Use of Electric Arc Furnace Dust (EAFD) in Alkali-Activated ... - ISWA

Use of Electric Arc Furnace Dust (EAFD) inAlkali-Activated Fly Ash BinderBruno Pavão 1 , Universidade Federal do Rio Grande do Sul (Brasil)Professor, PhD, Alexandre S. de Vargas, Centro de Estudos Superiores Feevale (Brasil)Professor, PhD, Angela B. Masuero, Universidade Federal do Rio Grande do Sul (Brasil)Professor, PhD, Denise C. C. Dal Molin, Universidade Federal do Rio Grande do Sul (Brasil)Professor, PhD, Antônio C.F. Vilela, Universidade Federal do Rio Grande do Sul (Brasil)EXECUTIVE SUMMARYThe scenario of increasing fly ash (FA) production in south of Brazil from mineral coal-fired inpower plants motivates the study of new alternatives for utilization of this by-product. Furthermore,Electric Arc Furnace Dust (EAFD) production in steel industries is also an environmental problemand alternatives for its disposal are needed. This study proposes the alkali-activation of fly ash as aneco-efficient binder and evaluates the mechanical properties and microstructure effects of EAFDaddition to its matrix.Class F ASTM, 1998 fly ashes used in this study had low calcium content, were predominantly inthe vitreous phase and had some crystalline inclusions of mullite, hematite and quartz. The alkaliactivatorused was sodium hydroxide 98% pure commercial type (NaOH). Alkali-activationcharacteristics for the region were studies before, and best compressive strength results wereobtained using a molar ratio Na 2 O/SiO 2 of 0.4 – main source of Na 2 O is the alkali solution ofNaOH, while fly ashes are the source of SiO 2 . Hygrotermic curing was set at 70°C during a periodof 24 hours and is necessary for accelerating strengthening reactions. EAFD used as addition wasobtained from a semi-integrated steel mill industry. In mortars, quartz sand was used as fineaggregate, divided in 4 granulometric bands (1.2; 0.6; 0.30; and 0.15 mm), representing each ¼ oftotal aggregate mass.Compressive strength test were carried out on mortars and XRD and FTIR analysis were performedon pastes samples. EAFD additions were made with 5%, 15% and 25% of fly ashes mass, andmortars were prepared at a 1:3 ratio (fly ash:sand)The influences of different addition ratios in compressive strength results indicate the following:5% EAFD increases strength of samples some 10% to 17% comparing with reference samples of0% addition; higher compressive strength results were obtained with 15% EAFD addition, whichpresented a maximum value of 21.15 MPa obtained at 180 days; samples with 25% of EAFD haveshown that this ratio exceeds the limit for dust addition, presenting negative influence after the ageof 180 days comparing to reference samples.XRD spectres of alkali-activated samples at the age of 1 day show that crystalline phases wereformed after curing at 70 °C temperature. Crystalline peaks of albite were not identified at the ageof 1 day in samples containing EAFD, and natron peaks tend to decrease as EAFD addition ratioincreases. Characteristic peaks of EAFD (quartz, mullite and hematite) are identified in all samplesand also tend to decrease with the addition ratio of EAFD. Same characteristics are identifiedthrough different ages of 28 and 180 days.1 Contact: bpavao@civil.uminho.pt

FTIR analysis shows that the process of alkali-activation is successful in all samples, regarding thecharacteristic differences with the fly ash in its original state. Fly ash band at 1084 cm -1 shifts toaround 1000 cm -1 in samples with EAFD addition, demonstrating no interference in polymerizationand formation of aluminosilicate gel. No wavelength difference was found between reference andEAFD added samples through different ages.Results demonstrate potentiality of binder proposed, especially for precast concrete production.Low carbon emission to atmosphere, sustainable characteristics, reduced energy demand and theuse of wastes and by-products are valuable positive aspects. Microstructural XRD analysis suggeststhe occurrence of reactions with the crystalline EAFD material. FTIR analysis indicates that theadditions have no interference in polymerization and aluminossilicate gel formation.INTRODUCTIONPortland cement, a material of choice for over the last centuries, praised for its mechanicalperformance, competitive price, and versatility, is also an issue of concern for its environmentalimpact. Resource and energy consumption and carbon dioxide emission are problems faced bycement industry. Production and study of environmental friendly binders are therefore needed, andalternative materials are listed as choice options for tackling the problem.Alkali-activated binders have been objective of several studies over the last half century(PACHECO-TORGAL et al., 2008) and are matter of interest in scientific communities for its ecoefficientproperties and the possibility of use of different types of industries by-products and wastesas source material, some of them already in use as additions to Portland cement (BUSCHWALDand SCHULZ, 2005). Alkali-activation briefly consists of submitting a source material withaluminates and silicates into an alkali aggressive environment in such way that the former existingbonds change structure into an alunimossilicate gel that provides mechanical strength. Thesebinders are produced by mixing amorphous materials rich in SiO 2 and Al 2 O 3 and submitting them toa strong alkaline medium, using activators such as NaOH, KOH and Na2SiO4. The alkalinemedium breaks the covalent bonds Si – O – Si – and Al – O – Al of the precursor material andpromotes the polycondensation of the Si – O – Al type into polymeric chains, forming sodiumaluminosilicate gel, which is responsible for the durability properties of this material (PALOMO etal., 1999). One such possible alternative is cement obtained from alkali-activated fly ashes (FA), asproposed for the materials in the south of Brazil in this and in previous work (VARGAS, 2006).The scenario in south of Brazil stimulates researching alternative uses for thermoelectric byproducts:the region possesses about 89% of the mineral coal reserves in the country, a total ofabout 28.8 billion tons. Nevertheless, these resources present a low calorific coefficient of 2600 to3200 kcal/kg and a high level of fly ashes production – 1 kg of fired coal generates approximately 1kWh and 400 g of fly ashes. Currently, coal-fired electrical power plants generate about 2 milliontons of fly ashes in the region each year. Of this total, 20% to 30% is used by cement and concreteindustries in the region, and the rest is disposed into coal strip mining trenches, which may causeextensive environmental damage, such as air, soil and groundwater contamination. Considering theprojects of new thermoelectric power plants, coal extraction in the region is predicted to increasefrom 1.6 to 4.2 million tons per year (ROHDE, 2006; VARGAS, 2006)Produced in steel mills, electric arc furnace dust (EAFD) is classified as a hazardous waste, notdisposable in the environment, according to the Resource Conservation and Recovery Act of theU.S. Environmental Protection Agency (UNITED STATES OF AMERICA, 2008). Its generalcomposition contains heavy metals such as Lead (Pb), Chromium (Cr) and Cadmium (Cd).

Techniques proposed for recycling these metals are still expensivee in the region, and further studiessof processes for immobilizing thematerial are still needed.Several alternative matrices were proposedfor immobilizing toxic waste. Palomo and Palacios(2002) evaluated the stabilisation/solidification capacityof alkali-activated flyash cementing matrixin the presence of toxic elements, and leaching tests carried out indicated that this new binder wasable to stabilise and solidify lead in a veryefficient way. Deja (2002) studied the properties ofalkali-activated slag pastes in the presence of Zinc (Zn), Cd, Crand Pb ions and found that thedegree of immobilization was very high (exceeding 99.9%). Shi and Fernanández-Jiménez (2006),reviewing the progresses in use of alkali-activatedbinders for stabilization/solidification ofhazardous and radioactive wastes, concluded that the leachabilityof contaminants in these bindersis lower than that from hardened Portlandcement stabilized wastes. Palomo and Fuente (2002)show that boron does not significantly alter the hardening processs of alkali-activated flyashes, thattits presence hardly modifies mechanical strength and, taking leaching resultss into account, that thealternative system was more effective thantraditionalones. Vargas (2002) studied theeffects ofEAFD additions in Portland cement concrete blocks and, although some heavy metals wereeimmobilized, the setting time of the binder was severelyaltered bythe presence of zinc.The purpose of this study was toevaluate the effects of three EAFD addition ratios to a previousstudied binder made of alkali-activated fly ashes (VARGAS,2006). Mechanical influence ofadditions was determined in compressivestrength tests and microstructurecharacterization wasmade usingXRD analysis and FTIR wavelength comparisons.EXPERIMENTALMaterialsClass F (ASTM, 1998) fly ash generated bya coal-firedpower plant located in southernBrazil wasused in this study. Ashes had low calcium content, were predominantly in the vitreousphase andhad some crystalline inclusions of mullite, hematite and quartz. XRD spectracontainingcrystallinepeaks are shown in Figure 1 and physical and chemical characterization of flyash in Table 1.Figure 1: XRD spectra of fly ash; Q = Quartz; M = Mullite; H = HematiteThe alkali-activator used was sodium hydroxide 98%pure commercial type (NaOH)with 2.13g/cm³ of density.Table 1: Chemical compositionof fly ash (%of mass)I.R. a L.O.I bSiO 2 Al 2 2O 3 CaO Fe 2 O 3 Na 2 O TiO 2 MgO K 2 O SO 4(%) (%)63.09 24. 02 1.15 6.85 0.211.77 0. 81 1.75 0.34 97.0aI.R. insolubleresidue.0.02Specific density c(g/cm 3 )2.17

L. .O.I. loss onignition.c NBR NM 23 (ABNT, 2000).EAFD used as addition was obtained from a semi-integrated steel mill industry, collected by filtersoutside theelectric arch oven. Chemical composition of the material indicates majority of Iron andZinc, according to Table 2. XRD spectraof EAFD present crystalline peaks of quartz (SiO 2 ),hematite ( Fe 2 O 3 ), franklinite (Fe 2 O 3 ZnO), zincite (ZnO) and cromite (FeO.Cr 2 O 3 ), asshown inFigure 2.Table 2: Chemical compositionof EAFD (%of mass)Na K Fe Mg CaPb Si Mn CdCr Zn3.35 1.0827.30 1.87 2.540.90 1.70 2.06 0.37 0. 13 37.25Figure 2: XRD spectra of EAFDIn mortars,quartz river sand wasused as fine aggregate, divided in4 granulometric bands (1.2; 0.6;0.30; and 0.15 mm), representingeach ¼ of total aggregate mass.Analysis of metal concentrationin leachingsamples, obtained according to a Brazilianproceduree(ABNT, 2004b), is presented in Table 3. Results of Cdand Pb concentrations are above the indexlimit (ABNT, 2004a),thus classifying EAFD as “Class I” hazardous waste - not disposable in theenvironment.TablElementConcentration (mg/l)Limit (ABNT, 2004a)nd: not defineddle 3: EAFDZn913ndD leaching teNa565ndest results (mCdmg/l)Cr6.9 0.0020.5 5Pb Cr+661 < 0.011 ndMethodsExperimental work was made to verify the influence of EAFD additions to alkali-activated fly ashmortar samples, evaluating compressive strength. Paste samples were submitted to XRD analysissand FTIR spectroscopy for wavelength comparisons.Alkali-activation characteristicsfor the region were studied before (VARGAS, 2006), and bestcompressive strength results were obtained using a molar ratio Na2 O/SiO 2 (N/S) of 0.4 – in whichhthe main source of Na 2 O is the alkali solution of NaOH, while fly ashes are the source of SiO 2 .Hygrotermic curing was set at70°C during a period of 24 hours necessary for acceleratingstrengthening reactions.

(a)(b)Figure 3: (a) Compressivee strength (MPa); (b) Strength gain(%)Microstructural AnalysisXRDXRD spectres of alkali-activatedsamples at the age of 1 day (Figure 2) show that crystalline phasesswere formed after curing at 70 °Ctemperature. Crystalline peaks of albite were not identified at theage of 1 day in samples containing EAFD,and natronn peaks tend to decrease as EAFD additionratio increases. Characteristic peaks of EAFD (quartz, mullite and hematite) are identified in alllsamples and also tend to decrease with the addition ratio of EAFD. Same characteristics wereeobserved at the age of180 days ( Figure 3).Figure 4: XRD spectres of: (a) EAFD; (b) FA; and paste samples at age of1 day with: (c) 0%EAFD; (d)5% EAFD; (e) 15% EAFD; and 25% EAFD

Figure 5: XRD spectres of: (a) EAFD; (b)FA; and paste samples at age of 180 day with: (c) 0%EAFD; (d) 15%EAFD; and(e) 25% EAFDObservation of XRDpeaks ofsamples with 25% EAFD through time (Figure 6) shows noformation of albite. Natron peaks, present inearly age of 1 day, diminishes at the age of 180 days,suggestingg that the carbonate maybe combining with other elements in the matrix.Figure 6:XRD spectres of: (a) EAFD; (b) FA; and paste sampless with N/S 0.4 and 25%EAFD attheages of: (c) 1day; (d) 28 days; (e)91 days and (f) 180 daysFTIRConsidering FA and alkali-activated samples, independently of age of analysis (Figure7 and 8),there are characteristic absorption bands common to fly ash before and after activation, namely3435, 1625, 796, 779, 693, 559 e 461 cm -1 . As observed in FTIRspectra of original flyash bands(Figure 7, sample b) 3450 cm -1 e a 1650 cm-1 frequencies indicatethe presence of water molecules,results that agree with previous studies (BARBOSAet al., 2000) that identified the presence ofwater in original metakaolin and in metakaolin alkali-activated samples.During alkali-activation, bands near 3450 cm-1 and 1650 cm -1 are more intense than original fly ashbands, as observed in figures 7 and 8. Bandnear 3450 cm -1 is related to vibration and stretching ofwater molecules H-O-H, and band near 1650 cm -1 isrelated to angular deformationsin H-O-HHbonds (PALOMO et al., 1999).

Figure7: FTIR Spectra of: ( a) EAFD; (b) FA; paste samples with 0% EAFD at: (c)1 day; (d)28 days; (e) 91 days; (f) 180days; paste samples with 15% EAFD at: (g) 1 day; (h) 28 days;(i) 91 days; (j) 180 daysFigure 8: FTIR Spectra of: (a) EAFD; (b) FA; paste samples with 180 days with: (c) 0%EAFD (d)5% EAFD; (e) 15% EAFD; and (f) 25%EAFDAbsorptionbands near 796 cm -1 e a 693 cm-1 are related to the presence of aluminium and quartz,respectively. Bakharev(2005) proposed thatt the band near 800 cm-1 is relatedto vibrations in AlO and that bands near 688 cm -1 4aredue to symmetrical stretching vibrations inSi-O-Si bonds in flyash samples.560 and 460 cm -1 bands were also identified in fly ashsamples before and after alkali-activation.Palomo et al. (1999) obtained similar resultss and proposed that band 560 cm -1 indicates the presenceeof mullite (octahedralformation) ). In XRD analysis these results are confirmed with the presence ofcrystalline peaks of mullite in figures 4, 5 and 6.Characteristic bands cited above (796, 779, 693, 559 e 461 cm -1 ) were identified both in original flyash and in alkali-activated fly ashsamples (Figures 7 and 8). As these bands are related to Si-O-Si eSi-O-Al bonds, FTIRresults confirm XRDanalysis, thus demonstrating thechemical stability ofcrystalline quartz and mullite in a strong alkali environment.Band 10900 cm -1 fromoriginal fly ash is altered to regions near to 1000 cm-1 in alkali activatedsamples. This is proposed to be related with vibrations in Si-O and Al-O bonds, and reducing thefrequency in this region suggests that vitreous components of flyash are reacting withthe alkali-activator and that different products arebeing formed – namely the aluminossilicate gel(FERNÁNDEZ-JIMENEZ et al., 2005).Comparisons of alkali-activatedfly ash samples andsamples with EAFDadditions show nodifference in frequencies of bands in FTIR analysis. These results indicate that alkali-activation issuccessful in all samples and that EAFD additions haveno harmful interference in the formation ofaluminosilicate gel and the in thepolymerization reactions.CONCLUSIONS

Results demonstrate potentiality of binder proposed here, especially for precast concrete production.Low carbon emission to atmosphere, sustainable characteristics, reduced energetic demand and theuse of wastes and by-products are valuable positive aspects for society and scientific communities.Mechanical properties of this binder can be improved using fly ash containing a higher level of SiO 2and using processes for reducing the fly ash mean particle size. Microstructural XRD analysissuggests that albite, natron and characteristic EAFD crystalline peaks decrease intensity accordingto EAFD addition ratio, indicating the occurrence of reactions with the crystalline material. FTIRanalysis indicates that original fly ash is successfully alkali-activated and that no significantdifference is found among the EAFD addition ratios, proving that the additions have no interferencein polymerization and aluminossilicate gel formation.REFERENCESASSOCIAÇÃO BRASILEIRA DE NORMAS TÉCNICAS, ABNT (1996): NBR 7215: cimentoPortland – determinação da resistência à compressão. Rio de Janeiro, 1996 (in Portuguese).ASSOCIAÇÃO BRASILEIRA DE NORMAS TÉCNICAS, ABNT (2000): NBR MN 23: cimentoPortland e outros materiais em pó – determinação da massa específica. Rio de Janeiro, 2000 (inPortuguese).ASSOCIAÇÃO BRASILEIRA DE NORMAS TÉCNICAS, ABNT (2003): NM 43: cimentoPortland – determinação da pasta de consistência normal. Rio de Janeiro, 2003 (in Portuguese).ASSOCIAÇÃO BRASILEIRA DE NORMAS TÉCNICAS, ABNT (2004a). NBR 10004: resíduossólidos - classificação. Rio de Janeiro, 2004a. (in Portuguese).ASSOCIAÇÃO BRASILEIRA DE NORMAS TÉCNICAS, ABNT (2004b). NBR 10005:Procedimento para obtenção do extrato lixiviado de resíduos sólidos. Rio de Janeiro, 2004b. (inPortuguese).AMERICAN SOCIETY FOR TESTING AND MATERIALS, ASTM (1998). “StandardSpecification for Coal Fly abd Raw or Calcined Original Pozzolan for Use as a Mineral Admixturein Concrete”. ASTM C618/98. In: Annual Book of ASTM, V. 04.02, Philadelphia, 1998.Bakharev, T. (2005): Geopolymeric materials prepared using Class F fly ash elevated temperaturecuring. Cement and Concrete Research, v. 35, p. 1224-1232, 2005.Barbosa, V.; Mackenzie, K.; Thaumaturgo, C. (2000): Synthesis and Characterisation of MaterialsBased on Inorganic Polymers of Alumina and Silica: Sodium Polysialate Polymers. InternationalJournal of Inorganic Materials, v. 2, p. 309-317, 2000.Buschwald, A.; Schultz M. (2005): Alkali-activated binders by use of industrial by-products.Cement and Concrete Research 35, pp.968–973.Deja J., (2002): Immobilization of Cr 6+ , Cd 2+ , Zn 2+ and Pb 2+ in alkali-activated slag binders.Cement and Concrete Research, 32 (2002), pp. 1971–1979.Fernández-Jiménez, A.; Palomo, A.; Criado, M. (2005): Composition and microstructure of alkaliactivated fly ash binder: effect of the activator. Cement and Concrete Research. Volume 35, Issue 6,p. 1204-1209, 2005.

Pacheco-Torgal, F.; Castro-Gomes, J.; Jalali, S. (2008): Alkali-activated binders: A review. Part 1.Historical background, terminology, reaction mechanisms and hydration products. Constructionand Building Materials 22, pp. 1305–1314.Palomo A., Palacios M. (2002): Alkali-activated cementitious materials: Alternative matrices forthe immobilisation of hazardous wastes - Part II. Stabilisation of chromium and lead. Cement andConcrete Research, 33(03), pp. 289–295.Palomo, A.; Fuente, J. I. L. (2003): Alkali-activated cementitous materials: Alternative matrices forthe immobilisation of hazardous wastes: Part I. Stabilisation of boron. Cement and ConcreteResearch 33 (2), Feb. 2003, pp. 281-288Palomo, A.; Grutzeck, M.; Blanco, M. (1999): Alkali-acivated Fly Ashes. A Cement for Future.Cement and Concrete Research, v. 29, p. 1323-1329, 1999.Rohde, G.M. (2006): Cinzas de carvão no Brasil: Restrições ambientais e medidas mitigadoras. In:Rohde GM. et al. Cinzas de Carvão Fóssil no Brasil – Aspectos Técnicos e Ambientais, PortoAlegre: CIENTEC, 2006. Chap. 2: 137-195 (in Portuguese).Shi, C.; Fernández-Jiménez, A. (2006): Stabilization/solidification of hazardous and radioactivewastes with alkali-activated cements. Journal of Hazardous Materials 137 (3), 11 October 2006, pp.1656-1663.United States of America, (2008): Environmental Protection Agency. RCRA Orientation Manual2008: Resource Conservation and Recovery Act (2008).http://www.epa.gov/osw/inforesources/pubs/orientat/rom.pdf (Jun. 02, 2009).Vargas, A. (2002). Estudo da viabilidade do uso do pó de aciaria elétrica na confecção de blocosde concreto para pavimentação. 2002. 148 f. Dissertação (Mestrado em Engenharia) - Programa dePós-Graduação em Engenharia Civil, Universidade Federal do Rio Grande do Sul, Porto Alegre (inPortuguese).Vargas, A. (2006): Cinzas volantes álcali-ativadas para a obtenção de aglomerantes especiais.2006. 154 f. Tese (Doutorado em Engenharia) – Programa de Pós-Graduação em Engenharia Civil.Universidade Federal do Rio Grande do Sul, Porto Alegre, 2007 (in Portuguese).