Communication Cite This: J. Am. Chem. Soc. XXXX, XXX, XXX−XXX
pubs.acs.org/JACS
Cationic Covalent Organic Framework Nanosheets for Fast Li-Ion Conduction Hongwei Chen,*,† Hangyu Tu,† Chenji Hu,‡ Yi Liu,§ Derui Dong,† Yufei Sun,† Yafei Dai,§ Senlin Wang,† Hao Qian,† Zhiyong Lin,† and Liwei Chen*,‡ †
College of Materials Science and Engineering, Huaqiao University, Xiamen 361021, People’s Republic of China i-Lab, CAS Center for Excellence in Nanoscience, Suzhou Institute of Nano-Tech and Nano-Bionics (SINANO), Chinese Academy of Sciences, Suzhou 215123, People’s Republic of China § Jiangsu Key Laboratory of Optoelectronic Technology, School of Physics Science & Technology, Nanjing Normal University, Nanjing 210023, People’s Republic of China ‡
S Supporting Information *
using Li salts that easily dissociate or adding metal oxide fillers that contribute to the dissociation of ion pairs.14−16 However, the ion aggregation problem still persists in these cases; and the addition of fillers may introduce new challenges such as dispersion of the fillers and the loss of energy density in the final assembled batteries. In the past few years, a new strategy based on ionic molecular skeletons has been pioneered to develop efficient Li+ conductors.12,17 Incorporation of ionic scaffolds in the structure gives rise to high dielectric constant materials, which can effectively screen Coulombic interactions and thus yield more free mobile ions and better Li + conductivity as have been shown in polymer solid-state Li+ conductors.17−19 Here we demonstrate for the first time that cationic moieties incorporated into the skeleton of a COF material can indeed split the ion pair of the Li salt, boost the concentration of free mobile Li+, and thus improve the Li+ conductivity in COFbased Li+ conductors (up to 2.09 × 10−4 S cm−1 at 70 °C). The cationic COF effectively screens the Coulombic interactions due to its greater polarizability compared to the COFs with neutral frameworks (Figure 1a). The concentration of free Li+ ions and consequently the Li+ conductivity are drastically improved without employing any solvent, quasi-solid-state, or solid plasticizers. The two-dimensional (2D) COF nanosheet (CONs) was selected as the COF matrix for this study due to its high surface area and unique 2D framework, both of which are beneficial to the exposure of the ionic moieties toward Li salts.20−22 The desired CON-based conductor was prepared in three steps and the synthetic scheme was shown in Figure 1b. The cationic CON with chloride counterions (CON-Cl) w first synthesized according to the literature (step I).23 The 13C cross-polarization-magic angle spinning (CP-MAS) solid-state nuclear magnetic resonance (NMR) spectra of the CON-Cl confirmed the chemical structure of the product.24 Fourier transform infrared (FTIR) spectra further confirmed the proposed chemical structure (see Figures S1−S3 in SI). The CON-Cl was then ion-exchanged with lithium bis (trifluoromethane) sulfonimide (LiTFSI) (step II), during which Cl− anions were
ABSTRACT: Covalent organic frameworks (COFs) with their porous structures that are accommodative of Li salts are considered to be potential candidates for solid-state fast Li+ conductors. However, Li salts simply infiltrated in the pores of solid-state COFs tend to be present in closely associate ion pairs, resulting in slow ionic diffusion dynamics. Here we incorporate cationic skeleton into the COF structure to split the Li salt ion pair through stronger dielectric screening. It is observed that the concentration of free Li+ ions in the resulting material is drastically increased, leading to a significantly improved Li + conductivity in the absence of any solvent (up to 2.09 × 10−4 S cm−1 at 70 °C).
S
olid Li+ conductors are key materials for the construction of solid-state lithium batteries with high energy, high power, and long cycle life.1−3 As a novel class of porous crystalline solids, covalent organic frameworks (COFs) are characterized by their stable organic frameworks and open channels, which can accommodate lithium salts and thus may exhibit ionic conduction behavior.4−6 However, little is known about the Li+ conduction properties of COF materials. Recently, fast Li+ conductivity has been demonstrated in two novel types of COFs.5,6 Considering the almost unlimited possibility in organic molecular design and the better electrochemical stability compared with other porous materials such as metal−organic frameworks (MOFs), the COFs-based Li+ conductors deserve further investigation. It is imperative to understand the Li+ conduction mechanism and the structural design guideline in this new type of solid ionic conductors. Generally, Li+ conductivity is the product of the Li+ concentration and the Li+ mobility.7 In solid conductors, the ionic components (including Li+ and corresponding anionic species) tend to be closely associated in the form of ion pairs or even ionic aggregates due to strong Coulombic interactions.8 These bulky ion pairs and aggregates thus result in slow ionic diffusion dynamics and low ionic conductivity.4,9,10 Promoting the dissociation of ion pairs in solid-state conductors has thus become an important approach for improving ionic conductivity in MOFs and polymers.11−13 It has been achieved by © XXXX American Chemical Society
Received: November 21, 2017 Published: January 5, 2018 A
DOI: 10.1021/jacs.7b12292 J. Am. Chem. Soc. XXXX, XXX, XXX−XXX
Communication
Journal of the American Chemical Society
Figure 1. (a) Schematic illustrations of ion association in COFs with neutral and cationic frameworks, respectively. (b) Synthetic scheme of the CONs with cationic framework for Li+ conduction. (c) 13C CP-MAS solid-state NMR of CON-Cl. (d) Comparison of XRD patterns of CON-Cl and Li-CON-TFSI. (e) SEM images of Li-CON-TFSI (inset shows a zoomed-in image; scale bars = 0.5 μm).
replaced and TFSI− anions were introduced into the cationic framework to form the intermediate CON-TFSI. The X-ray photoelectron spectroscopy (XPS) full-spectrum scans of the CON-TFSI show that the Cl signal almost disappeared after ion exchange (Figure S4). The final product (dubbed as Li-CON-TFSI) was obtained by mixing the CON-TFSI with LiTFSI salt in ethanol solution and then dried (step III). Elemental analysis results show that the molar ratio between N+ (corresponding to the cationic moiety in the framework) and Li+ in Li-CON-TFSI was approximately 1:1 (Table S1). The X-ray diffraction (XRD) pattern of Li-CON-TFSI reveals no crystalline peak of LiTFSI salt (Figure 1d), suggesting completely altered ion pairing situation of LiTFSI within Li-CON-TFSI. The XRD pattern of CON-Cl reveals a broad peak at 2θ = ∼11°, which corresponds to the (100) plane. The major broad peak at 2θ = ∼28° indicates poor π−π stacking between the vertically stacked 2D layers.23 After exchange of the Cl− with larger TFSI− ions, the intensity of the broad peak at 2θ = ∼28° decreases substantially in CON-TFSI, reflecting lowered crystallinity (Figure 1d), which is reasonable considering that the presence of the larger TFSI− anions and the loosely bound intrinsic positive framework could further disturb the weak π−π stacking interactions among the layers in the extended CON-TFSI structure (Figure S10). The peak intensity further diminishes in Li-CON-TFSI since more bulky TFSI− anions are introduced. The result is consistent with the simulated XRD pattern (Figure S11). The Li-CON-TFSI exhibited a loose and irregular 2D nanosheet morphology, as seen in the SEM images (Figure 1e, more in Figure S5). TEM images show that the Li-CON-TFSI are nearly transparent thin sheets (Figure S6). Atomic force microscopy (AFM) image and height profiles show that the thickness of the Li-CON-TFSI sheets are approximately 5 nm (Figure S7). The porosity of the samples was confirmed by N2 adsorption isotherm measurements (Figure S8). To study the ion conduction in Li-CON-TFSI, the browncolored powder sample was mechanically pressed into solid pellets for electrochemical testing (Figure S13). Electrochemical impedance spectroscopy (EIS) of Li-CON-TFSI pellets at different temperatures are shown in Figure 2a. All curves show a distorted semicircle with a pronounced tail at low frequencies, which is attributed to the blocking effect at the
Figure 2. (a) EIS of Li-CON-TFSI at different temperature (inset zooms in at the high frequency range). (b) Ionic conductivity of LiCON-TFSI as a function of different temperature.
electrode. The EIS spectra are consistent with ion migration in the solid pellet. The conductivity was calculated from the lowfrequency intercept on the real axis. The Li-CON-TFSI exhibited a conductivity of 5.74 × 10−5 S cm−1 at 30 °C; and the conductivity increased to 2.09 × 10−4 S cm−1 at 70 °C. The ionic conductivity is slightly higher than that of the MOF/ COF-based electrolytes prepared by adding various plasticizers, and is comparable with the best-reported COFs (see Table S2). The temperature dependence of the ionic conductivity is plotted in Figure 2b. A semilogarithm relation between conductivity and the inverse of temperature, which is typical of ceramic solid conductors, is observed for Li-CON-TFSI and the activation energy of Li+ diffusion is calculated to be 0.34 eV atom−1. Linear sweep voltammetry (LSV) scan of the Li-CONTFSI indicates that no obvious decomposition of any components occurs until 3.8 V versus Li+/Li (Figure S14a). An average Li+ transference number value of 0.61 ± 0.02 was determined for Li-CON-TFSI using the Bruce−Vincent−Evans (BVE) method (Figure S14b).25,26 The excellent Li+ conduction in Li-CON-TFSI arises from the unique loosely stacked ionic framework, which is beneficial to screening the Coulombic interactions due to its greater polarizability compared to the neutral frameworks. The crystalline peak of the Li-CON-TFSI almost disappeared compared to the original CONs (Figure 1d). Similar structural evolutions have also been reported in ionic polymers.18 The structure evolution suggests that there exist complicated Coulombic interactions between the cationic framework and the ion species, including free Li+ cations, free TFSI− anions, and the corresponding ion pairs (Li+TFSI−). The Li+ B
DOI: 10.1021/jacs.7b12292 J. Am. Chem. Soc. XXXX, XXX, XXX−XXX
Communication
Journal of the American Chemical Society TFSI− pairs may be restructured: the cationic framework is expected to interact with the TFSI− anions by electrostatic forces and thereby liberate the Li+ ions for faster ion conduction (Figure 3a).
Possible Li+ local environments within Li-CON-TFSI are marked in the inset of Figure 3c. Apart from the region close to the ionic center of cationic framework, the carbonyl oxygen atoms of Li-CON-TFSI are also expected to coordinate with Li+ ions (to form Li+OCO) and thus may contribute to the high Li+ conductivity, which has been reported previously for polycarbonate-based solid polymer electrolytes.30 This point is strongly supported by the FTIR spectra in Figure 3d. Carbonyl stretching vibrations are well-known to be highly sensitive to Li+ coordination.30 The original CON-Cl and CON-TFSI showed broad peaks at ∼1550−1700 cm−1, corresponding to CO and CC groups within the samples. When Li+ was introduced into the frameworks, the CC peak of Li-CONTFSI remained basically unchanged. However, the peak corresponding to the Li+OCO became relatively sharp and shifted to lower wavenumbers (at ∼1620 cm−1). DFT calculations further confirmed the possibility of Li+ binding to carbonyls (Figure S12). The carbonyl coordination provides abundant sites for Li+ transport and thus also contributed to the Li+ conductivity. In summary, we demonstrate a new class of solid Li+ conductors based on cationic CON. The special 2D planes as well as the cationic framework contribute to the improved Li+ conductivity up to 2.09 × 10−4 S cm−1 at 70 °C. Considering the great potential of COF materials with modular structural design, the ionic conductivity and electrochemical stability of the COF-based solid Li+ conductors can undoubtedly be further optimized. The structure may also be tailored for the conduction of other ions such as Na+, K+, Mg2+, or Al3+.
Figure 3. (a) Schematic of the ion pair around the neutral and cationic framework, respectively. (b) Raman spectra of CON-Cl, Li-CONTFSI, and LiTFSI in 650−800 cm−1. (c) 7Li MAS NMR spectra of LiCON-TFSI and LiTFSI. Inset: Schematic of the possible local Li+ environments in Li-CON-TFSI. (d) FTIR spectra of CONs and LiCON-TFSI, showing evolution of CO peak after introducing Li+. Left: the corresponding structure of the framework.
■
ASSOCIATED CONTENT
S Supporting Information *
Raman spectra of the Li-CON-TFSI and the control samples are thus compared because the Raman signals at the 650−800 cm−1 range relating to TFSI− anions are highly sensitive to the ion association status.27 As shown in Figure 3b, the Raman spectra of pure LiTFSI salt and CON-TFSI show characteristic sharp peaks at ∼750 and ∼730 cm−1, respectively, which are assigned to the aggregated ion pairs: TFSI− coordinating to Li+ cations or the cationic framework.27 In contrast, the spectrum of the Li-CON-TFSI shows a broad and slightly downshifted peak at ∼720 cm−1, which can be deconvoluted into two components of free and aggregated TFSI anions (Figure S9) with the prominent peak at ∼713 cm−1 is assigned to the “loose” or free TFSI− anions. On the basis of the integrated peak areas, an estimated ∼18% of the TFSI− anions are present in the form of aggregated ion pairs (Figure S9).28 The local chemical environments and dynamics of Li+ ions in Li-CON-TFSI and pure LiTFSI were further studied using 7Li solid-state NMR (Figure 3c). A single broad signal was observed in the spectrum of pure LiTFSI as a result of solidstate dipole−dipole and quadrupolar couplings between 7Li sites, hinting the limited mobility of Li+ in solid LiTFSI.9 In contrast, a narrow 7Li NMR signals at approximately −0.9 ppm was observed in Li-CON-TFSI, which is attributed to the presence of more mobile 7Li species resulting from the diminished solid-state pairing.5,29 In addition, multiple weak peaks are present in the 7Li spectrum of Li-CON-TFSI, indicating that Li+ may exist in different local chemical environments. The significant changes in the chemical shift of these 7Li peaks suggest different Li+ chemical environments in Li-CON-TFSI compared to that in the pure LiTFSI salt, consistent with the splitting of Li+TFSI− pairs in Li-CONTFSI.
The Supporting Information is available free of charge on the ACS Publications website at DOI: 10.1021/jacs.7b12292. Experimental methods, SEM, TEM, AFM, BET, DFT and detailed electrochemical studies (PDF)
■
AUTHOR INFORMATION
Corresponding Authors
*
[email protected] *
[email protected] Notes
The authors declare no competing financial interest.
■
ACKNOWLEDGMENTS This work was supported by the National Natural Science Foundation of China (Grant No. 21625304, 21733012, 21703072, and 21473242), the “Strategic Priority Research Program” of the CAS (Grant No. XDA09010600) and the Collaborative Innovation Center of Suzhou Nano Science and Technology. H.C. acknowledges the support from the Promotion Program for Young and Middle-aged Teacher in Science and Technology Research of Huaqiao University.
■
REFERENCES
(1) Bachman, J. C.; Muy, S.; Grimaud, A.; Chang, H.-H.; Pour, N.; Lux, S. F.; Paschos, O.; Maglia, F.; Lupart, S.; Lamp, P.; et al. Chem. Rev. 2016, 116, 140. (2) Wu, B.; Wang, S.; Evans, W. J., IV; Deng, D. Z.; Yang, J.; Xiao, J. J. Mater. Chem. A 2016, 4, 15266. (3) Tu, Z.; Nath, P.; Lu, Y.; Tikekar, M. D.; Archer, L. A. Acc. Chem. Res. 2015, 48, 2947.
C
DOI: 10.1021/jacs.7b12292 J. Am. Chem. Soc. XXXX, XXX, XXX−XXX
Communication
Journal of the American Chemical Society (4) Horike, S.; Umeyama, D.; Kitagawa, S. Acc. Chem. Res. 2013, 46, 2376. (5) Vazquez-Molina, D. A.; Mohammad-Pour, G. S.; Lee, C.; Logan, M. W.; Duan, X.; Harper, J. K.; Uribe-Romo, F. J. J. Am. Chem. Soc. 2016, 138, 9767. (6) Du, Y.; Yang, H.; Whiteley, J. M.; Wan, S.; Jin, Y.; Lee, S. H.; Zhang, W. Angew. Chem., Int. Ed. 2016, 55, 1737. (7) Park, M.; Zhang, X.; Chung, M.; Less, G. B.; Sastry, A. M. J. Power Sources 2010, 195, 7904. (8) Izutsu, K. Electrochemistry in nonaqueous solutions; John Wiley & Sons, 2009. (9) Yanai, N.; Uemura, T.; Horike, S.; Shimomura, S.; Kitagawa, S. Chem. Commun. 2011, 47, 1722. (10) Rhodes, C. P.; Frech, R. Solid State Ionics 1999, 121, 91. (11) Schmidtchen, F. P.; Berger, M. Chem. Rev. 1997, 97, 1609. (12) Wiers, B. M.; Foo, M.-L.; Balsara, N. P.; Long, J. R. J. Am. Chem. Soc. 2011, 133, 14522. (13) Fujie, K.; Otsubo, K.; Ikeda, R.; Yamada, T.; Kitagawa, H. Chem. Sci. 2015, 6, 4306. (14) Moganty, S. S.; Jayaprakash, N.; Nugent, J. L.; Shen, J.; Archer, L. A. Angew. Chem. 2010, 122, 9344. (15) Xu, K. Chem. Rev. 2004, 104, 4303. (16) Yuan, C.; Li, J.; Han, P.; Lai, Y.; Zhang, Z.; Liu, J. J. Power Sources 2013, 240, 653. (17) Strauss, E.; Menkin, S.; Golodnitsky, D. J. Solid State Electrochem. 2017, 21, 1879. (18) Zhang, P.; Li, M.; Yang, B.; Fang, Y.; Jiang, X.; Veith, G. M.; Sun, X. G.; Dai, S. Adv. Mater. 2015, 27, 8088. (19) Mecerreyes, D. Prog. Polym. Sci. 2011, 36, 1629. (20) Dogru, M.; Bein, T. Nat. Nanotechnol. 2011, 6, 333. (21) Dalapati, S.; Jin, E.; Addicoat, M.; Heine, T.; Jiang, D. J. Am. Chem. Soc. 2016, 138, 5797. (22) Khayum, M. A.; Kandambeth, S.; Mitra, S.; Nair, S. B.; Das, A.; Nagane, S. S.; Mukherjee, R.; Banerjee, R. Angew. Chem., Int. Ed. 2016, 55, 15604. (23) Mitra, S.; Kandambeth, S.; Biswal, B. P.; Khayum, A.; Choudhury, C. K.; Mehta, M.; Kaur, G.; Banerjee, S.; Prabhune, A.; Verma, S.; et al. J. Am. Chem. Soc. 2016, 138, 2823. (24) Kandambeth, S.; Mallick, A.; Lukose, B.; Mane, M. V.; Heine, T.; Banerjee, R. J. Am. Chem. Soc. 2012, 134, 19524. (25) Evans, J.; Vincent, C. A.; Bruce, P. G. Polymer 1987, 28, 2324. (26) Bruce, P. G.; Evans, J.; Vincent, C. A. Solid State Ionics 1988, 28, 918. (27) Yamada, Y.; Furukawa, K.; Sodeyama, K.; Kikuchi, K.; Yaegashi, M.; Tateyama, Y.; Yamada, A. J. Am. Chem. Soc. 2014, 136, 5039. (28) Seo, D. M.; Borodin, O.; Han, S.-D.; Boyle, P. D.; Henderson, W. A. J. Electrochem. Soc. 2012, 159, A1489. (29) Gobet, M.; Greenbaum, S.; Sahu, G.; Liang, C. Chem. Mater. 2014, 26, 3558. (30) Zhang, J.; Zhao, J.; Yue, L.; Wang, Q.; Chai, J.; Liu, Z.; Zhou, X.; Li, H.; Guo, Y.; Cui, G.; et al. Adv. Energy Mater. 2015, 5, 1501082.
D
DOI: 10.1021/jacs.7b12292 J. Am. Chem. Soc. XXXX, XXX, XXX−XXX
