Proton and Lithium Permeation Through Defective Graphene
May 19, 2020 – Carbon films one atom thick that are full of eight-atom rings still block helium, but let protons through about 1000 times more easily than perfect graphene, and lithium ions pass as well. Researchers from Manchester, Antwerp, Jena, Ulm, Singapore, Tokyo and the Chinese Academy of Sciences imaged the rings with a Thermo Fisher double-hexapole Cs-corrected TEM in Ulm at 80 kV, validated the images with a JEOL Delta-probe Cs-corrected STEM at AIST in Tsukuba at 60 kV and traced the selectivity to their low energy barriers, a promising route to membranes for fuel cells and lithium batteries.
Graphene is the tightest wall a single layer of atoms can build. Not even helium, the most elusive of all gases, gets through a flawless sheet in any measurable amount. Protons, the bare nuclei of hydrogen atoms, are the one exception, and they pass only slowly. Researchers from the Universities of Manchester, Antwerp, Jena and Ulm, the National University of Singapore, AIST and the University of Tokyo and the Chinese Academy of Sciences have now shown that the right kind of flaw can open the wall for small ions while it stays shut for everything else.
The team compared carbon films one atom thick that are full of rings with five, seven and eight atoms instead of six. These films still blocked helium, but let protons through about 1000 times more easily than perfect graphene, and lithium ions passed as well. The energy barrier for both ions dropped as the number of atoms in the ring grew, consistent with earlier theoretical results[1],[2]. Transport through the disordered films therefore depends on how many eight-atom rings they contain[3],[4]. These rings also explain why protons crossed the films about ten times faster than lithium ions. The ring structures were validated between a Thermo Fisher double-hexapole Cs-corrected TEM at 80 kV and a JEOL Delta-probe Cs-corrected STEM at 60 kV.
Why atomic-scale defects matter
Despite being only one atom thick, a defect-free graphene membrane a few micrometers across lets no more than a few gas atoms per hour pass, as experiments have shown[5],[6]. The dense electron clouds of the crystal lattice impose energy barriers of several electron volts on incoming atoms and molecules, which forbids their permeation under ambient conditions[1],[7]. Theory predicts that rings with seven or eight atoms in the lattice should greatly reduce the barrier faced by protons[1] and might even let small ions such as Li+ through, while graphene stays impermeable to atoms and molecules[7].
Until now, such defects could hardly be tested, because large graphene samples with a high density of atomic-scale defects did not exist. Holes were usually made by ion irradiation or by chemical and plasma etching[8],[9]. This removes carbon atoms locally and typically leaves pores about a nanometer wide[8] rather than defects of single atoms. These pores let through gases, ions and even large molecules such as DNA[8]. The team took another route and used films grown with the extended rings from the start, made with methods reported earlier[3],[4].
Four carbon films, from crystal to amorphous
The first film was nanocrystalline graphene, grown by quenching a hot platinum foil in ethanol: a patchwork of graphene crystallites only 3–4 nm across (Fig. 1a, b)[4]. Their boundaries consist of rings with five, six, seven and eight atoms, and because the grains are so small, the film contains a high density of these non-hexagonal rings. The second was monolayer amorphous carbon (MAC), deposited with laser-assisted chemical vapor deposition[3]: a one-atom-thick, random network of rings with five to eight atoms, without any visible graphene crystallites (Fig. 1c, d). Together the two are called disordered graphene.
Two reference materials completed the series. Carbon nanomembranes (CNMs) are amorphous carbon films about a nanometer thick, made from aromatic molecules that assemble on a gold surface and are cross-linked by electron irradiation[10],[11]. They are pierced by a dense network of pores, about 1014 per cm2 and roughly 0.7 nm in diameter[10]. The team used CNMs about 0.9 and 1.2 nm thick, called BPT and TPT after their precursors. The second reference was defect-free graphene peeled from graphite. For the first time, the permeability of two-dimensional carbon could thus be compared over the whole range from crystalline to disordered to amorphous.
Counting the rings
To link permeability to the ring structures, the films were imaged with atomic resolution on two microscopes with different aberration correctors: in Ulm, a Thermo Fisher double-hexapole Cs-corrected TEM (Titan 80-300, formerly FEI, corrector by CEOS) with a monochromator, operated as HRTEM at 80 kV, and at AIST in Tsukuba, a JEOL Delta-probe Cs-corrected STEM (JEOL 2100F), operated as HAADF-STEM at 60 kV. Both voltages are low enough to leave the carbon lattice intact. The Thermo Fisher TEM imaged nanocrystalline graphene (Fig. 1a), the JEOL STEM monolayer amorphous carbon, which was heated to 700 °C in the microscope to keep its surface clean (Fig. 1c). In both films, every ring could be identified and colored by its size (Fig. 1b, d). How the two microscopes validate each other is explained in the box below (Fig. 2).
Tight for helium, open for protons
The films were suspended over sub-micrometer holes in silicon nitride membranes. A mass spectrometer detected no helium flow through either disordered graphene or the nanomembranes, which puts their helium permeance below about 10−14 mol s−1 cm−2 Pa−1. That is as tight as commercial ion-conducting polymer membranes more than 100 µm thick.
For the ion measurements, the membranes were coated on both sides with an ion-conducting polymer and contacted with electrodes (Fig. 3a, b). All disordered films were about 1000 times more permeable to protons than defect-free graphene. Nanocrystalline graphene conducted best, followed by MAC at about half the value (Fig. 3c). Lithium ions also passed through all films except perfect graphene, which showed no lithium transport within the detection limit of about 10−5 S cm−2. In disordered graphene, the lithium conductivity was about ten times lower than the proton conductivity. This selectivity was missing in the nanomembranes: for them, protons and lithium ions gave similar values, and both scaled with the membrane thickness. How the ion currents were measured is explained in the box.
Eight-atom rings as gates
The electron clouds give a first picture. Around an eight-atom ring, the charge density is clearly thinner than around a six-atom ring of graphene, which should make the larger ring easier to cross (Fig. 3c, insets). Density functional theory (DFT) calculations confirmed it (Fig. 4): for a proton, the barrier dropped from 1.39 eV in the six-atom ring to 0.88 eV in the seven-atom rings of a Stone–Wales defect and to 0.78 eV in the eight-atom ring of a 5–8–5 defect. Because the conductivity depends exponentially on the barrier, the rare eight-atom rings dominate the transport. For a lithium ion, the eight-atom ring poses a barrier of about 1.5 eV, about twice that for a proton, which explains why disordered graphene prefers protons.
This interpretation is also consistent with the missing selectivity of the carbon nanomembranes, which have no eight-atom rings but pores about 0.7 nm wide. Their conductance for both ions grew with thickness, so the transport through the bulk of the film matters, not a single barrier at the entrance as in graphene[10]. This fits their structure, a dense network of sub-nanometer pores in which ions diffuse along winding paths, as is typical for porous media[12].
What it opens up
According to the measured temperature dependence, the proton conductivity of disordered graphene should exceed the industry benchmark of the polymer membrane Nafion 117, about 5 S cm−2, already at about 60 °C[13]. Defect-free graphene reaches this level only at about 200 °C, the temperature most desirable for fuel cells[13],[14]. Extrapolated to that range, disordered graphene membranes could reach about 100 S cm−2, well above the industry targets[15]. Disordered graphene can also be mass produced[3],[4].
The lithium permeability deserves special attention. Graphene is being explored as a host for highly reactive lithium metal[16] and lithium–silicon particles[17] in battery anodes and for lithium–sulfur cathodes[18], to protect them from reactions with the electrolyte, to stop lithium dendrites from growing and to make them mechanically stable. Such coatings need high lithium-ion conductivity combined with impermeability to reactive species. Since perfect graphene blocks lithium ions, defects are essential, and those that let only small ions pass, like eight-atom rings, offer considerable advantages.
“I've been working with two-dimensional amorphous structures since 2009 and we see now that disordered 2D materials are a very promising candidate for filtration of protons or ions,” says Ute Kaiser, head of the SALVE Microscopy Center at Ulm University[19].
“All the above indicates that defect engineering in graphene and other 2D materials could be a productive venue for optimizing their use in energy conversion and storage technologies,” the authors conclude. The research was funded by the European Research Council, the Graphene Flagship and the German Research Foundation, among others.
Resource:
Griffin, E., Mogg, L., Hao, G.-P., Kalon, G., Bacaksiz, C., Lopez-Polin, G., Zhou, T. Y., Guarochico, V., Cai, J., Neumann, C., Winter, A., Mohn, M., Lee, J. H., Lin, J., Kaiser, U., Grigorieva, I. V., Suenaga, K., Özyilmaz, B., Cheng, H.-M., Ren, W., Turchanin, A., Peeters, F. M., Geim, A. K., Lozada-Hidalgo, M. (2020).
Proton and Li-Ion Permeation through Graphene with Eight-Atom-Ring Defects.
ACS Nano 14(6), 7280–7286.
https://doi.org/10.1021/acsnano.0c02496
Download this news article as PDF – for reading on paper, taking notes or filing.
-
Miao, M., Nardelli, M., Wang, Q., Liu, Y. (2013). First Principles Study of the Permeability of Graphene to Hydrogen Atoms. Phys. Chem. Chem. Phys., 15, 16132. https://doi.org/10.1039/c3cp52318g
-
Xin, Y., Huang, A., Hu, Q., Shi, H., Wang, M., Xiao, Z., Zheng, X., Di, Z., Chu, P. K. (2018). Barrier Reduction of Lithium Ion Tunneling Through Graphene With Hybrid Defects: First‐Principles Calculations. Adv. Theory Simul., 1, 1700009. https://doi.org/10.1002/adts.201700009
-
Toh, C., Zhang, H., Lin, J., Mayorov, A., Wang, Y., Orofeo, C., Ferry, D., Andersen, H., Kakenov, N., Guo, Z., Abidi, I., Sims, H., Suenaga, K., Pantelides, S., Özyilmaz, B. (2020). Synthesis and Properties of Free-Standing Monolayer Amorphous Carbon. Nature, 577, 199–203. https://doi.org/10.1038/s41586-019-1871-2
-
Zhao, T., Xu, C., Ma, W., Liu, Z., Zhou, T., Liu, Z., Feng, S., Zhu, M., Kang, N., Sun, D., Cheng, H., Ren, W. (2019). Ultrafast Growth of Nanocrystalline Graphene Films by Quenching and Grain-Size-Dependent Strength and Bandgap Opening. Nat. Commun., 10, 4854. https://doi.org/10.1038/s41467-019-12662-z
-
Sun, P., Yang, Q., Kuang, W., Stebunov, Y., Xiong, W., Yu, J., Nair, R., Katsnelson, M., Yuan, S., Grigorieva, I., Lozada-Hidalgo, M., Wang, F., Geim, A. (2020). Limits on Gas Impermeability of Graphene. Nature, 579, 229–232. https://doi.org/10.1038/s41586-020-2070-x
-
Mogg, L., Zhang, S., Hao, G., Gopinadhan, K., Barry, D., Liu, B., Cheng, H., Geim, A., Lozada-Hidalgo, M. (2019). Perfect Proton Selectivity in Ion Transport Through Two-Dimensional Crystals. Nat. Commun., 10, 4243. https://doi.org/10.1038/s41467-019-12314-2
-
Leenaerts, O., Partoens, B., Peeters, F. (2008). Graphene: A Perfect Nanoballoon. Appl. Phys. Lett., 93, 193107. https://doi.org/10.1063/1.3021413
-
Wang, L., Boutilier, M., Kidambi, P., Jang, D., Hadjiconstantinou, N., Karnik, R. (2017). Fundamental Transport Mechanisms, Fabrication and Potential Applications of Nanoporous Atomically Thin Membranes. Nat. Nanotechnol., 12, 509–522. https://doi.org/10.1038/nnano.2017.72
-
Kotakoski, J., Krasheninnikov, A., Kaiser, U., Meyer, J. (2011). From Point Defects in Graphene to Two-Dimensional Amorphous Carbon. Phys. Rev. Lett., 106, 105505. https://doi.org/10.1103/physrevlett.106.105505
-
Yang, Y., Dementyev, P., Biere, N., Emmrich, D., Stohmann, P., Korzetz, R., Zhang, X., Beyer, A., Koch, S., Anselmetti, D., Gölzhäuser, A. (2018). Rapid Water Permeation Through Carbon Nanomembranes With Sub-Nanometer Channels. ACS Nano, 12, 4695–4701. https://doi.org/10.1021/acsnano.8b01266
-
Turchanin, A., Gölzhäuser, A. (2016). Carbon Nanomembranes. Adv. Mater., 28, 6075–6103. https://doi.org/10.1002/adma.201506058
-
Wiedenmann, D., Keller, L., Holzer, L., Stojadinović, J., Münch, B., Suarez, L., Fumey, B., Hagendorfer, H., Brönnimann, R., Modregger, P., Gorbar, M., Vogt, U., Züttel, A., Mantia, F., Wepf, R., Grobéty, B. (2013). Three‐Dimensional Pore Structure and Ion Conductivity of Porous Ceramic Diaphragms. AIChE J., 59, 1446–1457. https://doi.org/10.1002/aic.14094
-
Casciola, M., Alberti, G., Sganappa, M., Narducci, R. (2006). On the Decay of Nafion Proton Conductivity at High Temperature and Relative Humidity. J. Power Sources, 162, 141–145. https://doi.org/10.1016/j.jpowsour.2006.06.023
-
Bose, S., Kuila, T., Nguyen, T., Kim, N., Lau, K., Lee, J. (2011). Polymer Membranes for High Temperature Proton Exchange Membrane Fuel Cell: Recent Advances and Challenges. Prog. Polym. Sci., 36, 813–843. https://doi.org/10.1016/j.progpolymsci.2011.01.003
-
U.S. Department of Energy (2012). Multi-Year Research, Development and Demonstration Plan. https://www.energy.gov/sites/prod/files/2014/12/f19/fcto_myrdd_full_document.pdf (accessed 2018-06-01). https://www.energy.gov/sites/prod/files/2014/12/f19/fcto_myrdd_full_document.pdf
-
Zhao, J., Zhou, G., Yan, K., Xie, J., Li, Y., Liao, L., Jin, Y., Liu, K., Hsu, P., Wang, J., Cheng, H., Cui, Y. (2017). Air-Stable and Freestanding Lithium Alloy/Graphene Foil as an Alternative to Lithium Metal Anodes. Nat. Nanotechnol., 12, 993–999. https://doi.org/10.1038/nnano.2017.129
-
Li, Y., Yan, K., Lee, H., Lu, Z., Liu, N., Cui, Y. (2016). Growth of Conformal Graphene Cages on Micrometre-Sized Silicon Particles as Stable Battery Anodes. Nat. Energy, 1, 15029. https://doi.org/10.1038/nenergy.2015.29
-
Tan, G., Xu, R., Xing, Z., Yuan, Y., Lu, J., Wen, J., Liu, C., Ma, L., Zhan, C., Liu, Q., Wu, T., Jian, Z., Shahbazian-Yassar, R., Ren, Y., Miller, D., Curtiss, L., Ji, X., Amine, K. (2017). Burning Lithium in CS2 for High-Performing Compact Li2S–Graphene Nanocapsules for Li–S Batteries. Nat. Energy, 2, 17090. https://doi.org/10.1038/nenergy.2017.90
-
Geim, A. (2009). Graphene: Status and Prospects. Science, 324, 1530–1534. The TEM image in its Fig. 4D was taken from: Meyer, J. C., Chuvilin, A., Kaiser, U. (2009). Electron microscopic studies with graphene. Microsc. Microanal. https://doi.org/10.1126/science.1158877




