How clean is your surface?

Determining the cleanliness of a surface is an important task in many industrial situations.

In particular, any coating applied to a contaminated substrate surface will have defects such as poor adhesion, leading to subsequent failure. Monitoring to detect such contamination is only part of the solution. Precise identification of a contaminant not only indicates the best method for its removal, but more importantly allows its source to be traced and eliminated. Desktop instrument-ation is now available that can routinely produce these vital answers in minutes. When contamination is present as a surface film rather than as a particle, it may be only a monolayer in thickness and invisible to the naked eye. Some form of surface chemical analysis is therefore needed to detect it. Thin film contamination of this type is frequently organic rather than inorganic in nature. Many analysis techniques, which identify only the component elements, are therefore of limited use. However, extra information on molecular structure can be revealed by mass spectrometry, and this is vital in building a complete picture. It is possible to try to dissolve a contaminant and perform a bulk mass spectrometric analysis, but there is a more convenient and direct method of identifying surface contamination in situ. In addition, in situ imaging shows the distribution of the contaminant and may reveal further important clues about its origin. This analysis technique is known as Secondary Ion Mass Spectrometry (SIMS). In a SIMS analysis, a high-energy beam of ions is focused on to the sample surface [see Fig. 1], and the inter-atomic collisions eject elemental and molecular species from the extreme surface up to about 1nm depth. These 'secondary ions' are collected and analysed by mass spectrometry. Imaging is accomplished by scanning the incoming probe beam across the sample to build up an image, a process that can be completed in seconds rather than minutes. Longer irradiation of the sample starts to remove significant amounts of material, exposing each sub-surface layer in turn for analysis. Operating in this mode quickly shows whether a contaminant is a superficial layer or has penetrated the surface. This type of chemical analysis has traditionally required the use of large and complex instrumentation, which is not only very expensive but also difficult to operate. As an alternative, samples could be sent to a specialist contract analysis laboratory. Fortunately, recent years have seen the development of low-cost instrumentation such as the MiniSIMS developed by Millbrook Instruments [Fig. 2]. Operation is via a Windows interface, with the dedicated computer also monitoring instrument performance. This type of instrumentation is designed to allow high throughput analysis by less experienced operators. It is sufficiently compact to be treated as a mobile instrument, and can be operated remotely via a modem link. This is very useful where the need for analysis is shared within a company.

Cost-effective and practical This instrumentation harnesses the power of the SIMS technique in a form and at a price that allows its widespread use. It does not offer the same performance specification or flexibility in analysis as the more expensive conventional SIMS system, but reducing the cost of a SIMS analysis by up to 90% makes it cost-effective to generate SIMS data in a wide range of practical situations. This provides information that would be difficult or impossible to obtain by other analysis techniques. The MiniSIMS provides a fast method of detecting a wide range of common organic contaminants. The mass spectra not only identify the component elements, but the higher mass molecular ions also provide a characteristic ÔfingerprintÕ for each contaminant. For example, Fig. 3 shows the SIMS spectra acquired in less than 30 seconds from three contaminated metal surfaces. It also shows the negative SIMS spectra, and in each case there is a corresponding positive SIMS spectrum. For full identification, both spectra would normally be analysed in combination, but these negative spectra show the key points. The first spectrum was matched to a high-performance cutting fluid. The spectrum features a relatively low oxygen to carbon ratio, reflected in the peak ratio for CH- and O- at low mass. Even though the molecular weight of the contaminant is above the mass range of the MiniSIMS, characteristic fragment ions in the mass range m/z = 100 to 300 are seen, and these can be used as a specific 'fingerprint' to detect the presence of this fluid. The second spectrum is from a fluorinated lubricant additive, and now F- is the dominant elemental peak at low mass. The higher mass peaks (with the exception of the peak at m/z = 61) are carbon/fluorine cluster ions. The relative ratio of these CxFy- peaks alters between fluorocarbons, and this effect can be used to deduce further information about the molecular structure. The third spectrum is of a cleaning and degreasing agent. Once again the lower range of the spectrum identifies the elements that are present, in this case including sulphur from a sulphonated surfactant. Chlorine is notably absent from the spectrum. The higher mass range again provides a way to obtain a more specific match to a particular cleaning agent. As well as simply detecting these very different organic contaminants, the MiniSIMS can provide more detailed answers. The instrument provides a fast method of distinguishing between similar but different members of the same polymer family. In this second example, it will be seen that the mass spectra show clear and immediate differences as the side chains on a common polymer backbone are varied. Fig. 4 shows three SIMS spectra acquired from three aluminium surfaces contaminated with different silicones (siloxanes). The positive SIMS spectra are shown, and in each case there is a corresponding negative SIMS spectrum. Both spectra would normally be analysed in combination, but only the positive spectra are considered here. The first spectrum is of dimethyl siloxane (PDMS). This is the most widely used silicone, with two methyl groups on each silicon atom in the linear backbone. The elemental peak due to silicon is clearly visible (O- is seen in the negative spectrum). The peaks at m/z = 43 and m/z = 73 are characteristic of the end group of the molecule. The other characteristic peak at m/z = 147 represents the first two units in the polymer chain. The second spectrum is from methyl hydrogen siloxane, where each silicon atom in the chain has one methyl and one hydrogen atom. The tri-methyl end group (represented by the peak at m/z = 73) is still the same, but at low mass it can be seen that the molecule fragments differently. The peak at m/z = 147 is absent because the relevant structure no longer exists in the parent molecule. At higher mass there are more peaks than for PDMS, because some of the symmetry of the molecule has been lost. The third spectrum is of methyl glycol siloxane. Here the glycol side chain progressively fragments to give a set of characteristic peaks between m/z = 120 and 87. Again the tri-methyl end group is reflected in the peak at m/z = 73. These examples show that MiniSIMS can detect and identify common organic contaminants rapidly, and without the cost and complexity traditionally associated with SIMS analysis.