Atomic probe of bond dynamics and reactivity of individual HF and H2O atoms

September 12, 2020 - Permeation and transport of atoms, molecules or ions across membranes is one of the most fundamental phenomena in natural and manmade materials. A team of scientists from the Universities of Ulm, Germany and Nottingham, UK has directly imaged the atomic permeation across a defect in a carbon lattice.

Graphene being an ultimate membrane with nano‐ and sub‐nanometer vacancy defects has been subject of several studies for transport of ions, atoms and molecules.[1, 2] Pristine graphene is impermeable for all ions, atoms and molecules except H+,[3] while porous single‐layer graphene is more permeable[2] and can be used for e.g. molecular sieving.[4] The driving forces for these applications include pressure,[4] electric fields and osmosis[5] across porous graphene.

Analogous to graphene, sub‐nanometer defects may pre‐exist in the lattice of carbon nanotubes[6] or can be created by chemical and physical techniques such as treatment with an oxidant or microwaves.[7] But a systematic experimental study of premeation thorugh the curved carbon lattice had not yet been performed. Theoretical calculations predicted that a transition metal atom bonding to a sub‐nano vacancy defect on a single‐walled carbon nanotube (SWNT) externally is more stable than when it bonds to the defect internally,[7] which in principle may provide a mechanism for metal atom transport across a defect in SWNT, without any external field gradient.

A team of researchers at the Sub-Angstrom Low-voltage electron (SALVE) microscopy center at Ulm University in Germany and the University of Nottingham in the UK has now come up with a novel study. They prepared Palladium nanoparticles inside single-walled carbon nanotubes (SWNT) by inserting palladium hexafluoroacetylacetonate complex into SWNT from gas phase followed by thermal decomposition and formation of metallic nanoparticles inside the nanotube.[8, 9] (Figure 1).

Under the e-beam, the nanoparticles could be divided into three parts including a crystalline Part I, a different crystalline Part II, and an amorphous Part III (Figure 1b). It was observed that Part III binds to the vacancy defects in SWNT carbon lattice as indicated by the red arrows in Figure 1 a. The nanoparticle shown in this Figure consisted of 140 Pd atoms confined within a (13,7) SWNT. The TEM image simulated for this model (Figure 1 c,d) illustrates how the vacancy defect may bond to the Pd atoms (Figure 1 d).

"By imaging the permeation process of Pd atoms atom‐by‐atom, in real time and with atomic resolution we could say important things about the phenomenon and underlying atomistic permeation mechanism that are determining the filtration processes by porous carbon‐based membranes," said SALVE director Ute Kaiser about the research.

The research team's findings showed that the rate of permeation correlates with the phase state of the metal nanoparticle: the crystalline part of the Pd nanoparticle gradually changing into an amorphous state, which then provides accessible metal atoms for transport across the lattice defect. The results clearly show that amorphization of Pd nanoparticle is a prerequisite for the permeation.

As the samples could be captured and observed for a longer time by TEM, the permeation of each Pd atom adsorbed on the defect outside of the SWNT, before desorbing irreversibly into the vacuum, could be obtained.[9] The external Pd atoms were sputtered by electron beam immediately with a residence time of less than a second (Figure 2, Video 1).

Video 1: Time‐series AC‐HRTEM images showing the atoms of a Pd nanoparticle permeating across a sub‐nanometer defect under 80 keV electron beam irradiation with an exposure time of 0.5 s.

After that the Pd atoms began permeating across the defect and escaping into vacuum successively until 61 s (Figure 3). From 62 s to 136 s, the permeation rate significantly reduced, during which time the crystalline Part I was still generally stable. The next fast permeation process started at 137 s. The scientists showed that amorphization of Pd nanoparticle is a prerequisite for the permeation. Typically, a gradient of pressure, electric field or concentration is required to drive the permeation process.[3, 4] The amorphous Pd cluster has higher free energy per atom than Pd crystal (Figure 4 ), the barrier for removal of Pd atom from a Pd cluster [10] can only be overcome after amorphization, they found out. After 150 s, the Pd cluster started bonding with the host SWNT instead of permeating, which may be caused by the increased surface energy of such a small Pd nanocluster.

In addition, the flexible nature of intermetallic bonding within the nanoparticle makes it metastable under the electron beam which could even form dynamical metal atoms with coordination number of one as demonstrated by the authors in previous works.[9, 11] They had also shown previously that direct sputtering of Pd atoms from nanoparticles can take place under 80 keV e‐beam.[9] However, inside the nanotube the Pd nanoparticle clearly loses Pd atoms and emits them through a sub‐nanometer defect (Figure 2), indicating that the contact between the metal and dangling bonds of the defect in the carbon lattice is an essential factor enabling atomic permeation and sputtering. Due to a significant steric strain when the metal is bonded to the concave surface of the nanotube, with a difference in the bonding energy,[8, 12] Pd atom moves through the defect and forms stronger bonds with the convex surface of the nanotube (State 4). The energy barrier between State 3 and State 4 in Figure 4 is overcome by the transferred kinetic energy from incident electrons to the Pd atom, considering the electron beam is the only stimulus for the whole process. Thus, the permeation properties are very different from graphene. and the curvature of the carbon lattice is shown to drive the atomic permeation in a direction from the concave to convex side of the membrane, due to a difference between metal‐carbon bonding with two opposite sides of the carbon lattice.

„The application of nano‐filtration has recently become an important area of research, as for instance in NaCl filtration or the separation of gases. This methodology provides the first direct observation of the atomic permeation through a subnano‐pore, highlighting the importance of chemical bonding between the mobile atom and dangling bonds around the subnano‐pore orifice.“ says Ute Kaiser.

Resource: Cao, K., Skowron, S. T., Stoppiello, C. T., Biskupek, J., Khlobystov, A. N., & Kaiser, U. (2020) Direct Imaging of Atomic Permeation Through a Vacancy Defect in the Carbon Lattice. Angewandte Chemie International Edition, 14, 11178, doi: 10.1002/ange.202010630.

References
  1. L. Wang, M. S. H. Boutilier, P. R. Kidambi, D. Jang, N. G. Hadjiconstantinou, R. Karnik, Nat. Nanotechnol. 2017, 12, 509– 522.

  2. Griffin, E., Mogg, L., Hao, G. P., Kalon, G., Bacaksiz, C., Lopez-Polin, G., et al. (2020), ACS Nano 2020, 14 (6), 7280-7286.

  3. S. Hu, M. Lozada-Hidalgo, F. C. Wang, A. Mishchenko, F. Schedin, R. R. Nair, E. W. Hill, D. W. Boukhvalov, M. I. Katsnelson, R. A. Dryfe, I. V. Grigorieva, H. A. Wu, A. K. Geim, Nature 2014, 516, 227– 230.

  4. S. P. Koenig, L. Wang, J. Pellegrino, J. S. Bunch, Nat. Nanotechnol. 2012, 7, 728– 732;

  5. D. Cohen-Tanugi, J. C. Grossman, Desalination 2015, 366, 59– 70.

  6. K. Suenaga, H. Wakabayashi, M. Koshino, Y. Sato, K. Urita, S. Iijima, Nat. Nanotechnol. 2007, 2, 358– 360.

  7. C. G. Salzmann, S. A. Llewellyn, G. Tobias, M. A. H. Ward, Y. Huh, M. L. H. Green, Adv. Mater. 2007, 19, 883– 887.

  8. A. N. Andriotis, M. Menon, G. Froudakis, Phys. Rev. Lett. 2000, 85, 3193– 3196.

  9. K. Cao, T. Zoberbier, J. Biskupek, A. Botos, R. L. McSweeney, A. Kurtoglu, C. T. Stoppiello, A. V. Markevich, E. Besley, T. W. Chamberlain, U. Kaiser, A. N. Khlobystov, Nat. Commun. 2018, 9, 3382.

  10. R. V. Stuart, G. K. Wehner, J. Appl. Phys. 1962, 33, 2345– 2352;

  11. K. Cao, T. W. Chamberlain, J. Biskupek, T. Zoberbier, U. Kaiser, A. N. Khlobystov, Nano Lett. 2018, 18, 6334– 6339.

  12. Y. K. Chen, L. V. Liu, W. Q. Tian, Y. A. Wang, J. Phys. Chem. C 2011, 115, 9306– 9311.