Outgassing: Advanced Detection and Control Techniques
Today`s cleanroom processing methods demand more sophisticated and complex monitoring systems to detect and eliminate outgassing and contaminants that limit process, yields and profits.
By R.J. Simonson, S.M. Thornberg, and A.Y. Liang
Materials in cleanrooms continually outgas (release) molecular contamination (MC) into the cleanroom air and also absorb chemicals from the air, which may be released later. In contrast to particulates, molecular contaminants typically exist as airborne, single molecules which can condense or agglomerate on surfaces, including silicon wafers, within the cleanroom and other wafer environments. Recent studies have revealed detrimental impacts of MC on processing, reducing yields and profitability. An important step toward reducing MC levels is identifying outgassing sources and the chemicals outgassed.
Outgassing from surfaces within the wafer environments and fugitive vapors that have escaped from processing chemicals are the two major sources of MC in wafer-processing facilities. Most of the commonly known molecular contaminants are organic in nature. Polymeric construction materials for mini-environments, standard mechanical interface (SMIF) pods and cleanrooms often outgas mold-release compounds, filler additives, plasticizers, antioxidants, and ultraviolet (UV) blockers. In addition, MC outgassed from cleanroom building materials, such as paint, structural plastics, sealants, floors, ceilings, and walls is often organic in nature. Other organic MC occurring in the vapor phase are photoresist, wafer prime, and organic solvents such as acetone and isopropyl alcohol. Some molecular contaminants are inorganic in nature. Examples include fugitive vapors that have escaped from sources the acids and bases used to process wafers, such as HCl, HF, H2SO4, and NH4OH, and contaminants in the air from sources outside the cleanroom, which can include traces of H2S, NO2, and SO2 gases are examples.
In all vacuum and non-vacuum processing equipment, residual gases may originate in the process chamber and even from the contaminated silicon substrates. However, leaks and outgassing are the main sources of MC inside process chambers. Moreover, vacuum system construction materials are often sources of residual gases. Fluorocarbon elastomer seals, o-rings, and plastic and ceramic composite materials can contain organic and inorganic fillers, plasticizers, and mold-release compounds that are often unstable under vacuum, heat, or UV radiation.
Current Study of MC in Cleanrooms
The three main areas of current MC investigation are the chemical quality of the cleanroom air, the outgassing and cross-contamination potential of the wafer pods, and the assessment of condensed molecular contaminants on wafers.
Recently, Muller et al.4 used charcoal collection badges to sample organic contamination levels in cleanroom air. The badges were used as passive vapor collectors, and gas chromatograph/mass spectrometer (GC/MS) analysis was used to determine airborne concentrations and arrival rates, as well as to identify collected vapor species. Organic compounds identified include various aliphatic and aromatic hydrocarbons, trichloroethylene, xylene, toluene, butyl acetate, and cello solve acetate, among others. Muller et al. concluded that the surface arrival rate of these gas species is, on a mass basis, several orders of magnitude higher than that of particles. Suzuki et al.8 indicated the abundance of MC in cleanroom air with contact angle measurements. They demonstrated that the contact angle of water droplets on oxidized Si wafers stored for 150 hours in a SMIF increased less than when exposed in cleanroom air. The increase of water-contact angles on oxidized silicon normally indicates an increase in organic condensation on the surface. Kinkead et al.9 conducted corrosive-vapor-monitoring experiments using silver coupons exposed to cleanroom air for 30 days. The experiments revealed severe corrosive conditions caused by fugitive vapors of sulfuric, hydrochloric, and possibly hydrofluoric acids found in some areas of the cleanroom.
Fergason10 reported the detection and the subsequent identification of MC on wafers, using thermal desorption/gas chromatography (TD/GC) with flame ionization detectors (FID) and flame photometric detectors (FPD). Wafers were exposed in cleanroom air or near polyvinyl chloride (PVC) curtains, or stored within polyethylene or PVC storage boxes. All wafers showed organic contamination, and some organophosphates from PVC curtains. With a similar test setup, Mori et al.11 recently analyzed condensation of MC on Si wafers. They identified triehylphosphate (TEP) and other organics on wafers that had been exposed in cleanroom air for 24 hours. The TEP was found to have outgassed from the poly urethane sealant used to secure the filter media to the HEPA-filter modules.
In a recent IBM study, Miller et al.6 used thermal desorption/plasma chromatograph mass spectrometer (TD/PCMS) analysis to study the contamination of wafers stored in pods. Miller et al. observed a trace amount of BHT (2,6-di-ter-butyl-p-cresol), a polymer antioxidant, in used, cleaned, and even new pods. In addition, NMP (n-methyl pyrrolidone), a photoresist solvent used in processing, was found in all used pods. They also performed contact angle measurements on wafers stored in pods. Their results indicated that wafers stored in both laboratory air and in pods showed an increase of contact angle compared to control wafers stored in a sealed and flame-cleaned glass container. In a similar setup, using PCMS analysis, Silverman12 reported the detection of trace amounts of TEP, NMP, BHT, diethyl phthalate (DEP), dibutyl phthalate (DBP), and dioctyl phthalate (DOP) on used pods. The latter three phthalates are plasticizers that were contained in the polyurethane gaskets in the pods.
In an earlier study, Budde and Holzapfel5 used an ion mobility spectrometer (IMS) to detect and characterize the outgassed species from wafers stored in polypropylene boxes. They detected plasticizers and antioxidants condensed on wafers. Later, using a thermal de sorption/gas chromatograph mass spectrometer (TD/GCMS) setup, the same authors detected polypropylene oligomers, isopropanol, and toluene outgassed from some of the wafer storage boxes. Additional analysis of the constituents thermally desorbed from the box materials revealed oxidation and decomposition products of polypropylene, BHT, and an anisol-type oxidation stabilizer.
Goodman et al.13 reported that wafer storage boxes and processing system parts made of fluoropolymers can outgas corrosive fluorine compounds that may etch and damage stored silicon wafers. Submonolayer deposition of fluorine compounds has also been observed after storage times of less than one hour. Dusan Grman et al.14 recently observed considerable fluorine contamination on bonding pads of finished devices. Such contamination could adversely affect wire bond properties of packaged devices and result in device failures. They also demonstrated that when silicon wafers were stored in closed polypropylene boxes for six weeks, degassing in the closed boxes led to a similar fluorine contamination on the bonding pads.
Parameters that Affect Molecular Outgassing
Rates of outgassing are affected by several parameters, including temperature, vapor pressure, surface area, and vapor pressure gradients. Temperature can increase the diffusion coefficients of compounds through materials (e.g., plasticizers in polymers) or can cause compounds to desorb from a surface. The vapor pressure of a compound will influence both initial outgassing and subsequent reabsorption. For example, a high-vapor-pressure compound will be volatilized or outgassed at a low temperature but will not condense easily. Conversely, a low-vapor-pressure compound may be released slowly but can readily stick and accumulate on surfaces.
Methods for Determining Outgassing
Several methods are used to characterize outgassed organics from materials and surfaces. A sample can be introduced directly into the instrument or indirectly by being adsorbed to another material. Proper sample preparation and sampling methods can be the major challenge for measuring outgassing. Ideally, the sampling method will not alter the composition or introduce contamination into the sample. Extremely low analyte concentrations may require preconcentration methods to provide enough material for the analysis. In addition, instrument blanks (measuring the instrument background), sample blanks (measuring sampling technique cleanliness), and high-quality calibration standards are required for obtaining these trace measurements.
Sample preparation must be undertaken with great care, since a minor contamination from the preparation process can appear as an “outgassed” peak. Some methods require a vacuum for analysis (e.g., X-ray photoelectron spectoscopy–XPS; and Auger electron spectroscopy–AES), while other methods require that the sample be collected at ambient pressure (thermal desorption). In addition, temperature is often used to accelerate the outgassing, but excessive temperatures can cause thermal degradation of the sample or yeild unrealistic results. For example, if a polymer is outgassed at a temperature above its glass transition temperature, the amount of outgassing will probably be unrealistic. The surface area of the sample is another factor influencing the amount of outgassing. A sample with a large surface area will outgas more than one with a small surface area. Consequently, a ground polymer sample will appear to be “dirtier” than an unground piece of the same material. It is necessary, therefore, to prepare all samples using identical standards when the data sets are to be compared.
Collection of outgassed molecules typically involves sampling (in a flowing gas stream) directly into an instrument, or collecting and concentrating the molecules for a period of time. Direct introduction of a sample material into an instrument requires that the materials have a high enough outgassing rate to be detected. If outgassing rates are below detection limits, the outgassed molecules are commonly concentrated cryogenically or adsorbed onto some sorbent medium. Cryogenic collection involves passing a carrier gas containing the outgassed molecules through a collection region cooled to temperatures below the condensation or freezing temperatures of the molecules. The collector can be as simple as coiled tubing in a liquid nitrogen dewar or as complex as a closed-cycle helium cryostat. For adsorbent collection, the carrier gas is passed through a sorbent bed (typically a form of activated carbon, although many types of adsorbents are commercially available), where the analyte molecules “stick” to the bed material and are released into the instrument when the material is heated. This method is typically used to measure low levels of organics in air (e.g., environmental monitoring in cleanrooms and pods).
Techniques used to measure the outgassed compounds range from very simple totalizing meters to very complex analytical instruments. For example, a total hydrocarbon detection meter can provide relative levels of concentrations of all hydrocarbons summed together with no compound identification.
These instruments are typically small and inexpensive and are used for screening. More complex instruments provide a great deal of information and are much more expensive to buy, maintain, and run. Instruments in this category include the gas chromatograph (GC), GC/MS, ATR/FTIR, and the IMS. Since outgassing is usually at very low levels (resulting in analyte concentrations in the ambient of ppm or less), this discussion will concentrate on the more sensitive techniques.
Gas chromatography is widely used to separate outgassed compounds that are then analyzed by a variety of detectors. In gas chromatography, the sample is introduced into the chromatographic column where a carrier gas tends to push the sample along. The sample molecules interact with the column-coating materials and will adsorb and desorb at rates corresponding to the affinity of the molecule to the coating. Materials with low affinities will elute quickly, while those with high affinities will tend to “stick” and slow down.
Detectors are usually either general purpose (for looking at a wide variety of compounds) or selective (for measuring a specific element or class of compound in the presence of other compounds). Detection limits are usually ppm to ppb, depending on the detector and on the sample being analyzed. General-purpose detectors include the thermal conductivity detector (TCD) for general chromatography, the FID for hydrocarbon detection, and the mass spectrometer. The latter is very popular because it provides “fingerprint” spectra for sample compounds eluting from the gas chromatographic column.
Typically, these spectra are compared to reference library spectra for compound identification. Detectors that are sensitive to specific elements or classes of compounds can be very helpful in locating the source of specific contaminants. For example, a nitrogen/phosphorus detector (NPD) will selectively detect only nitrogen or phosphorus compounds, even though the sample may have many other hydrocarbons. Consequently, the chromatograms are greatly simplified, allowing the operator to focus on the compounds of interest.
Mass spectroscopy is sometimes used as a stand-alone analytical technique for outgassing measurements. The most applicable spectrometers for organic measurements are the quadrupole, ion trap, and ion cyclotron resonance spectrometers. These spectrometers have a large mass range capability and an adequate sensitivity (detection limits typically ppm to high ppb) for most measurements. The ion mobility spectrometer (IMS) is used to measure trace amounts of organics and is a very sensitive instrument (with parts per trillion detection limits). To provide identification of unknowns during IMS, a secondary mass spectrometer, such as a quadrupole or ion trap, can be used to get a “fingerprint” spectrum, provided the analyte concentration is great enough to be detected by the secondary mass spectrometer.
Outgassing in Cleanrooms
The measurement of outgassing in cleanrooms is of great importance and has been reviewed in a recent article.4 The sampling method is the primary difference between measuring outgassing from materials and in a cleanroom. Samples taken for measuring outgassing in a cleanroom are normally collected on sorbent media for subsequent analysis. The sorbent media can be in a badge worn by a worker, in a tube with an active air flow through it, or any other convenient form. The sample is then removed from the sorbent by solvent extraction, purge and trap, or thermal desorption. At this point, the sample can be introduced into the instruments mentioned previously and the analysis performed.
Other instruments are capable of performing active sampling and analyses of room air. The simple total hydrocarbon monitor sensor mentioned previously can be used for nonselective detection of organic contamination. A detector for a gas chromatograph can be configured to measure hydrocarbons in air without using the normal GC column. The air sample is drawn directly through the detector without separation so a total hydrocarbon measurement can be made. A limitation of this method is that, depending on the sensor, a changing air background can result in erroneous conclusions. A mass spectrometer can be modified to sample room air continuously through a flow restrictor functioning as an inlet to the ionization region. Alternate methods for ionization can be used to reduce background signals. Chemical ionization has been shown to effectively monitor compounds in lithography processing and eliminate the bulk of the air background signals.23 Atmospheric pressure ionization mass spectroscopy (APIMS) is a method in which ions form in a high-pressure region and accelerate into a lower-pressure region, where they are detected by standard mass spectrometry.24 This allows for direct sampling of atmospheric pressure samples. Finally, the gas chromatograph can be equipped with a sample loop capable of trapping room air and introducing this aliquot to the GC column. The latter method is of rather limited use because it is not sensitive enough for most trace compounds.
Surface Contamination Measurement Techniques
A convenient distinction can be made between “outgassing” and “contamination” in the cleanroom environment. Airborne molecular or aerosol species that adsorb or condense on production substrates, where they may interfere with subsequent process steps, can be classed as “contamination.” Production substrates may include optical, biotechnical, and other products, as well as semiconductor devices. The sampling and analysis methods outlined above are useful for determining the arrival rates of potential contaminants at the production substrate. Such knowledge should be combined with direct measurements of actual adsorbed contamination. The resulting data can be combined with statistical process monitoring methods to determine the severity of contamination effects on product yields and to help pinpoint the sources of individual problems.
The choice of the “best” method for monitoring surface contamination depends on a number of factors, with surface detection sensitivity being the foremost. An example is the contamination of Si wafers by organic vapors during semiconductor device processing. The Semiconductor Industry Association has established acceptable upper limits for organic contamination of wafer surfaces as 1013 C atoms/cm2 in 1995, and 1012 C atoms/cm2 in 1998.25 These coverage limits correspond to approximately 0.01 – 0.001 atomic monolayers. Similar levels of coverage have also been found to affect product performance in other industries, e.g., optics.21 Several categories of instrumental techniques are available with the required detection sensitivity. The choice of methods is then further narrowed by other factors: compatibility with device and process requirements; molecular contaminant species identification capability vs. contaminant atom detection sensitivity; and calibration requirements if quantitative coverage measurements are desired.
It has been suggested that the applicable (i.e., sufficiently sensitive) surface analytical methods can be classified according to whether or not they require vacuum conditions.5 While this distinction is relevant to the issue of process compatibility, from the analysis perspective, it is more valuable to consider whether a method requires removal of the analyte from the process substrate, or if the signal can be acquired from the adsorbed contaminant. The reason is that the desorption or removal of the analyte from the surface may not be quantitative for all species present, resulting in fractionation of contaminant mixtures and misleading analysis results. Therefore, readily available methods may be divided into one group that requires desorption of analytes and another group that does not. Mass spectroscopic methods obviously fall into the former category, while photon (FTIR, Raman, Ellipsometry) and charged-particle spectroscopies (AES, XPS, ion scattering) belong in the latter. Crystal microbalance or surface acoustic wave (SAW) sensors26 used to detect mass accumulation in micro-environments will not be discussed here, except to mention that molecular species identification and quantitative calibration of such sensors must be established using analyses by other techniques.
Considering the above caveat regarding fractionation during analyses, mass spectroscopy-based methods are among the most sensitive techniques available for analyzing surface contaminants and offer the best “stand-alone” molecular species identification capability. The list of techniques based on mass spectroscopy of products removed from the substrate boasts an extensive array of acronyms. These are derived from the method used to de sorb the analyte, the method of ionization of the desorption product(s), the type of mass analyzer used, whether or not methods of preconcentration (ion traps) or preseparation (ion mobility or plasma chromatography) of ions are employed prior to mass analysis, or any combination of the above. Any of these factors, singly or in combination, can affect the detection sensitivity and species identification capability of the overall technique. The resultant body of literature is enormous.27 Once desorption and ionization of the analytes are completed, the identification of molecular contaminant species can be performed with standard mass spectroscopic interpretation.
The simplest experimental method for removing adsorbates from a substrate surface for mass analysis is thermal desorption.28-30 One of the advantages of thermal desorption for sampling contaminants is that vacuum conditions are not required if the desorbing species are entrained in a suitable carrier gas stream. In such cases, ionization is often performed by collisional and/or chemical reaction processes.5 Preseparating ions prior to mass analysis can be accomplished in such cases by exploiting differences in “ion mobility” or “drift velocity” of the ions in the carrier gas under the influence of an applied field. This is the basis of ion mobility spectroscopy (IMS) or plasma chromatography; 5,31-36 detection sensitivity is excellent because organic surface contaminants typically are detectable at picogram quantities or below.35 Quantitative determination of contaminants from “first principles” requires accurate knowledge of desorption yields, ionization probabilities, and mass filter transmission. Such information is typically unavailable in practice, and calibrations must rely on comparison with analyses of secondary standards.
Another method for performing mass spectral analyses relies on energetic particle bombardment of the surface to effect removal or “sputtering” of analyte into the gas phase. Incident particles are most often ions, as in secondary ion mass spectroscopy (SIMS) 27, 37-42 or sputtered neutral mass spectroscopy (SNMS) 37, 38, 43-56 (See Figure 1, page 30), although incident neutral beams have been employed in the “fast atom bombardment” (FAB)27, 37 technique. Sensitivity and spectral interpretation are similar to the thermal desorption methods discussed above. The use of incident particle beams for sputtering the surface dictates the need for a vacuum. The most prevalent techniques, SIMS and SNMS, differ primarily in the method of ionization of the analyte. In SIMS, secondary (analyte) ions are produced by concerted or near-concerted mechanism(s) during the sputtering process.27, 38 The mechanisms involved are complex and poorly understood, and ionization cross sections depend strongly on the local chemical environment of atoms or molecules on the surface, resulting in the so-called SIMS “matrix effect.” 27, 37, 38, 57 The severity of this effect can be greatly reduced with SNMS rather than SIMS, because the sputtering and ionization processes in SNMS are separated. An additional advantage of SNMS is that the yield of neutrals sputtered from a surface greatly exceeds that of secondary ions.37,43 Several methods of ionizing sputtered neutrals have been applied, such as electron bombardment,27, 58-60 plasma ionization,61 and resonant47-53 or nonresonant43, 54-56 photoionization. Detection limits on the order of 10-8 monolayers (107 atoms/cm2) have been reported.43, 54, 55 The identification of adsorbed molecular species will still depend on the probability of the molecules remaining intact or fragmenting identifiably in the desorption (sputtering) and ionization events.43,61,62
Some nondesorption-based surface analytical techniques, that have been successfully used to monitor condensates on process substrates are contact angle measurements, optical techniques (total reflection X-ray fluorescence–TRXF, FTIR, Raman, ellipsometry) and electron spectroscopies (AES, XPS). The measurement of the contact angle of a drop of water or other solvent on a substrate can be used to sensitively determine the degree of contamination of the surface,63, 64 and standard test methods have been developed.65 This method is simple, fast, and inexpensive, but does not provide contaminant species identification, and it can be affected by other factors such as surface topography.
Optical methods of probing surface contaminants have the advantage that vacuum environments are not usually required. For the detection of contaminants containing heavier atomic species (Z>15), TRXF can provide sensitivity on the order of 1011 atoms/cm2 or less.63, 66 In some cases, optical techniques such as differential reflectance,37, 67-69 second harmonic generation,70,71 and ellipsometry 37, 63, 72 can detect changes in surface adsorption on the order of monolayers. However, these methods detect changes in optical properties of the interface rather than directly probing the adsorbate. Thus, the signals received are usually difficult to clearly interpret as resulting from adsorbate coverages or molecular identification. Vibrational spectroscopies can provide useful information about specific functional groups of adsorbed contaminant molecules. Infrared absorption spectroscopy can be used in either single reflection73, 74 or multiple reflection (attenuated total reflection, ATR) modes.25, 37, 75-77 Raman spectroscopy has also been successfully used to detect surface adsorbates.34, 78-81 Sensitivity, however, is usually not as good as the electron spectroscopies discussed below. However, in favorable cases, concentrations as low as 0.001 monolayers have been detected with infrared (IR) by exploiting polarization-dependent selection rules for adsorbates at interfaces.82
The two electron spectroscopies that are used the most for surface analyses are AES and XPS. XPS is often referred to as electron spectroscopy for chemical analysis (ESCA). Extensive reviews are available on these techniques.37, 38, 57, 83 Both techniques can detect submonolayer coverages of surface contaminants. AES has the added advantage that high lateral resolution (on the order of 100) can be achieved, depending on the electron beam excitation source used. On the other hand, the focused electron beam excitation typically used in Auger analysis can damage some samples, particularly organic materials. In addition, elemental analysis by AES is relatively straightforward, but chemical state information (in this case, contaminant molecular identification) is usually difficult to extract. XPS is less prone than AES to beam-induced damage and difficulties caused by electrostatic charging if insulating substrates are to be analyzed. Chemical information is typically easier to interpret for XPS than for AES. However, it is still difficult, if not impossible,to clearly distinguish between different organic compounds having similar structures in the absence of supporting data from other techniques.
Control of Outgassing/Contamination in Cleanrooms
Ideally, detecting outgassing and contamination can be combined with product/process yield and reliability monitoring, often referred to as “impact studies,” to identify specific contaminants (and their sources) that affect the profitability of a process. Once such identification is available, however the issue of controlling or correcting the process contamination remains. The ideal solution is to completely eliminate the contamination source. Unfortunately, true elimination of outgassing in the cleanroom can be achieved only by stringently selecting materials and design criteria before the cleanroom is constructed. For instance, carefully choosing construction and component materials can greatly reduce outgassing, and such screening standards are available.84 Carefully choosing the cleanroom design includes minimizing surface areas that will be in contact with the airstream, both through mechanical design and by specifying a minimum standard of surface finish.4 Components with sufficient temperature tolerance can be vacuum baked prior to installation, but for the majority of construction materials, this is not practical. For existing cleanroom facilities it is, of course, too late to consider construction choices. Also, in the majority of new construction, the cost of such measures cannot be justified unless prior knowledge of the specific impact of an outgassing source on a particular process is available.9 Thus, current efforts to control outgassing and contamination usually involve collecting information on the specific impact of an outgassing source, then “locally correcting” the problem within the cleanroom environment. Locally correcting the problem can be accomplished by using mini- or micro- environments (pods and clustering) or by adding activated charcoal or chemically specific filters to remove volatile materials from limited areas of the process environment.4, 9, 84, 85
Cleanroom processing in a number of industries has progressed to the point that the traditional performance criteria, based on particle counts, are no longer sufficient to ensure adequate process control. Sensitive, reliable means to detect and measure outgassing of materials are therefore becoming crucial to improving processing technologies.
Outgassing measurements must be combined with actual surface contamination measurements, and these combined measurements must be used in conjunction with process defect monitoring to pinpoint contamination problems that limit process yields. This combined monitoring approach is necessary to control contamination so that the ratio of cost-to-benefit can be minimized (since completely eliminating outgassing, while retaining currently accepted cleanroom materials and components, is impractical).
A great potential market exists for the development of materials and components with minimized outgassing to be used throughout cleanroom design, construction, and operation.
This work was supported by the U.S. Department of Energy under Contract DE-AC04-94AL85000.
A complete list of the references cited in this article is available from CleanRooms Magazine, (603) 891-9230. n
R. Joseph Simonson received a Ph.D. in physical chemistry from the Massachusetts Institute of Technology in 1987. He has worked in surface research and analysis at the Dow Chemical Co., the Los Alamos National Laboratory, and the U.S. Air Force Phillips Laboratory. He joined Sandia in 1994 and currently supervises the X-ray photoelectron spectroscopy and secondary ion mass spectroscopy laboratories within Sandia`s Materials and Process Sciences Center.
Steven M. Thornberg received a Ph.D. in analytical chemistry from the University of New Mexico–Albuquerque in 1984. He is currently a senior member of the technical staff at Sandia`s Materials and Process Sciences Center, supervises the Center`s Gas Chromatography/Mass Spectroscopy Laboratory, and is co-task leader for organic contamination studies at Sandia`s Contamination Free Manufacturing Research Center.
Alan Y. Liang received a Ph.D. in electrical engineering from The Johns Hopkins University in 1977. Before joining Sandia in 1988, he worked extensively in the semiconductor industry, including device/yield engineering manager at GE Semiconductor and at Signetics. He is currently co-task leader for organic contamination studies at Sandia`s Contamination Free Manufacturing Research Center.
A prototype system for monitoring volatile organics in real-time from a lithography wafer processing tool. Chemical ionization mass spectra are taken and individual species can be identified and monitored by tracking their characteristic ions.
Figure 1. Secondary Ion Mass Spectroscopy (SIMS) vs. Sputtered Neutral Mass Spectroscopy (SNMS). In SIMS, ions of the sample material produced during the sputtering event are detected directly. In SNMS, neutral particles sputtered from the surface are ionized by electron impact (e-), chemical ionization (M*), or photoionization (hv). The yield of sputtered neutrals typically exceeds that of secondary ions by several orders of magnitude.