|Max T. Wills received his Ph.D. in organic chemistry from the University of Washington. He joined the faculty of California Polytechnic State University in 1967 where he helped establish a polymers and coatings program. He is a member of ASTM Committee D01 on Paint and Related Coatings, Materials and Applications.
Dane Jones received his Ph.D. in physical chemistry from Stanford University in 1974. Since 1976 he has been on the faculty at California Polytechnic State University, San Luis Obispo, Calif. He was instrumental in developing the Polymers and Coatings Program at Cal Poly and served as director of the program until 2002.
Direct Methods for Analyzing Volatile Organic Compounds in Coatings
In the United States the volatile organic compound content of coatings is currently measured using the U.S. Environmental Protection Agency’s Method 24, Determination of Volatile Matter Content, Water Content, Density, Volume Solids, and Weight Solids of Surface Coatings. This VOC content is determined by subtracting the water and exempt compound content from the total volatile content of the coating. Total volatile content is determined by heating a coating sample for one hour at 110 °C and measuring weight loss per ASTM D 2369, Test Method for Volatile Content of Coatings. This technique works well for coatings that do not contain water (or exempt solvents) but gives poor precision for low VOC content coatings, particularly waterborne coatings.
Newer methods for coating VOC analysis involve direct gas chromatographic determination. ASTM D 6886, Test Method for Speciation of the Volatile Organic Compounds (VOCs) in Low VOC Content Waterborne Air-Dry Coatings by Gas Chromatograpy, is such a direct method and was published in 2003. The method is currently being used by many companies to verify the composition of the volatile components of many different types of coatings and was recently used collaboratively by Cal Poly, San Luis Obispo and California’s South Coast Air Quality Management District to determine the VOC content of various architectural coatings.
The International Organization for Standardization (ISO) uses a VOC measurement version similar to Method 24 but only for coatings containing more than 15 weight percent VOC. For coatings with a VOC content between 0.1 and 15 percent, a direct gas chromatographic method similar to that in D 6886 is used.
In the ISO gas chromatographic method, a boiling point marker is used to define what constitutes a VOC. For waterborne coatings that marker is diethyl adipate, which has a boiling point of 250 °C, and compounds eluting after this marker on a specified capillary column are not considered to be VOCs.
Recent investigations sponsored by the Emulsion Polymers Council and the Adhesive and Sealant Council involve the development of an improved static headspace method for VOC measurement. This is largely the result of recent advancements and availability of precision headspace instrumentation. Static headspace methodology has been applied to a wide variety of coating types including architectural, original-equipment-manufacturer, 2K, ultraviolet-cure, and powder coatings. The method works especially well for the determination of nearly all volatile compounds, including hazardous air pollutants and exempt solvents. It is anticipated that a single universal method will be created for measuring VOCs, HAPs, and exempt solvents and will be applicable to virtually any coating now being manufactured.
The EPA Method 24 Problem
Specific regulatory VOC limits have been set for most coatings to ensure that emissions from these materials will decrease and air quality will improve. As regulations have lowered limits of allowed VOCs, a significant problem with the enforceability of these regulations has developed because reliable methods for the analysis of these VOCs are not available. The California Air Resources Board contracted with the University of California at Davis to carry out a project to develop a plan to obtain information to guide and prioritize the development and improvement of test methods related to volatile organic compound emissions from coating applications and related operations.
One of the conclusions of the survey1 was as follows:
The ARB Test Method Survey held a telephone conference (October 17, 1995) on test method problems. Participants in this conference included five air districts (Bay Area AQMD, Sacramento Metropolitan AQMD, San Diego County APCD, San Joaquin Valley Unified APCD, and the South Coast AQMD) and represent areas comprising more than 80% of the population of California, and encompassing all of the regulatory categories for VOC emissions from coatings and coatings operations identified by the ARB Test Method Survey. The participants in the telephone conference were virtually unanimous in their agreement on the number one problem with current test methods for VOC emissions from coatings and coatings operations: the inability of EPA Method 24 (and related ASTM and district methods) to provide accurate results for coatings containing low VOC and high water content. The current methods cannot be used with confidence for water-borne coatings containing VOC < 100 g/L. The problem is not primarily with the analytical techniques involved, but with the method of calculating the VOC concentration. The problems with EPA Method 24 and related methods are not amenable to improvements in the various analytical techniques involved [for low VOC, high water content coatings]. Therefore, it appears necessary to develop a new, direct method for determining the VOC content of coatings that can be used for low VOC, high water-content coatings. The use of a direct method would address a number of problems in addition to the problem with low VOC coatings, including the proliferation of exempt compounds, the need to measure hazardous air pollutants (HAPs), and the proposals to base ozone control strategies on the atmospheric reactivity of individual VOCs, rather than the total VOC content.2
ASTM Test Method D 6886
When test method D 6886 was first conceived, it was postulated that the majority of high-sales-volume waterborne architectural coatings (flat, eggshell and semi-gloss) would contain less than five percent by weight VOC and that the number of specific solvents would be both small and consist of relatively common materials, i.e., ethylene glycol, propylene glycol, butoxyethanol, butoxyethoxyethanol, and Texanol®. In a round robin involving eight laboratories and five commercial coatings (a flat, an eggshell, a semi-gloss, a gloss and a primer) this was indeed the case.
In carrying out the method, a sample of coating is dispersed in tetrahydrofuran containing an internal standard (p-cymene, cyclohexanol and p-fluorotoluene have been used). An aliquot of this dispersion is then chromatographed and the amount of each volatile component is determined from peak areas. The sum of the components represents the total VOC content of the coating. More recently, the solvent system has been changed from tetrahydrofuran to water containing diethoxyethane as the internal standard. The advantage of this change is that water is more innocuous than tetrahydrofuran and does not give a gas chromatograph peak in the flame ionization detection mode.
The improvement in precision using D 6886 instead of Method 24 is approximately tenfold and improves further as the VOC content approaches zero. Negative VOC values are not obtained as is sometimes the case for low VOC coatings using Method 24.
In a recent determination of the VOC content of 28 architectural coatings using D 6886, the California South Coast Air Quality Management District also carried out the determination on 13 of the 28 samples using both Method 24 and D 6886. A comparison of the results obtained by them and Cal Poly, San Luis Obispo is presented in Table 1.
Static Headspace Analysis
In static headspace analysis, a relatively small sample of coating, generally 20 milligrams or less, is placed in a 20 millilitre vial and sealed with an aluminum crimp cap. Any volatile materials present in the sample are therefore confined within this sealed vial. To analyze the sample, the vial is transferred to a precision oven where it is heated at a predetermined temperature for a specified length of time, generally at 110 to 150 °C for 10 to 20 minutes. Other temperatures and time intervals may also be used. During the heating period virtually all of the volatile components evaporate into the headspace because the sample amount is relatively small compared to the available headspace volume.
At the end of the heating period a portion of the headspace is transferred via a heated line to a gas chromatograph where the components are separated on a suitable capillary column and measured by either flame ionization or mass spectral detection. After sample preparation, the entire method is computer controlled using static headspace/GC instrumentation available from various commercial vendors. This methodology has been used successfully on various coatings previously analyzed by method D 6886 and the results are essentially the same using either method.
The static headspace method is particularly useful for coatings systems that cure by chemical reaction. These include powder coatings, melamine-cure automotive coatings, various two-component coatings, and radiation-cure coatings. An example of a melamine-cure automotive primer that was analyzed for HAP content by static headspace and also in a recent National Paint and Coatings Association-sponsored round robin for EPA Method 311, Analysis of Hazardous Air Pollutant Compounds in Paints and Coatings by Direct Injection into a Gas Chromatograph, is given in Table 2.
In headspace analysis, the sample is heated at the same temperature as the actual application cure temperature for this particular coating, making it possible to determine both the cure HAP methanol as well as the HAP solvents that are actually added during the manufacture of the coating. A particular advantage to the headspace method is that the coating does not need to be dissolved in a solvent such as tetrahydrofuran prior to analysis as is the case in a Method 311 determination.
To analyze a powder coating, a sample of the powder is placed in a headspace vial along with an internal standard, is sealed with a crimp cap, and is then heated to its recommended cure temperature. Analysis of a glycidyl methacrylate powder showed that it emits methyl methacrylate, butyl methacrylate, glycidyl methacrylate, benzoin, and several other components.
Two-component coatings are analyzed by first mixing the components and placing a small sample of the mixture along with an internal standard into a headspace vial. The vial is sealed with a crimp cap and is then allowed to cure in the vial for 24 hours. After cure the volatiles retained in the vial are analyzed by heating the vial to an appropriate temperature and transferring to the gas chromatograph. In one experiment with a water-borne polyurethane 2K coating, the Method 24 result was identical to the headspace result.
An advantage of the headspace result is that the actual amounts of the individual components emitted may be determined. Experiments are under way with ultraviolet-cure coatings to measure VOC emissions. A thin film of the UV-cure coating is sealed into a headspace vial and is then cured by subjecting it to UV light. The VOCs are retained in the vial and are then analyzed using the headspace procedure.
Cal Poly, San Luis Obispo is under contract to the California Air Resources Board to develop direct analysis methods to replace EPA Method 24. Approximately 60 to 70 architectural coatings that are currently being manufactured will be analyzed using both ASTM Method D 6886 and the new static headspace procedure. The work will be shadowed by other laboratories that have agreed to analyze some of the same samples. It is anticipated that these new VOC analysis methods will become methods of choice in the future, particularly for low VOC coatings containing water. //
1 The findings were published in 1998 and may be found at www. arb.ca.gov/research/abstracts/93-344.htm
2 Russell, et al., Science, July 28, 1995, 269: 491-495; Bergin,et al., Environmental Science and Technology, December 1995, 29:3028-3037.