TOF‐SIMS: Introduction - mikee9265/SIMS-Wiki GitHub Wiki

Pulse the primary ion beam. The primary ion beam must be accelerated to a uniform energy and then filtered so that the ions all have the same mass, and will therefore arrive at the sample all at the same time. Each pulse of primary ions will produce a pulse of secondary ions resulting from the impact of the primary ions with the surface. As it turns out, the energies of the emitted secondary ions will all be fairly low, mostly less than 100 eV for the atomic ions, less than 10 eV for cluster and organic fragments, and even less for molecular ions. Apply an acceleration voltage to all of the secondary ions, in the process accelerating them towards the detector. Capture the time of arrival at the detector relative to the initial primary ion pulse for each secondary ion. The result is a TOF-SIMS spectrum.

Because all of the secondary ions with a single charge are given the same energy when they are accelerated towards the detector, and given the fact that E = 1/2 mv² where E is the energy, m is the secondary ion mass, and v is the ion’s velocity, the different masses of the secondary ions will result in their having different velocities. In fact v² = 2*E/m which means that the lightest singly charged ion H⁺ will be the fastest, while the heaviest ion you can detect will be the slowest. Fortunately, multiple charged ions are rare. They will receive multiples of the accelerating energy and will arrive at the detector that much faster. The software of the instrument can convert the raw spectrum (ion counts as a function of arrival time) to a mass spectrum (ion counts as a function of m/e, mass to charge ratio).

Any variation in the initial energies of the secondary ions will lead to a variation in the time of arrival at the detector. This leads directly to uncertainty in the mass of the secondary ion detected. For improved mass resolution and mass accuracy, commercial TOF-SIMS systems are all more sophisticated than the simple linear system described earlier. By inserting appropriate electrostatic fields into the secondary ion flight paths, some energy compensation can be achieved. Ions with greater initial energies will penetrate these fields a little further, thus taking a longer flight path and therefore arriving at the detector at the same time as sister ions starting out with the same mass but lower energies.

If you are scanning the primary ion beam across the sample, you can get an image of secondary ion intensities for any secondary ion or collection of secondary ions. Instruments have long had the capability of saving raw data, that is the X and Y position of the primary ion beam for every secondary ion that is detected. You can thus make images retrospectively from the raw data. Further, from any collection of pixels in an image, you can plot a complete spectrum.

Given properly designed extraction optics, the TOF-SIMS instrument is a remarkably sensitive tool. The transmission of ions from sample to detector is quite efficient, with as much as half the ions formed at the surface reaching the detector. This is in part because TOF-SIMS has an essentially parallel detection system, with all ions being measured for each pulse. There is no spectrometer that needs scanning.

Given the complexity of the sputter event, the question is, why invest in this expensive and difficult surface analytical technique?

• Given a chemical species that readily produces secondary ions under primary ion bombardment, the TOF-SIMS is an extremely sensitive technique.
• The data the modern TOF-SIMS instrument produces is highly reproducible. Given an identical sample, analysis performed at different times, even years apart, will be nearly identical. Conversely, differences in the data obtained from two different samples invariably reflect real differences in the samples, however subtle those differences may be.
• TOF-SIMS results often have a chemical specificity that is difficult to match. Even identical compounds present at the sample in different configurations (for instance, patchy coverage vs. a diffuse layer) will often give discernably different results.
• TOF-SIMS analyses are very fast. A large amount of data is generated within minutes of the start of acquisition. The answer to an analytical question is often clear within seconds.
• TOF-SIMS produces a wealth of image information about the samples analyzed. With modern ion sources, the primary ion spot sizes can be significantly less than 100 nm. With stage scanning, large samples can be imaged as well.
• Given a sufficient set of standards, TOF-SIMS can be used for quantification of both inorganic and organic species.

So with all this going for it, what else do you need besides a TOF-SIMS in your lab? Why use any other surface analytical technique?

• The complexity of the sputter event has placed a proper theoretical understanding for the technique still in the future. There is no way, even approximately, to answer the question as to how much of any detected species is present on the sample surface without standards. Signals can vary in intensity by orders of magnitude from one surface to the next, and the relative sensitivities to different species can vary widely as well.
• While SIMS images are usually meaningful, a significant amount of the time larger intensities do not indicate higher concentrations. As an example of this, a Au contact pad on a wafer will light up in SIMS images for a variety of contaminants adsorbed to the surface, but this definitely does not mean there are more of them on the Au pad than there is in the surrounding region.
• Sensitivities to both inorganic species and organic species vary by many orders of magnitude. There are materials that produce almost no significant signals in TOF-SIMS analyses. Elemental Zn has terrible ion yields in the SIMS. In our lab, we have found widely varying signal intensities from oxidized S compounds. Some compounds the SIMS could not detect but the Auger found easily. There were others for which the sensitivity in TOF-SIMS could not be matched. Most methyl silicones produce strong signals in the TOF-SIMS, but a silicone devoid of short chain species and with high molecular weights, and therefore few end-groups, produces remarkably weak spectra. Polymers that are well cross-linked transferred into the instrument through ordinary lab air will produce far stronger signals for the adventitiously adsorbed organics from the environment than from the polymer itself. In biological systems, the TOF-SIMS can be frustrating to use with its strong sensitivities to some biological materials (specific metabolites, phospholipids) and the near absence of signals from others (proteins).
• The mass spectral databases available today to the SIMS analyst are remarkably small. Even if you get a distinct spectrum for an unknown, there is a strong possibility that you will not be able to identify the material (although the MS/MS capabilities now available increases your chances). As of this writing the prospects for a better library are few. There is no funding available for such a project.
• Organic materials with a small surface area pose a particular problem. Analysis with a beam having a sufficiently small spot size for the task invariably leads to damage before the analysis is complete. Today’s cluster beams that do not leave damage at the surface preventing analysis after the consumption of a few monolayers do not have sufficient resolution.

It is a fact that there are enough applications for TOF-SIMS that most labs should have, or at least have access to, one. It is also a fact that TOF-SIMS is far from the single surface analysis tool for all occasions.