- Email: [email protected]
Mechanism of triethanolamine on Portland cement hydration process and microstructure characteristics
Construction and Building Materials 93 (2015) 457–462
Contents lists available at ScienceDirect
Construction and Building Materials journal homepage: www.elsevier.com/locate/conbuildmat
Mechanism of triethanolamine on Portland cement hydration process and microstructure characteristics Jianguo Han a,b,⇑, Kejin Wang b, Jiyao Shi c, Yue Wang c a
Department of Civil Engineering, Tsinghua University, Beijing 100084, China Department of Civil, Construction and Environmental Engineering, IOWA State University, 50014, USA c Technology Center of China Railway Tunnel Co., Ltd., Luoyang 471000, China b
h i g h l i g h t s TEA accelerates hydration of C3A, promoting hydration heat release rate. When TEA introduced, after depletion of calcium sulfate, AFt is converted to AFm. TEA retards hydration of C3S, which can jeopardize concrete strength at early age. Influence of TEA on cement paste pore structure is dosage dependent. Balance should be made when using TEA, on its accelerating and retarding effect.
a r t i c l e
i n f o
Article history: Received 19 August 2014 Received in revised form 30 March 2015 Accepted 8 June 2015 Available online 20 June 2015 Keywords: TEA Acceleration Hydration Microstructure AFt AFm
a b s t r a c t Influence of triethanolamine (TEA) on Portland cement hydration heat evolution process, hydration product, pore structure, setting time and concrete strength were investigated; and mechanism of TEA on the cement hydration process and microstructure is discussed. Research results indicate that TEA accelerates hydration process of C3A and retards that of C3S, such effects enhance with the increase of TEA dosage. TEA also promotes rate of AFt formation and its conversion to AFm. As a result, TEA accelerates setting of cement paste but jeopardizes strength of concrete. Depending on its dosage, TEA can either optimize or harm cement paste pore structure. The types of calcium sulfoaluminate hydrate (AFt or AFm) produced in cement paste containing TEA depends on the abundance of calcium sulfate, before depletion of calcium sulfate, AFt is produced, after that, AFm is produced (converted from AFt). Ó 2015 Elsevier Ltd. All rights reserved.
1. Introduction Triethanolamine (TEA) is a weak base surfactant. It can be used as grinding aid for cement manufacturing (typically at dosage of 0.1%, weight percentage of clinker), in order to prevent agglomeration of powder and coating formation on milling ball surface [1]. TEA can also be used as component of chemical admixture, such as shotcrete accelerator or water reducing admixture, in order to accelerate cement setting or counteract the retarding effect of other component [2,3]. Previous research indicates that: (1) effect of TEA on cement hydration is objective oriented: it accelerates hydration process of C3A but retard that of C3S [4]. Ramachandran conducted a series of studies to investigate the influences of TEA on hydration process of C3A, C3A–gypsum and ⇑ Corresponding author at: Department of Civil Engineering, Tsinghua University, Beijing 100084, China. http://dx.doi.org/10.1016/j.conbuildmat.2015.06.018 0950-0618/Ó 2015 Elsevier Ltd. All rights reserved.
C3S and found that TEA accelerates C3A hydration and C3A–gypsum reaction by promoting formation of aluminate hydrate and AFt, and retards hydration process of C3S by extending the induction period [5–7]; (2) effect of TEA on cement hydration is dosage dependent also. At small dosage (e.g. 0.02%, weight percentage of cement) it acts as setting accelerator, at higher dosage (e.g. 0.5%) it acts as setting retarder, at more higher dosage (e.g. 1.0%) it acts as setting accelerator once again [8]. Despite all the work have been done, the influence of TEA, especially at high dosage (higher than 1.0%), on Portland cement setting behavior and strength development still need further research, and the working mechanism of TEA on Portland cement hydration process and microstructure characteristics merit further investigation. Such research can facilitate TEA application and understanding of its effect in Portland cement manufacturing and utilization. In this paper, TEA dosage from 0.02% to 8% was adopted, and its influence on Portland cement hydration process, hydration
458
J. Han et al. / Construction and Building Materials 93 (2015) 457–462
products, setting time, pore structure and concrete strength were studied.
Table 2 Cement paste mix proportion and properties measured.
2. Materials and mix proportions Materials used include: type I Portland cement with chemical composition and fineness shown in Table 1; deionized water with resistivity no less than 10.0 MX cm; River sand with fineness modulus of 3.1; limestone with particle size ranging from 5 to 10 mm; polycarboxylate-based water reducing admixture (WRA); TEA was a liquid chemical reagent. Cement paste mix proportion and properties measured are shown in Table 2. Concrete mix proportion is shown in Table 3, and 100 mm cubes are used for compressive strength evaluation.
3. Testing methods Hydration heat evolution process of cement pastes was monitored by a differential scanning calorimeter (DSC) (ToniCAL, manufactured by Toni Company), under constant temperature of 25 °C. The preparation of cement pastes for microstructure investigation consists of following steps: (1) Conducting a complete DSC test (0–24 h) for a cement paste. (2) Selecting a few key time points on the DSC curve so as to capture features of the hydration process of the paste. (3) Performing a series of DSC tests as described in Step (1) for the same cement paste; but, when the test reaches each of the key time points as described in Step (2), took the cement paste sample out of the DSC instrument chamber. (4) Cracking the samples into small pieces and submerging these small pieces into ethanol so as to discontinue the hydration process (It took less than 2 min from taking the sample out of the DSC instrument chamber to submerging the cracked pieces into ethanol; the volume ratio of ethanol to cracked pieces was about 10:1). (5) Studying the microstructure (hydration product and pore structure) of the cement paste. Cement paste samples were taken out of ethanol and oven dried at 60 °C for 6 h. The oven dry process was to evaporate ethanol in capillary pore; then, pulverized using an agate pestle and mortar and passed through an 80 lm sieve, if powder was needed for measurement. Hydration products of the above cement pastes were identified by X-ray diffractometer (Automated D/max, Rigaku Corporation), with a Cu Ka source and a scanning speed of 2° per minute. TG–DSC–MS hyphenated method was used to verify the hydration product further. Using the TG–DSC instrument (STA 449 F3, Netzsch Company), under a controlled regime (from 50 to 1100 °C, heating rate of 20 °C/min; air atmosphere), the mass change and the difference of energy input between cement paste and reference material were measured simultaneously. Meanwhile, the gas released during TG–DSC testing process was transported to the MS instrument (QMS 403C, Netzsch Company) via capillary tube, and the input gas was ionized first and then classified and identified based on mass to charge ratio. Pore structure of the paste samples was measured by BET absorption method. Cement paste samples were vacuum
Table 1 Chemical composition and fineness of Portland cement. Oxide (wt.%) SiO2
Al2O3
Fe2O3
CaO
MgO
SO3
Na2Oeq
f-CaO
21.58 C3S 57.34
4.03
3.46 C2S 18.90
61.49
2.60 C3A 6.47
2.83
0.51 C4AF 11.25
0.67
Blaine fineness (m2/kg) 346 LOI (wt.%) 1.97
*
Water to cement ratio (weight percentage)
Dosage of TEA* (weight Properties measured percentage of cement) (%)
0.4
0, 0.02, 0.2, 2 and 6
0.4
0, 0.2 and 6
0.4
0, 2, 4, 6 and 8
Hydration heat evolution process Microstructure (including: hydration product and pore structure) Setting time
No water was deducted when TEA was introduced.
Table 3 Concrete mix proportion. No.
Cement (kg/m3)
Water (kg/m3)
W/C
Sand (kg/m3)
Aggregate (kg/m3)
WRA (%)
TEA** (%)
C-TEA-0 C-TEA-5
490 490
196 196
0.4 0.4
830 830
880 880
1.5 1.5
0 5
** Weight percentage of cement, and no water was deducted when TEA was introduced.
de-aerated for 6 h at 100 °C before measurement. Nitrogen (N2) gas was used as adsorbate. Measurement was performed by a surface area and pore analyzer (Quadrasorb SI, Quantachrome Company). Pore volume and pore size were calculated based on the amount of N2 adsorbed. Setting time of cement pastes with different dosages of TEA was measured according to ASTM C191 by Vicat apparatus, under temperature of 23 ± 2 °C. Compressive strength of concrete was measured at 1 day of age according to ASTM C39. 4. Results and analysis 4.1. Effect of TEA on hydration heat evolution process During Portland cement hydration process monitored by DSC method, there are usually 5 stages: (A) initial reaction period, (B) induction period, (C) acceleration period, (D) deceleration period, and (E) slow reaction period [9]. There are usually two peaks during the hydration process, the first peak, in initial reaction period, is mainly resulted from ion dissolution and reaction between C3A and calcium sulfate, usually resulting in formation of AFt; the second peak, at the end of acceleration period, is mainly resulted from hydration of C3S, forming C–S–H and calcium hydroxide (CH) [10]. Hydration heat release rate curves of cement pastes with different dosage of TEA are given in Fig. 1. It can be seen in the figure, as the TEA dosage increases, the hydration heat release rate in initial reaction period is promoted, the induction period is prolonged, and the acceleration period is postponed. When TEA dosage is higher than 2%, no acceleration period can be identified during the 24 h measurement. An affiliated peak is noted right after the main peak in the initial reaction period, which corresponds to the conversion of AFt to AFm as explained by XRD measurement in later part of this paper. The promoted hydration heat release rate in initial reaction period is due to the accelerating effect of TEA on reaction of C3A with calcium sulfate. TEA can chelate with some cation ions, such as Al3+, Fe3+ and Ca2+, under highly alkaline medium [11,12]. When chelated with Al3+, TEA can promote formation of calcium sulfoaluminate hydrate (AFt and AFm) and accelerate hydration process of C3A. The postponed hydration process of C3S after the introduction of TEA might be attributed to TEA’s chelating ability with Fe3+ and Ca2+ cations. Due to mineral structure characteristics, the
459
J. Han et al. / Construction and Building Materials 93 (2015) 457–462
(a) Without TEA
Fig. 1. Hydration heat release rate curves of cement pastes with different dosage of TEA.
hydration ability of C4AF, C3S and C2S are inferior to that of C3A. Before acceleration period of hydration process, majority of the Fe3+ in C4AF and Ca2+ in C3S and C2S are still in site with the original minerals. As a result, a surface complex is formed on C3S surface, which acts as a barrier, isolating C3S from water and hindering its hydration process [5]. Meanwhile, the accelerated hydration of C3A can speed up the depletion of calcium sulfate, as proved by the XRD measurement in the later part of this paper. After the depletion of calcium sulfate, C3A can hydrate alone to form C2AH8 and C4AH19, which belongs to AFm phase. AFm is poor crystalline hydration product, which can cover the surface of cement particle and hinder hydration process of C3S [13]. That is, the formation of C2AH8 and C4AH19, resulting from hydration of C3A, may be another cause for the retarding effect of TEA on C3S hydration process, especially at high dosage, in which the depletion of calcium sulfate can be easily achieved. Based on the characteristic of hydration heat release rate curves in Fig. 1, TEA dosage of 0%, 0.2% and 6% were chosen for investigating the influence of TEA on microstructure of cement paste. Time points selected on these curves are given in Fig. 2, and hydration was discontinued at these time points, by taking sample out of the DSC instrument chamber and submerging the cracked pieces into ethanol.
4.2. Effect of TEA on hydration products and pore structure At each of the time points shown in Fig. 2, the microstructure of cement paste, including hydration products and pore structure, was studied. XRD patterns of cement pastes are given in Fig. 3. From Fig. 3(a) (without TEA), it can be seen that anhydrite is the main calcium sulfate in cement. During cement hydration process, anhydrite is consumed and AFt is produced (at time point 5#, or the age of 24 h). As illustrated in Fig. 3(b), cement paste with 0.2% TEA possesses XRD pattern similar to that of control specimen (without TEA). However, AFt is identified at time point 4# (18.5 h of age), manifesting the accelerating effect of TEA on AFt formation. AFt peak disappeared at time point 5# in Fig. 3(b) is probably due to the conversion of AFt to AFm. AFm is usually of poor crystalline when compared with AFt, and it is more difficult to be identified by XRD. As shown in Fig. 3(c), when 6% TEA was used, XRD pattern of cement paste appears very different with that of control and 0.2%
(b) With 0.2% TEA
(c) With 6% TEA Fig. 2. Time points on hydration heat release rate curves.
TEA samples. This is probably due to the quick consumption (before and at time point 1#) of anhydrite during the accelerated C3A hydration. The quantity of CH produced in the whole hydration process is negligible, due to the retarded hydration process of C3S. Meanwhile, AFm, instead of AFt, is formed after time point 1#, as the peak representing calcium sulfoaluminate hydrate is shifted from AFt to AFm. The peak appeared at 26.6° in Fig. 3(b) and (c), which is not emerged in control specimen, needs further investigation. DSC curves (got from TG–DSC measurement) of cement paste with different dosage of TEA are given in Fig. 4. MS measurement was performed on the last time points in Fig. 2 during TG–DSC testing, and test results are given in Fig. 5. The endothermal peak appeared on DSC curve in Fig. 4 can be used to further identify hydration products. MS curve in Fig. 5 can be used to certify hydration product identification made by TG–DSC method. Base on DSC and MS results in Figs. 4 and 5, hydration products and their thermal behavior during the TG– DSC–MS measurement are summarized in Table 4. The following observations can be made from Fig. 4: (1) Comparison of endothermal peak intensity (between different time points, of same TEA dosage) produced by decomposition of CH is stronger in control sample than that in TEA containing samples. (2) TEA has retarded hydration process of C3S, and this effect is intensified with the increase of TEA dosage. (3) AFt is produced in control and 0.2% TEA samples, and AFm is produced in the 6% TEA sample, which is consistent with the findings resulting from XRD measurement.
460
J. Han et al. / Construction and Building Materials 93 (2015) 457–462
( a ) W i th o u t TE A
(b) With 0.2% TEA
(C) With 6% TEA Fig. 3. XRD pattern of cement pastes with different dosage of TEA.
(4) AFm in Fig. 4(a) and (b) is produced by calcination of AFt [14], and this AFm decomposes at 880 °C. (5) AFm in Fig. 4(c) is converted from AFt during hydration process, which loses water at 180 and 280 °C, and decomposes at 980 °C. (6) At high temperature, AFm will lose the interlayer sulfate and crystallizes into C3A and C12A7, releasing SO2 (gas) [15]. (7) The decomposition temperature of the two type of AFm, one from calcination of AFt and one from conversion of AFt during hydration, is different; the former is at about 880 °C and the latter is at about 980 °C. Pore structure of cement paste with different dosages of TEA at the age of 24 h is given in Fig. 6. Generally, pores with a diameter within 10–100 nm are considered as capillary pores, and the amount of capillary pores is related to the degree of hydration. As cement hydration progresses, the hydration products will fill in particle gaps and refine the pore structure. Pore structure is a key parameter which can bring influences to mechanical properties and anti-penetration ability. As seen in Fig. 6, cement paste with 0.2% TEA has narrower and finer pore size distribution than cement paste without TEA, probably due to the accelerating effect of TEA on C3A hydration process.
Cement paste with 6% TEA has broader and coarser pore size distribution, probably due to the retarding effect of TEA on C3S hydration process. 4.3. Effect of TEA on setting time and strength Influence of TEA on Portland cement setting time is shown in Fig. 7. The figure indicates that both initial and final setting is accelerated significantly with TEA addition, from more than 200 min to less than 50 min. The concave up shape caused by TEA dosage of 8% indicates that the acceleration effect is less active after the dosage of TEA reaches 6%. This is probably because the depletion of anhydrite in cement paste with 6% TEA, which occurs almost immediately after the mixing of cement paste (see Fig. 3(c)). Meanwhile, at a higher dosage, the increased liquid-to-solid volume resulted from the redundant TEA can lower density of cement paste, and the retarding effect of TEA on C3S hydration process may contribute to this result too. Compressive strength of concrete with 0% and 5% TEA at the age of 1 day was measured. Control concrete (without TEA) produced compressive strength of 16.2 MPa and 5% TEA containing specimen produced compressive strength of 3.5 MPa (21.6% of control specimen). The results indicate that although TEA can reduce cement
461
J. Han et al. / Construction and Building Materials 93 (2015) 457–462
(a) Without TEA
(b) With 0.2% TEA
(c) With 6% TEA Fig. 4. DSC curve of cement pastes with different dosage of TEA.
Table 4 Hydration products and its thermal behavior during TG–DSC–MS measurement. Temperature/°C (approximately)
Hydration product
Thermal behavior
Material released
120 180 and 280 460 720 880 980
C–S–H or AFt AFm CH CaCO3 AFm* AFm
Losing water Losing water Decomposition Decomposition Decomposition Decomposition
H2O (gas) H2O (gas) H2O (gas) CO2 SO2 SO2
*
Produced by decomposition of AFt during calcination. Endothermal peak is not obvious on DSC curve in Fig. 4(c).
further more, it was manifested in Ramachandran’s research that TEA can compromise concrete later age strength in addition to early age strength [7]. Fig. 5. MS spectrum of cement pastes with different dosage of TEA.
4.4. TEA working mechanism on Portland cement paste set time, TEA can significantly reduce early age strength of concrete, due to the retarding effect of TEA on C3S hydration process. Ramachandran has observed this phenomenon also. And
Base on the effect of TEA on cement hydration process, microstructure, setting time and concrete strength, the working mechanism of TEA on Portland cement can be concluded as following:
462
J. Han et al. / Construction and Building Materials 93 (2015) 457–462
pore structure. So, balance about the accelerating and retarding effect of TEA on different mineral in cement should be made when utilizing TEA in Portland cement. 5. Conclusion
Fig. 6. Pore size distribution of cement pastes with different dosage of TEA.
TEA accelerates reaction of C3A with calcium sulfate in Portland cement, and this effect enhances with the increase of TEA dosage. As a result, TEA can enhance hydration heat release rate during initial reaction period, promote AFt forming rate and conversion rate of AFt to AFm, and accelerate cement setting. In TEA introduced specimen, after depletion of calcium sulfate, AFt is converted to AFm, which is the same as that in common cement. However, TEA can speed up the conversion rate; thus, AFm can be identified shortly after the beginning of hydration at high dosage. TEA retards hydration process of C3S, and this effect enhances with the increase of TEA dosage also. AS a result, concrete strength at early age can be jeopardized by TEA. Influence of TEA on cement paste pore structure depends on its dosage; TEA can optimize pore size distribution at low dosage (e.g. 0.2% TEA) and injure pore size distribution at high dosage (e.g. 6% TEA). Balance should be made when utilizing TEA in Portland cement, by considering its accelerating effect on C3A and retarding effect on C3S.
References
Fig. 7. Effect of TEA on cement paste setting time.
(1) TEA accelerates reaction of C3A with calcium sulfate (gypsum, hemihydrate or anhydrite) to produce calcium sulfoaluminate hydrate (AFt and AFm), and this effect enhances with the increase of TEA dosage. (2) In TEA introduced specimen, after depletion of calcium sulfate, AFt is converted to AFm, which is the same as that of common cement: before depletion of calcium sulfate, AFt is produced; after that, AFm is produced. However, TEA can speed up the conversion of AFt to AFm, resulting in the appearance of AFm shortly after the beginning of cement hydration process. (3) TEA retards hydration process of C3S by prolonging the introduction period, and the retarding effect enhances with the increase of TEA dosage. (4) AS a result, introduction of TEA can enhance setting behavior; however, it can jeopardize early age strength; and depending on its dosage, TEA can either optimize or harm
[1] Sohoni S, Sridhar R, Mandal G. The effect of grinding aids on the fine grinding of limestone, quartz and Portland cement clinker. Powder Technol 1991;67(3):277–86. [2] Connolly JD, Hime WG. Analysis of concrete for triethanolamine. Cem Concr Res 1976;6(6):741–6. [3] Taylor HFW. Cement chemistry. 2nd ed. London, U.K.: Thomas Telford; 1997. p. 157–176 and p. 323–338. [4] Hewlett P. Lea’s chemistry of cement and concrete. 4th ed. U.K.: Elsevier; 2004. p. 881–887. [5] Ramachandran VS. Influence of triethanolamine on the hydration characteristics of tricalcium silicate. J Appl Chem Biotech 1972;22(11): 1125–38. [6] Ramachandran VS. Action of triethanolamine on the hydration of tricalcium aluminate. Cem Concr Res 1973;3(1):41–54. [7] Ramachandran VS. Hydration of cement—role of triethanolamine. Cem Concr Res 1976;6(5):623–31. [8] Gartner E, Myers D. Influence of tertiary alkanolamines on Portland cement hydration. J Am Ceram Soc 1993;76(6):1521–30. [9] Taylor HFW. Cement chemistry. 2nd ed. Thomas Telford: London, U.K; 1997. p. 150–225. [10] Bullard JW, Jennings HM, Livingston RA, et al. Mechanisms of cement hydration. Cem Concr Res 2011;41(12):1208–23. [11] Yilmaz VT, Menek N, Odabasogˇlu M. Quantitative determination of triethanolamine in cements. Cem Concr Res 1993;23(3):603–8. [12] Kong X, Lu Z, Yan J, et al. Influence of triethanolamine on elemental concentrations in aqueous phase of hydrating cement pastes. J Chin Ceram Soc 2013;41(7):981–6. [13] Cheung J, Jeknavorian A, Roberts L, et al. Impact of admixtures on the hydration kinetics of Portland cement. Cem Concr Res 2011;41(12):1289–309. [14] Taylor HFW, Famy C, Scrivener KL. Delayed ettringite formation. Cem Concr Res 2001;31(5):683–93. [15] Ukrainczyk N, Matusinovic T, Kurajica S, et al. Dehydration of a layered double hydroxide-C2AH8. Thermochim Acta 2007;464(1):7–15.
Contents lists available at ScienceDirect
Construction and Building Materials journal homepage: www.elsevier.com/locate/conbuildmat
Mechanism of triethanolamine on Portland cement hydration process and microstructure characteristics Jianguo Han a,b,⇑, Kejin Wang b, Jiyao Shi c, Yue Wang c a
Department of Civil Engineering, Tsinghua University, Beijing 100084, China Department of Civil, Construction and Environmental Engineering, IOWA State University, 50014, USA c Technology Center of China Railway Tunnel Co., Ltd., Luoyang 471000, China b
h i g h l i g h t s TEA accelerates hydration of C3A, promoting hydration heat release rate. When TEA introduced, after depletion of calcium sulfate, AFt is converted to AFm. TEA retards hydration of C3S, which can jeopardize concrete strength at early age. Influence of TEA on cement paste pore structure is dosage dependent. Balance should be made when using TEA, on its accelerating and retarding effect.
a r t i c l e
i n f o
Article history: Received 19 August 2014 Received in revised form 30 March 2015 Accepted 8 June 2015 Available online 20 June 2015 Keywords: TEA Acceleration Hydration Microstructure AFt AFm
a b s t r a c t Influence of triethanolamine (TEA) on Portland cement hydration heat evolution process, hydration product, pore structure, setting time and concrete strength were investigated; and mechanism of TEA on the cement hydration process and microstructure is discussed. Research results indicate that TEA accelerates hydration process of C3A and retards that of C3S, such effects enhance with the increase of TEA dosage. TEA also promotes rate of AFt formation and its conversion to AFm. As a result, TEA accelerates setting of cement paste but jeopardizes strength of concrete. Depending on its dosage, TEA can either optimize or harm cement paste pore structure. The types of calcium sulfoaluminate hydrate (AFt or AFm) produced in cement paste containing TEA depends on the abundance of calcium sulfate, before depletion of calcium sulfate, AFt is produced, after that, AFm is produced (converted from AFt). Ó 2015 Elsevier Ltd. All rights reserved.
1. Introduction Triethanolamine (TEA) is a weak base surfactant. It can be used as grinding aid for cement manufacturing (typically at dosage of 0.1%, weight percentage of clinker), in order to prevent agglomeration of powder and coating formation on milling ball surface [1]. TEA can also be used as component of chemical admixture, such as shotcrete accelerator or water reducing admixture, in order to accelerate cement setting or counteract the retarding effect of other component [2,3]. Previous research indicates that: (1) effect of TEA on cement hydration is objective oriented: it accelerates hydration process of C3A but retard that of C3S [4]. Ramachandran conducted a series of studies to investigate the influences of TEA on hydration process of C3A, C3A–gypsum and ⇑ Corresponding author at: Department of Civil Engineering, Tsinghua University, Beijing 100084, China. http://dx.doi.org/10.1016/j.conbuildmat.2015.06.018 0950-0618/Ó 2015 Elsevier Ltd. All rights reserved.
C3S and found that TEA accelerates C3A hydration and C3A–gypsum reaction by promoting formation of aluminate hydrate and AFt, and retards hydration process of C3S by extending the induction period [5–7]; (2) effect of TEA on cement hydration is dosage dependent also. At small dosage (e.g. 0.02%, weight percentage of cement) it acts as setting accelerator, at higher dosage (e.g. 0.5%) it acts as setting retarder, at more higher dosage (e.g. 1.0%) it acts as setting accelerator once again [8]. Despite all the work have been done, the influence of TEA, especially at high dosage (higher than 1.0%), on Portland cement setting behavior and strength development still need further research, and the working mechanism of TEA on Portland cement hydration process and microstructure characteristics merit further investigation. Such research can facilitate TEA application and understanding of its effect in Portland cement manufacturing and utilization. In this paper, TEA dosage from 0.02% to 8% was adopted, and its influence on Portland cement hydration process, hydration
458
J. Han et al. / Construction and Building Materials 93 (2015) 457–462
products, setting time, pore structure and concrete strength were studied.
Table 2 Cement paste mix proportion and properties measured.
2. Materials and mix proportions Materials used include: type I Portland cement with chemical composition and fineness shown in Table 1; deionized water with resistivity no less than 10.0 MX cm; River sand with fineness modulus of 3.1; limestone with particle size ranging from 5 to 10 mm; polycarboxylate-based water reducing admixture (WRA); TEA was a liquid chemical reagent. Cement paste mix proportion and properties measured are shown in Table 2. Concrete mix proportion is shown in Table 3, and 100 mm cubes are used for compressive strength evaluation.
3. Testing methods Hydration heat evolution process of cement pastes was monitored by a differential scanning calorimeter (DSC) (ToniCAL, manufactured by Toni Company), under constant temperature of 25 °C. The preparation of cement pastes for microstructure investigation consists of following steps: (1) Conducting a complete DSC test (0–24 h) for a cement paste. (2) Selecting a few key time points on the DSC curve so as to capture features of the hydration process of the paste. (3) Performing a series of DSC tests as described in Step (1) for the same cement paste; but, when the test reaches each of the key time points as described in Step (2), took the cement paste sample out of the DSC instrument chamber. (4) Cracking the samples into small pieces and submerging these small pieces into ethanol so as to discontinue the hydration process (It took less than 2 min from taking the sample out of the DSC instrument chamber to submerging the cracked pieces into ethanol; the volume ratio of ethanol to cracked pieces was about 10:1). (5) Studying the microstructure (hydration product and pore structure) of the cement paste. Cement paste samples were taken out of ethanol and oven dried at 60 °C for 6 h. The oven dry process was to evaporate ethanol in capillary pore; then, pulverized using an agate pestle and mortar and passed through an 80 lm sieve, if powder was needed for measurement. Hydration products of the above cement pastes were identified by X-ray diffractometer (Automated D/max, Rigaku Corporation), with a Cu Ka source and a scanning speed of 2° per minute. TG–DSC–MS hyphenated method was used to verify the hydration product further. Using the TG–DSC instrument (STA 449 F3, Netzsch Company), under a controlled regime (from 50 to 1100 °C, heating rate of 20 °C/min; air atmosphere), the mass change and the difference of energy input between cement paste and reference material were measured simultaneously. Meanwhile, the gas released during TG–DSC testing process was transported to the MS instrument (QMS 403C, Netzsch Company) via capillary tube, and the input gas was ionized first and then classified and identified based on mass to charge ratio. Pore structure of the paste samples was measured by BET absorption method. Cement paste samples were vacuum
Table 1 Chemical composition and fineness of Portland cement. Oxide (wt.%) SiO2
Al2O3
Fe2O3
CaO
MgO
SO3
Na2Oeq
f-CaO
21.58 C3S 57.34
4.03
3.46 C2S 18.90
61.49
2.60 C3A 6.47
2.83
0.51 C4AF 11.25
0.67
Blaine fineness (m2/kg) 346 LOI (wt.%) 1.97
*
Water to cement ratio (weight percentage)
Dosage of TEA* (weight Properties measured percentage of cement) (%)
0.4
0, 0.02, 0.2, 2 and 6
0.4
0, 0.2 and 6
0.4
0, 2, 4, 6 and 8
Hydration heat evolution process Microstructure (including: hydration product and pore structure) Setting time
No water was deducted when TEA was introduced.
Table 3 Concrete mix proportion. No.
Cement (kg/m3)
Water (kg/m3)
W/C
Sand (kg/m3)
Aggregate (kg/m3)
WRA (%)
TEA** (%)
C-TEA-0 C-TEA-5
490 490
196 196
0.4 0.4
830 830
880 880
1.5 1.5
0 5
** Weight percentage of cement, and no water was deducted when TEA was introduced.
de-aerated for 6 h at 100 °C before measurement. Nitrogen (N2) gas was used as adsorbate. Measurement was performed by a surface area and pore analyzer (Quadrasorb SI, Quantachrome Company). Pore volume and pore size were calculated based on the amount of N2 adsorbed. Setting time of cement pastes with different dosages of TEA was measured according to ASTM C191 by Vicat apparatus, under temperature of 23 ± 2 °C. Compressive strength of concrete was measured at 1 day of age according to ASTM C39. 4. Results and analysis 4.1. Effect of TEA on hydration heat evolution process During Portland cement hydration process monitored by DSC method, there are usually 5 stages: (A) initial reaction period, (B) induction period, (C) acceleration period, (D) deceleration period, and (E) slow reaction period [9]. There are usually two peaks during the hydration process, the first peak, in initial reaction period, is mainly resulted from ion dissolution and reaction between C3A and calcium sulfate, usually resulting in formation of AFt; the second peak, at the end of acceleration period, is mainly resulted from hydration of C3S, forming C–S–H and calcium hydroxide (CH) [10]. Hydration heat release rate curves of cement pastes with different dosage of TEA are given in Fig. 1. It can be seen in the figure, as the TEA dosage increases, the hydration heat release rate in initial reaction period is promoted, the induction period is prolonged, and the acceleration period is postponed. When TEA dosage is higher than 2%, no acceleration period can be identified during the 24 h measurement. An affiliated peak is noted right after the main peak in the initial reaction period, which corresponds to the conversion of AFt to AFm as explained by XRD measurement in later part of this paper. The promoted hydration heat release rate in initial reaction period is due to the accelerating effect of TEA on reaction of C3A with calcium sulfate. TEA can chelate with some cation ions, such as Al3+, Fe3+ and Ca2+, under highly alkaline medium [11,12]. When chelated with Al3+, TEA can promote formation of calcium sulfoaluminate hydrate (AFt and AFm) and accelerate hydration process of C3A. The postponed hydration process of C3S after the introduction of TEA might be attributed to TEA’s chelating ability with Fe3+ and Ca2+ cations. Due to mineral structure characteristics, the
459
J. Han et al. / Construction and Building Materials 93 (2015) 457–462
(a) Without TEA
Fig. 1. Hydration heat release rate curves of cement pastes with different dosage of TEA.
hydration ability of C4AF, C3S and C2S are inferior to that of C3A. Before acceleration period of hydration process, majority of the Fe3+ in C4AF and Ca2+ in C3S and C2S are still in site with the original minerals. As a result, a surface complex is formed on C3S surface, which acts as a barrier, isolating C3S from water and hindering its hydration process [5]. Meanwhile, the accelerated hydration of C3A can speed up the depletion of calcium sulfate, as proved by the XRD measurement in the later part of this paper. After the depletion of calcium sulfate, C3A can hydrate alone to form C2AH8 and C4AH19, which belongs to AFm phase. AFm is poor crystalline hydration product, which can cover the surface of cement particle and hinder hydration process of C3S [13]. That is, the formation of C2AH8 and C4AH19, resulting from hydration of C3A, may be another cause for the retarding effect of TEA on C3S hydration process, especially at high dosage, in which the depletion of calcium sulfate can be easily achieved. Based on the characteristic of hydration heat release rate curves in Fig. 1, TEA dosage of 0%, 0.2% and 6% were chosen for investigating the influence of TEA on microstructure of cement paste. Time points selected on these curves are given in Fig. 2, and hydration was discontinued at these time points, by taking sample out of the DSC instrument chamber and submerging the cracked pieces into ethanol.
4.2. Effect of TEA on hydration products and pore structure At each of the time points shown in Fig. 2, the microstructure of cement paste, including hydration products and pore structure, was studied. XRD patterns of cement pastes are given in Fig. 3. From Fig. 3(a) (without TEA), it can be seen that anhydrite is the main calcium sulfate in cement. During cement hydration process, anhydrite is consumed and AFt is produced (at time point 5#, or the age of 24 h). As illustrated in Fig. 3(b), cement paste with 0.2% TEA possesses XRD pattern similar to that of control specimen (without TEA). However, AFt is identified at time point 4# (18.5 h of age), manifesting the accelerating effect of TEA on AFt formation. AFt peak disappeared at time point 5# in Fig. 3(b) is probably due to the conversion of AFt to AFm. AFm is usually of poor crystalline when compared with AFt, and it is more difficult to be identified by XRD. As shown in Fig. 3(c), when 6% TEA was used, XRD pattern of cement paste appears very different with that of control and 0.2%
(b) With 0.2% TEA
(c) With 6% TEA Fig. 2. Time points on hydration heat release rate curves.
TEA samples. This is probably due to the quick consumption (before and at time point 1#) of anhydrite during the accelerated C3A hydration. The quantity of CH produced in the whole hydration process is negligible, due to the retarded hydration process of C3S. Meanwhile, AFm, instead of AFt, is formed after time point 1#, as the peak representing calcium sulfoaluminate hydrate is shifted from AFt to AFm. The peak appeared at 26.6° in Fig. 3(b) and (c), which is not emerged in control specimen, needs further investigation. DSC curves (got from TG–DSC measurement) of cement paste with different dosage of TEA are given in Fig. 4. MS measurement was performed on the last time points in Fig. 2 during TG–DSC testing, and test results are given in Fig. 5. The endothermal peak appeared on DSC curve in Fig. 4 can be used to further identify hydration products. MS curve in Fig. 5 can be used to certify hydration product identification made by TG–DSC method. Base on DSC and MS results in Figs. 4 and 5, hydration products and their thermal behavior during the TG– DSC–MS measurement are summarized in Table 4. The following observations can be made from Fig. 4: (1) Comparison of endothermal peak intensity (between different time points, of same TEA dosage) produced by decomposition of CH is stronger in control sample than that in TEA containing samples. (2) TEA has retarded hydration process of C3S, and this effect is intensified with the increase of TEA dosage. (3) AFt is produced in control and 0.2% TEA samples, and AFm is produced in the 6% TEA sample, which is consistent with the findings resulting from XRD measurement.
460
J. Han et al. / Construction and Building Materials 93 (2015) 457–462
( a ) W i th o u t TE A
(b) With 0.2% TEA
(C) With 6% TEA Fig. 3. XRD pattern of cement pastes with different dosage of TEA.
(4) AFm in Fig. 4(a) and (b) is produced by calcination of AFt [14], and this AFm decomposes at 880 °C. (5) AFm in Fig. 4(c) is converted from AFt during hydration process, which loses water at 180 and 280 °C, and decomposes at 980 °C. (6) At high temperature, AFm will lose the interlayer sulfate and crystallizes into C3A and C12A7, releasing SO2 (gas) [15]. (7) The decomposition temperature of the two type of AFm, one from calcination of AFt and one from conversion of AFt during hydration, is different; the former is at about 880 °C and the latter is at about 980 °C. Pore structure of cement paste with different dosages of TEA at the age of 24 h is given in Fig. 6. Generally, pores with a diameter within 10–100 nm are considered as capillary pores, and the amount of capillary pores is related to the degree of hydration. As cement hydration progresses, the hydration products will fill in particle gaps and refine the pore structure. Pore structure is a key parameter which can bring influences to mechanical properties and anti-penetration ability. As seen in Fig. 6, cement paste with 0.2% TEA has narrower and finer pore size distribution than cement paste without TEA, probably due to the accelerating effect of TEA on C3A hydration process.
Cement paste with 6% TEA has broader and coarser pore size distribution, probably due to the retarding effect of TEA on C3S hydration process. 4.3. Effect of TEA on setting time and strength Influence of TEA on Portland cement setting time is shown in Fig. 7. The figure indicates that both initial and final setting is accelerated significantly with TEA addition, from more than 200 min to less than 50 min. The concave up shape caused by TEA dosage of 8% indicates that the acceleration effect is less active after the dosage of TEA reaches 6%. This is probably because the depletion of anhydrite in cement paste with 6% TEA, which occurs almost immediately after the mixing of cement paste (see Fig. 3(c)). Meanwhile, at a higher dosage, the increased liquid-to-solid volume resulted from the redundant TEA can lower density of cement paste, and the retarding effect of TEA on C3S hydration process may contribute to this result too. Compressive strength of concrete with 0% and 5% TEA at the age of 1 day was measured. Control concrete (without TEA) produced compressive strength of 16.2 MPa and 5% TEA containing specimen produced compressive strength of 3.5 MPa (21.6% of control specimen). The results indicate that although TEA can reduce cement
461
J. Han et al. / Construction and Building Materials 93 (2015) 457–462
(a) Without TEA
(b) With 0.2% TEA
(c) With 6% TEA Fig. 4. DSC curve of cement pastes with different dosage of TEA.
Table 4 Hydration products and its thermal behavior during TG–DSC–MS measurement. Temperature/°C (approximately)
Hydration product
Thermal behavior
Material released
120 180 and 280 460 720 880 980
C–S–H or AFt AFm CH CaCO3 AFm* AFm
Losing water Losing water Decomposition Decomposition Decomposition Decomposition
H2O (gas) H2O (gas) H2O (gas) CO2 SO2 SO2
*
Produced by decomposition of AFt during calcination. Endothermal peak is not obvious on DSC curve in Fig. 4(c).
further more, it was manifested in Ramachandran’s research that TEA can compromise concrete later age strength in addition to early age strength [7]. Fig. 5. MS spectrum of cement pastes with different dosage of TEA.
4.4. TEA working mechanism on Portland cement paste set time, TEA can significantly reduce early age strength of concrete, due to the retarding effect of TEA on C3S hydration process. Ramachandran has observed this phenomenon also. And
Base on the effect of TEA on cement hydration process, microstructure, setting time and concrete strength, the working mechanism of TEA on Portland cement can be concluded as following:
462
J. Han et al. / Construction and Building Materials 93 (2015) 457–462
pore structure. So, balance about the accelerating and retarding effect of TEA on different mineral in cement should be made when utilizing TEA in Portland cement. 5. Conclusion
Fig. 6. Pore size distribution of cement pastes with different dosage of TEA.
TEA accelerates reaction of C3A with calcium sulfate in Portland cement, and this effect enhances with the increase of TEA dosage. As a result, TEA can enhance hydration heat release rate during initial reaction period, promote AFt forming rate and conversion rate of AFt to AFm, and accelerate cement setting. In TEA introduced specimen, after depletion of calcium sulfate, AFt is converted to AFm, which is the same as that in common cement. However, TEA can speed up the conversion rate; thus, AFm can be identified shortly after the beginning of hydration at high dosage. TEA retards hydration process of C3S, and this effect enhances with the increase of TEA dosage also. AS a result, concrete strength at early age can be jeopardized by TEA. Influence of TEA on cement paste pore structure depends on its dosage; TEA can optimize pore size distribution at low dosage (e.g. 0.2% TEA) and injure pore size distribution at high dosage (e.g. 6% TEA). Balance should be made when utilizing TEA in Portland cement, by considering its accelerating effect on C3A and retarding effect on C3S.
References
Fig. 7. Effect of TEA on cement paste setting time.
(1) TEA accelerates reaction of C3A with calcium sulfate (gypsum, hemihydrate or anhydrite) to produce calcium sulfoaluminate hydrate (AFt and AFm), and this effect enhances with the increase of TEA dosage. (2) In TEA introduced specimen, after depletion of calcium sulfate, AFt is converted to AFm, which is the same as that of common cement: before depletion of calcium sulfate, AFt is produced; after that, AFm is produced. However, TEA can speed up the conversion of AFt to AFm, resulting in the appearance of AFm shortly after the beginning of cement hydration process. (3) TEA retards hydration process of C3S by prolonging the introduction period, and the retarding effect enhances with the increase of TEA dosage. (4) AS a result, introduction of TEA can enhance setting behavior; however, it can jeopardize early age strength; and depending on its dosage, TEA can either optimize or harm
[1] Sohoni S, Sridhar R, Mandal G. The effect of grinding aids on the fine grinding of limestone, quartz and Portland cement clinker. Powder Technol 1991;67(3):277–86. [2] Connolly JD, Hime WG. Analysis of concrete for triethanolamine. Cem Concr Res 1976;6(6):741–6. [3] Taylor HFW. Cement chemistry. 2nd ed. London, U.K.: Thomas Telford; 1997. p. 157–176 and p. 323–338. [4] Hewlett P. Lea’s chemistry of cement and concrete. 4th ed. U.K.: Elsevier; 2004. p. 881–887. [5] Ramachandran VS. Influence of triethanolamine on the hydration characteristics of tricalcium silicate. J Appl Chem Biotech 1972;22(11): 1125–38. [6] Ramachandran VS. Action of triethanolamine on the hydration of tricalcium aluminate. Cem Concr Res 1973;3(1):41–54. [7] Ramachandran VS. Hydration of cement—role of triethanolamine. Cem Concr Res 1976;6(5):623–31. [8] Gartner E, Myers D. Influence of tertiary alkanolamines on Portland cement hydration. J Am Ceram Soc 1993;76(6):1521–30. [9] Taylor HFW. Cement chemistry. 2nd ed. Thomas Telford: London, U.K; 1997. p. 150–225. [10] Bullard JW, Jennings HM, Livingston RA, et al. Mechanisms of cement hydration. Cem Concr Res 2011;41(12):1208–23. [11] Yilmaz VT, Menek N, Odabasogˇlu M. Quantitative determination of triethanolamine in cements. Cem Concr Res 1993;23(3):603–8. [12] Kong X, Lu Z, Yan J, et al. Influence of triethanolamine on elemental concentrations in aqueous phase of hydrating cement pastes. J Chin Ceram Soc 2013;41(7):981–6. [13] Cheung J, Jeknavorian A, Roberts L, et al. Impact of admixtures on the hydration kinetics of Portland cement. Cem Concr Res 2011;41(12):1289–309. [14] Taylor HFW, Famy C, Scrivener KL. Delayed ettringite formation. Cem Concr Res 2001;31(5):683–93. [15] Ukrainczyk N, Matusinovic T, Kurajica S, et al. Dehydration of a layered double hydroxide-C2AH8. Thermochim Acta 2007;464(1):7–15.