Environ. Sci. Technol. 2007, 41, 3741-3746
Different Catalytic Effects by Copper and Chromium on the Formation and Degradation of Chlorinated Aromatic Compounds in Fly Ash TOMAS O ¨ B E R G , * B O B E R G B A¨ C K , A N D ERIK O ¨ BERG Department of Biology and Environmental Science, University of Kalmar, SE-391 82 Kalmar, Sweden
Fly ash from municipal solid waste incineration may catalytically enhance the formation and degradation of chlorinated aromatic compounds. The activities of three Deacon catalysts in this process were investigated in a statistically designed experiment. Chlorides of copper, chromium, and nickel were added to fly ash samples and the resulting samples heated at 300 °C for 2 h in an air atmosphere. The addition of copper increases the formation of all chlorinated aromatic compounds except the low chlorinated congeners of polychlorinated dibenzo-p-dioxins and dibenzofurans. The addition of chromium decreased the formation of most chlorinated aromatic compounds except the highest chlorinated species, where it was without effect. The addition of nickel did not show any significant effect. The outcome of the experiment can be interpreted as two competing processes: the chlorination of aromatic rings and the oxidation of carbon-carbon and carbonoxygen bonds. The delicate balance between chlorination and oxidation could probably be further exploited to minimize both the emissions and the net production of chlorinated aromatic compounds from combustion.
Introduction Polychlorinatedbenzenes(PCBz’s),dibenzo-p-dioxins(PCDDs), and dibenzofurans (PCDFs) were detected in municipal solid waste incinerator (MSWI) fly ash already in the late 1970s (1). The statistical correlation pattern between different chlorinated aromatic compounds in combustion samples (fly ash and gas) was later established (2). Subsequently, a low-temperature (300-400 °C) formation pathway was indicated, both in laboratory experiments with MSWI fly ash and by measurement results from full-scale combustion plants (3, 4). The availability of chlorine gas through the Deacon process was postulated to be of importance in the formation of PCDD (5), and copper is a well-known active oxychlorination catalyst in this process (6). Copper is also a catalyst in the Ullmann coupling reaction, which may have a role in the final formation of PCDDs and PCDFs (7). The catalytic activity of copper in the formation of chlorinated aromatic compounds has been frequently demonstrated in both laboratory experiments and full-scale trials (8-15). However, copper is not the only metal that is an active oxychlorination catalyst; a substantial number of other transition metals show similar properties (6). MSWI fly ash * Corresponding author phone: +46 480 446247; fax: +46 480 447305; e-mail:
[email protected]. 10.1021/es0625069 CCC: $37.00 Published on Web 04/17/2007
2007 American Chemical Society
is composed of a complex mixture of all elements used in society. It is therefore of interest to widen the scope to also examine the catalytic activity of other elements in the formation of chlorinated aromatic compounds. The elements investigated in laboratory studies (Na, K, Mg, Ca, Mn, Fe, Ag, Zn, Cd, Hg, Sn, and Pb) all seem to show less activity than copper (16, 17, 10, 18), but several active oxychlorination catalysts are missing from those investigated (e.g., chromium and nickel). Results from full-scale experimental studies indicate that catalytic activity could also be attributed to these “missing” elements (19, 20). Catalytic destruction of chlorinated aromatic compounds in fly ash by dechlorination (hydrogenation) has been shown to occur under oxygen-deficient conditions (21-25). Copper also seems to be the most active catalyst for the dechlorination reactions compared to other elements investigated. Under oxygen-rich conditions (air atmosphere), the degradation by oxidation may be substantial if the temperature is increased above 400 °C (3). The net outcome of the simultaneous and competing formation and degradation processes will thus be highly dependent upon the temperature and gas composition (26). In addition, the residence time and fly ash composition are also critical parameters in determining the balance between formation and degradation (27). In a previous full-scale experimental study, we found correlation patterns that indicated the importance of metals other than copper in the net formation of chlorinated aromatic compounds in MSWI fly ash (20). Chromium, nickel, and tungsten showed a substantial contribution to this statistical correlation. Chromium and nickel are both known catalysts of the Deacon process, and chromium is reported to be more active than copper (6). Tungsten is also mentioned in the patent literature and could act as a promoter (28). The concentrations of copper (220-1800 µg/g) and chromium (200-1600 µg/g) in MSWI fly ash were comparable in the previous study, while those of nickel (36-110 µg/g) and tungsten (12-50 µg/g) were lower. We therefore suggested applying a strategy centered on the statistical design of experiments to evaluate the relative importance and interactions of these elements. The present laboratory study aims at such an evaluation for copper, chromium, and nickel.
Experimental Section Fly ash was collected from the electrostatic precipitator (ESP) at the 20 MW biofuel incinerator operated by Borla¨nge Energi, Sweden. The fuel used during the sampling period consisted of a mixture of household and industrial wastes (paper and textiles). This is the same plant as in our previous correlation studies (20). Experimental Setup. Controlled heating experiments with the fly ash were carried out in an air atmosphere. Fly ash samples of 5 g were added to heat-resistant glass test tubes (300 × 25 mm), which were sealed with permeable polyurethane foam (PUF) plugs (40 × 30 mm) and placed for 2 h in a thermostatically controlled multitube block heater (Tecator Digestion System 40, model 1016) operated at 300 °C. Experimental conditions were selected on the basis of literature data and previous range-finding experiments. Prior to use, the PUF plugs were cleaned by Soxhlet extraction with hot toluene and checked for purity. The temperature settings for the heating block were verified by separate measurements with a K-type thermocouple. Repeated measurements were made to establish the time-temperature profile in the fly ash samples during the experiment. A stable equilibrium temperature in the fly ash VOL. 41, NO. 10, 2007 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
9
3741
TABLE 1. Experimental Conditions run
CuCl amt (mmol)
CrCl3 amt (mmol)
NiCl2 amt (mmol)
NaCl amt (mmol)
1 2 3 4 5 6 7 8 9 10
0 0 0 0 5 5 5 5 2.5 2.5
0 5 0 5 0 5 0 5 2.5 2.5
0 0 5 5 0 0 5 5 2.5 2.5
30 15 20 5 25 10 15 0 15 15
(293-299 °C) was reached after 20 min. The temperature around the PUF plugs was also checked and found to be below 40 °C at the end of the heating experiment. After heating, the test tube with fly ash and PUF plug was analyzed as a composite sample. An external accredited laboratory (ALcontrol AB) carried out the analyses for the PCDDs and PCDFs according to CEN 1948 (29). The same laboratory analyzed the chlorinated benzenes by highresolution gas chromatography and selected ion monitoring mass spectrometry (HRGC-MS/SIM). The analytical procedure has been described in an earlier paper (30). Analytical results for PCDDs and PCDFs were reported for the 17 specific congeners used to calculate the international toxic equivalent quantity (I-TEQ) (31). The more recent WHO weighting schemes to calculate TEQ will give similar results, but I-TEQ is preferred for backward comparability. Analytical results for chlorinated benzenes were reported for 11 congeners (di to hexa). Statistical Design of the Experiment. A statistical approach to the design of experiments facilitates the separation and estimation of effects in a multifactor setting (32). Factorial experiments carried out at two factor levels can be used to estimate both linear effects and interactions. The catalytic effects of copper, nickel, and chromium were evaluated by adding excess amounts of the chloride salts to the fly ash and applying a full factorial design, Table 1. Sodium chloride was added to balance the addition of chlorides to keep the total amount constant. High-grade chemicals were purchased from Sigma-Aldrich: CuCl (g97%), CrCl3 (anhydrous, >98%), NiCl2 (98%), and NaCl (99.5%). The oxidation states of catalytically active metals change during the heating process, so the initial state is probably not critical (33); however, this was not investigated further by us. Each sample was ground and thoroughly mixed before being heated in the test tubes, and excess amounts of the chloride salts were
used to avoid any effects of these operations. Two replicate trials were carried out at intermediate conditions to assess the model and experimental error. All ten runs were performed in one batch. Data Analysis. The outcome of the experiments was evaluated using standard statistical methodology (32). Linear polynomial models were fitted to the data using multiple linear regression (MLR), and the effects to include were selected from normal probability plots. Analysis of variance (ANOVA) was used to assess the statistical significance of the models and the various model parameters. The different congeners of chlorinated benzenes, PCDDs, and PCDFs are often correlated (2, 34). Partial least-squares regression (PLSR) is a suitable method to model such collinear data (35, 36). PLSR is based on linear transformations of the original variables to a limited number of orthogonal factors, attempting to maximize the covariance between a linear combination of the independent predictor variables and a linear combination of the dependent response variables. Cross-validation was used to establish the rank of the PLSR models, and all variables were standardized to zero mean and unit variance before the data analysis. An adequate model requires the errors to be normally and independently distributed with constant variance. To fulfill this requirement, a variance-stabilizing transformation of the dependent response variables may be needed. Suitable power transformations (YR) can be found empirically (37). Values below the detection limit were treated as half this limit. The data analysis was carried out using the software Design-Expert v6.0 (Stat-Ease Inc.), Statistica v6.1 (StatSoft Inc.), and Unscrambler v9.1 (CAMO software A/S).
Results The purpose of the factorial experiment was to evaluate the catalytic effects of chromium and nickel compared to copper for low-temperature formation of chlorinated aromatic compounds. This experiment would also provide an indication of the feasibility of the multifactor statistical approach. Chlorinated Benzenes. The experimental design is reported in Table 1 (see the Experimental Setup), with the results for chlorinated benzenes shown in Table 2. The first run, without any addition of metal chlorides, indicates a substantial formation of chlorinated benzenes. The experimental arrangement thus seems to be responsive, producing results similar to those obtained with flow reactors (17). A linear model was fitted to the experimental factor set points and the sums of the measured concentrations of chlorinated benzenes (per gram of fly ash). A logarithmic
TABLE 2. Chlorinated Benzene Concentrations (ng/g) in the Fly Ash and after Each Experimental Runa congener
fly ash
run 1
run 2
run 3
run 4
run 5
run 6
run 7
run 8
run 9
run 10
addition 13CBz 14CBz 12CBz 135CBz 124CBz 123CBz 1235CBz 1245CBz 1234CBz PeCBz HxCBz sum CBz’s
