Gerardo A. González-Martínez‡
a,
J. Antonio Zárate‡a,
Ana Martínez*a,
Elí Sánchez-Gonzáleza,
J. Raziel Álvareza,
Enrique Limaa,
Eduardo González-Zamora*b and
Ilich A. Ibarra*a
aLaboratorio de Fisicoquímica y Reactividad de Superficies (LaFReS), Instituto de Investigaciones en Materiales, Universidad Nacional Autónoma de México, Circuito Exterior S/N, CU, Del. Coyoacán, 04510, Ciudad de México, Mexico. E-mail: argel@unam.mx; egz@xanum.uam.mx; martina@unam.mx; Fax: +52-55-5622-4595
bDepartamento de Química, Universidad Autónoma Metropolitana-Iztapalapa, San Rafael Atlixco 186, Col. Vicentina, Iztapalapa, C. P. 09340, Ciudad de México, Mexico
First published on 10th May 2017
CO2 capture of MIL-53(Al) was enhanced by confining MeOH and i-PrOH within its micropores. Compared to MIL-53(Al), results showed an approximately 1.3 fold increase in CO2 capture capacity (kinetic isothermal CO2 adsorption experiments), via confining small amounts of both alcohols. Adsorption–desorption properties are investigated for MeOH and i-PrOH and the enthalpy of adsorption, for MeOH and i-PrOH, was measured by differential scanning calorimetry (DSC): ΔH = 50 and 56 kJ mol−1, respectively. Regeneration (CO2 adsorption–desorption cycles) of the sample MeOH@MIL-53(Al) exhibited a loss on the CO2 capacity of only 6.3% after 10 cycles and the desorption is accomplished by only turning the CO2 flow off. Static CO2 adsorption experiments (at 196 K) demonstrated a 1.25-fold CO2 capture increase (from 7.2 mmol g−1, fully activated MIL-53(Al) to 9.0 mmol g−1, MeOH@MIL-53(Al)). The CO2 enthalpy of adsorption for MIL-53(Al) and Me@OHMIL-53(Al) were estimated to be ΔH = 42.1 and 50.3 kJ mol−1, respectively. Computational calculations demonstrated the role of the hydrogen bonds formed between CO2 molecules and confined MeOH and i-PrOH molecules, resulting in the enhancement of the overall CO2 capture.
Focusing on gas oversolubility (or gas enhancement by confining solvents), Luzar and Bratko5 showed by computational simulations (molecular dynamics, MD) an increase of N2 and O2 solubility in water (from 5-fold to 10-fold) under confinement conditions in hydrophobic mesopores. By confining n-hexane, CHCl3, EtOH and H2O in mesostructured materials (MCM-41, SBA-15 and silica aerogel), remarkable H2 solubility enhancements, were reported by Pera-Titus and co-workers.1,6 Pellenq7 confined N-methyl-2-pyrrolydone (NMP) in MCM-41 and demonstrated a 6-fold increase in CO2 solubility. Interestingly, Llewellyn8 reported a 5-fold increase in CO2 capture by confining water in a mesoporous MOF material entitled MIL-100(Fe). This striking CO2 increase was achieved when approximately 40 wt% of H2O was pre-adsorbed in the mesopores of MIL-100(Fe).
In all the previous examples only mesoporous materials showed gas oversolubility properties. Interestingly, for microporous materials this phenomenon has not been observed as thoroughly demonstrated by Llewellyn and co-workers8 in a MOF microporous material UiO-66. Thus, in order to referring to gas oversolubility it is necessary to incorporate (pre-adsorb) considerably high amounts of solvent (e.g. H2O, n-hexane and EtOH) prior any gas adsorption. For example, Farrusseng et al.9 remarkably reported a 22-fold improvement on the H2 uptake (at 298 K and 30 bar) on MIL-101(Cr) by confining n-hexane. By a solvent wet impregnation method,9 they loaded n-hexane into the fully activated MOF material and this corresponded to the 60% of the pore volume. In other words, 60% of the pore volume of MIL-101(Cr), was filled with n-hexane.
However, it is possible to confine small amounts of solvents in micropores materials to enhance gas adsorption properties. Certainly, this phenomenon cannot be referred as gas oversolubility since there is not “enough” solvent to solubilise gas molecules. Walton et al.10 exhibited that small amounts of water can enhance CO2 capture in microporous MOF materials. These small amounts of H2O can interact with selected functional groups within the pores of these materials. In particular, they demonstrated that hydroxyl (–OH) functional groups act as a directing agent for water molecules inside the pores allowing a more efficient and ordered packing of H2O.11 In addition, Yaghi12 showed that these functional groups (–OH), significantly improve the affinity of MOF materials to water.
Previously, we have demonstrated that small amounts of pre-adsorbed H2O into microporous MOF materials considerably enhanced their CO2 capture properties.13 Additionally, our research group started to investigate the confinement of alcohols in MOFs to increase their CO2 capture capabilities and this was exemplified with the pre-adsorption of small amounts of EtOH in InOF-1.14 Therefore, we continue with the development of hybrid adsorbent MOF materials (via confining small amounts of alcohols within their micropores) which can contribute to new CO2 capture technologies.15 Among ‘the twelve principles of CO2 chemistry’ that Poliakoff16 proposed, CO2 capture constitutes one of these challenges (maximise integration). Therefore, the search for post-synthetically modified MOFs with high structural stability, adsorption capacity, solvent stability, fast sorption kinetics and mild regeneration conditions, is nowadays a very hot research field.17,18
Herein, we report the augmented CO2 capture properties of the microporous material MIL-53(Al), first reported by Serre and co-workers,19 upon confining small amounts of alcohols (methanol and isopropanol), together with MeOH and i-PrOH adsorption–desorption properties of MIL-53(Al).
Fig. 1 Methanol adsorption isotherm at 30 °C of MIL-53(Al) from %P/P0 = 0 to 85. Solid circles represent adsorption, and open circles show desorption. |
Later, a new calcined MIL-53(Al) sample was activated (as previously described) and an isopropanol adsorption–desorption isotherm was performed from %P/P0 = 0 to 85 at 30 °C (Fig. 2). To the best of our knowledge, this is the first time that the adsorption properties of isopropanol (i-PrOH) are described in MIL-53(Al). From 0 to 10 (%P/P0) the uptake of i-PrOH was rapidly increased (indicative of favourable host–guest interactions) achieving an i-PrOH uptake of approximately 2.5 mmol g−1 (15.3 wt%). Then, from %P/P0 = 10 to 85 a much slower (but constant) alcohol uptake was observed with a maximum of approximately 3.8 mmol g−1 (23.2 wt%). A relatively strong hysteresis was observed in the desorption phase (Fig. 2, open symbols). This hysteresis occurred mainly in the low pressure range from %P/P0 = 0 to 10. As in the previous case (MeOH), the kinetic diameter of i-PrOH is 4.6 Å which is considerably smaller than the pore openings of activated MIL-53(Al), (2.6 Å × 13.6 Å, ht form),19b,c suggesting again, the formation of hydrogen bonds with the bridging hydroxo functional groups (μ2-OH).19e
Fig. 2 Isopropanol adsorption isotherm at 30 °C of MIL-53(Al) from %P/P0 = 0 to 85. Solid circles represent adsorption, and open circles show desorption. |
When both alcohols (MeOH and i-PrOH) uptakes in MIL-53(Al) are compared at low loadings (%P/P0 = 0 to 10), there is a clear preference for i-PrOH over MeOH (2.5 vs. 1.4 mmol g−1, respectively) which can be attributed to a stronger interaction of i-PrOH with the material. In order to confirm this, the enthalpy of adsorption (ΔH) for i-PrOH was experimentally measured by differential scanning calorimetry (DSC) from room temperature to 600 °C (with a ramp of 5 °C min−1). The ΔH value was equal to 56 kJ mol−1 (see Fig. S3 ESI†). When the same experimental determination was performed for MeOH, the ΔH value was equal to 50 kJ mol−1 (see Fig. S4 ESI†), in good correlation with the alcohol uptakes (low loadings). When the total uptake of these alcohols is compared (MeOH = 8.1 mmol g−1 and i-PrOH = 3.8 mmol g−1), MIL-53(Al) can adsorb considerably more MeOH than i-PrOH, presumably, due to the size of the alcohol molecules (methanol is smaller than isopropanol; kinetic diameters = 3.6 Å and 4.6 Å, respectively).
Fig. 3 Kinetic CO2 uptake experiments performed at 30 °C with a CO2 flow of 60 mL min−1 in MIL-53(Al), (black curve); MeOH@MIL-53(Al), (red curve) and i-PrOH@MIL-53(Al), (blue curve). |
Later, a calcined sample of MIL-53(Al) was activated (180 °C for 2 h and under a flow of N2), cooled down to 30 °C (under N2) and saturated with MeOH (see ESI†). After an activation protocol (see ESI†) the residual amount of MeOH was equal to 2 wt%. The reproducibility of this protocol was confirmed by performing 5 independent experiment (see ESI†). Hereinafter, this sample will be referred as MeOH@MIL-53(Al), [Al(OH)(BDC)]·MeOH0.08. Similar procedure was carried out for sample i-PrOH@MIL-53(Al), [Al(OH)(BDC)]·iPrOH0.05, as it is described in the ESI.†
It was decided to only pre-adsorbed small amounts of alcohols (MeOH and i-PrOH) in MIL-53(Al) samples, based on the investigation of confined H2O in the micropores of MIL-53(Cr),21 (a MOF material which is isostructural to MIL-53(Al)). Then, Paesani and co-workers21 demonstrated by computational infrared spectroscopy that at low water loadings, these water molecules interact strongly with the hydroxo (μ2-OH) functional groups, via hydrogen bonding, which are located inside the pore walls of MIL-53(Cr). Continuing with the investigation of MIL-53, by different calculation methodologies, Haigis22 employed molecular dynamics (MD), to show that H2O molecules can form strong hydrogen bonds with the μ2-OH functional groups, in MIL-53(Cr), as a function of water loading. In addition, Maurin et al.23 corroborated by GCMC computational simulations, in MIL-53(Cr), that at low water loadings, these H2O molecules are regularly accommodated inside all the pores of the material.
Certainly, all the previous computational approaches indicated the role of the μ2-OH functional groups (inside the micropores of MIL-53(Cr)) to “pin” small amounts of water via hydrogen bonding. Complementing to these calculations, we experimentally demonstrated that small amounts of EtOH (2.6 wt%), confined within the micropores of InOF-1 (ref. 14) can: (i) hydrogen-bond to the similar hydroxo functional (In2(μ2-OH)) which was visualised by single crystal X-ray diffraction; and (ii) significantly enhance CO2 capture (2.7 fold).
Therefore, we rationalised the hypothesis of low MeOH and i-PrOH loadings where the micro-porous channels of MIL-53(Al) could efficiently accommodate these alcohols molecules. As a result of this very well ordered positioning of MeOH and i-PrOH, these confined alcohol molecules could help to pack more efficiently CO2 molecules and finally enhance the total CO2 capture.
Then, a kinetic CO2 experiment, at 30 °C, was carried out on the MeOH@MIL-53(Al) sample. The maximum amount of CO2 captured corresponded to 4.7 wt%, which was reached at approximately 4 min and it was constant until the end of the experiment (12 min), Fig. 3 (MeOH@MIL-53). It is important to mention that samples of MeOH@MIL-53(Al) were prepared with anhydrous methanol (<0.005% water) and methanol (reagent alcohol, 95%). The kinetic CO2 experiments showed no difference in the maximum amount of CO2 captured. Additionally, we also tried different small residual-amounts of MeOH: 3%, 4% and 5% and the best result was obtained with 2 wt% of MeOH.
Therefore, the CO2 capture was approximately 1.3-fold improved (from 3.5 wt% to 4.7 wt%), when small amounts of MeOH were pre-adsorbed in MIL-53(Al). Moreover, the 1.3-fold increase was reached at the same time (∼4 min) than the MIL-53(Al) sample, showing that the CO2 adsorption kinetics were highly improved due to the MeOH presence.
Later, the sample i-PrOH@MIL-53(Al) was prepared as previously described (vide supra and see ESI†) where the amount of pre-adsorbed isopropanol was equal to 2 wt%. Fig. 3 exhibits the kinetic CO2 experiment performed on i-PrOH@MIL-53(Al) and the maximum amount of adsorbed CO2 (captured) was equal to 4.5 wt%. This value is slightly smaller to the one observed for sample MeOH@MIL-53(Al), indicating that the pre-adsorption of MeOH in MIL-53(Al) favours the overall capture of CO2. It is worth to emphasise that the confinement of small amounts of both alcohols (2 wt%) enhances the CO2 capture properties of this MOF material.
Long-term regeneration capacity is a fundamental parameter for any CO2 capture material and it is desirable to show very low energy requirements for CO2 release.24 In industrial separation processes this step is typically very expensive and complicated.25 Among many current methodologies to this target, perhaps, the most common is the use of vacuum and temperature swing adsorption. For example, Long and co-workers26 reported a working CO2 capacity (total CO2 adsorption) of ∼7 wt% at room temperature (25 °C) on mmen-CuBTTri. This MOF material was regenerated by switching the flow (15% CO2 in N2) to a pure N2 stream followed by increasing the temperature up to 60 °C. Denayer and co-workers27 reported a total CO2 adsorption of 3.7 wt% for NH2-MIL-53 and the regeneration of this material was performed under purge flow at 159 °C.
In order to evaluate the regeneration properties of MeOH@MIL-53(Al), a new sample was prepared (by confining 2 wt% of MeOH in MIL-53(Al)) and kinetic CO2 adsorption–desorption experiments, at 30 °C, were carried on (Fig. 4). We decided to only evaluate the regeneration properties of MeOH@MIL-53(Al) since it showed the best CO2 capture result. Each cycle consists of an adsorption step (15 min) and a desorption step (15 min), providing a cycling time of only 30 min without the use of N2 purge nor increasing the temperature.
Simply, by turning off the CO2 flow (corresponding to the desorption step) and keeping the adsorption temperature (30 °C), the total regeneration of the MeOH@MIL-53(Al) sample was accomplished. Form the first cycle (4.7 wt% CO2 adsorption) to the tenth cycle (4.4 wt% CO2 adsorption) it was observed a loss of the CO2 capacity (from 4.7 to 4.4 wt%) which represents a loss on the CO2 capacity of only 6.3% (Fig. 4). Although there was a small loss in the CO2 capacity, this result is noteworthy since there is no need to use a purge gas (e.g. N2) and more significantly no thermal re-activation of the sample is required, resulting in a very low cost separation process.
In order to describe the CO2 adsorption properties of MeOH@MIL-53(Al), we performed static (increasing the partial pressure from 0 to 1 bar) and isothermal (196 K) CO2 adsorption experiments on MeOH@MIL-53(Al) samples. It was decided to perform these experiments at 196 K since the adsorption of CO2 at 30 °C (303 K) is problematic due to proximity to the critical temperature of CO2.28 The uncertainly of the δCO2 (density) of the CO2 adsorbed, and since at that temperature the CO2 saturation pressure is extremely high, the range of P/P0 is limited to 0.02 at sub-atmospheric pressures.29 It has been proposed that adsorption in well-defined micropores occurs by a pore-filling mechanism rather than surface coverage.29,30 As an example, N2 molecules at 77 K can fill these micropores in a liquid-like fashion at very low relative pressures (below 0.01). On the other hand, CO2 adsorbed at approximately ambient temperatures (298 or 303 K) can only form a monolayer on the walls of the micropores.30 Therefore, to accomplish pore-filling within the micropores of MOFs and a more accurate description of the CO2 adsorption properties of these microporous materials, CO2 gas adsorption experiments at 196 K are preferred.31
First, a CO2 adsorption experiment at 196 K was carried on a fully activated (180 °C for 1 h and 10−3 bar) sample of MIL-53(Al) exhibiting a total CO2 uptake of 7.2 mmol g−1 (31.7 wt%), (see Fig. 5, MIL-53(Al)). Later, a MeOH@MIL-53(Al) sample was placed in a high-pressure cell (Belsorp HP) and gently evacuated to remove any absorbed moisture. The CO2 uptake was measured from 0 to 1 bar at 196 K and the resultant CO2 adsorption exhibited a characteristic Type-I isotherm with a total CO2 capture of 9.0 mmol g−1 (39.6 wt%), (Fig. 5, MeOH@MIL-53(Al)). Then, at 1 bar and 196 K, the CO2 capture was approximately 1.25-fold increased (from 7.2 to 9.0 mmol g−1) when small amounts of MeOH are present within the micropores of MIL-53(Al). The evaluation of the BET surface area of MeOH@MIL-53(Al) was equal to 762 m2 g−1 with a pore volume of 0.39 cm3 g−1 (lower values than for the fully activated MIL-53(Al), BET = 1096 m2 g−1 and pore volume = 0.56 cm3 g−1). In addition, we measured CO2 adsorption isotherms at 212 K (dry ice and chloroform bath) and 231 K (dry ice and acetonitrile bath), in order to calculate the enthalpy of adsorption (ΔH) by isosteric method for both samples: MIL-53(Al) and MeOH@MIL-53(Al), (see Fig. S5–S8 ESI†). The values were estimated to be 42.1 and 50.3 kJ mol−1 for MIL-53(Al) and MeOH@MIL-53(Al), respectively. Thus, ΔH for CO2 was enhanced when small quantities of MeOH are confined within the material MIL-53(Al).
Fig. 5 Static CO2 adsorption performed from 0 to 1 bar at 196 K on MIL-53(Al), (red circles) and MeOH@MIL-53(Al), (blue circles). |
Fig. 6 Optimised structures of: (A) proposed model to represent the active site; (B) model interacting with MeOH and (C) model interacting with i-PrOH. |
Later, we introduced a single CO2 molecule in all the optimised structures: proposed model (empty) and the models interacting with MeOH and i-PrOH. In Fig. 7 is presented the optimised structures of all the models interacting with CO2. As expected, the single CO2 molecule forms hydrogen bonds in all the structures. It is very well studied that the hydrogen bond distance is an indication on the strength of the interaction between molecules.32 Thus, If we compare the μ2-OH⋯OCO bond distances, it is possible to observe that the structure containing MeOH establishes a shorter hydrogen bond (1.79 Å, Fig. 7A) than the structure with i-PrOH (1.85 Å, Fig. 7B). This slight difference, in the bond length, can actually explain the small dissimilarity in the overall capture of CO2 when small amounts of the alcohols are confined within the MOF material MIL-53(Al), see Fig. 3. Therefore, our computational calculations highlighted the significance of the hydrogen bonds formed between CO2 molecules and confined alcohol molecules: on increasing the strength (diminishing the distance) of the hydrogen bond between the μ2-OH group and the confined alcohol, the overall ability to capture CO2 increases.
Fig. 7 Optimised structures of: (A) proposed model interacting with MeOH and CO2 and (B) proposed model interacting with i-PrOH and CO2. |
Footnotes |
† Electronic supplementary information (ESI) available: TGA data, PXRD data, DSC data, activation protocol, isosteric enthalpy of adsorption data and theoretical calculations. See DOI: 10.1039/c7ra03608f |
‡ These authors contributed equally to this work. |
This journal is © The Royal Society of Chemistry 2017 |