Thaumasite formation by hydration of sulphosilicate clinker

Journal Pre-proof Thaumasite formation by hydration of sulphosilicate clinker Karel Dvořák, Dalibor Všianský, Dominik Gazdič, Marcela Fridrichová, Danute Vaičiukynienė PII: S2352-4928(20)32460-0 DOI: https://doi.org/10.1016/j.mtcomm.2020.101449 Reference: MTCOMM 101449 To appear in: Materials Today Communications Received Date: 11 May 2020 Revised Date: 3 July 2020 Accepted Date: 9 July 2020 Please cite this article as: { doi: https://doi.org/ This is a PDF file of an article that has undergone enhancements after acceptance, such as the addition of a cover page and metadata, and formatting for readability, but it is not yet the definitive version of record. This version will undergo additional copyediting, typesetting and review before it is published in its final form, but we are providing this version to give early visibility of the article. Please note that, during the production process, errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain. © 2020 Published by Elsevier. Thaumasite formation by hydration of sulphosilicate clinker Karel Dvořák 1, Dalibor Všianský 2,*, Dominik Gazdič 3, Marcela Fridrichová 4, Danute Vaičiukynienė 5 Brno University of Technology, Faculty of civil engineering, Czech Republic Masaryk University, Faculty of Science, Department of Geological Sciences, Brno, Czech Republic * Correspondence: [email protected]; Tel.: +420 549 49 7813 3 Brno University of Technology, Faculty of Civil Engineering, Czech Republic 4 Brno University of Technology, Faculty of Civil Engineering, Czech Republic 5 Kaunas University of Technology, Department of Building Materials, Lituania 1 2 Highlights Hydration of ternesite containing sulphosilicate clinkers was monitored for up to 810 days.  Thaumasite can be formed as one of the hydration products of sulphosilicate clinkers.  Besides indisputable ecological benefits of the application of sulphoaluminate oo f  A novel way of thaumasite synthesis is demonstrated. e-  pr cements, the possible risk of sulphate attack should also be considered. ur na l Pr Abstract: Extensive research on sulphosilicate clinkers has been performed, primarily given by the ecological aspects. The formation of thaumasite and ettringite has become an important issue due to possible concrete deterioration, and also reducing greenhouse gas emissions. This work was aimed at hydration of ternesite based clinkers at low temperatures, with respect to thaumasite formation. Five clinkers with various compositions corresponding to stoichiometric ratios of elements in thaumasite and ternesite were prepared. Clinker-water suspensions were stored at 5 °C for up to 810 days. X-ray diffraction, DSC/TG, and SEM/EDS were used as the analytical tools. All the clinkers contained ternesite, belite, anhydrite, and free lime. As shown by the results, ternesite and belite preferably hydrate to CSH gels, and sulphate ions forming ternesite become a part of gypsum. The remaining portion of gypsum originated from anhydrite hydration. Gypsum reacted with silica and carbon dioxide to form thaumasite. Thaumasite was clearly identified already after 28 days of hydration, and for one sample, its content finally reached approximately 34 wt. %. It should be considered that the application of sulphoaluminate cements may possibly be connected with the sever risk of sulphate attack. Jo Keywords: sulphosilicate clinker; hydration; ternesite; thaumasite; sustainability 1. Introduction Thaumasite and ettringite formation has become an important issue due to concrete deterioration and also reducing greenhouse gas emissions. The reduction of energy consumption motivates the research of non-traditional binders, which may strongly contribute to sustainable development of building industry. Extensive research on calcium sulphoaluminate (CSA), belite – calcium sulphoaluminate – ternesite (BST), and belite cements is currently conducted [1] as they are alternatives to ordinary Portland cement (OPC). Their production is connected to remarkably lower CO2 emissions, because the raw meal contains less CaCO3. The maximum burning temperature reaches only 1250 - 1350 °C, which is directly connected with lower energy consumption. They can also be manufactured from secondary raw materials. Clinkers of these new types of cements are usually based on the system of ye´elimite (C4A3𝐒̅) - belite (C2S), with a small amount of ferrite phase [2]. Ternesite (C5S2𝐒̅) is a sulphosilicate mineral analogous to sulphoaluminate ye´elimite (C 4A3𝐒̅). The basic difference is the substitution of Si by Al ions in the structure. Ternesite can be found in the green crust rings in the cooler areas of rotary kilns. Ternesite mostly forms by the reaction of anhydrite and belite at ca 1250 °C [3] and in an open system decomposes at 1350 °C [1]. The ternesite synthesis methodology based on stoichiometric belite and anhydrite mixture burning has been protected by the WO2013023729 patent [4]. Although ternesite has been considered to be almost non-reactive, several recent studies reported that ternesite hydrates to form AFt and C-S-H phases, contributing to binder hardening and strength and can be even more reactive than belite [5]. Some authors consider the AFt phase to be analogous to ettringite, but with silicate ions substituting the aluminate ones, and carbonate ions replacing some of the sulphate ions. The mineral corresponding to this composition is thaumasite (CaSiO3∙CaSO4∙CaCO3∙15 H2O) [6] with hexagonal symmetry [7]. Pr e- pr oo f Due to low aluminium content, thaumasite usually forms in ordinary and sulphate-resistant cements when stored at low temperatures. However, Diamond (2003) [8] proved that thaumasite can form in warm conditions as well. Others characterized various conditions for successful synthesis of thaumasite [9, 10, 11]. High humidity, temperature below 15 °C, ideally between 0 – 5 °C, pH over 11, and a source of silica, sulphate and carbonate ions are crucial. Calcium salts feature higher solubility at lower temperatures. Due to the necessity of low temperatures, synthesis of thaumasite typically takes several months. Thaumasite is stable in a relatively narrow range of pH between 9 to 13. At the pH over 13 it remains in the solution and below 9 decomposes to gypsum, carbonates, and several other phases [12, 13]. The detrimental effect of thaumasite to concrete was firstly identified by Erlin & Stark (1965) [14]; its effect can be similar to delayed ettringite formation (DEF). However, thaumasite attacks CSHs directly [9], due to which is it more dangerous than sulphate attack caused by DEF [15]. ur na l Several authors attempted to prepare thaumasite, mostly by mixing Portland cement with a source of (SO4)2- and (CO3)2- ions [16, 17, 18, 19, 20]. Aguilera et al. (2001) [10] prepared thaumasite by mixing the solutions of 10 % sugar with Na2SO4, Na2SiO3, and Na2CO3, with the additions of CaO, or Ca(OH)2. The mixture was exposed to the temperature of 5 °C. Purnell et al. (2003) [21] synthesised thaumasite by adding alitic clinker, alumina, and calcium carbonate into a MgSO4 solution which they stored for 100 days at 5 °C. Another method was reported by Skalamprinos et al. (2018) [22]; the solid state reaction approach to the synthesis of single ternesite phase is reported in this research. These authors also reported “thaumasite-like phases” forming already in the early stages of hydration. However, they used a solvent (acetone) exchange technique to terminate the hydration. They state that the formation of the “thaumasite-like phases” was most likely due to a secondary reaction of C-S-Hs in the presence of acetone, gypsum, and carbon dioxide. Jo The aim of this work was to study the hydration of ternesite containing sulphosilicate clinkers of various compositions at low temperature, with respect to thaumasite formation. Despite the fact that their production is more ecological than of OPC, application of sulphosilicate clinkers appears to possibly be sulphate-attack risky in this point. Furthermore, ternesite hydration process is still not fully understood. 2. Materials and Methods Five ternesite clinkers of various raw meal compositions (Tab. 1) were prepared and subsequently subjected to hydration. The samples were designated as follows. Raw meals: 1 – 5, clinkers: 1c – 5c, hydrated samples: 1h - 5h (+ number of days of hydration). The raw meal chemical composition of sample no. 1 corresponds to the stoichiometric ratio of elements in thaumasite, and the composition of sample no. 5 corresponds to ternesite. Precipitated gypsum (CaSO4·2H2O), calcium carbonate (CaCO3), and amorphous silica (SiO2), all in p.a. purity, were used as raw materials. Table 1. Phase composition of raw meals in wt. %, rounded to first decimal place. Sample no. Chemical composition [mol] CaO SiO2 SO3 Dosage [%] CaCO3 SiO2 CaSO4·2 H2O 1 3 1 1 46.30 13.89 39.81 2 5 1 1 63.29 9.49 27.22 3 7 2 1 67.26 13.45 19.28 4 6 2 1 63.13 15.15 21.72 5 2 1 57.80 17.34 24.86 The clinkers were prepared by the solid-state reaction. The raw meal was milled and homogenized in distilled water using a PULVERISETTE 6 planetary mill; the used water/powder ratio was 0.8. The volume of the hardened steel milling capsule, 25 grinding steel balls with 20 mm in diameter in which were inserted, was 0.5 dm3. The grinding time was 15 minutes and the speed was 400 rpm. The dosage of the raw meal slurry was 300 ml in total. The slurry was dried in a laboratory dryer at 200 °C for 2 hours, nodules of 10-15 mm in diameter during which formed naturally. Dried nodules were dosed (approximate amount of 50 g) into platinum crucibles. The burning process was carried out in a Classic 2017S laboratory kiln under the temperature of 1200 °C, 60 min soaking time, and 8 °C/min heating rate. The samples were subsequently quenched by air stream. Mineralogical composition of the clinkers was monitored by XRD. The burned samples were milled in the above mentioned planetary mill at 350 rpm until the grain size was under 0.063 mm. 200 g of the powders were mixed with 600 ml of distilled water for two minutes. The suspensions were placed into 1.1 dm3 plastic bottles and stored in a climate chamber at 5 °C. The atmospheric air dissolved in the water (being naturally replenished from the 250 ml free volume of atmosphere above the suspension) was used as the source of CO2 for thaumasite development. The samples were shaken regularly at weekly intervals and the atmosphere above the suspension was also changed to keep the concentration of CO2 common in the natural atmosphere. The pH value of all the samples was monitored with a AD8000 Professional Multi-Parameter pH- Bench Meter, with the resolution of 0.01 pH and accuracy of ±0.01. The prepared and examined samples were designated as follows: clinkers 1c – 5c, The hydrated samples were always collected by a pipette in the approximate amount of 25 ml of suspension and subsequently dried at 35 ° C in a laboratory dryer. All the produced samples were analysed by powder X-ray diffraction (XRD) after 28, 56, 180, and 810 days of hydration. Samples collected after 810 days of hydration were also studied by thermal analysis (TA). Both the samples after 180 and 810 days of hydration were examined by the means of scanning electron microscopy and energy-dispersive X-ray microanalysis (SEM/EDS). XRD analysis of pulverized samples was conducted using a Panalytical X´Pert PRO MPD diffractometer powered at 40 kV and 30 mA, equipped with a cobalt tube (λKα = 0.17903 nm), Fe filter, and 1-D RMTS (X’Celerator) detector at the Bragg–Brentano parafocusing reflection geometry. Step size: 0.016° 2Θ, time per step: 100 s, angular range: 6-100° 2Θ, and total scan duration of 75 min and 17 s. The acquired data was processed using the Panalytical HighScore 4.5 plus software and the ICSD 2012 databases. The quantitative phase analysis was done by the Rietveld method. The amorphous phase was refined and quantified using a hkl-fit after a previous calibration. Thermal analysis consisting of simultaneous thermogravimetry (TGA) and differential scanning calorimetry (DSC) was carried out using a Seteram Setsys Evolution 1750 instrument. The analysis was conducted in a dynamic air atmosphere with the constant heating rate of 10 °C/min between 40 and 1000 °C. The acquired data was processed using the Seteram Processing software. Jo ur na l Pr e- pr oo f 5 Gold coated samples were subjected to SEM/EDS examination using a Tescan Mira 3 instrument powered at 10 kV. The analysis was focused on the identification and morphology of thaumasite and the other hydration phases. 3. Results 3.1. Sulphosilicate clinkers oo f Five sulphosilicate clinker samples, designated 1c – 5c, were prepared as the starting material for hydration experiments. The dominant phase of all the samples was ternesite (C5S2S̅). The contents of other phases, belite (β-C2S), anhydrite (CaSO4), and lime (CaO) (Tab. 2) were derived from the raw meal composition (Tab. 1). Small amounts of cristobalite (SiO2), below 0.5 wt. %, were also detected in each clinker sample. The amounts of cristobalite did not change significantly during the subsequent hydration. Diffractograms of clinkers are shown in Fig. 1. Table 2. Phase composition of sulphosilicate clinkers in wt. %, rounded to first decimal place. Sample Ternesite Larnite Anhydrite Lime 1c 47.1 22 28.9 2.0 2c 51.1 6.5 20.5 3c 63.6 14.0 5.8 16.6 4c 59.4 23.9 7.0 9.7 5c 44.6 37.1 e- pr 22.0 1.9 Jo ur na l Pr 16.4 Figure 1. Diffractograms of synthetized sulphoslicate clinkers designated 1c – 5c (24.5 – 45 °2Θ range); Anh = anhydrite, Ter = ternesite, C2S = βC 2S (belite), C = CaO (lime), Crs = cristobalite. 3.2. Hydration of sulphosilicate clinkers The phase composition of hydration products of the clinkers 1c – 5c, designated 1h – 5h, as well as the pH values, were monitored after 28, 56, 180, and 810 days. The pH was reaching the value of 12.5 – 12.6 after 28 days and was relatively stable during the whole experiment. After 810 days, the pH values were still in the range between 11.9 and 12.2. The contents of thaumasite (CaSiO3∙CaSO4∙CaCO3∙15 H2O), portlandite (Ca(OH)2), gypsum (CaSO4·2 H2O), calcite (CaCO3), and amorphous phase were quantified. Based on the results, the amorphous phase is formed predominantly by calcium hydrosilicate (C-S-H) gels. In the samples rich in thaumasite, low contents of calcite are probably not detected due to overlapping of the individual peaks. Jo ur na l Pr e- pr oo f Variations of the phase composition during hydration are shown in Fig. 2 as quantified by Rietveld analysis of the XRD data collected at each sampling point. Numerical values are given in Appendix A. Figure 2. a) to e) development of all quantified phases’ contents in samples 1h to 5h during hydration; f) comparison of thaumasite content in all five samples; horizontal axis is in square root scale Pr e- pr oo f Comparison of the diffractograms of the samples after 810 days of hydration is shown in Fig. 3. Elevated background between ca 30 – 40° 2Θ indicates the presence of amorphous phase, the content of which gradually increases with time. na l Figure 3. Diffractograms of the samples after 810 days of hydration designated 1h – 5h (8 – 47 °2Θ range); Th = thaumasite, Gp = gypsum, CH = Ca(OH)2 (portlandite), C2S = βC2S (belite), Crs = cristobalite, Cal = calcite Jo ur Lime hydration is the most intense, as it immediately converts to portlandite. Subsequently, portlandite reacts to form the other hydration phases. Anhydrite starts to hydrate in the beginning phases of the experiments to form gypsum. The reaction accelerates between 28 and 56 days. Except for 1h sample, the belite content decreases substantially between 28 and 56 days. Ternesite hydration generally proceeds within the first 180 days. Thaumasite, as well as amorphous CSHs, appears already after 28 days of hydration and its content increases rapidly till 180 days, after which the individual increments of the thaumasite content decrease. On the contrary, the CSHs content is slightly reduced, or remains unchanged, after 180 days. The compositions of the samples after 810 days of hydration, with respect to thaumasite content, were also studied using thermal analysis. The amount of thaumasite was calculated based on the weight lost resulting from decarbonation (Fig. 4a); the thaumasite and calcite decarbonation effects are relatively well separated on the first derivative of the TG curve (dTG), which enables thaumasite quantification. The acquired results correspond to those of XRD measurements quite well (Tab. 3). Table 3. Comparison of thaumasite content quantified by XRD and thermal analysis (TG and dTG), and temperature range of thaumasite decarbonation Sample TG (810 d) Temperature (°C) XRD (810 d) 1h 32.7 694.4-744.7 36.2 2h 14.4 673.1-711.5 14.7 3h 23.7 648.7-711.6 24.4 4h 19.4 604.9-705.0 18.4 5h 11.9 594.5-669.8 11.7 Pr e- pr oo f The course of the heat flow curves (Fig 4b) reflects the phase composition of the analysed samples. The first three endothermic effects are due to gypsum dehydration to hemihydrate (basanite) and anhydrite II, subsequently. The effect of the first stage of thaumasite dehydration occurs between these two peaks, while the shallow effect of the second thaumasite dehydration stage can be observed below 300 °C. The strong endothermic effects between ca 420 and 500 °C are due to portlandite, and also amorphous CSHs dehydration. The last two endothermic peaks above ca 650 °C result from thaumasite decarbonation followed by calcite decarbonation. The exothermic effects with maximum at ca 590 – 600 °C and above 820 °C are likely associated with dicalcium silicate (C2S) formation and γ to β transition respectively, because no weight loss connected to these effects was observed. (b) (a) ur na l Figure 4. Samples after 810 days of hydration: TG) and dTG curves for sample 2h (a); heat flow curves (b). Effects observed on heat flow curves (b): I - loss of adsorbed water, II, III – gypsum and bassanite dehydration overlapping with dehydration of thaumasite, IV – portlandite dehydration, V – crystallization of γ-C2S, VI – thaumasite decarbonation, VII – decarbonation of calcite, VIII – γ to β-C2S transition Jo All the hydration phases identified by XRD, were observed by SEM/EDS as well (Fig 5 a – d,. examples of EDS spectra are given in Appendix B). (a) (b) (c) (d) oo f Figure 5. SEM images: comparison of thaumasite crystals’ morphology and size after 180 and 810 days of hydration (a); samples after 180 days of hydration: well developed thaumasite crystals (b), clusters of tiny hexagonal portlandite (CH) crystals and areas with amorphous CSHs (marked with circles), corroded gypsum (Gp) crystal and calcite (Cal) formed by portlandite hydration pr The appearance of thaumasite crystals after 180 and 810 days of hydration is shown in Fig 5 a. Some of the thaumasite crystals are almost perfectly shaped (Fig. 5 b). Portlandite crystals form rounded clusters, while amorphous CSHs have typical bowl-like shapes (Fig. 5 c). Corrosion of the majority of gypsum crystals, and local carbonation of portlandite resulting into calcite formation, can also be observed (Fig. 5 d). e- 4. Discussion Pr The hydration process starts by the conversions of lime to portlandite, and anhydrite to gypsum. Given by the high pH values of the system, the anhydrite hydration proceeds rapidly (Freyer et al. 2003) [23]. On the other hand, hydration of the other sulphosilicate clinker phases, ternesite and belite, proceeds markedly slower in the first stages. The results showed that the belite content decreases intensively between 28 and 56 days, which corresponds to the previously published data [24]. The belite hydration rate then decreases between 56 and 180 days, after which its content is eliminated or remains unchanged until the end of the experiment. na l The acceleration of ternesite hydration correlates with the increase in the amorphous phase content between 28 and 180 days from the beginning of the hydration experiment. Jo ur Based on the changes of the phase contents, the following statements can be made: ternesite and belite decompose to form CSH gels, and sulphate ions from ternesite become part of gypsum. It may be assumed that gypsum produced by the hydration of both, anhydrite and ternesite, react with silica and carbon dioxide to form thaumasite. In the experiments, the CO2, dissolving in water, was regularly replenished from the atmosphere above the solution. Analogous system of belite and ye´elimite was examined by El- Didamony et al. (2014) [25]. They revealed that β-C2S - C4A3S̅ inhibits the formation of crystalline calcium sulphoaluminate hydrates (AFt or AFm). Calcium aluminosilicate hydrate is formed by the reaction of β-C2S or CSH with the AH3 liberated from hydrated C4A3S̅. Similar process can be expected to proceed during ternesite hydration, during which portlandite instead of aluminum hydroxide is liberated and CSHs are formed. By the subsequent reaction of portlandite with gypsum and CO2, thaumasite is formed. After 180 days, the rate of thaumasite formation decreases or even stops, while gypsum remains present in the system. This phenomenon can be explained by the depletion of silica, except for 1h sample, the clinker composition in which respected the stoichiometry of thaumasite. Termination of thaumasite crystallization because of pH being below 9.5 [13] can be excluded, since in all the five cases, the pH values were between 11.9 and 12.2 even after 810 days of hydration. The content of carbonates was low during all the experiments; only the presence of calcite was detected. Although portlandite and CO2 were present, calcite content remained low due to high pH values in the batches. Calcite precipitates up to the pH value of 10 [26]. oo f A satisfactory correlation of the results of thaumasite quantification performed by two independent methods, TA and XRD, was acquired despite the fact that both the methods feature certain limitations. The identification of the beginning of the thaumasite decarbonation effect based on dTG curve is rather subjective. On the other hand, the XRD results are, among others, affected by peak overlaps and the data on preferred orientation of gypsum and portlandite, which is complicated to avoid at reflection geometry. However, the knowledge of the development of relative contents of thaumasite are more important than the exact percentages. Mirvald et al. (2019) [27] published a review on thaumasite detection using the DSC method based on dehydration. Quantification of thaumasite in the presence of gypsum is problematic in this way, because the first thaumasite dehydration effect partly overlaps with dehydration of gypsum. This procedure may successfully be used for qualitative detection of thaumasite only. Therefore, weight loss due to decarbonation was used for quantification of thaumasite in this research. pr SEM/EDS analysis showed no observable change of thaumasite crystal size and shape between 180 and 810 days of hydration (Fig 5 a). This fact corresponds to the conclusion of quantitative analysis by XRD, which proved no significant change of thaumasite content in this part of the experiment. Conclusions    Jo ur   e-  Pr  Lime hydrates immediately and the resulting high pH contributes to the conversion of anhydrite to gypsum. The content of other sulphosilicate phases, ternesite and belite, decreases intensively between 28 and 56 days and slows down between 56 and 180 days of hydration. Acceleration of ternesite hydration correlates with the increase in the amorphous phase content between 28 and 180 days. Ternesite and belite preferably produce CSH gels and sulphate ions from ternesite become a part of gypsum. Gypsum produced by the hydration of anhydrite and ternesite reacts with silica and carbon dioxide to form thaumasite. DSC or DTA is a useful tool for thaumasite detection, but not quantification. The quantification of thaumasite by weight loss due to its decarbonation in combination with XRD data and Rietveld refinement seems to be a promising way. Thaumasite can be formed as one of the hydration products of sulphosilicate clinkers. Besides indisputable ecological benefits of the application of sulphoaluminate cements, the possible risk of sulphate attack should also be considered. Furthermore, not only the cement composition, but also the environmental condition to which the construction is exposed play an important role. na l  Declaration of interests The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper. Acknowledgement: This work was financially supported by project number: GA17-24954S “The Conditions of Thermodynamic Stability and Transformation of AFt Phases”. Thanks for the language corrections are to Lenka Kunčická. na l ur Jo f oo pr e- Pr References 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. Jo 20. f 8. oo 7. pr 6. e- 5. Pr 4. na l 3. Ludwig, H.M., Zhang, W. Research review of cement clinker chemistry. Cem Concr Res 2015, 78, pp. 24-37. Horr, Y.A., Elhoweris, A.,Esam Elsarrag, E. The development of a novel process for the production of calcium sulfoaluminate. Int. J. Sustain. Built Environ. 2017, 6, pp 734-741. Montes, M., Pato, E., Carmona-Quiroga, P.M., Blanco-Varela M.T. Can calcium aluminates activate ternesite hydration? Cem Concr Res. 2018, 103, pp 204-215. Patent WO2013023729, Verfahren zu Herstellung von Ternesit/method for producing ternesite. Applicant HeidelbergCement AG 2013 Shen, Y., Qian, J., Huang, Y., Yang, D. Synthesis of belite sulfoaluminate-ternesite cements with phosphogypsum. Cem Concr Compos. 2015, 63, pp 67-75. Mcphee, D., Diamond, S., Thaumasite in cementitious materials. Cem Concr Compos. 2003, 25, 8, pp 805807. Norman, R.I., Dann, S.E., Hogg, S.C., Kirk, C.A., Synthesis and structural characterisation of new ettringite and thamasite type phases, Solid State Sci. 2013 25, pp 110-117. Diamond, S. Thaumasite in Orange County, Southern California: an inquiry into the effect of low temperature. Cem Concr Compos 2003, 25, 8, pp 1161-1164. Bensted, J. Thaumasite - background and nature in deterioration of cements, mortars and concretes. Cem Concr Compos. 1999, 21, 2, pp 117-121. Aguilera, J., Varela, M.T.B., Vázquez, T. Procedure of synthesis of thaumasite. Cem Concr Res. 2001, 31, pp 1163-1168. Pipilikaki, P., Papageorgiou, D., Teas, Ch., Chaniotakis, E., Katsioti, M. The effect of temperature on thaumasite formation. Cem Concr Compos. 2008, 30, 10, pp 964-969. Jallad, K. N., Santhanam, M., Cohen, M. D., Stability and reactivity of thaumasite at different pH levels, Cem Concr Res. 2003, 33, 433–437, doi: 10.1016/S0008-8846(02)00971-7 Yang, Ch., Liu, B., Xiang, X., Zhang, J. The influence of pH on the formation and stability of thaumasite. Appl. Mech. Mater. 2012, 174-177, pp 268-274 Erlin, B., Stark, D.C. Identification and occurence of thaumasite in concrete a discussion for the symposium on aggresive fluids. Highway research records 1965, 113, pp 108-113. Menéndez, E., Aldea, B., Formoso, M., García-Roves, R., Ruiz, S., de Frutos J., Influence of Temperature and Aggressive Solutions in the Formation of Thaumasite and Ettringite in Standard and Commercial Mortars, Sulphate Attack – Field Aspects and Lab Tests. RILEM Bookseries 2020, pp 55-70 doi:10.1007/978-3030-20331-3 Masárová, A. Verification of Synthetic Preparation of Thaumasite. Materials Science Forum 2016, 851, pp 80–85. Kollman, H., Strübel, G., Trost, F., Mineralsynthetische Untersuchungen zu Treibursachen durch Ca-AlSulfat-Hydrat und Ca-Si-Carbonat-Sulfat-Hydrat. Tonindustrie-Zeitung 1977, 101, pp 63-70. Crammond, N., Thaumasite in failed cement mortars and renders from exposed brickwork. Cem Concr Res. 1985,15, 6, pp 1039-1050. Bensted, J., Thaumasite a deterioration product of hardened cement structures. Il Cemento 1988, 85, pp 310. Crammond, N., Nixon, P.J. Deterioration of concrete foundation piles as a result of thaumasite formation. Sixth International Conference on Durability of Building Materials and Components, Omiya, Japan, 1993, pp. 1-33. Purnell, P., Francis, O.J., Page, C.L. Formation of thaumasite in synthetic cement mineral slurries. Cem Concr Compos. 2003, 25, 8, pp 857-860. Skalamprinos, S., Jen, G., Galan, I., Whittaker, M., Elhoweris, A., Glasser, F. The synthesis and hydration of ternesite, Ca5(SiO4)2 SO4. Cem Concr Res. 2018,113, pp 27-40 doi:10.1016/j.cemconres.2018.06.012 Freyer, D.,Voigt, W., Crystalization and phase stability of CaSO4 and CaSO4 salts. Monatshefte für Chemie 2003, 134, pp 693-719. Dienemann, W., Schmitt, D., Bullerjahn, F., Ben Haha, M. Belite-Calciumsulfoaluminate-Ternesite (BCT) A new low-carbon clinker Technology. Cem Int. 2013 11, 4, pp 3-10 El-Didamony, H., Heikal, M., El-sokkary, T.M., Khalil, Kh. A., Ahmed, I. Active belite β-C2S and hydration of calcium sulphoaluminates prepared from nano-materials. Ceram. -Silik. 2012 58, 2, pp 165171 ur 1. 2. 21. 22. 23. 24. 25. 26. Jo ur na l Pr e- pr oo f 27. Ruiz-Agudo, E., Putnis, C. V., Rodriguez-Navarro, C., & Putnis, A. Effect of pH on calcite growth at constant ratio and supersaturation. Geochim. Cosmochim. Acta 2011 75, 1, 284–296. doi:10.1016/j.gca.2010.09.034 Mirvalad, S., Nokken, M., & Banu, D., Detection of Thaumasite Formation Using Differential Scanning Calorimetry. J Mater Civil Eng, 2019 31, 9, pp 04019178-1-12. doi:10.1061/(asce)mt.1943-5533.0002820 Appendix A: Phase composition development quantified using XRD data Table A1. Sample 1h (CaO : SiO2 : SO3 molar ratio = 3 : 1 : 1) Thau Gyp Portlan Calcite Ter Larnite Days dite masite sum nesite Anhyd rite Lime Amor phous 0.0 0.0 0.0 0.0 47.1 22.0 28.9 2.0 0.0 28 8.4 12.1 0.0 0.0 42.2 10.4 21.6 0.0 5.3 56 13.8 42.0 1.0 0.0 22.9 2.4 0.4 0.0 17.5 180 15.1 46.2 2.9 1.8 2.8 1.3 0.6 0.0 29.3 810 34.2 31.8 2.6 5.6 0.0 0.0 0.0 0.0 25.8 Anhyd rite Lime Amor phous 22.0 Table A2. Sample 2h (CaO : SiO2 : SO3 molar ratio = 5 : 1 : 1) Thau Gyp Portlan Calcite Ter Larnite Days dite masite sum nesite f 0 0.0 0.0 0.0 0.0 51.1 6.5 20.5 28 2.6 2.1 27.7 0.0 33.9 8.8 12.8 0.0 12.1 56 13.0 24.4 13.7 0.0 16.2 3.9 0.3 0.0 28.5 180 14.0 24.5 22.2 1.8 0.0 0.0 0.0 0.0 37.5 810 14.7 31.9 22.8 0.3 0.0 0.0 0.0 0.0 pr 30.3 Anhyd rite Amor phous 14.0 5.8 16.6 0.0 0.0 28 4.0 0.1 15.3 56 12.9 3.9 12.5 180 19.2 8.3 24.3 810 24.0 11.1 22.9 0.0 54.4 17.1 3.4 0.0 5.7 0.0 40.5 10.0 0.4 0.0 19.8 1.5 0.0 0.0 0.0 0.0 46.7 1.2 0.0 0.0 0.0 0.0 40.8 Anhyd rite Lime Amor phous 0.0 63.6 Pr 0.0 Lime 0 0.0 na l 0.0 e- Table A3. Sample 3h (CaO : SiO2 : SO3 molar ratio = 7 : 2 : 1) Thau Gyp Portlan Calcite Ter Larnite Days masite sum nesite dite 0 0.0 oo 0 0.0 0.0 0.0 59.4 23.9 7.0 9.7 0.0 28 1.9 0.9 6.2 0.0 61.4 22.1 3.7 0.0 3.8 56 10.3 6.7 8.5 0.0 57.9 5.4 0.0 0.0 11.2 180 15.6 11.8 11.4 1.3 6.9 4.1 0.0 0.0 48.9 810 18.4 11.5 13.1 2.5 0.0 3.5 0.0 0.0 51.0 Anhyd rite Lime Amor phous Jo ur Table A4. Sample 4h (CaO : SiO2 : SO3 molar ratio = 6 : 2 : 1) Thau Gyp Portlan Calcite Ter Larnite Days masite sum nesite dite Table A5. Sample 5h (CaO : SiO2 : SO3 molar ratio = 5 : 1 : 1) Thau Gyp Portlan Calcite Ter Larnite Days dite masite sum nesite 0 0.0 0.0 0.0 0.0 44.6 37.1 16.4 1.9 0.0 28 2.8 3.6 0.0 0.0 39.8 35.3 10.6 0.0 7.9 56 7.8 16.4 0.9 0.0 33.3 19.9 0.0 0.0 21.7 180 10.4 25.0 2.8 1.6 0.0 0.5 0.0 0.0 59.7 810 11.7 23.8 4.4 1.5 0.0 1.9 0.0 0.0 56.7 Jo ur na l Pr e- pr oo f Appendix B: Examples of X-ray microanalysis EDS spectra