Recovery of titanium dioxide and other pigments from waste paint by pyrolysis.
Keywords Titanium dioxide recycling, Ti[O.sub.2], Paint waste recycling, Pigment, Extender, Pyrolysis
The U.S. Geological Survey estimated that in 2011 the total world mining production of ilmenite and rutile ore, the two major mined titanium minerals, was 6.7 million tons. (1) As much as 95% of the mined titanium minerals were) used in the production of titanium dioxide. (2) While several nonpigment uses exist for titanium dioxide (Ti[O.sub.2]), the most important use is as a white pigment in a wide range of products, for example, paint, plastics, and paper. About 90% of the Ti[O.sub.2] is used as a pigment and the majority of this is in paint and coatings. Due to the heavy use of titanium dioxide in architectural coatings, activity in the construction industry has an especially strong effect on titanium dioxide demand. (3)
When used in paint, the most important qualities of a white pigment are its ability to give the paint white color and to have a high opacity to cover the substrate on which it is applied. Ti[O.sub.2] appears white due to its ability to scatter visible light. Ti[O.sub.2] in the rutile form, however, is not perfectly white since it absorbs some of the light in the 400-500 nm range, thus giving creamtoned whites. (4) High whiteness has a direct correlation to chemical purity, especially the absence of transition metals and compounds thereof. To optimize the Ti[O.sub.2] performance as a pigment, it is washed, filtered, and milled to the right particle distribution size. Typically a Ti[O.sub.2] pigment is given a surface treatment to optimize desirable paint qualities such as dispersibility, opacity, and gloss. (5)
Mineral pigments or extenders have traditionally been used to dilute a more expensive product without compromising the desired qualities. Now almost all extenders can be seen as functional fillers with their own purpose and function in a coating formulation. The carbonates are probably the most used extenders based on weight consumed, but they can be replaced with other extenders more abundant in the local area. This gives large variation in coating formulations depending on which extenders are available locally. Common complements/replacements for carbonates are kaolin (kaolinite) and talc. (4)
Due to the high demand for Ti[O.sub.2], the increased price level, and the high environmental impact of producing virgin Ti[O.sub.2], new innovative ways of titanium dioxide production are needed. The European Union (EU) has recognized the environmental impacts of titanium dioxide production and has consequently set restrictions on the amount of titanium dioxide allowed to be used in paint formulations in order to produce a paint that qualifies for the voluntary EU Ecolabel. (6) A possible new source for titanium dioxide can be recovery from different waste materials, such as paint waste. Today, processes for the recovery of titanium dioxide are nonexistent. Recovery of Ti[O.sub.2] from paint waste is of interest due to the relatively high Ti[O.sub.2] content in the waste and its abundance as a reasonably homogeneous waste stream. The main incentive for a recycling process is legislative and environmental rather than monetary. In the future, paint producers may be responsible for their product from the cradle to the grave (so-called extended producer responsibility). This means, for example, that the producer is responsible for the take-back and recycling of leftovers from consumers. Not only could a recycling process provide paint manufactures with a new source of the pigment, it could also provide the industry with a way to manage production waste and leftover paint products.
From a wider perspective, the reuse of old paint, either directly or as a component in new batches of paint, is a straightforward approach and initiatives in this direction have been made. (7,8) A company in the United Kingdom (UK) collects leftover paint, processes it, and incorporates it in new paint. (7) Another initiative is the Community Repaint organization in the UK that is collecting usable waste paint and distributing it to people and organizations needing paint. (8) These options are not always applicable, however, due to dry nondispersible paint lumps, incompatible paint components, microbiological contamination, and changing chemical and biocide regulations that can make the new paint batch containing waste paint unusable from a legal perspective. Instead, the main existing waste management processes for paint have been designed with the aim of reducing the volumes of waste. In Sweden, paint cans and paint waste together are a major part of the hazardous waste stream from households and industry. The packaging is normally sorted separately, while the paint residues are incinerated and the resulting ashes are landfilled. (6)
Gasification or pyrolysis of paint waste or used plastic paint containers as a method for recycling of paint components and the value of these components have been described in the literature. (10-14) In the described methods, the organic components in the paint were viewed as the valuable part to be used for energy production. No work on the recovery of the inorganic parts (Ti[O.sub.2]-containing pigments) has been found in the literature to date.
In the present work, the inorganic pigments are considered to be the most valuable part of the waste paint. The aim of this work was to investigate if the inorganic Ti[O.sub.2]-containing residues (pigments) resulting from the pyrolysis of waste paint can be used as a replacement, or partial replacement, for virgin pigment and fillers in paint formulations. The gas and the oil produced from the pyrolysis fraction could potentially be used as an energy source or a raw material for the synthesis of chemicals, but at the moment that is outside the scope of the present work.
Materials and methods
White model paint
Ti[O.sub.2] in the rutile form, along with some common extender pigments, were procured from commercially available sources. The extenders chosen for this study were dolomite, kaolin, talc, and mica. Unless otherwise stated, the pigments were used as delivered from Akzo Nobel Decorative Paints, UK.
A model paint containing all the pigments/extenders was produced according to the formulation shown in Table 1. The equipment for mixing and paint preparation in the laboratory of Akzo Decorative Paints in Slough, UK was used. In addition, a mixed pigment sample corresponding to the pigment composition in the model paint was prepared by mixing dry pigments together (see Table 1 for composition). Two types of binders were also acquired--a vinyl acrylic latex copolymer with acrylic latex and a pure acrylic latex binder. These two binders correspond to the major types of binders that can be found in paint waste. Both contain 52 wt% polymer and 48 wt% water. A nonpigment quality of rutile (particle size <5 [micro]m, purity [greater than or equal to] 99.9%), to be used for comparison with pigment quality rutile, was acquired from Sigma-Aldrich. This rutile had no surface coating and was used to assess if the recovery process has an effect on the rutile itself or on the surface coating of the rutile pigment.
Methods for characterization of pigments and inorganic recovery residues
X-ray powder diffraction (XRD)
The main crystalline compounds in pigment samples were identified by qualitative X-ray powder diffractometry (XRD) using a Siemens D5000 X-ray powder diffractometer with an X-ray tube giving the characteristic Cu radiation and a scintillation detector. The 2[theta] range used was 10[degrees]-70[degrees] with a step size of 0.050[degrees] and a 1-s step time. The identification of compounds was performed through comparison with standards in the Joint Committee of Powder Diffraction Standards. (18) Identification of a compound in a mixture is generally possible if the compound is present at a concentration of 2 wt% or more. Amorphous compounds and compounds occurring in nano-sized crystals cannot be detected by XRD.
Thermo gravimetric analysis (TGA)
Thermogravimetric analysis (TGA) measurements of pigments and extenders were made in order to see how a thermal recycling process would affect the titanium dioxide and the extenders in the model paint. The nonpigment quality of rutile was also measured as a reference for the titanium dioxide pigment. A nitrogen atmosphere was used to simulate the oxygen-free atmosphere in a pyrolysis reaction. All of the pigments were measured on a NETZSCH ST A 409 PC Luxx instrument and the nonpigment rutile was measured on a TA Instruments Q500 within the temperature interval 25-1000[degrees]C, with a heating rate of 10[degrees]C/min. Before starting an experiment, the furnace was evacuated twice and then filled with [N.sub.2].
The high temperature stability of the two binder samples, a vinyl acrylic latex copolymer with acrylic latex and a pure acrylic latex (see the materials section for more information), was investigated using the TGA Q500 from TA Instruments within the temperature range 25-600[degrees]C, with a heating rate of 5[degrees]C per minute in inert atmosphere ([N.sub.2]) to simulate pyrolysis conditions.
Specific surface area BET measurements
The BET (16) specific surface areas of the pigments, the pigment mix used in the model paint and the pyrolysis ashes, before and after oxidative heat treatment, were determined by [N.sub.2] adsorption isothermal at 77 K using a Micromeritics ASAP2020. Before measurement, the samples were outgassed at 60[degrees]C under high vacuum (roughly 1 [micro]m Hg), until the samples were considered dry or for a maximum of 1500 min. The samples were considered dry when the measured pressured change due to the sample was lower than 5 pm Hg/min. The outgassing temperature has a significant effect on the BET result, (17) so a relatively low outgassing temperature was chosen in order not to alter the surface of the pigment.
Microwave pyrolysis experiments
The pyrolysis experiments were carried out in a pilotscale microwave pyrolysis furnace provided by Stena Metall AB, Goteborg. The dimensions of the pyrolysis chamber were 390 x 430 x 950 (mm) and the total capacity was 10 L. (18) The microwaves were produced by three magnetrons of model Samsung OM 75P with a fixed frequency at 2465 MHz (18) and an individual power output for each magnetron of 1.5 kW. A sample container made of mica (transparent to microwaves) was used. The oven was flushed with a flow of 5.6 L nitrogen per minute for 4 min before each experiment. The liquid residue fraction produced in the pyrolysis was continuously collected during the experiment, but the gas fraction was not collected. All temperatures were measured continuously at three points, inside the sample, at the surface of the reaction vessel, and in the gas flow out from the furnace (see Fig. 1). After the experiment, the dry residue and the liquid oil fraction were weighed and preserved for analysis. The amount of gas produced was estimated by subtraction of the dry and liquid residue fractions from the total weight prior to the experiments. In the present work, the focus was on the inorganic parts of the sample materials and therefore the oil and gas fractions were not analyzed in detail. However, an average heating value for the oil was determined at the laboratory of the Eon heat and power plant in Karlshamn, Sweden.
Four pyrolysis experiments were performed, three on the paint formulation shown in Table 1 and one on the same formulation but dried to a dry content of 65% prior to the pyrolysis experiment. This semi-dry paint still contained some water but was nonfluid. Experiments 1 (wet paint) and 4 (dry/semi-dry paint) were designed for the initial evaluation of microwave pyrolysis as a method to recover the inorganic components from a waste paint. Experiments 2 and 3 were performed using larger amounts of material in order to produce enough recycled pigment material to make it possible to manufacture a test paint based on a significant fraction of recycled pigments.
After-treatment and characterization of ash from pyrolysis experiments 1, 2, and 3
The solid residue from pyrolysis experiment 1 was homogenized, weighed, and heated in a muffle furnace from 22 to 400[degrees]C (the heating took 40 min) in air. The temperature was then held at 400[degrees]C for 230 min. After cooling, the sample was weighed and its mineralogy was analyzed by XRD. In addition, the effect of the after-treatment on the color of the pyrolyzed material was visually evaluated. The evaluation was done by placing six samples next to each other on a white background and visually comparing the whiteness of each sample. The samples included (1) ground pyrolysis ash, (2) ground heat-treated pyrolysis ash, (3) virgin rutile pigment, (4) virgin dolomite, (5) virgin mica, and (6) a mix of virgin pigment corresponding to the one in the pyrolysis ash.
The solid residues from pyrolysis experiments 2 and 3 were mixed together and homogenized. This material was then divided into 10 sub-samples, which were heat treated at 450[degrees]C in air for 3 h. The weights before and after the heat treatments were recorded. After heat treatment, all of the sub-samples were mixed together again into one sample and this material was analyzed for crystalline compounds and specific surface area using XRD and BET.
Evaluation of the properties of a paint based on recycled pigment material
To evaluate the potential of the pyrolysis ash as a pigment, the combined and after-treated material from pyrolysis experiments 2 and 3 was incorporated in two formulations with high pigment volume concentration (PVC) (see Table 2). To facilitate the formulation work, the pyrolysis residue was simplified as a mixture of 45.2 wt% Ti[O.sub.2] and 54.8 wt% dolomite, as these were the major components in the residues. Ti[O.sub.2] in each formulation was substituted with the corresponding amount of ash on a volume basis ([rho][TiO.sub.2] = 4.05 g/[cm.sup.3]; Pdoiomite = 2.85 g/[cm.sup.3]) so that the volume of Ti[O.sub.2] remained constant in standard paint and in the paint based on recycled material. Virgin dolomite was added when needed so that the total volume of extenders remained the same in the compared formulations. The properties of the paints made from recycled material were compared to the properties of standard paints with the same formulations but with virgin pigments and extenders. The experiment was classed as a success if the performance of the varieties with recycled material matched all of the key paint properties of opacity, gloss, whiteness, and durability of the standard paints. These properties were evaluated using test methods that are adapted from ISO tests (see Table 3). Test surfaces painted with paint based on virgin and recycled material, respectively, were also analyzed using a scanning electron microscope (SEM) with energy-dispersive X-ray (EDX) spectroscopic element detection (Hitachi TM 3000 with EDX, Quantax 70).
Results and discussion
Mineralogy and thermal stability of pigments
Three pigments and extenders; Ti[O.sub.2], Kaolin and Dolomite, were identified by XRD as the minerals pure rutile, kaolinite and dolomite, respectively. The extenders Talc and Mica were both shown to contain more than one mineral. The XRD spectrum for Talc identified both talc and clinochlore, as well as potentially a small amount of quartz. The Mica extender components were harder to identify, however the major part seemed to be muscovite along with some kaolinite. Impurities are not uncommon since the extenders are natural minerals. The XRD results are summarized in Table 4.
Results from the thermogravimetric analysis (TGA) of the Dolomite sample showed that this extender is stable at temperatures below 700[degrees]C. At higher temperatures (782 and 878[degrees]C), two endothermic reactions occurred, leading to the formation of magnesium oxide and calcium oxide as expected. (19 20) The XRD spectrum for the residues from the TGA experiments confirmed that the dolomite had decomposed into MgO (periclase) and CaO (lime). No other minerals were identified. Theoretically a sample consisting of pure dolomite should in the first reaction give a mass loss (due to C02 emission) of 23.9% and the second calcination reaction should bring the total mass loss to 47.7%. The TGA experiment showed a mass loss of 21.8% in the first reaction step and 25.6% in the second reaction, resulting in a total mass loss of 47.4%. Since there was no mass loss below 700[degrees]C, it could be concluded that no water was adsorbed on the pigment and no organic or inorganic coatings that are temperature sensitive below 700[degrees]C were present.
The TGA data gave a total mass loss of roughly 1.5% when the Rutile pigment was heated to 1000[degrees]C. According to the literature, rutile is stable up to temperatures near 1800[degrees]C.21 XRD analysis of the Rutile sample before and after the TGA experiments verified that the crystal structure was intact. Burfield 22 reported similar results when heating pigment quality titanium dioxide, with weight losses of 2-3%. This weight loss was found to be due to decomposition of organic surface coatings. Pigments are commonly treated with compounds to improve their dispersion in the paint systems. (21) The nonpigment quality rutile did not contain any surface modifications and showed no weight loss when heated to 1000[degrees]C.
The Kaolin sample underwent an endothermic decomposition starting at 542[degrees]C. A comparison with literature data (24-26) suggests this is most likely due to dehydration of kaolin and the formation of metakaolin. The total mass loss was roughly 13%. This fits well with the loss of 2 water molecules per kaolin molecule during the dehydration, which would give a mass loss of 14%. At 970[degrees]C, a strong exothermic reaction was observed. This was most likely the beginning of the transformation of meta-kaolin into a spinel and amorphous silica. (24-26) However, this could not be verified by XRD. Before the dehydration reaction of kaolin at 542[degrees]C, a small mass loss of around 0.7% was observed in the temperature range of 100-500[degrees]C. This could be due to vaporization of surface water, degradation of a surface coating, or an analytical uncertainty.
The TGA results for the Mica sample were comparable to those found in the literature. (27,28) Muscovite was dehydroxylated over the temperature range of 475-950[degrees]C to a final weight loss of 4-5%. The Mica still had a well-ordered crystal structure after the TGA experiments, but there was no complete agreement between the spectra obtained before and after the TGA experiments, so small changes in the crystal structure have probably occurred. This was not investigated further as these changes were assumed to occur at temperatures above 500[degrees]C, i.e., outside the temperature range for the pyrolysis.
The Talc sample decomposed via endothermic reactions at 607 and 835[degrees]C, followed by an exothermic reaction at 870[degrees]C. According to a review article, (16) the first major reaction during the heating of talc is an endothermic reaction that dehydrates the material, followed by various reactions where different crystalline and amorphous magnesium and silica oxides are formed, depending on the temperature. XRD analysis of the TGA residues showed only noncrystalline components. These are most likely not fully crystallized magnesium oxide and amorphous silica. (16) The transformation of talc into magnesium oxide and silica would give a theoretical mass loss of 4.8% in the form of water vapor. The total mass loss in the present experiment was closer to 9%. This result, as well as the XRD results described earlier, suggested impurities in the talc sample.
The thermal and mineralogical analysis of the pigment samples concluded that the weight losses that occurred at low temperatures are probably due to loss of surface water and decomposition of organic coatings on the pigments. A pyrolysis process at temperatures below 500[degrees]C was thus expected to leave the crystal structure of the major part of the inorganic pigments/extender unaffected.
Thermal stability of binders
The vinyl acrylic latex binder decomposed in two distinct steps at temperatures below 350[degrees]C in nitrogen atmosphere (see Fig. 2). The decomposition occurred gradually (no distinct steps due to low experimental resolution) over a temperature range of 350-500[degrees]C until a residue of 5% of the original sample weight remained. The binder based on pure acrylic polymer decomposed in a single step within the temperature range of 310-400[degrees]C to give a residue weighing 3% of the original sample weight in a nitrogen atmosphere (see Fig. 2).
The thermal decomposition of these binder samples could be studied in greater detail but the aim of this work was mainly to examine the temperature range at which the binder samples would start to decompose into volatile products. The results show that a thermal process, such as pyrolysis at 500[degrees]C, would be sufficient to make most of the binders in waste paint mixtures decompose.
Pyrolysis of the model paint
The pyrolysis experiments were carried out both on wet paint (experiments 1-3) and on semi-dry nonfluid paint (experiment 4) samples. The paint used in experiment 4 had been dried to a dry content of 65% (compared to 44% for the other experiments) prior to the pyrolysis. The weight fraction of the inorganic components (dolomite, kaolin clay, talc, mica, and rutile) of the wet paint make up 26.8 wt% of the total sample (see Table 1). In contrast, the semi-dry paint contains 36.9 wt% inorganic compounds. Theoretically, the dry pyrolysis residue should contain all of the inorganic components in the paint due to the volatile nature of the water and the organics at higher temperature.
The measured and calculated values of the weight of inorganic material remaining after pyrolysis relative to the weight of the original samples (dry fraction) presented in Table 5 show good agreement with the expected values for experiments 1 and 2. In experiment 3, the residue is somewhat greater (29%) but is still in the same range, indicating there was a small amount of organic matter that was not completely pyrolyzed in experiment 3. This could be confirmed as pieces of nonpyrolyzed paint were found in the residues. In experiment 4, the low amount of water in the sample made it difficult to achieve a uniform heating of the material by microwave heating, which resulted in incomplete pyrolysis. Large particles of nondecomposed paint were still present after the pyrolysis despite the experiment being continued for 100 min. Complete volatilization of the water and organic components would have resulted in a dry residue of 36.9%. The weight of solid residue from experiment 4 corresponded to 64% of the original sample weight (see Table 5). This is also an indication of incomplete pyrolysis and shows that this type of microwave pyrolysis is not suitable for the treatment of dry paint or nonfluid paint.
Roughly 90 vol% of the liquid fraction in Table 5 was a transparent liquid (assumed to be mainly water) while 10 vol% was oil. The lighter oil separated from the heavier, water-rich phase due to differences in density. The average effective heating value of the oil produced from the pyrolysis was 26 MJ/kg, which is higher than for wood fuels but significantly lower than that of heating oil. This indicates that the oil fraction may have a value and this will be taken into account in our further work.
Oxidative after-treatment of the dry residue from pyrolysis experiment 1
After the heat treatment in air, the solid residues from pyrolysis experiment 1 had lost 4% of their initial weights. Heat treatment of the solid residue in the presence of oxygen significantly increased the whiteness of the sample, as shown in Fig. 3. The weight change and the change in color are probably due to the oxidation of char carbon to carbon oxides. However, although the whiteness is improved by the heat treatment, the color is still far from the color of a mix containing virgin pigments.
The XRD spectra of the ground pyrolysis residue, before and after oxidative heat treatment, are shown in Figs. 4 and 5. The solid residue from the pyrolysis process mainly consists of the inorganic pigments of the pyrolyzed paint, see Table 1. Rutile and dolomite together make up almost 90 wt% of the total residue, which explains their dominance in the XRD spectrum. The difference in peak heights for some of the rutile peaks in the spectra collected before and after the heat treatment in air is most likely due to a preferred orientation of dolomite crystals and not to a change in mineral concentration during the after-treatment, since only some of the peaks have increased heights.
Based on the XRD results, the crystal structure of rutile and dolomite seem to be unaffected by both the pyrolysis process and the after-treatment in air. This is consistent with the previously presented results from the TGA experiments showing rutile and dolomite as the most temperature-durable pigments of those studied. Since it was shown that the pigments themselves are unaffected by the heat treatment in air, it should be possible to obtain a whiter pigment by oxidizing the remaining residues of carbon-containing compounds in the presence of oxygen or another oxidizing agent.
After-treatment of the dry residue from pyrolysis experiments 2 and 3
The recorded weights of samples of mixed material from pyrolysis experiments 2 and 3, before and after heat treatment in air, are shown in Table 6. In these experiments, the weight loss due to the oxidative heat treatment was roughly 6-7%. The variations in weight loss between samples are likely caused by inhomogeneity in the pyrolyzed material and slight variations in temperature during oxidation.
The XRD spectra for the mixed materials from pyrolysis experiments 2 and 3 after the heat treatment in air showed that dolomite and rutile are intact. This is similar to what was observed for the oxidized residues from pyrolysis experiment 1 (see Fig. 6). The spectra included in Fig. 6 were recorded for the oxidized residues from pyrolysis experiment 1 (upper panel), pyrolysis experiment 2 (middle panel), and pyrolysis experiment 3 (lower panel).
Effect of pyrolysis and oxidative after-treatment on specific surface area
The BET specific surface area data for the mixed residue from experiments 2 and 3 (Table 7) showed that there was no significant difference in the specific surface area before and after the oxidative heat treatment of the pyrolysis residue. A sintering during the oxidation should have given a reduced surface area. This result combined with the XRD results suggests that the heat treatment in air does not affect the pigments remaining after pyrolysis in a negative way. Flowever, the specific surface area of the pyrolysis residue is significantly smaller (a decrease of 33%) than that of the starting pigment mix used in the model paint, showing that an agglomeration of the pigment and extender mixture occurs during the pyrolysis.
Comparison of the performance of paint based on recycled pigments with paint based on virgin pigments
Figure 7 shows two surfaces of a paint formulation with a dry film thickness of 100 pm. The surface in the upper panel is the standard formulation containing only virgin pigments and extenders while the surface shown in the lower panel is a formulation containing recycled pigments, i.e., the inorganic residue from pyrolysis of the model paint and oxidative heat treatment of the pyrolysis ash. The standard paint is a paint that is used as a reference in paint performance evaluations. To facilitate the formulation work, the pyrolysis residue was simplified as a mixture of 45.2 wt% Ti[O.sub.2] and 54.8 wt% dolomite, as these were the major components in the residues. Ti[O.sub.2] in each formulation was substituted with the corresponding amount of ash on a volume basis ([rho]Ti[O.sub.2], = 4.05 g/[cm.sup.3] [P.sub.doiomite] = 2.85 g/[cm.sup.3]) so that the volume of Ti[O.sub.2] remained similar in the standard paint and the paint based on recycled material. Virgin dolomite was also added when needed so that the total volume of extenders remained the same in the compared formulations.
The surface painted with the test paint containing recycled components was also studied using an SEM with element mapping capability (SEM-EDX) in order to investigate what had caused the formation of lumps. A typical result is shown in Fig. 8. Only the EDX map for Ti is shown but maps for other elements were recorded and evaluated as well.
The photographs in Fig. 7 clearly show that the recycled pigment material was not fully dispersed into the paint system, resulting in the film defects (lumps) seen in the lower half of the figure. This effect was seen even when increased dispersant levels were added to the formulation. As was shown in the previous section, the specific surface area of the recycled pigment mix is 33% lower than that of a similar mix made from pigment grade Ti[O.sub.2] and virgin extenders (Table 7), which indicates some agglomeration of the mineral particles. The SEM-EDX results (Fig. 8) show that agglomeration of particles with high Ti content, i.e., probably rutile particles, during the pyrolysis is one reason for the rough paint surface.
However, it is possible that the Ti[O.sub.2]-rich lumps have been formed in the paint due to inferior dispersion. Destruction of the surface coating on the Ti[O.sub.2] particles could cause such agglomeration. The nature of the surface coating applied to the specific rutile pigment used here was not stated in the data provided by the manufacturer. Both inorganic coating, based on alumina or silica, and organic coatings are commonly used in the pigment industry. Since the pigment rutile used in this work showed a weight loss during the thermogravimetric experiments, it is highly probable that it had an organic coating and that this coating was destroyed in the pyrolysis. The reasons for the poor dispersion and the agglomeration of particles during pyrolysis shown in these preliminary recycling tests will be investigated in more detail in our continued work. Milling the pigments before incorporation into the mill base might also reduce some of the agglomeration in the final coating. In further studies, the particle size will be closely monitored before and after each step in the process.
To help describe the test paint results in terms of color and other related properties, the expected measurement values for different paint qualities are shown in Table 8. Table 9 presents the actual results for these quality parameters based on the measurement results. Substitution of the recycled materials for the virgin pigments has impacted the product performance by changing it from a Mid-Tier Matt performance to a Low-Tier Matt performance in terms of whiteness and yellowness. The point of interest that is highlighted by these results is that the effect on whiteness is more significant than the effect seen on the yellowness. Although whiteness and yellowness are generally secondary paint properties, an indicative change in acceptable results for these in absolute terms would be a whiteness decrease of 1-1.5 U and a yellowness increase of 0.3-0.5 U.
The true extent of performance impact of the recycled material is difficult to be determined based only on the whiteness and yellowness range. The results of the key properties can be expressed as follows:
* Opacity-a false positive was observed based on the gray hue of the paint, which influenced the spreading rate calculations (this method includes absorption and scatter contributions, so the color will have an impact on the observations).
* Gloss-equivalent for both formulations (the film defect problem is hidden by matt formulations).
* Whiteness and yellowness index-a significant drop in paint whiteness (a large impact because of the gray nature of the recycled Ti[O.sub.2]) but equivalent yellowness.
* Durability-slightly reduced performance (probably caused by the increased dispersant levels).
To summarize, these results show that the difference in color between the pigment mixes containing only virgin materials and the mix containing recycled pigments and extenders causes false positives in some of the paint testing results. The gray color impacts the opacity results as the calculation is based on the coverage over a white and black area, which can be impacted by paint color.
From these experimental results, it can be concluded that recycled pigments treated as described above cannot be used as a direct replacement in paint formulations. However, before the material can be approved or rejected for use in paint formulation, the issues of poor dispersion and color change need to be resolved. To ensure the pyrolysis residues have been properly used, the dispersant demand needs to be investigated to ensure that the correct level and type of dispersant is used when formulating with this material. The level and type of dispersant needed will be dependent on the surface chemistry of the recycled pigment.
Based on the quality issues discussed above, is it hard to judge if there is any monetary benefit of using a Ti[O.sub.2]-containing pigment mix produced from the processes described in this paper to virgin pigments. A life-cycle assessment of this process to evaluate potential environmental and economic gains will be done at a later stage but initial calculations suggest that the price per kg of recycled Ti[O.sub.2] could be around 10-50% of the price of virgin Ti[O.sub.2] based on today's energy and pigment prices. Further optimization of the process could give increased financial benefits, but at the same time the cost will increase as the pigments need to be further processed, such as by milling and separation of Ti[O.sub.2] from the extenders.
The aim of this work was to investigate if micro-waveheated pyrolysis could be used to recover and reuse the inorganic pigments and extenders in waste paint. A simulated waste paint was manufactured based on a common formulation including Ti[O.sub.2] in the rutile form together with dolomite, kaolin, mica, and talc.
Initial TGA experiments showed that a pyrolysis process at temperatures below 500[degrees]C should leave the crystal structure of the inorganic pigments/fillers unaffected, whereas the organic binder should be fully decomposed. The rutile pigment has an organic surface coating, however, which will be decomposed as well.
The microwave-heated pyrolysis process proved to be quite difficult to control and to get to work effectively, especially when a semi-dry (65% dry matter) paint was treated. It was also concluded that the pyrolyzed paint residue was too dark in color, since it contained uncombusted elemental carbon, and that a second oxidative heat treatment had to be added. The oxidation step improved the whiteness of the residue significantly without affecting the main crystal structure of the inorganic components.
When the heat-treated pyrolysis residues were used in paint formulations as pigment replacements, a significant drop in paint whiteness was observed but the opacity, gloss, and durability were almost equivalent to standard paint. A serious issue identified, however, was an inferior dispersibility of the recycled mix of pigments and extenders in the paint. This gave the painted surface an unwanted, rough texture that was mainly due to agglomeration of Ti[O.sub.2] particles. The conclusion reached was that the recycled material showed promise as a pigment but further work is needed to make it possible to incorporate recycled material into paint formulations in an effective way.
In addition, due to the difficulties in controlling the temperature distribution in the microwave-heated pyrolysis unit and the difficulties in treating dry paint, the microwave heating method will be replaced by an electrically heated oven in our further work. The possibility of a surface treatment to be added to the recovered pigments to improve their dispersibility and the effect of thermal treatment on the pigment particle size will also be investigated.
M. C. F. Karlsson (53), B.-M. Steenari Industrial Materials Recycling, Department of Chemical and Biological Engineering, Chalmers University of Technology, Kemivagen 4, 412 96 Goteborg, Sweden e-mail: firstname.lastname@example.org
D. Corr ICI Paints/AkzoNobel, Building 183 Lab 305, Wexham Road, Slough SL2 5DS, UK
C. Forsgren STENA METALL AB, Fiskhamnsgatan 8 D, Box 4088, 400 40 Goteborg, Sweden
Acknowledgments This work has been funded by Vinnova, Akzo Decorative Paints UK, and Stena Metal AB Sweden, and this is gratefully acknowledged. Dr. Stellan Holgersson, Chalmers University of Technology; Dr. Christopher Knee, Chalmers University of Technology; and Kim Henriksson, Stena Recycling International AB are all acknowledged for their assistance. The authors are also grateful to Dr. Phil Taylor at the Akzo Decorative Paint Slough facility lab for help with formulation and production of the test paints.
(1.) USG Survey, Mineral Commodity Summaries 2012. U.S. Geological Survey, Reston, 2012
(2.) Imam, MA, Froes, FH, Housley, KL, "Titanium and Titanium Alloys." In: Kirk-Othmer Encyclopedia of Chemical Technology. Wiley, New York, 2000
(3.) TZMI, Introduction to the Titanium and Zirconium Value Chains, 2013
(4.) Patton, TC, Lewis, PA, Pigment Handbook. Vol. 1: Properties and Economics. Wiley, New York, 1988
(5.) Fisher, J, Egerton, TA, "Titanium Compounds, Inorganic." In: Kirk-Othmer Encyclopedia of Chemical Technology. Wiley, New York, 2000
(6.) Jiannis, S, Kougoulis, RK, Walsh, B, Revision of EU European Ecolabel for Indoor and Outdoor Paints and Varnishes, 2012
(7.) RebornPaints, 2014. http://www.rebornpaints.co.uk/. Accessed 5 May 2014
(8.) CommunitePaints, 2014 http://www.communityrepaint.org.uk/. Accessed 5 May 2014
(9.) SAKAB. Om farligt avfall, 2014. http://www.sakab.se/sites/ default/files/attachment/vitbok_om-farligt_a vfall.pdf. Ac cessed 3 June 2014
(10.) Wang, Y, Namioka, T, Yoshikawa, K, "Research on Gasification and Power Generation from Waste Paint Dregs." J. Environ. Eng., 3 (1) 182-191 (2008)
(11.) Muniz, LAR, et al., "A Study of Paint Sludge Deactivation by Pyrolysis Reactions." Braz. J. Chem. Eng., 20 63-68 (2003)
(12.) Saft, R, "Life Cycle Assessment of a Pyrolysis/Gasification Plant for Hazardous Paint Waste." Int. J. Life Cycle Assess., 12 (4) 230-238 (2007)
(13.) Tukker, A, "Life Cycle Assessments for Waste, Part III: The Case of Paint Packaging Separation and General Conclusions: Strategic EIA for the Dutch National Hazardous Waste Management Plan 1997-2007." Int. J. Life Cycle Assess., 5 (2) 105-112 (2000)
(14.) Nakouzi, S, et al., "A Novel Approach to Paint Sludge Recycling: Reclaiming of Paint Sludge Components as Ceramic Composites and Their Applications in Reinforcement of Metals and Polymers." J. Mater. Res., 13 (1) 53-60 (1998)
(15.) JCPDS-ICCD:PDF-4, Joint Committee of Powder Diffraction Standards, 2013.
(16.) Wesolowski, M, "Thermal Decomposition of Talc: A Review." Thermochim. Acta, 78 (1-3) 395M21 (1984)
(17.) Urwin. D, "Surface Area and Porosity of Titanium Dioxide Pigments." J. Oil Colour Chem. Assoc., 52 (8) 697-710 (1969)
(18.) Andersson, M, et al., "Microwave Assisted Pyrolysis of Residual Fractions of Waste Electrical and Electronics Equipment." Miner. Eng., 29 105-111 (2012)
(19.) Samtani, M, Dollimore, D, Alexander, KS, "Comparison of Dolomite Decomposition Kinetics with Related Carbonates and the Effect of Procedural Variables on Its Kinetic Parameters." Thermochim. Acta, 392-393 135-145 (2002)
(20.) Gunasekaran, S, Anbalagan, G, "Thermal Decomposition of Natural Dolomite." Bull. Mater. Sci., 30 (4) 339-344 (2007)
(21.) Clark, RJH, Bradley, DC, Thornton P, The Chemistry of Titanium. Pergamon Texts in Inorganic Chemistry, 990125710-7; 19. Pergamon Press, Oxford, 1973.
(22.) Burfield. DR, "Compositional Analysis of Waterborne Paint Systems by Thermogravimetry." Thermochim. Acta, 106C 79-91 (1986)
(23.) Day, RE, "The Characterisation of the Surface of Titanium Dioxide Pigments." Prog. Org. Coat., 2 (4) 269-288 (1974)
(24.) Brindley, GW. Nakahira, M, "The Kaolinite-Mullite Reaction Series: II, Metakaolin." J. Am. Ceram. Soc., 42 (7) 314-318 (1959)
(25.) Brindley, GW. Nakahira. M, "The Kaolinite-Mullite Reaction Series: III, The High-Temperature Phases." J. Am. Ceram. Soc., 42 (7) 319-324 (1959)
(26.) Ptacek, P, et ah, "Isothermal Kinetic Analysis of the Thermal Decomposition of Kaolinite: The Thermogravimetric Study." Thermochim. Acta, 501 (1-2) 24-29 (2010)
(27.) Guggenheim, S, Chang, YH, Koster Van Groos, AF, "Muscovite Dehydroxylation: High-Temperature Studies." Am. Mineral., 72 (5-6) 537-550 (1987)
(28.) Mazzucato, E, Artioli, G, Gualtieri, A, "High Temperature Dehydroxylation of Muscovite-2Ml: A Kinetic Study by In Situ XRPD." Phys. Chem. Miner., 26 (5) 375-381 (1999)
(29.) ISO 6504-3:2006, Paints and Varnishes-Determination of Hiding Power-Part 3: Determination of Contrast Ratio of Light-Coloured Paints at a Fixed Spreading Rate, 2009
(30.) ASTM E313, Standard Practice for Calculating Yellowness and Whiteness Indices from Instrumentally Measured Color Coordinates. American Society for Testing and Materials, West Conshohocken, PA, 2005.
(31.) ISO 2813:1994, Paints and Varnishes-Determination of Specular Gloss of Non-metallic Paint Films at 20 Degrees, 60 Degrees and 85 Degrees, 1994.
(32.) ISO 11998:2006, Paints and Varnishes-Determination of Wet-Scrub Resistance and Cleanability of Coatings, 2006
(33.) ISO/CIE 11664-6:2014 (CIE S 014-6/E:2013), Colorimetry-Part 6: CIEDE2000 Colour-Difference Formula, 2014.
Table 1: Composition of the model paint used in the pyrolysis experiments Type of Wt% total Wt% pigment component Component formulation mixture Solvent Water 32.0 Additive Surfactants 0.7 Coalescent aid 0.6 Biocide 0.2 Organic polymer hiding additive 5.0 pH modifier 0.1 Cellulose rheology modifier 0.7 Rheology modifier 0.2 Binder Acrylic copolymer dispersion 33.6 Pigment Dolomite 11.1 41.3 Kaolin 1.2 4.5 Talc 1.2 4.6 Mica 1.2 4.3 Rutile 12.1 45.2 Table 2: Paint formulations used (expressed in wet volume in 100 mL of paint) Material High PVC formulation 1 High PVC formulation 1 with virgin material with pyrolysis ash Latex binder (50% 5.3 5.3 solids) Virgin Ti[O.sub.2] 1.5 0.0 Ti[O.sub.2]/doiomite 0.0 4.3 (recycled) Virgin dolomite 7.4 4.6 Other virgin 18.6 18.6 extenders Additives 0.6 0.6 Water 66.6 66.6 Material High PVC formulation 2 High PVC formulation 2 with virgin material with pyrolysis ash Latex binder (50% 4.4 4.4 solids) Virgin Ti[O.sub.2] 0.8 0.0 Ti[O.sub.2]/doiomite 0.0 2.4 (recycled) Virgin dolomite 2.1 0.5 Other virgin 19.4 19.4 extenders Additives 1.0 1.0 Water 72.3 72.3 PVC pigment volume concentration Table 3: Methods used to evaluate the paint produced from recycled material Test method Standard References Opacity (98% spreading rate) ISO 6504-3 29 Whiteness ASTM E313 30 Yellowness ASTM E313 30 Gloss (85[degrees]) ISO 2813 31 Durability (scrub resistance) ISO 11998 32 Color difference (dE20oo) ISO 11664-6 33 Table 4: Results from the identification of minerals in the pigments and extenders used in this work Name of Compound pigment identified sample by XRD Chemical formula Dolomite Dolomite CaMg[(C[O.sub.3]).sub.2] Kaolin Kaolinite [Al.sub.2][Si.sub.2][O.sub.5][(OH).sub.4] Talc * Talc [Mg.sub.3][Si.sub.4][O.sub.10][(OH).sub.2] Clinochlore [(Mg,Al).sub.6][(Si,Al).sub.4][O.sub.10] [(OH).sub.8] Quartz Si[O.sub.2] Mica * Kaolinite [Al.sub.2][Si.sub.2][O.sub.5][(OH).sub.4] Muscovite K[Al.sub.2]([Si.sub.3]Al)[O.sub.10][(OH,F) .sub.2] or K[KAl.sub.2][(SiAl).sub.2] [O.sub.10][(OH).sub.2] TiO.sub.2] Rutile Ti[O.sub.2] Samples marked with (*) are likely to contain some minor impurities Table 5: Weight of paint samples and pyrolysis products Sample Experiment 1 Experiment 2 wt (g) wt% wt (g) wt% Sample, prior to heating 241 100 1450 100 Liquid fraction 112 47 850 59 Dry fraction 66 27 398 27 Gas fraction (calculated) 63 26 202 14 Sample Experiment 3 Experiment 4 wt (g) wt% wt (g) wt% Sample, prior to heating 1624 100 245 100 Liquid fraction 1021 63 52 21 Dry fraction 468 29 157 64 Gas fraction (calculated) 135 8 37 15 Table 6: Weights of mixed solid residues from pyrolysis experiments 2 and 3 before and after heat treatment at 450[degrees]C for 180 min in air Weight before Weight after Remaining material heat treatment (g) heat treatment (g) after oxidation (wt%) 50.31 47.12 93.7 49.05 45.59 92.9 52.15 48.74 93.5 50.40 47.07 93.4 54.28 50.73 93.5 64.64 60.51 93.6 61.71 57.90 93.8 67.85 63.79 94.0 62.33 58.49 93.8 50.36 47.48 94.3 Table 7: Result from BET measurements made in triplicates on pigments and mixed pyrolysis residues from experiments 2 and 3 before and after heat treatment in air at 450[degrees]C for 180 min and a mix of virgin pigments corresponding to the pigments in the pyrolyzed paint Sample BET ([m.sup.2]/g) Average STD Pyrolysis residue 6.33 0.11 Heat-treated residue 6.53 0.27 Pigment mix 9.71 0.31 Table 8: Simplified overview of the expected whiteness and yellowness of different formulation types Expected Expected Ti[O.sub.2] whiteness yellowness Formulation type content range range Silk/soft sheen/premium matt High 91-86 1.0-2.0 Mid-Tier Matt Moderate 86-82 2.0-3.0 Low-Tier Matt Low 82-78 3.0-4.0 Standard data provided by AkzoNobel Decorative Paints Table 9: Results of paint parameters for paint based on virgin and recycled material Method Opacity (98% at spreading rate) ISO 6504-3 ([m.sup.2]/l) Whiteness ASTM E313 Yellowness ASTM E313 Gloss (85[degrees]) (GU) ISO 2813 Durability (scrub resistance) ISO 11998 at 200 cycles ([micro]m) Color difference ([dE.sub.2000]) ISO 11664-6 High PVC High PVC formulation 1 formulation 1 with virgin with pyrolysis material ash Opacity (98% at spreading rate) 8.3 11.0 ([m.sup.2]/l) Whiteness 83.15 78.11 Yellowness 2.88 3.34 Gloss (85[degrees]) (GU) 14.9 12.3 Durability (scrub resistance) 4.5 6.3 at 200 cycles ([micro]m) Color difference ([dE.sub.2000]) N/A 1.1 High PVC High PVC formulation 2 formulation 2 with virgin with pyrolysis material ash Opacity (98% at spreading rate) 5.6 7.6 ([m.sup.2]/l) Whiteness 80.90 74.75 Yellowness 3.50 3.80 Gloss (85[degrees]) (GU) 10.4 9.6 Durability (scrub resistance) 60 80 at 200 cycles ([micro]m) Color difference ([dE.sub.2000]) N/A 1.6
Please note: Some tables or figures were omitted from this article.
|Printer friendly Cite/link Email Feedback|
|Author:||Karlsson, Mikael C.F.; Corr, Daniel; Forsgren, Christer; Steenari, Britt-Marie|
|Publication:||Journal of Coatings Technology and Research|
|Date:||Nov 1, 2015|
|Previous Article:||On the role of water, temperature, and glass transition in the corrosion protection behavior of epoxy coatings for underground pipelines.|
|Next Article:||Comparison and characterization of biocompatible polymer coatings on AISI 316L stainless steel.|