Hazardous air pollutants formation from reactions of raw meal organics in cement kilns

Article abstract of DOI:10.1016/S0045-6535(00)00222-8

Thermally induced chlorination, condensation, and formation reactions of raw meal organic surrogates were investigated on different types of surfaces. The System for Thermal Diagnostic Studies provided a powerful tool to study these reactions under defined reaction conditions, which were related to typical conditions in the preheater zone of cement kiln. Experiments were conducted with benzene and benzene/myristic acid (C₆H₆/C₁₃H₂₇COOH) mixtures in a quartz reactor containing different chlorinating catalysts/reagents over a temperature range of 300-500°C. Reaction products were trapped in-line and analyzed by GC-MS. A mixture of chlorides of calcium, potassium, aluminium and iron was highly effective for chlorination/condensation reactions of benzene and benzene/myristic acid mix at temperatures above 300°C. The same behavior was observed only when calcium chloride and potassium chloride were used as chlorinating catalyst/reagent. This result showed that transition metal chlorides like FeCl₃ are not necessary for chlorination/condensation of organics under post-combustion conditions. Methylene chloride was the major chlorinated product followed by chloroform and various other C₁, C₂ and C₆ chlorinated products. Yields of chlorinated aliphatics were highest at 400°C for both benzene and benzene/myristic acid mix. C₆ products were mainly mono- to hexa-chlorinated benzenes with trace amounts of chlorinated phenols. The major chlorinated products observed in this study (i.e., methylene chloride, chloroform, chloroethanes and monochlorobenzene) were also present as major chlorinated hydrocarbons in the cement kiln field emission data.

Full text of DOI:10.1016/S0045-6535(00)00222-8

Chemosphere 42 (2001) 499±506
Hazardous air pollutants formation from reactions of raw meal
organics in cement kilns
a,
a
a
b
Sukh Sidhu *, Nabil Kasti , Phil Edwards , Barry Dellinger
a
Environmental Science and Engineering, University of Dayton Research Institute, Dayton, OH 45469-0132, USA
b
Department of Chemistry, Louisiana State University, Baton Rouge, LA 70803, USA
Abstract
Thermally induced chlorination, condensation, and formation reactions of raw meal organic surrogates were in-
vestigated on di€erent types of surfaces. The System for Thermal Diagnostic Studies provided a powerful tool to study
these reactions under de®ned reaction conditions, which were related to typical conditions in the preheater zone of
cement kiln. Experiments were conducted with benzene and benzene/myristic acid (C6H6/C13H27COOH) mixtures in a
quartz reactor containing di€erent chlorinating catalysts/reagents over a temperature range of 300±500°C. Reaction
products were trapped in-line and analyzed by GC±MS. A mixture of chlorides of calcium, potassium, aluminium and
iron was highly e€ective for chlorination/condensation reactions of benzene and benzene/myristic acid mix at tem-
peratures above 300°C. The same behavior was observed only when calcium chloride and potassium chloride were used
as chlorinating catalyst/reagent. This result showed that transition metal chlorides like FeCl3 are not necessary for
chlorination/condensation of organics under post-combustion conditions. Methylene chloride was the major chlori-
nated product followed by chloroform and various other C1, C2 and C6 chlorinated products. Yields of chlorinated
aliphatics were highest at 400°C for both benzene and benzene/myristic acid mix. C6 products were mainly mono- to
hexa-chlorinated benzenes with trace amounts of chlorinated phenols. The major chlorinated products observed in this
study (i.e., methylene chloride, chloroform, chloroethanes and monochlorobenzene) were also present as major chlo-
rinated hydrocarbons in the cement kiln ®eld emission data. Ó 2001 Elsevier Science Ltd. All rights reserved.
Keywords: Air toxics; HAPs; Cement kiln emissions; Humic substances
1. Introduction
world-wide infrastructure, and thus it is important to
devise ecient, yet cost-e€ective, methods for the pre-
vention of HAPs.
Portland cement plants have been identi®ed by the
USEPA as major sources of hazardous air pollutantsÕ
(HAP) emissions, including hydrocarbons, hydrochloric
acid (HCl), and polychlorinated dibenzo-p-dioxins and
polychlorinated furans (PCDD/F). As such they have
been targeted for additional regulation under the aus-
pices of the Clean Air Act Amendment (CAAA) and
EPAÕs Combustion Strategy. A€ordable cement is ob-
viously essential for the growth of both domestic and
To devise cost-e€ective methods for prevention of
HAPs, their formation mechanism must be understood;
this mechanism is the focus of this contribution. For
HAP formation to take place in any cement kiln zone, it
requires a chlorine source and an organic source. Basic
combustion theory suggests that cement kilns provide
sucient reaction time, oxygen concentration, and high
temperatures to destroy all the organics present in the
fuel feed with destruction and removal eciency greater
than 99.99%. An improper mixing of fuel and oxygen or
a ``failure'' in any of the above-described parameters can
potentially result in poor combustion and can thus lead
* Corresponding author. Fax: +1-937-229-2503.
E-mail address: sidhu@udri.udayton.edu (S. Sidhu).
0045-6535/01/$ - see front matter Ó 2001 Elsevier Science Ltd. All rights reserved.
PII: S 0 0 4 5 - 6 5 3 5 ( 0 0 ) 0 0 2 2 2 - 8

500
S. Sidhu et al. / Chemosphere 42 (2001) 499±506
to pollutant emissions. However, ®eld studies conducted
at full-scale kilns show that total hydrocarbon emissions
are not very sensitive to the changes in combustion
conditions (Trenholm et al., 1990; Dellinger et al., 1993).
Based on these results, it was hypothesized that organics
present in the raw materials (e.g., clay, shale, gypsum,
etc.) are the major sources of organic emissions from
cement kilns not burning hazardous waste. In our pre-
vious study, we veri®ed this hypothesis by collecting
eight raw meal samples from di€erent process points
(before and after each cyclone, bag house, grinding mill,
and silo) and thermally desorbing these samples in inert
atmosphere (helium) to determine their total organic
content (Sidhu and Dellinger, 1995). The results of these
experiments showed that most of the organics were de-
sorbed between temperatures of 250±500°C in ®rst two
preheating cyclones.
Chlorine required for HAP formation can be present
in the preheater zone as organic chlorides (RCl), HCl,
Cl2, alkali and alkaline earth chlorides (AClx), and
transition metal chlorides (MeClx). Any chlorine in the
coal or oil fuel enters the preheater zone as HCl and Cl2
(25:1) and RCl is only expected to be signi®cant when
chlorinated hydrocarbons are co-®red in solid wastes.
The HCl to Cl2 ratio is extremely important, as Cl2 is a
much superior chlorinating agent than HCl. The pres-
ence of chlorine as AClx/MeClx is due to the occurrence
of alkali and metal ions in the raw meal. The chlorine
may also be initially present in the raw meal or may be
captured from the gas-phase by alkali reactions of the
raw meal.
2. Experimental
To speciate these desorbed organics, cement kiln raw
meal samples were desorbed under cement kiln pre-
heater zone conditions (4% O2, temperature range of 30±
500°C). The results of these experiments showed that
cement kiln raw meal contains a wide range of organics
including benzene, toluene, naphthalene, alkanes, C14±
C18 carboxylic acids, phthalates and their derivatives
and natural products like cholesterol (Sidhu and Del-
linger, 1995). To gain a better insight into the speciation
of raw meal organics, the total organics thermally de-
sorbed from each sample were divided into di€erent
molecular weight ranges and di€erent chemical classes.
From these classi®cations, it was observed that most of
the organics are in the 200±300 amu molecular weight
range and are C14±C18 carboxylic acids. The second
largest group includes organics that are natural products
and that have molecular weights greater than 400 amu.
Aromatics and aliphatics (molecular weight less than
200 amu) form a small fraction of the total organics
desorbed from the raw meal. These results are consistent
with the hypothesis that organics are present in raw meal
as long chain carboxylic acids or as phthalates, choles-
terols, and their derivatives (Sidhu and Dellinger, 1995).
These results are also consistent with the work of Saiz-
Jimenez, which proposes that unsaturated fatty acids are
the major constituents of humic substances (Saiz-
Jimenez, 1994). Upon pyrolysis, fatty acids undergo
decarboxylation, fragmentation, cyclization and aro-
matization reactions, yielding a homologous series of
alkyl benzenes or alkyl furans (in the presence of
oxygen). Pyrolysis of humic substances has also been
investigated by Schulten in order to determine the three-
dimensional structures of soil organo-mineral complexes
(Schulten, 1995). The formation of the homologous
series of alkylated benzenes and naphthalenes from
aliphatic precursors is not only of interest regarding
structural con®guration of humic substances but also
has immense importance as an organic source in HAP
formation reactions in cement kiln preheater zone.
The raw meal sample was obtained from a local
preheater cement kiln. Semi-quantitative analyses of raw
meal and cement kiln dust were performed using X-ray
di€raction and infrared absorption (FTIR) techniques.
In X-ray di€raction analyses, it was assumed that all
elements were present as oxides. Raw meal showed small
amounts of chlorine. In addition to the elements indi-
cated in Table 1, sample contained a trace of sulfur,
however, the peak was below detection limit (0.2%). The
surface area of the sample was measured using B.E.T.
method. The total surface area was 2.5 m2/g.
To investigate HAP formation reactions from raw
meal organics, condensate was collected by desorbing
hydrocarbons from raw meal and collecting them in a
dry-ice cold trap. The carrier gas during desorption was
4% O2±He mix and the heating rate during these de-
sorption experiments was adjusted to emulate the heat-
ing rate of long dry kiln (room temperature to 730°C in
25 min). These desorption experiments were designed to
collect hydrocarbons greater than C5. The initial ex-
periments using this hydrocarbon condensate resulted in
inconsistent results. To understand these inconsistent
results, we re-examined the whole experimental proce-
dure and found that hydrocarbon condensate was in fact
a two-phase mixture (water + organics). However, the
volume of condensate collected was very small (250 ll
Table 1
Elemental analysis of cement kiln raw meal
Elements
Wt%
Mg
Al
Si
1.7
4.6
11.7
1.6
K
Ca
Fe
Cl
40.6
3.1
0.2

S. Sidhu et al. / Chemosphere 42 (2001) 499±506
501
from about 45 g of raw meal) and it was not possible to
di€erentiate between two phases. Even though we in-
jected 20 ll condensate in every experiment, we had
variable precursor concentrations due to di€erent phase
distributions in each experiment. Depending on the
position of organic±water interface and the position of
the syringe needle, it is possible to inject more water
phase (low inlet precursor concentration) in one exper-
iment and more organic phase (higher inlet precursor
concentration) in the next experiment. This inconsis-
tency in inlet precursor concentration is further magni-
®ed because of the non-linear dependence of certain
pollutant yields on the inlet precursor concentration.
To overcome the problems associated with water, a
benzene and myristic acid mixture (65%/35% by weight)
was used to simulate raw meal hydrocarbons. Myristic
acid represents the long chain carboxylic acid and
phthalate functionalities that constitute the raw meal
hydrocarbons. Benzene represents the aromatic decom-
position products of natural humic substances and also
acts as a solvent. We studied the chlorination of the
benzene and myristic acid mix over three di€erent sur-
faces: catalyst mix A (40 mg CaCl2, 5 mg KCl, 5 mg
AlCl3, and 5 mg of FeCl3), catalyst mix B (40 mg CaCl2
and 5 mg KCl), and catalyst mix A + thermally desorbed
raw meal. The catalyst mix A contains metal chlorides of
known chlorinating agents or promoters and their rela-
tive masses are based on the raw meal elemental com-
position (see Table 1). Catalyst mix B contains only
alkali and alkaline earth chlorides and should help elu-
cidate the role of transition and alkali earth metals in the
chlorination of raw meal organics. In a full-scale cement
kiln preheater zone, only 3±6% of total potassium or
calcium is chlorinated so we repeated catalyst A exper-
iments with 1.3 g of thermally desorbed raw meal to
represent conditions in the full-scale cement kiln pre-
heater zone. A complete experimental matrix is given in
Table 2.
All experiments were conducted on a high tempera-
ture ¯ow reactor system, referred to as the System for
Thermal Diagnostic Studies (STDS). The detailed design
of STDS has been published elsewhere (Rubey and
Grant, 1988). In brief, the STDS consists of four inte-
grated units: (1) a control console for precise adjustment
of temperature, pressure, residence time and carrier gas
¯ow; (2) a thermal reactor compartment that is housed
within an HP 5890 gas chromatograph (GC) to allow
precise temperature control of sample lines into and out
of the reactor; (3) a gas chromatographic chamber (HP
5890 GC) for separating products; (4) a mass spec-
trometer (HP 5970B MSD) for product identi®cation
and quantitation. A 3 mm inner diameter, packed-bed,
fused-silica reactor was used for this study. One-centi-
meter quartz wool plugs were placed on each side of the
catalyst to hold it in place and to avoid particle en-
trainment into the GC±MS.
The STDS was operated to simulate the conditions
occurring in the preheater zone of cement kiln. One ll of
reactant mix (benzene + myristic acid) was injected into
the STDS using a liquid syringe. In addition, a reaction
gas was also introduced into the STDS, which contained
4% O2, which is typical of cement kilns. The vaporized
reactant mix and the reaction gas were swept into the in-
line catalytic reactor of the STDS where the chlorination
reactions were studied. The experimental data were ac-
quired for temperatures of 300°C, 400°C, and 500°C,
with a gas-phase, mean-reaction time of 15 ms and
nearly constant total pressure of 1.1 atm. Chemical
analysis of the gaseous reactor e‚uent was performed.
The reactor e‚uents were cryogenically trapped and
then subjected to programmed temperature gas chro-
matograph using a DB-5 (J & W Scienti®c), 30 m, 0.25
micron ®lm thickness, 0.25 mm I.D. capillary column
held at a trapping temperature of )60°C. The GC was
temperature programmed from )60°C to 290°C (10 min
hold) at 10°C/min. Scanning (scanning 15±350 amu)
Table 2
Experimental test matrix
Exp. no.
1
Organic reactant
Surface
Temperature (°C)
Benzene (65 wt%) + Myristic acid (35 wt%)
Catalyst mix A (55 mg) (40 mg
CaCl2 + 5 mg KCl + 5 mg
FeCl3 + 5 mg AlCl3)
300, 400, 500
2
3
Benzene
Catalyst mix A (55 mg) (40 mg
CaCl2 + 5 mg KCl + 5 mg
FeCl3 + 5 mg AlCl3)
300, 400, 500
300, 400, 500
Benzene (65 wt%) + Myristic acid (35 wt%)
Catalyst mix A (55 mg) (40 mg
CaCl2 + 5 mg KCl + 5 mg
FeCl3 + 5 mg AlCl3 + desorbed
raw meal (1.3 g)
4
Benzene (65 wt%) Myristic acid (35 wt%)
Catalyst mix B (45 mg) (40 mg
CaCl2 + 5 mg KCl)
300, 400, 500

502
S. Sidhu et al. / Chemosphere 42 (2001) 499±506
mode was used to identify products formed from these
gas-surface reactions.
3. Results
In these experiments, organic mixture simulating raw
meal organics was exposed to temperatures of 300°C,
400°C and 500°C in 4% oxygen over three di€erent
surfaces. When benzene and myristic acid mix was
passed over catalyst mix A, hardly any benzene de-
composed, as its concentration remained constant over
the temperature range investigated. However, the my-
ristic acid, simulating long chain carboxylic acid func-
tionalities evolved from the raw meal, was degraded
selectively compared to benzene. Products formed in-
cluded mono- through tetra-chlorinated methanes and
Fig. 2. Relative abundance of chlorinated aromatic products
from benzene and benzene/myristic acid experiments over cat-
alyst mix A (CaCl2, KCl, AlCl3 and FeCl3) at 300°C, 400°C and
500°C.
ethylenes, which increased in concentration from 300°C
to 400°C, with maximum yields around 400°C (see
Fig. 1). The mono- through tri-chlorinated benzenes
also increased in concentration with a maximum for-
mation rate at about 400°C (see Fig. 2). However, the
increase in hexachlorobenzene yields was much smaller
than the increase observed in the yields of mono-
through tri-chlorinated benzenes. Also, hexachloroben-
zene was observed as a product in the 300°C exposure
whereas the mono- through tri-chlorinated benzenes did
not form until temperatures greater than 300°C.
When benzene was passed over catalyst mix A,
mono- through hexachlorobenzenes were observed even
at 300°C. In fact, the only major di€erence between
chlorination of benzene and benzene + myristic acid mix
is that benzene only experiments show much higher
chlorination at 300°C. This could perhaps be due to
preferential adsorption of myristic acid on chlorination
Fig. 1. Relative abundance of chlorinated aliphatic products
from benzene and benzene/myristic acid experiments over cat-
alyst mix A (CaCl2, KCl, AlCl3 and FeCl3) at 300°C, 400°C and
500°C.

S. Sidhu et al. / Chemosphere 42 (2001) 499±506
503
Fig. 3. Relative abundance of chlorinated aliphatic products
from benzene/myristic acid experiments over three di€erent
surfaces: Catalyst mix A (CaCl2, KCl, AlCl3 and FeCl3); Cat-
alyst mix B (CaCl2 and KCl); Catalyst mix A (CaCl2, KCl,
AlCl3 and FeCl3) + desorbed raw meal, at 300°C, 400°C and
500°C.
Fig. 4. Relative abundance of chlorinated aromatic products
from benzene/myristic acid experiments over three di€erent
surfaces: Catalyst mix A (CaCl2, KCl, AlCl3 and FeCl3); Cat-
alyst mix B (CaCl2 and KCl); Catalyst mix A (CaCl2, KCl,
AlCl3 and FeCl3) + desorbed raw meal, at 300°C, 400°C and
500°C.
temperatures. This result suggests that transition metals
(Fe) are not necessary for chlorination of organic ma-
terials, at least under the experimental conditions in-
vestigated in this study. When desorbed raw meal was
added to catalyst A, yields of chlorinated products
changed. No tri- or tetra-chlorobenzenes were observed
with the addition of the desorbed meal. The dichloro-
benzene concentration decreased from a maximum at
300°C to a minimum between 400°C and 500°C. In
contrast with this result, catalyst A without the meal
showed no formation of dichlorobenzene until temper-
atures of 400°C. The hexachlorobenzene did not form at
all under these conditions (with desorbed meal) until a
temperature of 500°C was reached. In the catalyst A
sites at 300°C. At higher temperatures, adsorption of
myristic acid is followed by fragmentation and decarb-
oxylation into light hydrocarbons that are readily
chlorinated. These light chlorinated compounds are then
free to cyclize and form chlorinated aromatic com-
pounds. Chlorinated phenols were only observed at one
temperature (400°C) when benzene + myristic acid mix
was passed over catalyst A. However, in benzene only
(no myristic acid) experiments (over catalyst A), di- and
tri-chlorophenol were observed with maximum at 300°C
and a minimum between 400°C and 500°C.
Figs. 3 and 4 show the e€ect of di€erent catalyst/re-
agent compositions on chlorination products. Catalysts
A and B show similar chlorination yields at all three

504
S. Sidhu et al. / Chemosphere 42 (2001) 499±506
experiments without the meal, the hexachlorobenzene
formed at the lowest temperature 300°C. Clearly the
addition of the desorbed raw meal decreased the variety
of chlorinated products (both aliphatic and aromatic) at
all temperatures. This is because addition of desorbed
raw meal adds oxides of potassium, magnesium and
calcium to the reaction system thus increasing the
competition for available chlorine.
the preheater exhaust, even though no fuel was
combusted in the preheater. The authors concluded that
the MCBz resulted from the reactions of organics from
the kiln raw meal with chlorinated species (HCl, CaCl2,
KCl, FeCl3, etc.). Further testing demonstrated a linear
relationship between the log of the chlorobenzene
emission rate and the total chlorine input rate to the kiln
and an absence of correlation with any other parameter
characterizing the input waste stream. This suggested
that a ®rst-order reaction of chlorine with organic
compounds desorbed from the raw meal in the preheater
was responsible for the observed MCBz emissions.
Furthermore, the rate of formation apparently de-
creased with increasing temperature and decreasing ox-
ygen content of the gas. This indicates the possibility of
a surface catalyzed reaction involving molecular oxygen.
The formation of any chlorinated hydrocarbon in a
cement kiln is somewhat surprising since it would be
expected that the primary source of chlorine for chlori-
nating reactions, viz. HCl, would be removed by reac-
tions with large excess of calcium (CaO and Ca(OH)2),
magnesium (MgO and Mg(OH)2) and potassium (KOH)
to form respective chlorides. However, our experiments
have shown that CaCl2/KCl are capable of chlorinating
organics even in the absence of iron and aluminum.
These results are in disagreement with the general con-
sensus that transition metals (e.g., iron) are required to
chlorinate organics under post-combustion conditions.
However, there is theoretical and experimental evi-
dence that CaCl2 can be used as a chlorinating agent.
Heterogeneous equilibrium calculation of the Deacon
reaction ꢀ4HCl ‡ O2 ˆ 2Cl2 ‡ 2H2O† in presence of
calcium with O2:H2O:Ca:Cl ratio of 240:90:20:1 shows
that ꢁ2% chlorine is present as Cl2 over the temperature
range of 200±1000°C. These results are consistent with
experimental observations of Chan et al. (1996), who
investigated CaCl2, MgCl2, AlCl3, FeCl3, and NaCl as
chlorinating agents for volatilizing the heavy metals of
concern, i.e., Pb, Cd, Zn, and Cu. Chan et al. (1996)
observed that tendency of chlorinating agents to release
chlorine varies based on their stability in air. This ten-
dency can be quanti®ed using the equilibrium partial
pressure of chlorine over the chlorinating agent under a
constant partial pressure of oxygen (0.21 atm). Chan
et al. found that equilibrium partial pressure of chlorine
decreases according to the sequence AlCl3, MgCl2,
FeCl2, CaCl2 and NaCl (see Fig. 6) in the temperature
range of 600±1200°C. These experimental results suggest
that CaCl2 must be reacting with oxygen to liberate
chlorine according to Eq. 1.
4. Discussion
In this paper, we have reported results of four sets of
experiments constituting two compositions of organic
precursors and three di€erent chlorinating reagent/cat-
alyst compositions. All these experiments have yielded
chlorinated methanes, ethylenes, benzenes and phenols
as major chlorinated products. The chlorinated products
observed in this study are similar to the major chlori-
nated products observed from full-scale cement kilns in
®eld studies. Fig. 5 shows 10 major chlorinated hydro-
carbons observed in stack gases of a wet cement kiln
(Dellinger et al., 1993). This ®gure shows that C1 to C6
chlorinated products are present in stack gases regard-
less of the fuel used which is consistent with the results
of this study.
The results of this study are also consistent with the
®ndings of Lamb et al. (1994), who used monochloro-
benzene (MCBz) as the principal organic hazardous
constituent of solid waste in a test for certi®cation of
compliance. They observed that, while measurements at
the preheater bypass veri®ed that the MCBz injected in
the sintering zone was virtually completely destroyed
(destruction and removal eciency of greater than
99.99%), measurable amounts of MCBz were found at
CaCl2 ‡ 1=2O2 ˆ CaO ‡ Cl2
ꢀ1†
Fig. 5. Comparison of yields for the 10 most highly concen-
trated hydrocarbons at wet cement kiln when burning liquid
and solid waste derived fuel, diesel oil/coal and coal only
(Dellinger et al., 1993).
This liberated chlorine can then chlorinate desired heavy
metals or organics. Chan and Kirk (1996) in a follow up
study showed that CaCl2 is in fact a more selective

S. Sidhu et al. / Chemosphere 42 (2001) 499±506
505
5. Conclusions
Based on the results of our experiments the following
conclusions can be drawn:
(1) At 300°C, presence of myristic acid seems to in-
hibit chlorination, as benzene experiments show higher
yields of chlorinated products than benzene/myristic
acid experiments. However, as the temperature increases
to 500°C, the yields of chlorinated products in benzene/
myristic experiments become comparable to the yields of
chlorinated products in benzene experiments. This sug-
gests that at higher temperatures, myristic acid breaks
into smaller hydrocarbons, which then chlorinate and
cyclize.
Fig. 6. Temperature dependence of equilibrium partial pressure
of chlorine on di€erent chlorination agents (assuming PO
0:21 atm, Chan et al. (1996).
(2) Comparison of catalyst mix A and catalyst mix B
data suggests that metals like Fe and Al are not essential
for chlorination of organics under the conditions in-
vestigated in this study. When desorbed raw meal was
added to catalyst mix A to make chlorinated metal
content approximately 3.5% of the total metal content, a
decrease in the total yields of chlorinated products was
observed at all temperatures. This is because addition of
desorbed raw meal adds oxides of potassium, magne-
sium and calcium to the reaction system thus increasing
the competition for available chlorine.
ˆ
2
chlorinating agent for volatilizing heavy metals (Pb, Cd,
Zn, and Cu) than Cl2.
Experimental study by Shigapov et al. of oxidative
dimerization of methane over CaO±CaCl2 catalytic
system also provides evidence for CaCl2 as chlorinating
agent at T ˆ 750°C. Shigapov et al. (1988) found that
CaCl2 catalyst experiments yielded methyl chloride as a
product along with ethylene and ethane. They concluded
that in the absence of gas-phase chlorine, presence of
methyl chloride means that CaCl2 is capable of gener-
ating chlorine according to the following reactions:
(3) Major chlorinated products observed in this study
were also present as major chlorinated hydrocarbons in
the cement kiln ®eld emissions data. The ease of chlo-
rination of raw meal organic surrogates in this study
demonstrates that raw meal organics can react with
chlorinated species in cement kiln preheater zone to
form chlorinated HAPs like benzenes, phenols, meth-
anes and ethylenes.
CaCl2 ‡ 1=2O2 ˆ CaO ‡ Cl2
CaCl2 ‡ H2O ˆ CaO ‡ 2HCl
2HCl ‡ 1=2O2 ˆ H2O ‡ Cl2
ꢀ1†
ꢀ2†
ꢀ3†
Acknowledgements
This research is partially supported by National
Science Foundation Grant No. CTS-9613422. We would
like to thank Project Ocer Dr. Farley Fisher for his
support. We would also like to thank Lou Cooper, Rich
Striebich and Cory Kirk for their help in preparing this
paper.
Shigapov et al. also studied the e€ect of water on the
product yields and found that presence of water in-
creased the yield of methyl chloride and C2 products and
also increased the eciency of ethane dehydrogenation
to ethylene. These results are consistent with the pres-
ence of HCl and water in reactor exit gas in our exper-
iments.
References
This alkaline-catalyzed chlorination represents an
interesting and potentially important ®nding, as addi-
tion of alkaline sorbent has been proposed as PCDD/F
control technique through control of HCl (Gullett et al.,
1994). However, since the Cl2 concentration can also
increase under certain conditions, this approach could
be counterproductive if used indiscriminately. Clearly,
further work on formation of molecular chlorine and
how it relates to process parameters such as tempera-
ture, oxygen concentration, and HCl to calcium or
magnesium ratio is required.
Chan, C., Jia, C.Q., Graydon, J.W., Kirk, D., 1996.
J. Hazardous Mater. 50, 75.
Chan, C., Kirk, D., 1999. J. Hazardous Mater. 64, 75.
Dellinger, B., Pershing, D.W., Saro®m, A.F., 1993. A Report of
the Scienti®c Advisory Board on Cement Kiln Recycling,
Cement Kiln Recycling Coalition, Washington, DC.
Gullett, B.K., Lemieux, P.M., Dunn, J.E., 1994. Environ. Sci.
Technol. 28, 107.
Lamb, C.W., Miller, F.M., Roth, A., Dellinger, B., Sidhu, S.S.,
1994. Hazard. Waste Hazard. Mater. 11, 201.

506
S. Sidhu et al. / Chemosphere 42 (2001) 499±506
Rubey, W.A., Grant, R.A., 1988. Rev. Sci. Instr. 59, 265.
Saiz-Jimenez, C., 1994. Environ. Sci. Technol. 28, 1773.
Schulten, H.-R., 1995. J. Anal. Appl. Pyrolysis 32, 111.
Shigapov, A.N., Novozhilova, M.A., Vereshchagin, S.N.,
Anshits, A.G., Sokolovskii, V.D., 1988. React. Kinet. Catal.
Lett. 37, 397.
Sidhu, S., Dellinger, B., 1995. In: Proceeding of the Interna-
tional Specialty Conference on Waste Combustion in
Boilers and Industrial Furnaces, A&WMA, Kansas City,
MO, March, p. 119.
Trenholm, A., Hlustick, D., Hansen, E., 1990. Presented at
A&WMA, 83rd Annual Meeting, Pittsburgh, PA.