Dan Tiana,
Hong-Zhou Zhangb,
Da-Shuai Zhanga,
Ze Chang*a,
Jie Hana,
Xue-Ping Gao*b and
Xian-He Bua
aDepartment of Chemistry, Collaborative Innovation Center of Chemical Science and Engineering (Tianjin), and Tianjin Key Lab on Metal and Molecule-based Material Chemistry, Nankai University, Tianjin 300071, China. E-mail: changze@nankai.edu.cn; Fax: +86-22-23502458
bInstitute of New Energy Material Chemistry, Nankai University, Tianjin 300071, China. E-mail: xpgao@nankai.edu.cn; Tel: +86-22-23500876
First published on 8th January 2014
In this work, three polyimides (PIs) are successfully synthesized by the condensation reaction of an aromatic triamine with three difunctional aromatic dianhydrides. These materials show electrochemical lithium ion storage properties as cathodes for lithium-ion batteries. The aromatic dianhydride monomers are found to have significant effects on their performances. In addition, the polymer PI-1 shows narrow micropore size distributions and selective adsorption of CO2 over CH4 and N2. Furthermore, PI-1 exhibits a high Qst value (11.7 kJ mol−1) for H2 adsorption.
1,4,5,8-Naphthalenete-tracarboxylicdianhydride (NTCDA) and 3,4,9,10-perylene-tetracarboxylicacid-dianhydride (PTCDA) are well-known organic molecule semiconductors with good crystalline properties.14 They can react with polyamine to generate PIs. The carbonyl groups of PIs based on NTCDA and PTCDA can efficiently conjugate with the aromatic rings and capture Li ions. For instance, Sun's group presented the aromatic carbonyl derivative polymers which showed both high capacity and high cycling stability.5a,5b Zhan et al. reported that PIs can be promising energy-storage materials.8a However, the application of PIs based on NTCDA and PTCDA as electrode materials has been still rarely reported. Herein, we reported three PIs, PI-1, PI-2 and PI-3, which have been synthesized by the condensation reaction of A3-type cross-linker 1,3,5-tris(4-aminophenyl)-benzene (TAPB) with a series of difunctional (B2) aromatic dianhydride (Scheme 1). Their lithium ion storage properties and gas adsorption properties have been investigated. It should be noted that though Wang and co-workers have reported similar polymers based on the reaction of TAPB with 1,2,4,5-benzenete-tracarboxylic anhydride (PMDA) and NTCDA,15 but the Li-ion storage and the CO2/CH4 separation still remain unexplored.
Based on TAPB as central building block, PMDA, NTCDA and PTCDA were used to expand the network of different PIs. All the PIs materials obtained are insoluble in water and common organic solvents such as hexanes, methanol, ethanol, acetone, tetrahydrofuran, and dichloromethane. The structures of polymers were characterized by FT-IR, solid 13C NMR spectra, thermogravimetric analysis, X-ray power diffraction (XRPD), and scanning electron microscope (SEM). Solid-state infrared spectroscopy (IR) was used to validate the chemical structures (Fig. S1, S2 and S3, ESI†). The characteristic bands at 1776, 1730 cm−1 for PI-1 are assigned to the symmetric and asymmetric vibrations of the CO groups. Compared with the signals for carbonyl of the reactant dianhydride, those of the corresponding PI-1 shift toward lower wavenumber, which indicates the formation of polyimide backbone. The typical C–N–C absorption at 1358 cm−1 for PI-1 manifests the generation of the five-membered rings. Meanwhile, the characteristic absorptions at 1717, 1675, 1340 cm−1 for PI-2 and 1701, 1655, 1336 cm−1 for PI-3 provide an evidence for the formation of six-membered imide rings. Moreover, the disappearance of N–H stretching from TAPB in the PIs also confirms the formation of imide bonds.
A detailed investigation of the structures of PIs was further carried out by 13C solid-state NMR spectroscopy (Fig. S4, ESI†). For PI-1, the signal at 165.3 ppm is consistent with the carbonyl carbon of imide ring. The aromatic carbon atoms from the reactants give rise to the signals around 127.0 and 137.7 ppm, respectively. PI-2 shows a peak at 161.7 ppm, corresponding to the carbonyl carbon of imide ring. The peaks around 128.5 and 141.3 ppm can be assigned to the aromatic carbon atoms of the polymer moieties. Similarly, PI-3 shows a carbonyl carbon signal at 162.4 ppm, whereas the aromatic carbon peaks appear at 141.2, 129.4, and 124.0 ppm, respectively.
To confirm the stability of the polymers, thermogravimetric analyses (TGA) of the PIs were performed (Fig. S5, ESI†). The results reveal that all the PIs have high thermal stability with decomposition temperatures above 450 °C. PXRD data suggest that all PIs are amorphous (Fig. S6, ESI†), which is consist with the SEM results for the materials (Fig. 1).
In view of the electrochemical redox properties of PMDA, NTCDA and PTCDA, Li-ion storage for the PIs were investigated in the coin cell by using Li metal as counter and reference electrodes. Fig. 2 shows the electrochemical performance of the samples, measured at the current density of 25 mA g−1. Potential profiles of the first three charge/discharge processes for each polyimide are also listed (Fig. 2a, b and c). In the initial cycle, the discharge capacities of the samples are 61.7, 103.4 and 78.1 mA h g−1, respectively. The initial coulombic efficiencies are 49.4, 92.7 and 90.5%. Obviously, PI-2 and PI-3 show relatively high discharge capacity and high initial coulombic efficiency than PI-1. Meanwhile, there is a relatively flat potential plateau at about 2.35 V (vs. Li/Li+) in the charge/discharge curves of PI-2 and PI-3, while a sloping potential plateau is observed for PI-1.
The CV curves of all these three materials on the coin cells at a scan rate of 0.1 mV s−1 are presented in Fig. S7, ESI.† The CV curve of PI-1 shows two pairs of distinguishable redox peaks in the range of 1.7–2.5 V, which is in good agreement with previous report,8a leading to the sloping potential plateau as shown in Fig. 2. For PI-2 and PI-3 samples, a pair of symmetrical redox peaks centered at ∼2.5 V are observed, in accordance with the appearance of the relatively flat potential plateau in the charge/discharge processes in Fig. 2. In addition, it should be noted that the redox peaks at above 4.0 V are much weak, which can be negligible compared with the large redox peaks in the range of 1.7–2.75 V, especially for PI-2 and PI-3. It implies that these samples may present a good stability without the disintegration of the polymer in the test condition. The CVs indicates that the safer electrochemical window of 1.5–3.0 V shall be set up in future to insure the long cycle stability of PIs. According to the redox mechanism shown in Scheme 2, under ideal conditions, each formula unit will transfer two electrons in each step. However, the capacity delivered from the PIs within 1.5–4.2 V (Fig. 2a, b and c) is far from achieving the theoretical value shown in Table S1 (ESI†). It is demonstrated from both experiment results and theory calculations5c,8a that this kind of polyimides materials can capture Li ions at positions of carbonyl groups and only transfer two electrons in the reversible charge/discharge process. The electrochemical reduction with the second-step two-electron transfer can be obtained by a deep discharging to below 1.5 V. This process could be achieved, however, accompanied with the collapse and passivation of the polymer,8a,8d which could be attributed to the repulsion of injected negative charges between the dianhydride units.
Scheme 2 Electrochemical redox reactions of Li ions with PI polymers based on anhydride functional groups. |
The cycling performances of the samples were investigated as shown in Fig. 2d. During long cycling, the discharge capacity stability of PI-3 is the best. After 65 cycles, the discharge capacity retention of the sample is about 74.1% (57.9 mA h g−1). For PI-2, after 30 cycles, there is only 66.2% (68.5 mA h g−1) of the capacity retention. The capacity of PMDA-based PI-1 decreases gradually with the increase of cycle number. Among all the samples, PI-3 shows the best electrochemical performance. The difference in the capacity of three PIs is usually caused by the diversity of the molecule structure and the active sites of the aromatic dianhydride monomer for lithium insertion/extraction, as well as the current density used in the measurement. For example, there are the same active sites for PI-2 and PI-3. PI-3 with the relatively large molecule structure delivers the low discharge capacity as compared with PI-2 with the relatively small molecule structure.
On the other hand, taking the structural feature of three PIs into consideration, the porosities of PI-1, PI-2 and PI-3 were discussed, N2 adsorption experiments were explored at 77 K (Fig. 3a). The Brunauer–Emmet–Teller (BET) and Langmuir surface areas were obtained as 508 and 685 m2 g−1 for PI-1, and 200 and 307 m2 g−1 for PI-2, respectively. Also, the fit of the adsorption data using the Horvath–Kawazoe method demonstrates the pore width distribution about 5.4 Å for PI-1 and 4.1 Å for PI-2 (Fig. 3b). Subsequently, by analyzing these sorption curves, the higher N2 uptake of PI-1 than PI-2 might be caused by the smaller size of linker PMDA than NTCDA. In contrast, in the case of PI-3, almost no adsorption was observed, which could be ascribed to the interpenetration of networks along with length increasing of the linker.
Due to the higher surface area and microporosity, PI-1 was selected to test its H2 storage and CO2 capture abilities. Fig. 3c exhibits that PI-1 can adsorb a moderate amount of H2 (74 cm3 g−1, 0.66 wt%) at 77 K and 1 atm, and it decreased to 57 cm3 g−1 (0.51 wt%) at 87 K. The enthalpy of H2 adsorption for PI-1 is calculated using the modified Clausius–Clapeyron equation (Fig. S8, ESI†) by analyzing the isotherms at 77 K and 87 K.16 The result reveals that PI-1 presents a high initial Qst value, that is 11.7 kJ mol−1 at initial loading, which is higher than or comparable to other porous organic materials, such as BILPs (7.8–8.3 kJ mol−1),17 COFs (6.0–7.0 kJ mol−1),18 –OH functionalized POFs (8.3–9.0 kJ mol−1),19 tetrazine-based TzFs (7.8–8.2 kJ mol−1),20 metalated POPs (8.1–9.6 kJ mol−1),21 APOPs (7.6–9.4 kJ mol−1).22 The high Qst value proves the strong interaction between pore wall and the absorbed H2 molecule. Furthermore, micropore size distribution for PI-1 (5.4 Å) is close to the size of about two H2 molecule (2.8 Å × 2), which may help enhancing the interaction between the H2 molecules and the framework. PI-1 may have application in H2 storage. For CO2 capture, 1 shows average CO2 adsorption capacity of about 41 cm3 g−1 at 273 K and 1 atm, 23 cm3 g−1 at 298 K and 1 atm (Fig. 3d), respectively. Additionally, PI-1 shows selective adsorption of CO2 over CH4 and N2. The CO2/CH4 selectivity at 1 atm increases from about 4.3 at 273 K to 8.1 at 298 K.
In conclusion, three polyimides polymers are successfully synthesized by the condensation reaction of A3-type cross-linker with a series of difunctional aromatic dianhydride. The reversible redox enolation of carbonyl groups of polyimides polymers provides a promising energy-storage system as cathode materials for Li-ion batteries. Though these materials only show moderate discharge capacities and cycle performance, such organic materials may exhibit broader application prospects than the conventional transition-metal-oxide-based materials. Meanwhile, the polymer PI-1 can selectively adsorbs CO2 over CH4 and N2 and exhibits the high Qst value of H2.
Footnote |
† Electronic supplementary information (ESI) available: Experimental details, characterization of the materials and gas sorption isotherms. See DOI: 10.1039/c3ra45563g |
This journal is © The Royal Society of Chemistry 2014 |