A simple method to clean ligand contamination on TEM grids
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)
Contamination mitigation strategies for scanning transmission electron microscopy
Micron
(2015)- et al.
Contamination of holey/lacey carbon films in STEM
Micron
(2012) - et al.
Radiation damage in the TEM and SEM
Micron.
(2004) - et al.
Scanning transmission electron microscopy under controlled low-pressure atmospheres
Ultramicroscopy
(2019) - et al.
Understanding individual defects in CdTe thin-film solar cells via STEM: from atomic structure to electrical activity
Mater. Sci. Semicond. Process.
(2017) - et al.
Metal–ligand bond strength determines the fate of organic ligands on the catalyst surface during the electrochemical CO2 reduction reaction
Chem. Sci.
(2020) - et al.
The surface science of nanocrystals
Nat. Mater.
(2016) - et al.
The role of ligands in the chemical synthesis and applications of inorganic nanoparticles
Chem. Rev.
(2019) - et al.
Contamination-free transmission electron microscopy for high-resolution carbon elemental mapping of polymers
ACS Nano
(2009) - et al.
Mitigating e-beam-induced hydrocarbon deposition on graphene for atomic-scale scanning transmission electron microscopy studies
J. Vacuum Sci. Technol. B, Nanotechnol. Microelectron.
(2018)