- Email: [email protected]
Preparation of autoclaved aerated concrete using copper tailings and blast furnace slag
Construction and Building Materials 27 (2012) 1–5
Contents lists available at SciVerse ScienceDirect
Construction and Building Materials journal homepage: www.elsevier.com/locate/conbuildmat
Preparation of autoclaved aerated concrete using copper tailings and blast furnace slag Xiao-yan Huang a, Wen Ni a, Wei-hua Cui b,⇑, Zhong-jie Wang a, Li-ping Zhu a a b
State Key Laboratory of High-Efficient Mining and Safe of Metal Mines, University of Science and Technology Beijing, Ministry of Education, Beijing 100083, China School of Water Resources and Environment, China University of Geosciences (Beijing), Beijing 100083, China
a r t i c l e
i n f o
Article history: Received 21 February 2011 Received in revised form 11 July 2011 Accepted 9 August 2011 Available online 20 September 2011 Keywords: Skarn-type copper tailings Autoclaved aerated concrete Reaction mechanism Tobermorite
a b s t r a c t Based on the background that large amount of copper tailings are stockpiled in China, the utilization of skarn-type copper tailings to prepare autoclaved aerated concrete (AAC) was studied. The AAC samples were prepared on a laboratory scale with a dry density of 610.2 kg m 3 and compressive strength of 4.0 MPa. Compared with the traditional AAC, lime was totally substituted by skarn-type copper tailings and blast furnace slag in order to develop a potential technique of reducing CO2 emission during the AAC production process. The samples of different curing stage were examined by XRD, FESEM as well as 29Si and 27Al NMR analyses. Results show that the main minerals in the AAC product are tobermorite-11 Å, anhydrite, augite, quartz, calcite and dolomite, with small amount of other minerals brought in by the copper tailings. It was also suggested that most minerals in the copper tailings participated in the hydration reaction during the procuring process, and the chemical elements in them got into the structure of platy tobermorite in the subsequent autoclaving process. Ó 2011 Elsevier Ltd. All rights reserved.
1. Introduction Copper tailings in China have been nearly totally piled up through the history of copper production. Now more than 2.4 billion tons of copper tailings are estimated to exist in China [1], posing a severe threat to the environmental condition. Skarn-type copper tailing (SCT) is one type of copper tailings to be most difficultly reused in the traditional building materials, due to its nature of very fine grain size and high CaO and MgO content. It has been reported that skarn-type copper tailings can be used to prepare autoclaved brick with good quality [2–4]. But the high transportation costs and low commodity prices restrict the broad commercialization of such brick. Autoclaved aerated concrete (AAC) is a lightweight and highly porous wall material with excellent insulation ability [5,6]. According to Chinese national standard, AAC is the only one type of wall materials owning the ability to meet 50% of the building energy saving request without adding other affiliated thermal insulation materials [7]. The background that Chinese government is promoting building energy saving and carbon emission reduction gives autoclaved aerated concrete a broad application prospect. The commercial AAC is usually produced with cement and lime as calcareous materials, and with quartz sand or fly ash as the siliceous materials. To extend the range of raw materials and lower the pro-
⇑ Corresponding author. Tel.: +86 137200160522; fax: +86 01082321081. E-mail address: [email protected] (W.-h. Cui). 0950-0618/$ - see front matter Ó 2011 Elsevier Ltd. All rights reserved. doi:10.1016/j.conbuildmat.2011.08.034
duction costs, several researchers have investigated the possibility of replacing the traditional raw materials of AAC by industrial waste, such as air-cooled slag [8], coal bottom ash [9], efflorescence sand and phosphorus slag [10], lead–zinc tailings [11] and iron ore tailings [12]. These studies mainly focused on exploring suitable alternative siliceous materials, and few researches involved the investigation into calcareous materials. In this study, skarn-type copper tailings and water-quenched blast furnace slag (BFS) were used as mainly raw materials to prepare AAC. Because the high CaO and MgO content in both SCT and BFS, SCT and BFS was used to replace lime as calcareous material. Besides, SiO2 in the BFS was considered as partly alternative siliceous resource to reduce the consumption of quartz sand. The object of the present work is to investigate the microstructural properties and phase compositions of the AAC prepared by SCT and BFS, and to make primary understanding of reaction mechanism during the process of precuring and autoclaving, especially the behavior of SCT in the hydrothermal reaction. 2. Experiments 2.1. Raw materials The AAC samples were prepared by the following raw materials: SCT, BFS, quartz sand (QS), cement clinker (CC) and natural gypsum. They were all ground in a SMU500 500 type ball mill. The results of chemical analyses and specific surface area tests of the raw materials are listed in Table 1. The specific surface area of natural gypsum is 402.6 m2 kg 1. XRD of the SCT sample shows that the main minerals are augite, dolomite, phologipite, amphibole and calcite, accompanied
2
X.-y. Huang et al. / Construction and Building Materials 27 (2012) 1–5
Table 1 Chemical compositions and specific surface area of raw materials.
3. Results and discussion
Oxide composition (wt%)
SCT
BFS
QS
CC
SiO2 Fe2O3 CaO Al2O3 MgO K2O Na2O IOT Specific surface area (m2 kg
44.52 1.94 13.56 5.36 19.92 1.20 1.00 9.26 656.8
32.7 0.4 38.79 15.4 8.97 0.36 0.23 0.76 562.6
82.83 0.53 1.83 7.13 1.14 2.69 1.55 0.18 792.7
21.98 5.13 60.38 5.54 3.03 2.17 0.25 – 380.6
1
)
3.1. Mechanical properties The result of compressive strength test of the AAC sample and its related raw materials proportions are given in Table 2. Superplasticizer dosage was 0.8% of the total solid mixture and aluminum powder dosage was 0.10%. The water/solid ratio was 0.38. At the designed proportions, the compressive strength of AAC samples could get 4.0 MPa and the dry density of AAC was 610.2 kg m 3. Additionally, the compressive strength of sample in the absolute dry condition (DCS) can reach up to 5.9 MPa, and its corresponding specific strength reaches as high as 9.7.
3.2. XRD analyses XRD analyses were performed to investigate the phase changes in the AAC samples during the process of precuring and autoclaving. The XRD patterns of the dry mixture (DM) of raw materials, hardened aerated concrete (HAC) sample precured under the saturated steam curing for 12 h after being mixed with warm water and the final AAC sample are shown in Fig. 2. As indicated in DM’s spectrum, most minerals in SCT are identified, except talc and clinochlore. The non-detection of these two minerals is likely due to the fact that the overall amount of those minerals originally comprising small amount in SCT are in more minor quantities after addition into the dry mixture with only 30% in weight, so that the XRD is not sensitive enough to allow detection at such low level. Quartz, gypsum and alite from the raw materials of quartz sand, natural gypsum and cement clinker, respectively, were detected as expected. A diffused band existing between 22° and 38° of 2h was caused by the incorporation of BFS with glassy nature. It can be seen from HAC’s spectrum, after being precured around 48 °C for 12 h, the diffraction peaks of alite and gypsum disappeared. The AFt, a new phase, was formed. The broad band at around 17–38° of 2h indicated the existence of C-S-H gels. The appearance of AFt and C-S-H gels and the disappearing of alite
Fig. 1. XRD pattern of skarn-type copper tailings. by minor phases including clinochrysotile, albite, talc and clinochlore (Fig. 1), in accordance with the chemical analysis result that most of the minerals included are rich in CaO and MgO. The XRD analysis of natural gypsum sample shows that the main mineral in it is CaSO42H2O and no other crystalline phases were identified. Aluminum powder was used as a gas producing agent for the slurry foaming, and naphthalene was used as superplasticizer. 2.2. Procedure Firstly, the prepared powder of the raw materials and the superplasticizer were thoroughly dry mixed. Then warm water (48 ± 1 °C) was added and mixed for 2 min. Finally, aluminum powder was added and mixed with the slurry for another 30 s. The obtained slurry was casted into preheated steel molds of 100 100 100 mm to allow it to expand and harden at the temperature of 48 ± 1 °C for 12 h under a steam saturated condition. After their swollen up surface being cut to flat, the samples were demolded and put into an industrial autoclave for hydrothermal reaction for 8 h at 13.5 bars. 2.3. Analysis The bulk density and compressive strength tests were conducted according to GB 11968-2008, ‘‘Autoclaved aerated concrete blocks’’ which specifies that bulk density should be determined by oven dry of 24 h at 60 ± 5 °C, and then another 24 h at 80 ± 5 °C, following by oven dry at 105 ± 5 °C until samples tested reached constant weight, and compressive strength tests were performed at loading rate of 2.0 ± 0.5 kN/s and on samples with moisture content of 8–12%. The X-ray diffraction (XRD) spectra of different samples were obtained with a D/Max-RC diffractometer (Japan) with copper Ka radiation at 30 mA and 50 kV. A step size of 0.02° was selected over a 2h range of 5–70°. The microstructure of the samples under different curing stage was observed with a SUPRA™55field emission scanning electron microscope (FESEM). The fractured surfaces of the samples were coated with carbon prior to examination. MAS NMR tests were conducted in a Bruker Avance III400 spectrometer operating at 59.62 MHz (29Si) and 104.0 MHz (27Al). The rotation frequency was 5 kHz for 29Si and 10 kHz for 27Al, and the delay time was 3 s for 29Si and 1 s for 27Al.
Fig. 2. XRD patterns of DM, HAC and AAC samples.
Table 2 Material proportions and mechanical properties of AAC sample. Mixture composition of AAC (wt%)
Dry density (kg m
SCT
BFS
QS
CC
Gypsum
30
35
20
10
5
610.2
3
)
Compressive strength (MPa)
DCS (MPa)
Specific strength
4.0
5.9
9.7
X.-y. Huang et al. / Construction and Building Materials 27 (2012) 1–5
and gypsum were the results of the hydration reaction between gypsum, clinker and BFS. The intensity reduction of quartz’s characteristic peaks indicated that the fine quartz particles participated in the pozzolanic reaction. Comparing spectra of DM and HAC, the peaks of dolomite, phlogopite, albite, augite and amphibole decreased significantly. That means these crystalline minerals in SCT, which were inert at room temperature, participated in the hydration reaction forming AFt or C-S-H gel during the precuring process. From spectrum of AAC, it can be seen that the major minerals in the final AAC products are tobermorite-11 Å, anhydrite, and some residual minerals including quartz, dolomite, calcite and augite, accompanied by phlogopite, amphibole, clinochrysotile and albite in minor quantities. The changes in peak intensity of minerals from SCT were not clearly detected. That is to say the minerals in SCT were not evidently involved into the hydrothermal reaction during the 8 h autoclaving process. The reason why the half width of tobermorite peaks was broad was that the structure of tobermorite crystals would be complicated by the incorporation of Mg2+, Al3+ and other minor composition from SCT. During the precuring treatment, most minerals from SCT participated in the hydration reaction forming AFt and C-S-H gel, and through autoclaving process, the hydration products of C-S-H gel and AFt were converted to tobermorite and anhydrite. The chemical elements decomposed from SCT were absorbed into the tobermorite crystals through the transfer of AFt and C-S-H gel. The presence of those elements in tobemorite leaded to numerous tobermorite crystals with different lattice parameters and broad half width of the tobermorite peaks in XRD patterns.
3.3. FESEM analyses The surface microstructure of artificial pores in HAC sample (Fig. 3) shows that, except for AFt and C-S-H gel, there was no other crystalline hydration products formed, which was in good accordance with the XRD results in Section 3.2. The SEM images of the AAC sample are shown in Fig. 4. Because of the low water/solid ratio, the micro-morphology of wall section between artificial pores was dense, which can be seen from Fig. 4a. The dense structure had good contribution to high compressive strength of the AAC samples. The surface microstructure of one artificial pore was presented in Fig. 4b. Long-strip shaped anhydrite interpenetrating with platy tobermorite crystals was observed. Combining Fig. 4b with Fig. 3, it can be seen that AFt and C-S-H gel were transformed into tobermorite and anhydrite. Further magnification of Fig. 4b is shown in Fig. 4c. It clearly shows that the platy tobermorite is well crystallized with a thickness of
AFt
C-S-H
Fig. 3. FESEM of hardened aerated concrete sample.
3
about 60–80 nm and a width of 1–2 lm. The platy tobermorite interfinger and overlap with each other forming firm skeleton and empty cavities between the plates. As a result of such microstructure, high compressive strength, good thermal insulation and heat preservation performance of this material would be guaranteed. 3.4.
29
Si and
27
Al NMR analyses
The 29Si NMR spectrum of the final AAC product is given in Fig. 5. As shown in Fig. 5, there is a broad 29Si resonance peak including three subsidiary signals at chemical shift of 79.1 10 6, 80.4 10 6, 81.2 10 6. The signal at 79.1 10 6 can be partly referred to Q2 (0Al) sites of the residual augite with single-chain silicate structure in SCT. But the content of augite is too small to make a peak at 79.1 10 6 so prominent. As we know, in the structure of augite, the negative charges created by Si–O structure are mainly balanced by Ca2+, Mg2+ and Fe2+, while in that of C-S-H gel those charges are balanced by Ca2+ and H+. Due to the fact that the electronegativity of H+ is greater than that of Ca2+, Mg2+ and Fe2+, the chemical shift of Q2 (0Al) sites in augite would have lower field shift than that in C-S-H gel and be approximate to that of Q2 (1Al) sites of C-S-H gel. So, the resonance signal at 79.1 10 6 is probably the overlap of both augite Q2 (0Al) and C-S-H gel Q2 (1Al) tetrahedrons. The signals at 80.4 and 81.2 10 6 can be assigned to Q2 (1Al) sites in tobermorite-11 Å. As discussed in Section 3.2, tobermorite-11 Å is the dominant phase in AAC samples. The reason that there are two subsidiary peaks assigned to Q2 (1Al) was that Ca in tobermorite-11 Å may be substituted by Mg from SCT and BFS. As noted above, these Mg rich minerals have participated in the hydrothermal reaction to produce tobermorite. If the Si–O tetrahedrons were linked by Mg rather than by Ca between tetrahedron chains, the electron density around the related silicon atoms would be getting lower. That is because the electronegativity of Mg is greater than that of Ca. Thus, the signal at 80.4 10 6 may be caused by the Q2 (1Al) sites where Mg ions are the linkages between the chains, and the signal at 81.2 10 6 can be assigned to Q2 (1Al) sites where Ca ions are the linkages. On the right shoulder of the major peak there are many minor subsidiary peaks. Among them, signal at 91.3 10 6 may be indicative of Q3 (1Al) structure state of tobermorite-11 Å [13], indicating that Al also had substituted in Q3 units in tobermorite structure. Compared with the intensity of signals at 81.2 10 6 and 80.4 10 6 assigned to Q2 (1Al) sites, the intensity of signal at 91.3 10 6 assigned to Q3 (1Al) sites is significantly lower, suggesting that the Si–O–Si branching sites between silicate double chains are quite few. Other subsidiary signals from 86.7 10 6 to 96.2 10 6 on the right side of the major peak can be assigned to Q3 (1Al) and Q3 (0Al) sites of residual phyllosilicates including phlogopite, clinochrysotile and amphibole. Evidently, all of the peaks with chemical shifts beyond 96.2 10 6 are very low, suggesting most minerals with Q3 (0Al) and Q4 (nAl) sites, which are quartz and field spar minerals, have been largely consumed. There seems to be a contradiction between the XRD spectrum which show the highest peak belongs to quartz and the 29Si NMR spectrum which shows very low content of residue quartz in the final AAC samples. This is because the crystallinity of quartz is well. Small amounts of quartz can make a very strong peak in the XRD spectrum. The 27Al NMR spectrum of the final AAC product (Fig. 6) shows a dominant resonance at 62.3 10 6 indicative of tetrahedral Al coordination and a lower resonance at 12.7 10 6 indicative of hexahedral Al coordination. This assignment of 27Al suggested that the Al atoms mainly served as substitution of Si atoms, which is in good accordance with the 29Si NMR analysis results. The relatively
4
X.-y. Huang et al. / Construction and Building Materials 27 (2012) 1–5
Fig. 4. FESEM of AAC sample (a: surface of wall section; b: inner surface of one artificial pore; c: further magnification of b).
Fig. 6. Fig. 5.
27
Al NMR spectra of the AAC products.
29
Si NMR spectra of the AAC products.
small amount of hexacoordinated Al atoms in the AAC samples probably existed as charge balancing ions in the silicates. 4. Conclusions The results show that skarn-type copper tailings and blast furnace slag can be used as substitution of lime to produce AAC products by providing CaO and MgO compositions. Using them can reduce the emission of CO2, for the elimination of limestone calcinations process. The AAC product with a dry density of 610.2 kg m 3 and compressive strength of 4.0 MPa was produced by the raw material composition of 30% skarn-type copper tailings,
35% high furnace slag, 10% cement clinker and 5% gypsum. The main minerals in the AAC are platy tobermorite and long-strip anhydrate besides quartz and other residual minerals from the copper tailings. Through the NMR analysis, it could be find that in the tobermorite structure, Ca atoms may be substituted by Mg atoms which were provided by the high furnace slag and minerals from the skarn-type copper tailings, and Al atoms significantly substituted Si atoms in the chain sites as well as chain bridge sites of tobermorite. Acknowledgements The authors gratefully acknowledge Beijing Jinyu AAC Co., Ltd. for providing access to industrial autoclave. We thank Li Hong
X.-y. Huang et al. / Construction and Building Materials 27 (2012) 1–5
and He Jianping of University of Science and Technology Beijing for XRD and FESEM test. We would also like to express thanks to Xiang Junfeng of institute of chemistry of Chinese Academy of Sciences for his assistance in NMR test. References [1] Yu Liang-hui, Jia Wen-long, Xue Ya-zhou. Survey and analysis of the copper tailing resources in China. Metal Mine 2009(8):179–81 [in Chinese]. [2] Zhao Feng-qing, Zhao Jing, Liu Hong-jie. Autoclaved brick from low-silicon tailings. Constr Build Mater 2009;23(1):538–41. [3] Zhao Feng-qing, Ni Wen, Wang Hui-jun. Preparation of load-bearing brick with copper tailings by steam curing method. Express Inf Min Ind 2006(4):34–6 [in Chinese]. [4] Zhao Feng-qing, Xiao Jin-yi, Liu Hong-jie. Autoclaved brick from low-silicon tailings: preparation and discussion. In: Proceedings of 3rd International Conference on Energy and Environment Materials ICEEM-3, China: Guangzhou; 2006. p. 137. [5] Narayanan N, Ramamurthy K. Structure and properties of aerated concrete: a review. Cem Concr Compos 2000;22(5):321–9.
5
_ [6] Karakurt Cenk, Kurama Haldun, Topçu Ilker Bekir. Utilization of natural zeolite in aerated concrete production. Cem Concr Compos 2010;32(1):1–8. [7] Tao Hong-sen. Research for developing aerated concrete by using fly ash with high calcium of Gehua power plant in Hongshan, Wuhan [Dissertation]. WuHan:Wuhan University of Technology; 2004 [In Chinese]. [8] Mostafa NY. Influence of air-cooled slag on physicochemical properties of autoclaved aerated concrete. Cem Concr Res 2005;35(7):1349–57. _ Karakurt C. Properties of the autoclaved aerated concrete [9] Kurama H, Topçu IB, produced from coal bottom ash. J Mater Process Technol 2009;209(2):767–73. [10] Ma Bao-guo, Xu Zheng. Study on a new kind of aerated concrete containing efflorescence sand-phosphorus slag-lime. J Build Mater 1999;2(3):223–8 [In Chinese]. [11] Li Fang-xian, Chen You-zhi, Long Shi-zong. Experimental investigation on aerated concrete with addition of lead–zinc tailings. J Southwest Jiaotong Univ 2008;43(6):810–5 [In Chinese]. [12] Wang Yan, Yin Jie, Chen Ji-chun, Peng Chang-qi. Aerocrete made with low silicon tailings of Cheng Chao iron ore mine. J Wuhan Univ Technol – Mater Sci 2000;15(2):58–62. [13] Komarneni Sridhar, Roy Rustum, Roy Della M. 29Si and 27Al magic angle spinning nuclear magnetic resonance spectroscopy of Al-substituted tobermorites. J Mater Sci 1985;20:4209–14.
Contents lists available at SciVerse ScienceDirect
Construction and Building Materials journal homepage: www.elsevier.com/locate/conbuildmat
Preparation of autoclaved aerated concrete using copper tailings and blast furnace slag Xiao-yan Huang a, Wen Ni a, Wei-hua Cui b,⇑, Zhong-jie Wang a, Li-ping Zhu a a b
State Key Laboratory of High-Efficient Mining and Safe of Metal Mines, University of Science and Technology Beijing, Ministry of Education, Beijing 100083, China School of Water Resources and Environment, China University of Geosciences (Beijing), Beijing 100083, China
a r t i c l e
i n f o
Article history: Received 21 February 2011 Received in revised form 11 July 2011 Accepted 9 August 2011 Available online 20 September 2011 Keywords: Skarn-type copper tailings Autoclaved aerated concrete Reaction mechanism Tobermorite
a b s t r a c t Based on the background that large amount of copper tailings are stockpiled in China, the utilization of skarn-type copper tailings to prepare autoclaved aerated concrete (AAC) was studied. The AAC samples were prepared on a laboratory scale with a dry density of 610.2 kg m 3 and compressive strength of 4.0 MPa. Compared with the traditional AAC, lime was totally substituted by skarn-type copper tailings and blast furnace slag in order to develop a potential technique of reducing CO2 emission during the AAC production process. The samples of different curing stage were examined by XRD, FESEM as well as 29Si and 27Al NMR analyses. Results show that the main minerals in the AAC product are tobermorite-11 Å, anhydrite, augite, quartz, calcite and dolomite, with small amount of other minerals brought in by the copper tailings. It was also suggested that most minerals in the copper tailings participated in the hydration reaction during the procuring process, and the chemical elements in them got into the structure of platy tobermorite in the subsequent autoclaving process. Ó 2011 Elsevier Ltd. All rights reserved.
1. Introduction Copper tailings in China have been nearly totally piled up through the history of copper production. Now more than 2.4 billion tons of copper tailings are estimated to exist in China [1], posing a severe threat to the environmental condition. Skarn-type copper tailing (SCT) is one type of copper tailings to be most difficultly reused in the traditional building materials, due to its nature of very fine grain size and high CaO and MgO content. It has been reported that skarn-type copper tailings can be used to prepare autoclaved brick with good quality [2–4]. But the high transportation costs and low commodity prices restrict the broad commercialization of such brick. Autoclaved aerated concrete (AAC) is a lightweight and highly porous wall material with excellent insulation ability [5,6]. According to Chinese national standard, AAC is the only one type of wall materials owning the ability to meet 50% of the building energy saving request without adding other affiliated thermal insulation materials [7]. The background that Chinese government is promoting building energy saving and carbon emission reduction gives autoclaved aerated concrete a broad application prospect. The commercial AAC is usually produced with cement and lime as calcareous materials, and with quartz sand or fly ash as the siliceous materials. To extend the range of raw materials and lower the pro-
⇑ Corresponding author. Tel.: +86 137200160522; fax: +86 01082321081. E-mail address: [email protected] (W.-h. Cui). 0950-0618/$ - see front matter Ó 2011 Elsevier Ltd. All rights reserved. doi:10.1016/j.conbuildmat.2011.08.034
duction costs, several researchers have investigated the possibility of replacing the traditional raw materials of AAC by industrial waste, such as air-cooled slag [8], coal bottom ash [9], efflorescence sand and phosphorus slag [10], lead–zinc tailings [11] and iron ore tailings [12]. These studies mainly focused on exploring suitable alternative siliceous materials, and few researches involved the investigation into calcareous materials. In this study, skarn-type copper tailings and water-quenched blast furnace slag (BFS) were used as mainly raw materials to prepare AAC. Because the high CaO and MgO content in both SCT and BFS, SCT and BFS was used to replace lime as calcareous material. Besides, SiO2 in the BFS was considered as partly alternative siliceous resource to reduce the consumption of quartz sand. The object of the present work is to investigate the microstructural properties and phase compositions of the AAC prepared by SCT and BFS, and to make primary understanding of reaction mechanism during the process of precuring and autoclaving, especially the behavior of SCT in the hydrothermal reaction. 2. Experiments 2.1. Raw materials The AAC samples were prepared by the following raw materials: SCT, BFS, quartz sand (QS), cement clinker (CC) and natural gypsum. They were all ground in a SMU500 500 type ball mill. The results of chemical analyses and specific surface area tests of the raw materials are listed in Table 1. The specific surface area of natural gypsum is 402.6 m2 kg 1. XRD of the SCT sample shows that the main minerals are augite, dolomite, phologipite, amphibole and calcite, accompanied
2
X.-y. Huang et al. / Construction and Building Materials 27 (2012) 1–5
Table 1 Chemical compositions and specific surface area of raw materials.
3. Results and discussion
Oxide composition (wt%)
SCT
BFS
QS
CC
SiO2 Fe2O3 CaO Al2O3 MgO K2O Na2O IOT Specific surface area (m2 kg
44.52 1.94 13.56 5.36 19.92 1.20 1.00 9.26 656.8
32.7 0.4 38.79 15.4 8.97 0.36 0.23 0.76 562.6
82.83 0.53 1.83 7.13 1.14 2.69 1.55 0.18 792.7
21.98 5.13 60.38 5.54 3.03 2.17 0.25 – 380.6
1
)
3.1. Mechanical properties The result of compressive strength test of the AAC sample and its related raw materials proportions are given in Table 2. Superplasticizer dosage was 0.8% of the total solid mixture and aluminum powder dosage was 0.10%. The water/solid ratio was 0.38. At the designed proportions, the compressive strength of AAC samples could get 4.0 MPa and the dry density of AAC was 610.2 kg m 3. Additionally, the compressive strength of sample in the absolute dry condition (DCS) can reach up to 5.9 MPa, and its corresponding specific strength reaches as high as 9.7.
3.2. XRD analyses XRD analyses were performed to investigate the phase changes in the AAC samples during the process of precuring and autoclaving. The XRD patterns of the dry mixture (DM) of raw materials, hardened aerated concrete (HAC) sample precured under the saturated steam curing for 12 h after being mixed with warm water and the final AAC sample are shown in Fig. 2. As indicated in DM’s spectrum, most minerals in SCT are identified, except talc and clinochlore. The non-detection of these two minerals is likely due to the fact that the overall amount of those minerals originally comprising small amount in SCT are in more minor quantities after addition into the dry mixture with only 30% in weight, so that the XRD is not sensitive enough to allow detection at such low level. Quartz, gypsum and alite from the raw materials of quartz sand, natural gypsum and cement clinker, respectively, were detected as expected. A diffused band existing between 22° and 38° of 2h was caused by the incorporation of BFS with glassy nature. It can be seen from HAC’s spectrum, after being precured around 48 °C for 12 h, the diffraction peaks of alite and gypsum disappeared. The AFt, a new phase, was formed. The broad band at around 17–38° of 2h indicated the existence of C-S-H gels. The appearance of AFt and C-S-H gels and the disappearing of alite
Fig. 1. XRD pattern of skarn-type copper tailings. by minor phases including clinochrysotile, albite, talc and clinochlore (Fig. 1), in accordance with the chemical analysis result that most of the minerals included are rich in CaO and MgO. The XRD analysis of natural gypsum sample shows that the main mineral in it is CaSO42H2O and no other crystalline phases were identified. Aluminum powder was used as a gas producing agent for the slurry foaming, and naphthalene was used as superplasticizer. 2.2. Procedure Firstly, the prepared powder of the raw materials and the superplasticizer were thoroughly dry mixed. Then warm water (48 ± 1 °C) was added and mixed for 2 min. Finally, aluminum powder was added and mixed with the slurry for another 30 s. The obtained slurry was casted into preheated steel molds of 100 100 100 mm to allow it to expand and harden at the temperature of 48 ± 1 °C for 12 h under a steam saturated condition. After their swollen up surface being cut to flat, the samples were demolded and put into an industrial autoclave for hydrothermal reaction for 8 h at 13.5 bars. 2.3. Analysis The bulk density and compressive strength tests were conducted according to GB 11968-2008, ‘‘Autoclaved aerated concrete blocks’’ which specifies that bulk density should be determined by oven dry of 24 h at 60 ± 5 °C, and then another 24 h at 80 ± 5 °C, following by oven dry at 105 ± 5 °C until samples tested reached constant weight, and compressive strength tests were performed at loading rate of 2.0 ± 0.5 kN/s and on samples with moisture content of 8–12%. The X-ray diffraction (XRD) spectra of different samples were obtained with a D/Max-RC diffractometer (Japan) with copper Ka radiation at 30 mA and 50 kV. A step size of 0.02° was selected over a 2h range of 5–70°. The microstructure of the samples under different curing stage was observed with a SUPRA™55field emission scanning electron microscope (FESEM). The fractured surfaces of the samples were coated with carbon prior to examination. MAS NMR tests were conducted in a Bruker Avance III400 spectrometer operating at 59.62 MHz (29Si) and 104.0 MHz (27Al). The rotation frequency was 5 kHz for 29Si and 10 kHz for 27Al, and the delay time was 3 s for 29Si and 1 s for 27Al.
Fig. 2. XRD patterns of DM, HAC and AAC samples.
Table 2 Material proportions and mechanical properties of AAC sample. Mixture composition of AAC (wt%)
Dry density (kg m
SCT
BFS
QS
CC
Gypsum
30
35
20
10
5
610.2
3
)
Compressive strength (MPa)
DCS (MPa)
Specific strength
4.0
5.9
9.7
X.-y. Huang et al. / Construction and Building Materials 27 (2012) 1–5
and gypsum were the results of the hydration reaction between gypsum, clinker and BFS. The intensity reduction of quartz’s characteristic peaks indicated that the fine quartz particles participated in the pozzolanic reaction. Comparing spectra of DM and HAC, the peaks of dolomite, phlogopite, albite, augite and amphibole decreased significantly. That means these crystalline minerals in SCT, which were inert at room temperature, participated in the hydration reaction forming AFt or C-S-H gel during the precuring process. From spectrum of AAC, it can be seen that the major minerals in the final AAC products are tobermorite-11 Å, anhydrite, and some residual minerals including quartz, dolomite, calcite and augite, accompanied by phlogopite, amphibole, clinochrysotile and albite in minor quantities. The changes in peak intensity of minerals from SCT were not clearly detected. That is to say the minerals in SCT were not evidently involved into the hydrothermal reaction during the 8 h autoclaving process. The reason why the half width of tobermorite peaks was broad was that the structure of tobermorite crystals would be complicated by the incorporation of Mg2+, Al3+ and other minor composition from SCT. During the precuring treatment, most minerals from SCT participated in the hydration reaction forming AFt and C-S-H gel, and through autoclaving process, the hydration products of C-S-H gel and AFt were converted to tobermorite and anhydrite. The chemical elements decomposed from SCT were absorbed into the tobermorite crystals through the transfer of AFt and C-S-H gel. The presence of those elements in tobemorite leaded to numerous tobermorite crystals with different lattice parameters and broad half width of the tobermorite peaks in XRD patterns.
3.3. FESEM analyses The surface microstructure of artificial pores in HAC sample (Fig. 3) shows that, except for AFt and C-S-H gel, there was no other crystalline hydration products formed, which was in good accordance with the XRD results in Section 3.2. The SEM images of the AAC sample are shown in Fig. 4. Because of the low water/solid ratio, the micro-morphology of wall section between artificial pores was dense, which can be seen from Fig. 4a. The dense structure had good contribution to high compressive strength of the AAC samples. The surface microstructure of one artificial pore was presented in Fig. 4b. Long-strip shaped anhydrite interpenetrating with platy tobermorite crystals was observed. Combining Fig. 4b with Fig. 3, it can be seen that AFt and C-S-H gel were transformed into tobermorite and anhydrite. Further magnification of Fig. 4b is shown in Fig. 4c. It clearly shows that the platy tobermorite is well crystallized with a thickness of
AFt
C-S-H
Fig. 3. FESEM of hardened aerated concrete sample.
3
about 60–80 nm and a width of 1–2 lm. The platy tobermorite interfinger and overlap with each other forming firm skeleton and empty cavities between the plates. As a result of such microstructure, high compressive strength, good thermal insulation and heat preservation performance of this material would be guaranteed. 3.4.
29
Si and
27
Al NMR analyses
The 29Si NMR spectrum of the final AAC product is given in Fig. 5. As shown in Fig. 5, there is a broad 29Si resonance peak including three subsidiary signals at chemical shift of 79.1 10 6, 80.4 10 6, 81.2 10 6. The signal at 79.1 10 6 can be partly referred to Q2 (0Al) sites of the residual augite with single-chain silicate structure in SCT. But the content of augite is too small to make a peak at 79.1 10 6 so prominent. As we know, in the structure of augite, the negative charges created by Si–O structure are mainly balanced by Ca2+, Mg2+ and Fe2+, while in that of C-S-H gel those charges are balanced by Ca2+ and H+. Due to the fact that the electronegativity of H+ is greater than that of Ca2+, Mg2+ and Fe2+, the chemical shift of Q2 (0Al) sites in augite would have lower field shift than that in C-S-H gel and be approximate to that of Q2 (1Al) sites of C-S-H gel. So, the resonance signal at 79.1 10 6 is probably the overlap of both augite Q2 (0Al) and C-S-H gel Q2 (1Al) tetrahedrons. The signals at 80.4 and 81.2 10 6 can be assigned to Q2 (1Al) sites in tobermorite-11 Å. As discussed in Section 3.2, tobermorite-11 Å is the dominant phase in AAC samples. The reason that there are two subsidiary peaks assigned to Q2 (1Al) was that Ca in tobermorite-11 Å may be substituted by Mg from SCT and BFS. As noted above, these Mg rich minerals have participated in the hydrothermal reaction to produce tobermorite. If the Si–O tetrahedrons were linked by Mg rather than by Ca between tetrahedron chains, the electron density around the related silicon atoms would be getting lower. That is because the electronegativity of Mg is greater than that of Ca. Thus, the signal at 80.4 10 6 may be caused by the Q2 (1Al) sites where Mg ions are the linkages between the chains, and the signal at 81.2 10 6 can be assigned to Q2 (1Al) sites where Ca ions are the linkages. On the right shoulder of the major peak there are many minor subsidiary peaks. Among them, signal at 91.3 10 6 may be indicative of Q3 (1Al) structure state of tobermorite-11 Å [13], indicating that Al also had substituted in Q3 units in tobermorite structure. Compared with the intensity of signals at 81.2 10 6 and 80.4 10 6 assigned to Q2 (1Al) sites, the intensity of signal at 91.3 10 6 assigned to Q3 (1Al) sites is significantly lower, suggesting that the Si–O–Si branching sites between silicate double chains are quite few. Other subsidiary signals from 86.7 10 6 to 96.2 10 6 on the right side of the major peak can be assigned to Q3 (1Al) and Q3 (0Al) sites of residual phyllosilicates including phlogopite, clinochrysotile and amphibole. Evidently, all of the peaks with chemical shifts beyond 96.2 10 6 are very low, suggesting most minerals with Q3 (0Al) and Q4 (nAl) sites, which are quartz and field spar minerals, have been largely consumed. There seems to be a contradiction between the XRD spectrum which show the highest peak belongs to quartz and the 29Si NMR spectrum which shows very low content of residue quartz in the final AAC samples. This is because the crystallinity of quartz is well. Small amounts of quartz can make a very strong peak in the XRD spectrum. The 27Al NMR spectrum of the final AAC product (Fig. 6) shows a dominant resonance at 62.3 10 6 indicative of tetrahedral Al coordination and a lower resonance at 12.7 10 6 indicative of hexahedral Al coordination. This assignment of 27Al suggested that the Al atoms mainly served as substitution of Si atoms, which is in good accordance with the 29Si NMR analysis results. The relatively
4
X.-y. Huang et al. / Construction and Building Materials 27 (2012) 1–5
Fig. 4. FESEM of AAC sample (a: surface of wall section; b: inner surface of one artificial pore; c: further magnification of b).
Fig. 6. Fig. 5.
27
Al NMR spectra of the AAC products.
29
Si NMR spectra of the AAC products.
small amount of hexacoordinated Al atoms in the AAC samples probably existed as charge balancing ions in the silicates. 4. Conclusions The results show that skarn-type copper tailings and blast furnace slag can be used as substitution of lime to produce AAC products by providing CaO and MgO compositions. Using them can reduce the emission of CO2, for the elimination of limestone calcinations process. The AAC product with a dry density of 610.2 kg m 3 and compressive strength of 4.0 MPa was produced by the raw material composition of 30% skarn-type copper tailings,
35% high furnace slag, 10% cement clinker and 5% gypsum. The main minerals in the AAC are platy tobermorite and long-strip anhydrate besides quartz and other residual minerals from the copper tailings. Through the NMR analysis, it could be find that in the tobermorite structure, Ca atoms may be substituted by Mg atoms which were provided by the high furnace slag and minerals from the skarn-type copper tailings, and Al atoms significantly substituted Si atoms in the chain sites as well as chain bridge sites of tobermorite. Acknowledgements The authors gratefully acknowledge Beijing Jinyu AAC Co., Ltd. for providing access to industrial autoclave. We thank Li Hong
X.-y. Huang et al. / Construction and Building Materials 27 (2012) 1–5
and He Jianping of University of Science and Technology Beijing for XRD and FESEM test. We would also like to express thanks to Xiang Junfeng of institute of chemistry of Chinese Academy of Sciences for his assistance in NMR test. References [1] Yu Liang-hui, Jia Wen-long, Xue Ya-zhou. Survey and analysis of the copper tailing resources in China. Metal Mine 2009(8):179–81 [in Chinese]. [2] Zhao Feng-qing, Zhao Jing, Liu Hong-jie. Autoclaved brick from low-silicon tailings. Constr Build Mater 2009;23(1):538–41. [3] Zhao Feng-qing, Ni Wen, Wang Hui-jun. Preparation of load-bearing brick with copper tailings by steam curing method. Express Inf Min Ind 2006(4):34–6 [in Chinese]. [4] Zhao Feng-qing, Xiao Jin-yi, Liu Hong-jie. Autoclaved brick from low-silicon tailings: preparation and discussion. In: Proceedings of 3rd International Conference on Energy and Environment Materials ICEEM-3, China: Guangzhou; 2006. p. 137. [5] Narayanan N, Ramamurthy K. Structure and properties of aerated concrete: a review. Cem Concr Compos 2000;22(5):321–9.
5
_ [6] Karakurt Cenk, Kurama Haldun, Topçu Ilker Bekir. Utilization of natural zeolite in aerated concrete production. Cem Concr Compos 2010;32(1):1–8. [7] Tao Hong-sen. Research for developing aerated concrete by using fly ash with high calcium of Gehua power plant in Hongshan, Wuhan [Dissertation]. WuHan:Wuhan University of Technology; 2004 [In Chinese]. [8] Mostafa NY. Influence of air-cooled slag on physicochemical properties of autoclaved aerated concrete. Cem Concr Res 2005;35(7):1349–57. _ Karakurt C. Properties of the autoclaved aerated concrete [9] Kurama H, Topçu IB, produced from coal bottom ash. J Mater Process Technol 2009;209(2):767–73. [10] Ma Bao-guo, Xu Zheng. Study on a new kind of aerated concrete containing efflorescence sand-phosphorus slag-lime. J Build Mater 1999;2(3):223–8 [In Chinese]. [11] Li Fang-xian, Chen You-zhi, Long Shi-zong. Experimental investigation on aerated concrete with addition of lead–zinc tailings. J Southwest Jiaotong Univ 2008;43(6):810–5 [In Chinese]. [12] Wang Yan, Yin Jie, Chen Ji-chun, Peng Chang-qi. Aerocrete made with low silicon tailings of Cheng Chao iron ore mine. J Wuhan Univ Technol – Mater Sci 2000;15(2):58–62. [13] Komarneni Sridhar, Roy Rustum, Roy Della M. 29Si and 27Al magic angle spinning nuclear magnetic resonance spectroscopy of Al-substituted tobermorites. J Mater Sci 1985;20:4209–14.