine and the Etbipyridine precursors, the CO2 adsorption is favored (as for the aminopyridine) for a large Etbipyridine molar fraction (15%), and for a low iPrbipyridine molar fraction (6%). Our results compare well with some of the studies which use amine-modified porous silicas for CO2 capture and the field has been comprehensively reviewed [40,41]. Note that these studies concern bulk materials. For instance, in our case, at 1 atm, the adsorbed CO2 amount obtained for the best materials is around 1 mmol/g which corresponds to 45 mg/g.   Despite the similarity of the chemical structures between the iPrbipyridine and the Etbipyridine precursors, the CO 2 adsorption is favored (as for the aminopyridine) for a large Etbipyridine molar fraction (15%), and for a low iPrbipyridine molar fraction (6%).

Selection of the Best PMOs for the CO 2 Capture
Our results compare well with some of the studies which use amine-modified porous silicas for CO 2 capture and the field has been comprehensively reviewed [40,41]. Note that these studies concern bulk materials. For instance, in our case, at 1 atm, the adsorbed CO 2 amount obtained for the best materials is around 1 mmol/g which corresponds to 45 mg/g. Using APTES-grafted MCM-41, Mello et al. obtained 33 mg/g at 20 • C [42] with APTES-grafted MCM-48; Kim et al. obtained 35 mg/g [43]. However, using MCM-41 with larger mesopores grafted with diethanolamine, 104 mg/g of CO 2 could be adsorbed at very low pressure (0.05 atm) [44].
Finally, according to the different types of synthesized PMO NPs, the highest CO 2 adsorption values were obtained at 298 K for: Compared to the 100% BTEE (0.81 mmol·g −1 ), higher CO 2 adsorption capacity was observed with 29% (iPrbipyPMO 6 NPs), 17% (EtbipyPMO 15 NPs) and 14% (pyPMO 15 NPs), which showed the strong impact of the polycondensation between the BTEE and the precursor containing aminopyridine groups, amines having an important affinity with the CO 2 molecule. On the one hand, the best CO 2 adsorption values were obtained with the bipyridine and the Etbipyridine precursors, proving the importance of the presence of two aromatic cycles with nitrogen for CO 2 capture. On the other hand, CO 2 adsorption was favored with a large aminopyridine (or Etbipyridine) molar fraction (15%), and with a low iPrbipyridine molar fraction (6%). In terms of CO 2 /PMO affinity, these three materials exhibited different features, as can be seen in Figure 17, which presents plots of the isosteric heat of adsorption.
In the case of the pyridine-based material, the enthalpy of adsorption was rather weak, close to the enthalpy of condensation of CO 2 at 298 K, that is 16.7 kJ·mol −1 at room temperature [47]. Furthermore, the shape of the enthalpic curve was rather flat, which indicates that this material is seen as homogeneous by carbon dioxide species. The other materials were more interesting in terms of affinity, as the enthalpy of adsorption was close to −30 kJ·mol −1 at low coverage, which indicates a high CO 2 /PMO interaction. However, the high interaction sites did not dominate the sorption processes, as shown by the fact that the enthalpy curves strongly decrease as coverage increases. This is especially true for Etbipyridine-based PMO, whereas in the case of the iPrbipyridine analogue, there is an inflexion in the curve (located at θ = 0.2) which suggests two sorption regimes. At very low coverage, sorption is likely to mostly happen on the iPrbipyridine moieties, while at higher coverage, sorption proceeds on the whole surface of the material. To investigate these observations, we performed two sorption cycles using the procedure described in the experimental section ( Figure 18). However, between the two cycles, iPrbipy PMO 6 was subjected to a re-activation stage at 60 • C for 2 h under vacuum. The two sorption isotherms were found to be close to one another, however with~15% difference at 10 5 Pa. This difference can therefore be accounted for by the interaction sites of iPrbipy PMO 6. In the case of the other materials, for which less interaction was found, it can be anticipated that consecutive sorption isotherms would lead to closer sorption isotherms. the bipyridine and the Etbipyridine precursors, proving the importance of the presence of two aromatic cycles with nitrogen for CO2 capture. On the other hand, CO2 adsorption was favored with a large aminopyridine (or Etbipyridine) molar fraction (15%), and with a low iPrbipyridine molar fraction (6%). In terms of CO2/PMO affinity, these three materials exhibited different features, as can be seen in Figure 17, which presents plots of the isosteric heat of adsorption. In the case of the pyridine-based material, the enthalpy of adsorption was rather weak, close to the enthalpy of condensation of CO2 at 298 K, that is 16.7 kJ·mol −1 at room temperature [47]. Furthermore, the shape of the enthalpic curve was rather flat, which indicates that this material is seen as homogeneous by carbon dioxide species. The other materials were more interesting in terms of affinity, as the enthalpy of adsorption was close to -30 kJ·mol −1 at low coverage, which indicates a high CO2/PMO interaction. However, the high interaction sites did not dominate the sorption processes, as shown by the fact that the enthalpy curves strongly decrease as coverage increases. This is especially true for Etbipyridine-based PMO, whereas in the case of the iPrbipyridine analogue, there is an inflexion in the curve (located at θ = 0.2) which suggests two sorption regimes. At very low coverage, sorption is likely to mostly happen on the iPrbipyridine moieties while at higher coverage, sorption proceeds on the whole surface of the material. To investigate these observations, we performed two sorption cycles using the procedure described in the experimental section ( Figure 18). However, between the two cycles, iPrbipy PMO 6 was subjected to a re-activation stage at 60 °C for 2 h under vacuum. The two sorption isotherms were found to be close to one another, however with ~15% difference at 10 5 Pa. This difference can therefore be accounted for by the interaction sites of iPrbipy PMO 6. In the case of the other materials, for which less interaction was found, it can be anticipated that consecutive sorption isotherms would lead to closer sorption isotherms. It is therefore interesting to discuss the properties of these PMOs not only in terms of their sorption capacity, but also in terms of their sorption efficiency, as these two aspects of the sorption process lead to the choice of the appropriate material. Where traces of CO2 are concerned, a high affinity is required, whereas if large amounts of CO2 must be ad- It is therefore interesting to discuss the properties of these PMOs not only in terms of their sorption capacity, but also in terms of their sorption efficiency, as these two aspects of the sorption process lead to the choice of the appropriate material. Where traces of CO 2 are concerned, a high affinity is required, whereas if large amounts of CO 2 must be adsorbed, large CO 2 sorption capacities should prevail.

Synthesis of BTEENPs
For CO 2 adsorption, PMO NPs were synthesized using the sol-gel procedure. A total of 250 mg of CTAB as the surfactant (6.86 × 10 −4 mol), 437 µL of NaOH solution (2 mol·L −1 ), and 333 µL of 1,2-bis(triethoxysilyl)ethane (or BTEE, Figure 1) (9 × 10 −4 mol) were dissolved in 60 mL of deionized water. The mixture was stirred for 2 h at 80 • C/750 rpm. The reaction mixture was cooled to room temperature and centrifuged for 20 min at 20,000 rpm. The CTAB surfactant was removed by 45 min sonication at 40 • C, with a 30 mL solution of ammonium nitrate (NH 4 NO 3 ) in EtOH (6 g·L −1 ), followed by one water wash and two EtOH washes (30 mL each). BTEENPs were dried in a vacuum oven at 70 • C.

DLS Analysis
Dynamic light scattering (DLS) is a non-destructive analysis technique for measuring the hydrodynamic diameter of particles suspended in a liquid. Analysis of the scattering of light from the laser of the device by particles was performed with the Cordouan Technologies DL 135.

DLS Analysis
Dynamic light scattering (DLS) is a non-destructive analysis technique for measuring the hydrodynamic diameter of particles suspended in a liquid. Analysis of the scattering of light from the laser of the device by particles was performed with the Cordouan Technologies DL 135.

FTIR Analysis
Fourier transform infrared spectroscopy (FTIR) is a non-destructive analysis technique based on the absorption of electromagnetic radiation between 2.5 and 25 µm (wavelength between 400 and 4000 cm −1 ). Absorptions in this zone correspond to the movement (vibrations, rotations, etc.) of organic functional groups and can be used to deduce the chemical bonds and the structural details of the material. Analyses were performed using the Spectrum Two FT-IR Spectrometer (Perkin Elmer, Waltham, MA, USA).

TEM Analysis
Transmission electron microscopy (TEM) is an analytical technique which enables observation of the local pore arrangement of a material by means of the interactions that occur when an electron beam accelerated by a high voltage travels through the material. For TEM characterization, the nanoparticles of the material were dispersed in EtOH and carefully deposited on a copper grid with porous carbon films. The JEOL 1400 Plus (120 kV) microscope was used to record the image size, the shapes and the pores of the synthetized nanoparticles.

Textural Analysis
Prior to the sorption measurements, the nanoparticles were evacuated under a secondary vacuum at 80 • C for 8 h. Because our materials are hybrid materials, the organic moieties in the materials cannot withstand high activation temperatures. We optimized the activation temperature, and 80 • C was found to be acceptable. The nitrogen sorption isotherms showed that the mesoporosity is accessible to nitrogen, which supports our choice.
The sorption isotherms were obtained at 77 K using a Micromeritics TriStar device. The specific surface area of the various materials was derived using the BET (Brunauer-Emmet-Teller) method, taking 0.162 nm 2 as cross sectional area for nitrogen. T-plot analysis of the sorption isotherms revealed that the prepared materials were not microporous, which allowed the use of the BET model for deriving the specific surface areas. The pore size distributions were derived using the BJH (Barrett-Joyner-Helenda) approach on the desorption branches of the sorption isotherms, starting from p/p • = 0.95 downwards. The pore geometry was assumed to be cylindrical.

RMN Analysis
Nuclear magnetic resonance (NMR) spectroscopy is an analytical technique for determining the structure of an organic molecule, exploiting the magnetic properties of certain atomic nuclei, such as hydrogen, carbon, or silicon. Solid 13 C-NMR analysis was carried out on 50 mg of NPs using a VARIAN 300 MHz (Wild Bore) solid spectrometer (3.2 mm MAS probe) device.

Elemental Analysis
Elemental analysis is a qualitative and quantitative technique for revealing, by means of combustion or pyrolysis, the elemental composition of an organic compound and therefore the mass percentage of elements such as carbon, hydrogen, nitrogen, sulfur and oxygen. Elemental analysis was carried out using an Elementar Vario Micro Cube device with a sample of 15 mg.

CO 2 Adsorption Measurement
The CO 2 adsorption isotherms were measured using a Micromeritics 3Flex device. Two sorption temperatures were investigated, namely 273 K and 298 K, in order to provide the isosteric heat of adsorption. Prior to the sorption measurements, the materials underwent a degassing stage at 80 • C for 8 h under a secondary vacuum.