The coatings industry as a whole is learning regarding non-visible surface contaminants, primarily chlorides and sulfates. Often people in the industry either want to test to determine contamination levels, or to test results of de-contamination efforts or at times simply desire to contaminate panels for further testing of one kind or another.
The two most often asked questions are how to contaminate the panels and after contamination, how are the test results determined. These are actually two distinctly different procedures, each will be addressed individually here. The first involves contaminating panels, the second involves measuring contamination.
Many different methods of contamination have been tried and the results of other methods have been reviewed. The actual results vary widely. Keeping in mind what we are trying to accomplish is a replication of field conditions, we have to remember that what we encounter in the field is the result of long term contamination, not something that is recently applied or contaminated. In the field, the metal surfaces encountered often are pitted, either deeply or slightly, but either way the contaminants are found to be concentrated at the pit bottoms. While extremely high concentrations can be found at the pit bottoms, often a mere fraction of an inch away they may be so small in quantity as to be almost non-detectable. This is something extremely hard to duplicate in laboratory conditions. As example, while we may duplicate the content of seawater, it is very difficult to duplicate the long term effects of exposure. Corrosion engineers doing coatings testing are aware of this and consequently test sites have been developed to expose painted panels for extended time spans. Often this same knowledge is not used in determining methods used to contaminate uncoated panels for testing. Too often panels are doped with saline solution, sulfuric acid, hydrochloric acid or another chemistry to instigate contamination levels acceptable for testing. This often results in a thin layer of contaminants placed somewhat evenly across the surface, almost what could be called a thin film, that has not had the time experience to form a tightly adhered electro-chemical attachment to the substrate such as is commonly found in the field. Hence, these contaminants are usually relatively easily flushed from the substrate with a simple water wash. Though this may be an inexpensive quick method of contamination, if we remember the reason we are doing it, we realize these methods simply do not replicate field conditions to any great extent. Because of the ease with which surface film type contaminants are removed, it is hard to achieve meaningful test results. Since no industry standard methods have been established, many laboratories use this type procedure and then wonder at the results.
Our own testing within our firm has found the method which most closely reproduces field conditions within a reasonable time is to place the metal coupons in a salt fog cabinet for a matter of days, rotating position and placement daily to achieve as even a contamination as possible. There are still several problems even with this method. The main one is that it appears extremely difficult to attain contamination levels commonly seen in the field. After ~8 to 10 days exposure the contaminant levels seems to level off and not increase. If they do it is by almost imperceptible degrees, if at all. On two occasions we placed 30 standard bead blasted panels in a salt fog cabinet, rotating them as described below. After 10 days, 15 of them were removed and the remaining 15 remained for an additional 10 days for a total of 20 days exposure. When they had all been abrasive blasted to a Near White Metal standard and then the contaminants extracted and measured, we found no significant difference between the two samplings. Far more work needs to be done in this area and many questions need to be answered, but we feel the results obtained are closer to field conditions and far more reliable than dosing methods. Below is the outlined methods used by our firm in developing contaminated panels for testing. The number of panels may vary for different testing samplings, but the procedure remain the same.
Use thirty (30) metal panels (3 inch X 6 inch is common, but other sizes may be used) abrasive blasted or bead blasted with clean abrasive, to a SSPC SP-5 "White Metal Blast". (Note: Usually factory produced panels come with a film of inhibitor on them to stop rusting during storage. This inhibitor must be removed by a solvent wash or wipe prior to blasting or use. Contact the panel supplier for the best method of removal.) Expose these panels to a salt fog for 144 hours, per ASTM B 117, using a solution of 5% sodium chloride and a 2% sodium sulfate solution. Periodically reverse, rotate and invert panels during contamination for uniform exposure. Pitting of panels will probably occur.
Note: The sodium sulfate solution may be deleted from the contamination of these panels if only the chloride ion is required.
After exposure, allow panels to dry. Abrasive blast these contaminated panels (both sides) to a "Near White Metal" SSPC SP-10 or whatever standard is most pertinent to your project. Once the panels are blasted, latex gloves should be worn to minimize contamination from your hands. During the blast operation the panels should be mounted or supported on a rack and not blasted on the ground or a floor surface as this will lead to cross contamination also.
You now have 30 panels contaminated close to actual field conditions. From this point you can use them for several purposes, such as testing various paint films for durability or for testing the effectiveness of de-contamination methods or products. If using liquids for de-contamination, whether it be multiple water washes, various water qualities or liquid soluble salt remover products I would suggest the following:
The thirty panels are divided into three sets of ten each and treated as outlined below.
Set aside the first 10 panels for the control. These will be boiled later to determine contamination levels.
Wash the second 10 panels (both sides) with 3000 psi pressure washer with potable water to simulate field conditions. Allow to air dry or blow dry with a hot air gun.
Wash 10 panels (both sides) with 3000 psi pressure washer with a solution of liquid soluble salt remover and the same type water as used previously. Allow to air dry or blow dry with a hot air gun. (Or use another liquid type procedure, such as de-ionized water wash)
Now you have a control, which will give you the original level of contamination, a water wash set of samples showing the degree of cleanliness attained with a water wash and a set of samples utilizing the cleaning method to be tested. To attain the actual results you must first extract the contaminants from the metal samples and then measure the level of contamination in the extraction fluid. This is outlined below:
Extraction From Contaminated Panel
The boiling extraction method described below typically will extract 90 to 100% of the contaminants from the metal surface. This is considered total extraction. Caution should be used to thoroughly clean and decontaminate all equipment prior to use and between uses by a minimum triple rinse in de-ionized water. Latex gloves should be worn to keep from getting salts from a persons hands on the test panels or the equipment.
Quantitative analysis is completed by using an ion specific testing device, such as an ion specific probe or a titrator, and the extract solution. Convert readings to ug/cm squared, determined by the surface area of the test coupons. Once the measuring device provides a reading, usually in parts per million (P.P.M.), you multiply that reading by the volume of liquid (milliliters), divided by the surface area of the test panel, in centimeters square.
Example: P.P.M. X milliliters divided by surface area in centimeters2 = micrograms/cm2
Frequently the question is asked "Why not simply use conductivity to measure the salts?" Since so many have believed for so long that conductivity accurately measures soluble salts I would like to show test results for solutions in which conductivity has been measured, then the solutions measured by ion specific methods for chlorides. Twenty steel panels were contaminated in a salt fog cabinet. Each day the panels were rotated to a different position to keep contamination as even as possible. After 10 days in the salt fog cabinet the panels were removed and abrasive blasted to a near white metal standard. The panels were then placed in individual Pyrex dishes, filled with D. I. water to a level of 350 ml and boiled for one hour. D.I. water was added as needed to maintain the volume. After the boiling was completed, the volume was brought up to 500 ml on all sample solutions. One sample was damaged by breakage and was discarded. The solutions were then tested both by ion specific electrode and by conductivity. The results are shown in Table 1.
Column 1 shows conductivity, column 2 shows specific chloride ions as tested with an ion specific electrode.
Chloride in micrograms/cm2
|Panel 12||Damaged||No Analysis|
As the reader can see, it is quite easy for overall conductivity to be increased, while chlorides, measured by ion specific methods, are actually reduced. Take for example panel 2 versus panel 15. The overall conductivity of the two is very close, yet the contamination levels of chlorides differs greatly. This testing was not performed in our laboratory, but was done by an independent third party laboratory which is nationally known. Our firm has been involved in testing performed by three laboratories, all nationally known reputable facilities, and each has arrived at typically the same results.
The coatings industry has, with old knowledge, always considered anything conductive to be salts, even such things as metal ions are considered salts. Using conductivity does measure everything conductive, but gives absolutely no indication what portions are the ions we are most interested in when dealing with corrosion or protective coatings. Today we have test devices, such as titration strips and tubes, reagent chemicals and others readily available which offer ion specific measurements. We no longer must live with "close guesses".