PHIL

line

CHEMISTRY OF DIOXIN FORMATION

line

FAIRY-TALE

Chlorinated plastics, principally PVC (Polyvinyl Chloride), are a major source of chlorine which is necessary for dioxin generation in incinerators, and at present most hospital waste contains more than twice the amount of chlorinated plastic as does regular municipal waste..

Physician Consensus Statement, Physicians For Social Responsibility and Health Care Without Harm, March 1998.


Parts of this page

line

FACTS AND FIGURES

The Physicians For Social Responsibility are completely wrong with their conclusion. Of course you need chlorine to make dioxins, as good as you need carbon and oxygen. But all three are very abundant in incinerators, compared to the amount of dioxins which are really formed.

The following information is a summary of an article from H. Huang and A. Buekens, Department of Chemical Engineering and Industrial Chemistry, Free University of Brussels, Belgium [33]. This article describes the most probable routes how dioxins are formed in incinerators. Most dioxins are formed from the carbon in fly ash during cooling of the off-gases.


Findings from different experiments:

Carbon morphology dependence:
Some carbon types generate PCDD/F's (dioxins): sugar coal, charcoal, soot, active coke,...
Others generate little PCDD/F under the same conditions: graphite and unactivated amorphous carbon.

Catalyst dependence:
Cu2+ ions have a strong catalytic effect on PCDD/F formation, Fe3+, Pb2+ and Zn2+ have a minor effect, and a lot of others have no observable effect.

Oxygen dependence:
Molecular O2 must be present in the gas stream to form PCDD/F's. The formation rate increases with the O2 concentration by a reaction order of about 0.5.

Temperature dependence:
Maximum PCDD/F formation occurs at 300-325 °C, little can be detected below 250 or above 450 °C.

Other flue gas components:
HCl, Cl2, SO2, CO2, CO and H2 present in the concentrations normally found in Municipal Solid Waste Incinerators (MSWI) flue gases have no significant influence on PCDD/F formation. The effect of H2O is less certain.

Cogeneration of other (chlorinated) compounds:
A variaty of other (chlorinated) compounds like (polychlorinated) benzenes, phenols, naphthalenes and biphenyls are generated together with PCDD/F's.

Relationship with carbon gasification:
PCDD/F formation has been shown to be closely related to low-temperature carbon gasification. This too is oxygen dependent and increases by a reaction order of about 0.5 with O2 concentration.


Basic reactions:

The main carbon source for PCDD/F generation seems to be deformed and degenerated graphitic structures. Neither completely ordered graphite nor amorphous carbon generate much PCDD/F. For regular graphite the amount of PCDD/F generated is four orders of magnitude lower than for deformed graphite.

The basic reaction is the oxydation of microcrystallite carbon. At low temperatures (below 700 °C) this occurs mainly at the imperfect edges of the layer, which forms active sites. Without a catalyst, this is a slow process. Several metal oxydes catalyse the oxydation, including copper and iron. This forms or leaves some (oxydised) ring structures, including benzene, phenol, biphenyl, dibenzodioxins and -furans. The latter three can be formed by coupling elementary ring structures like the former two, called the Ullmann coupling reaction. Alternatively, the DD/DF structures are directly built by oxydation of the carbon layer and subsequent oxydative degradation of surplus carbon rings.

Chlorination of these elementary and condensed structures or directly at the carbon layer is happening concurrently and is strongly catalysed by some metal salts, especially copper (in all forms). Also the Ullmann reaction of chlorinated elementary rings is strongly catalysed by copper. At the same time dechlorination and decomposition reactions also occur and are also catalysed by copper. Which reactions are preffered is mainly a matter of temperature. Higher temperatures favour decomposition.

The net yield of all these reactions is:
For each app. 30-ring (100 carbon) graphite sheet of a 5-10 layer microcrystallite, app. one polychlorinated aromatic structure is formed [personal note: the rest is transformed into CO, CO2 and non-chlorinated aromatics]. For each app. 200 layers, one PCCD/F is formed. It should be noted that the oxydation is on a layer-by-layer basis. That results in the fact that regular graphite, which contains 5x10^4 regular carbon rings per layer, produces only a very little amount of aromatics and hence very little PCDD/F.

Metals like copper, which catalyses all three reactions leading to the formation of PCDD/F's will form a lot of them. Iron has a strong catalytic effect for oxydation reactions, but a weaker for chlorination and Ullmann-type reactions. Many other metals have some strong effect for one reaction but no or much weaker for the others. None of them produce as much dioxins as copper and iron do.


Experiments.

Turbulence:

Lowering the amount of primary air and working with a continuous feed reduced the disturbances in temperature and amounts of fly ash in an experiment at an incinerator in Flanders. With secondary air, temperatures were continuously held over 950 degr.C. This measures alone were enough to reduce dioxin emissions with a hundredfold. But also in several other experiments and renewed designs, high dioxin reductions were obtained by reducing turbulence and upsets in the hearth and more turbulence and constant temperatures in the secondary burning zone.
 

Temperature:

In one experiment [1], at the incinerator outlet, very little chlorinated dioxins/furans were detected. Between the incinerator outlet and the boiler outlet, each pass gave a spectacular rise in chlorobenzenes, chlorophenols and PCDD/F's: A factor 3 between 900 and 500 degr.C, a factor 7 with further cooling to 270 degr.C and a factor 10 at the boiler outlet (250 degr.C, all compared to the incinerator outlet). That means that dioxins/furans are formed de novo. Further experiments revealed that these are formed on fly ash.
Real fly ash was heated during 2 hours at different temperatures in air [2]. Up to 200 degr.C no differences were found with the original fly ash. At 300 degr.C the amounts were 10 times higher, at 400 degr.C 3.5 times, at 500 degr.C 30 times lower (PCDD's) and 30% lower (PCDF's) and finally at 600 degr.C PCDD levels were below detection limits and PCDF levels were 10 times lower than original.
 

Oxygen:

When fly ash was heated at 300 degr.C, 2 hours in different atmospheres [2], PCDD/F's are destroyed in pure nitrogen, but doubled already with 1% O2, 4 times higher with 4% O2 and 8 times with 10% O2. If steam was added at 10% O2, the amounts were 20 folded...
 

Chlorine:

A synthetic fly ash (1% C, 1% KCl+, 0.4% Cu as chloride, rest Mg-Al-silicate) was spiked with different amounts of KCl [3], between 0.5 and 6% total Cl. Again heated at 300 degr.C during 2 hours. The sum PCDD/F doubled from 0.5% to 6% chlorine.
 

Time:

PCDD/F levels roughly augment linear with time with a similar mixture as in previous point (but with 7% Cl). Moreover, although only inorganic chloride was added, amounts of total organic chlorine augmented from zero to 0.5 mg/g of which 0.1 mg/g was extractable, after 4 hours of heating.
 

Metals/Copper:

With a similar mixture as in the previous points, 2 hours at 300 degr.C, but with 1% KCl base and a blank or 1% chlorides of different metals showed interesting results: without an extra metal <0.5 ng/g PCDD and <0.5 ng/g PCDF. From the several additions, CaCl2, FeCl2 and CdCl2 didn't show higher levels. Further:
 
Metal chloride 1% PCDD PCDF
no <0.5 <0.5
MgCl2.6H2O 2.0 3.6
CaCl2 <0.5 <0.5
ZnCl2 8.1 10.4
SnCl2.2H2O 1.2 7.4
FeCl2.4H2O <0.5 <0.5
FeCl3.4H2O 7.6 53.0
MnCl2.4H2O 3.0 4.9
NiCl2.6H2O 0.4 5.5
CdCl2.H2O <0.5 <0.5
HgCl2 0.5 7.4
PbCl2 5.0 20.0
CuCl2.2H2O 679.0 4340.0

As you can see, copper is by far the best catalyst to form dioxins...
That was confirmed in another test, where different amounts of copper were added:
 

Congeners additional % Cu2+
0.00 0.08 0.24 0.40
PCDD 4.50 101 770 1448
PCDF 22.6 760 3640 8480

Again, copper plays a crucial role in dioxin formation...

In all these tests only INorganic chlorine was used, together with amorphous carbon (active coal) and an inactive carrier. Crystalline carbon like graphite doesn't form dioxins. Further, chlorinated organic molecules, including dioxins/furans, are formed from inorganic chloride salts. Thus the assumption that only organic chlorine is involved in dioxin formation is proven false. And the amount of chlorine plays a very limited role, compared to other factors like turbulence, time, temperature, oxygen and by far copper content...
 

line

THE ALTERNATIVES

Several alternatives for incineration exist, each having their own advantages and drawbacks. One of the alternatives for direct incineration is thermolyses, using no air in the first step. The waste, mixed with some lime to trap acid gases, is heated to app. 500 °C without burning. All organic material is carbonised. Because there is no oxygen in the atmosphere, no dioxins are formed. The gases are incinerated to carbonise the waste. No dioxins or other nasty stuff are detected above detection limits (according to the makers!), because the main source, carbon cristallytes are absent in the gases.
The remaining solid carbon is separated from non-combustibles, washed to remove salts and can be used as coke-replacement in high-temperature processes like cement manufacturing. Alternative (Siemens, Germany) is on-site incineration (at 1,300 °C) for boiler heating. I haven't seen dioxin figures for that part of the process...


CONCLUSION

PCDD/F production is mainly a catalytic reaction of carbon with oxygen and chlorine in the fly ash, strongly dependent of type and amount of catalyst, temperature, type and amount of carbon, weaker dependent of concentration of oxygen in the gase phase, weak of the amount of chlorine in the ash and independent of chlorine/HCl and other gases (except water) in the gas phase.

This explains why co-burning of PVC results in not more PCDD/F formation/emission than other materials in incinerators (or accidental fire), because most of the chlorine from PVC is split off as HCl, which has no measurable effect on PCDD/F formation at the concentrations found in MSW incinerators.

line

You are at level three of the Chlorophiles pages.

Created: March 30, 1998.
Last update: November 25, 2001.

Welcome page

To the Home Page of the Chlorophiles

Up: Chlorine input and dioxin emissions

Left: The ASME Research Report summary on incineration

Right: An (Italian) example of state-of-the-art municipal and hospital waste incineration

For any comment on this or other pages, especially on dioxin formation:

Mail to: Chlorophiles@ping.be