Applications of Time-of-Flight Secondary Ion Mass Spectrometry

Time-of-flight secondary ion mass spectrometry (ToF-SIMS) is extremely surface sensitive and offers superior chemical selectivity, making it a powerful and often unique technique for identifying chemical structures and exploring surface chemistry. With its imaging capability, ToF-SIMS is particularly useful in materials failure analyses, where both the identification of chemicals and pinpointing their localities are required. Assisted by a sputter ion beam, ToF-SIMS is also capable of depth profiling both organic and inorganic materials. This allows for the exploration of chemical variability in the depth direction without prior knowledge of the sample under investigation, thanks to its parallel detection of all generated ions.

• probing the outermost 1-3 nm surface
• identifying organic molecules
• imaging and depth profiling

I hope the examples presented here will help you appreciate how ToF-SIMS can assist your research on surface chemistry, solve surface-associated problems, and develop analytical approaches to understanding materials failure mechanisms.

How ToF-SIMS works?

In ToF-SIMS, such as the IONTOF TOF-SIMS IV at Surface Science Western, a pulsed (~1 ns) primary ion beam (e.g., 25 keV Bi₃⁺) is used to bombard the sample surface. This bombardment generates secondary particles, including ions, electrons, and neutral particles, which carry chemical information about the surface. Depending on the polarity of the electric field applied between the extractor and the sample stage, either positive or negative ions are extracted one at a time by an electric field (e.g., 2 kV). The extracted ions gain energy of 2 keV (assuming they are singly charged). These energetic ions then travel through a flight tube (2 m in length). Lighter ions fly faster and thus reach the detector before the heavier ones. The basic concept of ToF-SIMS is illustrated below (in the case of positive ion extraction; the polarity of the extraction electric field is reversed for negative ion extraction).

The ions arrive at the detector based on their flight times, which are determined by their mass-to-charge ratio (m/z). As illustrated in the following figure, the measured time events of the ions striking the detector are converted to m/z through calibration with known species such as hydrogen, carbon, and hydrocarbons, as well as any other detected ions. Depending on the size of the scanned area, mass resolutions (defined as the center of a peak divided by its full width at half maximum) can be as high as 10,000.

After the extraction of the secondary ions is complete, a low-energy (18 eV) electron beam floods the scanned area for charge compensation until the next pulse of the primary ion beam is shot. This capability allows ToF-SIMS to be applied to highly insulating materials. All these events occur within the cycle time. Within each cycle time, the first ion to arrive at the detector is a hydrogen ion, and the heaviest ion arrives last, determined by the time the system allows for ion detection. For example, with a cycle time of 100 μs (or a repetition rate of 10 kHz), the cutoff is around m/z 900—anything heavier than this will not reach the detector before the system starts the next cycle. Therefore, to detect higher mass species, the cycle time must be increased (e.g., m/z up to 1600 can be covered by a cycle time of >140 μs). The base pressure of the analysis chamber of a ToF-SIMS instrument can be as good as 1×10^-9 mbar. However, it will work with no problems even if at pressures around 1×10^-5 mbar.

• parallel detection: all generated ions are recorded.
• high mass resolution: up to 10,000.

The primary beam rasters the sample surface pixel by pixel (e.g., 128 pixels × 128 pixels) over the scanned area (e.g., 500 μm × 500 μm). Each pixel has a mass spectrum, enabling ion imaging when the intensities of selected ions are plotted against the pixels. Ion images reveal the spatial distributions of chemicals over the scanned area. The spatial resolution is a couple of microns in high mass resolution mode (the so-called high current bunched mode). The primary ion fluence is usually controlled at ~10¹¹ ions/cm², which is much less than the static limit of ~10¹³ ions/cm². This is why ToF-SIMS is referred to as static SIMS, meaning that the surface is not significantly altered after measurement, allowing the same surface to be subjected to other analyses. Accordingly, with a sputter ion beam (e.g., Cs⁺ and C₆₀⁺) used to remove a controllable portion of materials, ToF-SIMS is also capable of elemental and molecular depth profiling.

ToF-SIMS is highly surface-sensitive, probing only the outermost 1-3 nm of the surface. Due to varying ion yields for different species and changes in these yields based on the chemical environment (matrix effect), ToF-SIMS is not a quantitative technique. Quantification of specific elements requires a reference. Therefore, it is not practical to compare different species within the same sample. However, ToF-SIMS results, specifically ion intensities, can be conveniently used to compare the same species across different samples. For example, one might find that sodium (Na⁺) is several times more abundant in one sample compared to another.

ToF-SIMS secondary ion mass spectra are rich in chemical information, capturing all ions generated during the bombardment by the primary ion beam. This technique often provides unique chemical selectivity for organic materials through characteristic ions, often molecular ions. This capability allows for the exploration and development of unique analytical approaches to reveal surface chemistry and identify chemical structures.

ToF-SIMS is also referred to as static SIMS because it collects chemical information from the surface while sputtering off 1% or less of the topmost layer of the specimen, thus leaving the surface chemistry essentially undiminished. The term “static” is used to contrast with dynamic SIMS, which was developed prior to static SIMS. In dynamic SIMS, the primary ion beam sputters through the specimen while detecting its chemical constituents, making it suitable for applications in depth profiling elements (e.g., dopants in semiconductors and precious metals in ores).

Ion mass spectra

Surface sensitivity and chemical selectivity of ToF-SIMS

ToF-SIMS is characterized by its surface sensitivity and chemical selectivity. For instance, if there is only a monolayer of some molecules, many other techniques may lack either surface sensitivity or chemical selectivity, or both. In contrast, a monolayer of a substance is more than enough to be detected by ToF-SIMS. As an example, the negative (upper panel) and positive (lower panel) secondary ion mass spectra shown below were obtained from a stain visible on a metal substrate, which only showed slightly increased carbon levels when detected by two other techniques. ToF-SIMS helped identify the stain as being due to the presence of diacylglycerides (also called diglycerides). The fatty acid chains are seen as negative ions corresponding to fatty acids, such as palmitate C₁₆H₃₁O₂¯ at m/z 255 and myristate C₁₄H₂₇O₂¯ at m/z 227. With the removal of a hydroxyl group (OH)  from the glyceryl link, glyceryl dipalmitate (C₁₆H₃₁O₂)₂C₃H₅⁺ is detected as a positive ion at m/z 552. As seen in the spectra, other combinations with different fatty acids are also detected.

Another example for identification of Cyanox 1790, whose molecular formula is [M=C₃O₃N₃(C₁₃H₁₉O)₃], is shown below. The molecular ions M⁺ and [M-H]¯ are detected at m/z 699 and 698, respectively, though both are rather weak. The more abundant ions are C₁₃H₁₉O⁺ at m/z 191 and C₃O₃N₃(C₁₃H₁₉O)₂¯ at m/z 508. This example, along with the previous one, demonstrates the superior chemical selectivity of ToF-SIMS.

• extremely surface sensitive
• superb chemical selectivity

Vancomycin. with the molecular formula, M=C₆₆H₇₅Cl₂N₉O₂₄, is a Gram-positive bacteria antibiotic. Shown below are negative and positive secondary ion mass spectra for a vancomycin layer spin-coating on a cleaved mica substrate. There is an abundant negative ion at m/z 155 and an an abundant positive ion at m/z 100. These peaks have been identified as C₇H₄O₂Cl¯ and C₆H₁₄N ⁺, respectively.

These two ions are fragmented from the two structural moieties of the vancomycin molecule illustrated below. Additionally, though at much lower intensities, the protonated molecular ion [M+H]⁺ (m/z 1448) and decarboxylated molecular ion [M-COOH]¯ (m/z 1402) were also detected.

Other chemicals, such as pigments, antibiotics, and surfactants, have been examined. Their molecular ions are often detected using ToF-SIMS, and they usually also have diagnostic fragmented ions. To reiterate, the superior chemical selectivity and surface sensitivity of ToF-SIMS often make it powerful in investigating surface contaminants. These contaminants may be as thin as a couple of nanometers, yet they can be either visible and/or impact the surface chemistry of the contaminated substrate.

Isotopes of elements

ToF-SIMS is also useful for detecting isotopes of elements and molecules containing them. For substances that are not isotope-enriched, the ion mass spectra show the natural abundance of the isotopes. In fact, the isotope distribution pattern of an element is a useful property for identifying elements and molecules/fragments. Shown below are ion mass spectra for tin, copper, and a fragment ion AuCl₂.

Shown below is the negative secondary ion mass spectrum of Pigment Yellow 110 (M=C₂₂H₆Cl₈N₄O₂). The pigment molecule has eight chlorine atoms, each of which has two isotopes at m/z 35 and 37, resulting in a complex group of peaks for the deprotonated molecular ion. The most abundant peak is called the monoisotopic peak, which in this case is at m/z 641. These isotope patterns can be checked using the Isotope Distribution Calculator and Mass Spec Plotter provided by Scientific Instrument Services.

Ion mass spectra of silicones

Silicones are widely used in plastic industry as release agents and paint systems as wetting agents. They are also common in many personal cosmetic products. Consequently, through plastic packaging and handling silicones often contaminate the surfaces of various objects. Due to their low surface energy (~24 mJ/m²), silicones cab cause adhesion failures and paint defects. Even trace amounts of silicones can significantly alter the surface energy of the contaminated substrate. ToF-SIMS is particularly powerful in identifying siloxane contamination as the cause of automative paint craters, thanks to its superior surface sensitivity and chemical selectivity.

Shown below are negative secondary ion mass spectra of different silicones. While both spectra contain identical ions, the ion intensity ratios between some ions differ for the two silicones. This is crucial chemical information because it helps differentiate molecules that contain the same elements. For example, for the two silicones, the ion intensity ratios for ions at m/z 89 (SiC₃H₉O¯) and 91 (SiC₂H₇O₂¯), as well as those at m/z 163 (Si₂C₅H₁₅O₂¯) and 165 (Si₂C₄H₁₃O₃¯), are reversed. This difference is indeed a reflection of the abundance and lack of Si(CH₃)₃ in polydimethylsiloxanes and cyclosiloxane, respectively. This trend is also observed for Sylgard 184, a silicone elastomer kit from Dow Corning, with varying degrees of cross-linking.

Shown below is a list of the major negative and positive ions (and their m/z) fragmented from silicones, along with the fragmentation patterns based on the molecular formula of poly(dimethyl siloxane) (PDMS), a silicone oil.

Negative ions

• Si¯ (m/z 27.977)
• SiCH₃O¯ (58.995)
• SiC₂H₅O¯ (73.011)
• SiCH₃O₂¯ (74.991)
• SiC₃H₉O¯ (89.043)
• SiC₂H₇O₂¯ (91.021)
• Si₂C₃H₉O₃¯ (149.009)
• Si₂C₅H₁₅O₂¯ (163.061)
• Si₂C₄H₁₃O₃¯ (165.041)
• Si₃C₅H₁₅O₄¯ (223.027)
• Si₃C₇H₂₁O₃¯ (237.081)

Positive ions

• Si⁺ (m/z 27.977)
• SiC₃H₉⁺ (73.047)
• Si₂C₅H₁₅O⁺ (147.065)
• Si₃C₅H₁₅O₃⁺ (207.033)
• Si₃C₇H₂₁O₂⁺ (221.086)
• Si₄C₇H₂₁O₄⁺ (281.051)
• Si₅C₇H₂₁O₅⁺ (325.024)
• Si₅C₉H₂₇O₅⁺ (355.070)

• Silicones, due to their low surface energy, are widely used as wettability enhancers and release agents.
• However, this same surface property can cause failures when silicones contaminate a substrate that is subjected to adhesion or painting applications.

Ion images

Paint craters

ToF-SIMS imaging is invaluable for studying the localization of chemicals on a surface. This technique is particularly useful when the presence of a chemical is critical but may be overlooked due to the weak intensities of its ions, which can result from its minimal size when only considering the ion mass spectra collected over a scanned area. For instance, paint craters can be caused by tiny perfluorocarbon particles. These particles, with their low surface energy (approximately 18 mJ/m²), obstruct the wetting of the paint, and their contribution to the mass spectrum collected over a large area may be buried or hardly noticeable. However, by mapping the scanned area, say 500 μm × 500 μm, even a perfluorocarbon particle just a few microns across can be detected due to its contrast. This approach is essential in ToF-SIMS studies on the causes of paint craters. Shown below are positive ion images of an automotive paint crater caused by a perfluorocarbon particle. Perfluorocarbons are characterized by C⁺, CF⁺, CF₃⁺ and C₃F₅⁺, as well as other CxF_y⁺ ions.

• imaging critical for identifying tiny low-surface-energy particles that cause paint craters.

Other chemicals classified to cause automotive paint cratering include silicones, fatty acids and detergents. The common property of those agents lies in their low surface energy, perhaps ranging in 20-25 mJ/m². It is worth noting that silicones are also used as an ingredient in some paint systems to enhance wettability. In this case, silicones lower the surface tension (measured in mN/m, numerically equivalent to surface energy in mJ/m²) of the paint, allowing it to better wet a substrate that does not necessarily have high surface energy. Most plastics, unless modified to increase surface energy, typically have a low surface energy, around 25-35 mJ/m². Paint craters form when the paint cannot wet a spot on the substrate due to its lower surface energy, which may be caused by the presence of a particle with even lower surface energy. Therefore, both aggregates of silicones in the paint system and contamination by foreign silicones can cause paint craters. It is sometimes necessary to determine which is the case when silicone is found to cause the crater. In such instances, comparing negative silicone ions at m/z 89 and 91 and at m/z 163 and 165, as described earlier , can be quite helpful.

The automotive painting process, which includes substrate preparation, primer, basecoat and clearcoat, accounts for roughly one-third of the cost of manufacturing vehicles. The appearance of a car significantly influences customers’ purchasing decisions. Paint cratering can lead to a shutdown of the production line. ToF-SIMS is the most effective technique for identifying the causes of paint cratering, as it often relates to tiny particles or extremely thin layers of materials with low surface energy. Detecting these defects requires the surface sensitivity and chemical selectivity that only ToF-SIMS can provide.

Differentiating fatty acids and their salts

Shown here is a composite image of Si⁺ (blue), C₁₆H₃₃O₂⁺ (green) and C₁₆H₃₃O₂Na₂⁺ (red). This image demonstrates the excellent chemical selectivity of ToF-SIMS, as it differentiates palmitic acid from its salt (sodium palmitate). In the salt form, almost no C₁₆H₃₃O₂⁺ is detected. Both the acid and its salt, however, have abundant negative ion C₁₆H₃₁O₂¯ (not shown). Thus, the chemical selectivity for this acid and its salt lies in the positive ions rather than the negative ones. This also applies to other fatty acids and their salts. This observation can be explained by the fact that for a fatty acid to become a protonated molecular ion, it only needs to acquire one hydrogen atom or proton. In contrast, for the salt of a fatty acid to become a protonated molecular ion, it must acquire two hydrogen atoms or protons. This ion formation process is hindered by the presence of the cation (e.g., Na_⁺). The ToF-SIMS results suggest that it is easier for the salt to acquire one more cation to become a positive ion of the salt with an extra cation than to capture two hydrogen atoms or protons.

Phase separation

Phase separation of the antibiotic vancomycin and its polymeric matrix poly(lactide-co-glycolide) (PLGA) was observed when dimethyl sulfoxide (DMSO) was used as the solvent. The sample was prepared by placing the solution on a cleaved mica substrate. The phase separation is evident in the ion images of C₇H₄O₂Cl¯ (at m/z 155, representing vancomycin) and C₆H₇O₅¯ (representing PLGA). For the vancomycin molecule, C₇H₄O₂Cl¯, along with its isotope ion C₇H₄O₂^{37Cl¯ (not shown), is a diagnostic ion that is both abundant and a structural entity of the molecule. Additionally, sulfur species were found in the vancomycin phase, suggesting a possible interaction between vancomycin and DMSO.}

Depth profiling

Metal oxides

Assisted by a sputter ion beam, ToF-SIMS can be used to depth profile both inorganic and organic materials. In practice, an area is sputtered for a predetermined period, followed by analyzing the newly generated surface within the sputtered area, which provides a data point in the depth profile. This sputter/analysis cycle is repeated until the desired depth is profiled. In most cases, the depth of the crater generated from depth profiling is measured using either a mecahnical stylus or an optical profilometer to calibrate the depth. Otherwise, the depth information is expressed by sputter time. Subnanometer depth resolutions are readily achievable. Shown below are an optical image (top left corner) of the crater after depth profiling and depth profiles (right) of a layered structure of ZnO/In₂O₃/Si using a 3 keV Cs⁺ sputter beam and a 25 keV Bi₃⁺ analysis beam. The zinc oxide layer is represented by ZnO¯, indium by InO¯ and silicon by Si¯. AAlso shown is a cross-section image of the profiled depth. Note that the vertical dimension represents the profiled depth, while the lateral dimension represents the scanned area.

• dual beam approach for depth profiling
• subnanometer depth resolutions
• powerful for investigating deposited thin films and native oxide layers

Shown here are depth profiles for a structure consisting of alternating layers of different oxides of metals M1, M2, and M3. Carbon and carbon-related ions are also displayed. A peak of carbon is observed in the middle of the film, suggesting increased carbon levels due to an interruption during the deposition process (i.e., accumulation of hydrocarbons in vacuum). Therefore, ToF-SIMS depth profiling is powerful for investigating the quality and integrity of layered structures.

Layered organics

Depth profiling organic materials often requires appropriate sputter ion beam that will not significantly degrade the molecules. C₆₀⁺ is an effective sputter beam for many organic materials. Vancomycin is used to showcase the ability of ToF-SIMS to perform three-dimensional molecular imaging. Shown below are the depth profiles of C₇H₄O₂Cl¯and C₃H₃O₃¯, representing vancomycin and PLGA, respectively. The ion intensities for a single depth are the sums of those over the scanned area. Since each pixel has a full spectrum over the scanned area for all probed depths, ToF-SIMS provides a three-dimensional distribution of chemical information. For example, shown on the right-hand side in the following figure is a cross-section, i.e., an x-z plane cut from the profiled depth. Note that the vertical direction represents the depth (z, 290 nm) and its scale is different from the dimension (x, 200 μm) of the scanned area.

Quantifying cross-linking degrees of PMMA

An approach based on ToF-SIMS was developed to determine the degree of cross-linking on the surface and its variations in a nanometer-scale depth of organic materials. ToF-SIMS is extremely surface-sensitive and capable of depth profiling using a sputter ion beam to remove controllable amounts of substance. Poly(methyl methacrylate) (PMMA) films spin-coated on a Si substrate were cross-linked using a recently developed, surface-sensitive, hyperthermal hydrogen projectile bombardment technique. The ion intensity ratio between two ubiquitous hydrocarbon species C₆H¯ and C₄H¯, detected in ToF-SIMS, denoted as ρ, was used to assess the degree of cross-linking of the PMMA films. The cross-linking depth of the PMMA films was revealed by depth profiling ρ into the polymer films using the C₆₀⁺ sputter beam.

The control PMMA film spin-coated on a Si substrate was characterized by ρ=32% on its surface when using a 25 keV Bi₃⁺ primary ion beam. This parameter on the PMMA films subjected to HHIC treatment for 10, 100, and 500 seconds increased to 45%, 56%, and 65%, respectively. The depth profiles of ρ obtained using a 10 keV C₆₀⁺ ion beam resembled an exponential decay, from which the cross-linking depth was estimated to be 3, 15 and 39 nm, respectively, for the three cross-linked PMMA films.

As shown in panel (a) of the figure, the characteristic PMMA ion C₄H₅O₂¯ indicates the variation of cross-linking degree of PMMA. Panel (b) shows that the ion intensity ratio ρ between C₆H¯ to C₄H¯ detected in ToF-SIMS provides a unique and simple means to assess the degree of cross-linking of the surface of PMMA films cross-linked by the surface sensitive hyperthermal hydrogen projectile bombardment technique. The ToF-SIMS approach is capable of depth profiling ultra-thin organic films with subnanometer resolutions.

• opportunities to developing new analytical approaches

Principal component analysis of C_nH¯ ions

The rich chemical information provided by ToF-SIMS ion mass spectra lies in the form of fragmented ions, often numbering in the hundreds. This abundance of ions enables the identification of chemicals and the exploration of surface chemistry. Given the daunting number of ions (i.e., variables) detected in ToF-SIMS, dimensionality reduction techniques such as principal component analysis (PCA), a multivariate data analysis method, have proven useful. PCA reveals similarities or differences in ions based on the variability of their intensities and can differentiate polymers with different molecular weights or different types of polymers.

In PCA, the original variables (i.e., ions in the ToF-SIMS case) are transformed into a much smaller number of new orthogonal variables (i.e., principal components or PCs), which are linear combinations of the original variables. The transformation is done such that the first PC accounts for as much of the variance in the original data as possible, with subsequent PCs capturing the remaining maximum variances. Although there are as many PCs as the original variables, one can discard the higher PCs without losing much information, as the first few PCs explain the vast majority of the variance (thus achieving dimensionality reduction). However, PCA is merely a mathematical transformation of the original data—one still needs to interpret what the first PCs represent, which is inherently data- or problem-specific.

• As a dimensionality reduction technique, PCA offers visualization of clustering (or lack thereof) among observations, contributions of variables to clustering, and relationships among variables.

PCA allows for the comparison of numerous observations (data) across multiple variables in a biplot. This biplot is constructed by plotting the scores of the observations on two principal components (PCs), often the first two, overlapped with a plot of the loadings of the variables on the same two PCs. A biplot visualizes the similarities and differences among the multivariate observations, their relationships with the variables, and the correlations between the variables.

Shown in the biplot below is a covariance biplot for the scores of the observations as points and loadings of the variables as arrowed lines, both on PC1 and PC2, for polyethylene (PE), polypropylene (PP), polyisoprene (PIP) and polystyrene (PS). The PCA results were obtained from a data set of 39 observations for the four polymers against the 10 C_nH¯ (n=1 to 10) ions, using the prcomp() function in the open-source R language. This function is based on singular value decomposition , ( svd() function in R), returning scores of observations, loadings of variables, and eigenvalues (i.e., variances of PCs).

The score of an observation on a PC is the sum of the individual ion intensity of each variable multiplied by the loading of the corresponding variable on the PC. In other words, it is the projection of all the C_nH¯ intensities of an observation on a PC, representing how an observation is expressed as a single point on the PC axis. A score plot is used to determine similarities or differences among observations. The score plot in the figure shows that the four polymers are separated into four distinct groups.

An arrowed line of a variable may be called a “variable vector,” indicating the direction of the variable in the PC coordinates. The smaller the angle between two variable vectors of two C_nH¯ions, the more similarly the two variables behave in terms of their variability in ion intensities. In other words, two variables are positively or negatively correlated if the angle between their variable vectors is close to 0 or 180°, respectively. Conversely, an angle close to 90° indicates that the two variables are not correlated at all.

The loading plot in the figure clearly shows the similarities among C₆H¯ to C₁₀H¯ due to their significant contributions to PC1. In contrast, the variable vectors of C₂H¯ and C₃H¯ point in approximately the opposite direction and have increased loadings on PC2. This indicates that the intensities of the two groups of ions vary in opposite directions; that is, when the ion intensities of one group measured on a polymer sample increase, those of the other group decrease.

• PC1 captures the variability of "carbon density" of the four polymers.

Assisted by the PCA results, we confirmed that polymers with lower and higher “carbon density” favor the formation of smaller (e.g., C₂H¯ and C₃H¯ and larger (i.e., C₆H¯ to C₁₀H¯) C_nH¯, respectively. Moreover, we found that the variability of the C₄H¯ intensity across different polymers is relatively small compared to other C_nH¯ intensities. The PCA results shown in the figure verified our argument that with increased "carbon density", the ion intensities of larger C_nH¯ increase while those of smaller C_nH¯ decrease. Therefore, it is the "carbon density" of polymers that dictates the variability of C_nH¯ captured by PC1.