Wojciech Kukulka*a,
Krzysztof Cendrowskia,
Beata Michalkiewiczb and
Ewa Mijowska*a
aNanomaterials Physicochemistry Department, West Pomeranian University of Technology, Szczecin, Piastów Av. 45, Szczecin 70-311, Poland. E-mail: wojciech_kukulka@zut.edu.pl; emijowska@zut.edu.pl
bInstitute of Chemical and Environment Engineering, West Pomeranian University of Technology, Szczecin, Pulaskiego St. 10, Szczecin 70-322, Poland
First published on 12th June 2019
In our study we prepared MOF-5 derived carbon to reveal the thermodynamics of CO2 absorption processes in great detail. Porous carbon material was prepared from a metal–organic framework (MOF-5) via carbonization at 1000 °C. The obtained structure consists only of carbon and exhibits a BET specific surface area, total pore volume and micropore volume of 1884 m2 g−1, 1.84 cm3 g−1 and 0.59 cm3 g−1, respectively. Structural analysis allowed the assumption that this material is an ideal candidate for efficient CO2 absorption. The CO2 uptake was 2.43 mmol g−1 at 25 °C and 1 bar. Additionally, the absorption over a wide range of temperatures (25, 40, 60, 80 and 100 °C) and pressures (in range of 0–40 bar) was investigated. It is shown that the CO2 absorption isotherm fits a multitemperature Sips model. The calculated Sips equation parameters allows the isosteric heat of adsorption to be obtained. The isosteric heat of adsorption for CO2 decreased substantially with an increase in surface coverage by gas molecules. This indicates a negligible intermolecular interaction between CO2 molecules. A decrease in the isosteric heat of adsorption with surface coverage is a result of the disappearance of favourable adsorption sites.
Many different types of materials have already been tested in this regard. Zeolites3–8 and porous carbon materials9–14 were of particular interest. Recently, another group of potential candidates for absorption of gases has appeared – metal–organic frameworks. They have been examined for the absorption of the abovementioned gases.15,16 The first reports on the absorption of gases in MOFs was published already in the middle of the last decade.17–19 Since then, various MOF structures have been investigated for gas absorption. Until now, the absorption properties have been tested for such metal–organic materials as MOF-5,20–22 MOF-177,23–25 ZIF-8,26,27 IRMOF,28,29 HKUST,30,31 MIL-100/101 (ref. 31–34) and UiO-66/Zr-MOF.35,36 The most interesting property of metal–organic structures is the ease to obtain highly porous carbon materials after simple carbonization. This enables a number of new materials to be tested for various applications, including gas absorption. Metal–organic structures are often unstable and prone to damage, for example as a result of prevailing humidity in the air. So, carbonization is a very simple solution to obtain a stable material under various conditions. The carbonization of MOF also can generate highly porous carbon products and hence the high specific surface area and total pore volume.37 The optimization of the carbonization parameters of metal–organic structures is widely studied in the literature.38 The first absorption measurements of hydrogen, methane and carbon dioxide in the carbonized metal–organic structures were also performed. The absorption measurements of different gases for carbonized MOF-1,39 MOF-5,40 MIL41–43 and ZIF44–49 structures were performed. The detailed data are presented in Table 4.
Carbonized MOF structure shows the advantage over other carbon structures, like amorphous carbon, with the uniform structure and control over the MOF structures properties with the synthesis parameters. Recent articles showed that with control of the synthesis and carbonization parameters, different properties can be tuned depending on the needs.38,50 The MOF-5 structure are not without disadvantages. The main disadvantages are the stability of the MOF-5 in the presence of even trace amounts of moisture and the compounds toxicity during synthesis. Both of this problems can be eliminated by carbonization of the MOF-5 structure directly after synthesis. The MOF-5 structure shows high surface but after exposition to the trace amounts of moisture, their crystal structure starts to decomposed.51 In results of that specific surface area of MOF-5 structures drops radically. The thermal transformation of MOF-5 to the MOF-5 derived carbon structures allows to maintain the pristine crystals size and shape with similar surface area. After carbonization obtained MOF-5 derived carbon structures are immune to water and shows stability at higher temperature.52,53 Second disadvantage was the high toxicity of DMF used for the synthesis of MOF-5 and due to this high cost of the material production. Our recent publication showed that DMF used for MOF-5 synthesis, after separation from obtained structure and by-products can be reused.54 Additionally, recent presented data show recovering and synthesis MOF structures from terephthalic acid from polyethylene terephthalate (PET) waste like used plastic bottles.54 Since MOF-5 can be synthesized from DMF and PET waste the MOF productivity problem can be resolved.
Detailed research on the mechanism of MOF-5 carbonization, were previously reported elsewhere.38,55,56 In our previous research, TGA analysis was performed in the temperature range from 25 to 1000 °C under inert gas flow (argon).38 The observed weight loss were assigned to the: removal of water and residual solvent molecules (in the range from 25 to 200 °C); thermal decomposition of organic ligand molecules and the formation of CO2 and benzene after the breaking of carboxylic bridges between benzene rings and Zn4O clusters (in the range from 400 to 550 °C),57 carboreduction of ZnO, in which carbonaceous materials deoxidize ZnO and later evaporate forming mainly CO2 and CO. This process starts at about 750 °C and intensify with the temperature.38 The effect of zinc oxide nanostructure formation (spherical and rod-like structures), growth and thermal extraction induces cracking and cavity formation in the carbonized structure. Such structural distortions may reduce the surface area of the whole microporous structure, causing a deterioration of the physical properties of the carbonized MOF-5. Cracking and the appearance of voids take place regardless of the shape of zinc oxide nanostructures; however; the concentration of defects in the carbonized MOF-5 increases with nanorod formation.38 This negative effect can be minimalize by long thermal carbonization at high temperatures.38,58
In this work, we focus on the absorption properties of carbon dioxide by the carbonized MOF-5 structure. The thermodynamics of the absorption process in our material has also been investigated in great details. Therefore, we could fill the gap in current state of the art – there is a lack of report on it.
Fig. 1 SEM images of MOF-5 (A and B) and carbonized MOF-5 (D and E) and particle size distribution of MOF-5 (C) and carbonized MOF-5 (F). |
Additionally, Raman spectroscopy confirms the presence of carbon in the sample. As shown in Fig. 2F, the Raman spectrum shows two strong peaks at around 1300 and 1600 cm−1 corresponding to the D and G bands, respectively. The D band is ascribed to the vibration of carbon atoms with dangling bonds in the plane with termination by disordered graphite. The G band indicates the E2g mode in carbon with high graphitization degree and it is related to the vibration of sp2-hybridized carbon atoms. The D band has much higher intensity than G band which suggests that carbonized MOF-5 have a lot of defects and it consist mostly of amorphous carbon.
The crystal structure of MOF-5 before and after carbonization was investigated by X-ray diffraction (XRD) (Fig. 3A). All reflections in XRD pattern of MOF-5 before carbonization can be attributed to the reference standard card (CCDC – 256965). The reflections at 2θ angle of ∼7°, 9.8°, 13.8° and 15.6° correspond to the (002), (022), (004), and (024) planes, respectively.59 There are no significant peaks corresponding to zinc oxide in XRD pattern of MOF-5 after the carbonization. The carbonized sample shows broad peaks between 20°, 25° and at ∼45°, related to the disorderly oriented carbon with low crystallinity. The presented XRD data are in good accordance with morphology analysis demonstrating: (i) metal species extraction and (ii) high purity of carbon material obtained during the carbonization process.
Fig. 3 Pristine and carbonized MOF-5 XRD spectrum (A), TGA (B), nitrogen adsorption/desorption profile (C) and pore volume distribution in carbonized MOF-5 (D). |
Furthermore, thermogravimetric analysis was performed in the temperature range from room temperature to 1000 °C under inert argon gas flow and presented in the Fig. 3B. A small weight loss assigned to the removal of water and residual solvent molecules is observed in the temperature from 100 to 200 °C (up to 2 wt%). Next weight loss starts around 385 °C and ends around 600 °C. It is related to the thermal decomposition of carbonized MOF-5 and the formation of carbon dioxide. The weight loss was ∼90 wt% at 1000 °C which indicates that some residual metal species are still present in the sample.
The porosity of the pristine and carbonized MOF-5 was tested by the N2 adsorption–desorption experiment. The typical IV isotherms with H3 hysteresis loops are observed in the samples, which is typical of mesoporous materials (Fig. 3C). The hysteresis loop for pristine MOF-5 is not clearly seen because the porosity of this sample is considerably lower than the porosity of carbonized MOF-5. The pore size distribution curves show the coexistence of micro- and mesopores below 10 nm in both samples but the micropore and mesopore volumes are much higher for carbonized MOF-5. The calculated BET specific surface area, total pore volume and micropore volume was 1884 m2 g−1, 1.84 cm3 g−1 and 0.59 cm3 g−1, respectively. The specific surface area, total pore volume and micropore volume estimated for pristine MOF-5 were equal to 477 m2 g−1, 0.33 cm3 g−1 and 0.24 cm3 g−1, respectively. During carbonization the porous structure was built up.
Based on the above results we can conclude that during carbonization some kind of activated carbon was obtained. XRD and Raman results confirmed formation of amorphous carbon. N2 adsorption–desorption measurements showed that highly porous carbon was obtained during carbonization. The changes of textural properties of MOF-5 are presented at Fig. 3C and D.
Fig. 4 CO2 adsorption isotherms in carbonized MOF-5 (points – experimental data, lines – calculated using the Sips equation). |
The carbon dioxide adsorption isotherms were fitted using Freundlich, Langmuir, Sips and Toth empirical equations. More detailed information about these models can be found elsewhere.60,61
The adsorption isotherm data were fitted to the isotherm models by non-linear regression method. The reduced chi-squared was applied to test the how the models fit to experimental data. The smaller the reduced chi-squared value, better fitting is assumed. Basing on it, it was found that the Sips model provided the most accurate fit to the CO2 adsorption data. The Sips model is also called the Langmuir–Freundlich model because it contains the elements of Langmuir and Freundlich models. Sips model is similar to Freundlich model, but it has a finite limit of the sufficiently high pressure:
(1) |
Table 1 shows the Sips isotherms parameters and their errors and reduced chi-sqr and adjusted R-square (R2).
Temp. [K] | qm [mmol g−1] | b [1/bar] | n | Chi-sqr | R2 | |||
---|---|---|---|---|---|---|---|---|
Value | Error | Value | Error | Value | Error | |||
298 | 44.6 | 2.7 | 0.0147 | 0.0023 | 0.672 | 0.013 | 0.0097 | 0.9998 |
313 | 42.5 | 5.2 | 0.0124 | 0.0037 | 0.694 | 0.025 | 0.0241 | 0.9993 |
333 | 35.1 | 5.1 | 0.0111 | 0.0061 | 0.706 | 0.041 | 0.0324 | 0.9991 |
353 | 30.3 | 6.2 | 0.0110 | 0.0051 | 0.759 | 0.038 | 0.0399 | 0.9990 |
373 | 28.6 | 2.5 | 0.0098 | 0.0017 | 0.806 | 0.017 | 0.0025 | 0.9998 |
The parameter b is a measure of strongness of adsorbate and sorbent interaction. The values of parameter b decreased with the increase in the temperature. Therefore, the adsorbate and sorbent interaction is weaken when the temperature is elevated.
The parameter n is a measure of heterogeneity of the adsorbate and sorbent system. If the n is equal 1 the system is homogenous and the Sips equation is reduced to the Langmuir equation. The lower is n parameter the more heterogeneous is the system. In our study the values of n parameter increased along with the temperature. It is indication that the heterogeneity of our system increased at higher temperature. The changes of b and n parameters were in agreement with the multitemperature Sips equation. The temperature dependence of the affinity constant b is the following:54
(2) |
The temperature dependence of the exponent n is the following:54
(3) |
The saturation capacity qm can be consider as a constant. However, it can also following the temperature dependence:62
(4) |
In order to calculate the parameters in (2)–(4) equations, they were linearized and plotted:
ln(qm) = f1(T) | (5) |
ln(b) = f2(1/T) | (6) |
n = f3(1/T) | (7) |
Fig. 5 show plots of functions: ln(qm) = f1(T), ln(b) = f2(1/T), n = f3(1/T). Table 2 present the results of the linear fitting. The slope is denoted as a1 and the intercept as a0.
Function | a1 | a0 | R2 | ||
---|---|---|---|---|---|
Value | Error | Value | Error | ||
ln(qm) = f1(T) | 5.73 | 0.22 | −0.00646 | 0.00066 | 0.9597 |
ln(b) = f2(1/T) | −6.10 | 0.27 | 547 | 88 | 0.9027 |
n = f3(1/T) | 1.310 | 0.094 | −193 | 31 | 0.9034 |
On the basis on the values listed in Table 2 the optimal parameters for the temperature dependent Sips equations ((2)–(4)) were calculated (Table 3). T0 was the reference temperature that was equal to 298 K as the lowest temperature set in the sorption investigations of carbonized MOF-5.
Parameter | Value | Error |
---|---|---|
Q [J mol−1] | 4550 | 68 |
b0 [1/bar] | 0.0141 | 0.0058 |
n0 | 0.662 | 0.035 |
α | 0.648 | 0.026 |
qm0 [mmol g−1] | 45.0 | 2.2 |
χ | 1.925 | 0.016 |
As it is presented, the multitemperature Sips model (eqn (1)–(4)) provided a very successful fit of the adsorption observed in the experimental data. Therefore, the calculated Sips equation parameters can be used for isosteric heat of adsorption calculation. Isosteric heat of adsorption (Qis) is one of the basic requirements for the characterization and optimization of the adsorption process. It is defined as the heat of adsorption determined at constant surface coverage (θ). A surface coverage is defined as:
(8) |
The isosteric heat of adsorption was calculated using the Clausius–Clapeyron eqn (9):
(9) |
The calculated parameters of Sips equation (Table 1) were used to calculate the pressure values at different temperatures, at constant surface coverage from (10):
(10) |
The logarithm of the equilibrium pressures (ln(p)) was plotted against the reciprocal temperature (1/T) at the constant coverage θ (Fig. 6). The surface coverage was varied from 0.01 to 0.07 with interval of 0.01.
The isosteric heat of adsorption was calculated utilizing the slopes (A) of the linear functions ln(p) = f(1/T):
Qis = −RA | (11) |
The isosteric heat of adsorption for CO2 decreased substantially with increase in the gas surface coverage (Fig. 7). Such course of the curve indicates the negligible intermolecular interaction between CO2 molecules. A decrease in the isosteric heat of adsorption with the surface coverage is a result of the disappearance of favorable adsorption sites.
Carbon dioxide molecules preferred to adsorb onto the high-energy sites. Increasing the coverage caused adsorption onto the sites of low-energy which results in a slow increase in the amount of adsorbed vs. pressure. This was also in agreement with the slope of adsorption isotherm (Fig. 4).
Table 4 summarize the detailed review of the data, including our data (2012–2018), on carbonized metal–organic frameworks studied as material for CO2 absorption. Many reports describe CO2 absorption results measured at 273 K (0 °C). As the temperature decreases, the absorption results increase significantly. The same applies to the use of different pressure units. Using an atmosphere or millimeters of mercury raises the result in comparison to the bars. The result then looks better (conversion factor given in the title of the Table 4). However, it is certainly necessary to unify the results in order to reliably compare them. In addition, the data including measurements at room temperature and higher temperatures seem to be more suitable for discussions on the potential in industrial application of obtained materials for CO2 absorption. The wide range of measured pressures and the thermodynamics of the absorption process are also valuable for testing of the potential applications of the materials. Taking these factors into account, the obtained result are among the best of materials based on carbonized metal–organic frameworks structures described in the literature. Materials based on ZIF and their modifications44–49 were most often discussed in the literature. Ma et al. obtained four different samples of carbonized MOF-5 at different time.40 They carried out carbonization at 600, 700, 800 and 900 °C for 5 h. Additionally, their materials were doped with nitrogen by adding urea prior carbonization. Our results of CO2 uptake (2.43 mmol g−1 at 25 °C and 1 bar) are basically very similar. This was possible by increasing the carbonization temperature to 1000 °C which led to more effective zinc oxide evaporation process38 and favorable development of porosity. It was realized in 2 h carbonization without any additional modifications. However, doping with nitrogen and activation of carbonaceous materials (for example using KOH) is a popular method for improving their electrochemical and absorption properties. Therefore, we do not rule out further investigation of the obtained carbonized MOF-5 after its further treatment. In this work we focused primarily on the description of the precise thermodynamics of the CO2 absorption process and the presentation of results in a wide range of temperatures and pressures of the pristine carbonized MOF-5.
Sample name | Original material | CO2 uptake experiment parameters | Result (mmol g−1) | Source |
---|---|---|---|---|
BM-900 | bio-MOF-1 | 273 K and 1 bar | 4.62 | 39 |
298 K and 1 bar | 3.55 | |||
KBM-700 | bio-MOF-1 | 273 K and 1 bar | 4.75 | 39 |
298 K and 1 bar | 3.29 | |||
MUC600 | MOF-5 | 0 °C and 1 bar | 3.55 | 40 |
25 °C and 1 bar | 2.44 | |||
25 °C and 0.15 bar | 0.73 | |||
MUC900 | MOF-5 | 0 °C and 1 bar | 3.71 | 40 |
25 °C and 1 bar | 2.31 | |||
25 °C and 0.15 bar | 0.43 | |||
AAC-2W | MIL-100(Al) | 273 K and 1 bar | 6.5 | 41 |
298 K and 1 bar | 4.8 | |||
298 K and 0.2 bar | 1.74 | |||
C800 | MIL-100(Al) | 273 K and 1 atm | 4.1 | 42 |
298 K and 1 atm | 2.6 | |||
NC800 | MIL-100(Al) | 273 K and 1 atm | 5.7 | 42 |
273 K and 0.15 atm | 2.3 | |||
298 K and 1 atm | 3.8 | |||
N-HPCMs-5-0.6-973 | Al-based composite | 273 K and 780 mmHg | 2.35 | 43 |
298 K and 780 mmHg | 1.82 | |||
C700 | ZIF-8 | 273 K and 1 bar | 3.70 | 44 |
298 K and 1 bar | 2.76 | |||
C1000 | ZIF-8 | 273 K and 1 bar | 4.64 | 44 |
298 K and 1 bar | 3.39 | |||
C700W | ZIF-8 | 273 K and 1 bar | 5.51 | 44 |
298 K and 1 bar | 3.80 | |||
NC900 | ZIF-8 | 273 K and 1 atm | 5.1 | 45 |
298 K and 1 atm | 3.9 | |||
AC-CB700 | ZIF-8 | 25 °C and 1 bar | 2.0 | 46 |
1000 | ZIF-8 | 25 °C and 0.15 bar | 0.99 | 47 |
25 °C and 1 bar | 3.22 | |||
25 °C and 20 bar | 10.21 | |||
C68 | ZIF-68 + FA | 273 K and 1 atm | 4.76 | 48 |
C69 | ZIF-69 + FA | 4.54 | ||
C70 | ZIF-70 + FA | 5.45 | ||
C68 | ZIF-68 + FA | 298 K and 1 atm | 4.00 | 48 |
C69 | ZIF-69 + FA | 3.86 | ||
C70 | ZIF-70 + FA | 4.49 | ||
CZIF8a | ZIF-68 + FA | 273 K and 1 atm | 4.04 | 49 |
CZIF68a | ZIF-69 + FA | 4.49 | ||
CZIF69a | ZIF-70 + FA | 4.76 | ||
Pristine MOF | MOF-5 | 25 °C and 0.15 bar | 0.26 | This work |
25 °C and 1 bar | 1.30 | |||
Carbonized MOF | CMOF-5 | 25 °C and 0.15 bar | 0.57 | This work |
25 °C and 1 bar | 2.43 | |||
25 °C and 10 bar | 9.73 | |||
25 °C and 20 bar | 13.55 | |||
25 °C and 40 bar | 18.56 | |||
40 °C and 0.15 bar | 0.39 | |||
40 °C and 1 bar | 1.95 | |||
40 °C and 10 bar | 8.09 | |||
40 °C and 20 bar | 11.76 | |||
40 °C and 40 bar | 16.27 | |||
100 °C and 0.15 bar | 0.12 | |||
100 °C and 1 bar | 0.66 | |||
100 °C and 10 bar | 3.86 | |||
100 °C and 20 bar | 6.15 | |||
100 °C and 40 bar | 9.18 |
In order to determine the regeneration performance of the obtained samples, pure and carbonized MOF-5 were subjected to a fifteen cycles of adsorption and desorption. The data for adsorption of carbon dioxide in pristine and carbonized MOF-5 were collected in the temperature of 25 °C and pressure up to 1 bar. As presented in the Fig. 8, the performance of adsorption after fifteen cycles does not change for both materials. It also turns out that the material obtained by carbonization allows adsorption to be 1.87 times higher (at 1 bar) than in the case of pure MOF-5. This analysis indicated that the adsorption of CO2 did not influence the efficiency of MOF-5 and carbon derived MOF-5 structures.
The total pore volume after carbonization increased five times, specific surface area increased four times and micropore volume two and half times. This is the reason that CO2 uptake at carbonized MOF-5 was nearly two times higher than at pristine MOF-5.
Additionally, the absorption in a wide range of temperatures (25, 40, 60, 80 and 100 °C) and pressures (in range of 0–40 bar) was investigated. These temperature range is less popular because it can result in lower values. However, it is at the same time this where the potential applications of technology in the industry have place. The points on the absorption chart were experimental data, and the isotherms were calculated using the Sips equation. In this work, thermodynamics of the absorption process was also precisely described. This gives a broader perspective on the potential use of material in industrial practice.
This journal is © The Royal Society of Chemistry 2019 |