Junwen Wangab,
Yichao Linb,
Qunfeng Yue*a,
Kai Taob,
Chunlong Kong*b and
Liang Chen*b
aCollege of Chemistry & Chemical Engineering, Harbin Normal University, China. E-mail: qfyue@hrbnu.edu.cn
bInstitute of New Energy Technology, Ningbo Institute of Material Technology and Engineering, China. E-mail: kongchl@nimte.ac.cn; chenliang@nimte.ac.cn
First published on 25th May 2016
N-doped porous carbon has a wide range of applications in many fields, such as gas adsorption and super-capacitor, which has stimulated active research for developing efficient strategies to fabricate N-rich porous carbon. In the present study, a series of N-rich porous carbons are derived from polyamine-incorporated metal–organic framework materials (MOFs). Results show that the N content of as-prepared porous carbon is greatly increased by loading polyethyleneimine (PEI) into ZIF-70 frameworks because the loaded PEI can adsorb onto the pore walls of the ZIF-70 and simultaneously supply carbon and N sources during the carbonization process. As a result, the surface area of the as-prepared sample is also increased. In addition, the N content of the porous carbon can be tuned by PEI loadings and carbonization temperature. The as-prepared N-rich porous carbons exhibit greatly enhanced CO2-selective adsorption capacity compared to ZIF-70 and porous carbon derived from ZIF-70. Here the CO2 capture capacity of the as-prepared N-rich porous increases with increasing N content due to considerable interaction affinity between doped N and CO2 molecules. Thus, the as-prepared porous carbon with N content of 11.08 wt% displays high CO2 uptake of 4.86 mmol g−1 at 0 °C and 1 bar, albeit it has a moderate surface area of 652 m2 g−1. Moreover, the N-rich porous carbon clearly shows an excellent separation performance for CO2-over-N2 and CO2-over-CH4. Overall, polyamine-incorporated MOFs are an efficient strategy for controllable fabrication of N-rich porous carbon. The resulting products display a high CO2-selective capture performance. It should be noted that, to achieve the optimal CO2 capture ability, a comprehensive optimization of the polyamine-MOFs-derived porous carbon should be performed.
In the last years, a series of solid porous adsorbents such as amine-modified silicas,8 zeolites,9 porous carbons,10 and MOFs11 have been developed for CO2 capture. Among them, porous carbon is one of the most attractive and important materials,12 which has a wide range of applications in many fields, such as adsorption and separation,13 super-capacitor and fuel cells.14 Thus, different strategies such as chemical vapor deposition, laser ablation, electric-arc technique and template method have been explored to generate porous carbon with controlled pore structure.15 To the best of our knowledge, the template method has been considered as one of the most effective approaches to synthesize porous carbon with controllable architecture and desirable physical and chemical properties.16 Therefore, a variety of template materials such as zeolites,17 mesoporous silicas,18 colloid crystals19 and aluminium oxide membranes have been developed for porous carbon materials synthesis. However, this traditional template method usually involves template removal, which limits its practical application, because the resulting porous carbon structure could be partially destroyed during the process of template removal using acid or alkali.20 In addition, this process is unfriendly for the environmental development.
Recently, research on MOFs-derived porous carbon has received considerable attention because of their high porosity and the ability to control their pore textures.21 In addition, their organic ligands can serve as a precursor and contribute to the formation of high-quality porous carbon.22 More importantly, the textural properties and application performances of the resulting porous carbons can be tuned by properties of the parental MOFs.23 For instance, Xu et al.24 reported that nano-porous carbons derived from NH4OH-incorporated ZIF-8 show high CO2 uptakes. However, porous carbons obtained directly from carbonization of MOFs only show moderate CO2 capture capacity as well as low selectivity.25 Indeed, it has been found that the properties of porous carbons can be tuned by integrating heteroatom (i.e., N, B, P) to access a wide range of applications including heterogeneous catalysis, energy storage and gas adsorption and separation.26 Thus, incorporation of heteroatoms with Lewis basic properties into porous carbons can simultaneously enhance their CO2 capture capacity and selectivity due to considerable interaction between Lewis base and CO2 molecules.25
In a previous study, we have successfully fabricated N-rich polyamine-MOF porous materials by incorporation of polyethyleneimine (PEI) molecules into MOF frameworks.27 The resulting composite materials exhibit a moderate porosity and very competitive CO2 capture properties.28 On the basis of this study, we propose that N-rich porous carbons with high CO2 capture capacities can be facilely derived from PEI-incorporated MOF materials,29 as shown in Fig. 1. Here linear PEI (with an average molecular-weight of 300) and ZIF-70 are chosen as the model system to fabricate polyamine-MOF composite materials. ZIF-70 is chosen because of its high porosity and large pore size (∼1.6 nm),30 so that many PEI can be readily incorporated into its framework pores. Moreover, ZIF-70 framework has many N elements, which can be partly doped into carbon network during carbonization process.31
Therefore, both PEI and ZIF-70 can work as carbon and N precursors to generate N-rich porous carbon materials. The resulting porous carbons are expected to exhibit a very competitive CO2 adsorption capacity owing to the high content of N doped into their framework. In this work, a series of N-rich porous carbons are fabricated by tuning the content of PEI loadings. In addition, the effect of carbonization conditions on the textural properties and the CO2 capture capacities are also studied. The CO2, N2 and CH4 adsorption properties of the as-prepared N-rich porous carbons are evaluated at 0 °C and 25 °C, respectively. Results show that porous carbons derived from PEI-loaded ZIF-70 display much higher CO2 adsorption capacity than prepared directly by ZIF-70. Moreover, the CO2 capture capacities of the as-prepared porous carbons increase with their N content increasing.32
The adsorption isotherms of CO2, N2 and CH4 were measured using volumetric technique by an apparatus from Micromeritics America (ASAP-2020). The amount of gas adsorbed as a function of pressure was determined using the Langmuir–Freundlich fit for the isotherms:
Here, Q and Qm are the uptake and the maximum uptake, respectively. P is the equilibrium pressure, B and t are the equation constants. To obtain the exact pressures P, corresponding to constant amount of gas adsorbed, the above equation can be rearranged to:
The CO2/N2 and CO2/CH4 adsorption selectivities (α) of all samples were calculated by the following equation, according to a previous report:34
Fig. 3 shows the XRD patterns of the as-prepared ZIF-70, PEI loaded ZIF-70 and porous carbon. Clearly, the structure of ZIF-70 synthesized in this work matches well with that reported in the literature.35 The powder XRD patterns of ZIF-70 and PEI loaded ZIF-70 samples show that their Bragg diffraction angles indicating that ZIF-70 structure is well maintained after PEI loading. However, the peaks below 10° of PEI-loaded ZIF-70 almost disappear after PEI loading due to the filling of ZIF-70 pores of PEI. Apparently, the intensity of peaks above 10° of PEI-loaded ZIF-70 remains intact. As for the porous carbon, it can be seen that two broad diffraction peaks at 2θ values of about 25° and 44° corresponding to carbon (002) and (101) diffractions, respectively, can be found clearly from the wide-angle XRD patterns. Note that, results indicate that most of obtained carbon species are amorphous carbons because the peak (101) of graphitic crystallinity is very weak. In addition, no diffraction peaks of ZIF-70 can be observed in Fig. 3C, indicating that carbonization of ZIF-70 is complete during the carbonization process.
![]() | ||
Fig. 3 XRD patterns of the as-prepared ZIF-70 (A), PEI-loaded ZIF-70 (B) and porous carbon (C) derived from PEI–ZIF-70 composite. |
To evaluate pore structure of the resulting porous carbon, the N2 adsorption–desorption isotherms of the samples are measured at 77.3 K. Here three PEI-loaded ZIF-70 samples, corresponding to 25 wt%, 50 wt% and 100 wt%, are prepared for porous carbon preparation. The carbonization temperatures are in the range of 700–900 °C. The resulting samples are denoted as C-0%-900, C-25%-900, C-50%-900, C-100%-900, C-100%-800 and C-100%-700, respectively. As shown in Fig. 4, all samples exhibit a type I mixed with type IV behaviours. As for KOH activated porous carbon, many microporous can be generated between the mesoporous pores during activation process, resulting in surface area increasing greately.36 Generally, the isotherm at low relative pressure (P/P0 < 0.3) is attributed to N2 adsorption in micropores. The hysteresis of desorption isotherm at relative high pressure reveals the existence of mesopores in the sample. It should be noted that the presence of micropore and mesopore depends on the carbon gasification and high evaporation etching of the Zn.37 Here the N2 uptakes of samples prepared at 900 °C increase distinctly with PEI loading increasing. This is ascribed to the fact that loaded PEI could adsorb onto pore wall of the ZIF-70. Thus, it can supply carbon source during carbonization process, reducing the sacrifice of template framework (ZIF-70). Actually, a similar phenomenon has been observed in literatures.24 The corresponding BET surface area and pore volume for all samples are calculated from the isotherms and summarized in Table 1. Clearly, the surface area and pore volume of porous carbon samples increase with PEI loading increasing. On the other hand, N2 uptakes of the samples decrease distinctly with carbonization temperature decreasing. This could be explained because the loss of carbon source at lower temperature is lower. It has been found that KOH can clean the slag on the surface of the porous carbon to release some extra pores and expand some micropores. Therefore, KOH are further used to activate the C-100%-900 sample. The N2 adsorption analysis indicates that both BET surface area and pore volume are greatly increased. Above results show that PEI loading, carbonization temperature and activation process greatly affect the pore structure of PEI–ZIF-70 derived carbon materials.
Samples | Surface areaa (m2 g−1) | Pore volumeb (cm3 g−1) | N contentc (wt%) |
---|---|---|---|
a The specific surface area is calculated in the P/P0 range of 0.05–0.2.b Values at P/P0 = 0.99.c Obtained from XPS analysis. | |||
C-0%-900 | 158.17 | 0.10 | 1.23% |
C-25%-900 | 445.97 | 0.22 | 4.23% |
C-50%-900 | 523.46 | 0.24 | 7.37% |
C-100%-900 | 652.44 | 0.33 | 11.08% |
C-100%-800 | 437.90 | 0.20 | 6.45% |
C-100%-700 | 415.70 | 0.23 | 5.56% |
C-100%-900-KOH | 2107.50 | 1.09 | 2.65% |
The nature of N moieties on the as-prepared porous carbon surface is analyzed by XPS. Fig. 5 shows the N 1s XPS spectra of representative samples. For all samples, the XPS N 1s spectra can be deconvoluted into three peaks with the binding energies centered at 398.3, 400.1, and 401.5 eV, respectively. The peak at 398.3 eV is attributed to the pyridinic-N (N-6), while the peaks at 400.1 and 401.5 eV can be assigned to pyrrolic-/pyridonic-N (N-5) and quaternaryN (N-Q), respectively. Note that, the peaks (N-6, N-5, and N-Q) of N species is very weak after KOH activation, which is consistent with the reported ref. 38. Quantitative analysis reveals that N present in the form of N-5 is higher than that present in the form of N-6 and N-Q, in the order of N-5 > N-6 > N-Q for all N-doped samples. This result is beneficial to CO2 capture, because it has been found that N-5 generally has a much greater contribution to CO2 capture than N-6 and N-Q.39 The N compositions of the as-prepared porous carbons are calculated and summarized in Table 1. Apparently, the N contents of the samples prepared by PEI-loaded ZIF-70 are much higher than that prepared directly by carbonization of ZIF-70. For instance, the N content of porous carbon prepared by carbonization of PEI-loaded ZIF-70 can reach 11.0 wt%. However, the N content of porous carbon prepared directly by carbonization of ZIF-70 is only 1.32 wt%. Results clearly indicate that PEI loading is an effective strategy for improving the N content of MOFs-derived porous carbons. In addition, the N contents of the samples increase with PEI loading increasing. In this work, the N contents of the samples increase with carbonization temperature increasing. It might be ascribed to the fact that much more carbon loss happens at higher temperature. However, the N contents of the sample activated by KOH is greatly decreased, because a large number of N species on the carbon surface would decompose during the activation process.36
It should be noted that the sample activated by KOH shows higher CO2 capture capacities in high test pressure (>0.4 bar), whereas it shows a lower CO2 capture capacities at low test pressure (<0.4 bar), compared to unactivated sample. It is ascribed to the lower N content and higher surface area after KOH activation. The CO2 adsorption properties of the samples prepared at different temperatures are shown in Fig. 6b. Clearly, their CO2 adsorption capacities increase with carbonization temperatures increasing because the higher N content and surface area can be achieved at higher temperature (see Table 1). For instance, the CO2 capture capacity of C-100%-800 is distinctly higher than that of C-100%-700 due to its higher N content, albeit their surface areas are slightly different. It should be noted that, the highest CO2 adsorption difference between C-100%-800 and C-100%-700 appears at around 0.4 bar. It might suggest that the affinity between doped N and CO2 molecular are moderate. The CO2 adsorption difference between C-100%-900 and C-100%-700 increases with the test pressure increasing because both N content and BET surface area of C-100%-900 are higher than those of C-100%-700. In brief, it indicates that the N contents of porous carbons greatly affect their CO2 adsorption capacity. By the comparison of C-100%-900 and C-100%-900-KOH, apparently, C-100%-900 displays higher CO2 adsorption capacities at low adsorption pressure (<0.6 bar) due to its much higher N content, albeit it has much smaller BET surface area than that of C-100%-900-KOH (see Table 1). This result further indicates that the affinity between doped N and CO2 are moderate, which can greatly improve CO2 adsorption capacity at low pressure. In high adsorption pressure, C-100%-900-KOH exhibits higher CO2 adsorption capacity than that of C-100%-900 due to its much higher surface area. The CO2 adsorption capacities of samples at low temperature (0 °C) are also investigated, as shown in Fig. 7. It can be seen that the CO2 adsorption capacities of all samples increase greatly, compared with that of 25 °C. The highest CO2 uptakes of sample can reach 5.4 mmol g−1 at 1 bar and 0 °C. The CO2 uptakes of samples at 0.15 bar (0 °C, 25 °C) and 1 bar (0 °C, 25 °C) are calculated and summarized in Table 2. As shown in Fig. 8 and 9, N2 and CH4 uptakes of all samples are further measured at 0 °C and 25 °C, respectively. Clearly, the N2 and CH4 uptakes are much lower than CO2 at both temperatures. Since N-doped porous carbon can interact moderately with CO2 molecules because doped N has Lewis basic property.
Sample | CO2 uptakes (mmol g−1) | |||
---|---|---|---|---|
0.15 bar | 1 bar | |||
0 °C | 25 °C | 0 °C | 25 °C | |
C-0%-900 | 1.17 | 0.81 | 2.70 | 2.08 |
C-25%-900 | 1.60 | 1.04 | 3.97 | 2.51 |
C-50%-900 | 2.13 | 1.42 | 4.77 | 3.21 |
C-100%-900 | 2.20 | 1.36 | 4.86 | 3.56 |
C-100%-800 | 1.65 | 1.13 | 4.13 | 2.62 |
C-100%-700 | 1.49 | 0.99 | 3.53 | 2.44 |
C-100%-900-KOH | 1.73 | 0.75 | 5.46 | 4.40 |
Isosteric adsorption enthalpies as a function of the quantity of gases adsorbed were calculated via variant of the Clausius–Clapeyron equation. As show in Table 3, we calculate the heats of adsorption of all samples. Here the increase of carbonization temperature, the heats adsorption of samples also will increase. In addition, the heats adsorption of C-100%-900 is much more than that of C-0%-900, which proves a point that the amount of CO2 adsorbed correspondently increases to that of N. It is indicated that the CO2 adsorption capacity of N-rich materials is superior to that of usual carbon materials, because of its quadrupole interactions with N in N-doped carbon material in addition to the van der Waals interactions. Hence, the heats adsorption of CO2 supports the key role of the N-doping for stronger interactions with CO2 molecules in the material.
Sample | Heats of adsorption (kJ mol−1) |
---|---|
C-0%-900 | 19.97 |
C-25%-900 | 26.61 |
C-50%-900 | 26.10 |
C-100%-900 | 26.97 |
C-100%-800 | 25.93 |
C-100%-700 | 25.00 |
C-100%-900-KOH | 15.22 |
In addition, the quadrupole moment of CO2 (−14 × 10−40 C m2) is much higher than that of N2 (−4.6 × 10−40 C m2) and CH4 (nonpolar). The synergetic effects of the two factors lead to higher uptakes of CO2 than N2 and CH4. Here the samples show higher adsorption capacity for CH4 over N2, which is probably ascribed to the larger molecular size of CH4. The adsorption results of CO2, N2 and CH4 indicate that the as-prepared samples exhibit excellent CO2-selective adsorption ability. In addition, separation performance of the as-prepared N-rich porous carbons increase with N content increasing. The separation performances of the as-prepared samples for CO2-over-N2 and CO2-over-CH4 mixtures are further corroborated by breakthrough experiments performed at 25 °C, as shown in Fig. 10. Results apparently show that N2 and CH4 elutes rapidly through the column, whereas CO2 only starts to elute after a period of time. Since CO2 molecules are selectively adsorbed by the as-prepared sample (C-100%-900), which elutes last in the sequence at Fig. 10. Results further indicate that the as-prepared N-rich porous carbon displays an excellent separation performance for CO2-over-N2 and CO2-over-CH4. We noted that there is a time interval that we can get pure CH4 or N2 using the N-rich carbon.
![]() | ||
Fig. 10 Breakthrough curves of C-100%-900 with an equimolar CO2/CH4 mixture and CO2/N2 mixture at 25 °C. |
This journal is © The Royal Society of Chemistry 2016 |