Monday, November 29, 2010

Finding The Optimal Analytical Test Part 1

Barbara Kanegsberg
Ed Kanegsberg
What analytical test should you use? A recent teachable and successful case study does not provide a miracle all-purpose analytical technique (there is no such thing) but rather illustrates the process of how a decision might be reached, and the considerations involved. It illustrates the importance of exploring beyond the methods that are in one’s immediate comfort zone, of understanding the features and limitations of any analytical method, and of adopting a scientific, rational, and defensible approach.
Testing involves “somehow perturbing the area to be analyzed and observing the results.”1 The key to successful analytical testing involves perturbing and observing the material to be analyzed in a meaningful, reproducible manner and in such a manner that meets the requirements at hand. This is easier said than done; and the process of selecting, developing, and validating a rational analytical method can involve interminable testing and discussion.
The case study involved collaboration between regulators at the South Coast Air Quality Management District (SCAQMD) in southern California and the Independent Lubricant and Manufacturers Association (ILMA). The approach to selecting the method involved determining and emulating likely manufacturing situations, testing and evaluation, assessment of method practicality, refinement of a standardized method, and consensus building.
In this instance, the goal was to determine evaporative loss of relatively volatile materials from metal working fluids so that a method could be included in an SCAQMD regulation with the goal of reducing VOC air emissions.
John M. Burke is Director of Engineering Services at Houghton International, Inc. in Valley Forge, PA. In 2008, as Chair of the Safety, Health, Environmental, and Regulatory Affairs (SHERA) Committee at ILMA, “I received a call from an ILMA member in southern California who described discussions with SCAQMD regarding all sorts of metalworking fluids, including vanishing oil, rust preventatives, and other protective fluids. As we learned more about the efforts to determine the VOC content, we knew we needed to be involved.”
The ILMA involvement developed into a major scientific effort, one that benefited the environment and considered requirements of industry. Naveen Berry, Planning and Rules Manager at SCAQMD, Diamond Bar, CA, described the association with ILMA as “a truly collaborative effort.”
Berry explains that “we considered EPA Method 24,3 it has many strengths. However, it is recognized as not being an optimal method to use for coatings but not for use with metalworking fluids and lubricants.” Burke adds that “Method 24 requires application of heat. While the temperature conditions are appropriate for aqueous solutions, there is a point when you heat a lubricant, which you go past the point of evaporating it, you are cooking it. The breakdown products are artifactual and do not truly emulate VOCs that would be released with normal, lower temperature, evaporation. In addition, Method 24 uses a laboratory oven. Placing a given sample at four different positions within the oven might yield four different results.”
Readers might want to note these limitations, because EPA Method 24 has been known to be used in manufacturing applications. In addition, loss and/or modification of a mixture of analytes during sample preparation are important considerations for everyone involved in manufacturing. Many complex mixtures can be readily modified during evaporation, especially with heating. For example, years ago, one of us (BK) participated in aerospace studies of Non-Volatile Residue (NVR) to qualify higher-boiling point solvents as replacements for lower-boiling point chlorinated and fluorinated solvents. We found that the evaporation temperature had to be lowered; and that the position of the sample in the laboratory oven had to be specified.
“In 2008 there was no standardized method for measuring VOCs in metalworking fluids,” continues Burke. “SCAQMD proposed a test method that they termed Method 313L.4 It called for determining VOCs by separation using gas chromatography (GC) with a flame ionization detector (FID).” Burke recounts that “after a year of testing with 313L, we could not get consistent test results among laboratories. In fact, we saw run to run variability even using the same GC system and using the same technician.”
Berry explains that SCAQMD decided not to pursue Method 313L, GC/FID for the current version of Rule 1144 because it did not meet the requirements of ASTM E-691 in terms of reproducibility. He adds that the project “raises the issue of how to treat semi-volatiles. When you test a mixture of multiple solvents in a gas chromatogram, generally you see distinct peaks. Some of the semi-volatiles did not form distinct peaks. They formed a smear.”5
Burke recounts that “we were in a quandary. How could we agree to regulatory limits? How could we find a consistent test method? What constitutes volatility? Then, one of our company scientists suggested using TGA. TGA use is called out in military standards to determine the volatility in hydraulic oil. The thought was that perhaps TGA could be adapted to the problem at hand.”
On the surface, TGA might seem like EPA Method 24; and Burke notes that if TGA was used at the temperature called out for EPA Method 24, the results were indeed similar. ILMA contended that 110°C +/- 5°C was too hot, and that modification of the fluid was occurring. Burke recounts that “we finally said that we really need to define what constitutes volatility.”
The SCAQMD/ILMA committee designed a test to characterize the evaporative properties of oils, using a time/temperature combination that might emulate realworld conditions. Burke notes that “it was important that we agreed to methods and to pass/fail criteria in advance. We tested three common naphthenic oils, with three different viscosities,* obtained from a single refinery. We selected a 26 week time span using the rationale that “two turns” of oil per year would replicate actual production conditions. The temperature was 40°C, emulating warm but realistic factory conditions. We agreed that we would observe the volatility curves (rate of change of fluid weight). If the curves flattened out before 26 weeks, we would stop earlier. We were surprised, not very pleasantly, by the results.”
The samples were still evaporating at 26 weeks. Burke explains that while the most viscous oil was beginning to asymptotically approach a leveling-off point, the other oils were still evaporating steadily. We thought that the medium viscosity oil would not be very evaporative; but it was almost 100% evaporative.”
“Since a 26 week testing protocol would be a bit cumbersome, the next step was to determine time and temperature conditions that emulated behavior of the oils used in the 26 week protocol. We tested five temperatures (71, 81, 91, 101, and 111°C) over shorter time periods. For each temperature, 120 minute by minute data points were collected.” Burke explains that they were able to set analytical conditions that matched results of the six month study to within 1%.
In summary, the group characterized the evaporative properties of oils used in lubricants, evaluated analytical approaches, and selected a promising method, TGA. In the next column, we continue with how the TGA method was tested and optimized, the current status, and a look to the future.
* The oils were characterized as 40 second, 60 second, and 100 second oils. The units refer to Universal Saybolt Seconds, a viscosity unit. It is the time for a given volume of oil to flow through a certain orifice. The longer the time, the more viscous the fluid.
Metalworking Fluid Restrictions May Affect You
We suggest that you, the reader, peruse Rule 1144.2 Metalworking fluids are being restricted to less evaporative materials in the SCAQMD area. SCAQMD rules sometimes are precedents for regulations in other areas and states. Features of the rule may impact your manufacturing facility either by current or future regulatory mandate or by corporate edict. If metalworking fluids that you or your suppliers use are reformulated or restricted, critical cleaning processes and approaches to determining surface residue on high-value product may also need to be revisited.
  1. B. Schiefelbein in Kanegsberg and Kanegsberg, “Find the Contaminant by Perturbing the Surface: XPS and Auger (Part 1),” Controlled Environments Magazine, November, 2007.
  2. SCAQMD Rule 1144, “Metalworking Fluids and Direct Contact Lubricants,” Amended July 9, 2010
  3. EPA Method 24, Method 24 – “Determination Of Volatile Matter Content, Water Content, Density, Volume Solids, and Weight Solids Of Surface Coatings”
  4. Appendix I SCAQMD Method 313: Determination of Volatile Organic Compounds (VOC) by Gas Chromatography/Mass Spectrometry (GC/MS)
  5. Clean Air Solvent Certification Protocol and

Barbara Kanegsberg and Ed Kanegsberg, Ph.D.“The Cleaning Lady” and “The Rocket Scientist,” are independent consultants in surface quality including critical/precision cleaning, contamination control, and validation. They are editors of The Handbook for Critical Cleaning, CRC Press; an expanded second edition is scheduled for publication in the 4th quarter of 2010. Contact BFK Solutions LLC, 310-459-3614;

No comments: