Elsevier

Ultramicroscopy

Volume 221, February 2021, 113195
Ultramicroscopy

A simple method to clean ligand contamination on TEM grids

https://doi.org/10.1016/j.ultramic.2020.113195Get rights and content

Highlights

  • A simple method reduces ligand contamination of NPs on TEM grids.

  • Improve the quality of high-resolution TEM imaging and spectroscopy.

  • Facile, requires no dedicated experimental set-up.

  • Avoid heating, plasma reaction, and extra electron-beam irradiation.

  • Can be applied for many types of materials that suffer from ligand contamination.

Abstract

Colloidal nanoparticles (NPs) including nanowires and nanosheets made by chemical methods involve many organic ligands. When the structure of NPs is investigated via transmission electron microscopy (TEM), the organic ligands act as a source for e-beam induced deposition and this causes substantial build-up of carbon layers in the investigated areas, which is typically referred to as “contamination” in the field of electron microscopy. This contamination is often more severe for scanning TEM, a technique that is based on a focused electron beam and hence higher electron dose rate. In this paper, we report a simple and effective method to clean drop-cast TEM grids that contain NPs with ligands. Using a combination of activated carbon and ethanol, this method effectively reduces the amount of ligands on TEM grids, and therefore greatly improves the quality of electron microscopy images and subsequent analytical measurements. This efficient and facile method can be helpful during electron microscopy investigation of different kinds of nanomaterials that suffer from ligand-induced contamination.

Introduction

Nanoparticles (NPs) including nanowires and nanosheets show interesting optical, chemical, electrical, and physical properties that are very different from their bulk counterparts, and therefore play an important role in a wide range of fields. Very often, (organic) ligands are used to control the morphology and stability of the NPs as well as to modify their properties [1,2]. However, when the structure of these NPs is investigated by transmission electron microscopy (TEM), the incident electrons attract the organic ligands and break the bonds of the hydrocarbons, forming thick carbon layers in the area of interest. This phenomenon is often referred to as contamination in the field of electron microscopy [3,4]. The formation of such e-beam induced contamination is similar to e-beam induced deposition used as a lithography (EBL) technique, although for the latter a precursor gas is intentionally induced for deposition [5,6]. When colloidal NPs are drop-cast onto TEM grids, not only the ligands bonded with the NPs but also the ligands in the solution will leave residues on TEM grids and contribute to contamination. This contamination problem is more severe for scanning TEM (STEM), in which the electron beam is focused into a small probe and tends to have a high dose rate, forming e-beam induced contamination relatively rapidly. The illustrations in Fig. 1 demonstrate this process. In general, the thickness of the contamination increases when the magnification and acquisition time increase [7]. The higher the magnification, the faster contamination builds up, which makes atomic-resolution imaging very challenging. Similar problems occur for analytical analyses such as electron energy loss spectroscopy (EELS) and energy-dispersive X-ray spectroscopy (EDX) that require long acquisition times.

A broad range of methods for decontamination treatments of TEM grids have been proposed, including plasma cleaning (on PtNi and PbS NPs [7]; carbon films [7,8]; Si [7,9], InAs [10], SrTiO3 [9] and stainless steel [9,11] foils), beam showering (on PtNi and PbS NPs [7]; carbon films [7]; Si [7] and SiN [12] foils), heating in vacuum or in different atmospheres (on PtNi and PbS NPs [7]; graphene [4] and carbon films [7,8]; NiO thin film [7]), baking with activated carbon (on graphene [13]), cooling (on PtNi NPs [7]; organic specimens [14]), exposure to ultraviolet light and ozone (on Au and Sn particles [15]; carbon nanotubes [15]; graphene and carbon film [16]; Li-polymer [16]) and etching by exposure to low-pressure gas atmospheres under e-beam irradiation (on graphene [17]). All methods have their own advantages and disadvantages, some of which are summarized in Table 1 and TEM grids drop-cast with CuAg colloidal NPs capped by tetradecylphosphonic acid (TDPA) are used as testing material. The main conclusions are:

  • (1)

    Plasma cleaning using O2 successfully removed hydrocarbon contamination. However, for materials that might react with O2, including CuAg NPs, this is not a good solution. This limitation similarly applies for ozone treatment. Plasma using Ar or H2 without O2 did not show a significant sign of contamination removal. Another problem is that when the time used for the plasma cleaning was too long, the carbon support on the TEM grids showed partial damage and bending.

  • (2)

    Beam showering (shining a large electron flux on the samples at low magnification in TEM mode) worked very effectively for removing the contamination, albeit only for the relatively small areas (μm scale) that the electron beam can cover. Moreover, the cleaning effect was restricted in time (~ 30 min), as ligands from the surrounding non-showered contaminated area diffused back to the clean area. Finally, beam-damage occurred if the beam shower was carried out too long.

  • (3)

    Heating the TEM grids at 160 °C for 8 h in vacuum (~ 10−2 Pa range) did not remove the contamination for CuAg colloidal NPs. However, this method was applied in previous research and has sufficiently cleaned airborne hydrocarbon contamination on TEM grids [18,19]. These different results are likely related to the stronger bonding between the organic ligands and the CuAg NPs as well as the TEM grids in comparison to the bonding between airborne hydrocarbon and TEM grids. Moreover, a major concern for the heat treatment is that it might cause structural and compositional changes (i.e. alloying) in the samples.

  • (4)

    Heating TEM grids inside activated carbon was originally developed to clean graphene [13]. However, heating the grids up to 200 °C for 3 h was not sufficient to remove the ligand contamination for CuAg colloidal NPs. Again, a major concern is that heating might cause structural and compositional changes in the samples, especially as oxidation might occur while the samples are heated in air.

In summary, it is important and urgent to find a new method that can effectively remove organic ligand-induced contamination on nanomaterials, and meanwhile does not involve heating or induce oxidization. In this paper, we develop a new and facile method to clean TEM grids containing nanomaterials surrounded by ligands. Our results show that this method drastically removes excess ligands on the TEM grid, thereby enabling atomic resolution imaging and spectrum imaging. By comparing with control experiments, the mechanism behind the novel method is explained. We furthermore discuss the parameters that can be tuned for different situations and different material systems.

Section snippets

Materials and microscopy information

The colloidal CuAg NPs used in this paper were synthesized using a two-step procedure [20]. First, polycrystalline Cu cores were synthesized by reacting 0.6 mmol of Cu(OAc)2·H2O in 10 ml of trioctylamine with 0.3 mmol of TDPA. The reaction mixture was heated to 100 °C and degassed by multiple vacuum-argon filling cycles. Then, the mixture was heated at ~4 °C/min to 180 °C, held at that temperature for 30 min and then heated to 250 °C, followed by an additional dwell time of 30 min. Afterwards,

Cleaning method development

In our search for a cleaning method that is effective for cleaning ligands but does not involve heating or oxidization, we were inspired by the behavior of a sample containing CuSn NPs on carbon black. When the solution-based CuSn NPs were drop-cast onto TEM grids, a high level of contamination was observed during microscopy investigation (Fig. S1(a1-a2)). Then, to obtain a good conducting electrode for electrochemical measurements, the CuSn NPs were dispersed on carbon black and ethanol was

Discussion: mechanism and applications

To investigate the influence of ethanol and activated carbon in this method, we performed control experiments using different time, different solvents, and without using activated carbon.

As already mentioned in Section 3, we also submerged TEM grids in activated carbon + ethanol for longer periods including 2 and 24 h (control experiment 01 in Table 2). Although both treatments cleaned the grids, they did not show obvious improvements in comparison to submerging the grid for 10 min, which is

Conclusions

To remove the unwanted effect of contamination during electron microscopy investigations of NPs surrounded by ligands, an effective and facile method has been developed to clean the drop-cast TEM grids. A combination of activated carbon and non-solvent such as ethanol is hereby used. By performing and comparing control experiments, a mechanism for the approach is proposed: pouring non-solvent liquid into activated carbon pushes the air to escape from the high density of pores in activated

Author contribution

CL developed the method and wrote the paper. CL and APT tested this method on different materials. CL and DW performed the control experiments. APT checked activated carbon with tomography. DC and KVD supplied the CuAg and CuSn NPs samples, respectively. TB and SB conducted the project.

Declaration of Competing Interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Acknowledgment

This research was funded by the University Antwerp GOA project (ID 33928). DW acknowledges an Individual Fellowship funded by the Marie Sklodowska-Curie Actions (MSCA) in Horizon 2020 program (grant 894254 SuprAtom).

Reference (30)

  • M. Huth et al.

    Focused electron beam induced deposition: a perspective

    Beilstein J. Nanotechnol.

    (2012)
  • W.F. van Dorp et al.

    A critical literature review of focused electron beam induced deposition

    J. Appl. Phys.

    (2008)
  • T.C. Isabell et al.

    Plasma cleaning and its applications for electron microscopy

    Microsc. Microanal.

    (2002)
  • A.J.V. Griffiths et al.

    Quantification of carbon contamination under electron beam irradiation in a scanning transmission electron microscope and its suppression by plasma cleaning

    J. Phys.

    (2010)
  • N.J. Zaluzec et al.

    Reactive gas plasma specimen processing for use in microanalysis and imaging in analytical electron microscopy

  • Cited by (0)

    View full text