Environ. Sci. Technol. 2010, 44, 7678–7684

Deactivation of Metal Chlorides by Alkaline Compounds Inhibits Formation of Chlorinated Aromatics T A K A S H I F U J I M O R I , * ,†,‡ YASUKA FUJINAGA,† AND MASAKI TAKAOKA† Department of Urban and Environmental Engineering, Graduate School of Engineering, Kyoto University, Katsura, Nisikyo-ku, 615-8540, Kyoto, Japan

Received June 16, 2010. Revised manuscript received August 19, 2010. Accepted September 1, 2010.

The inhibitory mechanisms of alkaline compounds on the formation of chlorinated aromatic (aromatic-Cl) compounds in postcombustion fly ash from thermal processes such as municipal solid waste (MSW) incineration are not fully understood. Here, we report quantitative and X-ray spectroscopic evidence that deactivation of metal chloride promoter activity by alkaline compounds inhibits the formation of aromatic-Cl compounds. The formation of aromatic-Cl compounds such as chlorobenzenes and polychlorinated biphenyls in real MSW fly ash was inhibited by the addition of NaOH, Ca(OH)2, or NaHCO3, either dry or in solution, with the fly ash. With optimal conditions, the formation of aromatic-Cl compounds was inhibited by more than 95% in comparison with formation in reheated raw MSW fly ash. We prepared simplified model fly ash samples to estimate the influence of alkaline compounds on trace Cu, Fe, Pb, and Zn chlorides, which strongly promote aromatic-Cl compound formation. More than 99% inhibition was observed in some model samples. Cl K-edge X-ray absorption and X-ray diffraction provided clear evidence of promoter deactivation, as NaOH or NaHCO3 changed to NaCl, and Ca(OH)2 changed to CaCl2 or CaClOH by reaction with the metal chlorides. NaOH was the most reactive and useful of the three alkaline compounds tested. We recommend deactivation of metal chlorides as an environmentally friendly method of inhibiting the formation of aromatic-Cl compounds, with the added benefit of changing the alkaline compounds and metal chlorides into harmless chemicals such as NaCl and metal oxides.

savings, chemical inhibitory methods using alkaline compounds have been studied since the late 1980s (5-15). Two major hypotheses regarding inhibition by alkaline compounds have been proposed. The first is neutralization of gaseous chlorine. Calcium-containing compounds such as CaO, Ca(OH)2, and CaMg(CO3)2 (dolomite) sprayed onto or mixed into real MSW or model fly ash inhibited the formation of aromatic-Cl compounds such as polychlorinated dibenzo-p-dioxins (PCDDs), furans (PCDFs), and chlorobenzenes (CBzs) as well as chlorinated aliphatic compounds after heating or combustion (5-8). Stro¨mberg attributed the inhibition of aromatic-Cl compound formation to the neutralization of gaseous chlorine by the calcium compounds (6). Indeed, the concentrations of HCl (5-8) and Cl2 (8) in the postcombustion gas phase were decreased in each experiment. Therefore, gaseous chlorine is thought to be a key factor in the generation of aromatic-Cl compounds in the gas phase. The second hypothesis involves interactions with precursors such as chlorophenols (CPs), as the generation of gas-phase aromatic-Cl compounds from precursors was inhibited by calcium-containing compounds (9-12). This was attributed to adsorbent (10) or acid-base reactions (11) of the precursors with the calcium-containing compounds. However, no acute inhibitory effects were observed with the addition of calcium oxide (13). Weber et al. suggested that the inhibitory effect was not due to the calcium concentration, but rather to the high pH of the fly ash (14). In the present study, trace metal compounds promoted the formation of PCDDs, PCDFs, polychlorinated biphenyls (PCBs), and CBzs in fly ash (15). Our recent studies have provided direct X-ray absorption fine structure (XAFS) spectroscopic evidence of direct chlorination by cupric and ferric chlorides (16-18). Although the trace metal compounds clearly promoted the formation of aromatic-Cl compounds, the mechanism of which has been described in depth, there have been no previous studies on the mechanism by which alkaline compounds inhibit the activity of trace metal compounds, with the exception of one report by Hinton and Lane (19). Here, we present quantitative and X-ray spectroscopic evidence of a mechanism of inhibition involving deactivation of metal promoters by alkaline compounds. Each of three alkaline compounds, calcium hydroxide [Ca(OH)2], sodium hydroxide (NaOH), and hydrogen carbonate (NaHCO3), was added to real and model fly ashes by three different methods, that is, mixing, dripping, or shaking. The content of aromaticCl compounds in the resulting fly ash was measured by gas chromatography/mass spectrometry (GC/MS), and the chemical form of chlorine was measured by Cl K-edge nearedge X-ray absorption fine structure (NEXAFS) spectroscopy.

Introduction

Materials and Methods

Toxic chlorinated aromatic (aromatic-Cl) compounds are generated during various thermal processes such as those in municipal solid waste (MSW) incinerators (1, 2) and iron ore sintering plants (3, 4) and are concentrated in fly ash collected in the postcombustion zone. To prevent the toxic and detrimental environmental effects of aromatic-Cl compounds, methods of inhibiting their formation are applied at thermal facilities. From the viewpoint of cost and energy

Real Fly Ash. We collected real fly ash (RFA) from an electrostatic precipitator at a MSW incinerator in Japan; the fly ash had not been sprayed with activated carbon or hydrated lime, Ca(OH)2. The approximate contents of the study-related main elements in the RFA, as determined by X-ray fluorescence spectrometry (XRF-1700; Shimadzu Corp., Kyoto, Japan), were (%): C, 8.3; Cl, 13; O, 28; Ca, 21; Na, 5.2; K, 5.2; Cu, 0.2; Fe, 1.7; Pb, 1.0; and Zn, 4.2. Using a total organic carbon (TOC) analyzer (TOC-5000; Shimadzu Corp.), the TOC in the RFA was determined to be ca. 3.2%. Concentrations of PCBs and CBzs are showed in Table 1. Model Fly Ash. A simplified model sample called model fly ash (MFA) was prepared by referencing the chemical composition of the RFA. The MFA contained KCl (10% Cl)

* Corresponding author e-mail: [email protected]. † Kyoto University. ‡ Present address: Research Center for Material Cycles and Waste Management, National Institute for Environmental Studies, 16-2 Onogawa, Tsukuba, 305-8506, Ibaraki, Japan. 7678

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 44, NO. 19, 2010

10.1021/es102055v

 2010 American Chemical Society

Published on Web 09/14/2010

TABLE 1. Concentrations of CBzs and PCBs in MSWI Real Fly Ash (RFA) after Heating at 300 °C under Various Additive Conditionsa CBzs sample name RFA RFA RFA RFA RFA RFA RFA RFA RFA RFA RFA RFA RFA RFA RFA RFA RFA RFA b

with grounding + NaOH(2%) + NaOH(5%) + NaHCO3(2%) + NaHCO3(5%) + Ca(OH)2(2%) + Ca(OH)2(5%) without grounding + NaOHaq(2%) + NaOHaq(5%) + NaHCO3aq(2%) + NaHCO3aq(5%) + NaOHsk(2%) + NaOHsk(5%) + NaHCO3sk(2%) + NaHCO3sk(5%) + Ca(OH)2sk(2%) + Ca(OH)2sk(5%)

alkaline cmpd.

additive method

additive percentage (%)

concn (ng/g)

no NaOH NaOH NaHCO3 NaHCO3 Ca(OH)2 Ca(OH)2 no NaOH NaOH NaHCO3 NaHCO3 NaOH NaOH NaHCO3 NaHCO3 Ca(OH)2 Ca(OH)2

no addition mixing mixing mixing mixing mixing mixing no addition dripping dripping dripping dripping shaking shaking shaking shaking shaking shaking

0 2 5 2 5 2 5 0 2 5 2 5 2 5 2 5 2 5

4800 3900 210 3600 2400 3500 1600 2100 940 310 1100 970 2100 1800 3100 2600 2100 1300

PCBs IEb (%)

concn (ng/g)

IEb (%)

base 19 96 25 50 27 67 base 55 85 48 54 0 14 -48 -24 0 38

260 76 14 140 46 150 39 100 21 7.7 24 13 73 50 110 85 78 48

base 71 95 46 82 42 85 base 79 92 76 87 27 50 -10 15 22 52

a Alkaline compounds, NaOH, NaHCO3, and Ca(OH)2 were added to pre-heated RFA by mixing, dripping, or shaking. Inhibitory effect defined by eq 1.

as a chlorine source, activated carbon (3.0%) as a carbon source, and SiO2 (remainder) as the base material. We selected KCl as the main chlorine source in the MFA, because it has been reported as one of the major chlorine forms in RFA (20, 21). Before use, the activated carbon was heated at 500 °C for 60 min under N2 gas at a flow rate of 100 mL/min, to remove volatile organic compounds. To examine the promotion of aromatic-Cl compound formation by trace metal chlorides, cupric chloride (CuCl2 · 2 H2O, 0.2% Cu), ferric chloride (FeCl3 · 6 H2O, 0.2% Fe), or a mixture of four metal chlorides [CuCl2 (0.2%Cu) + FeCl3 (0.5% Fe) + PbCl2 (1.0% Pb) + ZnCl2 (2.0% Zn)] was added to the MFA. The resulting samples were named MFA(Cu), MFA(Fe), or MFA(Mix), respectively. Model samples appropriate for Cl K-edge NEXAFS spectroscopic measurements were prepared. To avoid background absorption by silicon, the base material was changed from SiO2 to boron nitride (BN). KCl was not added to the MFA, as our goal was to determine the atomic-level interactions between chlorine from trace metal chlorides and the alkaline compounds. These model samples, containing activated carbon (5%) and BN (remainder) plus CuCl22 · H2O (5%), FeCl3 (5%), or a four-metal mixture [CuCl2 (5%) + FeCl3 (5%) + PbCl2 (5%) + ZnCl2 (5%)] were named MFA(Cu)X, MFA(Fe)X, or MFA(Mix)X, respectively. We also prepared model samples for X-ray diffraction (XRD) analysis. No base materials such as SiO2 or BN were used, to avoid interference with the diffraction patterns of the alkaline and metal compounds. The samples were composed of activated carbon and CuCl2 · 2H2O (1:1 on a per weight basis), FeCl3 · 6H2O (1:1), or a four-metal mixture (CuCl2 + FeCl3 + PbCl2 + ZnCl2) (1:1:1:1:1) and were designated as MFA(Cu)D, MFA(Fe)D, or MFA(Mix)D, respectively. Addition of Alkaline Compounds. We added each of three alkaline compounds, NaOH, NaHCO3, and Ca(OH)2, to the RFA and model samples using three different application methods, before performing the heating experiment described in the following section. The first application method was mixing, which involved grinding the sample along with the alkaline compound in a mortar for 10 min. The second method, called dripping, consisted of dropping an aqueous solution of the alkaline compound onto the sample. The third method was shaking and was accomplished by shaking

the sample with an alkaline compound in a 250 mL plastic bottle for 5 min. With reference to a report by Addink et al. (22), additive percentages of alkaline compounds were set at 2, 5, and 6 weight% per dry weight of sample. For the dripping method, the alkaline compounds were dropped evenly onto the samples to adjust the weight%. For the model samples used for XRD analysis, we added larger amounts of alkaline compounds, using alkaline compound:activated carbon ratios of 1:1, 1:1, 3:1, on a per weight basis, with the MFA(Cu)D, MFA(Fe)D, and MFA(Mix)D samples, respectively. As little Ca(OH)2 could be dissolved in water, we did not use Ca(OH)2 with the dripping method of addition. Quantitation of Chlorinated Aromatic Compounds in Residue after Post-Combustion Heating. Real and model samples, with or without alkaline compounds, were heated at 300 °C for 30 min under a stream of 10% oxygen and 90% nitrogen gas (flow rate, 50 mL/min), to simulate the temperature conditions associated with maximum formation of aromatic-Cl compounds in MSW-incinerator fly ash (23). To estimate the formation of aromatic-Cl compounds after the samples were heated, high-resolution gas chromatography/ low-resolution mass spectrometry (HP-6890/HP-5973) was used to analyze PCBs and CBzs in the residue; the analyses were performed in duplicate (n ) 2) for each experiment. The PCBs consisted of two condensed benzenes bonded with chlorine. The CBzs were one-benzene aromatics bonded with chlorine and precursors of PCDDs and PCDFs. The details of the heating device and pretreatment procedures were described in our previous report (15). Cl K-Edge NEXAFS. The Cl forms present in the heated samples were determined by Cl K-edge NEXAFS spectroscopy. We heated MFA(Cu)X, MFA(Fe)X, or MFA(Mix)X with an alkaline compound at 200 °C, 300 °C, and 400 °C for 30 min under 10% O2/90% N2 at 50 mL/min. The heated sample residue was sealed as quickly as possible and analyzed by Cl K-edge NEXAFS spectroscopy using BL-11B and BL-9A at the Photon Factory synchrotron radiation facility (Tsukuba, Japan). Powdered samples were mounted on carbon tape, and the NEXAFS spectra were collected in total fluorescence yield (TFY) mode under vacuum. Reference NEXAFS spectra of standard materials were also obtained. All Cl K-edge NEXAFS spectra were calibrated against the absorption intensity of KCl at 2822.8 eV. Reflecting these spectral features, VOL. 44, NO. 19, 2010 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

7679

TABLE 2. Concentrations of CBzs and PCBs in Model Fly Ash [KCl (10%Cl) + Activated Carbon (3%) + SiO2 (rest)] (MFA) after Heating at 300 °C under Various Additive Conditionsa CBzs sample name

metal compound

alkaline cmpd.

additive method

MFA MFA(Cu) MFA(Cu) + NaOH MFA(Cu) + NaHCO3 MFA(Cu) + Ca(OH)2 MFA(Cu) + NaOHaq MFA(Cu) + NaHCO3aq MFA(Fe) MFA(Fe) + NaOH MFA(Fe) + NaHCO3 MFA(Fe) + Ca(OH)2 MFA(Fe) + NaOHaq MFA(Fe) + NaHCO3aq MFA(Mix) MFA(Mix) + NaOH MFA(Mix) + NaOHaq

no CuCl2 · 2H2O CuCl2 · 2H2O CuCl2 · 2H2O CuCl2 · 2H2O CuCl2 · 2H2O CuCl2 · 2H2O FeCl3 · 6H2O FeCl3 · 6H2O FeCl3 · 6H2O FeCl3 · 6H2O FeCl3 · 6H2O FeCl3 · 6H2O CuCl2 · 2H2O + FeCl3 · 6H2O + PbCl2 + ZnCl2 CuCl2 · 2H2O + FeCl3 · 6H2O + PbCl2 + ZnCl2 CuCl2 · 2H2O + FeCl3 · 6H2O + PbCl2 + ZnCl2

no no NaOH NaHCO3 Ca(OH)2 NaOH NaHCO3 No NaOH NaHCO3 Ca(OH)2 NaOH NaHCO3 No NaOH NaOH

no addition no addition Mixing mixing mixing dripping dripping no addition mixing mixing mixing dripping dripping no addition mixing dripping

additive concn percentage (%) (ng/g) 0 0 2 2 2 2 2 0 2 2 2 2 2 0 6 6

25 15000 47 1800 300 130 620 3500 26 32 25 41 46 9800 24 780

PCBs ISb

base

concn (ng/g)

5.9 1400 SS 1.7 52 8.5 S 17 15 180 SS 11 SS 23 SS 16 SS 15 SS 11 440 SSS 6.2 26

ISb base SSS S SS S S SS S S S SS SS S

a Cupric, ferric, and four- metal chlorides were grounded with and alkaline compounds were mixed or dripped to MFA before heating. b Inhibition score noted by eq 2.

analyses were perform by linear combination fit (LCF) with reference materials of chlorine using REX 2000 ver 2.5.5 software (Rigaku Corp., Tokyo, Japan). The measurement procedures and analyses by Cl K-edge NEXAFS were explained in detail in our previous reports (17, 21). X-ray Diffraction. The trace metals in the crystal structures of MFA(Cu)D, MFA(Fe)D, and MFA(Mix)D mixed with alkaline compounds were examined by XRD (RINT-UltimaPC; Rigaku Corp.) at room temperature, after heating at 300 °C. Crystal information was identified using MDI Jade 6j software (Rigaku Corp.) and the powder diffraction file was maintained by the International Centre for Diffraction Data (ICDD).

Inhibition of Aromatic-Cl Compounds by Alkaline Compounds. The presence of alkaline compounds added to the RFA by mixing or dripping inhibited the formation of CBzs and PCBs in RFA upon heating at 300 °C (Table 1). We defined the inhibitory effect (IE) as:

shaking, suggesting that the shaking method might not have sufficiently mixed the additive compounds with the RFA. Therefore, it was concluded that sufficient mixing between the sample and alkaline compound, as with the mixing and dripping methods, was required to achieve a strong inhibitory effect. These findings indicate that the inhibition of aromaticCl compound formation resulted from sufficient contact between the alkaline compounds and the RFA, throughout the aqueous or solid sample. To determine the trigger event, simplified model fly ash samples focused on specific elements from the RFA were created and studied. Using the model samples MFA, MFA(Cu), MFA(Fe), and MFA(Mix), we found marked inhibition of trace metal chloride-promoted formation of aromatic-Cl compounds upon the addition of alkaline compounds by mixing or dripping, as seen in the experiments using RFA. Table 2 shows the concentrations of CBzs and PCBs for the different conditions and the inhibition score (IS) for each, calculated as shown:

IE(%) ) (1 - sample/base) × 100

IS ) s/b

Results and Discussion

(1)

where sample is the concentration of CBzs or PCBs in the RFA with the addition of alkaline compounds. Base is the concentration of CBzs or PCBs in the original RFA without the addition of alkaline compounds. The IE for each alkaline compound was calculated separately for each method of addition, that is, mixing by grinding, dripping, or shaking. In all additive conditions, the concentrations of CBzs and PCBs decreased with increasing percentage of alkaline compounds added, from 2 to 5%. The IEs for PCBs were nearly the same or greater than the IEs for CBzs. For example, mixing NaOH with RFA resulted in decreases in the concentrations of CBzs and PCBs, from 3900 (IE ) 19%) and 76 ng/g (71%), respectively, with a 2% addition, to 210 (96%) and 14 ng/g (95%), respectively, with a 5% addition. The IE of NaHCO3 was less than those of NaOH and Ca(OH)2 under mixing conditions. With the addition of NaHCO3 to RFA by mixing, CBzs and PCBs were inhibited e50%, with the exception of the PCB concentration with a 5% addition. Compared with mixing, addition by dripping provided slightly greater IEs. The best IEs for CBzs (85%) and PCBs (92%) were achieved by adding NaOH at 5% by dripping (sample RFA+NaOHaq(5%) in Table 1). However, poor or no inhibition was observed with each alkaline compound added by 7680

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 44, NO. 19, 2010

(2)

where s is the concentration of CBzs or PCBs in MFA(Cu), MFA(Fe), or MFA(Mix), respectively. b is the concentration of CBzs or PCBs in MFA. The IS was divided into four ranks: SSS, IS < 1; SS, 2 > IS > 1; S, 10 > IS > 2; and blank, IS > 10. We used the IS as the index to judge the level of inhibition by deactivation of metal chlorides. When IS ) 1, concentrations of CBzs and PCBs were same as that in MFA including no metal chlorides. So, near or below 1 of IS indicated that metal chlorides did not promote the formation of chlorinated aromatic compounds in solid phase, that is, deactivation. Without the addition of alkaline compounds, the CBz and PCB concentrations in MFA(Cu), MFA(Fe), and MFA(Mix) after heating were 15 000 and 1400 ng/g; 3500 and 180 ng/g; and 9800 and 440 ng/g, respectively, which are 102 or 103 times those in MFA (25 and 5.9 ng/g, respectively). This confirms the promotion of aromatic-Cl compound formation by trace metal chlorides, as reported in our previous study (15). When NaOH was added to the model samples by mixing, the best IS ranks were SS or SSS, as shown in Table 2 (+NaOH); CBzs and PCBs were strongly inhibited, by as much as 99%, except for PCBs in MFA(Fe) + NaOH. Therefore, the promotion of aromatic-Cl compound formation by trace metals was almost abolished by NaOH added by mixing.

FIGURE 1. Cl K-edge NEXAFS spectra of model fly ash admixed with cupric chloride, ferric chloride, or four-metal chlorides [MFA(Cu)X, MFA(Fe)X, or MFA(Mix)X] under various additive and heating conditions between room temperature (rt) and 400 °C. NaOH (5%) was added to the preheated (A) MFA(Cu)X, (B) MFA(Fe)X, or (C) MFA(Mix)X, respectively, by mixing or dripping [indicated by (aq)]. NaHCO3 (5%) was added to the preheated (D) MFA(Cu)X or (E) MFA(Fe)X, respectively, by mixing or dripping [indicated by (aq)]. Ca(OH)2 (5%) was added by mixing to (F) MFA(Cu)X or (G) MFA(Fe)X, respectively. (H) Cl K-edge NEXAFS spectra of representative reference compounds. Pre-edge peak (pre) around 2817-2818 eV was shown in chlorine surrounded by copper or iron. NaCl had characteristic dip (ca. 2828 eV) and postedge peak (post) at ca. 2835 eV. Dripping NaOH into the model samples, designated as +NaOHaq in Table 2, also inhibited trace metal promotion of aromatic-Cl compound formation, with an IS of the same rank or one rank lower than that with mixing conditions, except for CBzs in MFA(Mix)+NaOHaq. When Ca(OH)2 was added to the model samples by mixing, CBz and PCB formation was inhibited by 91.1-99.4% (IS e SS). The inhibitory effects of NaOH and Ca(OH)2 on metal chloridepromoted aromatic-Cl compound formation were both the same under mixing conditions. NaHCO3, added by mixing or dripping, selectively inhibited aromatic-Cl promotion by ferric chloride (FeCl3 · 6 H2O), with an IS of SS or S for MFA(Fe) + NaHCO3 or +NaHCO3aq (Table 2), but only weakly inhibited CBz and PCB formation promoted by cupric chloride (CuCl2 · 2H2O), with IS rankings of blank and S, respectively. The optimum temperature for the generation of aromatic-Cl compounds by cupric chloride was reported previously to be around 300 °C (17), whereas ferric chloride was reported to promote aromatic-Cl compound formation at temperatures higher than 300 °C (18, 24). Therefore, the difference in the scores for inhibition by NaHCO3 between cupric and ferric chlorides may reflect their thermal behavior.

We examined this difference using X-ray spectroscopic data, as described below. These quantitative results showed strong inhibitory effects by alkaline compounds, especially NaOH and Ca(OH)2, on aromatic-Cl compound formation promoted by trace metal chlorides. Given that trace metals are generally present in RFA, their promotion of aromatic-Cl compound formation activity is an important target for reducing the aromatic-Cl compounds in RFA. X-ray Spectroscopic Evidence of Metal Chloride Deactivation by NaOH, NaHCO3, and Ca(OH)2. We used Cl K-edge NEXAFS and XRD to examine why aromatic-Cl compound formation was inhibited by alkaline compounds. Figure 1 shows Cl K-edge NEXAFS spectra of model samples with addition of alkaline compound (Figures 1A-G) and the reference (Figure 1H). Without the addition of alkaline compounds, Cl was bonded with Cu in CuCl2( · 2H2O) or Fe in FeCl3( · 6H2O) in the ground model samples containing these metals (cupric or ferric chloride + activated carbon + boron nitride), as determined from the high degree of similarity in the shape of the Cl K-edge NEXAFS spectra between the samples and reference. Reference spectra of compounds with ClsCu or ClsFe bonds had characteristic VOL. 44, NO. 19, 2010 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

7681

FIGURE 3. Positive correlation between percentage of Cl bonded with aromatic and aliphatic carbon (ClsC) and (A) concentration of chlorobenzenes (CBzs) or (B) polychlorinated biphenyls (PCBs) in model samples added with three types of alkaline compounds, NaOH, NaHCO3, Ca(OH)2 after heating at 300 °C.

FIGURE 2. Percentage of Cl form of MFA(Cu)X, MFA(Fe)X, or MFA(mix) heating conditions between room temperature (rt) and 400 °C. Content explained by letters A-G corresponds with the caption of Figure 1. pre-edge peaks at ca. 2817-2818 eV on Cl K-edge NEXAFS, indicated by “pre” in Figure 1H. We identified the pre-edge peaks as being derived from ClsCu and ClsFe bonds and used this information for LCF analysis to determine the existence of Cl bonded with Cu, Fe, Pb, or Zn in the fourmetal model sample (Cu + Fe + Pb + Zn chlorides, activated carbon, and boron nitride). When alkaline compounds were added to the model samples by mixing or dripping, the Cl K-edge spectra changed markedly, reflecting a change from Cl-metal bonds to compounds with ClsNa or ClsCa bonds (Figures 1A-G). The mathematical calculations of the LCF analyses (see Figure 2) were cross-checked by correlation between the ratio of Cl bonded with carbon (ClsC) as determined by LCF and the concentrations of CBzs and PCBs at 300 °C as determined by GC/MS (Table 2). Figures 3A and 3B show the positive relationship using three types of alkali. In the case NaCl, we 7682

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 44, NO. 19, 2010

used the depth of the dip at ca. 2828 eV and the height of the postedge peak (post) at ca. 2835 eV (Supporting Information (SI) Figure S1). The data set was normalized by dividing by the respective values of reference NaCl. The calculated normalized dip and post (SI Figure S2) showed strong positive and negative correlations with the ratios of NaCl and Cl-metal bonded compounds by LCF analysis, respectively (SI Figure S3). Therefore, measuring the dip and post was an easy and useful method for determining the ratio of NaCl as well as Cl-metal analysis. We carefully identified the chemical forms of chlorine based on a combination of the shape of the Cl K-edge NEXAFS spectrum, the results of the LCF analysis, the crystal information determined by XRD, the pre-edge features of ClsCu and ClsFe compounds, and the measurements of the post and dip derived from NaCl. The Cl K-edge NEXAFS spectra of samples with NaOH added by mixing or dripping (Figure 1A-C) were nearly the same as the spectrum of NaCl (Figure 1H), and about 70% of all Cl was in the form of NaCl based on LCF analyses (Figure 2A-C). Thus, with the addition of NaOH, the Cl that was bonded with metal reacted immediately with Na, and the metal chloride was deactivated. Moreover, as shown in Figure 4A-C, the XRD patterns also revealed NaCl crystals in the samples mixed with NaOH and detected metal oxides (CuO, Cu2O, Fe2O3, PbO, or ZnO) in the samples before and after heating at 300 °C. These observations suggest that upon the addition of NaOH by mixing or dripping, the metal chlorides in the model samples were rapidly deactivated owing to the formation of metal oxides. NaCl has been reported to show almost no promotion of aromatic-Cl compound formation (21, 25), and NaCl itself is not harmful, in contrast to NaOH. Although copper and iron oxides have also been shown to promote the formation of aromatic-Cl compounds in the solid phase, the concentrations of

FIGURE 4. X-ray diffraction patterns of various model samples added by mixing of alkaline compounds. Diffraction patterns of MFA(Cu)D, MFA(Fe)D, or MFA(Mix)D added NaOH (mixing) were presented in (A), (B), or (C), respectively. (D) or (E) showed MFA(Cu)D or MFA(Fe)D added NaHCO3 (mixing), respectively. Ca(OH)2 was added by mixing to (F) MFA(Cu)D or (G) MFA(Fe)D, respectively. aromatic-Cl compounds promoted by these oxides were more than 103 times lower than those induced by their chlorides, as measured in the same model system (15). The chemical forms of metals such as copper, iron, and sodium in fly ash were changed to safe and stable forms with weak activity. Thus, the inhibition of aromatic-Cl compound formation by deactivation of metal chlorides is an environmentally friendly technique. Although NaHCO3 partly reacted with Cu or Fe before heating of the sample, LCF results showed that around 40% or 60% of all Cl was present in compounds with ClsCu or ClsFe bonds, respectively, at room temperature (rt), as shown in Figure 2D and E, respectively. The percentage of NaCl increased and that of Cl-metal decreased with increasing temperature. An increasing NaCl ratio was visualized by growth of the characteristic dip at ca. 2828 eV and postedge peak at ca. 2835 eV (post) (Figure 1D, E, and H; SI Figure S1D, E, and F). The XRD patterns indicated that NaHCO3 was present before heating, but disappeared after heating, with the concomitant increase in the NaCl peak and appearance of CuO or Fe2O3 (Figure 4D and E). Therefore, the thermal chemical reaction between NaHCO3 and metal chlorides progressed gradually. Figure 2D and E show that the lowest ratios of Cl bonded with carbon (ClsC) were observed at 300 and 400 °C in MFA(Cu)X and MFA(Fe)X mixed with NaHCO3, respectively, and in MFA(Cu)X + NaHCO3 added by dripping. The temperature dependence of the ClsC ratio gave the same patterns as those observed in our previous X-ray spectroscopic research regarding CuCl2 (17) and FeCl3 (18) without NaHCO3. As the temperature pattern originated from cupric or ferric chloride, approximately 40% of the CuCl2( · 2H2O) and 60% of the FeCl3( · 6H2O) were not deactivated by NaHCO3 at room temperature, but promoted the formation of aromatic-Cl compounds. These observations support the suggestion above, that the lower inhibitory effect of NaHCO3 added to MFA(Cu) (Table 2) was caused by differences in thermal behavior between CuCl2 and FeCl3. Weak inhibition

resulted from weak deactivation of metal chloride by NaHCO3, indicating that deactivation levels of metal chlorides by the alkali compounds relates with an effective inhibition. More than 75% of the chlorine bonded with calcium upon mixing of Ca(OH)2 with model samples containing cupric or ferric promoters at room temperature (rt) (Figures 2F and 2G). The chemical form of the compound containing ClsCa bonds was identified as CaCl2, for which the reference Cl K-edge NEXAFS spectrum is shown in Figure 1H. Meanwhile, the crystal structure determined by XRD indicated calcium chloride hydroxide (CaClOH) as the ClsCa bonded compound at rt (Figure 4F and G). As we did not measure reference Cl spectra other than CaCl2, it will be necessary to determine the reference CaClOH spectrum. However, we also performed Ca K-edge NEXAFS of MFA(Cu)X + Ca(OH)2 containing CaCl2 and CaClOH as references to support this characterization. The chemical form of Ca was identified from the sum of CaClOH and CaCl2 from rt to 400 °C, as shown in SI Figure S4. Therefore, based on three types of characterization, we conclude that when Ca(OH)2 was mixed with fly ash, it immediately reacted with and deactivated metal chlorides, forming CasCl bonded compounds such as CaCl2 or CaClOH. After heating, no changes were observed in the Cl K-edge NEXAFS spectra. However, the XRD patterns revealed CaCO3 at 300 °C under MFA(Cu)D + Ca(OH)2 conditions (Figure 4F). CasCl chemicals were thought to be partly carbonated by CO2 gas derived from the consumption of activated carbon. Metal oxides (CuO and Fe2O3) were also identified by XRD at 300 °C (Figures 4F and 4G). To inhibit aromatic-Cl compound formation in the solid phase, sufficient contact between the metal chloride and the alkali was more important than shaking. It was necessary to physically grind the barely soluble Ca(OH)2 to achieve effective inhibition. The easyto-use dripping method could also achieve sufficient contact, but it was not applied to Ca(OH)2. In the present study, the quantitative GC/MS results regarding the inhibition of aromatic-Cl compound formation VOL. 44, NO. 19, 2010 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

7683

were clarified by X-ray spectroscopic characterization using Cl K-edge NEXAFS and XRD. The inhibitory effects (or scores) were strongly linked to the deactivation activities of three alkalis, NaOH, NaHCO3, and Ca(OH)2, as the ratio of NaCl or ClsCa bound compounds. Comparison of the inhibitory effects with regard to the ease of deactivation and the flexibility of the additive method indicated that NaOH was the easiest to use and the most effective of the three alkaline compounds examined. Although we will test spraying of aqueous NaOH and various other methods of application in future studies, the mechanism of deactivation of metal promoters demonstrated in the present study is suitable as an environmentally sustainable method for inhibiting the formation of aromatic-Cl compounds in MSW fly ash.

Acknowledgments We thank Y. Tanino, K. Oshita, K. Shiota, S. Morisawa, and T. Sasaki for supporting this study; Y. Kitajima (BL-11B) and Y. Inada (BL-9A) are thanked for helping with Cl K-edge NEXAFS measurement at Photon Factory (Proposal Nos. 2007G069 and 2009G044).

Supporting Information Available Four supporting figures are available. This material is available free of charge via the Internet at http://pubs.acs.org.

Literature Cited (1) Olie, K.; Vermeulen, P. L.; Hutzinger, O. Chlorodibenzo-p-dioxins and chlorodibenzofurans are trace components of fly ash and flue gas of some municipal incinerators in The Netherlands. Chemosphere 1977, 6, 455–459. (2) Ballschmiter, K.; Braunmiller, I.; Niemczyk, R.; Swerev, M. Reaction pathways for the formation of polychloro-dibenzodioxins (PCDD) and -dibenzofurans (PCDF) in combustion processes: II. Chlorobenzenes and chlorophenols as precursors in the formation of polychloro-dibenzodioxins and -dibenzofurans in flame chemistry. Chemosphere 1988, 17, 995–1005. (3) Anderson, D.; Fisher, R. Sources of dioxins in the United Kingdom: The steel industry and other sources. Chemosphere 2002, 46, 371–381. (4) Cieplik, M. K.; Carbonell, J. P.; Mun ˜ oz, C.; Baker, S.; Kru ¨ ger, S.; Liljelind, P.; Marklund, S.; Louw, R. On dioxin formation in iron ore sintering. Environ. Sci. Technol. 2003, 37, 3323–3331. (5) Takeshita, R.; Akimoto, Y. Control of PCDD and PCDF formation in fluidized bed incinerators. Chemosphere 1989, 19, 345–352. (6) Stro¨mberg, B. Low temperature formation and minimization of chlorinated hydrocarbons. Chemosphere 1991, 23, 1515–1525. (7) Hay, D. J.; Finkelstein, A.; Klicius, R.; Bridle, T. The national incinerator testing & evaluation program air pollution control technology. Chemosphere 1987, 16, 1923–1928. (8) Gullett, B. K.; Lemieux, P. M.; Dunn, J. E. Role of combustion and sorbent parameters in prevention of polychlorinated dibenzo-p-dioxin and polychlorinated dibenzofuran formation during waste combustion. Environ. Sci. Technol. 1994, 28, 107– 118.

7684

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 44, NO. 19, 2010

(9) Naikwadi, K. P.; Karasek, F. W. Prevention of PCDD formation in MSW incinerators by inhibition of catalytic activity of fly ash produced. Chemosphere 1989, 19, 299–304. (10) Liu, W.; Zheng, M.; Zhang, B.; Qian, Y.; Ma, X.; Liu, W. Inhibition of PCDD/Fs formation from dioxin precursors by calcium oxide. Chemosphere 2005, 60, 785–790. (11) Qian, Y.; Zheng, M.; Liu, W.; Ma, X.; Zhang, B. Influence of metal oxides on PCDD/Fs formation from pentachlorophenol. Chemosphere 2005, 60, 951–958. (12) Lu, S. Y.; Chen, T.; Yan, J. H.; Li, X. D.; Ni, Y. L. M. J.; Cen, K. F. Effects of calcium-based sorbents on PCDD/F formation from pentachlorophenol combustion process. J. Hazard. Mater. 2007, 147, 663–671. (13) Lenoir, D.; Kaune, A.; Hutzinger, O.; Mu ¨ tzenich, G.; Horch, K. Influence of operating parameters and fuel type on PCDD/F emissions from a fluidized bed incinerator. Chemosphere 1991, 23, 1491–1500. (14) Weber, R.; Sakurai, T.; Hagenmaier, H. Formation and destruction of pcdd/pcdf during heat treatment of fly ash samples from fluidized bed incinerators. Chemosphere 1999, 38, 2633–2642. (15) Fujimori, T.; Takaoka, M.; Takeda, N. Influence of Cu, Fe, Pb and Zn chlorides and oxides on formation of chlorinated aromatic compounds in MSWI fly ash. Environ. Sci. Technol. 2009, 43, 8053–8059. (16) Takaoka, M.; Shiono, A.; Nishimura, K.; Yamamoto, T.; Uruga, T.; Takeda, N.; Tanaka, T.; Oshita, K.; Matsumoto, T.; Harada, H. Dynamic change of copper in fly ash during de novo synthesis of dioxins. Environ. Sci. Technol. 2005, 39, 5878–5884. (17) Fujimori, T.; Takaoka, M. Direct chlorination of carbon by copper chloride in a thermal process. Environ. Sci. Technol. 2009, 43, 2241–2246. (18) Fujimori, T.; Takaoka, M.; Morisawa, S. Chlorinated aromatic compounds in a thermal process promoted by oxychlorination of ferric chloride. Environ. Sci. Technol. 2010, 44, 1974–1979. (19) Hinton, W. S.; Lane, A. M. Effect of zinc, copper, and sodium on formation of polychlorinated dioxins on MSW incinerator fly ash. Chemosphere 1992, 25, 811–819. (20) Zhu, F.; Takaoka, M.; Shiota, K.; Oshita, K.; Kitajima, Y. Chloride chemical form in various types of fly ash. Environ. Sci. Technol. 2008, 42, 3932–3937. (21) Fujimori, T.; Tanino, Y.; Takaoka, M.; Morisawa, S. Chlorination mechanism of carbon during dioxins formation by using Cl-K near edge X-ray absorption fine structure. Bunseki Kagaku 2009, 58, 221–229 (in Japanese). (22) Addink, R.; Paulus, R. H. W. L.; Olie, K. Prevention of polychlorinated dibenzo-p-dioxins/dibenzofurans formation on municipal waste incinerator fly ash using nitrogen and sulfur compounds. Environ. Sci. Technol. 1996, 30, 2350–2354. (23) Everaert, K.; Baeyens, J. The formation and emission of dioxins in large scale thermal processes. Chemosphere 2002, 46, 439– 448. (24) Ryan, S. P.; Altwicker, E. R. Understanding the role of iron chlorides in the de novo synthesis of polychlorinated dibenzep-dioxins/dibenzofurans. Environ. Sci. Technol. 2004, 38, 1708– 1717. (25) Addink, R.; Espourteille, F.; Altwicker, E. R. Role of inorganic chlorine in the formation of polychlorinated dibenzo-p-dioxins/ dibenzofurans from residual carbon on incinerator fly ash. Environ. Sci. Technol. 1998, 32, 3356–3359.

ES102055V