The QE-TPDA profiles of nonane, observed for high-silica zeolites ZSM-5 and Y using the new system, are close to those recorded with the standard one [4] (Fig. 2). The high-temperature desorption maxima and adsorption minima correspond to emptying and filling of the structural micropores in both zeolitic frameworks, respectively. The low-temperature maxima result from desorption from the external surface and interparticle mesopores. The fact that the new profiles are shifted to higher temperatures should be explained in terms of Le Chatelier's principle, taking into account the higher saturator temperature and, consequently, the higher partial pressure of nonane in the carrier gas.
The adsorption isobars, calculated by averaging of the integral adsorption and desorption curves, calculated from the new profiles exhibit similar behavior as those calculated from the old profiles. However, the saturation sorption capacities based on the new isobars are underestimated by about 20%. The reason for this systematic discrepancy is not clear; for both series of experiments (i.e., performed in the old and new QE-TPDA systems), quantitative interpretation was based on the calibration procedure described earlier. However, these problems may be overcome by the application of well-defined microporous materials (e.g. the commercial zeolites) as standard reference materials.
The QE-TPDA profiles of nonane recorded for the selected ordered mesoporous silicas SBA-15 are shown in Fig. 4. They consist of single adsorption minima, resulting from capillary condensation in the mesopores, and desorption maxima corresponding to emptying of the mesopores. As these effects are very intensive and appear in the low temperature range, which is quite narrow (20–80°C). For this reason only low heating/cooling rates may be used in the QE-TPDA measurements for mesoporous materials (e.g. 2°C/min – much lower than 10°C/min usually used in the case of zeolites). Both the new and the old ones conform to the rule ‘the smaller the pores, the higher the desorption temperature’, since for SBA-15 silicas the increase in the hydrothermal treatment results in a larger pore size.
In addition to the qualitative interpretation of the QE-TPDA profiles of nonane, related to the pore sizes, a quantitative method of the pore size analysis is available. The pore size distributions, calculated according to the modified BJH scheme [9] from the desorption parts of the new QE-TPDA profiles of nonane nonane obtained for the studied SBA-15 silicas, are shown in Fig. 5, together with the PSDs calculated from desorption branches of the low-temperature sorption isotherms (shown in Fig. S5), reported earlier [15]. Good agreement of the corresponding PSDs may be noticed. Although QE-TPDA-based pore sizes (determined as positions of the PSDs’ maxima) are systematically underestimated by about 10%, they show a very good correlation with the N2 desorption-based values (with the R2 value above 0.99). This is consistent with previous findings concerning a good correlation of the pore sizes obtained from the QE-TPDA profiles of nonane, measured using the standard system, with those based on N2 sorption data [15].
The new QE-TPDA system was also used to characterize sorption of water vapor in selected MOFs. The framework of Al-fumarate MOF (Al-fum) is similar to that of MIL-53(Al), containing terephthalate anions as linkers. The framework of Al-fum forms 1D system of micropores with 0.57 × 0.60 nm2 openings [19]. Unlike MIL-53 (Al), Al-fum does not exhibit “breathing” behavior upon dehydration and rehydration [19]. The structure of Al-fum synthesized in this study was verified by means of XRD (Fig. S6) and its adsorption properties toward water – in measurements of the adsorption-desorption isotherms, with the use of the standard manometric system (Fig. 6). The QE-TPDA profiles of water observed for Al-fum are consistent with the adsorption isotherms. The main desorption maximum observed in the temperature range 48–70°C (for the heating rate of 2°C/min) corresponds to the first step in the desorption isotherm, observed in the relative pressure range of 0.2–0.3, and should be attributed to the emptying of the structural micropores. The fact that the main desorption peaks present in the QE-TPDA profiles recorded for different heating rates partially overlap may indicate occurring of a sorption driven phase transition, in which partial pressure is the only intensive parameter. Such transitions, similar to capillary condensation and evaporation, have been reported to take place in the microporous carbons [24]. The low-temperature desorption peak clearly corresponds to the second step in the desorption isotherm (p/ps -0.6-0.9) and should be attributed to adsorption on the external surface and/or in the interparticle mesopores, if they are present.
It should be mentioned that the QE-TPDA profiles in Fig. 7 were actually obtained from the relative humidity signal, originating from the SHTC3 sensor. It has been found that the profiles of partial pressure in the carrier gas, calculated from both signals of the detector (either the RH signal or the thermal conductivity signal from STC31), are practically the same (Fig. S8). Since the RH signal exhibited a better signal-to-noise ratio, it was chosen as a source of data in the QE-TPDA of water.
The new system was also used in modified water QE-TPDA experiments aimed at characterizing the hydrothermal stability of the MOFs studied. The sample temperature consisted of numerous cycles, comprising slow heating in the temperature range of 30–80°C (during which ‘diagnostic’ thermodesorption maxima were recorded), isothermal heating equivalent to hydrothermal treatment, and prolonged cooling to the ambient temperature (when readsorption occurred). After each 3-cycle sequence, the temperature of the hydrothermal treatment was increased by 25°C, starting from 200°C (Fig. S9). The entire hydrothermal test consisted of 31 cycles and lasted 5 days.
The results of the hydrothermal stability test of Al-fum are shown in Fig. 8. Only small changes in the intensity of the low-temperature desorption peak were observed up to 325°C: small increases in the range of 200–275°C and gradual decreases in the range of 275-325C. These changes may be attributed to the degradation of the surface features that contribute to the high external surface. The onset of a decrease in the intensity of the high-temperature desorption peak, indicating damage of the structural micropores, was noticed at 350°C, but only after the treatment at 400°C the micropore volume was considerably decreased.
Another MOF investigated in this study was MIL-88A. Its framework, based on trimers of Fe(III) octahedra connected by fumarate linkers, forms a 3D system of interconnected channels and cages with sizes of 0.5–0.7 nm [26]. The XRD pattern obtained for MIL-88A synthesized in this study (shown in Fig S7) contains all main reflexes of the literature patterns [27, 28], thus confirming the presence of this framework. The low intensity of the XRD data may result from the low crystallinity of the material (Fig. S9). Additional reflexes at 2θ = 9 and 11 ° can indicate the presence of another phase, which could not be identified.
Both the isotherms (Fig. 9) and QE-TPDA profiles (Fig. 10) reveal complex patterns of adsorption and desorption of water, with four distinct steps and substantial hysteresis. This behavior is different from that reported by Luo et al. [29], who observed a three-step adsorption isotherm, with a considerably lower saturation sorption capacity of 340 mg/g. Although MIL-88A has been known since 2005, we could not find any other reports on water sorption in this material. Although the origin of the multistep adsorption and desorption is not clear, the shape of the QE-TPDA profiles recorded at different heating and cooling rates, which practically do not contain any overlapping fragments, suggests a pore filling mechanism.
The results of the hydrothermal stability test of MIL-88A are shown in Fig. 11. It was performed in similar way as in the case of Al-fum, only the initial temperature of the hydrothermal treatment was much lower (125°C). The first notable change appeared after heating at 175°C, when the smallest high temperature peak (visible at 60–70°C) disappeared. This may be related to the presence of another phase indicated by the XRD results. The material retained most of its adsorptive properties after the hydrothermal treatment at 200 and 225°C. Some minor symptoms of hydrothermal damage could be observed after heating at 250°C, but further increase of the temperature to 275°C resulted in a complete loss of the adsorption properties due to structure degradation.