Copper speciation of combustion ashes

Start date 01/12/2009
End date The project is closed: 31/12/2015

​Copper is one of the more abundant trace metals in municipal solid waste (MSW). It is potentially toxic and may pose as a problem in managing of the ashes from combustion of MSW. It may pose as a potential asset as well, since copper has a high value and could be recovered and recycled. Copper has also been identified as one of the most important elements contributing to dioxin formation within the boiler. With increased knowledge of the copper chemistry in combustion and in the ashes, it might be possible to reduce the amount of dioxins formed and thereby reducing the toxicity of the ashes as well as the flue gases. It could also increase the understanding of how to best manage different ashes, and support the development of recycling techniques.

The aims of this project are to investigate further the applicability of the synchrotron based X-ray absorption spectroscopy (XAS) for the speciation of trace metals (e.g. Cu) in combustion ashes and to contribute to the understanding of the chemistry of copper in combustion of biomass and waste derived fuels, sewage sludge and combinations thereof. This includes clarification of mechanisms involved in the binding of copper and correlation of different copper species to the presence of dioxins in the ashes and to copper leachability.

Methods and Materials

​The XAS method has earlier been applied to the speciation of cadmium in fly ashes in a PhD thesis and showed promising results. The positive features of the method is that it is element specific, it can show the oxidation state of the metal and in favorable cases the compounds present can be identified and approximately quantified. The nearest neighbor atoms around the target atom (the investigated element) can be identified by theoretical modeling. Semi-quantitative determination of metal compounds is possible with linear combinations of XAS data from pure (reference) compounds obtained at the same beamline as the sample data. Thus, the quality of the quantitative result is dependent on the amount and quality of such standard data. A compound that is not included in the database cannot be found as a component in a sample. Therefore, an important part of this work has been the building up of a database of XAS data for copper compounds. In many cases this involved synthesis of reaction products that may occur in combustion residues. Our database now comprises 23 copper compounds. In addition, for a scientist to be able to do XAS data collection at a suitable synchrotron beam line the research project has to be approved by the scientific committee reviewing and choosing projects. The available beam time at synchrotron is very limited and therefore additional methods were investigated as well in this work.

Additional methods being used are:

- X-ray powder diffraction (XRD) gives information about the crystalline compounds present in amounts larger than 2% (w/w).

- Scanning electron microscopy with element detection by energy dispersive X-ray spectrometry (SEM-EDX) is used to investigate the spatial distribution of elements in the ash.

- Diffuse reflectance Fourier transform infrared spectroscopy (DR-FTIR) gives information about compounds that are not necessarily crystalline. In this project it mainly works as a complement to the XRD technique.

- Thermodynamic modeling of ash chemistry in furnaces, for predicting the ash composition in theoretical scenarios. Software used for this purpose is e.g. FactSage.

- High temperature reactions, as an investigation of what types of copper compounds could be expected from incineration. Some of the reaction products have also been used as reference compounds in the XAS study.

Ashes from three different boilers have, so far, been analyzed in this project. At a commercial bubbling fluidized bed (BFB) boiler, fired with sorted MSW, ash samples were collected at four different locations (the bottom bed, a hopper directly after the combustion chamber, a cyclone for separation of coarser particles in the flue gas and a textile filter for separation of finer particles). At a commercial grate fired (GF) boiler, fired with MSW, one sample of filter (electrostatic precipitator) ash was collected. At a research boiler with a circulating fluidized bed (CFB) ash samples were collected at four different locations (bottom bed, return leg, cyclone and filter) at seven different occasions (cases a-g). During the seven occasions the fuel varied somewhat. In cases a-d the fuel consisted of bark pellets, refuse derived fuel (RDF) and additives (kaolin, phosphate and sulfate). In cases e and f the fuel consisted of wood chips and RDF, with or without lime injection to the flue gas before the filter. In case g the fuel consisted of wood chips, shredder light fraction (SLF, also known as “carfluff”) and sewage sludge. Two filter ash samples (from the BFB and the GF) had previously been tested for leaching with acidic solution as well as ammonium nitrate solution. The solid residues from these leaching tests have also been investigated in this project.

In a parallel project, the BFB filter ash mentioned above, along with a filter ash from a fourth boiler, has been analyzed by sequential extraction. We have collected XAS data from the residues from each of the steps of the sequential leaching. These data has not yet been fully analyzed, but it looks promising for an increased knowledge in both the composition of the filter ashes and what actually happens during the sequential leaching.

Preliminary Results

​In the samples collected at several positions within the boilers (BFB and CFB), it is possible to see two trends. Firstly, the copper speciation generally follows a gradual shift from the bottom bed to the filter. Secondly, the average copper oxidation state usually increases gradually from the bottom bed to the filter. There is, however, a large difference in copper speciation between the different boilers. In the BFB ashes we found a transition from Cu metal, Cu (I and II) oxides and mixed oxides (such as the CuCr2O4) in the bottom bed, to almost 100% Cu(II) in a combination of sulfate, hydroxides and chlorides in the filter ash. In the GF filter ash we found a large part Cu(II) phosphate and/or silicate along with minor parts of Cu metal, CuO and Cu(I) sulfide and/or chloride. From cases a-f from the research CFB boiler we could see no significant change in copper speciation due to fuel composition or lime injection. In all six bottom ashes (a-f) we could see mostly Cu metal and Cu sulfide along with possibly Cu (I and II) with oxygen as nearest neighbor in the bottom ash. In the filter ashes (a-f) we found a small part (5-10%) Cu metal and about 90% Cu(II) with oxygen as nearest neighbor (probably a significant fraction is CuO but CuSO4 and Cu3(PO4)2 are also possible). Case g from the CFB boiler showed a completely different result, with a strong CuFe2O4 signal from the filter ash, which can be due to the relatively high iron (Fe) content of both the SLF and the sludge. The sulfides in the CFB bottom ashes are due to the low air-to-fuel ratio in the bed (50-55% primary air supplied to the bed and the rest supplied as secondary air further up in the freeboard).

A high solubility of the copper compounds in the BFB filter ash is reflected in the leaching results, where almost 100% of the copper was released. The GF filter ash, on the other hand, only released 60% of its copper. The residues of both ashes showed remarkable similarities in copper speciation, from both acid leaching as well as ammonium nitrate leaching. The residues did also show large similarities, in copper speciation, with the untreated GF filter ash, which helps to explain the lower solubility of copper in that sample.

The Swedish Energy Agency, through The Swedish Research Council
Chalmers Area of Advance, Energy

Published: Fri 08 Nov 2013. Modified: Tue 23 Sep 2014