A number of different procedures for preparing nanoparticles for surface analysis are presented (drop casting, spin coating, deposition from powders, and cryofixation). We discuss the challenges, opportunities, and possible applications of each method, particularly regarding the changes in the surface properties caused by the different preparation methods.
Nanoparticles have gained increasing attention in recent years due to their potential and application in different fields including medicine, cosmetics, chemistry, and their potential to enable advanced materials. To effectively understand and regulate the physico-chemical properties and potential adverse effects of nanoparticles, validated measurement procedures for the various properties of nanoparticles need to be developed. While procedures for measuring nanoparticle size and size distribution are already established, standardized methods for analysis of their surface chemistry are not yet in place, although the influence of the surface chemistry on nanoparticle properties is undisputed. In particular, storage and preparation of nanoparticles for surface analysis strongly influences the analytical results from various methods, and in order to obtain consistent results, sample preparation must be both optimized and standardized. In this contribution, we present, in detail, some standard procedures for preparing nanoparticles for surface analytics. In principle, nanoparticles can be deposited on a suitable substrate from suspension or as a powder. Silicon (Si) wafers are commonly used as substrate, however, their cleaning is critical to the process. For sample preparation from suspension, we will discuss drop-casting and spin-coating, where not only the cleanliness of the substrate and purity of the suspension but also its concentration play important roles for the success of the preparation methodology. For nanoparticles with sensitive ligand shells or coatings, deposition as powders is more suitable, although this method requires particular care in fixing the sample.
Nanomaterials are defined as materials having any external dimension between 1 nm and 100 nm or having an internal or surface structure on this scale1. Due to the unique properties arising from their small scale and correspondingly large surface area (among other factors), they find increasing use in a wide variety of fields including agriculture, chemistry, automotive construction, cosmetics, environment, medicine, printing, energy, and textiles. This increased use means that both humans and the environment will be exposed, on a hitherto unknown scale, to these materials whose toxicological properties are not yet fully known, and whose size enables their facile integration into biological or environmental systems2.
After the fundamental properties of surface area and particle size/size distribution, surface chemistry and coatings were identified as the most crucial property of nanomaterials3; smaller particles have a higher surface area per unit mass, and therefore a higher ratio of surface to bulk atoms. Indeed, for nanoparticles of 1 nm size, over 70% of atoms can be found at corners or edges; this strongly influences surface properties such as chemisorption which is highly dependent on the atomic-scale surface morphology4. Regulations dealing with nanomaterials require accurate data regarding physicochemical properties and reliable estimates of the toxicological properties of these materials. In order to efficiently estimate toxicological properties from physical and chemical properties of nanomaterials, the nanomaterials community requires reliable, standardized, and verified analytical procedures. Projects such as ACEnano5 aim to collect and correlate accurate and verifiable physical data from nanoparticles in a framework allowing better regulation and characterization of nanomaterials. This drive towards standardized analytical procedures has also been supported by the editors of ACS Nano, wishing “to consolidate and to agree on methods of characterization and minimum levels of analysis of materials6”. Furthermore, XPS and ToF-SIMS offers new possibilities for elucidating the particle architecture of core-shell nanoparticles7,8.
X-ray photoelectron spectroscopy (XPS) and Time-of-Flight Secondary Ion Mass Spectrometry (ToF-SIMS), compared in Table 1, are well-established methods for investigation of surface atoms. In XPS, the sample is irradiated with x-rays having an energy between 1 and 2 keV, causing emission of electrons due to the photoelectric effect. These emitted electrons, having a kinetic energy in the same range, correlate to the binding energy of the electrons in the solid; the appearance of photoelectrons at these defined binding energies and measurable intensities therefore allows quantitative analysis of the composition. Since the mean free pathway of these photoelectrons is below 10 nm, XPS is a highly surface sensitive technique for quantitative analysis. Furthermore, detailed analysis of the binding energies in highly resolved spectra enables the quantitative determination of the valence states of these electrons.
In ToF-SIMS the surface is sputtered with a focused ion beam (primary ions), with the ions ejected from the material (secondary ions) collected and analyzed in a time-of-flight mass spectrometer. The mass/charge pattern obtained allows the determination of the elemental, isotopic, or molecular composition. Due to the mean free pathway of the secondary ions, this technique is also highly surface sensitive and has an information depth of 1–2 nm but is at best semi-quantitative, due to the matrix effect whereby the ionization probability (and therefore yield) of secondary ions is strongly influenced by their surrounding matrix. ToF-SIMS can be operated in either static or dynamic mode; the difference between the two is the primary ion flux impacting the surface. Static SIMS keeps the primary ion flux to a level that impacts (i.e., fragments) a maximum of 1%-10% of the surface; the surface remains relatively undisturbed, which allows analysis of the top atomic layers of material. Since even static SIMS causes some destruction to the surface, it is considered to be less “non-destructive” of the two methods.
These surface-sensitive techniques allow analysis of the first few nanometers of the material, including intentional or unintentional coatings, which, for nanomaterials, can significantly influence material properties. Examples of intentional coatings are capping layers on quantum dots to improve photoluminescence quantum yields and reduce environmental reactivity9, alumina or silica coatings for the prevention of photocatalytic activity of titania nanoparticles in sun blockers10, surface functionalization to enable bioconjugation and subsequent biological activity11, coatings for diagnostic and drug delivery applications12, and fluorocarbon coatings on magnetic particles for ferrofluids and core-shell metallic systems to enhance catalyst properties13. Unintentional coatings, such as oxidation, surface contamination, or protein coronas in biological systems have a similarly strong influence on nanoparticle properties and it is crucial that experimental preparation procedures ensure that the coating and more generally the surface chemistry of the nanomaterial is not destroyed or transformed. It is also crucial to evaluate the properties of the nanoparticles as they are in-situ, as their properties can be drastically altered by the change2,14,15. In addition, the concentration of stabilizers in the nanoparticle suspension can dramatically influence the analysis and structural integrity of the nanoparticles; the presence of a stabilizer can result in large unwanted signals (for example, C, H, O, and Na) in the analysis, while its removal can result in damage or agglomeration of the nanoparticles.
Due to their size and surface area, the storage conditions of nanoparticles also affect their behavior, both as stored powders/suspensions and as prepared samples. The effect of sub-optimal storage conditions, particularly room-temperature storage and exposure to light, have been shown in various studies to cause degradation of the nanoparticles which has been shown to alter the particles’ physical, chemical, and/or toxicological properties14,15,16,17,18. Smaller nanoparticles have been shown to oxidize more rapidly than larger ones with oxidation/degradation rates dependent on storage conditions15 as well as surface chemistry14. The effects of the nanoparticle degradation during storage have been shown to significantly affect physicochemical properties including toxicity14, while the oxidative growth can proceed inwards at the expense of the core15.
The careful storage and preparation of nanomaterials is therefore essential for an accurate surface analysis, and any factors which could influence the sample surface and/or the quality of the measurements should be carefully considered. It should be noted that due to the relatively low spatial resolution of XPS (in the µm range) and ToF-SIMS (a few hundred nm), only a small subset of the nanoparticles can be investigated; these methods average over an area and do not have the ability to image single particles as is possible with techniques such as electron microscopy. For this reason, any analysis requires deposition of the nanoparticles in a continuous layer to ensure no interference from the substrate. Electron microscopy and XPS/ToF-SIMS are therefore often used together as complementary methods for nanomaterial analysis.
Aside from changes in surface chemistry, the main challenges for preparation of nanoparticle samples for XPS and ToF-SIMS analysis are to prepare a layer that is: homogeneous, to increase reproducibility; gapless, to minimize the contribution of the substrate to the spectra; thin enough to avoid charging effects (for non-conductive samples); and securely fixed to the substrate, to avoid free nanoparticles entering and damaging ultrahigh vacuum instruments
Nanoparticles can be deposited onto the substrate from suspension or as a powder. Firstly, we will discuss the different methods for depositing nanoparticles from suspension. Silicon wafers are a commonly used substrate for suspension deposition, because they are relatively cheap, readily available as a highly pure product consisting of pure or doped silicon (doping avoids charging effects), and for most nanoparticles the spectral peaks do not overlap with peaks typical for nanoparticles. This last point is important; before analysis it should be ensured that the substrate peaks are well separated from the peaks expected from the nanoparticles, otherwise interpretation of the spectra is complicated or impossible and the continuous coverage of the substrate by the nanoparticles cannot be verified. Before using silicon wafers, an extensive cleaning procedure (described in this publication) is necessary to remove (organic) contaminants and to increase the surface wettability. Other suitable substrates such as gold films, highly ordered pyrolytic graphite (HOPG), or indium foils have been successfully used, but a discussion about their preparation is beyond the scope of this work19,20,21,22.
Secondly, we present methods for depositing nanoparticle powders on a substrate for XPS and ToF-SIMS analysis and present the advantages and disadvantages of each method, allowing researchers new to the techniques to find the optimal preparation method for their purposes. Thirdly, we discuss cryofixation, which is a suitable preparation method to conserve features such as the agglomeration behavior, organic corona, solid/aqueous interface23,24 or distribution in biological media25 of NPs. Cryofixation, typically fast freezing of material in a liquid nitrogen–cooled cryogen and analysis in the frozen-hydrated state, allows the analysis and visualization of nanoparticles directly in complex matrices. This procedure does not cause ice crystal formation but forms amorphous ice that keeps membranes and cellular and tissue structures in their native biological state, avoiding damage caused by water crystallization processes and enabling the exact chemical distribution of all cell metabolites and cell membrane compounds to be maintained26,27,28. This preparation method may be of particular interest for presenting an exact chemical map of the actual NP agglomerate or heteroagglomerate, visualizing the exact chemical space in close proximity to the nanoparticle directly in suspension, or correlating either cell tissue-specific features or intra-cellular compartments within NP agglomerates or heteroagglomerates.
As shown through the results presented in this work, the most suitable procedure in a particular case is dependent on a variety of parameters such as the nanoparticles’ hydrophilicity, stability, conductivity, state (e.g., powder or suspension) and the analytical question at hand (e.g., size, bulk properties, or surface coatings). A variety of methods are presented here that can be used for preparation of NPs for surface analysis, as well as a comparison of their advantages and disadvantages.
A number of methods have been presented for the preparation of nanoparticles for surface analysis using XPS and ToF-SIMS. We have summarized the advantages and disadvantages of these methods, as well as possible sources of error and suitability for different materials, in Table 2. As shown in the representative results, the preparation of nanoparticles can strongly influence the success of the resulting surface analysis. In addition, not all methods are suitable for all particle types due to factors such as signal interference with the substrate or mounting materials, charging effects in non-conducting thick films, state of the nanoparticles as a powder or suspension, potential damage to sensitive outer layers, destruction of biological structures and information on aggregation and interfaces, or vulnerability of sensitive ultrahigh-vacuum instruments to free nanoparticles.
Because XPS and ToF-SIMS measurements average over an area rather than measuring single particles, it is only possible to obtain reproducible results from homogeneous layers; aggregation or agglomeration of the particles on the substrate should therefore be avoided. Additionally, too-thick layers of non-conductive materials cause charging effects during analysis, which can lead to undesired artefacts in the spectra, especially partial charging which cannot be compensated with a flood gun. On the other hand, incomplete films show strong signals from the substrate or mounting materials (e.g., adhesives), which can interfere with sensitive peaks from the particle surface. The ideal thickness of the film is material dependent and should be determined experimentally by analysis of films of different thicknesses. In particular, samples prepared using spin coating should be analyzed with SEM to ensure completeness of the coating.
Working with NP suspensions presents fewer exposure hazards and safety requirements compared to working with NP powders. Drop-casting is a relatively simple method with low equipment requirements and is particularly suitable for conductive nanoparticles in suspension where film thickness is not a concern. While the samples can easily be dried under atmospheric conditions, the vacuum desiccator serves to reduce the drying time for the droplets as well as protecting the wafers from contamination. The Viton ring is used to modify the evaporation patterns of the droplet and thereby minimize the formation of coffee-rings. Evaporation patterns can also be influenced by varying the substrate hydrophilicity using cleaning protocols or by application of alternative coatings51,52, by evaporating in solvent atmospheres53, or even by heating the substrate54. Spin-coating is recommended for suspensions of non-conductive nanoparticles in suspension because it is capable of generating a homogeneous particle layer that is thin enough to avoid charging effects but still thick enough to prevent the Si substrate from contributing to the XPS and ToF-SIMS spectra. For each individual NP system and concentration, both the centrifuge and spin-coating parameters must be optimized but can then very reliably be reproduced even on different instruments. Because the spin-coated drop is always in the middle of the wafer, the radius of rotation is irrelevant and the unit “revolutions per minute” (rpm) can be used. The suspension could alternatively be deposited on the wafer after starting the program; however, this would require different spin-coating parameters and a greater amount of suspension to obtain a thicker coating.
Because of their extremely small size, nanoparticles may detach from the substrate and move freely inside the ultrahigh-vacuum chamber when impacted with an ion or x-ray beam. This is a particular problem for samples prepared with powder. In some cases, the nanoparticles can penetrate into the sensitive components of the instrument requiring expensive and time-consuming maintenance. Due to the applied acceleration voltage, the danger of damaging sensitive parts is larger with ToF-SIMS than with XPS. Powdered samples, particularly those prepared using the “stick and go” method, should be carefully checked to ensure the powders are fixed securely enough, especially for ToF-SIMS analysis. This can be confirmed by, for example, holding the sample upside down and blowing a stream of gas (e.g., N2) across it. Prior to analysis, the samples can also be left overnight in the airlock or other initial sample entry chamber of the instrument, where a stable vacuum can indicate that there are no loose particles from the sample. Nanoparticles prepared as pellets, however, can even be sputtered (at low acceleration voltages) without damaging the instrument; this method can eliminate contaminants, particularly hydrocarbons, introduced from the press and can also enable bulk analysis of the particles.
Preparation of NP powders in the sample holder stub allows for the preparation of samples with defined geometry and a macroscopically flat surface. Critical points are the cleanliness of the tool for pressing the sample, and the use of a low pressure to avoid changes in the nanoparticle surface due to this procedure. It has the disadvantages of needing a relatively high amount of material, and potential problems with loss of material in high-vacuum instruments. We do not recommend this method for ToF-SIMS analysis, as the particles are not compressed or secured in any way.
Regarding the NP material, the first consideration for sample preparation is elimination or minimizing of interference between NPs and substrates of similar material; for example, Si wafers are an unsuitable substrate for analysis of SiO2 NPs using XPS and ToF-SIMS, even with sufficient sample coverage. Metallic or inorganic nanoparticles can be readily analyzed as powder on an adhesive (assuming they contain no organic layers or coatings) due to the lack of signal interference between the nanoparticles and the double-sided adhesive, a preparation method which would be unsuitable for polymeric NPs. Metallic nanoparticles have more flexibility in terms of possible film thickness used due to the absence of charging effects, and may be drop-cast with relatively little equipment; however, they are likely to contain large amounts of impurities and stabilizers from their synthesis, which must be carefully removed without damage to the particles. Polymeric nanoparticles may more easily be damaged by die pressing but may also more readily hold together in the pellet, depending on the pressures used. Pellets or soft organic coatings on the NP surface also can be damage sensitive. Direct deposition from the solution has the potential to damage sensitive coatings either through the suspension or the drying process but is advantageous for analyzing NPs already present in suspension. Cryofixation is a suitable method for the analysis of chemical structures, surfaces or interfaces in suspension which would be damaged or destroyed by various other sample preparation techniques, but requires a specialized cryoequipment for both XPS and ToF-SIMS46’47.
While this paper describes several exemplary methods that can be used for sample preparation, in every case the method should be optimized and validated using alternative analytical methods. A detailed overview of the influence of different factors was recently published22. Besides the development and validation of suitable preparation methods, the documentation of these steps is also of paramount importance40. This publication presents some easy-to-handle methods and is a guide to modify or develop new methods according to the requirements of the specific task.
The authors have nothing to disclose.
This project has received funding from the European Union Horizon 2020 Programme (H2020) under grant agreement no. 720952 (ACEnano). The authors would like to thank Sigrid Benemann for SEM measurements, Markus Schneider for ToF-SIMS measurements and PCA, and Philipp Reichardt for assistance with filming.
4-figure Laboratory balance | Kern & Sohn GmbH | ADB200-4A | |
5 mm Pellet die | Specac | GS03060 | |
Alkali glass cleaning solution | Sigma-Aldrich | Hellmanex™ III Z805939 | Special cleaning solution for cuvettes |
Carbon adhesive tabs | Plano | "Leit-Tabs" G3347 | |
Clean laboratory beakers | any | e.g. 300 mL | |
Cryo-freezer | Electron Microscopy Sciences | EMS-002 Cryo Workstation | |
Dialysis tube with fasteners | Medicell Membranees Ltd | DTV12000.06.30 | Molecular weight cut-off (MWCO) 12-14 kDa |
Die press | any | Capable of 2 kN force | |
Disposable syringe, 1 mL, Luer-slip | TH Geyer | Labsolute 7657545 | Any appropriate volume can be used |
Double-sided adhesive | 3M | Removable Repositionable Tape 665 | |
Dry ice | Linde AG | ICEBITZZZ® | For short term storage/cooling |
Eppendorf transfer pipette and tips | Eppendorf | various | Check correct size for planned pipetting volume |
Ethanol, ACS grade | Merck KGaA | 1009832500 | |
FFP2 or FFP3 mask | various | For working with nanoparticles from non-hazardous materials, when not in a fume hood or glove box | |
Isopropanol, ACS grade | Merck KGaA | 1096342500 | |
Lab coat, gloves and goggles | any | ||
Laboratory centrifuge | Eppendorf | Centrifuge 5430 | |
Laboratory fume hood | any | necessary for working with nanoparticles | |
Laboratory stirrer & stirrer bar | NeoLab | D-6010 | |
Lint-free wipes | Kimberley Clark Professional | Kimtech Science Precision wipes | Recommended for working with Si wafers |
Liquid Nitrogen | Linde AG | Stickstoff flüssig 5.0 | Only for cooling of the cryogen. |
Microtube/centrifuge tube 1,5 mL | T.H. Geyer GmbH & Co. KG | Labsolute 7696751 | |
Nitrogen 5.0 | any | 99.999% purity | |
Pasteur pipette, PE, plastic 3 mL | TH Geyer | Labsolute 7 691 203 | |
Pasteur pipette, PE, plastic 3 mL | TH Geyer | Labsolute 7 691 203 | |
Powder sample holder | BAM workshop | "Home-made" sample holder | |
Propane | Sigma-Aldrich | 769037 | The cryogen should be of highest possible purity. |
Sample vial or centrifuge tube 1 mL | Greiner Bio-One GmbH | Cellstar 188 261 | Should be capable of being fixed in the Vortexer |
Silicon wafers | any | ideally 1cm2 pre-cut | |
Spin-coater | SPS Europe | SPIN150i-NPP | |
Syringe filter 0,45 µm | Th Geyer | Labsolute 7699803 | For smaller samples; larger versions exist for larger sample volumes |
ToF-SIMS | IONTOF GmbH | ToF-SIMS IV or V, equipped with Bi LMIG and flood gun | |
Tweezers for handling Si wafers | any | ||
ultrapure water | TKA | MicroPure 08.1202 | |
Ultrasonicator | Bandelin | Sonorex Super | |
UV/Ozone cleaner | NanoBioAnalytics | UVC-1014 | |
Vacuum dessicator | any | ||
Vacuum pump (membrane/diaphragm) | Vacuubrand GmbH | Type MD-4T | |
Viton O-ring 6.07 x 1.78 mm | Betech GmbH | 2-010, FKM 80 | |
Vortexer | Heathrow Scientific | Vortexer HS120212 | |
Wafer Holder 25mm coin style | Semiconductor Production Systems Europe | eWB0091-ASSY-1 | |
XPS | Kratos | Kratos Axis Ultra DLD |