Comparative study on different methods of activation of recycled powder grouts

RP was supplied by Yangzhou Huimin Renewable Resources Co., Ltd (Yangzhou, P. R. China). The SEM image and particle size distribution of RP are shown in Figures 1 and 2, respectively. The D₅₀ and specific surface area of RP are 29.01 μm and 378 m²/kg, respectively. The RP was placed in a muffle furnace and calcined at 700°C for 120 min, if necessary. P.I. 52.5 Portland cement, which was purchased from Taizhou Yangwan Conch Cement Co., Ltd. (Taizhou, P. R. China), was used in this article. The chemical compositions of RP and Portland cement are listed in Table 1. Analytical-grade sodium hydroxide (NaOH), sodium carbonate (Na₂CO₃), and CH were supplied by Sinopharm Chemical Reagent Co., Ltd (Shanghai, P. R. China).Commercial-grade DEIPA and TIPA were obtained from Sigma-Aldrich Chemical Reagent Co., Ltd (Saint Louis, USA), and the chemical structure and composition of DEIPA and TIPA are illustrated in Scheme 1. Polycarboxylic ether (PCE) superplasticizer (P540) was purchased from Shanghai Chenqi Chemical Technology Company (Shanghai, P. R. China). The defoaming agent (P80) was obtained from MÜNZING Chemie Corporation (Münzingstraße, Germany). The expanding agent (YH-S) was supplied by Tianjin Weihe Technology Admixture Company (Tianjin, P. R. China).

Figure 1

SEM image of RP.

Figure 2

Particle size distribution of RP.

Scheme 1

Chemical structure and composition of DEIPA and TIPA.

Both grout pastes were prepared with a water-to-binder (w/b) ratio of 0.22. The replacement ratios of RP were set at 0, 10, 20, 30, and 40% by mass. The dosages of P540, P80, and YH-S were 0.7, 0.1, and 0.05%, respectively. Cement and RP were mixed in a V-shaped blender according to the predetermined mix ratios to achieve a homogeneous consistency. Subsequently, water and chemical additives were added successively into the NJ-160 cement paste mixer, immediately followed by the addition of the homogeneous mixture. The paste was mixed according to the specified procedure and placed into a test mold measuring 40 mm × 40 mm × 160 mm. The specimens were removed from the mold after 24 h and placed in a standard curing box with a relative humidity of 98% and a temperature of 20 ± 2°C. Compressive strength was tested in accordance with the Chinese Standards GB/T 17671-2021 [30] and was measured at 1, 3, and 28 days, with average results obtained from at least three samples. The mix design and group numbers of grouts are listed in Table 2.

The fluidity was assessed in accordance with the Chinese Building Industry Standards JG/T 408-2013 [31]. Initially, the mixed grouting paste was poured into a truncated cone mold. Subsequently, the mold was gradually elevated to allow for the unrestricted flow of the paste until it ceased to spread. The maximum diameter of the paste’s spread was then measured and recorded as the initial fluidity. After a 30-min interval, the grouting paste was re-mixed and measured again to determine the fluidity at the 30-min mark. It is essential that each fluidity assessment was conducted at least three times, with the average value calculated from these measurements.

An ADVANCE Bruker D8 X-ray diffractometer was used to identify the mineral compositions (Cu Kα radiation, 2θ from 5° to 65°).

SEM images were recorded using an S-4800 scanning electron microscope (Hitachi, Japan), and the fracture surface was sputter-coated with gold before observation.

The influence of RP content on the fluidity of cement-based grouts is illustrated in Figure 3. It can be observed that fluidity increases with the addition of RP content; however, this increasing trend diminishes beyond a certain threshold. The grout without RP exhibits good fluidity and no time loss. Besides, the initial fluidity of grouts containing up to 10% RP exhibited an increasing trend with minimal time loss. Besides, the initial fluidity of grouts containing up to 10% RP shows an increasing trend with minimal time loss. However, when the RP content exceeds 20%, there is a significant decrease in initial fluidity, accompanied by an increase in time loss as the RP content is increased. On one hand, RP may fill voids [32] resulting from aggregate accumulation, optimize grain composition [33], and enhance grout fluidity. On the other hand, RP typically exhibits high water absorption [34,35,36], which can adversely affect the workability of fresh grouts. Furthermore, RP possess a large specific surface area, a rough texture, and an irregular shape. Consequently, when the RP content exceeds 20%, fluidity may decrease significantly. Moreover, the internal absorption of RP by PCE may further diminish the fluidity of grouts containing a high proportion of RP [37].

Figure 3

Influence of RP content on the fluidity of cement-based grouts.

The compressive strength of grouts with varying RP contents is depicted in Figure 4. It is evident that the compressive strength of the grouts at 1, 3, and 28 days decreases as the RP content increases. Furthermore, the reduction in strength exhibits a linear correlation with increasing RP content. Notably, the strength of grout containing 40% RP at each age falls below the specified value outlined in JG/T 408-2013 (Cementitious grout for coupler of rebar splicing).

Figure 4

Compressive strength of grouts with different contents of RP.

The influence of activation modes on the fluidity of RP grouts is depicted in Figure 5. In the case of single activation mode, the fluidity of samples S2, S3, and S7 decreased, while the fluidity of sample S6 increased compared to the non-activated sample S1. Besides, the fluidity of samples S4 and S5 was close to that of the S1 sample. Specifically, NaOH, Na₂CO₃, and thermal activation resulted in a reduction in the fluidity of RP grouts. Furthermore, CH and DEIPA activation showed no significant effect on the fluidity. Notably, TIPA activation may substantially enhance the fluidity. For example, the initial fluidity of samples S2 and S3 decreased by 27 and 29 mm, respectively, compared to that of sample S1. The inclusion of the thermally activated RP adversely affects the fluidity of the grouts, which decreased to 305 and 254 mm for the initial and 30-min fluidity measurements, respectively. In contrast, the initial and 30-min fluidity values of sample S6 were 370 and 356 mm, respectively. Among the two combined activation groups, the fluidity of sample S9 was higher than that of sample S8, attributed to the presence of TIPA.

Figure 5

Influence of activation modes on the fluidity of RP grouts.

The dissolution of strong inorganic alkalis, such as NaOH, Na₂CO₃, and Ca(OH)₂, can increase the concentration of hydroxide ions (OH⁻) while decreasing the concentration of calcium ions (Ca²⁺) in the pore solution [38]. This process may facilitate early hydration, accelerate the growth and recrystallization of ettringite (AFt), and subsequently disrupt the thin AFt film surrounding the surface of cement grain. The dissolution of the AFt thin film may further enhance the hydration rate, assist in the setting, and ultimately lead to a significant reduction in grout fluidity [39]. Following thermal treatment, the quantity of dehydration products increased, and the hydration rate of calcium oxide (CaO) was elevated, which resulted in the consumption of a substantial amount of mixing water [40]. Conversely, the fluidity of TIPA-activated RP grout was noticeably improved. The inclusion of TIPA, which contains a lipophilic methyl in its molecular chain, may introduce a certain amount of air, provide a ball-bearing effect, and consequently enhance the fluidity [39]. Furthermore, the combined activation of DEIPA, TIPA, and thermal treatment for reactive RP can further improve the grout fluidity.

The influence of activation modes on the compressive strength of grouts is illustrated in Figure 6. The strength of NaOH- and Na₂CO₃-activated RP grouts increased at 1 day but decreased at 3 and 28 days. The strength of the CH-activated group (S4) increased at each age, particularly at 28 days. Additionally, the strength of DEIPA-activated samples at 1, 3, and 28 days was all improved, although the 28th day strength is lower than the specified value of 85 MPa. In contrast, the strength of the TIPA-activated sample increased and exceeded the specified values. The strength of grout containing thermally treated RP also increased, with the performance at 3 and 28 days surpassing that of single chemical additives. Furthermore, the strength of grouts with combined-activated RP at each age was higher than that of grouts with single-activated RP. Notably, the strength of grouts with CH, DEIPA, and thermally treated combined-activated RP at 3 and 28 days reached the highest values of 73.3 and 95.7 MPa, respectively, representing increases of 19.5 and 17.7%, respectively. Meanwhile, the 1-, 3-, and 28-day compressive strength of the S9 sample was 43.1, 71.6, and 92.2 MPa, which were 20.4, 17, and 13.4% increases compared to the S1 sample, respectively. The addition of inorganic alkalis may disrupt Si–O–Si bonds and enhance the dissolution of quartz [41], thereby promoting the early hydration process of RP grouts. Alkanolamines, such as DEIPA and TIPA, can accelerate the hydration of C₃S, resulting in the formation more additional C–S–H gels. Elevated temperature treatment of RP may produce more reactive components due to the decomposition of the existing hydration products and calcite, leading to the formation of new hydration products and a denser microstructure [42].

Figure 6

Influence of activation modes on the compressive strength of grouts.

In conclusion, the impact of a single chemical additive activated mode on the activity of RP is limited, and the strength of the grout at various ages can only be marginally improved. Furthermore, NaOH- and Na₂CO₃-activated RP can significantly influence the later strength while reducing the fluidity of the grouts. Although the addition of thermally activated RP can enhance the strength of grout at all ages, it also leads to a notable decrease in fluidity. The strength of grout with CH, DEIPA, and thermally treated combined-activated RP exceeded the specified values at all ages; however, the fluidity after 30 min does not meet the required standards. In contrast, the compressive strength and fluidity of grouts with DEIPA, TIPA, and thermally treated combined-activated RP comply with the specified values outlined in the Chinese standards JG/T 408-2013.

Figure 7 shows the XRD patterns of the hydration products of the RP grout paste under various activation modes. The primary hydration products in the crystal phase include calcite, portlandite, AFt, quartz (SiO₂), and incompletely hydrated C₃S and C₂S.

Figure 7

XRD patterns of hardened samples of RP activated by different methods at 1 (a) and 28 (b) days.

On the first day, there was no significant change in the peaks of SiO₂ and portlandite in the hydration products of each of the inorganic alkali-activated samples (S2 and S4) compared to the non-activated group (S1) (Figure 7a). Possibly, the solubility of SiO₂ in the alkali activator was low [43]. Notably, the pozzolanic effect of the RP in the alkali-activated grouts was not apparent, which aligns with the results of the strength testing. In contrast, the peak intensity of SiO₂ in the hydration products of the alkanolamine-activated groups (S5 and S6) decreased. Additionally, the peak intensity of C₃S decreased, while that of AFt increased. On one hand, the dissolution of SiO₂ was facilitated by alkanolamines [44]. On the other hand, the hydration of C₃S may have been accelerated by alkanolamines, resulting in the formation of additional C–S–H gels. It is noteworthy that a new diffraction peak appears in the XRD pattern of the TIPA-activated group, which may be attributed to hemicarboaluminate [45,46]. In the XRD patterns of the thermally and combinatorially activated groups (S7, S8, and S9), the peak intensity of SiO₂ was lower than that of the other groups. Some SiO₂ may have transformed into amorphous silica under elevated temperature treatment and reacted with alkanolamines, which exhibited weak alkalinity. A certain relationship exists between the peak intensities of portlandite and calcite in the hydration products and the degree of hydration [43]. For the hydration products of S7, S8, and S9 groups, the portlandite peaks were substantially weakened compared to the other activated groups. The peaks of C₃S and C₂S were lower than those of the other samples, indicating that these three groups contained few unhydrated cement particles and had a higher degree of hydration than the other samples. Thus, thermal and alkanolamines activation was beneficial for the early strength development of the system.

On the 28th day, the peak intensity of portlandite in each activated group decreased and was lower than that of the S1 group (Figure 7b). Notably, the SiO₂ peaks of the S2 and S6 groups at 2θ = 27° did not exhibit significant changes compared to the other five activated groups (S4, S5, S7, S8, and S9). This observation may be attributed to the hydration of cement grains and RP. The change in peak intensity suggests that activation may have accelerated the pozzolanic reaction between portlandite and SiO₂. Furthermore, the content of AFt decreased due to the presence of OH⁻, which may facilitate the transformation of AFt to AFm [39]. Similarly, the peak intensity of CH in the S5, S6, S8, and S9 groups decreased, possibly due to the chelation of alkanolamines with Ca²⁺. This chelation effect can reduce the solubility of Ca²⁺ and the crystallinity of CH in the pore solution, potentially inhibiting the nucleation and growth of CH [44].

The microstructures of the hydration products in various grout pastes on the first day are illustrated in Figure 8. The microstructure of the non-activated group (S1) exhibited a loose structure with several cracks (Figure 8a), resulting in low strength. In addition to the C–S–H gel, small amounts of portlandite and AFt were detected in the pores. The microstructure of the NaOH-activated sample (S2) displayed a significant amount of C–S–H gels, along with numerous pores and several voids (Figure 8b). The SEM image of the S4 sample revealed a large number of hexagonal plate-like CH crystals (Figure 8c), which may be attributed to the nucleation of additional Ca²⁺ and OH⁻ ions. The morphology of the S5 sample showed a substantial amount of fibriform C–S–H (Figure 8d), which may contribute to a loose microstructure. The microstructure of the S6 sample exhibited several voids (Figure 8e), likely due to the air-entraining effect of DEIPA [45]. Besides, there were several anchor-shaped C–S–H gels and a small number of needle-like AFt crystals, which also contributed to a loose microstructure. The morphology of the S7 sample displayed a large number of hexagonal plate-like CH crystals and C–S–H gels (Figure 8f), which adhered to other hydration products and unhydrated cement grains. The hydration products consisted of anchor-shaped C–S–H, tetragonal CH, and tufted AFm (Figure 8g).

Figure 8

SEM images of samples S1 (a), S2 (b), S4 (c), S5 (d), S6 (e), S7 (f), S8 (g), and S9 (h) at 1 day.

The microstructure of the S9 sample presented several plate-like and hexagonal CH crystals, along with anchor-shaped and continuous C–S–H gels (Figure 8h). In summary, the internal structural integrity of the C–S–H gel in the activated samples differed from that of the non-activated sample. However, the width of the microcrack, pore size, and density of the microstructure reduced in the combined activated groups compared to the singly activated groups, resulting in increased compressive strength for the former compared to the latter.

Figure 9 illustrates the microstructures of hydration products in various grout pastes on the 28th day. The microstructure of the S1 sample on the 28th day exhibited a loose and porous morphology (Figure 9a). The SEM image of the S2 group revealed an even looser morphology, characterized by voids not filled by the continuously formed C–S–H gels (Figure 9b) when compared to the S1 group. The microstructures of the hydration products in the S4 group exhibited high portlandite content (Figure 9c), resulting in a denser microstructure. In the morphologies of the alkanolamines-activated groups (Figure 9d and e), hydration products such as C–S–H gels, CH, and AFm phases were further formed, interlocked, and created a stable microstructure [47]. Conversely, the morphology of the S7 group (Figure 9f), which was activated by calcination, contained a significant amount of C–S–H gel but exhibited a loose microstructure. Comparing Figure 9g and h, it is evident that more pores and voids were filled with continuous C–S–H gels, resulting in hydration products that interlocked to form a stable microstructure. Additionally, the morphology of S8 displayed a greater presence of the untransformed metastable phase AFm, which may contribute to its lower strength compared to S9. Furthermore, when comparing pastes with the non-activated paste and single additive-activated RP, the pastes with combined-activated RP consistently demonstrated a denser morphology.

Figure 9

SEM images of samples S1 (a), S2 (b), S4 (c), S5 (d), S6 (e), S7 (f), S8 (g), and S9 (h) at 28 days.