IV REUNIÓN NACIONAL DE
DIOXINAS, FURANOS Y
COMPUESTOS ORGÁNICOS
PERSISTENTES RELACIONADOS
http://web.ua.es/dioxinas
Copyright © 2013
Universidad de Alicante
IV REUNIÓN NACIONAL DE DIOXINAS, FURANOS Y COMPUESTOS ORGÁNICOS PERSISTENTES
RELACIONADOS
Edición:
Juan A. Conesa
Ignacio Aracil
Departamento de Ingeniería Química
Universidad de Alicante
Ap. 99 E-03080 Alicante
Diseño de la portada: Mª Francisca Gómez-Rico
Impresión y encuadernación:
Imprenta Universidad de Alicante
Depósito Legal: A 286-2013
P13
DE NOVO SYNTHESIS OF BROMINATED DIOXINS AND FURANS BY
THERMOGRAVIMETRY
Ortuño N, Conesa JA, Molto J, Font R
Chemical Engineering Department, University of Alicante, P.O. Box 99, E-03080 Alicante, Spain
e-mail: [email protected]
Introduction
Several authors found that the residual carbon and high molecular weight compounds from
incomplete combustion are the precursor material for this so-called de novo synthesis. The
formation of organohalogen compounds from fly ash of municipal waste incinerators and
sintering processes has been recognized as an important pathway for the release of
polychlorinated dibenzo-p-dioxins (PCDD), furans (PCDF) and some other chlorinated
1-11
compounds .
The formation of brominated organocompounds has also been observed in the low temperature
decomposition of carbonaceous materials, but only few studies investigated the de novo
synthesis of PBDDs/PBDFs and PXDDs/PXDFs in the presence of bromide. In such a way,
12
Heinbuch and Stiegliz substituted chloride in a fly ash by bromide without changing the
catalytic properties of the fly ash, and found formation of a variety of aromatic brominated
compounds during the thermal treatment of the ash. Among the organobrominated species, the
authors found bromo-benzenes, -benzonitriles, -thiophenes, -naphthalenes, and -dibenzofurans.
The formation of organobrominated compounds begins at 250 °C, with the maximum production
at ca. 350 ºC, and during a very long time (up to 120 min.). The formation of brominated
compounds increases with the amount of added bromide. Above 550 °C, nearly quantitative
destruction occurs for all compound classes.
This phenomenon has been observed in different studies. The presence of chlorine and
bromine in de novo synthesis experiments with model fly ash results in the formation of mixed
13
chlorinated–brominated PXDDs/ PXDFs . The ratio of bromine and chlorine substitution was
shown to depend on the temperature, being the bromine substitution less active at higher
temperatures and/or higher residence time. The isomer patterns of chlorinated, brominated–
chlorinated and brominated PXDDs and PXDFs in the de novo experiments were similar,
suggesting that the substitution mechanism of bromine and chlorine during de novo synthesis is
14
also similar .
12
In a closer look at the pattern of the dibromobenzenes, Heinburg and Stieglitz suggested a
two-stage mechanism of bromination of the residual carbon on the fly ash surface. The first step
is a bromination of the carbon surface yielding side by side orientation. The second step is the
oxidative decomposition of the carbon particle separating side by side substituted species.
1
Stieglitz et al. also found that, in the presence of bromine and chlorine, a mix of chloro-bromo
species is formed.
8
In a previous paper de-novo formation of chlorinated dioxins and furans was studied, both in
thermobalance and in a horizontal laboratory furnace, using model mixtures. In a similar way,
the present work investigates the de novo synthesis of brominated compounds. The following
aspects are considered:
1.
Effect of CuBr2 in the combustion of carbon, at different loads.
2.
Detailed thermogravimetric kinetic study of such processes. Especial interest has been
taken in the reproducibility of the runs, and on the fitting of the runs performed at various
conditions at the same time, as suggested in the specialized literature.
Materials and Methods
The material employed was an activated carbon (100-200 mesh) made by Euroglas analytical
Instruments with very low halide content; this activated carbon is used for the analysis of AOX
(adsorbable organic halides), so it is expected to have a little amount of chloride.
CuBr2 was mixed in different proportions and conditions with the activated carbon. A total of four
samples were studied:
Activated carbon (Sample C).
Activated carbon + a very little amount of CuBr 2 (approx. 0.15 %), with no further
treatment (Sample C+traceB).
Activated carbon + 1 % CuBr2, with no further treatment (Sample C+1B).
Activated carbon + 50 % CuBr2, with no further treatment (Sample C+50B).
Activated carbon + 50 % CuBr2, that later is pyrolyzed at 700 ºC for 17 min (1000 s),
2+
and washed for removal of CuBr2 excess. The operation is stopped when no Cu is detected in
the washing solution. This will be named sample ‘py+wash(C+50B)’.
Pure CuBr2 and CuCl2 were also decomposed in the thermobalance for comparing the different
behaviors. The thermogravimetric study was performed in a thermobalance, at various heating
rates (5, 10 and 20 K/min) and using sample amounts close to 8 mg. Twelve duplicated runs
have been performed in this equipment for the present study.
Runs were performed in a Mettler Toledo thermobalance (model TGA/SDTA851e/LF/1600) with
a horizontal furnace and a parallel-guided balance. The atmosphere used was synthetic air. Gas
-1
flowed at 100 mL min (STP), according to the specifications of the equipment. The sample
temperature was measured with a thermocouple directly at the crucible, i.e., next to the sample.
The reproducibility of the experiments was very good (temperature deviations < 0.1 %) and the
experimental data presented in this paper corresponding to the different conditions are the
mean values of the runs carried out.
Results and Discussion
The effect of copper and bromine on the combustion
Figure 1 presents an outline of the combustion runs performed at 10 K/min. Furthermore, some
15-17
samples were treated at three heating rates, in order to perform a detailed kinetic analysis
.
Figure 2 presents these combustion runs at 5, 10 and 20 K/min. The residue formed in each run
performed is almost the same (approx. 8.8 %, which would be the ash content of the carbon).
Figure 1. Combustion runs
performed at 10 K/min with
the following samples:
activated carbon ‘C’,
activated carbon with trace
amounts of CuBr2
‘C+traceB’), activated
carbon with 1 % CuBr2
‘C+1B’, activated carbon
with 50 % CuBr2 ‘C+50B’,
activated carbon + 50 %
CuBr2 pyrolyzed and
washed ‘py+wash(C+50B)’,
cupper chloride CuCl2 and
cupper bromine CuBr2.
The
maximum
weight loss rate,
commonly used
in literature to
characterize char
reactivity, occurs
at
different
temperatures for
the
different
samples. Table 1
presents
the
temperatures of
maximum
decomposition
rate (Tmax) for the
different
samples, as well
as
those
Table 1.Comparison between temperatures of maximum decomposition
rate of activated carbon in the presence of different cupper halides.
RUNS PERFORMED (10 K/min) and T max (ºC)
Sample
Material
Tmax Tmax
C
Activated carbon (C)
655
-
-
C + 50 % CuCl2 pyrolyzed* and washed
485
-170
-
C + 0,15 % CuCl2
604
-51
C + 50 % CuBr2
371
-284
508
-147
535
-120
C+50B
py+wash(C+50B) C + 50 % CuBr2 pyrolyzed* and washed
C+1B
C + 1 % CuBr2
C+traceB
C + 0.15 % CuBr2
577
-78
* Prepared in a horizontal furnace at 700 °C for 17 min in N2.atmosphere.
8
obtained in a previous work when studying chlorinated dioxins production by addition of CuCl2 .
Kinetic Study
As pointed out previously, a kinetic study of the combustion of the different model mixtures
prepared was performed. As already commented, in some cases the samples were washed with
an acid solution of nitric acid/nitrate until there was no detection of Cu in the washing fluid, in
order to eliminate the excess of CuBr2 before the thermogravimetric analysis. Note that the
previous pyrolytic decomposition took place separately from the combustion considered in this
section.
Figure 2 presents the
combustion runs
performed at three
different heating rates.
The kinetic model
chosen for the
analysis was a single
reaction of first order.
Figure 2. Combustion
runs performed at 5,
10 and 20 K/min with
the following samples:
‘C’, ‘C+50B’ and
‘py+wash(C+50B)’.
The results of the optimization are presented in Table 2. The first part of this table represents
the better fit obtained for first order reactions. In the table, the values of the variation coefficients
(V.C. (%)) are presented, which are quite low in all cases, bearing in mind that all three curves
obtained at different heating rates are considered.
For the fitting of the curves obtained with the sample C+50B (i.e. activated carbon with a 50 %
of CuBr2), only the decomposition of the activated carbon is considered (temperature range 265
– 440 ºC), by comparison with the curves obtained for CuBr2 alone (Figure 1).
Better fits are obtained when allowing for varying reaction orders, having a minimum in the O.F.
for an order close to 0.9. The values of k 0 and E, although correlation parameters, could be
related, respectively, to the amount of active sites in the activated carbon and to the energy of
the bond formed. The values of the pre-exponential factor indicate that the active sites of the
activated carbon are progressively occupied by bromine as the amount of this specie in the
sample increases. In previous work analyzing the effect of CuCl2 we found similar pre6
-1
exponential factors for all samples (in the order of 10 min ), indicating that the active sites were
not occupied by chlorine. The values of the activation energy clearly indicate that the C-H bond
(sample C) is quite stronger than the C-Br bond (sample C+50B). The effect of the pyrolysis and
washing of C+50B sample is to eliminate some Br excess and then to reduce the amount of CBr bonds; in this way, an intermediate value of activation energy is found.
Table 2.Values of kinetic parameters from TG curves, and V.C. obtained.
-1
Sample
K0 (min )
E (KJ/mol)
V.C. (%)
6
C
3.31·10
158.4
1.1
3
C+50B
8.90·10
78.7
1.0
5
Py+wash(C+50B)
2.47·10
119.4
1.9
Acknowledgements
Support for this work was provided by the Generalitat Valenciana (Spain) with projects
PROMETEO/2009/043/FEDER, and by the Spanish MCT CTQ2008-05520.
References
1. Stieglitz L, Zwick G, Beck J, Bautz H, Roth W. Chemosphere,1989; 19 (1-6): 283-290.
2. Schüler D, Jager J. Chemosphere,2004; 54 (1): 49-59.
3. Ni M, Xiao H, Chi Y, Yan J, Buekens A, Jin Y, Lu S. Waste Management,2012; 32 (3): 568574.
4. Du B, Zheng M, Tian H, Liu A, Huang Y, Li L, Ba T, Li N, Ren Y, Li Y, Dong S, Su G.
Chemosphere,2010; 80 (10): 1227-1233.
5. Weber R, Kuch B. Environment International,2003; 29 (6): 699-710.
6. Yu X, Zennegg M, Engwall M, Rotander A, Larsson M, Wong Ming H, Weber R.
Organohalogen Compounds,2008; 70: 813-816.
7. Xiao H-x, Ma Z-y, Huang Q-x, Jin Y-q, Chi Y, Yan J-h, Ni M-j In Combustion of electronic
wastes using a drop tube furnace, Electric Technology and Civil Engineering (ICETCE), 2011
International Conference on, 22-24 April 2011; 2011; pp 2988-2991.
8. Conesa J A, Fullana A, Font R. Environmental Science and Technology,2002; 36: 263-269.
9. Conesa J A. Fuel,2001; 80 (14): 2123-2124.
10. Caballero J A, Font R, Marcilla A, Conesa J A. Industrial & Engineering Chemistry
Research,1995; 34 (3): 806-812.
11. Fullana A, Font R, Conesa J A, Blasco P. Environmental Science & Technology,2000; 34
(11): 2092-2099.
12. Heinbuch D, Stieglitz L. Chemosphere,1993; 27 (1–3): 317-324.
13. Yasuhara A, Ito H, Morita M. Environmental Science & Technology,1987; 21 (10): 971-979.
14. Gullett B K, Bruce K R, Beach L O. Chemosphere,1990; 20 (10–12): 1945-1952.
15. Fermo P, Cariati F, Santacesaria S, Bruni S, Lasagni M, Tettamanti M, Collina E, Pitea D.
Environmental Science & Technology,2000; 34 (20): 4370-4375.
16. Gullett B K, Bruce K R, Beach L O In International Conference on Municipal Waste
Combustion, Hollywood, FL, April 11 - 14, 1989.
17. Froese K L, Hutzinger O. Chemosphere,1994; 28 (11): 1977-1987.