Environ. Sci. Technol. 2009, 43, 2184–2189
Kinetic Rate Laws of Cd, Pb, and Zn Vaporization during Municipal Solid Waste Incineration QUENTIN FALCOZ,† D A N I E L G A U T H I E R , * ,† ´ PHANE ABANADES,† STE GILLES FLAMANT,† AND FABRICE PATISSON‡ ´ es ´ Materiaux ´ Laboratoire Proced et Energie Solaire (CNRS-PROMES), 7 Rue du Four Solaire, Odeillo, 66120 Font-RomeusFrance, and Laboratoire de Sciences et ´ ´ ´ Genie des Materiaux et de Metallurgie (LSG2M), Ecole des Mines de Nancy, Parc de Saurupt, 54000 Nancy
Received November 3, 2008. Revised manuscript received January 5, 2009. Accepted January 23, 2009.
The kinetic rate laws of heavy metal (HM) vaporization from municipal solid waste during its incineration were studied. Realistic artificial waste (RAW) samples spiked with Pb, Zn, and Cd were injected into a fluidized bed reactor. Metal vaporization was tracked by continuous measure of the above metals in exhaust gases. An inverse model of the reactor was used to calculate the metal vaporization rates from the concentration vs time profiles in the outlet gas. For each metal, experiments were carried out at several temperatures in order to determine the kinetic parameters and to obtain specific rate laws as functions of temperature. Temperature has a strong influence on the HM vaporization dynamics, especially on the vaporization kinetics profile. This phenomenon was attributed to internal diffusion control of the HM release. Two types of kinetic rate laws were established based on temperature: a fourth- or fifth-order polynomial rate law (r(x) ) k0e-EA/RTp(x)) for temperatures lower than 740 °C and a first-order polynomial (r(x) ) k0e-EA/ RT(q-qf)) for temperatures higher than 740 °C.
1. Introduction Incineration as a waste treatment alternative with volume reduction, stabilization, sanitation, and energy generation benefits is playing an increasingly important role in municipal solid waste (MSW) management. However, heavy metals contained in MSW are concentrated in the incineration byproducts, such as bottom ash, boiler ash, filter ash, and air pollution control (APC) residues (1, 2). Many factors influence the heavy metal (HM) partitioning during incineration (3-6): their physicochemical properties, which influence their evaporation or reaction kinetics; the physicochemical conditions influencing the incineration, such as temperature, chlorine content in the waste, moisture content, etc., and the parameters influencing combustion kinetics, such as temperature, retention time, or mixing conditions. High-temperature thermal treatment does not destroy metals. A fraction of the toxic metal compounds vaporizes and then condenses to form particulates during flue gas cooling or * Corresponding author phone: +33 468 307 757; fax: +33 468 302 940; e-mail:
[email protected]. † Laboratoire Proce´de´s Mate´riaux et Energie Solaire. ‡ Laboratoire de Sciences et Ge´nie des Mate´riaux et de Me´tallurgie. 2184
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 43, NO. 6, 2009
deposits on available surfaces (7). The submicrometer metal particulates and gaseous metals may pass through the APC devices, allowing some vaporized metals, which are extremely hazardous for both human health and the environment, to be emitted. It is thus essential to understand the release mechanism of metals during high-temperature waste treatment in order to better understand their behavior and to more effectively control their emissions. Most research has dealt with model wastes to investigate the factors influencing HM fate (transfer and partitioning) during incineration (8-11) or with thermodynamic calculations to discuss and predict their chemical speciation and fate (12-15). Oxygen and chlorine contents have been identified as the main parameters influencing the final HM partitioning. However, calculations based on equilibrium thermodynamics cannot predict the temporal evolution of the system and thus that of the HM. This evolution can be investigated only by kinetic studies. Rate laws were determined by Ho et al. (16) to describe the metal behavior during thermal treatment of soil. They carried out experiments in order to identify kinetic parameters in their model. Abanades et al. (9) studied the kinetics of HM vaporization from model wastes in a fluidized bed. Both organic and mineral model wastes were used to study the influence of operating conditions on the extent of HM release in fumes. Liu et al. (17) determined the rate laws of toxic metal release during thermal treatment of model waste. A first-order rate was determined for a mineral matrix, and a second-order rate was determined for realistic model waste. The objective of this study was to identify the kinetic rate law for metal release from realistic artificial waste. The behaviors of three metals of most concern (Cd, Pb, and Zn) were studied. The inverse method, developed and validated previously (18), was used to determine vaporization rates at the particle level based on experimental concentration profiles in the outlet gas of a fluidized bed reactor. Such a reactor was chosen for better control of both temperature and mass transfer. The concentration profiles were obtained by online analysis according to a measurement method involving customized ICP-OES spectrometry (inductively coupled plasma- optical emission spectrometry), which was described in detail by Falcoz (19). For calibration, the different standard gases are created by nebulizing and vaporizing liquid solutions of known metal concentrations. Experiments were carried out at several temperatures in order to determine the kinetic parameters and to obtain specific rate laws as functions of temperature. The main improvements with respect to our previous results (17) are the following: (i) the accuracy of the experimental curves is significantly improved because the new online analysis system provides measurement resolution of one point per second, thus permitting us to take into account the short-term variations of the vaporization rate, and (ii) the measurements are now quantitative due to the latest developed calibration protocol.
2. Experimental Setup and Procedure 2.1. Fluidized Bed Reactor. The experimental setup scheme is shown in Figure 1. The high temperature reactor is a fluidized bed made of AISI 316 L stainless steel, 4.5 × 10-3 m thick. It is a 0.105 m i.d. and 0.4 m high cylinder topped by a 0.2 m disengaging height. The reactor is insulated and electrically heated by two half-cylinder radiative shells. The bed is composed of sand with mean particle diameter of 0.7 × 10-3 m (bed mass: 1.6 kg; initial bed height: 15 cm), into which a given mass of reactive metal-spiked particles is injected when the reactor is at thermal steady state. The particles are directly 10.1021/es803102x CCC: $40.75
2009 American Chemical Society
Published on Web 02/18/2009
FIGURE 1. Experimental setup.
TABLE 1. Instrument and Operating Conditions ICP generator
Jobin Yvon JY 38S frequency: 40.68 MHz power: 750-1500 W three turns; H ) 15.5 mm
induction coil torch
outer quartz tube: Le 23 mm; Li 21 mm; L 71 mm intermediate quartz tube: Le 18 mm; Li 16 mm; L 74 mm argon flow: plasma gas 14 L/min; auxiliary gas 0 L/min injector (for liquid analysis): Le 4 mm; Li 3 mm; L 88 mm; alumina injector (for gas analysis): Le 2 mm; Li 1.5 mm; L 88 mm; alumina sampling line (continuous analysis of gases) line primary pump secondary pump gas flow rate
TABLE 2. Properties of Waste Particles parameters
values
particle mean diameter (mm) density (kg · m-3) pore mean diameter (µm) porosity (%)
10 0.614 200 65.1
injected inside the reactor thanks to a compressed air inlet, to prevent vaporization of the spiked metals before the particles enter the reactor. The experimental setThe fluidizing gas is preheated through a series of two electrical resistances. In order to measure the metal concentration in exhaust gases online, the nearby ICP (distance: 6 m) is connected to the gas outlet. The gas to be analyzed (sample gas) is carried to the ICP through a heated and insulated line (temperature: 250 °C). A second sampling line, bypassing the reactor, is used to obtain clean gas (i.e., metal free gas) for calibration protocol.
stainless steel tube: Le 8 mm; Li 4 mm; L ≈ 6 m heating resistance L ) 5 m, P ) 600 W (Tmax: 450 °C) diaphragm pump KNF double-headed peristaltic pump, Masterflex 6-600 rpm; pump tubing Le 5 mm; Li 1.6 mm sample gas: 0.03 L/min
2.2. Online Gas Analysis System. The ICP spectrometer (Horiba Jobin Yvon JY 38S), classically used for liquid analysis, was adapted to allow gas injection into argon plasma. The interface, consisting of two sampling stages, is shown in Figure 1. It was previously used (17, 18) for similar studies without quantitative measurement because of the lack of a validated calibration method. The calibration method developed by Falcoz et al. (20) now allows quantitative measurements. The instrument characteristics and the operating conditions are listed in Table 1. The first isokinetic sampling stage aspirates the sample gas from the reactor outlet. The gas, pumped by a diaphragm pump protected by a filter and a condenser, flows through a tube heated at 250 °C to prevent water condensation; indeed, condensation would trap most metal species by depositing on the walls. Then, a secondary sampling stage reduces and controls the gas flow rate to the admissible flow in the plasma, which depends on the power of the induction plasma torch (1500 W maximum). The secondary pump is a 2-head peristaltic pump with opposite-phase heads. This configuration is very VOL. 43, NO. 6, 2009 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
9
2185
FIGURE 2. Time course of Pb concentration in the outlet gas from waste particle, q0 ) 756 mg/kg.
TABLE 3. Values of the Sigmoid Equation Parameters, Case of Lead, 650-800°C parameters
y0
xc
a
w1
w2
w3
800 °C 770 °C 740 °C 710 °C 680 °C 650 °C
0.26 0.04 0.02 0.01 0.01 -0.01
18.90 20.45 20.85 27.26 28.97 39.56
30 20.82 9 5.66 3.68 0.86
0 0 0 0 0 28.84
0.79 1.09 1.85 3.06 3.52 4.10
9.30 9.86 17.88 14.65 15 14.43
important for plasma stability because it minimizes the flow modulation due to the pump rollers. 2.3. Sample Preparation and Procedure. Experiments were carried out with a realistic artificial waste (RAW) made from real municipal solid waste (thus containing organic species). A special procedure was developed to make spiked homogeneous particles suitable for fluidization. In this procedure, the real municipal solid waste (available as flakes) is mixed with a liquid solution of metal salt, sand (diameter: 0.5 mm), and wallpaper glue to give an MSW/sand weight ratio of 1:1. The mixture is then shredded and pressed into cylindrical particles (diameter: 10 mm, height: 6-12 mm, density: 600 kg/ m3) using a hydraulic press, and finally dried. The properties of this realistic artificial waste are listed in Table 2. The experimental procedure is as follows. Once the reactor is at steady state (desired temperature reached and operating conditions maintained), a given amount of RAW waste is injected into the bed. The time course of the metal concentration in exhaust gases is measured online by the adapted ICP-OES technique. With such burnable particles, the initial metal concentration in the waste is measured by classical ICP spectrometry after acid digestion of particles.
3. Kinetic Rate Law 3.1. Method. Experiments were carried out at several temperatures: 650, 680, 710, 740, 770, and 800 °C. For each metal, the objective was to identify the rate laws governing the vaporization rate (r ) dq/dt) as functions of the metal concentration in the waste (q), taking into account the influence of temperature. The different steps of the method were measuring the profile concentrations for each temperature (by the online analysis method developed previously), determining the vaporization rate at the particle level by the inverse method (described in ref 18) from the online diagnostic results, and identifying a kinetic rate law from the vaporization rates determined at various temperatures. 2186
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 43, NO. 6, 2009
This approach leads to a rate law that depends on temperature (such as dq/dt ) k(T)qn). Therefore, it could be used at temperatures different from those considered in our study, provided that they are within the same validity domain. The identification method is detailed for the case of lead. For zinc and cadmium, only the final results, i.e., the equations of the kinetic rate laws, are presented. 3.2. Online Analysis Results. The online analysis method was used to measure lead concentration during waste combustion in a fluidized bed, from which no sample can be drawn. Experiments were carried out with Pb-spiked RAW (see section 2.3). The initial lead concentration in waste (q0) was 756 mg/kg. The quantitative results from the combustion of Pb-spiked model waste at various temperatures are plotted in Figure 2. These quantitative results are consistent. For all experiments, both the number of spiked particles and the initial lead concentration in the particle were kept rigorously constant. The total quantity of lead released by the burning particles during waste combustion increases with temperature, which is in agreement with the Arrhenius relationship. However, the increase of metal vapor diffusion from particle to gas phase may also contribute to this phenomenon. The diffusion influence on metal vaporization would have possibly been studied by running more experiments with different particle sizes; so far, it was not for technical reasons. 3.3. Inverse Model Simulation. The lead vaporization flux at the particle level was determined by applying the inverse method (18) using the gaseous concentration profile as an inlet parameter. First, the concentration profile had to be fitted by the following sigmoid equation before being injected into the inverse model. y ) y0 + a
1 1 + e-
x-xc+w1 ⁄2 w2
(
1
11 + e-
x-xc-w1 ⁄2 w3
)
The sigmoid equation parameters are listed in Table 3 for all temperatures in the case of lead. The inverse method results in the case of lead are plotted in Figure 3. The lead vaporization rate (r ) dq/dt) is plotted as a function of time for the six temperatures. In all cases, the area under the curve increases with temperature, or in other words, the higher the temperature is, the higher the total amount of vaporized metal. The vaporization rate increases with temperature, and the peak value is reached earlier. 3.4. Kinetic Rate Law. It is essential to identify the six experimental points for each mathematical relationship (same type and same order) to validate the method. As shown in Figure
FIGURE 3. Inverse method results, time course of Pb vaporization rate (r ) dq/dt) from waste particle, q0 ) 756 mg/kg.
FIGURE 4. Fitting of experimental results by kinetic rate law, lead at T < 710 °C.
TABLE 4. Kinetic Parameters and Theoretical Laws for Pn, Zn, and Cd at Temperature Lower than 740°C metal
parameters
Pb
values theoretical law values theoretical law values theoretical law
Zn Cd
Ea(kJ/mol)
k0 (mg.kg-1 · s-1 · m-2)
qf ) f(T)
3.25 × 1011 qf ) q0/e-13938/T +12.175 + 1 r ) k0e-EA/RT(-7.8x4+20.8x3-20.8x2+7.8x+0.06) 214.5 1.7 × 1015 qf ) q0/5.781ln (T)-38.571 r ) k0e-EA/RT(14.9x5-47.1x4+55.2x3-32.4x2+9.1x+0.065) 11 140.6 4.25 × 10 qf ) q0/e-13811/T +12.802 + 1 r ) k0e-EA/RT(12.2 x5 - 40.5x4 + 52.2x3 - 33.5x2 + 9.6x + 0.004) 143.7
3, the experimental profiles vary with temperature, and all of them cannot be described by one single rate law. The determination method must distinguish between high temperature and low temperature cases: low temperature (T < 710 °C) curves were fitted to an nth-order polynomial, and high temperature (T > 740 °C) curves to a straight line (i.e., a first-order polynomial). The maximum rate (rmax) and the final concentration (qf) should only be influenced by temperature. Hence, the relationships of rmax and qf must be determined as functions of temperature. Moreover, the experimental kinetics may depend on the initial particle size. This phenomenon is integrated in the
theoretical laws by defining the vaporization rate with respect to the external surface (unit: mg kg-1 s-1 m-2). The necessary steps to obtain a rate law with temperature as a parameter are (1) fitting the rate law rmax ) f(T) to the Arrhenius equation by plotting ln(rmax) versus 1/T: thus, the activation energy Ea and the pre-exponential factor k0 are identified; (2) determining the relation qf ) f(T); (3) applying the following change of variables x ) q0-q/q0-qf to obtain a dimensionless metal concentration in waste; (4) plotting r/k0e-EA/RT versus x for each temperature: thus, normalized f(x) curves, integrating the temperature influence, are obtained; (5) fitting all curves f(x) to an nth-order polynomial P(x); and (6) finally, the kinetic rate law is given by r(x) ) k0e-EA/RTP(x). VOL. 43, NO. 6, 2009 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
9
2187
FIGURE 5. Fitting of experimental results by kinetic rate law, lead at T > 740 °C.
FIGURE 6. Validation of the kinetic rate laws, lead case.
TABLE 5. Kinetic Parameters and Theoretical Laws for Pn, Zn, and Cd at temperature higher than 740°C metals Pb Zn Cd
Ea (kJ/mol)
k0 (mg · kg-1.s-1 · m-2)
74.7 63.8 85.0
9.7 × 10 1.4 × 105 2.9 × 106 5
experimental kinetics may be fitted to a first-order rate law of general expression:
-Ea r ) k0e RT (q - qf)
qf ) f(T) qf ) q0/e-13531/T qf ) q0/e-15345/T qf ) q0/e-22132/T
+12.117
+1 +1 +21.584 +1 +14.14
Since the start of vaporization is not described by this equation, its validity domain excludes the first 10% loss of the initial concentration; i.e., the rate law is valid for qf e q e 0.9q0
Low Temperature (T < 710 °C). Figure 4 plots the comparison between the experimental kinetics and the kinetic rate law (r(x) ) k0e-EA/RTp(x)) for lead, mathematically identified when T < 710 °C. The rate laws and the corresponding parameters are listed in Table 4 for Cd, Pb, and Zn. The close agreement between the determined rate law and the experimental kinetics confirms that the lead vaporization is correctly predicted by the determined equation. Moreover, the temperature influence on the vaporization dynamics is correctly described. High Temperature (T > 740 °C). For T > 740 °C, the start of vaporization may be neglected because of its insignificant physical meaning and because it represents less than 10% of the total metal vaporized. For the three metals, the 2188
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 43, NO. 6, 2009
This restriction is not penalizing, since the first moments of the vaporization are less important for modeling the phenomenon. Figure 5 plots the comparison between the experimental kinetics and the rate law mathematically identified when T > 740 °C. The rate laws and the corresponding parameters are listed in Table 5 for Cd, Pb, and Zn.
4. Validation The kinetic rate laws were validated by carrying out other experiments. The experimental curves obtained during these experiments were not used in identifying the rate laws in order to guarantee the independency of the results. The experimental conditions were strictly identical, with the exception of the initial
metal concentrations in waste. The final concentrations are those calculated by the inverse method, applying the equations proposed in Section 3.4. Experiments were performed at 650 and 800 °C. For lead (Figure 6), the rate law successfully predicted the experimental profile at 800 °C, whereas experimental results were overestimated (up to 50% at maximum vaporization rate) by the rate law at 650 °C. For Cd and Zn, the experimental results at both temperatures were well predicted by the proposed rate laws. Thus, the specific rate laws proposed in this study are general relationships that can be used to predict the heavy metal vaporization kinetics from burning particles.
5. Discussion The experimental study revealed a strong influence of temperature on the HM vaporization dynamics. Temperature affects the vaporization kinetic profile; in other words, the HM release is controlled by internal diffusion. (1) At high temperatures, the chemical reaction controls HM vaporization. Internal diffusion is negligible in this case because of very fast particle combustion. (2) At low temperatures, two factors may explain the global kinetic slowdown and the resultant polynomial profile of the curves the reaction kinetics is intrinsically slowed down according to the Arrhenius equation; the global kinetics is limited by the HM internal diffusion because of the slower particle combustion and the reduced value of the metal diffusion coefficient. Indeed, the vaporized metal may cross an ash layer in that case, thus reducing the vaporization rate. Thus, depending on the temperature, the global kinetics may be described by two identified rate laws: a fourth- or fifthorder polynomial at temperatures below 740 °C, and a firstorder polynomial at temperatures above 740 °C. At high temperatures, the kinetic rate laws for all three metals are simple (linear relations), but their use is limited to their validity domain. Indeed, they do not describe the start of HM vaporization. In order to integrate the starting phase into the formulation, the concentration qm can be defined as the boundary value of the validity domain (qm ) 0.9q0), corresponding to the maximum vaporization rate value (rmax). Beyond qm, the vaporization rate may be supposed to be constant and equal to rmax. The rate law formulation for high temperatures becomes qf e q e qm f r ) k0eqm e q e q0 f r ) k0e
EA
⁄RT
EA
(q - qf)
- ⁄RT
(qm - qf) ) rmax
These specific kinetic rate laws may be integrated into a global model of municipal solid waste combustion in order to simulate the effects of operating parameters on heavy metal behavior.
Acknowledgments ThisstudywassupportedbyADEME(Agencedel’Environnement et de la Maıˆtrise de l’Energie) and Languedoc Roussillon Regional Council.
Appendix A NOMENCLATURE Ea k0 q0 q
activation energy, kJ/mol pre-exponential factor, mg kg-1 s-1 m-2 initial metal concentration in waste, mg/kg metal concentration in waste, mg/kg
qf qm r rmax x
final metal concentration in waste, mg/kg metal concentration in waste at the maximum vaporization rate, mg/kg metal vaporization rate, mg kg-1 s-1 m-2 maximum vaporization rate, mg kg-1 s-1 m-2 metal dimensionless concentration in waste
Literature Cited (1) Chandler, A, J.; Eighmy, T, T; Hartle´n, J; Hjelmar, O; Kosson, D, S; Sawell, S, E; Van der Sloot, H. A.; Vehlow, J. Municipal Solid Waste Incinerator Residues, The International Ash Working Group (IAWG); Elsevier Science: New York, 1997. (2) Zhang, H.; He, P. J.; Shao, L. M. Fate of heavy metals during municipal solid waste incineration in Shanghai. J. Hazard. Mater. 2008, 156, 365–373. (3) Toledo, J. M.; Corella, J.; Corella, L. M. The partitioning of heavy metals in incineration of sludges and waste in a bubbling fluidized bed: 2. Interpretation of results with a conceptual model. J. Hazard. Mater. 2005, 126 (1-3), 158–168. (4) Corella, J.; Toledo, J. M. Incineration of doped sludges in fluidized bed. Fate and partitioning of six targeted heavy metals. I. Pilot plant used and results. J. Hazard. Mater. 2000, 80 (1-3), 81– 105. (5) Belevi, H.; Langmeier, M. Factors determining the element behavior in municipal solid waste incinerators. 2. Laboratory experiments. Environ. Sci. Technol. 2000, 34 (12), 2507–2512. (6) Wey, M. Y.; Hwang, J. H.; Chen, J. C. The behavior of heavy metals Cr, Pb and Cd during waste incineration in fluidized bed under various chlorine additives. J. Chem. Eng. Jpn. 1996, 29, 494–500. (7) Barton, R. G.; Clark, W. D.; Seeker, W. R. Fate of metals in waste combustion systems. Combust. Sci. Technol. 1990, 74, 327–342. (8) Abanades, S.; Flamant, G.; Gauthier, D.; Tomas, S.; Huang, L. Development of an inverse method to identify the kinetics of vaporization of heavy metal during waste incineration in fluidized bed. J. Hazard. Mater. 2005, 124, 19–26. (9) Abanades, S.; Flamant, G.; Gauthier, D. Kinetics of heavy metal vaporization from model wastes in fluidized bed. Environ. Sci. Technol. 2002, 36 (17), 3879–3884. (10) Scharnhorst, W.; Ludwig, C.; Wochele, J.; Jolliet, O. Heavy metal partitioning from electronic scrap during thermal end-of-life treatment. Sci. Total Environ. 2007, 373 (2-3), 576–584. (11) Marias, F.; Puiggali, J. R.; Flamant, G. Modeling for simulation of fluidized-bed incineration process. AIChE J. 2001, 47 (6), 1438–1460. (12) Frandsen, F.; Dam-Johansen, K.; Rasmussen, P. Trace elements from combustion and gasification of coalsAn equilibrium approach. Prog. Energy Combust. Sci. 1994, 20, 115–138. (13) Abanades, S.; Flamant, G.; Gauthier, D. Modeling of heavy metal vaporisation from a mineral matrix. J. Hazard. Mater. 2001, 88 (1), 75–94. (14) Abanades, S. Behavior of heavy metals during municipal solid waste incineration (in French), Ph.D. Dissertation, University of Perpignan, France, 2001. (15) Me´nard, Y.; Asthana, A.; Patisson, F.; Sessiecq, P.; Ablitzer, D. Thermodynamic Study of Heavy Metals Behaviour During Municipal Waste Incineration. Process Saf. Environ. Protect. 2006, 84 (4), 290–296. (16) Ho, T. C.; Chu, H. W.; Hopper, J. R. Metal volatilization and separation during incineration. Waste Manage. 1993, 13 (5-7), 455–466. (17) Liu, J.; Abanades, S.; Gauthier, D.; Flamant, G.; Zheng, C.; Lu, J. Determination of kinetic laws for toxic metals release during thermal treatment of model waste in a fluid-bed reactor. Environ. Sci. Technol. 2005, 39, 9331–9336. (18) Abanades, S.; Flamant, G.; Gauthier, D. The kinetics of vaporization of a heavy metal from a fluidized waste by an inverse method. Combust. Flame 2003, 134 (4), 315–326. (19) Falcoz, Q. Heavy metal vaporization during municipal waste incineration: kinetics and transfer processes (in French), Ph.D. Dissertation, University of Perpignan, France, 2008. (20) Falcoz, Q.; Gauthier, D.; Abanades, S.; Flamant, G. On-line analysis of vaporized cadmium in high temperature processes. Int. J. Energy Clean Environ. 2007, 8 (3), 259–272.
ES803102X
VOL. 43, NO. 6, 2009 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
9
2189
