Volatilization of Heavy Metals (As,Pb,Cd) During Co-Processing in Cement Kilns

Abstract

Heavy metals emitted by flue gas during co-processing hazardous waste (HW) in cement kilns presents a serious threat to the environment. Volatilization characteristics of three semivolatile metals (As, Pb, and Cd) in different chemical forms (As₂S₃, NaAsO₂, Na₂HAsO₄, PbS, PbO, PbCl₂, CdO, CdS, CdCl₂) during co-processing with cement raw materials were studied. Thermogravimetric experiments were performed to tentatively predict the lower limits of volatilization temperatures. Experiments involving co-processing of these compounds with cement raw meals was carried out at different temperatures over different time periods in an electric heated tube furnace to gain more insight into the macroscopic kinetic characteristics of volatilization. Volatilization of As, Pb, and Cd during co-processing with cement raw materials increases with time initially, and after 25 min, they do not volatilize. Temperature and compound speciation significantly affect metal volatilization. For Pb and Cd, volatilization was higher at higher temperatures. Volatilization of As, however, decreased as temperature increased at temperatures above 1,000°C, which can be explained by three parallel reactions. Kinetic models on the volatilization of these compounds during co-processing with cement raw materials were obtained. These models can provide a preliminary theory basis for predicting the release of heavy metals (Pb, Cd, As) and be used to control the environmental risk posed by heavy metals through restricting their feeding rate during co-processing HW in cement kilns.

Introduction

With continued industrialization and development of the economy, the quantity of hazardous waste (HW) produced globally is rapidly increasing. The development of solutions and techniques for the disposal of HW has received much attention across the world; currently, incineration is one of the most efficient options. Combustion in cement kilns is an alternative to disposal of certain HW streams. Co-processing of HW in cement kilns has been used in developed countries for more than 30 years. However, the environmental impacts associated with this technique are becoming a concern for the cement industry globally. Cement kilns possess certain advantages over incinerators that make them suited for HW treatment, for example, they allow for high incineration temperatures and long retention times, they provide an oxidizing and alkaline combustion environment, they have destruction and removal efficiencies, and they provide the ability to recover energy and materials from HW. Co-processing in cement kilns is, therefore, a commonly used technique. However, there are various environmental security problems associated with this technique, such as the accumulation of toxic heavy metals introduced by wastes and raw materials, which have gained attention in recent years (Guo and Eckert, 1996; Trezza et al., 2000; Yan et al., 2004; Li, 2012). Co-processing HW may release heavy metals into the environment, since their volatility is enhanced at high temperatures. Metal flue gases formed through the combustion process are the main sources of metal emissions, and are extremely hazardous for both human health and the environment. Concern about the formation of fine inhalable particles enriched in toxic metals or gaseous emissions has led to the development of stringent control measures; regulations of permissible effluent concentrations of heavy metals during co-processing of HW in cement kilns have been established by the European Union (UNEP, 2010), the United States (US EPA, 2013), and China (MEP, 2013).

Currently, domestic and international researchers have predominantly focused on the solidification properties of heavy metals, and particularly on leaching behavior (Eckert and Guo, 1998; Zhang, 2009; Siddique and Rajor, 2012; Van der Sloot and Kosson, 2012). Unfortunately, there have been few investigations of detailed processes of heavy metals being vented into flue gas at different temperatures over different time periods during co-processing of HW in cement kilns, although significant environmental concerns have been raised regarding metal emissions through cement kiln stacks. Hence, it is essential to understand the volatilization behavior of harmful metals during the various processes involved in co-processing to better understand their fate in cement kilns. Heavy metal volatilization has been studied mainly by the direct analysis of cement raw materials and clinker produced by calcining at 1,450°C; the volatilization is calculated based on the proportion of heavy metals fixed in clinker (Murat and Sorrentino, 1996). These analyses are not completely satisfactory since they can only give a view of heavy metals volatilized over a relatively long time period at a constant temperature. For the past few years, extensive studies have been developed with respect to the distribution characteristics of heavy metals under equilibrium conditions. However, it is nonequilibrium reactions that occur in the first stage and equilibrium status may not be reached at all during co-processing of heavy metal wastes in cement kilns. Kinetic analyses can show how metal volatilizations vary with time and temperature of co-processing in cement kilns. This information is very important in developing strategies to optimize flue gas treatment and to minimize the impact of metal pollutants. However, few kinetic analyses of metal emissions in flue gas at different temperatures over different time periods exist in available literature.

Lead (Pb), cadmium (Cd), and arsenic (As) are common heavy metal pollutants, and the impacts of these metals on human health and the environment have increasingly captured the attention of researchers in recent years (Linak and Wendt, 1993). Pb and Cd are typically semivolatile heavy metals and evaporate more easily during co-processing in cement kilns. As is considered to be nonvolatile matter, according to a reference document on best available techniques in the cement and lime manufacturing industries (EC, 2010). A recent study confirmed that As volatilizes easily and is released along with flue gas in cement kilns (Yan et al., 2009). Because of their hazardous effects and volatile characteristics, Pb, Cd, and As are strictly controlled in flue gas in accordance with the pollutant emission control standard of co-processing HW in cement kilns. Since the chemical forms of heavy metals directly affect their volatilization behavior during co-processing in cement kilns, As₂S₃, NaAsO₂, NaAsO₄, PbS, PbO, PbCl₂, CdO, CdS, and CdCl₂, which are the most common forms of As, Pb, and Cd in HW, were selected to conduct the laboratory experiments. Thermogravimetric experiments were performed to predict the lower limits of volatilization temperatures for heavy metal compounds. Since most of the heavy metal elements were incorporated in the clinker, their volatilization could be restrained in the sintered clinker phases (Barros et al., 2004). Then experiments involving co-processing of these compounds with cement raw meals were carried out to further study the characteristics of volatilization of these heavy metals. The behavior of these heavy metals during co-processing with cement raw meals was experimentally and theoretically studied in a laboratory scale tube furnace. A method was implemented to continuously monitor the concentrations of heavy metals in combustion flue gas to track the metal release process, based on successively collecting flue gas during co-processing. This is the first study to simulate the volatilization process of heavy metals through direct quantitative analysis of heavy metal concentrations in the flue gas at different temperatures and different times during co-processing in cement kilns. This article also found the abnormal volatilization behavior of As during co-processing with cement raw materials and reasonably explained it by three parallel reactions. This novel theory solved the contradiction from different literatures on the volatilization characteristics of As. In addition to experimental research, the kinetic models of heavy metal volatilization were established for the first time to simulate heavy metal volatilization tendencies under nonequilibrium conditions in the cement system during co-processing of heavy metal compounds with cement raw materials. These models can provide a preliminary theory basis for predicting the release of heavy metals (Pb, Cd, As) and be used to control the environmental risk posed by heavy metal release through restricting their feeding rate during co-processing of HW in cement kilns.

Experimental Materials and Methods

Experimental materials

For this study, analytical reagents of sodium arsenite, sodium arsenate, lead oxide, lead sulfide, lead chloride, cadmium oxide, cadmium sulfide, and cadmium chloride were chosen as heavy metal wastes because oxides, sulfides, and chlorides are the main existing forms of As, Pb, and Cd in HW. The cement raw meal used in the experiments was collected from a homogenizing and storage silo of a cement plant. Basic data of cement raw meal are shown in Tables 1 and 2.

Table 1.

Basic Ingredients in Raw Meal (%)

Chemical components										Cement modulus
SiO₂	Al₂O₃	Fe₂O₃	CaO	MgO	K₂O	Na₂O	SO₃	Loss	Sum	KH	SM	IM
13.87	2.52	1.74	42.56	2.18	0.68	0.06	0.48	35.02	99.11	0.9645	3.256	1.448

IM, iron modulus; KH, lime saturation factor; SM, silica modulus.

Table 2.

Proportioning of Raw Meal (%)

Limestone	Sandstone	Bauxite	Steel slag	Desulfurization gypsum	Total
87.2	7.1	2.3	2.8	0.6	100

Inductively coupled plasma mass spectrometry (ICP-MS; Agilent 7500a) was used to measure the heavy metal concentrations. Simultaneous differential technique (SDT Q600; TA Company) was used to obtain the thermal gravity (TG) curves, and an electric heated tube furnace was used for clinker calcining experiments.

Experimental methods

TG experiments

The pure reagents NaAsO₂, Na₂HAsO₄, As₂S₃, PbO, PbS, PbCl₂, CdO, CdS, and CdCl₂ were analyzed with the TG method. The experiments were carried out in air atmosphere with a flow rate of 100 mL/min, and a heating rate of 20 K/min from room temperature to a final temperature of 1,400°C. The sample mass used was ∼10 mg in all cases.

Experiments of clinker calcining

The mixtures of reagents NaAsO₂, Na₂HAsO₄, As₂S₃, PbO, PbS, PbCl₂, CdO, CdS, or CdCl₂ and cement raw meal were used in calcining experiments. The quantities of the heavy metal reagents required for co-processing with cement raw meal were chosen to be low enough to prevent the influence of heavy metals on the quality of clinker and to be high enough for the concentrations of heavy metals in flue gas to be above the detection limits of instrumentation. As a consequence, mass fractions of Pb and Cd in cement raw meal were 1%, and that of As was 0.5%. All of the above compounds and cement raw meal were kept uniformly mixed before the calcination experiments. Figure 1 shows the schematic of calcination experiment device. When the tube furnace reached the set temperature, the porcelain combustion boat with 30 g of the mixed sample was pushed into the middle of the tube furnace and the set temperature was kept constant throughout the calcination process for around 40 min in the air atmosphere. During the calcination process, the gaseous emissions were collected once every 3 min until this calcination process finished using the mixed aqueous acidic solution of 5% H₂O₂ and 10% HNO₃, which is an adsorption solution according to United States Environmental Protection Agency Method 29 (US EPA, 1996). Finally, the combustion boat was removed from the tube furnace and cooled to room temperature.

FIG. 1.

Schematic of clinker calcining experiment device.

In this study, the proportion of heavy metals volatilized is defined as α: \documentclass{aastex}\usepackage{amsbsy}\usepackage{amsfonts}\usepackage{amssymb}\usepackage{bm}\usepackage{mathrsfs}\usepackage{pifont}\usepackage{stmaryrd}\usepackage{textcomp}\usepackage{portland, xspace}\usepackage{amsmath, amsxtra}\pagestyle{empty}\DeclareMathSizes{10}{9}{7}{6}\begin{document} \begin{align*}\alpha = { \frac { \rm V \times C } { { \rm M } } } \ast 100 \% \tag { 1 } \end{align*} \end{document}

where V is the volume of adsorption solution (L), and C is the concentration of heavy metal in the adsorption solution (g/L). M is the mass of the heavy metal in the mixed sample (g).

Results and Discussion

Analysis of TG experiments

Figure 2 shows the TG curves and corresponding derivative thermal gravity (DTG) curves for As compounds. The DTG curve was gained by derivating the corresponding TG curve.

FIG. 2.

TG and DTG curves for As compounds (weight (%) represents the residual mass fraction of the sample). DTG, derivative thermal gravity; TG, thermal gravity. (A) AS₂S₃; (B) NaAsO₂; (C) Na₂HAsO₄·12H₂O.

Figure 2A shows the TG and DTG curves for As₂S₃. The initial weight loss temperature is ∼330°C and the final weight loss temperature is ∼600°C, with a sharp DTG peak at 419°C. As₂S₃ begins to melt at 300°C, and boil without decomposition when heated to 707°C (Dean et al., 1973). It can be deduced that volatility temperature of As₂S₃ is higher than its melting point, but lower than the boiling point. Figure 2B demonstrates the TG and DTG curves of NaAsO₂. The initial reaction temperature is ∼800°C and the final reaction temperature is ∼1,150°C, with a sharp DTG peak at 1,030°C. Figure 2C shows the TG and DTG curves of Na₂HAsO₄. There is a distinct decline around 100°C in the TG curve owing to the vaporization of the moisture, which occupies 53.7% of the total mass. After that, with the increase in temperature, there is no remarkable weight loss. It can be concluded that As in Na₂HAsO₄ is not volatile during co-processing in cement kilns.

Figure 3A–C show the TG and DTG curves for PbO, PbS, and PbCl₂, respectively. Figure 3A shows that the weight loss temperature range of PbO is 900–1,300°C, in which an obvious DTG peak is at 1,144°C. The volatilization temperature of PbO is lower than its boiling point of 1,516°C. Figure 3B demonstrates that the weight loss temperature of PbS ranges from 800°C to 1,300°C, and one DTG peak appears at 1,155°C. Figure 3C demonstrates that the weight loss temperature of PbCl₂ ranges from 500°C to 700°C, and one DTG peak appears at 606°C. The volatilization temperature of PbCl₂ is above its melting point of 501°C, but far below the boiling point of 953°C.

FIG. 3.

TG and DTG curves for Pb compounds and Cd compounds. (A) PbO; (B) PbS; (C) PbCl₂; (D) CdO; (E) CdS; (F) CdCl₂.

Figure 3D shows the TG and DTG curves of CdO. The initial weight loss temperature is ∼1,000°C and the final weight loss temperature is about 1,300°C, with a sharp DTG peak at 1,248°C. The TG and DTG curves of CdS are similar to those of CdO (Fig. 3E). Figure 3F shows the TG and DTG curves of CdCl₂. Because CdCl₂•2.5H₂O was used in the experiment, the sample was completely dehydrated before reaching 200°C, which can be observed from the first peak of the DTG curve. The actual weightlessness temperature of CdCl₂ ranges from 500°C to 700°C, and an obvious peak appears at 650°C. The volatilization temperature of CdCl₂ is above its melting point of 568°C, but far below the boiling point of 961°C.

The TG experiments demonstrate that weightlessness temperatures vary between different chemical compounds of an element. The initial weight loss temperature is usually not the boiling point temperature, but is slightly higher than the melting point temperature. In this study, the initial weight loss temperature was used as the reference to determine the temperature range to be used in the following experiment of co-processing with cement raw materials. In addition, these results show that chlorides (PbCl₂, CdCl₂) are much easier to volatilize than the corresponding oxides (NaAsO₂, Na₂HAsO₄, PbO, CdO) and sulfides (PbS, CdS), and the volatilization characteristics of oxides (PbO, CdO) are similar to those of the corresponding sulfides (PbS, CdS).

Analysis of heavy metal volatilization characteristics during co-processing with cement raw meals

Figure 4A shows the impact of time and temperature on the volatilization of As in As₂S₃. Volatilization increases with time over the first 25 min; after that, As does not volatilize. Volatilization increases as the temperature increases below 1,000°C, while beyond 1,000°C, volatilization decreases sharply as temperature increases. The maximum volatilization proportion below 1,000°C is 60%, and above 1,000°C is 18%. Because the TG experiment indicated that NaAsO₂ is not volatile below 900°C, the calcination experiments were conducted starting at a temperature of 900°C. As shown in Fig. 4B the variation in volatilization of As in NaAsO₂ with time is similar to that of As₂S₃, and volatilization always decreases as temperature increases.

FIG. 4.

Effect of temperature and time on the volatilization of As. (A) AS₂S₃; (B) NaAsO₂.

In summary, the volatilization proportion of As in both As₂S₃ and NaAsO₂ decreased as temperature increased in the high temperature zone. Volatilization proportion of As is below 20% when the temperature is higher than 1,000°C, which indicates that As is mainly solidified in the clinker. In fact, research has shown that CaO produced when the cement raw materials are calcinated could react with As to form stable solid Ca₃(AsO₄)₂, restraining the volatilization of As (Wang et al., 2011). Volatilization of As in As₂S₃ increases until a threshold temperature is reached, then it decreases as temperature increases, which can be explained in more detail in Fig. 5.

FIG. 5.

Possible reactions of As during co-processing in cement kilns. (A) Reactions below 1,000°C; (B) reactions over 1,000°C.

Possible reactions of As during co-processing in cement kilns include: \documentclass{aastex}\usepackage{amsbsy}\usepackage{amsfonts}\usepackage{amssymb}\usepackage{bm}\usepackage{mathrsfs}\usepackage{pifont}\usepackage{stmaryrd}\usepackage{textcomp}\usepackage{portland, xspace}\usepackage{amsmath, amsxtra}\pagestyle{empty}\DeclareMathSizes{10}{9}{7}{6}\begin{document} \begin{align*}{ \rm AB} + { \rm AsX} \rightarrow { \rm AsY} \ { \rm ( g ) } \uparrow \tag{2}\end{align*} \end{document} \documentclass{aastex}\usepackage{amsbsy}\usepackage{amsfonts}\usepackage{amssymb}\usepackage{bm}\usepackage{mathrsfs}\usepackage{pifont}\usepackage{stmaryrd}\usepackage{textcomp}\usepackage{portland, xspace}\usepackage{amsmath, amsxtra}\pagestyle{empty}\DeclareMathSizes{10}{9}{7}{6}\begin{document} \begin{align*}{ \rm AB} + { \rm AsX} \rightarrow { \rm AsZ} \ { \rm ( s ) } \downarrow \tag{3}\end{align*} \end{document} \documentclass{aastex}\usepackage{amsbsy}\usepackage{amsfonts}\usepackage{amssymb}\usepackage{bm}\usepackage{mathrsfs}\usepackage{pifont}\usepackage{stmaryrd}\usepackage{textcomp}\usepackage{portland, xspace}\usepackage{amsmath, amsxtra}\pagestyle{empty}\DeclareMathSizes{10}{9}{7}{6}\begin{document} \begin{align*}{ \rm AsZ ( s ) } \rightarrow { \rm AsZ} \ { \rm ( g ) } \uparrow \tag{4}\end{align*} \end{document}

where AB represents the cement raw meals, AsX and AsY represent compounds of As added to the cement meals and a gaseous compound of As produced by reaction (2), respectively; AsZ represents a solid compound of As that is difficult to volatilize and is produced by reaction (3); and reaction (4) represents the slow volatilization process of compound AsZ.

As shown in Fig. 5A and B, the X-axis represents reaction time, and the Y-axis represents the proportion of As, and curves (2) and (3) represent reactions (2) and (3), respectively. Segments AB, BC, CD represent As that is volatile, unreacted, and solidified, respectively.

For As₂S₃, reaction (2) is the main reaction below 1,000°C, whereas reaction (3) is also occurring, but slowly, which can be seen in Fig. 5A. Therefore, the volatilization proportion of As in As₂S₃ increased as the temperature increased below 1,000°C.

However, when the temperature exceeds 1,000°C, the reaction rate of reaction (3) increases intensely, which is shown by curve (3) in Fig. 5B, at the same time, the reaction rate of reaction (2) also increases, which is shown by curve (2) in Fig. 5B. After a short period of time, which is indicated as intersection E of curve (2) and (3) in Fig. 5B, reactions (2) and (3) are completed, since there is no unreacted As left and all As left in the sample is in the form of AsZ, which is not easy to volatilize. After that, only reaction (4), which is shown by curve (4) in Fig. 5B occurs. When the temperature increases continually, the reaction rates of reactions (2) and (3) increase simultaneously, as indicated by curves (2)′ and (3)′ in Fig. 5B. Then intersection E of curves (2) and (3) transfers to the intersection E′ in the lower left corner compared with its original one, as shown in Fig. 5B. This causes curve (4) to decline, as indicated by curve (4)′, leading to a decrease in volatilization with an increase of temperature in the higher temperature zone.

Volatilization character of As in NaAsO₂ can also be explained by the same theory. The only difference between As₂S₃ and NaAsO₂ is that both reactions (2) and (3) are very slow in the lower temperature zone for NaAsO₂.

Figure 6 A, B, C show the impact of time and temperature on the volatilization of Pb in PbO, PbS, and PbCl₂. Volatilization increases with time in the first 25 min; after that, Pb does not volatilize. The maximum proportions of Pb volatilized in PbO, PbS, and PbCl₂ are 96%, 93%, and 97%, respectively. Pai-Haung Shih (Shih et al., 2005) pointed out that Pb is a light volatile metal and the volatilization rate surpassed 90% in the clinker sintering process, which is in agreement with this study. Figure 6A shows that the volatilization of Pb in PbO increases with the temperature consistently throughout the entire process and increases sharply when the temperature reaches 1,200°C. Figure 6B and C also indicates that the volatilization of Pb in PbS and PbCl₂ increases with temperature. However, PbCl₂ appears to be more volatile than PbS. Related research has indicated that Pb is substantially lost by volatilization at high temperatures, particularly in the presence of chlorine; chlorine possibly promotes Pb volatilization (Chiang et al., 1997; Rio et al., 2007).

FIG. 6.

Effect of temperature and time on the volatilization of Pb. (A) PbO; (B) PbS; (C) PbCl₂.

Figure 7A–C shows the impact of time and temperature on the volatilization of Cd in CdO, CdS, and CdCl₂. Volatilization increases with time for the first 25 min; after that, Cd does not volatilize. The maximum proportions of Cd volatilized in CdO, CdS, and CdCl₂ are 98%, 94%, and 97%, respectively. Figure 7A indicates that volatilization of Cd in CdO increases with temperature consistently throughout the process. Figure 7B and C also indicates that the volatilization of Cd in CdS and CdCl₂ increases with temperature. However, CdCl₂ appears to be more volatile than CdS, which is the same as Pb.

FIG. 7.

Effect of temperature and time on volatilization of Cd. (A) CdO; (B) CdS; (C) CdCl₂.

Kinetic models of heavy metal volatilization during co-processing with cement raw meals

There are many existing kinetic models, but most of these models are related to the TG experiments in which the temperature is increasing linearly with time (Coats and Redfern 1964; Font et al., 2011). However, in each set of these calcination experiments, the temperature was staying the same, and the heavy metal volatilization in the calcination experiment was an isothermal process. Then different sets of calcining experiments under different temperatures were conducted. Based on the concentrations of As, Pb, and Cd in the flue gas from laboratory scale calcining experiments, volatilization kinetic models of these heavy metals during co-processing were established according to the isothermal kinetics, which has not been widely used in recent literatures. Although the isothermal kinetics needs more set of isothermal experiments under different temperatures compared with nonisothermal kinetics, it does not need temperature to be increased linearly with time and is very suitable to the experiments conducted in tube furnaces.

Many heterogeneous reactions corresponding to volatilization are described by the model of pseudohomogeneous kinetics (Helsen and Van den, 2000; Adnadjevic et al., 2007), resulting in the following rate equation: \documentclass{aastex}\usepackage{amsbsy}\usepackage{amsfonts}\usepackage{amssymb}\usepackage{bm}\usepackage{mathrsfs}\usepackage{pifont}\usepackage{stmaryrd}\usepackage{textcomp}\usepackage{portland, xspace}\usepackage{amsmath, amsxtra}\pagestyle{empty}\DeclareMathSizes{10}{9}{7}{6}\begin{document} \begin{align*} \frac { { \rm d } \alpha } { { \rm dt } } = { \rm k ( T ) ( 1 - \alpha ) ^n } \tag { 5 } \end{align*} \end{document}

where α represents the volatilization proportion [in the present context, Eq. (1)], t represents the calcination time (min), k represents the reaction rate constant, T represents the absolute temperature of the reaction (K), and n represents the reaction order.

The variation of the rate constant with temperature is approximated by the Arrhenius equation, because this relationship dominates physical and chemical phenomena (Carrasco, 1993) \documentclass{aastex}\usepackage{amsbsy}\usepackage{amsfonts}\usepackage{amssymb}\usepackage{bm}\usepackage{mathrsfs}\usepackage{pifont}\usepackage{stmaryrd}\usepackage{textcomp}\usepackage{portland, xspace}\usepackage{amsmath, amsxtra}\pagestyle{empty}\DeclareMathSizes{10}{9}{7}{6}\begin{document} \begin{align*}{ \rm k ( T ) = { \rm Aexp} \left( - {{ \rm E} \over{ \rm RT}} \right) } \tag{6}\end{align*} \end{document}

where A represents the preexponential factor (min⁻¹), E represents the activation energy (KJ/mol), and R represents the universal gas constant (8.314 J/[mol·K]).

Under isothermal conditions, the integration of reaction (2) between t₀ and t is easily carried out assuming t₀=0, α_t ₌ _t0=0:

for n=1 \documentclass{aastex}\usepackage{amsbsy}\usepackage{amsfonts}\usepackage{amssymb}\usepackage{bm}\usepackage{mathrsfs}\usepackage{pifont}\usepackage{stmaryrd}\usepackage{textcomp}\usepackage{portland, xspace}\usepackage{amsmath, amsxtra}\pagestyle{empty}\DeclareMathSizes{10}{9}{7}{6}\begin{document} \begin{align*}- \ln ( 1 - \alpha ) = { \rm k ( T ) t} \tag{7}\end{align*} \end{document}

and for n≠1 \documentclass{aastex}\usepackage{amsbsy}\usepackage{amsfonts}\usepackage{amssymb}\usepackage{bm}\usepackage{mathrsfs}\usepackage{pifont}\usepackage{stmaryrd}\usepackage{textcomp}\usepackage{portland, xspace}\usepackage{amsmath, amsxtra}\pagestyle{empty}\DeclareMathSizes{10}{9}{7}{6}\begin{document} \begin{align*}- { \frac { \rm ( 1 - \alpha ) ^ { 1 - n } } { 1 - n } = { \rm k ( T ) t } } \tag { 8 } \end{align*} \end{document}

A plot of the left-hand side of Equations (7) or (8) as a function of t, for a constant temperature T, should produce a straight line with the rate constant k as the slope.

Equation (6) can be transformed to Equation (9) as below: \documentclass{aastex}\usepackage{amsbsy}\usepackage{amsfonts}\usepackage{amssymb}\usepackage{bm}\usepackage{mathrsfs}\usepackage{pifont}\usepackage{stmaryrd}\usepackage{textcomp}\usepackage{portland, xspace}\usepackage{amsmath, amsxtra}\pagestyle{empty}\DeclareMathSizes{10}{9}{7}{6}\begin{document} \begin{align*}\ln { \rm k ( T ) } = - { \frac { \rm E } { \rm R } } \ast { \frac { \rm I } { \rm T } } + \ln { \rm A } \tag { 9 } \end{align*} \end{document}

A plot of the left-hand side of Equation (9) as function of 1/T should produce a straight line or polygonal line (if the reaction mechanism changes with temperature) with an intercept of lnA and a slope of E/R.

Kinetic models of As

Figure 8A shows the linear regression plot of −ln(1−α) as a function of t, assuming a reaction order of 1 for As in As₂S₃ using four calcining temperatures (450°C, 500°C, 550°C, 600°C). The best fit straight lines and their correlation coefficient (R) and standard deviation values for the four temperatures based on linear regression analysis are presented in this figure. The correlation coefficient and standard deviation values show good fitting effects, meaning that As volatilization can be described by a single reaction of order 1. The slopes of these lines are the reaction rate constants (k) at these particular temperatures (T). Figure 8B shows the linear regression plot of lnk as a function of 1/T. The coefficient of linear regression is 0.9769. The rate frequency factor (A) and activation energy (E) calculated are 1.77 min⁻¹ and 26.26 KJ/mol, respectively. Therefore, the kinetic equation obtained is α=1 − exp (−1.77 exp (−3158/T)t) (10).

FIG. 8.

Linear regression plots of (A) −ln(1−α) versus t and (B) lnk versus 1/T for As in As₂S₃.

When the calcination temperature is lower than 1,000°C, volatilization of As in As₂S₃ at any temperature and at any time can be calculated by Equation (10). Figure 9A shows the volatilization of As in As₂S₃ when the temperature is above 1,000°C. In this figure, b is an enlarged view of the small black box of a; and points A, B, C, D, E, and F are the intersections of the extension lines of experimental curves [reaction (4)] and theoretical calculation curves of reaction (2) by Equation (10) at the temperatures of 1,000°C, 1,100°C, 1,200°C, 1,300°C, 1,400°C, and 1,450°C, respectively. When the temperature is above 1,000°C, reactions (2) and (3) are completed in a very short period of time (less than 0.08 min), following which reaction (4) becomes the dominant reaction. Reaction (4) plays a vital role at higher temperatures, leading to a decrease in volatilization as temperature increases.

FIG. 9.

Volatilization of As in (A) As₂S₃ above 1,000°C and (B) NaAsO₂ above 1,000°C.

NaAsO₂ is not volatile below 900°C according to the TG experiments. When the temperature is above 1,000°C the volatilization behavior of As in NaAsO₂ is similar to that of As₂S₃. The dynamic curve for the volatilization behavior of As in NaAsO₂ is provided in Figure 9B, which can be described in the same way as As₂S₃. The volatilization proportion is always decreasing as the temperature rises, which indicates that reactions (2) and (3) are completed in a very short period of time.

Kinetic models of Pb

Table 3 shows the linear regression results of −ln(1 − α) as a function of t assuming a reaction order of 1 for Pb in PbCl₂. The correlation coefficients show that the hypothesis of the reaction order of 1 is suitable.

Table 3.

Linear Regression Results of −ln(1−α) Versus t for Pb in PbCl₂

T (°C)	Fitting curve	Correlation coefficient	Standard deviation
500	y=0.00367x−4.6052	0.9915	0.0240
550	y=0.00457x−4.6091	0.9974	0.0297
600	y=0.00688x−4.6183	0.9970	0.0447
650	y=0.0089x−4.6069	0.9965	0.0568
700	y=0.01055x−4.6125	0.9982	0.0685
800	y=0.0136x−4.6357	0.9951	0.0884
900	y=0.0481x−4.7558	0.9975	0.3123
1000	y=0.09366x−4.8198	0.9980	0.6080
1200	y=0.1266x−5.0270	0.9924	0.8255
1300	y=0.0149x−5.1443	0.9894	0.9756
1450	y=0.1851x−5.3050	0.9891	1.2107

Figure 10 shows the linear regression plot of lnk as a function of 1/T of Pb in PbCl₂. The plots consist of two straight lines which are broken at 800–900°C. This means that the volatile kinetic equations are different when the temperature is below 800°C and when the temperature is above 900°C, which is shown in Fig. 10.

FIG. 10.

Linear regression plots of lnk versus 1/T for Pb in PbCl₂.

The kinetic equations for Pb in PbO and PbS were obtained in the same way and are also shown in Table 4.

Table 4.

First-Order Reaction Equations for Pb

Compounds	T (°C)	E (KJ/mol)	A (min⁻¹)	Kinetic equation
PbO	900–1450	88.73	0.017	α=1−exp(−0.017exp(−10672.5/T)t)
PbS	800–1450	78.29	27.9	α=1−exp(−27.9exp(−9416.6/T)t)
PbCl₂	500–900	31.37	0.49	α=1−exp(−0.49exp(−3772.6/T)t)
	900–1450	40.24	3.23	α=1−exp(−3.23exp(−4048.4/T)t)

Kinetic models of Cd

Kinetic equations for Cd in CdO, CdS, CdCl₂ were obtained in the same way and are summarized in Table 5.

Table 5.

First-Order Reaction Equations for Cd

Compounds	T (°C)	E (KJ/mol)	A (min⁻¹)	Kinetic equation
CdO	900–1200	61	2.764	α=1−exp(−2.764exp(−7341.9/T)t)
	1250–1450	184.6	39775	α=1−exp(−39775exp(−22205/T)t)
CdS	800–1450	129	1032	α=1−exp(−1032exp(−15518/T)t)
CdCl₂·2.5H₂O	500–1450	124	898	α=1−exp(−898exp(−14901/T)t)

Conclusions

The processing temperature and compound speciation have significant effects on the volatilization of heavy metals during co-processing in cement kilns. Na₂HAsO₄ does not volatilize, and the volatilization of As₂S₃, NaAsO₂, PbS, PbO, PbCl₂, CdO, CdS, and CdCl₂ increases with the calcination time initially. However, after 25 min they no longer volatilize. The volatilization of Pb and Cd increases with an increase in temperature. However, the volatilization of As in As₂S₃ increases until a threshold temperature around 1,000°C; following this, volatilization decreases as temperature increases. NaAsO₂ volatilizes only when the temperature is above 900°C and the volatilization of As in NaAsO₂ always decreases as the temperature increases. The particular volatilization behavior of As can be explained by three parallel reactions. The volatility of different chemical forms of Pb and Cd is in the order chlorides>sulfides>oxides. Kinetic equations of As, Pb, and Cd in different chemical forms were obtained according to first-order kinetics. According to these equations, the volatilization proportion of these heavy metals contained in HW at any time and any temperature during co-processing in cement kilns can be accurately calculated.

Footnotes

Acknowledgments

This work was supported by the Environmental Public Welfare Project of China: “Environmental risk control technology study on co-processing HWs in boilers and industrial furnaces” (201209023) and the Sino-Norwegian project phase II: “Environmentally Sound Management of Co-processing Hazardous and Industrial Wastes in Cement Kilns in China” (CHN 2150 09/059).

Author Disclosure Statement

No competing financial interests exist.

References

Adnadjevic

, Jovanovic

, and Drakulic

(2007). Isothermal kinetics of (E)-4-(4-metoxypheny I)-4-oxo-2-butenoic acid release from poly (acrylic acid) hydrogel. Thermochim. Acta., 466, 38.

Barros

A.M.

, Tenório

J.A.S.

, and Espinosa

D.C.R.

(2004). Evaluation of the incorporation ratio of ZnO, PbO and CdO into cement clinker. J. Hazard. Mater., 112, 71.

Carrasco

(1993). The evaluation of kinetic parameters from thermogravimetric data: Comparison between established methods and the general analytical equation. Thermochim. Acta., 213, 115.

Chiang

K.Y.

, Wang

K.S.

, Lin

F.L.

, and Chu

W.T.

(1997). Chloride effects on the speciation and partitioning of heavy metal during the municipal solid waste incineration process. Sci. Total Environ., 203, 129.

Coats

A.W.

, and Redfern

J.P.

(1964). Kinetic parameters from thermogravimetric data. Nature, 201, 68.

Dean

J.A.

, Ed. (1973). Lange's Handbook of Chemistry. 11th edition. New York: McGraw-Hill.

Eckert

J.O.

, and Guo

Q.Z.

(1998). Heavy metals in cement and cement kiln dust from kilns co-fired with hazardous waste-derived fuel: Application of EPA leaching and acid-digestion procedures. J. Hazard. Mater., 59, 55.

EC. (2010). Integrated Pollution Prevention and Control (IPPC), Reference document on best available techniques in the cement and lime manufacturing industries. Available at http://eippcb.jrc.ec.europa.eu

EU. (2010). Official Journal of the European Union. Directive 2010/75/EU of the European Parliament and of the Council on Industrial Emissions. Available at http://eur-lex.europa.eu

10.

Font

, Moltó

, Egea

, and Conesa

J.A.

(2011). Thermogravimetric kinetic analysis and pollutant evolution during the pyrolysis and combustion of mobile phone case. Chemosphere, 85, 516.

11.

Guo

Q.Z.

, and Eckert

J.O.

(1996). Heavy metal outputs from a cement kiln co-fired with hazardous waste fuels. J. Hazard. Mater., 51, 47.

12.

Helsen

, and Van den Bulck

(2000). Kinetics of the low-temperature pyrolysis of chromated copper arsenate-treated wood. J. Anal. Appl. Pyrolysis, 53, 51.

13.

, Wang

, Zhang

, Wang

J.J.

, and Ouyang

(2012). The industrial practice of co-processing sewage sludge in cement kiln. Proc. Environ. Sci., 16, 628.

14.

Linak

W.P.

, Wendt

J.O.L.

(1993). Toxic metal emissions from incineration: Mechanisms and control. Prog. Energy Combust. Sci., 19, 145.

15.

MEP China. Standard for pollution control on co-processing of solid wastes in Cement kiln. GB. 30485–2013.

16.

Murat

, and Sorrentino

(1996). Effect of large additions of Cd, Pb, Cr, Zn, to cement raw meal on the composition and the properties of the clinker and the cement. Cem. Concr. Res., 26, 377.

17.

Rio

, Verwilghen

, Ramaroson

, and Nzihou

(2007). Heavy metal vaporization and abatement during thermal treatment of modified wastes. J. Hazard. Mater., 148, 521.

18.

Shih

P.H.

, Chang

J.E.

, Lu

H.C.

, and Chiang

L.C.

(2005). Reuse of heavy metal-containing sludges in cement production. Cem. Concr. Res., 35, 2110.

19.

Siddique

, and Rajor

(2012). Use of cement kiln dust in cement concrete and its leachate characteristics. Resour. Conserv. Recycl., 61, 59.

20.

Trezza

M.A.

, and Scian

A.N.

(2000). Burning wastes as an industrial resource: Their effect on Pórtland cement clinker. Cem. Concr. Res., 30, 137.

21.

US EPA. (1996). Emission measurement center; Methord 29-Determination of metal emission from stationary sources, Technical support division, OAOPS, EPA. Available at www.epa.gov

22.

US EPA. (2013). National emission standards for hazardous air pollutants from hazardous waste combustors. CFR Title 40 Part 63 Subpart EEE. Available at www.epa.gov

23.

Van der Sloot

H.A.

, and Kosson

D.S.

(2012). Use of characterisation leaching tests and associated modelling tools in assessing the hazardous nature of wastes. J. Hazard. Mater., 207, 36.

24.

Wang

, Jin

Y.Y.

, and Nie

Y.E.

(2011). Characteristics of arsenic migration in MSW I fly ash during co-processing with a cement kiln. Acta Sci. Circumst., 31, 407.

25.

Yan

D.H.

, Li

, and Huang

Q.F.

(2009). Distribution of heavy metals during co-processing hazardous wastes in new dry cement kilns. China Environ. Sci., 29, 977.

26.

Yan

D.H.

, Peng

, Karstensen

K.H.

, Ding

, and Wang

(2004). Destruction of DDT wastes in two preheater/precalciner cement kilns in China. Sci. Total Environ., 476, 250.

27.

Zhang

, Liu

, Li

, Jin

, Nie

, and Li

(2009). Comparison of the fixation effects of heavy metals by cement rotary kiln co-processing and cement based solidification/stabilization. J. Hazard. Mater., 165, 1179.