We periodically find ourselves engaged in debates over whether local (in-situ) or extractive techniques are of greater value. We see utility in three sorts of observation: overall scans (often called direct measurements), localized observation, and extractive techniques. Trade-offs include speed, cost, disruption of product flow, and whether the analysis is for general process control or to investigate a specific product performance problem.

DIRECT, OVERALL SURFACE SCANS

Analytical chemists can become enmeshed in pondering small features on a surface; considering the overall appearance of the surface may seem mundane. However, overall surface analysis is a valuable monitoring technique.

Visual observation
Start with the obvious. If there is an apparent change in the surface, further action is called for. Actually, it is wise to be on the lookout for changes in the product or component, and also to be aware of changes in process fluids. Changes in the surface can include color, films, visible particles, and texture. Other changes may be localized, such as corrosion, pitting, and discolorationat intermetallic areas.

Black light
Black light provides increased discrimination of contaminants. Many particulate contaminants will fluoresce when they absorb UV or black light, emittingvisible light that makes the particle shine like a star on a clear night.

Water break/contact angle
A water break or contact angle test can frequently determine if an organic contamination film is present. However, residual surfactants can create a false negative condition, where the contaminant contributes to a favorablecontact angle.

Optically Stimulated Electron Emission (OSEE)
In OSEE, a metallic surface is irradiated with UV light that stimulates emission of electrons through the photoelectric effect. The resulting electron current is a function of surface contamination levels. OSEE can be implemented inan on-line production monitoring station.

LOCALIZED (MICROSCOPIC) SCANNING

Optical scanning
Microscopic scanning includes a large number of techniques. In general, they identify the presence and location of contaminants but not the type of contaminant. The ability to pin-point location or identify contaminant particle size dependson the type of scanning used. Optical scanners range from large wavelength infrared (IR) through visible and ultraviolet (UV) down to the extremely small wavelengths associated with X-rays and even smaller with electron bombardment. The wavelength of the light is not the only discriminator on how precise a location can be scanned; one must also consider how the probe beam is directed and focused. IR scanning (e.g., FTIR) provides contaminant identity information sincedifferent organic molecules absorb different wavelengths of IR light.

EXAMPLES OF MORE COMPLEX SURFACE SCANS
Atomic Force Microscopy (AFM) is not an optical microscopy technique but provides the most precise location scanning. AFM employs a microscopic stylus that can detect movements down to atomic dimensions as it is scanned across asurface, analogous to a phonograph needle on a vinyl record.

Auger Electron Spectroscopy (AES) is particularly valuable for probing oxide layers in a localized, very shallow region of the surface. It is rapid, relatively low in cost, and may provide an early warning sign of trouble. It is a process where electrons stimulate the emission of other electrons; the energy of the emitted electrons can be used to identify the emitting element (contaminant).

EXTRACTION
Extraction provides a means of determining overall contamination where scanning the surface directly is impractical due to product size or configuration. Many analytical tools can be employed directly on the component surface as well as on the extracted material. In extractive techniques, the component is exposed to water or some organic solvent or blend. The extract is concentrated and then analyzed by an appropriate technique. The most common, workhorse techniques are non-volatile residue and par-ticulates. For non-volatile residue (NVR), most often, the extract is taken to dryness and weighed. Alternatively, particles are first separated, typically by filtration and occasionally by sedimentation. In this case, NVR consists of thin-film contamination. Ofcourse, thin film contamination could also include small particles that are nottrapped by the filter or that remain suspended.

Contamination on complex components
While it is possible to characterize the extracted particles by size, shape, and even by composition, after extraction it is usually impossible to determine where on the part the particle or other contaminant came from originally.

Extractive techniques are particularly valuable for particulate analysis because, even with careful observation or automated techniques, it is very difficult to count particles directly on a surface. With ornate components that have blind holes and surface texture, it is not feasible to count, quantify, or identify the particles or thin film contaminant directly.

EXTRACTION TECHNIQUES
While most people would distinguish extraction from cleaning, extraction is,in effect, a cleaning process. Variables include:

Extraction medium
The same rules of solvency that apply to cleaning apply to extraction. For polar soils, water or isopropyl alcohol are more appropriate. For non-polar soils, cyclohexane is more appropriate. For mixed soils, a cyclohexane/IPA azeotrope may act as a broad spectrum solvent. The ability of a given extraction medium to penetrate tightly-spaced components is also a consideration. As with cleaning agents, any extraction medium must be low-residue, particularlywhere low levels of contaminants are being evaluated.

Temperature
A higher temperature results in more complete extraction.

Time
The longer the time, the more complete the extraction.

Force
Forces such as agitation and ultrasonics enhance the extraction by moving thecontaminant away from the surface.

Substrate/materials of construction
We generally want to do a thorough extraction so that as much surface residue as possible is obtained. As with cleaning, extraction has to be balanced with potential substrate damage. This is true even if the part being analyzed is to be sacrificed. With excessively harsh extraction conditions, you are extracting the part from the part. This also means that particles may begenerated.

Extracting particles
How do you know you have removed all of the particles? Extracting particles is a particularly thorny issue in that particles are held to the surface by physical forces; any thin film contamination will also influence particle adherence. This means that, for example, three extraction cycles with a given fluid is unlikely to remove as many particles as will three extraction cycleswith three fluids each of different polarities.

In our opinion, therefore, there are no absolutes in extraction techniques, only guidelines. The most productive approach is to develop a logical extraction technique, relative to the soils, materials of construction and actual experimentalresults. Then record the extraction protocol and stick with it.

FORESTS OR TREES
Extractive or direct surface analysis? It depends on the problem at hand or on what you need to monitor. We think it is useful to periodically step back and critically assess even routine protocols to make sure the correct questionsare being asked.

Barbara Kanegsberg and Ed Kanegsberg are independent consultants in critical and precision cleaning,surface preparation,and contamination control.They are the editors of Handbook for Critical Cleaning,CRC Press.Contact them at BFK Solutions LLC., 310-459-3614;info@bfksolutions.com;www.bfksolutions.com.