3-Aminopropyltrimethoxysilane (AP-TMS) was purchased from Aldrich. Hexadecyltrimethylammonium chloride (C₁₆TMACl) and tetramethoxysilane (TMOS) were purchased from Tokyo Kasei (Japan). Methanol and 1N sodium hydroxide solution were purchased from Wako Pure Chemical Co (Japan). All materials were used as received.

Amino propy-modified nanoporous silica particles were obtained according to the literature [22]. In a typical synthesis, 3.52 g of C₁₆TMACl and 2.28 mL of 1 M sodium hydroxide solution were dissolved in 800 g of methanol/ water (50/50 = w/w) solution. A mixture of 1.19 g (7.81 mmol) of TMOS and 0.16 g (0.87 mmol) of AP-TMS (AP-TMS/(TMOS + AP-TMS) = 10 mol%) were then added to the solution with vigorous stirring at 298 K. After the addition of the TMOS and APTMS, the clear solution suddenly turned opaque and resulted in a white precipitate. After 8 h of continuous stirring, the mixture was aged overnight. The white powder was filtered and washed with distilled water at least three times, and then dried at 45˚C for 72 h. The powder obtained was heated in a 60 ml of ethanol solution containing a 1 ml of concentrated hydrochloric acid at 333 K for 3 h to remove the surfactant. Then, the powder was filtered, washed several times with ethanol, and dried at 318 K. Then, the amino propyl modified material was treated with an ammonia solution to remove any residual Cl⁻ ions and to neutralize the protonated amines in the sample [22]. A 0.35 g of the modified sample was suspended in a 20 ml of methanol solution containing a 1 ml of ammonia solution (28%) at room temperature for 8 h. The solid was recovered by filtration, washed with methanol, and finally dried in a vacuum at 423 K for 12 h.

Scanning electron micrographs (SEMs) were obtained with a SU3500 (Hitachi). Powder X-ray diffraction measurement was carried out with a Rigaku Rint-2200 X-ray diffractometer using Cu-Kα radiation. Nitrogen adsorption/ desorption isotherms were measured using a BELSORP-mini II (Bell-Japan) at 77 K. Samples were evacuated at 423 K under 0.13 Pa before the measurement. Pore volume was estimated from the amount of adsorbed nitrogen at the relative pressure of 0.95. Pore diameter was calculated by the BJH method for the adsorption branch. CO₂ adsorption isotherms were measured using a BELSORP-MAX-12-N-T-HL at different temperatures. N elemental analyses (EA) were carried out on an Elementer varioEL elemental analyzer.

The mobility of amino moieties was evaluated by pulsed NMR measurements that measure relaxation time of ¹H which is one of the main constituent elements of amino moieties. The pulsed NMR measurements were performed with JEOL-JNM-MU25A spectrometer operating at 25 MHz for protons (¹H) in the phase-sensitive detection mode. Spin-spin relaxation time (T₂) was measured for ca. 1 g of sample in a 10 mmf sample tube [29]. Samples were evacuated at 423K under 0.13 Pa before the measurement to remove adsorbed water. The pulse sequences for T₂ measurements were the solid echo pulse sequences (90˚_xτ90˚_y: 90˚ pulse width = 2 μs, τ = 8 μs, pulse repetition time = 4 - 10 s, cumulated number = 64 - 128). The measurements were conducted from 30˚C to 130˚C with 20˚C interval. The data for the decay process of the traverse magnetization decay signal M(t) is analyzed as the sum of the one Gaussian component for short T₂ and one single exponential component for long T₂ as expressed by Eq. (1):

M ( t ) = A S exp [ − ( t T 2 S ) 2 ] + A L exp [ − t T 2 L ] (1)

where A_S and A_L are the amplitude for the short and long component. T_2S and T_2L are the corresponding T₂ values. The fraction of short and long component, f_S and f_L are determined from A_S and A_L values.

Nanoporous silica samples obtained by co-condensation using 10, 30, and 50 mol % of AP-TMS are denoted as AP10, AP30, and AP50, respectively, according to the AP-TMS ratio in the synthesis. The content of nitrogen increases with increasing the ratio of AP-TMS, and results are listed in Table 1. Nitrogen adsorption-desorption isotherms of samples are shown in Figure 1. A steep increase between P/P₀ of 0.2 and 0.3 indicates the existence of mesopores in AP10. However, AP30 and AP50 are found to adsorb much less nitrogen than AP10. Physical properties of samples are also summarized in Table 1. Specific surface area, pore size and pore volume of AP10 are comparable to that in ref 20. Those values are much smaller for AP30 and AP50. Powder XRD measurements were conducted and corresponding XRD patterns are shown in Figure 2. A sharp peak around 2˚ ascribed to d₁₀₀ periodicity of mesopores is observed for AP10. The pattern is typical of mesoporous silica. No peak is observed for AP30 and AP50. This indicates that no regular pore structure exists in AP30 and AP50.

Figure 3 shows SEM images of samples. All particles are in sub-micron range between 0.41 μm and 0.68 μm.

CO₂ adsorption isotherms of samples at different temperatures are shown in Figure 4. As for AP10, adsorbed amount of CO₂ decreases with increasing

temperature. This tendency is quite normal. In contrast, the amount of CO₂ adsorbed increases with increasing temperature for AP30 and AP50 in some temperature range. The amount is the largest at 60˚C and decreases at higher temperature for AP30. This behavior is non-thermodynamic and does not follow the Clausius-Clapeyron equation. As for AP30, the pores are mostly fulfilled with amino moieties, and a few spaces exist at lower temperature. In this case, only a limited amount of CO₂ is captured. However, once amino moieties are more mobile and active at higher temperature, CO₂ molecules could easily penetrate into inside pores to be captured. The optimum temperature for AP50 is 80˚C which is higher than the temperature for AP30. Since much more amount of AP molecules are contained in pores of AP50, higher temperature would be needed for amino-moieties to be mobile.

It is found that CO₂ adsorption behavior changes drastically by adjusting the amount of amino moieties in nanoporous silica. AP10 adsorbs more CO₂ at lower temperature while AP30 and AP50 adsorb more CO₂ at higher temperature. It is interesting that the CO₂ adsorption behavior is highly affected by the ratio of AP loaded.

It is mentioned that amino-moieties of AP30 or AP50 are more mobile at higher temperature. To investigate the effect of temperature on the mobility of amino moieties in samples, the pulsed NMR measurements were conducted for AP10 and AP30 that show different temperature dependence in CO₂ adsorption. Normalized T₂ decay curves at different temperatures are shown in Figure 5. A curve is deconvoluted to a short and long component according to Equation (1). Table 2 summarizes spin-spin relaxation time for short (T_2S) and long (T_2L) component and corresponding fractions (f_S, f_L) for AP10 and AP30. A short relaxation time means that molecules are rigid, and long relaxation time reveals molecules are more mobile.

As for AP10, T_2S and T_2L are almost the same value against temperature, indicating that the mobility of the AP molecules unchanged. However, the fraction of rigid component, f_S, decreases and the more mobile component, f_L, increases with increasing temperature. The behavior can easily be understood by plotting the data (Figure 6). F_S is 52% at 30˚C and the value decreases to 38% at 130˚C.

The exact opposite tendency is observed for AP30. F_S and f_L remain unchanged against temperature, and T_2L increases drastically with increasing temperature. The f_S:f_L ratio is almost 74:26 and the ratio remains unchanged upon temperature increase. The f_S of AP30 is much larger than that of AP10, implying that the amount of rigid AP molecules is much larger for AP30. T_2S slightly increases from 11 μs to 16 μs whereas T_2L increases from 32 μs to 97 μs. The increase of the relaxation time for the mobile component is significant. T_2S for the rigid component of AP30 is much smaller than that of AP10, suggesting that many of AP molecules in AP30 are confined more tightly. This could be caused

by the incorporation of larger amount of AP molecules into nanopores. Conversely, T_2L of AP30 at higher temperature is bigger than that of AP10 although the values are lower at temperatures lower than 90˚C. Therefore, the increase of the amount of CO₂ adsorbed at higher temperature could be related to the increase of the mobility of AP molecules. Although the optimum temperature for CO₂ adsorption is different from the result obtained in the NMR experiments, the difference could be caused by the difference in atmosphere, with or without CO₂, or the length of time for the measurement.

Since the fraction and mobility of soft components in AP10 changed little, the diffusion of CO₂ was not affected by the temperature. As a result, CO₂ adsorption occurred thermodynamically. In contrast, the mobility of soft components in AP30 increased drastically with increasing temperature, leading to non-thermodynamic adsorption behovior.