There are a large number of surface analytical methods, each with their own strengths and limitations. This review provides tools and an amalgamated source of information to those new to surface characterization or to those who would like to select the most appropriate, commonly accessible, surface characterization technique for their sample. It includes a quick technique selection guide for narrowing down suitable methods for obtaining a range of compositional, structural, and surface properties. Considerations and recommendations for approaching instrument technicians and field experts are also discussed. To further aid in technique selection, comparative tables for complementary techniques are included in addition to case studies in technique selection and application, and finally, a brief overview of commonly employed analysis methods is provided, including related important considerations.

The sheer number of surface characterization techniques can be daunting to those new to the field, and the nuances of each technique can take a significant amount of time to discover and to master. In this article, we condense the required information for selection of an appropriate surface characterization technique with the limitations of techniques in mind. This article is laid out as follows: Sec. I provides an introduction to characterization method selection and important considerations including method sensitivity, types of spatial measurements, sample limitations, use of complementary techniques, approaching experts, and instrument availability. Section II provides comparative tables to aid in selecting between techniques, which provide similar information. Section III then provides practical examples of method selection for specific research questions. Section IV provides further technique specific information and resources intended for use, following short listing of techniques of interest using the tools provided in Secs. I and II.

To select an appropriate surface characterization method, the first question that should be considered is “What information is required to support the investigation and to draw sound conclusions” whether in research and development, manufacturing, or failure analysis. In this review, we group this information into the composition (elemental and chemical makeup), structure (physical arrangement and form), and surface properties (interactions with light, chemistry, and other materials). If it is already known what type of information is required absent specific resolution or sample requirements, a quick selection guide is provided in Fig. 1 to assist in narrowing down potential characterization methods. However, when choosing between techniques that provide similar information, the following considerations should be addressed before a final decision is made.

FIG. 1.

Quick technique selection guide based on the required information with some technique specific notes. The first two columns can be used to narrow down potentially useful techniques. Techniques that provide several types of information may appear more than once and are color-coded to the type(s) of information they provide: composition (blue), structure (magenta), surface properties (gold). Techniques are also listed in the order of frequency of utilization (within each “specific” information type based on publication numbers in the last 5 years, explored further in Sec. II, Fig. 2) with the most commonly used techniques appearing at the top. For each technique, the capability of mapping (M) and depth profiling (D) is indicated if available; additionally, whether measurements take place under vacuum (V) or are destructive (X) is also shown. Note that otherwise nondestructive techniques may be destructive when depth profiling.

FIG. 1.

Quick technique selection guide based on the required information with some technique specific notes. The first two columns can be used to narrow down potentially useful techniques. Techniques that provide several types of information may appear more than once and are color-coded to the type(s) of information they provide: composition (blue), structure (magenta), surface properties (gold). Techniques are also listed in the order of frequency of utilization (within each “specific” information type based on publication numbers in the last 5 years, explored further in Sec. II, Fig. 2) with the most commonly used techniques appearing at the top. For each technique, the capability of mapping (M) and depth profiling (D) is indicated if available; additionally, whether measurements take place under vacuum (V) or are destructive (X) is also shown. Note that otherwise nondestructive techniques may be destructive when depth profiling.

Close modal

Highly relevant to the characterization of surfaces is the sensitivity of the methods employed. Two types of sensitivity must be considered when determining whether an analyte of interest will be detectable: the detection limit and surface sensitivity. The detection limit is a measure of the lowest concentration a technique can detect, and surface sensitivity can be represented by the approximate analysis depth. Techniques that are highly surface sensitive (have a lower analysis depth) can detect lower concentrations of elements/molecules of interest localized to a sample surface than techniques with an equivalent (or even sometimes lower) overall detection limit. Consider two techniques with varying analysis depths 10 or 1000 nm. Interrogating a 1 nm thick layer deposited onto a flat substrate, upward of 10% of the signal in the first technique (10 nm analysis depth) will be from the deposited layer, whereas closer to 0.1% of the signal in the second technique would originate from the layer of interest, noting these numbers are provided for illustrative purposes and that signal tends not to attenuate linearly. In the above cases, the first technique would only require a detection limit of ∼10% ( 100 000 ppm) to be useful, whereas the second would need a detection limit around 0.1% (1000 ppm), assuming the surface layer is composed entirely of the substance of interest. This improvement in sensitivity, however, also comes with a proportional increase in sensitivity to surface contamination. As such, extra handling precautions (although often good practice regardless) are often required for the most surface sensitive techniques. Resources for comparing technique detection limit and analysis depth are provided in Secs. II and IV. It should also be noted that while detection limits (whether a substance can be measured) are critical, precision (measurement reproducibility) and accuracy (correct quantification) should also be considered. If comparing results from one instrument between similar samples with analysis performed in the same way, the precision of the measurements may be more important than accurate quantification. However, when comparing data from different instruments and/or different techniques, accurate quantification may be more important.

Many characterization techniques are capable of depth profiling, enabling systematic interrogation of sample regions below the outermost layer. As the name suggests, these measurements provide spatial information along the z-axis. The method by which this is performed is technique dependent. Many techniques perform profiling by ablating samples with, e.g., an ion gun and performing serial measurements, and alternatively, some techniques can selectively focus on different sample depths (e.g., confocal microscopy). Many characterization methods inherently provide some depth related information, which can be resolved using multiple measurement angles, teased apart using modeling, or a combination thereof [e.g., angle resolved x-ray photoelectron spectroscopy (XPS), Auger electron spectroscopy background fitting, and x-ray reflectometry (XRR)]. Mapping is a related method for obtaining spatial data across an instrument’s xy-plane, either by scanning or imaging. Sectioned samples can also be loaded into instruments such that a cross section is visible and can be mapped to provide sample depth information. It should be noted that both depth profiling and mapping, or a combination thereof, often come with significant time costs that may become prohibitive. The depth profiling and mapping capability of selected techniques is indicated in Fig. 1. The resolution of both depth profiling and mapping is also highly technique dependent; resources for comparison are provided in Sec. II.

Obtainable characterization data are also highly sample dependent. If the sample is a large irregular shape, cannot be placed in a vacuum, and is so precious that the technique cannot be destructive, then choice is very limited in the techniques available. When approaching instrument technicians/scientists, sample limitations for the purposes of protecting instruments may also be highlighted. For example, restrictions on loading compounds that become volatile under ultrahigh vacuum into instruments require this level of vacuum for operation (such as XPS). This can result in contamination that is difficult if not impossible to remove. Another example, magnetic samples (especially, loose powders) pose a significant risk of contamination and instrument damage for electron microscopy-based techniques because magnetic fields are necessary for focusing electrons. Similar magnetic lenses are also used as a component of charge neutralization systems in Kratos XPS instruments, again proving problematic for magnetic powder samples.1 It may still be possible to analyze a magnetic sample; however, additional checks may be required and not all samples will be suitable. If necessary, these experts will also be able to recommend more niche techniques or advise when to seek access to techniques with much greater infrastructure requirements (e.g., neutron reflectometry and other neutron-based techniques or synchrotron sources) not covered in this review. The authors do note that the intense and highly controlled radiation of a synchrotron source can greatly improve signal, data collection rates, and spot size of many of the techniques discussed. However, many technique-based limitations are conserved between lab-based and synchrotron sources while access to these sources is more restricted, and additional care must be taken to avoid damaging samples (due to the source intensity). The other aspect of sample nature that must be considered is susceptibility to environmental factors. This is important not just from a degradation/contamination perspective but also because preparing/modifying a sample under a controlled environment (e.g., a glovebox or within a vacuum chamber) may provide useful experimental results for a particular application. For this reason, it is increasingly common to find instruments equipped with this capability in laboratories.

As shown in Fig. 1, when obtaining specific types of information, generally, there are still several relevant techniques. Choosing between these can be difficult, and it is useful to know the difference between them and their limitations. As an example, x-ray photoelectron spectroscopy (XPS) and energy dispersive x-ray spectroscopy (EDS), two methods that provide quantitative elemental composition, almost always provide different numerical answers, both of which can be correct if assuming the analysis was performed correctly. This is because of the different sampling depths analyzed by the techniques with approx. 10 nm for the XPS method and approximately 1000 nm for the EDS method. For this reason, these methods are generally complementary rather than redundant with XPS providing a greater surface sensitivity (relevant for thin surface coatings and oxide layers) and EDS providing information about the surface and much deeper layers. Section II provides resources for comparing and contrasting similar methods, and Sec. IV provides more instrument specific details. When using multiple techniques to provide complementary information on a single sample, it is also worth considering the order in which analyses are run. It may be obvious that destructive techniques should not be carried out first, but other techniques generally considered to be nondestructive such as scanning electron microscopy (SEM), can still degrade surface information both through e-beam damage, and deposition of carbon limiting the usefulness of subsequent surface sensitive measurements.

Improper sample handling is also an activity that can result in significant surface information degradation. Stevie et al.2 discussed some of the “tricks of the trade” in sample handling and preparation for surface analysis (that typically go unpublished) and highlighted the utility of enlisting the expertise of analysts. They also note that the American Society for Testing and Materials (ASTM) and International Standards Organization (ISO), publish guidelines on the preparation and handling of samples for surface analysis ASTM (E1829-14, E1028-02) and ISO (18116, 18117, 20579-4) relevant to both experts and those preparing samples for analysis. Relating to discussions with instrument technicians and surface analysis experts, or sometimes lack thereof, there is a growing concern in the literature of poorly analyzed data appearing in publications even in journals of high repute that could be avoided through increased consultation.3,4 This is particularly challenging for both multidisciplinary researchers and journals where a wide breadth of techniques applied may result in the use of a particular characterization method with limited depth of knowledge and potentially shallower or erroneous conclusions. Similarly, reviewers may also have limited knowledge of specific surface analysis techniques being utilized. Major et al. present an in-depth perspective and expand on this issue with trends, implications, and advice.4 The authors of this publication would suggest researchers should seek out expertise prior to experimentation to ensure that conclusions drawn will stand up to the scrutiny of those versed in surface characterization. This approach will not only increase trust in experimental conclusions but will also ensure that experiments are performed in both a time and sample efficient manner. When approaching surface analysis experts for an initial meeting, the following points are provided as a checklist of information to have on hand:

  • A prioritized list related to what you would like to get out of the experiment—for discussion.

  • A list of samples, including appropriate control samples.

  • A clear description of your samples, including their expected chemical structure, processing steps, regions to be analyzed, as well as any relevant safety and handling information.

  • A list of questions—related to, e.g., what complementary analytical methods can be used?

Unfortunately, cost is often another consideration when it comes to selecting characterization methods. This is not covered in great detail here as cost changes significantly based on whether instruments are already in the lab or available at a platform with subsidized, at cost, or business rates. This is further adjusted by whether experiments are user or facility run and who is responsible for analyzing the results. It may also be cost/time efficient to carry out one-off high-cost analysis to calibrate/validate lower cost methods. Related to availability and cost, it is generally worth considering the different ways a question may be answered, when selecting a surface characterization technique. For example, a number of techniques are listed for investigating questions specifically related to biological interactions; however, many biological interactions also result in measurable changes in surface composition. In this case, techniques that provide chemical or even elemental composition may be useful. A common example of this is using XPS to measure protein adsorption onto substrates exposed to biological solutions ranging from single protein buffer solutions to serum. Depending on the composition of the substrate, different chemical signatures such as the nitrogen elemental composition can be used to measure the amount of protein present. However, if the substrate is similar in composition to the protein binding from solution, this can quickly become difficult and more specialized techniques, such as those listed for measuring biological interactions, may be more suitable.

When choosing between techniques that provide similar information, it is useful to look at what others have used. Even without knowing all the reasons a particular method is employed, it is likely the most frequently used techniques are going to have instrumentation available nearby. Delving a little deeper, the frequency with which a technique is employed is likely to be a product of the advantages of the method, popularity of the fields in which it is useful, and time/resource cost. The utilization frequency of the techniques displayed in Fig. 1 over the last 5 years by publication number is shown in Fig. 2. If in doubt, starting with the most frequently employed technique that provides the information you are looking for and determining if it is suitable for your sample is a reasonable way of quickly arriving at a promising candidate. The most common use case for each technique is indicated by color, but it should be noted that the overall use frequency is shown in Fig. 2 and techniques may have multiple use cases. There are also significant differences between surface related studies [broader, Fig. 2(a)] and those from surface focused journals where surface analysis is likely the primary focus [Fig. 2(b)]. Both of these graphs are presented, as (a) presents a better indication of what equipment is likely to be available coupled with ease of use, but (b) provides a better indication of what is most commonly used specifically for surface analysis, which may also require additional expertise.

FIG. 2.

Recent characterization methods use frequency relating to “surface” studies or surface specialized journals. Frequency is given by publication number (2019–2024) obtained from article searches on Web of Science (Ref. 5) (WoS, Clarivate, accessed 11 April 2024), these were performed using exact (“”) nonabbreviated technique names AND “Surface” for all journals (a) or without the surface search term within selected WoS-indexed, surface focused journals (Applied Surface Science, ACS Applied Materials Interfaces, and Langmuir) (b).

FIG. 2.

Recent characterization methods use frequency relating to “surface” studies or surface specialized journals. Frequency is given by publication number (2019–2024) obtained from article searches on Web of Science (Ref. 5) (WoS, Clarivate, accessed 11 April 2024), these were performed using exact (“”) nonabbreviated technique names AND “Surface” for all journals (a) or without the surface search term within selected WoS-indexed, surface focused journals (Applied Surface Science, ACS Applied Materials Interfaces, and Langmuir) (b).

Close modal

The specifications provided in the following tables are to educate and provide an introductory overview, rather than absolute limits of a technique. For example, although depth profiling with Fourier transform infrared (FTIR) spectroscopy is possible to some extent, this is not commonplace for surface characterization; likewise, some of the numbers presented are general approximations provided with references to more detailed studies on the subject. A comparison of parameters for techniques that provide compositional information is provided in Table I. Provided in Table II are useful parameters to consider when choosing a technique for measuring layer thicknesses.

TABLE I.

Comparison of typical resolution, analysis depth, and sensitivity for common compositional analysis techniques.

Composition techniqueApproximate lateral resolutionApproximate analysis depthDepth profilingMappingDetection limit (ppm)Notes
Atomic force microscopy-infrared analysis (AFM-IR) 10 nm (Ref. 610 μNo Yes Monolayer/single molecule Not quantitative 
Auger electron spectroscopy (AES) 10 nm (Ref. 71.5–15 nm Yes Yes 1 × 103–1 × 104 Quantification generally relies on standards due to strong matrix effects 
Energy dispersive x-ray spectroscopy (EDX) 30–1000 nm (Ref. 81 μYes Yes 1 × 103 (Ref. 9Resolution dependent on e energy nominally closer to 1 μ
Fourier transform infrared spectroscopy (FTIR) 3–75 μm (Ref. 100.5–5 μm (Ref. 11No Yes 1 × 103–3 × 104 (Ref. 12ATR depth analysis dependent on prism material
Not quantitative 
Low energy ion scattering (LEIS) 10 μm (Ref. 1310 nm (depth resolvable data) (Ref. 14Inherent/Yes Yes 1 × 103–3 × 104 (Ref. 13Able to analyze outermost atomic layer 
Matrix-assisted laser desorption/ionization (MALDI) 10–100 μm (Ref. 152–100 nm (Refs. 16 and 17Limited Yes ∼10−3 (Ref. 18Requires addition of a matrix material 
Raman spectroscopy 250 nm (Ref. 191 μLimited Yes 0.001–5 × 103 (Refs. 20 and 21Lower detection limit indicative of surface enhanced Raman
Limited quantitation 
Secondary ion mass spectrometry (SIMS) 20 nm–10 μm (Refs. 22 and 231 nm (Ref. 24Yes Yes 0.2 × 10−4–1 (Ref. 25Resolution and detection limit highly instrument, method, and sample dependent 
Wavelength dispersive x-ray spectroscopy (WDS) 30–1000 nm (Ref. 81 μYes Yes 1 × 102 (Ref. 9Resolution dependent on e energy 
X-ray fluorescence (angle resolved) (XRF) 100 μm (Ref. 2610 nm–10 μInherent/Yes Limited 1–4 × 102 (Ref. 27Angle dependent analysis depth
Quantitation requires calibration 
X-ray photoelectron spectroscopy (XPS) 10 μm (Ref. 2810 nm Yes Yes 1 × 103–1 × 104 (Ref. 29Sensitivity high element/matrix dependent 
Composition techniqueApproximate lateral resolutionApproximate analysis depthDepth profilingMappingDetection limit (ppm)Notes
Atomic force microscopy-infrared analysis (AFM-IR) 10 nm (Ref. 610 μNo Yes Monolayer/single molecule Not quantitative 
Auger electron spectroscopy (AES) 10 nm (Ref. 71.5–15 nm Yes Yes 1 × 103–1 × 104 Quantification generally relies on standards due to strong matrix effects 
Energy dispersive x-ray spectroscopy (EDX) 30–1000 nm (Ref. 81 μYes Yes 1 × 103 (Ref. 9Resolution dependent on e energy nominally closer to 1 μ
Fourier transform infrared spectroscopy (FTIR) 3–75 μm (Ref. 100.5–5 μm (Ref. 11No Yes 1 × 103–3 × 104 (Ref. 12ATR depth analysis dependent on prism material
Not quantitative 
Low energy ion scattering (LEIS) 10 μm (Ref. 1310 nm (depth resolvable data) (Ref. 14Inherent/Yes Yes 1 × 103–3 × 104 (Ref. 13Able to analyze outermost atomic layer 
Matrix-assisted laser desorption/ionization (MALDI) 10–100 μm (Ref. 152–100 nm (Refs. 16 and 17Limited Yes ∼10−3 (Ref. 18Requires addition of a matrix material 
Raman spectroscopy 250 nm (Ref. 191 μLimited Yes 0.001–5 × 103 (Refs. 20 and 21Lower detection limit indicative of surface enhanced Raman
Limited quantitation 
Secondary ion mass spectrometry (SIMS) 20 nm–10 μm (Refs. 22 and 231 nm (Ref. 24Yes Yes 0.2 × 10−4–1 (Ref. 25Resolution and detection limit highly instrument, method, and sample dependent 
Wavelength dispersive x-ray spectroscopy (WDS) 30–1000 nm (Ref. 81 μYes Yes 1 × 102 (Ref. 9Resolution dependent on e energy 
X-ray fluorescence (angle resolved) (XRF) 100 μm (Ref. 2610 nm–10 μInherent/Yes Limited 1–4 × 102 (Ref. 27Angle dependent analysis depth
Quantitation requires calibration 
X-ray photoelectron spectroscopy (XPS) 10 μm (Ref. 2810 nm Yes Yes 1 × 103–1 × 104 (Ref. 29Sensitivity high element/matrix dependent 
TABLE II.

Comparison of resolution and thickness limits/sensitivity for common thickness measurement techniques.

Layer thickness techniqueApproximate lateral resolutionMeasuring thickness viaLayer thickness rangeThickness mappingRequirements and notes
Ellipsometry 1 μm–25 mm (Refs. 30 and 31Absorbance, reflectance, and polarization 0.1 nm–μm (Refs. 32–34Yes Thickness range conditional on sample optical properties
Resolution highly instrument dependent
Requires knowledge of materials for modelling 
Low energy ion scattering (LEIS) 10 μm (Ref. 13Scattering and reionization 1–10 nm Limited UH vacuum 
Profilometry (Mechanical) a0.05–50 μm (Ref. 35Edge step/scratch test 10 nm–μLimited Requires soft layer on hard underlying substrate
Resolution stylus dependent 
Profilometry (Optical) 0.5–5 μm (Ref. 35Reflectance 10 nm–μYes Strongly absorbing samples limit maximum layer thickness measurements 
Scanning electron microscopy (SEM) 1–50 nm (Ref. 36FIB/sample cross section 1 nm–μLimited Vacuum
Limitation to cross-sectional measurements can limit samples 
Secondary ion mass spectrometry (SIMS) 20 nm–10 μm (Refs. 22 and 23Ablative profiling 0.5 nm–μm (Ref. 37Limited UH vacuum 
Quartz crystal microbalance (with dissipation) [QCM(-D)] N/A Weight (and dissipation) 2 nm–μm (Ref. 38No Thickness range is highly dependent on layer stiffness sensitive to mass changes >2 nm but thickness quantification presents challenges 
X-ray photoelectron spectroscopy (XPS) ∼10 μm (Ref. 28Ablative profiling/attenuation or angle resolved 10–1000 nm/1–10 nm Limited UH vacuum 
X-ray reflectometry (XRR) 5 mm–1 cm (Refs. 26 and 39Interference induced oscillations in x-ray reflectivity 2–300 nm (Ref. 39Limited Can model layer thickness and roughness
Higher resolution is achievable by synchrotron sources 
Layer thickness techniqueApproximate lateral resolutionMeasuring thickness viaLayer thickness rangeThickness mappingRequirements and notes
Ellipsometry 1 μm–25 mm (Refs. 30 and 31Absorbance, reflectance, and polarization 0.1 nm–μm (Refs. 32–34Yes Thickness range conditional on sample optical properties
Resolution highly instrument dependent
Requires knowledge of materials for modelling 
Low energy ion scattering (LEIS) 10 μm (Ref. 13Scattering and reionization 1–10 nm Limited UH vacuum 
Profilometry (Mechanical) a0.05–50 μm (Ref. 35Edge step/scratch test 10 nm–μLimited Requires soft layer on hard underlying substrate
Resolution stylus dependent 
Profilometry (Optical) 0.5–5 μm (Ref. 35Reflectance 10 nm–μYes Strongly absorbing samples limit maximum layer thickness measurements 
Scanning electron microscopy (SEM) 1–50 nm (Ref. 36FIB/sample cross section 1 nm–μLimited Vacuum
Limitation to cross-sectional measurements can limit samples 
Secondary ion mass spectrometry (SIMS) 20 nm–10 μm (Refs. 22 and 23Ablative profiling 0.5 nm–μm (Ref. 37Limited UH vacuum 
Quartz crystal microbalance (with dissipation) [QCM(-D)] N/A Weight (and dissipation) 2 nm–μm (Ref. 38No Thickness range is highly dependent on layer stiffness sensitive to mass changes >2 nm but thickness quantification presents challenges 
X-ray photoelectron spectroscopy (XPS) ∼10 μm (Ref. 28Ablative profiling/attenuation or angle resolved 10–1000 nm/1–10 nm Limited UH vacuum 
X-ray reflectometry (XRR) 5 mm–1 cm (Refs. 26 and 39Interference induced oscillations in x-ray reflectivity 2–300 nm (Ref. 39Limited Can model layer thickness and roughness
Higher resolution is achievable by synchrotron sources 

Nanoscale structures have extremely broad applications in a wide range of fields including sensors, biomaterials, energy storage, and many more. When these structures have a homogeneous composition, many of the compositional techniques listed above (Fig. 1) may be suitable; however, nanostructures with heterogeneous compositions have much stricter characterization requirements. These requirements come in the form of resolution, both lateral and depth, in combination with mapping capabilities.

In this case study, Ogaki et al. fabricate and characterize a micrometer-scale cup array (Fig. 3) using multiple deposition methods including particle assembly, plasma etching, plasma polymerization, sputter deposition, evaporative deposition, and, finally, particle removal. Demonstration of site selective binding of ferritin directed by repellant/adherent self-assembled monolayer regions is also achieved.40 In this case, not only is the characterization of the array a challenge but demonstrating site selective binding of a nonlabeled compound on a patterned array presents an additional challenge.

FIG. 3.

Microcup arrays featuring multiple chemical regions. (a) SIMS chemical maps from microcup arrays with heterogeneous composition and various region sizes. (b) AFM images of patterned arrays showing ferritin binding only to hydrophobic regions coated with 1,7-octadiene plasma polymer (ppOct) or octadecyl trichlorosilane (OTS). Reproduced with permission from Ogaki et al., Adv. Mater. 23, 1876–1881 (2011). Copyright 2011 Wiley.

FIG. 3.

Microcup arrays featuring multiple chemical regions. (a) SIMS chemical maps from microcup arrays with heterogeneous composition and various region sizes. (b) AFM images of patterned arrays showing ferritin binding only to hydrophobic regions coated with 1,7-octadiene plasma polymer (ppOct) or octadecyl trichlorosilane (OTS). Reproduced with permission from Ogaki et al., Adv. Mater. 23, 1876–1881 (2011). Copyright 2011 Wiley.

Close modal

Although scanning electron microscopy SEM is more than capable of imaging the features of interest in this work, there are few techniques with the lateral resolution capable of measuring the compositions of the chemical regions expected in Fig. 3 (resolution greater than 70 nm). From Table I, the options for composition analysis in this range are EDS/WDS, AFM-IR, and SIMS. The nominal resolution of EDS/WDS is around 1 μm—much greater than the feature sizes of interest. In addition, the analysis depth of EDS/WDS is also around a micrometer, whereas in this study, the layers of interest vary in thickness from monolayers to several nanometers, leaving only AFM-IR and SIMS as viable options. Either one of these techniques would be reasonable to apply in this study, and the authors employed SIMS. Notably, AFM-IR is a relatively new technology that has only seen significant adoption recently. The results of SIMS analysis are shown in Fig. 3. The results show that the resolution limit of this particular instrumentation is close to the feature size of interest in the array, and some of the expected ring structure is visible; however, some of the smaller features apparent through SEM imaging blend with their surroundings.

To demonstrate the site-specific binding of ferritin, the authors employ a combination of atomic force microscopy (AFM) and flat analog samples of homogeneous composition. Using the flat analog samples, they were able to show that ferritin binds the substrates SiO2, bare Au, and hydrophobic plasma polymer, whereas it is repelled by a self-assembled monolayer with a triethylene glycol headgroup. The use of flat homogeneous sample analogs enabled the application of x-ray photoelectron spectroscopy (XPS) (which has relatively low lateral resolution), for the interrogation of substrate-specific ferritin binding. AFM of these same flat samples showed a significant increase in roughness originating from ∼40 nm features only on the same samples that XPS indicated ferritin bound to. As such, these features were attributed to ferritin aggregates. When performing AFM on the patterned micro-array analogs, 40 nm features were found only in regions where binding was expected, Fig. 3.

The development of new and improved coatings is of outstanding interest in a broad range of biomedical device applications, ranging from biosensors to implantable medical devices. In this context, a common target is the reduction of biofouling. As the adhesion of cells and bacteria is dependent on the adsorption of biomolecules, the modulation of protein adsorption at interfaces is a critical aspect of the surface/coating design. While a range of graft polymer-based technologies have been developed over the last few decades aimed at achieving this goal, more recently, cross-linked polymer coatings have emerged as highly effective alternatives. These have the advantage of being applied in a single step, making them attractive from an industrial perspective.

In a case study published by Yoshikawa et al.,41 photoreactive copolymers of 2-hydroxypropyl acrylamide (HPA) and N-benzophenone acrylamide (BPA) were deposited from solution onto polymer substrates using a spin coating process and cross-linked by UV irradiation. This leads to covalent anchoring of the coating on the polymer substrate materials, without the need for specific functional groups.

Cross-linked coatings representing different molar ratios of the copolymer components were characterized by XPS to demonstrate the successful deposition of the coating based on the changes in elemental composition compared to the polystyrene (PS) substrate material, and to confirm the composition of the copolymer when compared to the theoretical composition, which can be easily deduced from its structure, Fig. 4. FTIR spectroscopy was also used to qualitatively confirm this modification, as penetration of ATR is in the order of a few micrometers, the underlying substrate was still clearly visible. Furthermore, static water contact angle measurements were carried out to evaluate the hydrophilicity of the coating, to compare the values to previously reported homopolymer coatings of HPA, and to investigate the effect of different molar ratios of the BPA component on the wettability of the coatings.

FIG. 4.

(a) Structure of the photoreactive copolymers of 2-hydroxypropyl acrylamide (HPA) and N-benzophenone acrylamide (BPA). (b) (i) Primary human foreskin fibroblast cell attachment observed after 24h, following gentle washing on TCPS, (ii) a coating representing 10 mol. % of the BPA component, (iii) 5 mol. % of the BPA component, and (iv) 1 mol. % of the BPA component. Scale bars represent 50 μm. Reproduced with permission from Yoshikawa et al., J. Mater. Chem. B 7, 3520 (2019). Copyright 2019 The Royal Society of Chemistry.

FIG. 4.

(a) Structure of the photoreactive copolymers of 2-hydroxypropyl acrylamide (HPA) and N-benzophenone acrylamide (BPA). (b) (i) Primary human foreskin fibroblast cell attachment observed after 24h, following gentle washing on TCPS, (ii) a coating representing 10 mol. % of the BPA component, (iii) 5 mol. % of the BPA component, and (iv) 1 mol. % of the BPA component. Scale bars represent 50 μm. Reproduced with permission from Yoshikawa et al., J. Mater. Chem. B 7, 3520 (2019). Copyright 2019 The Royal Society of Chemistry.

Close modal

Moreover, the thickness of the coatings was evaluated by ellipsometry. Here, coatings were spin coated on silicon wafer substrates rather than PS substrates to provide a more suitable substrate for this characterization method. The use of alternative substrate materials is frequently needed in the characterization of coatings in order to obtain high quality data. In the context of coating thickness measurements, profilometry measurements would be an alternative, but these also require a hard substrate material such as silicon wafers or glass, to be able to create a scratch for subsequent step height measurements. However, the use of these alternative substrate materials also often creates further issues that need to be resolved. Here, the inorganic substrate material does not provide C–H bonds that are required to achieve covalent anchoring of the copolymer coating. In this study, this was addressed by first modifying the silicon wafer substrates with 3-aminopropyltriethoxysilane (AMP) prior to coating.

The ellipsometry data not only allowed the measurement of the coating thickness but also allowed the evaluation of the coating thickness after swelling in PBS buffer. Here, an essential piece of information was to demonstrate that the swelling ratio changed dramatically with the cross-linking density, reflected by the molar ratio of the BPA component, ranging from 1.7 for a copolymer containing 10 mol. % of this component to 6.7 for a copolymer containing 1 mol. % of this component.

Atomic force microscopy (AFM) measurements carried out on the same samples also allowed the measurement of Young’s modulus. Here, again significant differences were observed for the different cross-linking densities, with values ranging from 1.3 GPa for a copolymer containing 10 mol. % of the BPA component to 15.6 Pa for a copolymer containing 1 mol. % of this component.

Importantly, the comprehensive surface characterization of the copolymer coatings provides potential explanations for the different biological responses observed with confocal microscopy. Notably, although all coatings were found to be of similar hydrophilicity (measured via contact angle), highly cross-linked coatings containing 10 mol. % of the BPA component, which were shown to be much stiffer (AFM) and swell less (ellipsometry), were not able to reduce the adhesion of primary human foreskin fibroblast cells in comparison to a tissue culture polystyrene (TCPS) control surface. In contrast, a 5 mol. % coating resulted in rounded, less adherent cells, and a 1 mol. % coating swelled significantly more, producing a soft and diffuse solid-liquid interface resulting in the complete prevention of cell attachment, Fig. 4.

This section contains a brief description and uses of the characterization techniques listed in Fig. 1. Accompanying these are additional considerations and limitations for deciding whether a sample is suitable for each technique. For more in-depth information, readers are directed to selected review articles on each technique.

AFM measures the deflection of a sensitive cantilever tip when in contact or close to a sample surface using a reflected laser. This technique can be used to obtain an extremely broad range of information, though most commonly used to measure surface topography, it is also possible to measure mechanical properties, and using a chemically modified tip, chemical interactions, at very high resolution. Conductive tips are also available for measuring electrical properties with high spatial resolution and heated tips are available to measure thermal transitions. Bowen and Hilal provide an excellent introduction to the applications of AFM.42 Nguyen-Tri et al. show some of the recent advances and applications of AFM including infrared coupled instruments in the context of polymer science and Liang shows advances in the context of characterizing biological systems.43,44

Additional considerations/limitations:

  • Can be performed in vacuum, ambient, or solution conditions.

  • Various cantilever tips (both in terms of composition and geometry) are available for specific applications and chemical modification is an active area of research, especially for biological studies.

  • When performed in ambient conditions, humidity can present challenges.

  • Is not well suited for surfaces which have complex 3D morphology.

  • Negative deflection is also observable and can be used to indicate attractive forces.

Auger electron spectroscopy (AES) uses electrons to excite a atom, which in the process of undergoing relaxation, and ejects another electron. This technique is useful for surface sensitive elemental analysis of conductive samples and can be paired with an SEM to provide this information at an extremely high resolution (scanning auger nanoprobe SAN). Unger et al. provide a good introduction to AES.7 

Additional considerations/limitations:

  • Lateral resolution: ∼10 nm.

  • Analysis depth: 1.5–15 nm.

  • Depth profiling/mapping: Yes/Yes.

  • Detection limit: 1 × 103–1 × 104 ppm.

  • Vacuum required.

  • Charging is a significant challenge for signal quality and many of the neutralization solutions employed in other instruments are not available to AES; therefore, analysis is generally performed on conductive samples.

  • Quantification requires standards to correct for matrix effects.

Contact angle provides measurements of wetting between a solid surface and a liquid droplet. This wettability is dependent on the surface-liquid interaction strength and surface roughness. This technique is particularly useful for quickly showing changes in surface properties following modification steps and during the development of either liquid repellant or adherent surfaces. Contact angle instruments come in a range of forms, the most common of which is the optical tensiometer, sometimes referred to as an optical goniometer. Hebbar et al. provide a textbook introduction to contact angle measurements,45 and Eral et al. provide an in-depth review of contact angle hysteresis.46 Behnecke et al. provide a useful introduction to the Wilhelmy balance and its utility for interrogating nonplanar samples.47 

Additional considerations/limitations:

  • Although dedicated setups offer advantages, it is possible to obtain some measurements manually using a camera and syringe.

  • Significant hysteresis is present between advancing and receding contact angles.

  • Contact angle measurements can be carried out with a range of liquids besides water.

  • Contact angle measurements are heavily influenced by surface contamination and clean solvents are crucial.

  • Optical tensiometers are generally limited to planar samples, Wilhelmy balances provide a nonoptical method of measuring apparent contact angles with more complex geometries.

EDS (also referred to as EDX spectroscopy) and WDS both use high energy electrons to produce vacancies in core orbitals, as electrons from outer orbitals relax characteristic x rays are generated, which can be used to quantify elemental composition (Fig. 5). These techniques differ in their detectors, EDS uses a detector that measures x-ray energy, whereas WDS instrument uses a component such as a diffraction grating to separate x rays based on wavelength. Both of these techniques are useful for measuring elemental composition and distribution maps, especially in combination with electron microscopy. Goldstein et al. provide a comprehensive and practical textbook introduction to these techniques in combination with scanning electron microscopy (SEM).48 

FIG. 5.

Schematic of energy dispersive x-ray and wavelength dispersive x-ray (EDX/WDX) measurements. (a) Electrons interact with a sample material producing characteristic x rays. (b) Example of an EDS spectrum of silica/gold composite nanoparticles on a silicon substrate. Reproduced with permission from Gu et al., Optik 221, 165274 (2019). Copyright 2020 Elsevier.

FIG. 5.

Schematic of energy dispersive x-ray and wavelength dispersive x-ray (EDX/WDX) measurements. (a) Electrons interact with a sample material producing characteristic x rays. (b) Example of an EDS spectrum of silica/gold composite nanoparticles on a silicon substrate. Reproduced with permission from Gu et al., Optik 221, 165274 (2019). Copyright 2020 Elsevier.

Close modal

Additional considerations/limitations:

  • Lateral resolution: 30–1000 nm (nominally closer to 1000 nm).

  • Analysis depth: 1 μm.

  • Depth profiling/mapping: Yes/Yes.

  • Detection limit: 1 × 103 ppm.

  • Vacuum required.

  • Although WDS gives higher energy resolution, detectors are less common, and higher operating current and voltage can present a challenge to some samples.

  • EDS is often also coupled with transmission electron microscopes (TEMs); however, this places additional constraints on sample preparation but can achieve higher lateral resolution.

  • Depth profiling is achievable with focused ion beam (FIB) etching.

  • Changing the energy of the electrons can yield very different resolutions and sensitivities.

  • Almost always coupled with electron microscopy, generally electron microscopy images can be collected at a higher resolution than possible for EDS maps.

  • TEM/STEM coupled electron spectroscopy can obtain very high-resolution maps but require an electron transparent sample, e.g., a very thin foil sample.

Electrochemistry is a field concerned with the flow of electrons at an electrode surface and is a much broader topic than a collection of techniques useful for surface characterization. That said, within this field, there are many techniques that can be used to infer particular surface properties, and in some cases, composition. Impedance spectroscopy is particularly notable for its application to a wide range of surface science applications. Electrochemistry is commonly used in surface analysis for thin film corrosion studies, characterizing sensor surfaces and battery applications. Cesiulis et al. provide a good overview of the application of electrochemical impedance spectroscopy (EIS) for various types of thin films,50 although a degree of background knowledge is assumed and a more general introduction to EIS such as that provided by Wang et al. is probably a better starting point to those new to the subject.51 

Additional considerations/limitations:

  • Requires conductive substrate, although passivation following deposition of nonconductive materials can also be interrogated.

  • As quartz crystal microbalances (QCM) also require electroactive surfaces, they are often paired with electrochemical measurement capability.

  • Obtainable information is highly dependent on the sample system.

  • Useful for characterizing properties in aqueous solutions.

  • Techniques can be either destructive or nondestructive.

Ellipsometry uses a combination of absorbance and changes in polarization of light reflected off a sample to build a model of surface layers which can provide an indication of layer thickness or optical properties (such as refractive index) of thin films. This technique is useful for interrogating a wide range of layer thicknesses deposited onto reflective surfaces nondestructively; it also has the potential to model multilayers of optically distinct materials and measure real and imaginary components of the refractive index. Aspenes provides an introduction to the theory and progress of ellipsometry52 and Tompkins and Hilfiker provide a practical textbook introduction to the application for thin film characterization.53 

Additional considerations/limitations:

  • Lateral resolution: 1 μm–25 mm.

  • Layer thickness range: 0.1 nm–μm.

  • Thickness mapping: Yes.

  • Measured layers must vary optically.

  • Strongly absorbing materials may not be suitable.

  • Multispectral and angle resolved measurements can improve model quality and may be necessary for more complex samples.

  • Sample uniformity is generally assumed; however, a degree of roughness can also be modeled with some systems.

  • Measurements are generally performed under ambient conditions.

Fluorescence imaging uses light to excite electrons in atoms and molecules. These electrons then relax losing some energy to nonradiative processes while also emitting light at a longer wavelength to that absorbed. This technique is useful for observing compounds and structures of interest, following a fluorescence labeling process or compounds that have native fluorescence. Fluorescence imaging is also commonly used for optical sensors. For the characterization of surfaces, reflective modalities, such as confocal reflectance microscopy, are most commonly employed. Lavrentovich provides a good introduction to confocal fluorescence microscopy including theory,54 and Ribbe et al. provide an early example of characterization of a thin film using this method.55 

Additional considerations/limitations:

  • Many atoms and molecules are not fluorescent, specifically designed dyes are often used to label chemical/biological moieties of interest.

  • Confocal microscopes enable 3D imaging.

  • Autofluorescence—this can be a challenge where the sample matrix produces unwanted background fluorescence.

  • Photobleaching—a process where under illumination molecules lose their fluorescent properties and signal decreases over time, this can be either reversible or irreversible.

  • Various super resolution methodologies are available for carrying out this type of microscopy at resolutions higher than the diffraction limit of light would normally allow.

FTIR uses light in the infrared range to carry out spectroscopy and investigates the vibrational modes of molecules. FTIR selects for polar bond vibrations whereas Raman spectroscopy is sensitive to nonpolar bonds. FTIR is useful for quickly confirming the presence of characteristic polar bonds following surface modification. Mohamed et al. provide a textbook introduction to FTIR,56 while Yan and Wakamatsu provide an early example of the utility of grazing angle FTIR for the characterization of surfaces.57 

Additional considerations/limitations:

  • Lateral resolution: 3–75 μm.

  • Analysis depth: 0.5–5 μm.

  • Depth profiling/mapping: No/Yes.

  • Detection limit: 1 × 103–3 × 104 ppm.

  • Grazing angle can provide more surface sensitive measurements.

  • Generally limited to qualitative measurements.

  • Has been coupled with both AFM and microscopy to provide spectroscopic imaging capabilities at high resolution.

In IPES, samples are irradiated with low energy electrons, and radiation is detected when electrons at a vacuum level couple with high energy unoccupied electronic states; decay between unoccupied electronic states can also be observed. This technique is useful for probing the energy of unoccupied electronic states and, as such, is commonly used with photoemission spectroscopy such as UPS, which provides information on occupied electronic states. More recently, low energy IPES (LIPES) has been developed to carry out higher resolution measurements without sacrificing sensitivity, while also reducing sample damage for organic materials.58 

Additional considerations/limitations:

  • Vacuum required.

  • Is able to measure dielectric properties of nanofilms, and possibly, nanostructures.59 

Ion scattering is a broad topic that covers a range of ion projectile energies (∼10 eV–MeV) and their interaction with samples. As different energy ranges require different instrumentation and cover different types of material interactions, these ranges are generally categorized into low, medium, or high energy ion scattering (LEIS, MEIS, and HEIS). LEIS, sometimes referred to as ion scattering spectroscopy (ISS), is useful for obtaining compositional information of the outermost atomic layer and inherently produces data that can be modeled to provide depth information of the top ∼10 nm. LEIS also tends to be the most commonly available technique due to the increasing size and cost of instrumentation with ion energy. HEIS, also referred to as Rutherford backscattering analysis (RBS), interrogates a greater sample depth due to the increased energy of ions, and thus, penetration. Bird and Williams have a comprehensive textbook on ion beams for material analysis,60 and Cushman et al. provide an excellent practical introduction to LEIS.13 Průša et al. also recently published a guide to interpreting LEIS spectra aimed at nonexpert users.61 

Additional considerations/limitations:

  • LEIS lateral resolution: 10 μm.

  • LEIS analysis depth: 10 nm.

  • LEIS depth profiling/mapping: Inherent/Yes.

  • LEIS layer thickness range: 1–10 nm.

  • LEIS detection limit: 1 × 103–3 × 104 ppm.

  • Vacuum required.

  • LEIS provides inherent depth profile information of approximately the top 10 nm, beyond this depth, profiles can be obtained using sputtering.

  • LEIS surface sensitivity means sample handling and history are critical for obtaining useful measurements.

  • ISS is a relatively inexpensive add-on to XPS instruments with the addition of a He source.

MALDI is a mass spectrometry technique that relies on a matrix material to aid in the desorption and ionization of organic material and is able to generate significantly larger molecular weight fragments than SIMS. Although traditionally applied to solution based analytes, this technique has been adapted for surface analysis and is useful for the interrogation of proteins and polymers on a surface that can be dissolved into a matrix.62 Griesser et al. provided a useful introduction to the application of MALDI for the surface analysis of biomaterials.63 

Additional considerations/limitations:

  • Lateral resolution: 10–100 μm.

  • Analysis depth: 2–100 nm.

  • Depth profiling/mapping: Limited/Yes.

  • Detection limit: 1 × 10−3 ppm.

  • Vacuum required.

  • Destructive measurement.

  • A range of matrices are available for analysis and are generally tailored to the compounds of interest.

  • For surface analysis, a solvent and matrix material capable of removing/dissolving the compounds of interest from the surface is employed as dispersion throughout the matrix is required.

Optical microscopy is a widely understood method by which lenses are used to magnify small objects observed with visible light. Common variations include bright-field, dark-field, reflected light, polarized, and phase contrast microscopy. Although of limited utility in nanoscale surface characterization, this technique is useful for obtaining morphological information of thin samples with feature sizes larger than the wavelength of visible light. Davidson and Abramowitz provide an excellent overview of the types of optical microscopy available and the history of the field.64 

PESA uses photons to free sample electrons similar to XPS and UPS. In contrast to these techniques, the electron is not detected directly and, instead, travels through air where it binds to an oxygen molecule which is then detected via an open counter detector. This unique setup allows photoelectron spectroscopy to be carried out at atmospheric pressure and the energy range used provides electronic configuration information such as the work function and ionization potential. Yamashita provides a good introduction to this technique and the information it can provide.65 

Additional considerations/limitations:

  • Measuring liquids is possible.

  • Also possible to observe some film thickness effects.

In this article, we distinguish AFM, sometimes referred to atomic force profilometry, from optical and mechanical profilometry which operates at a larger scale range with faster scan speeds but lower resolution than AFM. These techniques are useful for measuring reasonably large feature sizes over larger areas, for example, etched features across an entire silicon wafer. Laeri and Strand show an early example of optical profilometry that serves as a decent introduction.66 

Additional considerations/limitations:

  • Lateral resolution: 500 nm–5 μm (optical), 50 nm–50 μm (mechanical).

  • Layer thickness range: 10 nm–μm.

  • Thickness mapping: Yes (optical), Limited (mechanical).

  • Mechanical profilometry resolution limited by stylus tip size and shape.

  • Optical profilometry lateral resolution is limited by light wavelength depth resolution that can be much higher due to phase based measurements.

  • Optical profilometry can access mechanically inaccessible surfaces.

QCM uses the resonant frequency of a piezoelectric quartz crystal to measure changes in the mass of deposited layers. This technique is useful for interrogating surface absorption and deposition in a range of environments from an aqueous solution to vacuum. Many instruments also measure dissipation (QCM-D), which allows for the modeling of viscoelastic energy loss effects important for softer layers. Another variant, electrochemical QCM (EQCM) allows electrochemical measurements and processes to be carried out during measurements. Easley et al. provide a good introduction to the use of QCM-D for interrogating polymer films.67 

Additional considerations/limitations:

  • Layer thickness range: 2 nm–μm.

  • Thickness mapping: No.

  • Temperature must be tightly controlled to prevent frequency drift.

  • Measurements can be performed in solution, vacuum, or ambient conditions.

Raman uses the light that is scattered nonelastically off a sample to indicate chemical properties such as vibrational modes. These measurements are useful for providing a fingerprint type measurement of different compounds. Han et al. provide an excellent primer to both Raman and surface-enhanced Raman spectroscopy (SERS), a rapidly developing form of the technique with several advantages.68 

Additional considerations/limitations:

  • Lateral resolution: 250 nm.

  • Analysis depth: 1 μm.

  • Depth profiling/mapping: Limited/Yes.

  • Detection limit: 1 × 10−3–5 × 103 ppm.

  • Sensitivity of Raman is often limited; however, surface-enhanced Raman spectroscopy (SERS) can improve sensitivity by several orders of magnitude.

  • Raman generally has better sensitivity to nonpolar bond vibrations than FTIR.

REELS measures the energy of electrons reflected off a sample; the energy lost during sample interaction can provide an indication of valence level electronic transitions such as π–π transitions. The technique is useful for filling gaps in the sample energy level diagram in combination with UPS and IPES. REELS analysis tends to be quite theoretically complex and one of the best introductions comes from Tougaard et al. illustrating their data analysis software QUEELS.69 

  • Additional considerations/limitations:

  • • Vacuum required.

  • • Data analysis relies heavily on theoretical algorithms employed.

SEM uses an electron beam scanned across a sample and measures either the backscattered or secondary electrons (Fig. 6) produced to build an image. This technique is useful for obtaining morphological information on a wide range of samples. Images can be built by measuring either backscattered electrons, which produce more contrast between regions of distinct atomic composition, or secondary electrons, which produce more contrast from topography. The electron source in SEM is also commonly used for secondary analysis techniques including energy dispersive x-ray analysis (EDX), electron backscatter diffraction (EBSD), and as a fluorescence probe. Goldstein et al. provide a comprehensive and practical textbook introduction to SEM and associated analysis techniques.48 Mutalib et al. provide useful insight into the use of SEM for the characterization of membranes.71 

FIG. 6.

Scanning electron microscopy (SEM) imaging schematic, high energy electrons are directed toward the sample surface. These electrons can then interact elastically, in which case, they are referred to as backscattered electrons and detected via a backscatter detector (BSD), alternatively, the electron beam can excite sample electrons, which are then ejected with lower energy and detected via a secondary electron detector. (a) Example of secondary electron image sensitive to topology. (b) Example of the same region imaged with backscattered electrons providing more contrast between regions of distinct composition. SEM images reproduced with permission from Fernandez Bordín et al., Mater. Charact. 195, 112525 (2023). Copyright 2023 Elsevier.

FIG. 6.

Scanning electron microscopy (SEM) imaging schematic, high energy electrons are directed toward the sample surface. These electrons can then interact elastically, in which case, they are referred to as backscattered electrons and detected via a backscatter detector (BSD), alternatively, the electron beam can excite sample electrons, which are then ejected with lower energy and detected via a secondary electron detector. (a) Example of secondary electron image sensitive to topology. (b) Example of the same region imaged with backscattered electrons providing more contrast between regions of distinct composition. SEM images reproduced with permission from Fernandez Bordín et al., Mater. Charact. 195, 112525 (2023). Copyright 2023 Elsevier.

Close modal

Additional considerations/limitations:

  • Lateral resolution: 1–50 nm.

  • Layer thickness range: 1 nm–μm (cross sectional).

  • Thickness mapping: Limited.

  • Nonconductive samples generally have to be coated with a thin conductive layer (C, Au, Pd, or Ir) choice of coating should be informed by imaging and EDX requirements.

  • Electron backscatter diffraction (EBSD) is a technique often coupled with SEMs that is able to interrogate a local crystal structure and orientation.

  • Electron energy influences the surface sensitivity of imaging and should also be tailored to a sample material.

  • Environmental SEM (ESEM) is a variant that allows wet and nonconductive samples to be images, generally at the cost of image quality, as electrons are scattered more by gases present.

  • Magnetic samples pose instrument contamination risks and will degrade the image quality by deflecting electrons; however, analysis is often still possible.

SIMS is a surface sensitive mass spectrometry (MS) technique that uses an ion beam to ionize sample surface components (Fig. 7). This technique is useful for detecting trace elements present at a sample surface due to the capability to achieve attomolar detection limits. SIMS also has one of the highest spatial resolutions of the composition measurement techniques. It is worth noting that SIMS has many variants: static, dynamic, cluster, time-of-flight (ToF), gentle, and Orbi SIMS. These variants tend to be specialized for particular applications/characteristics including biological studies, depth profiling, or surface sensitivity as such, not all instruments may be suitable for a specific sample. McPhail provides a summary of the use of SIMS37 in material science while Seebauer and Barlaz discuss the recent advances in nano-SIMS.74 

FIG. 7.

Secondary ion mass spectrometry (SIMS) measurement schematic and data. (a) SIMS uses an ion beam (M+) to sputter the top layer of the sample down to ∼1 nm, ionized sample atoms and molecules termed secondary ions (Ms+) are then passed through a detector capable of resolving ion weight with extremely high resolution. (b) A typical annotated readout is reproduced with permission from Kingshott et al., Biomaterials 23, 4775–4785 (2002). Copyright 2002, Elsevier. (c) Example of 3D depth mapping of corrosion products possible with the time of flight SIMS reproduced with permission from Yue et al., Corros. Sci. 210, 110833 (2023). Copyright 2023 Author(s) licensed under a Creative Commons Attribution (CC BY) license.

FIG. 7.

Secondary ion mass spectrometry (SIMS) measurement schematic and data. (a) SIMS uses an ion beam (M+) to sputter the top layer of the sample down to ∼1 nm, ionized sample atoms and molecules termed secondary ions (Ms+) are then passed through a detector capable of resolving ion weight with extremely high resolution. (b) A typical annotated readout is reproduced with permission from Kingshott et al., Biomaterials 23, 4775–4785 (2002). Copyright 2002, Elsevier. (c) Example of 3D depth mapping of corrosion products possible with the time of flight SIMS reproduced with permission from Yue et al., Corros. Sci. 210, 110833 (2023). Copyright 2023 Author(s) licensed under a Creative Commons Attribution (CC BY) license.

Close modal

Additional considerations/limitations:

  • Lateral resolution: 20 nm–10 μm.

  • Analysis depth: 1 nm.

  • Depth profiling/mapping: Yes/Yes.

  • Layer thickness range: 0.5 nm–μm.

  • Thickness mapping: Limited.

  • Detection limit: 2 × 10−5–1 ppm.

  • Vacuum required.

  • Destructive measurement.

  • Various ion etching modes are available.

  • Instruments tend to be highly specialized.

  • Most common analysis is qualitative due to the challenges in correcting for ion yield of different species.

  • Depth profiling is generally carried out at a much lower lateral resolution to increase accuracy.

Surface plasmon resonance is an optical technique that is sensitive to changes in the refractive index within ∼300 nm of a metallic sensor surface. It is useful, and commonly used for measuring biological interactions and kinetics of macromolecules, cells, and nanoparticles in aqueous systems. Van Der Merwe provides an excellent practical introduction to the technique and applications,75 while Tang et al. provide a concise introduction to the theory behind this technique.76 

Additional considerations/limitations:

  • Incorporation into flow systems makes affinity and kinetic measurements.

  • Systems often have limited throughput.

UPS measures electrons produced under UV irradiation. These electrons originate from valence shells and give an indication of the electronic structure of occupied orbitals. This information is useful for applications in catalysis, photovoltaics, and organic electronics. In combination with IPES, material bandgaps can be measured. Whitten provides a practical introduction to UPS.77 

Additional considerations/limitations:

  • Depth profiling possible.

  • Vacuum required.

  • Elimination of charging important for meaning full data.

  • Possible to measure sample work function.

  • Preparation of layers of multiple thicknesses may be necessary to isolate contributions from film vs film-substrate interface.

XRD uses the diffraction of x rays as they travel through a sample to interrogate atomic arrangement/structure. This includes crystal grain size and interatomic distances. Using XRD, it is also possible to solve the structure of crystallized organic molecules such as proteins, and as such, it is extremely widely used outside of surface analysis. Variants include single crystal, powder, and grazing angle XRD (GI-XRD). GI-XRD is a useful variant for obtaining more surface sensitive information. Kobayashi et al. provide a good introduction to XRD for the analysis of thin films.78 Widjonarko discusses more advanced techniques and their application to polymeric films.79 

Additional considerations/limitations:

  • Sensitive to film thickness, but XRR is generally superior for smooth surfaces.

  • Libraries are available to match spectra with known compounds.

  • Synchrotron sources provide many benefits to XRD for surface analysis.

X-ray fluorescence uses x rays to produce element characteristic x-ray fluorescence that can be used to obtain sample composition. Grazing angle and total-reflection XRF (GI-XRF and TXRF) are useful variants of XRF for obtaining more surface sensitive information. Schwenke et al. present a useful introduction to TXRF,80 and Waychunas provides a broad introduction to GI-XRF.81 

Additional considerations/limitations:

  • Lateral resolution: 100 μm.

  • Analysis depth: 10 nm (angle resolved)–10 μm.

  • Depth profiling/mapping: Yes (angle resolved)/Limited.

  • Detection limit: 1–4 × 102 ppm.

  • Possible to measure film thickness but XRR is generally superior for smooth surfaces.

  • Calibrations are often necessary for XRF instruments.

  • Limited sensitivity for light elements.

XPS measures electrons produced by materials when excited by x rays (Fig. 8); these electrons have characteristic energies that can be used to quantify atomic and, often, chemical composition. Signal attenuation can be used to estimate thin film thicknesses <10 nm assuming distinct compositions. XPS is extremely useful for obtaining surface composition for a wide range of nonvolatile samples. Stevie and Donley provide an excellent introduction to XPS and theory,83 while Baer et al. provide a more in-depth collection of guides and advice for obtaining and analyzing XPS data.84 

FIG. 8.

XPS measurement schematic (a) x rays (γ) excite electrons of characteristic energy based on the electron configuration of the element of origin and to a lesser extent chemical state. Although x rays penetrate much further into the sample only electrons in the top ∼10 nm of the sample are able to escape. (b) Example XPS narrow scan of adventitious carbon on copper with chemical species fitted reproduced with permission from Gengenbach et al., J. Vac. Sci. Technol. A 39(1), 013204 (2021). Copyright 2021 American Vacuum Society.

FIG. 8.

XPS measurement schematic (a) x rays (γ) excite electrons of characteristic energy based on the electron configuration of the element of origin and to a lesser extent chemical state. Although x rays penetrate much further into the sample only electrons in the top ∼10 nm of the sample are able to escape. (b) Example XPS narrow scan of adventitious carbon on copper with chemical species fitted reproduced with permission from Gengenbach et al., J. Vac. Sci. Technol. A 39(1), 013204 (2021). Copyright 2021 American Vacuum Society.

Close modal

Additional considerations/limitations:

  • Lateral resolution: 10 μm.

  • Analysis depth: 10 nm.

  • Depth profiling/mapping: Yes/Yes.

  • Layer thickness range: 10–1000 nm (profiling), 1–10 nm (attenuation/angle resolved).

  • Thickness mapping: Limited.

  • Detection limit: 1 × 103–1 × 104 ppm.

  • Vacuum required.

  • Technique is destructive when depth profiling.

  • Sample charging can complicate data analysis, especially for nonconductive samples, instruments have incorporated various methods for neutralizing charging effects.

  • Longer measurement times, especially for depth profiles and mapping.

  • Some instruments are fitted with a liquid nitrogen cooled sample stage.

  • Angle resolved XPS can provide composition profiles of very thin films.

  • Detection limits are dependent on both trace and matrix elements.

  • Hard x-ray photoelectron spectroscopy (HAXPES) is a variant that uses higher energy x rays to probe deeper into samples and accesses higher energy core electron excitations.

XRR, also referred to as x-ray specular reflectivity, uses interference peaks in angle resolved x-ray reflection to measure film thickness and/or surface roughness. Chason and Mayer provide a good introduction to x-ray reflectivity.85 

Additional considerations/limitations:

  • Lateral resolution: 5 mm–1 cm.

  • Layer thickness range: 2–300 nm.

  • Thickness mapping: Limited.

  • Rough interfaces can significantly complicate data analysis while measuring films; low roughness can, however, be modeled.

In summary, there are many surface characterization techniques available, and many provide similar information. However, more often than not, this data will be complementary, providing deeper insights into sample surfaces at various sample depths and resolutions. Additionally, the use of sample analogs is often useful to characterize surfaces to enable the application of complementary techniques to support conclusions or to fill in blanks. The tools provided here should assist in narrowing down techniques and instruments of interest, but consultation with surface analysis experts remains crucial for ensuring that experiments are conducted efficiently and that data and conclusions will withstand robust interrogation.

K.R. acknowledges funding from the Commonwealth Scientific and Industrial Research Organisation (CSIRO) in the form of an early career research (CERC) postdoctoral fellowship.

The authors have no conflicts to disclose.

Kye J. Robinson: Conceptualization (equal); Data curation (equal); Formal analysis (equal); Investigation (equal); Visualization (equal); Writing – original draft (equal); Writing – review & editing (equal). Helmut Thissen: Conceptualization (equal); Supervision (equal); Writing – original draft (equal); Writing – review & editing (equal).

The data that support the findings of this study are available from the corresponding author upon reasonable request.

1.
D. R.
Baer
et al,
J. Vac. Sci. Technol. A
38
,
031204
(
2020
).
2.
F. A.
Stevie
,
R.
Garcia
,
J.
Shallenberger
,
J. G.
Newman
, and
C. L.
Donley
,
J. Vac. Sci. Technol. A
38
,
063202
(
2020
).
3.
G. H.
Major
et al,
J. Vac. Sci. Technol. A
38
,
061204
(
2020
).
4.
G. H.
Major
et al,
J. Vac. Sci. Technol. A
41
,
038501
(
2023
).
5.
“Document Search—Web of Science Core Collection,” Clarivate, accessed 11 April 2024.
6.
E.
Lipiec
,
A.
Wnętrzak
,
A.
Chachaj-Brekiesz
,
W.
Kwiatek
, and
P.
Dynarowicz-Latka
,
J. Colloid Interface Sci.
542
,
347
(
2019
).
7.
W. E. S.
Unger
,
T.
Wirth
, and
V.-D.
Hodoroaba
, “
Auger electron spectroscopy
,” in
Characterization of Nanoparticles
, edited by
V.-D.
Hodoroaba
,
W. E. S.
Unger
, and
A. G.
Shard
(
Elsevier, Amsterdam
,
2020
), Chap.
4.3.2
, pp.
373
395
.
8.
S.
Hashimoto
,
T.
Sakurada
, and
M.
Suzuki
,
J. Surf. Anal.
14
,
428
(
2008
).
9.
V.-D.
Hodoroaba
, “
Energy-dispersive X-ray spectroscopy (EDS)
,” in
Characterization of Nanoparticles
(
Elsevier, Amsterdam
,
2020
), pp.
397
417
.
10.
A.
Dazzi
and
C. B.
Prater
,
Chem. Rev.
117
,
5146
(
2017
).
11.
H.
Tiernan
,
B.
Byrne
, and
S. G.
Kazarian
,
Spectrochim. Acta. A
241
,
118636
(
2020
).
12.
A.
Lanzarotta
,
Appl. Spectrosc.
69
,
205
(
2015
).
13.
C. V.
Cushman
,
P.
Brüner
,
J.
Zakel
,
G. H.
Major
,
B. M.
Lunt
,
N. J.
Smith
,
T.
Grehl
, and
M. R.
Linford
,
Anal. Methods
8
,
3419
(
2016
).
14.
H. H.
Brongersma
, “
Low-energy ion scattering
,” in
Characterzation Materials
(
Wiley, Hoboken
,
2012
), pp.
1
23
.
15.
X.
Zhu
,
T.
Xu
,
C.
Peng
, and
S.
Wu
,
Front. Chem.
9
,
782432
(
2022
).
16.
S.-F.
Wang
,
X.
Li
,
R. L.
Agapov
,
C.
Wesdemiotis
, and
M. D.
Foster
,
ACS Macro Lett.
1
,
1024
(
2012
).
17.
T.
Satoh
,
H.
Niimi
,
N.
Kikuchi
,
M.
Fujii
,
T.
Seki
, and
J.
Matsuo
,
Int. J. Mass Spectrom.
404
,
1
(
2016
).
18.
B. O.
Keller
and
L.
Li
,
J. Am. Soc. Mass Spectrom.
12
,
1055
(
2001
).
19.
Y.
Kim
,
E. J.
Lee
,
S.
Roy
,
A. S.
Sharbirin
,
L.-G.
Ranz
,
T.
Dieing
, and
J.
Kim
,
Curr. Appl. Phys.
20
,
71
(
2020
).
20.
A. M.
Dowgiallo
and
D. A.
Guenther
,
J. Agric. Food Chem.
67
,
12642
(
2019
).
21.
P. M.
Rydzak
,
K. E.
Elwick
,
N.
Damaso
, and
J. M.
Robertson
,
J. Forensic Sci.
67
,
2278
(
2022
).
22.
l. J.
Gamble
and
C. R.
Anderton
,
Microsc. Today
24
,
24
(
2016
).
23.
N.
Klingner
,
R.
Heller
,
G.
Hlawacek
,
S.
Facsko
, and
J.
von Borany
,
Ultramicroscopy
198
,
10
(
2019
).
24.
M. P.
Seah
,
S. J.
Spencer
,
I. S.
Gilmore
, and
J. E.
Johnstone
,
Surf. Interface Anal.
29
,
73
(
2000
).
26.
S.
Terada
,
H.
Murakami
, and
K.
Nishihagi
, “
Thickness and density measurement for new materials with combined X-ray technique
,” in
2001 IEEE/SEMI Advanced Semiconductor Manufacturing Conference, IEEE Cat No01CH37160
Munich, 2001 (IEEE, Piscataway,
2001
), pp.
125
130
.
27.
D.
Richard
and
R.
Rousseau
,
Rigaku J.
18
,
33
(
2001
).
28.
W. E. S.
Unger
,
J. M.
Stockmann
,
M.
Senoner
,
T.
Weimann
,
S.
Bütefisch
,
C.
Passiu
,
N. D.
Spencer
, and
A.
Rossi
,
J. Vac. Sci. Technol. A
38
,
053206
(
2020
).
29.
A. G.
Shard
,
Surf. Interface Anal.
46
,
175
(
2014
).
30.
G.
Jin
,
R.
Jansson
, and
H.
Arwin
,
Rev. Sci. Instrum.
67
,
2930
(
1996
).
31.
C.
Chen
,
X.
Chen
,
C.
Wang
,
S.
Sheng
,
L.
Song
,
H.
Gu
, and
S.
Liu
,
Opt. Express
29
,
32712
(
2021
).
32.
A.
Rothen
,
Rev. Sci. Instrum.
16
,
26
(
2004
).
33.
A.
Crovetto
,
P. R.
Whelan
,
R.
Wang
,
M.
Galbiati
,
S.
Hofmann
, and
L.
Camilli
,
ACS Appl. Mater. Interfaces
10
,
25804
(
2018
).
34.
N.
Cella
,
H.
El Rhaleb
,
J. P.
Roger
,
D.
Fournier
,
E.
Anger
, and
A.
Gicquel
,
Diam. Relat. Mater.
5
,
1424
(
1996
).
35.
J.
Schmit
,
K.
Creath
, and
J. C.
Wyant
, “
Surface profilers, multiple wavelength, and white light intereferometry
,” in
Optical Shop Testing
(
Wiley, Hoboken
,
2007
), pp.
667
755
.
37.
38.
D.
Johannsmann
,
A.
Langhoff
,
C.
Leppin
,
I.
Reviakine
, and
A. M. C.
Maan
,
Sensors
23
,
1348
(
2023
).
39.
M. R.
Sardela
, “
X-ray diffraction and reflectivity
,” in
Practical Materials Characterization
, edited by
M.
Sardela
(
Springer
,
New York
,
2014
), pp.
1
41
.
40.
R.
Ogaki
,
M. A.
Cole
,
D. S.
Sutherland
, and
P.
Kingshott
,
Adv. Mater.
23
,
1876
(
2011
).
41.
C.
Yoshikawa
,
B.
Delalat
,
F.
Huang
,
S.
Braun
,
N.
Nishijima
,
N. H.
Voelcker
,
P.
Kingshott
, and
H.
Thissen
,
J. Mater. Chem. B
7
,
3520
(
2019
).
42.
W.R.
Bowen
and
N.
Hilal
,
Atomic Force Microscopy in Process Engineering: An Introduction to AFM for Improved Processes and Products
(
Butterworth-Heinemann, Burlington
,
2009
).
43.
P.
Nguyen-Tri
,
P.
Ghassemi
,
P.
Carriere
,
S.
Nanda
,
A. A.
Assadi
, and
D. D.
Nguyen
,
Polymers
12
,
1142
(
2020
).
44.
W.
Liang
,
H.
Shi
,
X.
Yang
,
J.
Wang
,
W.
Yang
,
H.
Zhang
, and
L.
Liu
,
Soft Matter
16
,
8962
(
2020
).
45.
R. S.
Hebbar
,
A. M.
Isloor
, and
A. F.
Ismail
, “
Contact angle measurements
,” in
Membrane Character
, edited by
N.
Hilal
,
A. F.
Ismail
,
T.
Matsuura
, and
D.
Oatley-Radcliffe
(
Elsevier, Amsterdam
,
2017
),
Chap. 12
, pp.
219
255
.
46.
H. B.
Eral
,
D. J. C. M.
’t Mannetje
, and
J. M.
Oh
,
Colloid Polym. Sci.
291
,
247
(
2013
).
47.
M.
Behnecke
,
E.
Berghaus
,
A.
Wildeis
,
E.
Brandes
, and
S.
Petersen
,
Surf. Interface Anal.
56
,
212
(
2024
).
48.
J. I.
Goldstein
,
D. E.
Newbury
,
J. R.
Michael
,
N. W. M.
Ritchie
,
J. H. J.
Scott
, and
D. C.
Joy
,
Scanning Electron Microscopy and X-Ray Microanalysis
(
Springer
,
New York
,
2018
).
49.
C.
Gu
,
S.-Q.
Man
,
J.
Tang
,
Z.
Zhao
,
Z.
Liu
, and
Z.
Zheng
,
Optik
221
,
165274
(
2020
).
50.
H.
Cesiulis
,
N.
Tsyntsaru
,
A.
Ramanavicius
, and
G.
Ragoisha
, “
The study of thin films by electrochemical impedance spectroscopy
,” in
Nanostructures and Thin Films for Multifunctional Applications: Technology, Properties and Devices
, edited by
I.
Tiginyanu
,
P.
Topala
, and
V.
Ursaki
(
Springer International Publishing
,
Cham
,
2016
), pp.
3
42
.
51.
S.
Wang
,
J.
Zhang
,
O.
Gharbi
,
V.
Vivier
,
M.
Gao
, and
M. E.
Orazem
,
Nat. Rev. Methods Primer
1
,
1
(
2021
).
52.
D. E.
Aspnes
,
Thin Solid Films
571
,
334
(
2014
).
53.
H. G.
Tompkins
and
J. N.
Hilfiker
,
Spectroscopic Ellipsometry: Practical Application to Thin Film Characterization
, 1st ed. (
Momentum
,
New York
,
2016
).
54.
O. D.
Lavrentovich
, “
Confocal fluorescence microscopy
,” in
Characterization Materials
(
Wiley, Hoboken
,
2012
), pp.
1
15
.
55.
A. E.
Ribbe
,
M.
Hayashi
,
M.
Weber
, and
T.
Hashimoto
,
Polymer
39
,
7149
(
1998
).
56.
M. A.
Mohamed
,
J.
Jaafar
,
A. F.
Ismail
,
M. H. D.
Othman
, and
M. A.
Rahman
, “
Fourier transform infrared (FTIR) spectroscopy
,” in
Membrane Character
, edited by
N.
Hilal
,
A. F.
Ismail
,
T.
Matsuura
, and
D.
Oatley-Radcliffe
(
Elsevier, Amsterdam
,
2017
),
Chap. 1
, pp.
3
29
.
58.
H.
Yoshida
,
Rev. Sci. Instrum.
85
,
016101
(
2014
).
59.
S.
Hajati
,
O.
Romanyuk
,
J.
Zemek
, and
S.
Tougaard
,
Phys. Rev. B
77
,
155403
(
2008
).
60.
R.C.
Bird
and
J.S.
Williams
,
Ion Beams for Materials Analysis
(
Academic, Marrickville
,
1989
).
61.
S.
Průša
,
M. R.
Linford
,
E.
Vaníčková
,
P.
Bábík
,
J. W.
Pinder
,
T.
Šikola
, and
H. H.
Brongersma
,
Appl. Surf. Sci.
657
,
158793
(
2024
).
62.
P.
Kingshott
,
H. A. W.
St John
,
R. C.
Chatelier
, and
H. J.
Griesser
,
J. Biomed. Mater. Res.
49
,
36
(
2000
).
63.
H. J.
Griesser
,
P.
Kingshott
,
S. L.
McArthur
,
K. M.
McLean
,
G. R.
Kinsel
, and
R. B.
Timmons
,
Biomaterials
25
,
4861
(
2004
).
64.
M. W.
Davidson
and
M.
Abramowitz
,
Encycl. Imaging Sci. Technol.
2
,
120
(
2002
).
65.
D.
Yamashita
,
Y.
Nakajima
,
A.
Ishizaki
, and
M.
Uda
,
J. Surf. Anal.
14
,
433
(
2008
).
66.
F.
Laeri
and
T. C.
Strand
,
Appl. Opt.
26
,
2245
(
1987
).
67.
A. D.
Easley
,
T.
Ma
,
C. I.
Eneh
,
J.
Yun
,
R. M.
Thakur
, and
J. L.
Lutkenhaus
,
J. Polym. Sci.
60
,
1090
(
2022
).
68.
X. X.
Han
,
R. S.
Rodriguez
,
C. L.
Haynes
,
Y.
Ozaki
, and
B.
Zhao
,
Nat. Rev. Methods Primer
1
,
1
(
2022
).
69.
S.
Tougaard
,
N.
Pauly
, and
F.
Yubero
,
Surf. Interface Anal.
54
,
820
(
2022
).
70.
S. P.
Fernandez Bordín
,
J. M.
Ranalli
,
V.
Galván
,
G.
Castellano
, and
S.
Limandri
,
Mater. Charact.
195
,
112525
(
2023
).
71.
M.
Abd Mutalib
,
M. A.
Rahman
,
M. H. D.
Othman
,
A. F.
Ismail
, and
J.
Jaafar
, “
Scanning electron microscopy (SEM) and energy-dispersive x-ray (EDX) spectroscopy
,” in
Membrane Character
, edited by
N.
Hilal
,
A. F.
Ismail
,
T.
Matsuura
, and
D.
Oatley-Radcliffe
(
Elsevier
, Amsterdam,
2017
),
Chap. 9,
pp.
161
179
.
72.
P.
Kingshott
,
S.
McArthur
,
H.
Thissen
,
D. G.
Castner
, and
H. J.
Griesser
,
Biomaterials
23
,
4775
(
2002
).
73.
X.
Yue
,
P.
Malmberg
,
E.
Isotahdon
,
V.
Ratia-Hanby
,
E.
Huttunen-Saarivirta
,
C.
Leygraf
, and
J.
Pan
,
Corros. Sci.
210
,
110833
(
2023
).
74.
E. G.
Seebauer
and
D. E.
Barlaz
,
Curr. Opin. Chem. Eng.
12
,
8
(
2016
).
75.
P. A.
Van Der Merwe
, “
Surface plasmon resonance
,” in
Protein-ligand interactions: hydrodynamics and calorimetry
, edited by B. Chowdhry and S. Harding (
Oxford University
,
Oxford
,
2001
)
76.
Y.
Tang
,
X.
Zeng
, and
J.
Liang
,
J. Chem. Educ.
87
,
742
(
2010
).
77.
J. E.
Whitten
,
Appl. Surf. Sci. Adv.
13
,
100384
(
2023
).
78.
S.
Kobayashi
and
K.
Inaba
,
Rigaku J
32
,
1
(
2016
).
80.
H.
Schwenke
,
P. A.
Beaven
, and
J.
Knoth
,
Fresenius J. Anal. Chem.
365
,
19
(
1999
).
81.
G. A.
Waychunas
,
Rev. Mineral. Geochem.
49
,
267
(
2002
).
82.
T. R.
Gengenbach
,
G. H.
Major
,
M. R.
Linford
, and
C. D.
Easton
,
J. Vac. Sci. Technol. A
39
,
013204
(
2021
).
83.
F. A.
Stevie
and
C. L.
Donley
,
J. Vac. Sci. Technol. A
38
,
063204
(
2020
).
84.
D. R.
Baer
,
G. E.
McGuire
,
K.
Artyushkova
,
C. D.
Easton
,
M. H.
Engelhard
, and
A. G.
Shard
,
J. Vac. Sci. Technol. A
39
,
021601
(
2021
).
85.
E.
Chason
and
T. M.
Mayer
,
Crit. Rev. Solid State Mater. Sci.
22
,
1
(
1997
).