1 Introduction

Tricalcium aluminate (C3A) is one of the primary clinker phases in ordinary Portland cement (OPC), and typically presents in amounts of 2–12 wt%. The crystal structure of cubic C3A (c-C3A or Ca3Al2O6) can accommodate the Na+ element, resulting in an orthorhombic C3A (o-C3A or NaxCa3-xAl2O6) polymorph [1,2,3]. This modification of c-C3A’s mineralogy can be also induced from a mechanism in which Ca2+ is partially replaced by Na+. The clinker produced during the calcining process usually contains c-C3A, but the phase can change to o-C3A depending not only on the raw material used but also on the storage environment of OPC [2, 3]. With these intrinsic properties of C3A, the hydration properties of C3A also differ considerably depending on the type of polymorph [4,5,6].

The reactivity of both polymorphs of C3A is significant within clinker, and it has been reported to have an impact on the initial hydration reaction of OPC [3, 7]. When water and C3A come into contact, a substantial amount of hydration heat is generated, and significant quantities of hydration products (such as gibbsite (AH3), Al-hydrogarnet (C3AH6), and OH-AFm (C4AHx) phases) are produced [6, 8]. Due to its high reactivity, it has been reported that C3A can actively react with water under ambient conditions, which is known as the pre-hydration of OPC. Pre-hydration can affect the initial hydration properties of OPC, including its setting behavior [9,10,11]. Therefore, inhibiting the effects of C3A pre-hydration is quite important factor, as it can potentially alter the mechanical and hydration behaviors of OPC.

A grinding agent (GA) is typically added to OPC before the grinding stage to improve its grinding performance [12,13,14,15]. Moreover, the application of GA significantly changes the hydration behavior of clinker phases [16,17,18,19]. Alkanolamine-based GAs, such as triethanolamine (TEA), diethanol isopropanolamine (DEIPA), and Triisopropanolamine (TIPA), are widely used in the cement production. When TEA is used, the aluminate reactions (C3A and ferrite [C4AF]) are improved, suggesting that even a small amount of TEA can cause remarkable variations in the hydration mechanism of aluminate phases [16]. Similarly, DEIPA accelerates the aluminate reactions in the early stage, and it has been concluded that the complex hydration mechanism can affect the mechanical properties of concrete [19]. TIPA enhances the aluminate reactions, and it has been reported that the C3A reaction is significantly improved with TIPA [20]. Based on this research, it is worth studying the variation of hydration reaction of C3A according to different alkanolamine-based GAs in the base of different polymorphs of C3A.

As mentioned above, it has been found that alkanolamine-based GAs can change the hydration characteristics of the C3A phase [19]. However, the effects of alkanolamine-based GAs on the surface of C3A and their impact on the hydration behavior of C3A phases have not fully understood. Therefore, the aim of this study is to investigate the effects of two dosages (0.1% and 0.3%) of TIPA on the crystallographic variations and surface properties of C3A, as well as the modification of the hydration and mechanical properties of C3A caused by these effects. The investigation was carried out on the two different crystal structures of cubic and orthorhombic using X-ray diffraction (XRD) and X-ray photoelectron spectroscopy (XPS) to analyze the crystallographic variations and surface properties of C3A induced by TIPA, respectively. In addition, inductively coupled plasma optical emission spectroscopy (ICP OES), isothermal calorimetry, thermogravimetric analysis (TGA), and quantitative XRD (QXRD) were used to investigate the hydration properties along with compressive strength measurement.

2 Materials and experimental methods

2.1 Sample preparation

In this study, samples of c-C3A and o-C3A (Construction Technology Laboratories, Inc., Skokie, IL) were used. Pure TIPA was obtained from a Korean concrete admixture company (Silkroad, Seoul, Korea). The C3A nodules were obtained by calcining at 1,300° for 2 h, and coarse C3A powders were obtained by hand grinding for 5 min, followed by sieving under 500 µm. The QXRD results of c-C3A and o-C3A are provided in Table 1. The c-C3A and o-C3A series were distinguished by adding TIPA contents (0%, 0.1%, and 0.3%) before the grinding program (Table 2). The water-to-C3A ratio for the paste was set to 1.0, with the water content adequately applied to the C3A content.

Table 1 QXRD results of unhydrated c-C3A and o-C3A
Full size table
Table 2 Mix design for C3A powder (before grinding program)
Full size table

2.2 Experimental method

In this study, lab-scale micro ball mill equipment (McCrone Micronizing Mill; McCrone Scientific Ltd., London, UK) was used to grind the C3A powder. The grinding conditions of C3A (10 g) powder and the amount of TIPA (0%, 0.1%, and 0.3%) were determined based on several preliminary experiments. The grinding time and rotation speed were set to 14 min and 1,166 rpm, respectively. The particle size distribution was measured using a microparticle size analyzer (Malvern Instruments Ltd., Malvern, UK) with isopropyl alcohol, which does not react with C3A, as the dispersion medium [21]. The effect of TIPA on the grinding degree of C3A was evaluated based on these results.

In this study, an AXIS Supra + spectrometer for chemical analysis (Kratos Analytical Limited, Manchester, UK) was used to measure the XPS spectra. The measuring range was set from 0 to 1,200 eV, with data collected at one point per eV [22]. The sample pre-processing method was as follows: first, the powders were dehydrated for 24 h in a vacuum environment at 20 °C Then, the dried powders were molded into a 7 mm diameter pressed pellet and stored in a vacuum environment before the experiment started. XPSPEAK 4.1 software was used to analyze the XPS spectra, with a Shirley type background adopted due to the complexity of XPS spectra. The Gaussian–Lorentzian ratio function (3:2) was used to fit the peak position, full width at half maximum (FWHM), and area factors [23].

Five identical 10 \(\times\) 10 \(\times\) 10 \({{\text{mm}}}^{3}\) paste specimens, where the water-to-C3A ratio was 1:1, were chosen to determine the compressive strength of the specimens cured for 1 day and 3 days. Based on several preliminary tests with identical sizes of OPC specimens, a loading rate of 0.05 mm/min was carefully assigned.

ICP OES analysis was conducted using an Agilent 5100 SVDV ICP OES device (Agilent Technologies, Santa Clara, CA, USA). The water to C3A ratio was set to 10:1 by weight. After 30 min of stirring, the solutions were filtered through a 200 nm filter. Prior to the experiment, the plasmid argon gas was subjected to an induction magnetic field, and the preprocessed liquid was sprayed onto the particles and injected [24].

To investigate the exothermic reaction between C3A, TIPA, and water, a micro-calorimeter (TAM Air 8-channel, TA Instruments, New Castle, DE, USA) was used. The water-to-C3A ratio was also set to 1:1 by weight. The device was calibrated before the experiment by holding it at a constant 20 °C for at least 24 h. After mixing 6 g of water and 6 g of C3A powder for 1 min, the 5 g of resulting slurry was injected into the ampoule. The ampoule was then placed into the device. The heat flow and cumulative heat data were normalized to the amount of C3A powder and water contained in the ampoule.

A D2 Phaser X-ray diffractometer (Bruker Co. Ltd., Land Baden-Württemberg, Germany) was used to measure the XRD patterns, with Cu Kα radiation (λ = 1.5418 Å) as the source. The tube current and generator voltage were set at 10 mA and 30 kV, respectively [25, 26]. For the samples that were hydrated for 1 day and 3 days, the hydration stop method was used. The detailed process is as follows: The paste samples that had cured for the specified periods (i.e., 1 day and 3 days) were ground for 10 min to attain fine powders. Isopropyl alcohol and diethyl ether were then used as exchange solutions to remove free water that was still present in the powders [27, 28].

The XRD patterns obtained were analyzed using TOPAS software version 7.0 (Bruker Co. Ltd., Land Baden-Württemberg, Germany). The background of the XRD patterns was corrected using the Chebyshev polynomial and 1/X terms equipped in TOPAS software. The scale, unit cell, and FWHM factors were the profiled fitting parameters [29, 30]. The internal standard method was used to analyze the hydration properties of C3A. Al2O3 (NIST SRM 676a) was selected as the internal standard material because it does not overlap with the other crystals. C3A powder and an internal standard material were mixed in a 1:9 ratio by weight, and the mixture was mixed for 30 min [28, 31, 32]. The Rietveld refinement method was then used to perform all of the quantitative evaluations for all crystalline phases. The quantified mineral phases were corrected using the amount of internal standard material added, which is shown in Eq. (1). The amount of amorphous phase was calculated using the actually added amount of internal standard material. Finally, the quantitative analysis results of all phases were calibrated using the value of chemically bound water (CBW), which will be mentioned later (Eq. (2)) [27].

$${\text{Corr}}\left({w}_{\alpha }\right)= {w}_{\alpha }\frac{{{\text{STD}}}_{{\text{known}}}}{{{\text{STD}}}_{{\text{measured}}}}$$
(1)
$${w}_{{\text{amorphous}}}=1- \sum_{j=1}^{n}{\text{Corr}}\left({w}_{j}\right)$$
(2)

where \({\text{Corr}}\left({{\text{w}}}_{\mathrm{\alpha }}\right)\), \({{\text{w}}}_{\mathrm{\alpha }}\), \({{\text{STD}}}_{{\text{known}}}\), and \({{\text{STD}}}_{{\text{measured}}}\) indicate the corrected weight percentage of \(\mathrm{\alpha }\), weight percentage of \(\mathrm{\alpha }\), analyzed weight percentage of internal standard material, and analyzed weight percentage of internal standard material, respectively.

To obtain TGA data, a differential scanning calorimeter (DSC)-TG system (SDT Q600, TA Instruments Ltd., Newcastle, DE, USA) was used. The temperature range was from 30 °C to 1,000 °C, and the heating rate was 10 ℃ per minute [27]. In addition, the hardened paste samples underwent the hydration stop procedure at 1 day and 3 days, and 25–30 mg of powder was measured to analyze the hydration properties of each sample.

3 Results

3.1 Particle size distribution

Variations in the particle size distribution of the c-C3A and o-C3A series are presented in Fig. 1a, b. It is difficult to identify a clear trend with the c-C3A series. However, a noticeable particle size reduction was confirmed with the o-C3A series. When 0.1% of TIPA was added, a new first mode was detected, and the second mode was shifted to the left position. In addition, a weakened effect was exhibited with 0.3% TIPA compared to the sample with 0.1% TIPA. The results obtained are consistent with those reported in a previous study where a large amount of GA was added [33].

Fig. 1
figure 1

Particle size distributions of a c-C3A and b o-C3A series

Full size image

3.2 XPS results

Figure 2a, b shows the XPS results obtained from both C3A polymorphs after the grinding process. Noticeable variations were observed according to the amount of TIPA added. In particular, variations in peak properties were observed in the scan ranges of ~ 72.5 eV and 290.0 eV, which correspond to the states of Al and C atoms, respectively.

Fig. 2
figure 2

XPS results of a c-C3A and b o-C3A series

Full size image

Similar to previous studies, the Al 2p spectra of c-C3A_0% and o-C3A_0% showed two contributions, with the spectra located at 72.93 and 74.31 eV for c-C3A_0% and at 73.34 and 73.94 eV for o-C3A_0% [9, 34]. This difference indicates that the Al binding energies are closely related to the coordination number [35]. In the case of o-C3A_0%, the replacement of Ca2+ by Na+ occurred, resulting in a difference in the binding energies of Al 2p. Meanwhile, an interesting trend was observed with the TIPA-added samples: the position of the Al 2p spectra shifted to a lower energy when the amount of TIPA added was increased. The binding energies of Ca 2p (346.48 to 346.93 eV) were almost identical and in good agreement with previous studies [9, 34]. With the Ca 2p spectra, it is difficult to observe a noticeable trend because the spectra are less susceptible to compositional variations (e.g., calcium aluminate hydrate and calcium silicate hydrate) [34, 36]. Detailed explanations will be provided later.

3.3 Compressive strength results

The compressive strength results of all C3A pastes are presented in Fig. 3a, b. Overall, the compressive strengths of the c-C3A series are higher than those of the o-C3A series. When TIPA was added, the compressive strength of the samples cured for 1 day increased. In particular, a significant effect was observed in the samples with 0.1% TIPA. However, in the samples cured for 3 days, the strengths were rather decreased compared to those cured for 1 day. Furthermore, depending on the type of C3A, differences in the development of strength at 3 days were confirmed: in the case of the c-C3A series, enhanced strength was obtained with TIPA application, but an inverse relationship between the amount of TIPA added and the compressive strength was confirmed with the o-C3A series. These trends could be directly related to the modified hydration behavior of C3A induced by TIPA.

Fig. 3
figure 3

Compressive strength results of a c-C3A and b o-C3A series

Full size image

3.4 ICP OES results

ICP OES results are presented in Fig. 4a, b. This analysis allowed for the evaluation of the reactivity degree of ions. The effect of TIPA on the ion elution degree of C3A was completely different depending on the type of C3A. In the case of the c-C3A series, when the amount of TIPA was increased, the elution amount of Al ions was almost similar, but the detected amount of Ca ions decreased. Meanwhile, with the o-C3A series, the elution of Al and Na ions remarkably increased with TIPA addition. In particular, it should be noted that the total elution content of ions is also different; the total elution content is lower than 500 ppm with the c-C3A series, but that of the o-C3A series is higher than 6,000 ppm. These findings are different from previous studies [4,5,6], and it could be confirmed that doped-Na is rapidly eluted from o-C3A.

Fig. 4
figure 4

ICP OES measurements of a c-C3A and b o-C3A series

Full size image

3.5 Calorimetric results

The impact of TIPA (at doses of 0, 0.1, and 0.3%) on C3A phases was investigated, and the results are shown in Fig. 5a–d. After water and c-C3A series were mixed, a substantial amount of heat was released, which improved with increasing amounts of TIPA (Fig. 5a). This result is consistent with the trend found in previous studies, which reported that the reactivity of C3A in the OPC system is enhanced by TIPA [37, 38]. In contrast, surprising results were observed with the o-C3A series upon TIPA application (Fig. 5c, d): an accelerated hydration reaction was observed until 10 h, followed by a remarkable heat absorption effect from 1 to 3 days. Consequently, the total released heat decreased compared to the o-C3A_0% sample. Therefore, to elucidate the aforementioned effects, it is necessary to clarify the variation in the hydration behavior of C3A that was observed during the measurement time.

Fig. 5
figure 5

Hydration heat evolution of the slurry: (a and b) c-C3A series and (c and d) o-C3A series

Full size image

3.6 TGA results

TG curves of C3A paste cured for 1 day and 3 days are presented in Fig. 6a, b. In the present study, weight loss below 180 °C indicated the decomposition of OH-AFm (C4AHX) [39], while weight loss between 250 and 280 °C could be attributed to the decomposition of gibbsite (AH3) [40,41,42,43]. Furthermore, the weight loss between 200 and 300 °C indicated the decomposition of C3AH6 [44, 45]. Based on the previous studies, the hydration mechanism of the hydrated samples can be estimated.

Fig. 6
figure 6

TG results of a c-C3A and b o-C3A series

Full size image

In the absence of TIPA, a large amount of C4AHX phases but a small amount of C3AH6 were observed in c-C3A (Fig. 6a). However, upon the addition of TIPA, a significant amount of C3AH6 was confirmed, while C4AHX was hardly detected. Furthermore, the amount of C3AH6 increased with longer curing durations, from 1 to 3 days. Since C3AH6 phase is a fairly thermodynamically stable mineral, the predominantly produced amount of C3AH6 phase indicates that C3A hydration is almost complete [46]. In summary, these results indicate that TIPA acts as an accelerator in the c-C3A system.

A different trend was observed with the o-C3A series. In the o-C3A_0% sample, substantial amounts of both C3AH6 and C4AHX phases were observed regardless of curing duration. When TIPA was added to o-C3A, an interesting result was observed: a significant amount of C4AHX phases was confirmed in the samples cured for 1 day, but these phases were hardly detected in the samples cured for 3 days (Fig. 6b). Instead, a remarkable amount of C3AH6 was obtained, indicating that the phase transitions from C4AHX to C3AH6 might occur. Although C3AH6 phase is stable in an environment of 20 °C, it is commonly reported that C4AHX phases are predominantly produced in short curing durations [6, 47, 48]. Therefore, these findings are not typical trends. Since it is challenging to determine this only with TGA, more detailed explanations will be provided with QXRD results.

3.7 XRD results

The XRD patterns of unhydrated C3A are presented in Fig. 7a, b). No significant changes in the XRD pattern were observed in the c-C3A series upon the addition of TIPA, while substantial modifications were observed in the o-C3A series. Specifically, the peaks located at 33.0° and 33.3° for the o-C3A series decreased with increasing amounts of TIPA. It is likely that the crystal structure of the minerals was altered due to the combination of grinding program [49]. Further details will be discussed in Sect. 4.2.

Fig. 7
figure 7

X-ray diffraction patterns of the investigated C3A according to the amount of TIPA added: a c-C3A and b o-C3A series

Full size image

The XRD patterns of the hydrated C3A are presented in Fig. 8a–f. For the c-C3A_0% sample, a significant amount of C4AHX phases was observed. Specifically, the C4AH11 and C4AH13 phases located at 11.8° and 10.5°, respectively, were detected in the sample cured for 1 day (Fig. 8a). After two days, the C4AH13 phase was not observed, but a noticeable amount of C4AH19 phase located at 8.5° was detected. This suggests the rehydration effect of the C4AH13 phase [50]. Meanwhile, upon the addition of TIPA, the hydration mechanism was completely altered, with almost no C4AHX phases detected and a significant amount of C3AH6 produced instead (Fig. 8c, e).

Fig. 8
figure 8

XRD patterns of all investigated C3A (cured for 1 day and 3 days): a c-C3A_0%, b o-C3A_0%, c c-C3A_0.1%, d o-C3A_0.1%, e c-C3A_0.3%, and f o-C3A_0.3%

Full size image

In the case of o-C3A_0%, significant amounts of C4AHX and C3AH6 phases were produced. Similar to the c-C3A_0% sample, the C4AH11 and C4AH13 phases were also observed in the sample cured for 1 day (Fig. 8b). However, with o-C3A_0% cured for 3 days, it was clearly detected that not only C4AH19 and C4AH11 phases but also the C4AH13 phase were present. Moreover, interesting results were obtained when TIPA was added. When 0.1% TIPA was added to o-C3A, the C4AH11 and C4AH19 phases were confirmed in the sample cured for 1 day (Fig. 8d). When the o-C3A_0.3% sample was cured for 1 day, it was detected that the C4AH13 and C4AH11 phases were present (Fig. 8f). After two days, the C4AH19 and C4AH13 phases were completely decomposed (Fig. 8d, f). It should be noted that there was a difference between the samples cured for 1 day and 3 days; as the curing duration increased from 1 to 3 days, the C4AH19 and C4AH13 phases might be transformed to C3AH6 phase, which could be related to the heat absorption effect observed in Sect. 3.5. These findings will be systemically verified with QXRD and thermodynamic database.

4 Discussion

4.1 Effect of TIPA on surface compositions of unhydrated C3A

In this section, the effects of TIPA on C3A phases are discussed in relation to carbonation and pre-hydration effects, which are directly related to C3A reactivity. Variations in the XPS spectra of C 1s were clearly observed and are shown in Fig. 9a, b. As the amount of TIPA added increased, the C 1s spectra decreased, but the Ca 2p spectra showed an anomalous trend Fig. 9c, d. Previous studies have reported that the Ca 2p spectra are less sensitive to changes in composition [34, 36]. As reported in the previous study, the increase in the C 1s peak indicates the formation of monocarboaluminate through the reaction of CO2 and C4AHx in the atmosphere [9]. Moreover, in XPS analysis, given that the measurement encompasses the surface of the C3A particle (approximately 1 to 10 nm thick) [51], this result suggests that the reaction occurs at the particle surface, and it can be inferred by the addition of TIPA. Therefore, to analyze the carbonation effect quantitatively, it is necessary to compare the relative ratios of the Ca 2p and C 1s spectra.

Fig. 9
figure 9

XPS spectra of C and Ca of unhydrated (a and c) c-C3A and (b and d) o-C3A series

Full size image

To investigate the carbonation effect of TIPA on the unhydrated C3A surface, the relative area ratios of the C 1s and Ca 2p spectra were calculated as presented in Fig. 10a, b) and Table 3 and 4. The C/Ca ratios of C3A phases without TIPA were calculated to be approximately 0.17 (c-C3A_0%) and 0.37 (o-C3A_0%), which is equivalent to being exposed to moist air for approximately 12 h [9]. The calculated values indicate that surface carbonation could be accelerated during the grinding process. On the other hand, it should be noted that the effect was mitigated by the addition of TIPA. This phenomenon can be attributed to the uniform coating of the C3A surface by TIPA during the grinding process [15]. In particular, this effect was predominantly observed in the c-C3A series. This is considered to be because the initial reactivity of c-C3A is relatively lower than that of o-C3A. Nevertheless, based on the results, it can be concluded that TIPA is effective in inhibiting surface carbonation regardless of the type of C3A phase.

Fig. 10
figure 10

Normalized ratios of C and Ca of a c-C3A and b o-C3A induced by TIPA

Full size image
Table 3 Fitting results of XPS spectra of c-C3A series
Full size table
Table 4 Fitting results of XPS spectra of o-C3A series
Full size table

The binding energies of Al 2p can be used to evaluate the degree of pre-hydration of the C3A surface. It has been reported that high binding energies in the Al 2p spectra indicate hydrated C-A-H phases [9]. For example, binding energy of 73.8 eV has been reported for C4AH13, while a range of values between 74.3 and 74.5 eV has been reported for C3AH6 [52], which was also observed in the present study. Overall, the addition of TIPA led to decreases in binding energies for both polymorphs (Table 3 and 4, and Fig. 11a, b). Based on these findings, it can be suggested that although pre-hydration of the C3A surface was confirmed in all samples, TIPA reduced the degree of pre-hydration to some extent.

Fig. 11
figure 11

XPS spectra of Al of a c-C3A and b o-C3A series

Full size image

4.2 Modifications of crystallographic information of unhydrated C3A induced by TIPA

Partial modifications of the XRD pattern were observed in the unhydrated c-C3A and o-C3A due to the addition of TIPA, as shown in Fig. 12a, b. Specifically, in the c-C3A series, the peak located at 33.2° slightly increased as the amount of TIPA added increased. However, the opposite trend was observed in the o-C3A series: the two peaks located at 33.0° and 33.3° decreased as the amount of TIPA added increased. It is well-known fact that variations in crystallographic information are typically observed with grinding programs [53,54,55]. Therefore, the variations of the XRD patterns induced by the grinding program and TIPA were regarded as partial modifications of the crystal structure.

Fig. 12
figure 12

XRD patterns of the unhydrated a c-C3A and b o-C3A series

Full size image

As presented in Table 5, modifications in the crystal structure of unhydrated C3A phases were confirmed with TIPA application. Interestingly, completely opposite trends were observed between the c-C3A and o-C3A series. With the increase in the amount of TIPA, the lattice parameters of c-C3A increased, resulting in a massive unit cell volume. Meanwhile, in the case of o-C3A series, the unit cell parameters decreased with the addition of TIPA, and the unit cell volume also decreased. Previous studies have suggested that these changes may affect the inherent characteristics of the mineral [56, 57]. Therefore, it could be suggested that the hydration properties of C3A phases changed with the aforementioned effect.

Table 5 Profile fitting results of unhydrated c-C3A and o-C3A series
Full size table

4.3 Variations of hydration behavior of C3A due to TIPA application

Table 6 and Fig. 13a, b show the quantitative TG and XRD results of all C3A samples (cured for 1 day and 3 days). The o-C3A_0% series exhibited a higher degree of hydration compared to the c-C3A_0% series, and a completely different hydration mechanism was also observed. In the c-C3A_0% series, significant amounts of C4AHX phases were quantified, while in the o-C3A_0% series, both C4AHX and C3AH6 were detected. This could be due to the presence of Na+ ions in o-C3A, which can affect the hydration behavior by increasing the pH in the paste matrix [9].

Table 6 Weight loss and CBW values of C3A cured for 1 day and 3 days
Full size table
Fig. 13
figure 13

QXRD results of the a c-C3A and b o-C3A series cured for 1 day and 3 days

Full size image

Surprisingly, remarkable variations in hydration properties were observed with the addition of TIPA. As mentioned above, significant amounts of amorphous phases (mostly considered to be C4AHX phases) were quantified in the c-C3A_0% sample. However, with the addition of TIPA (i.e., c-C3A_0.1% and c-C3A_0.3%), a remarkable amount of C3AH6 was quantified. This result is in contrast to what was found in previous studies [58, 59]. In addition, similar modifications of the hydration behavior of o-C3A were also observed with TIPA application, with C3AH6 phase being predominantly produced. However, a clear phase transition from C4AHX to C3AH6 was observed during the curing period from 1 to 3 days, along with consumption of AH3 (Fig. 14a, b). From these observations, it can be suggested that C4AHX and AH3 react to produce a significant amount of C3AH6. Therefore, it is suggested that the hydration behavior of o-C3A with TIPA during this particular curing period (1–3 days) can be described by Eq. (3–5). The CEMDATA used for enthalpy calculations is as follows: Gibbsite was used for AH3, and liquid water was used for H.

Fig. 14
figure 14

Phases transition effects of o-C3A from 1 to 3 days according to TIPA application: a o-C3A_0.1% and b o-C3A_0.3%

Full size image

Among the thermodynamic analysis results, the positive enthalpy change observed in Eq. (4 and 5) has been verified, indicating an endothermic reaction that absorbs heat from the surroundings [60]. Although only a trivial quantity of C4AH19 was detected in the o-C3A_0.3% 1 day sample, an enhanced heat absorption effect was observed compared to the o-C3A_0.1% sample. However, it might be reasonable to observe more endothermic reactions in o-C3A_0.3% due to its significant content of C4AH13. Furthermore, this could be attributed to the presence of C4AH19 in a state of poorly crystallinity within the o-C3A_0.3% 1 day sample [61]. This may be attributed to the partial modifications of the crystallographic information of o-C3A induced by TIPA, alongside mitigated pre-hydration and/or carbonation effects. These findings hold significance as alkanolamine-based substances are present in nearly all OPC formulations.

$${\text{3C}}_{{4}} {\text{AH}}_{{{11}}} + {\text{ AH}}_{{3}} \to {\text{4C}}_{{3}} {\text{AH}}_{{6}} + {\text{ 12H}},\;\Delta {\text{H}}_{{{\text{rxn}}}} : \, - { 28},{544}.0{\text{7 J}}/{\text{mol}}$$
(3)
$${\text{3C}}_{{4}} {\text{AH}}_{{{13}}} + {\text{ AH}}_{{3}} \to {\text{4C}}_{{3}} {\text{AH}}_{{6}} + {\text{ 18H}},\;\Delta {\text{H}}_{{{\text{rxn}}}} : \, + { 72},{789}.{\text{55 J}}/{\text{mol}}$$
(4)
$${\text{3C}}_{{4}} {\text{AH}}_{{{19}}} + {\text{ AH}}_{{3}} \to {\text{4C}}_{{3}} {\text{AH}}_{{6}} + {\text{ 36H}},\;\Delta {\text{H}}_{{{\text{rxn}}}} : \, + { 191},{114}.{\text{5 J}}/{\text{mol}}$$
(5)

4.4 Alterations in the structural stability of C3A composite with TIPA

Figure 15a–d presents the relationship between the total amount of hydration products and compressive strength. It is a well-known fact that compressive strength and the total amount of hydration products are generally proportional. However, for the samples cured for 1 day, the strength improvement was higher in the c-C3A_0.3% sample, despite the total amount of hydration products being lower compared to the c-C3A_0% sample (Fig. 15a). This might be due to differences in the produced hydration products and their amounts, as well as the ratio among the hydration products (i.e., C3AH6, AH3, C4AHX phases) produced by the addition of TIPA. These results suggest that the structural stability of hydrated C-A-H phases might vary.

Fig. 15
figure 15

Relationship between total amounts of hydration products and compressive strength: a c-C3A cured for 1 day, b c-C3A cured for 3 days, c o-C3A cured for 1 day, and d o-C3A cured for 3 days

Full size image

However, it is difficult to identify a consistent trend between the compressive strength of samples cured for 1 day and 3 days. Although the total amount of hydration products significantly increased in all samples (from 1 to 3 days), the compressive strength decreased instead. In the case of c-C3A_0% and o-C3A_0%, the rehydration of C4AHX phases was observed, indicating that a phase transition from C4AH13 to C4AH19 occurred [8, 62]. Since the structural stability of C4AHX might decrease as the water content increases, this effect could weaken the mechanical properties of the samples cured for 3 days compared to those cured for 1 day. As previously discussed, phase transition effects were observed in all samples with TIPA during the curing duration from 1 to 3 days. When C4AHX phases were converted to C3AH6 phase, the breaking of initial bonds and transportation of the hydration products occurred [63, 64]. Therefore, the transformation of already formed hydration products during a short period of time could weaken the structural stability of C3A composites.

5 Conclusions

In this study, two C3A polymorphs were examined to elucidate the hydration characteristics of “TIPA-C3A systems”. The effects of TIPA on the surface composite and crystal structure of unhydrated C3A were investigated using XPS and XRD analyses, respectively. The hydration behavior of C3A modified due to TIPA was characterized using ICP OES, TGA, and QXRD techniques. In short, the heat absorption behavior in o-C3A was experimentally confirmed for the first time. Furthermore, its result was thermodynamically supported by the thermodynamic data available in databases like CEMDATA. To optimize various material performance of cement-based materials, tailoring nano-structure of cement mixture is critical which can have much higher impact on material properties compared to other relatively expensive methods (such as using chemical admixture). In this sense, understanding on the crystallographic changes and resulting hydration characteristics of cement minerals under the use of functional grinding agent is important for the development of suitable grinding agent for cement-based materials.

XPS analysis confirmed pre-hydration and carbonation effects in both C3A phases during the grinding process. However, these effects were prevented under the use of TIPA. Notably, these effects were more prominent in c-C3A than in o-C3A. Since C3A hydration is closely related to the early hydration properties of OPC, such as setting and hardening, these results can be extended to control the material behavior of OPC. Based on XRD analysis, partial modifications of the crystallographic information of C3A were observed with the addition of TIPA. This was directly interpreted by changes in the unit cell parameters of C3A, resulting in variations in the Al–O state.

Aforementioned effects led to significant modification of hydration behavior of C3A as compared to previously reported research. Despite the short curing period (3 days), the dominant production of C3AH6 instead of C4AHX phases was confirmed. In particular, the phase transition of C4AHX phases to C3AH6 was clearly observed in o-C3A, resulting in an interesting heat absorption effect which was well explained further by thermodynamic data available in databases like CEMDATA. With these analyses of hydration behavior, the variations in compressive strength were also explained.