Characterization of Hydrocarbon Emissions from Green Sand Foundry

Oct 10, 2007 - These binders are used in the metal casting industry for making cores that are used to create ... screening tool for the foundries to c...
0 downloads 0 Views 160KB Size
Environ. Sci. Technol. 2007, 41, 7922-7927

Characterization of Hydrocarbon Emissions from Green Sand Foundry Core Binders by Analytical Pyrolysis

useful screening tool for the foundries to compare the relative emissions of alternative core binders and to choose proper materials in order to comply with air-emission regulations.

Y U J U E W A N G , † F R E D S . C A N N O N , * ,† MAGDA SALAMA,‡ JEFF GOUDZWAARD,§ AND JAMES C. FURNESS| Department of Civil and Environmental Engineering, The Pennsylvania State University, 212 Sackett Building, University Park, Pennsylvania 16802, Material Research Institute, The Pennsylvania State University, 199 MRI Building, University Park, Pennsylvania 16802, Neenah Foundry Company, 2121 Brooks Avenue, Neenah, Wisconsin 54956, and Furness-Newburge, Incorporated, 376 Crossfield Drive, Versailles, Kentucky 40383

Introduction

Analytical pyrolysis was conducted to study a relative comparison of the hydrocarbon and greenhouse gas emissions of three foundry sand binders as follows: (a) conventional phenolic urethane resin, (b) biodiesel phenolic urethane resin, and (c) collagen-based binder. These binders are used in the metal casting industry for making cores that are used to create internal cavities within castings. In this study, the core samples were flash pyrolyzed in a Curie-point pyrolyzer at 920 °C with a heating rate of about 3000 °C/sec. This simulated some key features of the fast heating conditions that the core binders would experience at the metal-core interface when molten metal is poured into green sand molds. The core samples were also pyrolyzed in a thermogravimetric analyzer (TGA) from ambient temperature to 1000 °C with a heating rate of 30 °C/min, and this simulated key features of the slow heating conditions that the core binders would experience at distances that are further away from the metal-core interface during casting cooling. Hydrocarbon emissions from flash pyrolysis were analyzed with a gas chromatography-flame ionization detector, while hydrocarbon and greenhouse gas (CO and CO2) emissions from TGA pyrolysis were monitored with mass spectrometry. The prominent hazardous air pollutant emissions during pyrolysis of the three binders were phenol, cresols, benzene, and toluene for the conventional phenolic urethane resin and biodiesel resin, and they were benzene and toluene for the collagen-based binder. It was also found that volatile organic compound and polycyclic aromatic hydrocarbon emissions considerably decreased in order from conventional phenolic urethane resin to biodiesel resin to collagenbased binder. These results have shown some similarity with those for stack emission testing conducted at demonstration scale and/or full-scale foundries, and the similar trends in the two sets of results offered promise that bench-scale analytical pyrolysis techniques could be a * Corresponding author fax: (814) 863-7304; e-mail: fcannon@ engr.psu.edu. † Department of Civil and Environmental Engineering, The Pennsylvania State University. ‡ Material Research Institute, The Pennsylvania State University. § Neenah Foundry Company. | Furness-Newburge, Incorporated. 7922

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 41, NO. 22, 2007

The metal casting industry represents an important manufacturing component in the United States. Each year, millions of tons of metal are poured into molds to create numerous casting products that play important roles in virtually every aspect of life (1). A majority (about 90%) of the castings are produced by employing green sand molding processes. As the most popular molding process, green sand molding is adopted by more than 85% of the U.S. foundries. The green sand is a black mixture usually composed of silica sand (8595%), bentonite clay (4-10%), carbonaceous additives (usually bituminous coal) (2-10%), and water (2-5%) (2). The silica sand forms the bulk of green sand molds, the clay and water act as the binder to hold the mold together, and the coal improves the casting finish. When casting complex metal shapes, sand cores are used to create internal cavities within the casting. The sand cores must be much stronger than the clay-bound green sand mold to produce acceptable castings and sufficiently collapsible to allow removal from the casting after it has cooled. Thus, organic resins or chemical binders are usually used to make the sand cores. Over the years, core binders have evolved and adapted to changing production, casting quality, and environmental needs. Particularly, much more attention has recently been paid to the environmental concerns regarding the air emissions from foundries. Indeed, among the 189 hazardous air pollutants (HAPs) listed in the 1990 Clean Air Act Amendment, some 40 compounds have been identified in the air emissions from foundries (3), and for high-production foundries, the amount of HAP emissions is significant (4, 5). Previous research has shown that over 90% (by weight) of the foundry HAPs are organic hydrocarbons (6), and the pyrolysis of carbonaceous additives and core binders during the casting process is the predominant sources of the HAPs (6-10). Once the molten metal is poured into the green sand mold, the carbonaceous additives (e.g., bituminous coal) and core binders will be exposed to the intense heat of metal castings and undergo thermal decomposition reactions (11). A variety of organic compounds and certain HAPs could be generated during this process and then released into the air during metal pouring, mold cooling, and casting shakeout. In an emission test in which phenolic urethane cores were used, data have shown that the HAP emissions from the core binders’ pyrolysis could range from about 30% (for average cored castings) to 70% (for very heavily cored castings) of the total HAP emissions (6). As a response to the ever more stringent environmental regulations, foundries have been seeking alternative raw materials that are more environmentally friendly than the traditional ones to diminish their pollutants and operation costs. Many new core binder systems have been developed recently, predicated on the foundries’ requirement. The need exists for foundries to evaluate these new products as a part of adopting them in full-scale production. Stack testing conducted at exhaust pipes at the demonstration scale or full-scale foundries can be used for measuring emissions from casting processes, e.g., monitoring volatile organic compound (VOC) emissions by EPA method 18 or total gaseous organics by EPA method 25A. However, such stack 10.1021/es071657o CCC: $37.00

 2007 American Chemical Society Published on Web 10/10/2007

TABLE 1. Hydrocarbon Emissions (mg/g core Binder) of the Three Core Binders during Curie-Point Flash Pyrolysis; Four Replicates for Flash Pyrolysisa phenolic urethane resin (Sigma Cure 7121 and 7516) entry 1

2 3 4 5 6, 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21

emissions C1-C5as methane VOCC6-C16 as benzene sum of individual HAPs PAHs (sum of peaks 17-20) benzene pyrrole toluene ethylbenzene m-, o-, p-xylene aniline phenol tert-butylbenzene o-cresol p-, m-cresol 2,6-dimethylphenol C10H14 isomersb 2,4-dimethylphenol 2,5-dimethylphenol 3,5-dimethylphenol naphthalene 2-methylnaphthalene 1-methylnaphthalene dimethylnaphthalene (isomers) methyl oleate (C17H33COOCH3)

biodiesel resin (Sigma Cure 705 and 305)

collagen-based binder

mean

STD

mean

STD

mean

STD

42.0 107.8 32.1 28.4

12.2 11.6 1.7 3.4

80.7 78.6 39.1 3.1

13.5 7.4 4.1 0.2

72.6 41.5 7.7 0.34

5.6 5.4 1.3 0.06 0.62 0.17 0.35 0.03 0.01 0.19 0.13

individual hydrocarbons 6.31 1.49 ND 2.71 0.32 0.18 0.05 0.77 0.12 0.78 0.08 11.92 1.47 0.62 0.08 7.69 1.76 1.10 0.30 2.27 0.33 4.11 0.65 0.52 0.11 ID 0.81 9.20 4.86 13.55 14.27

0.24 1.43 0.61 1.12 2.13

7.61 ND 3.61 0.33 0.94 1.79 13.88 0.64 9.27 1.53 2.47 16.27 0.58

0.76 0.45 0.10 0.13 0.28 0.92 0.07 1.75 0.20 0.40 1.05 0.08

2.24 11.26 2.64 0.21 0.23 0.65 1.04 ID 0.39 0.32 ND 0.48 ID

ID 0.47 0.74 0.45 1.39 39.92

0.04 0.06 0.02 0.10 7.14

ID 0.20 0.07 0.07 ND 1.59

0.03 0.01 0.02

0.04 0.01 0.01 0.30

a

Hazardous air pollutants listed in Title III of the 1990 CAAA are indicated by the numbers with italic font. Abbreviation key: ND, not detected; ID, intermittently detected. b C10H14 isomers included tetramethylbenzene isomers, 1-methyl-2-(1-methylethyl)-benzene, 1-methyl-3-(1-methylethyl)benzene, and 1-methyl-4-(1-methylethyl)-benzene etc; GC-FID results calibrated with sec-butylbenzene (C10H14).

testing is fraught with considerable variability and statistical uncertainty, and there are true variations in foundry operation conditions (5). Thus, emission factors acquired at pilot-scale stack testing conditions could not be unilaterally applied to full-scale productions where numerous process factors could significantly affect the emissions. Meanwhile, the high cost and complexity of collecting emissions data in full-scale foundries hampers foundries from conducting multiple emission testings for these alternative binder systems. In this light, we proposed that it would be useful to utilize the analytical pyrolysis techniques presented herein as a preliminary screening tool to compare the relative emission characteristics of several alternative core binder systems. Laboratory analytical pyrolysis can be readily conducted and analyzed under strictly controlled conditions that simulate key features of the thermal conditions of metal casting. In the past, many analytical pyrolysis studies have been conducted to study the roles of the carbonaceous additives in the casting process. These studies have also revealed important insights regarding the emission characteristics of the carbonaceous additives (10, 12-15). In comparison, analytical pyrolysis studies on core binders’ emissions are rare (16, 17). Thus, the main objective of this research was to develop analytical pyrolysis methodologies for comparing relative air emissions from several core binders that are used in full-scale foundries. Three core binder systems were tested in this research. One was the conventional phenolic urethane resin that represented the binders that have been widely used in foundries for many years. The other two represented relatively new developments in core binder systems, i.e., a biodiesel phenolic urethane resin and a collagen-based biopolymer binder. The biodiesel resin was introduced into the foundries in the late 1990s. Instead of using high-boiling aromatic hydrocarbon solvent, biodiesel resin used plant-based solvents (methyl esters of vegetable oils), and it aimed to further

improve productivity, performance, and resin stability while reducing hazardous emissions (18, 19). The collagen-based biopolymer binder was designed more recently to address the collapsibility issues and environmental concerns for the metal casting industry, and it is undergoing full-scale trials at a handful of foundries (20).

Methods and Materials The conventional phenolic urethane resin tested in this study was prepared with Sigma Cure 7121 (part 1) and Sigma Cure 7516 (part 2) from HA International (Westmont, IL). They were shipped to us from a full-scale foundry (Neenah Foundry, Neenah, WI) where they have been used for several years. The biodiesel resin (Sigma Cure 705 as part 1 and Sigma Cure 305 as part 2) was also shipped from the same foundry, where they have been used for about half a year. The collagen-based biopolymer binder (Hormel GMBOND) was from Hormel Foods Corp. (Austin, MN). The sand was washed silica sand from Wedron Silica Company (Wedron, IL). The core samples were made with the sand that was sieved to pass a 150 µm sieve (U.S. mesh #100). The sieved sand was thoroughly washed with deionized water, dried at 105 °C in the air overnight, and then stored in a desiccator before it was used to make the cores. Core Samples Preparation. The conventional phenolic urethane core samples were prepared according to the procedure described by Neenah Foundry’s personnel: 0.55 g of Sigma Cure 7121 (part 1) and 0.45 g of Sigma Cure 7516 (part 2) were mixed with 99 g of silica sand, then the catalyst dimethylethylamine (DMEA) was added using air to blow the DMEA through the core. The hardened core was then ground to individual sand particles using a mortar and pestle before the pyrolysis test. The biodiesel core samples were prepared in the same way as described above, with the only exception that 0.55 g of Sigma Cure 705 and 0.45 g of Sigma VOL. 41, NO. 22, 2007 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

7923

FIGURE 1. GC-FID responses to the hydrocarbon emissions of the core binders and bituminous coal during Curie-point pyrolysis (peak 1 truncated for the collagen-based binder and bituminous coal; hazardous air pollutants listed inTitle III of the 1990 CAAA are indicated by the numbers with bold font). Cure 305 were used to replace the Sigma Cure 7121 and Sigma Cure 7516. The GMBOND core sample was provided by Hormel Foods Corp. (Austin, MN), with roughly the same percent of organic binder as for the others. All the core samples were stored in airtight bags before the pyrolysis test. Loss on Ignition (LOI) Analysis. The LOI values of the three core samples were measured on the same day when these samples were Curie-point pyrolyzed. About 20 g of each sample was heated at 600 °C in the air for 2 h in a muffle furnace. The temperature was lower than the 982 °C that is employed in the American Foundry Society LOI test that measures the total organics in green sand. Because the organic binders tested herein were much easier to decompose than the coals and/or cokes in the green sand, 600 °C was a sufficiently high temperature for this study. The LOI was then calculated from the mass loss as a percent value. Duplicates of LOI tests were performed for each core sample. The experimental errors were within 4%. The average of the LOI was taken as the percent of core binders in the core samples when they were Curie-point pyrolyzed. Curie-Point Pyrolysis-Gas Chromatograph-FlameIonization Detector (GC-FID) Analysis. About 10 mg of core sample was tightly wrapped in a ferromagnetic foil that was then placed in a small quartz tube (4.0 mm i.d.) in the Curiepoint pyrolyzer (Japan Analytic Ind., JHP-22). The foil was rapidly heated to its specific Curie-point temperature of 920 °C in a helium atmosphere. The heating rate was about 3000 °C/sec. This simulated key features of the intensely fast heating conditions that the core experiences at the vicinity of metal-core interface when it is surrounded by the molten 7924

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 41, NO. 22, 2007

metal after the metal pouring. The pyrofoil was held at 920 °C for 3 s and then kept in the tube as it cooled. The gases emitted via this abrupt heating were carried through a heated (200 °C) transporting tube to the GC-FID (Hewlett-Packard 5890 Series II, packed column). The GC column (Crossband, Rtx-1) was a 60 m long × 0.32 mm i.d. capillary column with a 1.0 µm 100% dimethyl polysilosane film. The temperature ramp started with an isothermal step of 4 min at 50 °C, followed by a gradient of 5 °C/min to 150 °C, and then a gradient of 10 °C/min up to a final temperature of 250 °C that was held constant for 26 min. The total run time was 60 min. Preliminary tests with a total run time of 90 min revealed that no major peaks were found after 45 min. The identification and quantitation of compound peaks were conducted by comparing the sample retention times and peak areas with those of standard solutions. No mass spectrum identifications were made for the compounds. Although other possible compounds might theoretically have the same retention times as the compounds listed in Table 1, they were not usually present in the emissions from foundries. Calibrations of FID responses were conducted with standard solutions that contained a variety of chemicals with known concentrations. Four replicates of Curie-point pyrolysis were conducted for each core sample. This protocol’s gas chromatograph did not distinguish among the hydrocarbon molecules that hosted one to five carbons. Thus, the hydrocarbons detected before hexane were summarized as C1-C5 (see Figure 1 and Table 1), while roughly 20 specific species peaks have been identified and quantified subsequent to those of C1-C5. Also, all the C6-

FIGURE 2. Comparion of full-scale stack emission testing results for a conventional phenolic urethane resin and a biodiesel resin. C16 peaks were summed as VOC hydrocarbons that were detected between hexane (C6) and hexadecane (C16). The C6-C16 sum has often been taken as a measure for the VOC emissions from American foundries (4, 6). In Table 1, the C1-C5 hydrocarbons were normalized to the standard of methane, and the C6-C16 hydrocarbons were normalized to the standard of benzene. To convert from methane equivalent to benzene equivalent, the readers can divide the methane-equivalent number by 1.46. It is noted herein that one of the major emissions of collagen-based binder cores was pyrrole (C4H5N) that was detected after hexane and thus was included in the C6-C16 group in this test. The emission quantification results from the Curie-point pyrolysis tests have been summarized in Table 1. Thermogravimetric Analyzer (TGA) Pyrolysis-Mass Spectrometer (MS) Analysis. About 100 mg of core sample was placed in a TGA 2050 (TA Instruments, Newcastle, DE) and pyrolyzed from ambient temperature to 1000 °C with a heating rate of 30 °C/min under an argon atmosphere. This simulated some key features of the slow heating conditions that the core experienced at distances that were further away from the metal-core interface. The gaseous effluent from the TGA flowed to a downstream mass spectrometer (Thermostar GSD 301T, Pfeiffer Vacuum, Inc., Nashua, NH) for emission kinetics analysis. For species identification, the mass spectra were compared to those in the NIST Mass Spectral Library (Chem SW, version 2.0, Fairfield, CA).

Results and Discussion Curie-Point Pyrolysis-GC-FID Analysis. The hydrocarbon emission characteristics of the three core binders during Curie-point pyrolysis are shown in Figure 1 and Table 1. For reference, the emissions from bituminous coal have also been included herein (10). The emissions were calculated on the basis of the LOI of these core samples. The LOIs were 1.07, 1.00, and 0.97% for the three core samples that were bound with the conventional phenolic urethane, biodiesel resin, and collagen-based binders, respectively. As shown in Figure 1 and Table 1, the three core binders had their distinct emission characteristics that were also different from those of bituminous coal. A general trend was that the fraction of C6-C16 that contained many HAPs and environmentally pertinent polycyclic aromatic hydrocarbons (PAH) considerably decreased in order from conventional phenolic urethane resin to biodiesel phenolic urethane to collagen-based binder. Compared with the conventional phenolic urethane resin, biodiesel resin generated about twice as much mass of light gases (C1-C5) that were mainly low-molecular-weight alkane and alkene (e.g., methane, ethene, propene, etc.) during Curie-point pyrolysis. In contrast, the conventional phenolic urethane resin generated more than 10 times the mass of PAH emissions as did the biodiesel resin. The different

FIGURE 3. TGA and DTG profiles of the core samples in an argon gas environment (30 °C/min). emission characteristics of the conventional and biodiesel phenolic urethane resin were mainly attributed to the different solvents in these two binder systems. The conventional phenolic urethane resin used high-boiling aromatic hydrocarbons as a solvent that generated most of the PAHs upon pyrolysis. In comparison, the biodiesel resin used methyl esters of vegetable oils as the solvent. Correspondingly, the biodiesel resin generated more light hydrocarbon gases (C1-C5) and methyl ester species (e.g., methyl oleate) during pyrolysis than did the conventional phenolic urethane resin. With regard to the hazardous air pollutants (e.g., BTX compounds), the conventional phenolic urethane resin had comparable or slightly lower levels than the biodiesel resin tested herein. The most prominent HAP emissions from the two core samples were phenol, o-cresol, benzene, and toluene. From the environmental perspective, the main advantage of the biodiesel resin over the conventional phenolic urethane resin was the tremendous decreases of PAH emissions that would occur during the intense heating at the molten metal-core interface. The VOCC6-C16 emissions also decreased 27%. During Curie-point pyrolysis, the collagen-based binder released 50-60% less VOCC6-C16 and 75-80% HAPs emissions than the other two core binders tested herein. The collagenbased binder appeared to be a more environmental-friendly core binder. The results presented herein have shown some similarity with those of stack testing conducted at full-scale and/or VOL. 41, NO. 22, 2007 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

7925

FIGURE 4. Mass spectroscopy responses to the gas emissions during TGA-MS pyrolysis under an argon atmosphere. demonstration scale foundries. Figure 2 shows the stack emission testing results conducted at Neenah Foundry. The emissions were captured off the primary mold cooling line of Neenah’s Plant 3 line, and in all tests, core sand set up with 1.1% core binder, on a weight basis as measured at the time of adding the binder. Emissions were monitored by routing a slip stream of the exhaust air through a granularactivated carbon cartridge, extracting the sorbed VOCs in a solvent, and then passing this extract through GC-FID in accordance with EPA method 18. The emissions have been normalized as pounds of total VOC’s as hexane per ton of metal poured, and this has been plotted versus the pounds of core sand per ton of metal poured. As shown in Figure 2, when a conventional phenolic urethane binder (NFCO 459 and 859, Ashland, Inc., Covington, KY) was employed, emissions averaged 0.67 lb/ton metal poured. In comparison, the biodiesel resin (Sigma Cure 705 and 305) emissions averaged 0.35 lb/ton metal poured. This represented about less than half as many emissions. Also, casting results at a demonstration scale foundry have shown that biodiesel resin generated 45% less VOC and HAP emissions than did conventional phenolic urethane resin, and the collagen-based binder resulted in about a 90% decrease in VOC and HAP emissions as compared with the conventional phenolic urethane binders (18, 20). The similarity in emission trends of bench-scale analytical pyrolysis and full-scale stacking 7926

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 41, NO. 22, 2007

tests offered promise that analytical pyrolysis techniques could be a useful screening tool for comparing the relative emissions of alternative core binders before they are adopted in full-scale production. Yet, it was worth noting that the Curie-point pyrolysis was conducted under inert atmosphere conditions that only modeled the metal pouring and casting cooling phases of the foundry process. The thermochemistry changes during the casting shakeout phase of the process when the mold is broke open and air contacts the hot core and mold materials. The oxidative atmosphere may change the relative amounts of the specific analytes and result in the formation of more oxygenated species such as aldehydes that are not readily formed during the pouring and cooling phases. This was possibly part of the reason that the emission reduction extents observed herein were not exactly the same as those observed in the stack testing for the biodiesel resin and collagen-based binder. TGA Pyrolysis-MS Analysis. The mass loss and derivative thermogravimetric (DTG) profiles during TGA-MS tests are shown for the three core samples in Figure 3. It was found that the conventional phenolic urethane core sample lost most of its mass in the temperature range lower than 500 °C: about 70% of the total mass loss occurred before 500 °C. In comparison, about 50% of the mass loss of biodiesel core samples occurred above 500 °C and particularly between 600 and 900 °C. This indicated that the biodiesel resin

maintained more thermal stability to higher temperature than did the conventional phenolic urethane resin. Compared with the other core binders tested herein, the thermal decomposition of collagen-based binder occurred in a relative narrow temperature range of 250-500 °C: about 65% of the total mass loss occurred in this temperature range. The emissions of the three core samples during TGA pyrolysis were monitored by MS. An array of MS responses (i.e., m/z values) were monitored and assigned to the species monitored herein, as shown in Figure 4 and Supporting Information Figure S1. It is important to note that fragments of higher-molecular-weight compounds may yield MS responses of the same m/z values as those assigned to the species observed herein, presumably at lower (but unknown) intensities. Thus, the TGA-MS plots mainly served to reveal the temperature ranges where these species were released from the core samples when they were slowly heated. The MS plots may also serve to semiquantitatively compare the relative abundances of these emission species among the three core binders, while deriving quantification of the emissions from the peak areas of the MS plots should be taken cautiously. It was found that the conventional phenolic urethane core and biodiesel core released emissions at almost the same temperature regimes. The major HAP emissions were released at lower temperatures (500-600 °C). The collagen-based binder also released CO and CO2 at temperatures higher than 600 °C, while it released the hydrocarbons primarily in the temperature range of 300-600 °C where its thermal decomposition occurred most vigorously. Compared with the conventional phenolic urethane core and biodiesel core, the MS responses of phenol and cresol were very low for the collagen-based binder. This was consistent with the results observed in Curie-point pyrolysis. Also, compared with the conventional phenolic urethane resin and biodiesel resin, the MS response of CO2 was very low for the collagenbased binder, and this highlighted the opportunities afforded by the collagen-based binder for diminishing greenhouse gases. In comparison, toluene and aniline MS responses were higher for the collagen-based binder core samples than for the other two core samples. This was a different result than for Curie-point pyrolysis results and indicated that the heating conditions (e.g., heating rate) might affect the distribution of pyrolysis products of the core samples.

Acknowledgments

(3) (4)

(5) (6)

(7) (8)

(9) (10)

(11)

(12) (13)

(14)

(15)

(16)

(17)

This study was funded by NSF Grant No. 0524940 and was conducted in collaboration with the Neenah Foundry.

(18)

Supporting Information Available

(19)

MS of some other species during TGA pyrolysis. This material is available free of charge via the Internet at http:// pubs.acs.org.

Literature Cited (1) U.S. Department of Energy. Trends Effecting [sic] R&D in the Metalcasting Industry; U.S. Department of Energy, Office of Industrial Technologies: Washington, D.C., March, 1996. (2) U.S. Environmental Protection Agency. Profile of the Metal Casting Industry; EPA/310/R-97/004, EPA Office of Compliance

(20)

Sector notebook project; U.S. Environmental Protection Agency: Washington, DC, 1998. Technikon Environmental Development Center. Pre-production Air Emission Test Report; Technikon, LLC: McClellan, CA, 2000; Parts 1-3. Goudzwaard, J. E.; Kurtti, C. M.; Andrews, J. H.; Cannon, F. S.; Voigt, R. C.; Firebaugh, J. E.; Furness, J. C.; Sipple, D. L. Foundry emissions effects with an advanced oxidation blackwater system. Am. Foundry Soc. Trans. 2003, 111, 1191-1211. Land, J. D.; Cannon, F. S.; Voigt, R. C.; Goudzwaard, J. Perspectives on foundry air emissions: a statistical analysis approach. Am. Foundry Soc. Trans. 2004, 112, 1075-1081. Glowacki, C. R.; Crandell, G. R.; Cannon, F. S.; Voigt, R. C.; Clobes, J. K.; Furness, J. C.; McComb, B. A.; Knight, S. M. Emissions studies at a test foundry using an advanced oxidation-clear water system. Am. Foundry Soc. Trans. 2003, 111, 579-598. Allen, G. R.; Archibald, J. J.; Keenan, T. Hazardous air pollutants: a challenge to metal casting industry. Am. Foundry Soc. Trans. 1991, 99, 585-593. McKinley, M. D.; Jefcoat, I. A.; Herz, W. J.; Frederick, C. Air emissions from foundries: a current survey of literature, suppliers and foundrymen. Am. Foundry Soc. Trans. 1993, 101, 979-990. Fox, J. R.; Adamovits, M.; Henry, C. Strategies for reducing foundry emissions. Am. Foundry Soc. Trans. 2002, 110, 12991309. Wang, Y. J.; Huang, H.; Cannon, F. S.; Voigt, R. C.; Komarneni, S.; Furness, J. C. Evaluation of volatile hydrocarbon emission characteristics of carbonaceous additives in green sand foundries. Environ. Sci. Technol. 2007, 41, 2957-2963. Wang, Y. J.; Cannon, F. S.; Voigt, R. C.; Komarneni, S.; Furness, J. C. Effects of advanced oxidation on green sand properties via iron casting into green sand molds. Environ. Sci. Technol. 2006, 40, 3095-3101. Bachmann, J.; Baier, D. Some aspects of gas evolution from carbonaceous materials used in foundry molding sands. Am. Foundry Soc. Trans. 1982, 90, 465-471. LaFay, V. S.; Neltner, S. L.; Taulbee, D. N.; Wellbrock, R. J. Evaluating emission characteristics of sea coal and sea coal supplements using advanced analytical techniques. Am. Foundry Soc. Trans. 2000, 108, 713-718. Wang, Y. J.; Cannon, F. S.; Neill, D.; Crawford, K.; Voigt, R. C.; Furness, J. C.; Glowacki, C. R. Effects of advanced oxidation treatment on green sand properties and emissions. Am. Foundry Soc. Trans. 2004, 112, 635-648. Wang, Y. J.; Cannon, F. S.; Komarneni, S.; Voigt, R. C.; Furness, J. C. Mechanisms of advanced oxidation processing on bentonite consumption reduction in foundry. Environ. Sci. Technol. 2005, 39, 7712-7718. Lytle, C. A.; Bertsch, W.; McKinley, M. D. Determination of thermal decomposition products from a phenolic urethane resin by pyrolysis-gas chromatography-mass spectrometry. J. High Resolut. Chromatogr. 1998, 21, 128-132. Dungan, R. S.; Reeves, J. B. Pyrolysis of foundry sand resins: a determination of organic products by mass spectrometry. J. Environ. Sci. Health, Part A 2005, 40, 1557-1567. Eppley, M. C.; Laitar, R. A.; Pahr, E. R.; Roush, D. C.; Tse, R.; Zaretskiy, L. S. Improved phenolic urethane coldbox foundry resin system. Am. Foundry Soc. Trans. 2005, 113, 505-510. Trinowski, D. M.; Ladegourdie, G.; Lo¨chte, K. New coldbox binder system for improved productivity. Am. Foundry Soc. Trans. 1999, 107, 51-57. Giese, S. R.; Thiel, G. R.; Herreid, R. M.; Eastman, J. D. Influence of protein-based biopolymer-coated olivine core sands on olivine green sand molding properties. Am. Foundry Soc. Trans. 2002, 110, 595-601.

Received for review July 5, 2007. Revised manuscript received September 4, 2007. Accepted September 5, 2007. ES071657O

VOL. 41, NO. 22, 2007 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

7927