Introduction

Hydraulic lime is more durable and environmentally friendly than cement and air-setting lime. Hence, it has been widely used in construction and building materials in ancient and modern architecture [1,2,3,4,5,6,7,8]. Natural hydraulic limes have been researched and implemented in European countries, including modification and application in the conservation of historical buildings [1, 3, 8,9,10,11,12,13,14]. In China, hydraulic limes were not researched until recent decades. With the discovery of several hydraulic limes in ancient architectural ruins, Chinese hydraulic limes achieved a breakthrough. Ginger nut, discovered as ground material in the Dadiwan site in Qin’an County, Gansu Province, China [8, 12, 15,16,17], is a type of calcite concretion in the Quaternary sedimentary ore deposit, primarily comprising calcite and clay minerals [18]. However, the original ginger nut is not a proper construction material because no cementitious components are present. This material is considered the first evidence of natural hydraulic lime in China. Another Chinese hydraulic lime, the aga tu, a unique building material of the Tibetan ethnic group, abbreviated as AGA soil [16], is a calcite concretion formed in a semiarid plateau area in south Tibet [19]. Shell lime is another hydraulic lime calcined from shells, and it was widely used in ancient buildings in southeast China [20]. These three hydraulic limes are found mainly in ancient Chinese buildings and have been researched to restore rock heritages in recent years [21].

However, higher requirements for restoring mortars have been proposed because stone relics are in complex and changing environments (Fig. 1). These stone relics require urgent rescue. The shrinkage and early strength of Chinese hydraulic lime mortars have not yet reached the demands of grouting when dealing with uncertain conservations, leading to poor grouting compactness and secondary cracks. However, previous research has suggested that cement and lime mixed with metakaolinite have better strength, higher weather resistance abilities, and less shrinkage [22,23,24,25,26,27]. The metakaolinite was transformed from kaolinite under a sintering process (600 °C–900 °C), which primarily comprised amorphous SiO2 and Al2O3. This is not surprising because SiO2 and Al2O3 in metakaolinite react with Ca(OH)2 to generate calcium aluminosilicate, improving the mortar strength [28]. A growing number of studies have focused on the modification effect of metakaolinite on concrete in terms of structure, mortar working ability, and durability [29,30,31,32,33,34,35]. However, no information on the improvement mechanism of the metakaolinite and Chinese hydraulic limes mentioned above is available. This information could guide potential applications in mortar design that may lead to a better understanding of Chinese hydraulic limes.

Fig. 1
figure 1

Typical stone relics requiring restoration intervention in China: a interpenetrated crack of stone tablet in Yunju Temple, Beijing; b dehiscence of Sumeru in Mountain Resort, Chengde; c cracks of baluster in Mountain Resort, Chengde; and d structural fracture of cliff carvings in Yuanjue Cave, Sichuan Province

Full size image

Therefore, to figure out how metakaolinite modified Chinese hydraulic lime mortars, we conducted tests on Chinese hydraulic lime mortar mixed with metakaolinite. The mortar working abilities including setting time, fluidity, and stone body behavior, such as contracting rate, porosity, and mechanical properties were also tested. In addition, we conducted weathering resistance ability tests on the specimens prepared using a new ingredient. Moreover, the characteristics of carbonation and hydration were analyzed over 900 days using X-ray diffraction (XRD).

Materials and methods

Binders and aggregates

The ginger nut primarily comprised calcite and clay minerals (Fig. 2a). The AGA soil was found as a tomb-building material in south Tibet and is now widely used in houses and temple buildings (Fig. 2b). Shell lime is another hydraulic lime calcined from shells (Fig. 2c) and can be traced to the Spring and Autumn Period and the Warring States Period. Currently, this material is widely used to maintain ancient buildings in the southwest coastal area of China. The shell calcined in this study was a type of oyster sampled from Zhejiang province, southeast China. The leading Chinese hydraulic lime currently used comprises ginger nut, AGA soil, and shell lime (Table 1). The calcium carbonate and silica composition can produce hydraulic lime through calcination. Table 2 lists the chemical compositions of the calcinated materials. The Chinese hydraulic limes are composed mainly of air-setting (CaO) and hydraulic (β-CaSiO3 and Ca2Al2SiO7) components. The aggregate was quartz sand (2 mm) in equal proportion.

Fig. 2
figure 2

Raw materials: a ginger nut, b AGA soil, and c shell

Full size image
Table 1 Chemical composition of the raw materials
Full size table
Table 2 Materials composition after calcination at 1000 °C for 2 h
Full size table

Mixture ratio

The mortar performance and consolidation properties largely depend on the water binder ratio and aggregates. Generally, several criteria, including the water binder ratio and aggregates, should be considered before hydraulic lime is applied to heritage conservation. The water binder ratio and mortar setting time should be neither too large nor too small, considering handleability and fluidity. Therefore, some criteria must be satisfied before it can be considered: (1) the water binder ratio should not be too small because of grouting fluidity; (2) the setting time should be proper for mortar mixing and grouting; (3) the final setting time should be short because early strength is necessary for grouting; (4) the strength due to consolidation should be high enough for reinforcement. Different mixture and water binder ratios were used to determine the optimal mixture ratio of these three materials. According to the design of the masonry mortar [36], the optimal mixture and water binder ratios were determined based on the evaluation of mortar fluidity, setting times, consolidation shrinkage, and age strengths. Table 3 lists the mortar composition and proportion by weight.

Table 3 Mixture ratios and water binder ratios
Full size table

Experimental procedure

The slurry fluidity and setting times were determined according to the mixture ratio in Table 3. The consolidation age strengths of 3, 7, 14, and 28 days were tested, including the compressive and flexural strengths. After curing for 50 days, weather tests were conducted under different weathering destruction conditions [37], and the weather resistance performance was evaluated from the compressive and flexural strength variations according to weather ability test standards [38].

This study examined how materials carbonate and hydrate in a moist environment. After curing for 28 days, the consolidations were placed in Feilaifeng Cliffside Sculptures in Hangzhou, Zhejiang Province (Fig. 3). The annual average rainfall and relative humidity in the meteorological environment (Fig. 3a) are approximately 1100–1600 mm and 70%, respectively. The carbonation and hydration of the consolidations were monitored by XRD after curing for 5, 300, 600, and 900 days.

Fig. 3
figure 3

Specimens and curing environment: a temperature and rainfall in Hangzhou, Zhejiang Province and b specimens

Full size image

Freeze–thaw resistance test: The samples were frozen (40 × 40 × 160 mm) at − 30 °C for 12 h and thawed at 25 °C and a relative humidity of 70% for 12 h. Dry–wet resistance test: The samples (40 × 40 × 160 mm) were dried at 100 °C for 12 h and cooled to 25 °C at a relative humidity of 70% for 12 h. Salt erosion resistance test: The samples (40 × 40 × 160 mm) were immersed in a saturated Na2SO4 solution for 20 h and dried at 105 °C for 4 h. Alkalinity resistance test: The samples (40 × 40 × 160 mm) were immersed in a NaOH solution (2%) for 12 h and dried at 105 °C for 4 h. Water stability test: The water stability of the specimens mixed with metakaolinite was determined by immersing the samples (40 × 40 × 160 mm) in water (25 °C) for 24 h, followed by natural air drying; these were called the water-immersed dry samples. Uniaxial compression test: The uniaxial compression test was conducted using the suggested methods for determining the uniaxial compressive strength and deformability of rock materials [39]. Six samples were tested under the same experimental conditions. Mercury intrusion test: Sample porosity tests were performed using an AutoPore 9500 high-performance automatic mercury porosimeter. The samples were cut into cubes and dried in an oven. Each undisturbed specimen was divided into two equal parts. Multiple cube specimens smaller than 1 cm3 were taken for testing in each part.

Results

Basic properties of mortars

The mortars were mixed and evaluated in terms of mortar setting times and shrinkage (Table 4). The fluidity of the mortars did not change considerably when metakaolinite was added. However, the initial and final setting times were shortened after mixing with metakaolinite, particularly the initial setting time of HP. Hence, metakaolinite would improve the early strength of the mortar. In addition, the porosity of the consolidation mixed with metakaolinite increased and shrinkage decreased. Therefore, the mortar behavior of the restoration functions was improved in terms of water and air permeability. The shrinkage of the consolidations was decreased so that it would barely produce secondary cracks during the grouting process.

Table 4 Basic properties of the mortar samples
Full size table

Figure 4 shows the 3-, 7-, 14-, and 28-day age strengths of the mortar specimens. The age strengths, especially the early strengths, improved when metakaolinite was added to the mortar. The 3-day compression strength increased by 47.8% (LP), 165.2% (AP), and 302.5% (HP). The strengths of the shell lime specimens improved considerably when metakaolinite was added. The final compression and flexural strength of the shell lime mortar increased approximately 9 and 6 times, respectively. However, the strength of shell lime mixed with metakaolinite (HP) appeared to continue to increase in the later period. Metakaolinite first reacts with CaO in shell lime and produce Ca2Al2SiO7 in the carbonation process. Ca2Al2SiO7 is then transformed into Ca2Al2SiO7·nH2O during the hydration process. Hence, the consolidation process will last longer than those mortars without metakaolinite.

Fig. 4
figure 4

Specimens with and without age strength in 28 days: a compression strength and b flexural strength

Full size image

Weather resistance abilities

Environmental changes are important factors in the deterioration of cultural relics. Hence, evaluating the weathering resistance ability is necessary to restore materials. After curing for 50 days at 25 °C and relative humidity of 70%, strength tests were conducted on the specimens under different environmental destruction conditions to evaluate their weathering resistance ability. These evaluations include freeze–thaw resistance ability, dry–wet resistance stability, water stability, salt erosion resistance ability, and alkalinity resistance ability. The evolution of the specimen strength is shown in Fig. 5.

Fig. 5
figure 5

Strength evolution of specimens: a L, b A, and c H. (In figures: 1. Original specimen. 2. Freeze-thawed specimen. 3. Dry-wetted specimen. 4. Salt-eroded specimen. 5. Alkalinity-eroded specimen. 6. Water-immersed dry specimen) [36, 45]

Full size image

Freeze–thaw resistance ability: Freeze–thaw and soluble salt are the most significant erosion agents for construction materials [40], and performance degradation caused by freeze–thaw cycles and salt erosion was more conspicuous for the lime materials. Therefore, after 18 freeze–thaw cycles, we conducted compression strength test. The compressive strength of all specimens decreased after 18 freeze–thaw cycles (Fig. 5). The compression strength of the modified ginger nut specimens (L) decreased only by 17.40%, while that of the specimens mixed with metakaolinite (LP) decreased by 15.7%. Specimen A decreased by 17.16%, but AP decreased by only 8.62%. For the H and HP specimens, the compression strength did not change considerably. Thus, mortars mixed with metakaolinite have superior freeze–thaw resistance ability to those without metakaolinite.

Dry–wet resistance ability: After 18 dry–wet cycles, we conducted compression strength test. The results are shown in Fig. 5. The compression strength of L increased from 8.6 MPa to 18.3 MPa after the tests. The strengths of A and H were similar, which could be explained by the promotion of carbonation during dry–wet cycles [41]. However, the strength of the mortars mixed with metakaolinite changed slightly after 18 cycles. Therefore, mortars with metakaolinite were more adapted to the alternate wetting and drying environments. Indeed, many studies have shown that the frost resistance of cement could be improved by metakaolinite by reducing the pore size [42, 43].

Salt erosion resistance ability: Compressive strength tests were conducted after five cycles. The compression strengths of all specimens improved after five cycles (Fig. 5). This may be due to the possible recrystallization of Na2SO4 at high temperatures, which filled the pore spaces. Comparing these six kinds of materials, the shell lime mixed with metakaolinite has better salt erosion resistance ability than the others.

Alkalinity resistance ability: Water resistance is a crucial indicator for mortars because they will be exposed to an alkaline environment. Compressive strength tests were conducted after drying. The compression strength of specimens without metakaolinite changed slightly after the tests (Fig. 5). However, the mortars mixed with metakaolinte showed poor alkalinity resistance compared to those without metakaolinite. This was attributed to the higher porosity and larger pore size after adding metakaolinite to the mortars. The alkaline environment may eventually trigger a dedolomitization process inside the aggregate grains [44].

Water stability: To ascertain the degree to which the different mortar mixes permitted better water resistance ability, compressive strength tests were conducted before and after water immersion. Figure 5 shows that mortars mixed with metakaolinite have better mechanical strength than those without metakaolinite.

Carbonation and hydration characteristic analysis

The strength of the hydraulic mortar increased due to concretion, carbonation, and hydration. The carbonation and hydration of hydraulic lime are affected by many factors, such as temperature, relative humidity, and CO2 content [46,47,48]. To monitor the carbonation and hydration of the consolidations, we placed these mortars in Feilaifeng Cliffside Sculptures in Hangzhou, Zhejiang Province, and analyzed using XRD after curing for 5, 300, 600, and 900 days.

Figure 6 shows the XRD patterns of the specimens after curing. The original composition of the modified ginger nut was CaO (39.20%), β-CaSiO3 (26.70%), and Ca2Al2SiO7 (18.90%). After curing for 5 days, XRD revealed Ca(OH)2 (22.70%), CaCO3 (32.60%), β-CaSiO3· nH2O (22.27%), and Ca2Al2SiO7·nH2O (14.48%). The chemical reactions can be summarized as Eqs. (1), (2), (3), and (4), respectively. After 300 days, the Ca2Al2SiO7·nH2O content increased to 19.02% during the hydration process, and the CaCO3 content was 38.46% (Fig. 6). However, the CaO content was reduced to 7.78% due to carbonation, which decreased to 1.06% after 900 days. Regarding the modified ginger nut mortar with metakaolinite, the β-CaSiO3·nH2O and Ca2Al2SiO7·nH2O contents were higher than those of mortar L after 900 days. Correspondingly, the CaO, β-CaSiO3, and Ca2Al2SiO7 contents were much lower than those of the mortar without metakaolinite. The increased Ca2Al2SiO7 contents can be explained by the reaction (formula (5)) of Al2O3·SiO2 in metakaolinite and Ca(OH)2 produced by the CaO and water reaction in the early stage.

Fig. 6
figure 6

Carbonation and hydration results of the specimens by XRD: a ginger nut without metakaolinite, b ginger nut with metakaolinite, c AGA without metakaolinite, d AGA with metakaolinite, e shell lime without metakaolinite, and f shell lime with metakaolinite. (Peak 1-CaCO3. 2-CaO. 3-Ca(OH)2. 4-βCaSiO3. 5-Ca2Al2SiO7. 6-βCaSiO3·H2O. 7-Ca2Al2SiO7·H2O)

Full size image

The modified AGA clay mortars showed similar behavior to the modified ginger nut in 900 days, but the Ca2Al2SiO7·nH2O contents increased more slowly than the modified ginger nut mortars. The modified AGA clay mortars mixed with metakaolinite produced more β-CaSiO3·nH2O than those without metakaolinite (formula (3)). This is a cause of a higher early strength increase of mortars when adding metakaolinite.

$${\text{CaO}}+{{\text{H}}}_{2}{\text{O}}\to {\text{Ca}}{\left({\text{OH}}\right)}_{2}$$
(1)
$${\text{Ca}}{\left({\text{OH}}\right)}_{2}+{{\text{CO}}}_{2}\to {{\text{CaCO}}}_{3}+{{\text{H}}}_{2}{\text{O}}$$
(2)
$$\upbeta -{{\text{CaSiO}}}_{3}+{\text{n}}\cdot {{\text{H}}}_{2}{\text{O}}\to\upbeta -{{\text{CaSiO}}}_{3}\cdot {{\text{nH}}}_{2}{\text{O}}$$
(3)
$${{\text{Ca}}}_{2}{{\text{Al}}}_{2}{{\text{SiO}}}_{7}+{\text{n}}\cdot {{\text{H}}}_{2}{\text{O}}\to {{\text{Ca}}}_{2}{{\text{Al}}}_{2}{{\text{SiO}}}_{7}\cdot {{\text{nH}}}_{2}{\text{O}}$$
(4)
$${{\text{Al}}}_{2}{{\text{O}}}_{3}\cdot {{\text{SiO}}}_{2}+{\text{Ca}}{\left({\text{OH}}\right)}_{2}\to {{\text{Ca}}}_{2}{{\text{Al}}}_{2}{{\text{SiO}}}_{7}$$
(5)

The carbonation and hydration of shell lime mortar (Fig. 6e, f) were modified when metakaolinite was added. In the first 5 days, the mortar without metakaolinite generated 48.14% CaCO3, and the CaO was retained at 27.60% due to carbonation. The β-CaSiO3·nH2O content increased from 0 to 5.26% due to hydration. However, the mortar mixed with metakaolinite (Fig. 6f) produced 5.25% Ca2Al2SiO7·nH2O, which was not found in the mortar without metakaolinite (Fig. 6e). Therefore, Ca2Al2SiO7 is generated by the action of active metakaolinite and the CaO of shell lime. Accordingly, the sections of mortars with metakaolinite showed higher degrees of carbonation (thickness of the carbonized layer: approximately 5 mm on the surface) than those without metakaolinite (thickness of the carbonized layer: approximately 2 mm on the surface), as shown in Fig. 7. After 900 days, the mortar mixed with metakaolinite produced more Ca2Al2SiO7·nH2O and β-CaSiO3·nH2O than the shell lime mortars. As a result, the CaCO3 contents decreased in shell lime mortars mixed with metakaolinite because it generated more hydration products.

Fig. 7
figure 7

Section of H and HP after curing for 300 days: a shell lime without metakaolinite and b shell lime with metakaolinite

Full size image

The variations of the relative percentages of the contents during the carbonation and hydration process were examined by XRD to determine how the components transformed in the mortars. Figure 8 shows the changes in the relative components. The CaCO3, β-CaSiO3·nH2O, and Ca2Al2SiO7·nH2O contents increased rapidly in the first 5 days, while the CaO, β-CaSiO3, and Ca2Al2SiO7 contents decreased because of consumption during the carbonation and hydration processes. After 900 days, the carbonation and hydration product increased at a slower rate than in the early stage, and the consumption of CaO, β-CaSiO3, and Ca2Al2SiO7 decreased.

Fig. 8
figure 8

Chemical composition variations during carbonation and hydration: relative percentages of a CaCO3, b β-CaSiO3·nH2O, c Ca2Al2SiO7·nH2O, d CaO, e β-CaSiO3, and f Ca2Al2SiO7

Full size image

The mortar-modified ginger nut mixed with metakaolinite (LP) produced more Ca2Al2SiO7·nH2O than the mortar without metakaolinite (L), and the modified AGA clay mortar (A and AP) showed results similar to those of the modified ginger nut and modified AGA clay. However, the mortar shell lime mixed with metakaolinite (HP) produced less CaCO3 but more β-CaSiO3·nH2O and Ca2Al2SiO7·nH2O than the mortar shell lime without metakaolinite (H), and the rate of CaO consumption was higher. This could be explained by the reaction of Al2O3·SiO2 and Ca(OH)2 (formula (5)). In addition, the relative percentage of Ca2Al2SiO7 initially increased and then decreased. CaO consumption in the shell lime mortar was lower than in the modified ginger nut and AGA clay, which was likely due to the higher CaO contents in the original shell lime than in the other two limes.

The energy dispersive spectrometry (EDS) results (Fig. 9) showed that the weight percentage (Wt) of Ca increased from 4.39% to 61.06% when modified using metakaolinite (HP), while the percentage of Si decreased from 80.39% to 24.31%. This further proves that the lime mortars produced more carbonation with the modification of metakaolinite, which could also be inferred from the SEM results (Fig. 10).

Fig. 9
figure 9

EDS results of H and HP after curing for 900 days

Full size image
Fig. 10
figure 10

SEM images of H and HP after curing for 900 days: a shell lime without metakaolinite and b shell lime with metakaolinite

Full size image

Conclusion

In this study, we tried to modify Chinese hydraulic limes using metakaolinite. The following conclusions were drawn concerning the mechanical properties and weather resistance of lime combined with metaklinite. The physical properties of the Chinese hydraulic lime mortars, such as shrinkage, porosity, mechanical strength, and weather resistance, were considerably improved. The mortar fluidity and initial setting time decreased because metaklinite can replace some of the lime. The porosity increased because active SiO2 in metaklinite promoted the hydration and carbonation processes, resulting in more pores during consolidation. Adding metakaolin resulted in additional hydrate phases that contributed to the mechanical strength and weather resistance ability. XRD and SEM indicated that more carbonation and hydration products were produced after mixing with metakaolinite, and more uniform and consolidated structures were formed from the effects of metakaolinite. The shell lime mortars were considerably modified when adding metakaolinite compared to ginger nut and AGA soil. The mortars modified in this study provided a selection for crack grouting or repair works. The application issue and evaluation of the restoration results will be examined in future work.