Suppression of PCDD/Fs by Raw Meal in Cement Kilns

Raw meal of the cement production process could probably suppress the formation of polychlorinated dibenzo-p-dioxins and dibenzofurans (PCDD/Fs). In this study, these suppression effects were studied and compared during the co-processing of municipal solid waste (MSW) with conventional feed materials such as pure CaCO3, and pure CaO. A top suppression efficiency of 96% was achieved for a ratio of MSW to raw meal of 1:15. Moreover, NaCl and CuCl2 were added to the mix of feed material and MSW, to reveal the impact of additional chlorine and cations on PCDD/F formation. Similarly, the raw meal was mixed with the kiln dust to investigate possible suppression effects of those PCDD/Fs formed through the precursor or the de novo synthesis at a temperature of 350°C. The raw meal suppressed the formation of PCDD/Fs from kiln dust. Furthermore, the raw meal could not suppress the formation of PCDD/Fs on mixtures of kiln dust and NaCl or CuCl2, which could be associated with a potential increase in the formation effects of PCDD/Fs on the raw meal. The results can help in potential decision-making on the applicability of certain cement kilns on the co-processing of MSW containing high levels of Cl and Cu.


INTRODUCTION
The National Bureau of Statistics of China (NBSC, 2005(NBSC, , 2016) ) reports that the cement production in China has substantially amplified, from 970 Mt in 2004 to 2360 Mt in 2015.Such strong surge poses serious risks to both human health and the natural environment, given the considerable emissions of dust, and a host of organic and inorganic pollutants (Cheruiyot et al., 2016;Lee et al., 2016;Li et al., 2016a;Xu et al., 2018).During conventional cement production (Li et al., 2016b;Thuong et al., 2014) polychlorinated dibenzo-p-dioxins and dibenzofurans (PCDD/Fs) emerge at the level of 0.99 µg I-TEQ ton -1 clinker.Nevertheless, their concentration in the flue gas from cement kilns co-processing municipal solid waste (MSW) is still low compared to those from MSW incineration plants (Karstensen, 2008;Usón et al., 2013).The reasons for a lower PCDD/Fs concentration in cement kilns could be a higher processing temperature, higher turbulence and faster mass transfer in the cyclonic heat exchanger batteries and longer retention time in the process (Lohman and Seigneur, 2001).Moreover, the formation of PCDD/Fs might be suppressed under an alkaline atmosphere (Li et al., 2012) as in the cement kilns with -mainly CaCO 3 and CaO as neutral or basic raw meal (Jing et al., 2015).A high reduction efficiency of PCDD/Fs (84.8%) can be achieved when the flue gas passed through the raw mill of cement kiln due to the existence of CaCO 3 (Zhan et al., 2016a).CaO addition was also proven efficiently to enhance the reduction rate of PCDD/Fs in fly ashes during the thermal treatment at a low temperature, as well as suppress the formation of PCDD/Fs (Wang et al., 2012).However, the presence of other elements in the raw meal calls for a systematic investigation of the corresponding PCDD/Fs suppression and stimulation effects.
A large number of studies have been conducted regarding the possible suppressing effects of CaCO 3 and CaO on the formation of PCDD/Fs because of the ability of calcium containing compounds to bind to hydrochloric acid, thus avoiding a fraction of later chlorination and the consequent formation of PCDD/Fs (Jozewicz and Gullett, 1995;Shemwell et al., 2001;Lu et al., 2007).For example, PCDD/F formation was strongly suppressed by CaCO 3 during the incineration of polyvinyl chloride (PVC); also the HCl concentration in the flue gas significantly decreased (Sun et al., 2003).In addition, alkaline sodium silicate hydrate was applied for HCl removal to mitigate PCDD/Fs (Tsuyumoto et al., 2006).The reaction time and temperature largely affect the suppressing ability of CaO.The suppression efficiency of PCDD/Fs by CaO surpassed 90% in a range of 280-450°C and a reaction time of 2 h using pentachlorophenol as a model precursor (Liu et al., 2005).
MSW is normally introduced together with the conventional raw meal through a precalciner.The precalciner provides only a short retention time at the highest temperature of 850°C.Therefore, suppression efficiency should be studied in the precalciner zone, where the combustion of MSW actually occurs.Furthermore, also the effect of NaCl and CuCl 2 in MSW should be observed since MSW often contains certain amounts of NaCl and copper salts.Finally, Zhan et al. (2016b) showed that PCDD/F formation in recycled kiln dust can be suppressed when the kiln dust was exposed to the atmosphere formed by the raw meal.However, their ability to sufficiently suppress PCDD/Fs should be studied.As advised by the European IPPC Bureau (2013), kiln dust is often recycled into the first stage of the cyclone preheater.
The present study thoroughly investigates these suppression efficiencies of PCDD/Fs investigated in laboratory conditions closely replicating the conditions of a full-scale cement plant.Furthermore, these suppression efficiencies were studied when simultaneously using a Soxhlet-extracted kiln dust (SKD) and either NaCl or CuCl 2 .The results of this study should allow elucidating the issue of a potential co-processing of MSW with a high content of NaCl and CuCl 2 in cement plants.

Experimental Materials
In this study, the composition of MSW was replicated using a modeled municipal solid waste (M-MSW), thus substantially improving the homogeneity of the test samples compared to a real MSW (Fängmark et al., 1994).The waste fractions and their shares used in the preparation of the M-MSW employed in this study were given in the Table 1.Each waste fraction was thoroughly milled to a powder and then sieved to retain only the waste fraction with mesh size below 80 µm for further analysis in order to minimize the impact of a potential variation in the chemical composition.All milled waste fractions were mixed in a vertical rotary mixer for 24 h following the procedure described elsewhere (Zhan et al., 2016c).
The kiln dust and the raw meal were collected from a commercial full-scale operating cement kiln with a clinker capacity of 5000 t day -1 .The cement kiln represents a state-of-the-art dry process configuration with a preheater/ precalciner zone containing four cyclone stages (Li et al., 2015).Kiln dust was collected from a bag filter and usually is recycled into the first stage of the cyclone preheater along with the raw meal.Refuse-derived fuel (RDF) prepared  from MSW is fed into the precalciner at a constant feeding rate of 15 t h -1 .
The kiln dust was Soxhlet extracted for 48 h in toluene and termed SKD.The concentration of the remaining PCDD/Fs on the SKD was 0.03 ng g -1 .The contents of Cu and Cl in the SKD were 0.0013 and 0.011 wt.%, respectively.Likewise, the contents of Cu and Cl in the raw meal were 0.0012 and 0.033 wt.%, respectively.There was no need for the extraction of the raw meal since the initial concentration of PCDD/Fs did not exceed 0.02 ng g -1 .

Experimental Set-up
Fig. 1 illustrates the experimental set-up used to replicate the commercial cement plant co-processing MSW.The core part of the apparatus was a tubular furnace (4) composed of three consecutive 0.5 m long sections a, b, and c.Each successive section was independently heated; to 850, 600, and 350°C, respectively, and the furnace temperatures were controlled using three temperature controllers T a , T b , and T c (9).An S-type thermocouple concurrently ensures the reliability of the temperature readings.During the experiments, a temperature fluctuation lower than 5°C was observed.The conditions in Section a replicated those in the 4 th stage of the cyclonic preheater, where the actual combustion occurs.The Sections b and c replicated intermediate preheating stages.
The furnace consists of an internal quartz reactor tube (2) with 30 mm diameter and an external quartz reactor tube (5) with 45 mm diameter.Each sample was placed in a quartz boat (3) introduced into Section a. Two mass flow controllers (1) were used to generate a stable air flow of 300 mL min -1 containing 21% O 2 diluted in N 2 .The sample was then combusted, generating gases containing PCDD/Fs.The retention times in the Sections a, b and c was 9, 39 and 75 s, respectively.The flue gases generated were aspirated through a collection zone consisting of an XAD-II resin (6) and toluene solution (7), submerged into ice baths (8).The retention time in the collection zone was 112 s.Both the residual sample and the flue gas were collected for subsequent analyses.
The experiments on suppression by SKD and the activation by additional sources of chlorine used the same apparatus (Fig. 1) and the same procedures, unless mentioned otherwise.Target temperature in all three sections was now 350°C in order to model only the 1 st stage of the cyclone preheater, where the kiln dust and the raw meal are fed in the cement kilns.Each sample was placed in the Section b when the target temperature was achieved and stabilized.The oxygen content in the simulated flue gas was 6%.

Experimental Design
Table 2 lists the experiments performed in the study and their conditions.The experimental series A identified the preferred blending ratio of the M-MSW and raw meal selected to achieve the highest PCDD/F suppression efficiency (A1-A4).It also revealed the impact of CaCO 3 and CaO on suppression efficiency (A5-A6).
The second series B investigated the impact of NaCl and CuCl 2 on the formation of PCDD/Fs during MSW combustion (B1-B2), and studied their suppression efficiencies by raw meal (B3-B4).
The experiments C investigated the suppression efficiency of PCDD/Fs formed on SKD.
The experiments D specified the potential contribution of NaCl and CuCl 2 to the total mass of PCDD/Fs formed from SKD (D1-D2).Also they investigated the suppression effect of the raw meal on PCDD/Fs formed from the SKD with the high content of NaCl and CuCl 2 (the experiments D3-D4).

PCDD/Fs Analysis
The USEPA 1613 method was followed for the cleanup procedure of the PCDD/Fs samples.Identification and quantification of PCDD/Fs were performed by high-  Duration: 30 min 1.5 g M-MSW A-2 1.5 g M-MSW + 1.5 g raw meal A-3 1.5 g M-MSW + 4.5 g raw meal A-4 1.5 g M-MSW + 22.5 g raw meal A-5 1.5 g M-MSW + 1.5 g CaCO 3 5 g RDF + 5 g CaO A-6 1.5 g M-MSW + 1.5 g CaO B-1 1.  (2008): a splitless injection of 1 µl at the initial temperature of 150°C for 1 min was followed by the gradual temperature increase at 25 °C min -1 to 190°C and the final increase at 3 °C min -1 to 280°C, where the sample was kept for 20 min.The mean recoveries of the standards for PCDD/Fs ranged 55-125%, which was within the acceptable range of 25-150%.The toxic equivalents (TEQ) were calculated using the NATO/CCMS factors.

Suppression of PCDD/Fs at High Temperature MSW Co-processing and Suppression Efficiency
Fig. 2 shows the suppression efficiencies achieved.During experiment A-1, the concentration of PCDD/Fs generated was 9.12 ng g -1 and 1.17 ng I-TEQ g -1 for the I-TEQ value.During experiment A-2, M-MSW was blended with raw meal at a 1:1 ratio.Although the concentration of PCDD/Fs increased, the I-TEQ value decreased by 31%.Compared with experiment A-2, a further increase of the ratio of raw meal to M-MSW to 3:1 resulted in a consequent increase in the suppression efficiencies of PCDD/Fs to 22 and 36% for the I-TEQ value.
The largest suppression efficiency was achieved when the ratio of raw meal to M-MSW reached to 15:1.The suppression efficiency of PCDD/Fs climbed to 96% and to 99% for the I-TEQ value.That's main reason for the low concentration of PCDD/Fs in the flue gas of the cyclone preheater in cement kiln (Li et al., 2016).The concentration of PCDD/Fs formed decreased to 0.41 ng g -1 and to 0.004 ng I-TEQ g -1 for the I-TEQ value.Thus, the ratio of raw meal to M-MSW can significantly affect the suppression efficiency of PCDD/Fs.
Fig. 3 and Table 3 illustrate the congener distributions of PCDDs and PCDFs during the series A experiments.The share of PCDFs in the total PCDD/Fs was 79%, thus signifying the dominance of the de novo synthesis over the precursor formation pathways (Conesa et al., 2002).The formation mechanism was consistent with the model fly ash at a single temperature, gas flow rate and reaction time (Mubeen et al., 2017).OCDD accounts for 66% of the total mass of PCDDs, whereas 2,3,4,7,8-PeCDF, 1,2,3,4,6,7,8-HpCDF, and OCDF contributed most to the PCDFs.Converting the results to the I-TEQ values it was observed that a 2,3,4,7,8-PeCDF congener alone contributed 83% to the total I-TEQ value.
During the experiment A-2, the share of PCDFs in the total mass of PCDD/Fs enlarged to 88% and 93% in I-TEQ units.OCDD was the largest single contributor for of PCDDs, accounting for 46% whereas 1,2,3,4,6,7,8-HpCDF contributed 20% to PCDFs.The weight average level of chlorination of PCDD/Fs decreased slightly from 6.7 to 6.6.The results expressed showed that PCDFs contributed to the total toxicity, whereas 2,3,4,7,8-PeCDF was the largest single contributor accounting for 57%.
Unlike the domination of the congener distribution of PCDD/Fs by PCDFs in the experiments A-2 and A-3, PCDDs accounted for the largest share of PCDD/Fs in the experiment A-4, suggesting elimination of the de novo synthesis with the addition of the raw meal (Wikström et al., 2003).The dominant PCDDs congener was OCDD, accounting for 99%.At the same time, the domination of PCDFs shifted from 1,2,3,4,6,7,8-HpCDF to TCDF.The weight average level of chlorination of PCDD/Fs increased from 6.7 to 7.9, indicating a weakened dechlorination effect of PCDD/Fs in the presence of the raw meal ( Zhan et al., 2016a).In I-TEQ units, PCDFs contributed 64% of total toxicity, with 2,3,4,7,8-PeCDF was the largest single contributor, accounting for 39%.
These findings are consistent with 7 results obtained by Tagashira et al. (1999) who found that the higher chlorinated PCDD/Fs are more easily suppressed by Ca-containing compounds.In the I-TEQ units, PCDFs contributed 94%-95% to the total toxicity and 2,3,4,7,8-PeCDF was the largest single contributor accounting for 54%-57% of the total I-TEQ value in both experiments.Also, it was found that CaCO 3 and CaO were not the only functional components of the raw meal for the PCDD/Fs suppression, but also other components, such as CuO, Fe 2 O 3 etc., had an impact (Yasuhara et al., 2002).Moreover, high fraction of CuO in the ashes may lead to relatively high PCDD/F concentrations in the fly ashes (Tuan et al., 2012).

Additional Chlorine and Suppression Efficiency
NaCl was added during experiment B-1, yet the concentration found was almost as in experiment A-1, when only M-MSW was combusted.The addition of NaCl did not contribute to the PCDD/F formation at the combustion temperature of 850°C since the major chlorine source for the PCDD/F formation -Cl 2 -originates from hydrochloric acid at 300-600°C via the Deacon reaction (Chin, 2012).
Experiment B-2 presents a different situation, when CuCl 2 served as a source of additional chlorine.On one hand, CuCl 2 has been proven to be the key metal catalyst for the PCDD/F formation (Luijk et al., 1994).However, CuCl 2 also acts as an oxidizer, promoting the combustion efficiency, which in turn results in a lower PCDD/F formation (Hatanaka et al., 2001).In the experiment, the concentration of PCDD/Fs reduced to 4.78 ng g -1 or 0.82 ng I-TEQ g -1 with the addition of CuCl 2 , which was in line with the findings of Hatanaka et al. (2004).When the raw meal was used in the experiment B-3 for the PCDD/F formation, the concentration of PCDD/Fs decreased from 9.18 ng g -1 or 1.65 ng I-TEQ g -1 to 5.85 ng g -1 or 0.98 ng I-TEQ g -1 ; the corresponding suppression efficiencies were 36.3% and 40.7%.Regarding the addition of CuCl 2 , no suppression effect was found.The additional chlorine could have been absorbed by the raw meal to form stable CaCl 2 , which affected the suppression ability of the raw meal.As it is showed, the emissions of PCDD/Fs from the full-scale cement kilns are expected to increase with the increasing contents of NaCl or CuCl 2 in MSW resulting in the need for a strict control of the NaCl and CuCl 2 contents in the MSW co-processed.
Fig. 4 and Table 4 illustrate the congener distribution in the series of experiments B. During the experiment B-1, the fraction of PCDDs was 45%.A larger number of precursors were detected due to the higher initial concentration of chlorine in the raw meal, which further formed TCDD and OCDD.Similarly, the fraction of TCDF increased from 1% to 25%, thus indicating generation of lower chlorinated PCDD/Fs with the addition of NaCl.Moreover, the most pronounced contributors to the I-TEQ values also changed.Specifically, the contribution of PCDFs to the I-TEQ value was only 37% and TCDD was the largest single contributor accounting for 56% of the total PCDDs.During the experiment B-3, the fraction of PCDDs further increased to 70%, while the congener distribution was similar with the experiment without raw meal added.Moreover, the addition of the raw meal did not affect the contributors to the I-TEQ value since TCDD contributed 46% to the I-TEQ value.A comparison between the congener distributions during the experiments with the addition of the M-MSW, NaCl, and CuCl 2 showed the same influence of the raw meal on the PCDD/F formation.

Suppression of PCDD/Fs at Low Temperature
Formation and Suppression of PCDD/Fs Fig. 5 shows the concentration and I-TEQ value of PCDD/Fs formed during the series of experiments C and D. During the experiment C-1, the concentration of PCDD/Fs formed from the SKD at 350°C was 0.16 ng g -1 or 0.02 ng I-TEQ g -1 .During the experiment C-2, the concentration of PCDD/Fs formed from the raw meal at 350°C was 0.09 pg g -1 or 0.006 ng I-TEQ g -1 .The results of the experiment C-3 showed that the raw meal suppressed the emissions of PCDD/Fs formed from the SKD to 0.089 ng g -1 or 0.008 ng I-TEQ g -1 , thus indicating the suppression efficiency of 46 or 61% for the I-TEQ value.The suppression efficiency of 61% corresponds to the suppression efficiencies of other common suppressants, such as urea, SO 2 , NH 3 , amongst others (Raghunathan and Gullett, 1996;Pandelova et al., 2006).Therefore, the present study shows that the suppression of PCDD/Fs could be expected not only under the high-temperature conditions corresponding to the last stage of the cyclone preheater, but also when the process temperature is low, i.e., the first stage of the cyclone preheater.
Fig. 6 and Table 5 show the congener profiles of PCDD/Fs during the series of experiments C.During the experiment C-1, PCDFs contributed 86% to the total mass of PCDD/Fs, thus indicating a pronounced formation of PCDD/Fs via the de novo synthesis.PCDFs were dominated by TCDF, which contributed 59%.PCDDs were dominated by 1,2,3,4,6,7,8-HpCDD and OCDD, which accounted for 23 and 43%, respectively.
During the experiment C-2, more PCDDs were generated when comparing to PCDFs, thus showing the prevalence of the PCDD/F formation via the precursor synthesis.Precursor synthesis is supported by fast condensation of chlorophenols (CP) to PCDDs and PCDFs congeners, or, following a longer route, starting from chlorobenzenes (CBzs) or other suitable structures such as Polycyclic Aromatic Hydrocarbons (PAHs).The leading PCDD congener was OCDD accounting for 85% of the total mass of PCDDs, which was much different from the result of the experiment C-1.The PCDD/Fs emissions during the experiment C-3 were dominated by OCDD, which contributed 80% to the total PCDDs.Furthermore, 2,3,7,8-TCDD and 2,3,4,7,8-PeCDF contributed 69% to the total I-TEQ value of the SKD, which was similar to the experiments from the series A.

Additional Chlorine on Suppression Efficiency
Fig. 5 shows that the addition of NaCl in the experiment D-1 did not have a pronounced effect on the PCDD/F formation from the SKD.However, during the experiment D-2, where CuCl 2 was added, a significant increase in the PCDD/Fs concentration from 0.16 ng g -1 or 0.02 ng I-TEQ g - 1 in the experiment C-1 to 3.98 ng g -1 or 0.21 ng I-TEQ g -1 was observed.The reason for such a controversial behavior could be attributed to the fact that NaCl only acted as an additional sink of chlorine, whereas CuCl 2 also acted as a catalyst (Zhan et al., 2016d).Furthermore, the addition of CuCl 2 at the low temperature resulted in a controversial effect when compared to the addition of CuCl 2 at the high temperature as shown in the experiment B-2.
Analyzing the results of the experiment D-3, where the raw meal was added to suppress the PCDD/F formation, it was noticed that the concentration of PCDD/Fs formed actually increased from the initial 0.16 ng g -1 or 0.02 ng I-TEQ g -1 to 0.55 ng g -1 or 0.03 ng I-TEQ g -1 under the presence of NaCl.Furthermore, an even more substantial increase of the initial concentration to 36.9 ng g -1 or 1.79 ng I-TEQ g -1 was recorded in the experiment D-4, when CuCl 2 served as a source of additional chlorine.Such a dramatic increase in the PCDD/Fs concentration under the presence of CuCl 2 could be explained by the catalytic activity of Cu on the PCDD/F formation (Addink et al., 1996;Kuzuhara et al., 2005).Therefore, the increasing contents of NaCl or CuCl 2 in the real MSW might escalate the emissions of PCDD/Fs in the full-scale cement kilns.Furthermore, additional chlorine is expected to adhere to the surface of cement kilns, thus potentially leading to more often shutdowns of cement kiln required for the kilns maintenance (Sutou et al., 1999).Fig. 6 and Table 5 show the congener profiles of PCDD/Fs during the series of experiments D. During the experiment D-1, the fraction of PCDFs was 53%, thus showing a greater formation of PCDD/Fs via the de novo synthesis.PCDDs were dominated by OCDD, whereas PCDFs were dominated by TCDF, 1,2,3,4,6,7,8-HpCDF, and 1,2,3,4,7,8,9-HpCDF.Moreover, the weight average level of chlorination of PCDD/Fs slightly increased from 6.4 to 7.1, thus suggesting the enhancement of the PCDD/Fs chlorination with the addition of NaCl.In the I-TEQ units, the contribution of PCDFs was 45%, while the most dominant contributors were TCDD (20%) and 1,2,3,7,8-PeCDD (26%).During the experiment D-3, the fraction of PCDDs significantly increased to 71% since the precursor synthesis was enhanced when the raw meal was mixed with NaCl.The increased fraction of OCDD and the weight average level of chlorination of PCDD/Fs also suggest the presence of a reaction between the raw meal and NaCl.In the I-TEQ units, the contribution of PCDFs was 66%, whereas 2,3,4,7,8-PeCDF was the largest single contributor the I-TEQ value accounting for 36%.
During the experiment D-2, more PCDFs was formed compared to PCDDs since the de novo synthesis was enhanced under the presence of CuCl 2 .Moreover, the fraction of higher chlorinated PCDD/Fs was significantly increased.The cumulative contribution of 1,2,3,4,6,7,8-HpCDD and OCDD to the total PCDDs was 83%, whereas the contribution of 1,2,3,4,6,7,8-HpCDF to the total PCDFs was 38%.Thus, the weight average level of chlorination of PCDD/Fs increased from the initial 5.2 in the experiment C-1 to 6.8.In the I-TEQ units, the contribution of PCDFs was 86% and 2,3,4,7,8-PeCDF was the largest single contributor to the I-TEQ value accounting for 33%.The trend revealed in this study was consistent with the experiments on the PCDD/F formation with the addition of CuCl 2 performed elsewhere (Huang and Buekens, 1995).During the experiment D-4, the contribution of PCDFs to the total PCDD/Fs reached 85% since the de novo synthesis was similarly promoted under the presence of CuCl 2 .The fraction of higher chlorinated PCDD/Fs increased further.1,2,3,4,6,7,8-HpCDD and OCDD accounted for 93% of the total content of PCDDs, whereas 1,2,3,4,6,7,8-HpCDF accounted for 43% of the total mass of PCDFs.In the I-TEQ units, the contribution of PCDFs was 95%, and 2,3,4,7,8-PeCDF was the largest single contributor to the I-TEQ value accounting for 26%.

CONCLUSIONS
The results of the present study on the suppression efficiencies of PCDD/Fs formed during experiments on MSW co-processing in cement kilns showed that: 1.The highest suppression efficiency of 96% was achieved when the ratio of the MSW to the raw meal was 1:15, whereas almost no suppression was achieved when equal masses of MSW and the raw meal were co-processed, indicating the importance of the share of the MSW coprocessed; 2. The suppression efficiency might decline with addition of either NaCl or CuCl 2 , indicating the need for a strict control of the NaCl or CuCl 2 content in the MSW being co-processed; 3. The raw meal suppressees the de novo synthesis of PCDD/Fs, resulting in an excessive generation of PCDDs.Furthermore, the results of the present study showed that the suppression of PCDD/Fs formed during the kiln dust recycling into the first stage of the cyclone preheater of the cement plant could also be expected.At the same time, the content of chlorine in both the MSW co-processed and the kiln dust recycled should be controlled.

Fig. 2 .
Fig. 2. The suppression efficiency of PCDD/Fs during the series of experiments A and B in relation to the initial PCDD/Fs concentration.

Fig. 3 .
Fig. 3. Congener distribution of PCDD/Fs during the series of experiments A.

Fig. 4 .
Fig. 4. Congener distribution of PCDD/Fs during the series of experiments B.

Fig. 5 .
Fig. 5.The concentration of PCDD/Fs during the series of experiments C and D.

Table 1 .
The composition of the M-MSW, wt.%.

Table 3 .
Level and distribution of PCDD/Fs in experiments A.

Table 4 .
Level and distribution of PCDD/Fs in experiments A.