Study on the Mechanical Properties and Durability of Hydraulic Lime Mortars Based on Limestone and Potassium Feldspar

Abstract

Natural hydraulic lime (NHL) can be used as an inorganic cementitious material, as it exhibits low shrinkage, salt-alkali resistance, moderate strength, and good durability with cultural relics. There has been increasing interest in NHL, as it is considered an appropriate material for the restoration and reinforcement of architectural cultural relics. In this study, limestone and potassium feldspar were mixed and calcined at different ratios and high temperatures, and artificial hydraulic lime (HL) was produced. According to the X-ray diffraction (XRD) results, the resulting products after high-temperature calcination were mainly composed of calcium oxide, dicalcium silicate (C₂S), and dicalcium aluminosilicate (C₂AS). As a compromise, when potassium feldspar accounted for 30% of the total mass, HL contains a more suitable air-hardening component and hydraulic component. Scanning electron microscope (SEM) and Energy Dispersive Spectrometer (EDS) analyses show that the phases of calcium carbonate (CaCO₃) and hydrated calcium silicate (C-S-H) gradually increased with prolonged curing time for HL. To study the partial mechanical properties and durability of HL, a comparison was made with NHL. The mechanical properties were investigated with the flexural and compressive strengths and shrinkage. The results show that HL has higher strength than NHL, but NHL has smaller shrinkage. Accelerated aging tests indicated that HL and NHL5 led to higher resistance to water immersion, fluctuations in temperature and humidity, sulphate decay, an alkali environment, and frost–thaw action than NHL2. HL has excellent mechanical properties and durability and can be considered a conservation material for stone relics in the future.

1. Introduction

Lime was the most widely used inorganic binder in ancient times before the invention of Portland cement. Lime was found to have been used by ancient people to handle floors and walls approximately 4000 years ago in ancient China. However, it cannot meet the cultural architectural heritage reinforcement construction engineering requirement due to the lower water resistance, the lower strength, and the slower air-hardening reaction of lime [1]. NHL with hydraulic (C₂S) and air-hardening phases (Ca(OH)₂) has been used to protect architectural cultural relics in the 20th Century [2,3]. Mechanical properties and water resistance of the NHL could be improved due to the presence of C₂S compared to that of air lime. For example, NHL can be mixed with sand to form a mortar, which can be used as a grouting material for the protection and reinforcement of stone relics [4]. It is used as an inorganic cementitious material, as it exhibits low shrinkage, salt-alkali resistance, and moderate strength, in addition to good durability with cultural relics [5]. With the increasing use of NHL in the field of heritage conservation and its increasing demand, researchers have studied natural hydraulic lime in-depth and have carried out modification studies. For example, Silva et al. studied the effect of adding NHL to aerial lime-based mortar, and the mortar mixture of lime and NHL is more suitable as a restoration material for cultural relics [1,6]. The addition of metakaolin to NHL can improve the mechanical properties of mortar [7,8]. The use of zeolite to replace NHL in an appropriate proportion and the addition of glass fiber can improve the compactness, mechanical properties, and shrinkage resistance of mortar [9,10]. Polypropylene fibers can improve the toughness of NHL mortar [11]. Graphene oxide addition in NHL mortar led to a slight improvement in the mechanical and physical characteristics [12]. Due to the many advantages of NHL, conservators have utilized natural hydraulic lime to protect cultural heritage [13,14].

Natural hydraulic lime is common in Europe, but the mineral resources of preparation NHL are scarce in China, and it is primarily imported for the restoration of ancient buildings if it is used extensively. The most well-known case is the use of NHL2 produced by The German Hessler to protect the Guangxi Huashan rock paintings in China, which has achieved good results [15]. Therefore, Chinese scholars have been developing hydraulic lime. However, there are various problems with the currently produced hydraulic lime. Some materials contain cement components [16], and cement contains large amounts of soluble salts that can cause significant harm to cultural stone relics [17]. Another researcher used marlstone as a raw material to produce hydraulic lime, similar to NHL2. The material was difficult to scale due to its high preparation cost [18]. In addition, lead and zinc mine tailing were included in the research as raw materials for the preparation of hydraulic lime [19]. However, lead and zinc mine tailing are not widely distributed in China, and the components of raw materials differ from one region to another, making them difficult to use on a large scale.

According to the European Building Lime Standards [20], natural hydraulic lime is lime with hydraulic properties produced by burning more or less argillaceous or siliceous limestones with a reduction to powder by slaking with or without grinding. In this paper, limestone and potassium feldspar were used as raw materials, which are more common in nature, to try to develop hydraulic lime. According to the XRD results, HL primarily contains calcium oxide (CaO), calcium hydroxide (Ca(OH)₂), dicalcium silicate (C₂S), and dicalcium aluminosilicate (C₂AS). NHL contains CaO, Ca(OH)₂, and C₂S, and has similar phases to HL [21]. Therefore, NHL2 and NHL5 were selected for comparison with HL to analyze their mechanical strength, shrinkage, and durability.

2. Materials and Methods

2.1. Raw Materials

Potassium feldspar and limestone were used as raw materials to prepare the hydraulic lime (HL) in this work. Potassium feldspar was purchased from Lingshou County, Hebei Province, China, and limestone was purchased from Xinxiang City, Henan Province, China. The chemical compositions of potassium feldspar and limestone are shown in

Table 1. Chemical compositions of potassium feldspar and limestone (Mass/%).

Composition	SiO2	Al2O3	K2O	Na2O	CaO	Fe2O3	MgO	Total	LOI
Potassium feldspar	63.36	15.74	15.32	2.36	0.53	0.91	0.47	98.69	1.31
Limestone	1.34	0.29	0.00	0.00	54.13	0.17	1.53	57.46	42.54

. The chemical compositions of potassium feldspar contain SiO₂, Al₂O₃, and K₂O. The limestone was primarily composed of CaCO₃.

2.2. Sample Preparations

2.2.1. Calcination

Potassium feldspar and limestone were crushed and ground with a ball mill. The particle size of powders was approximately 80 µm. Potassium feldspar and limestone were subsequently mixed at mass percentage ratios of 1:9, 2:8, 3:7, and 4:6 and then fabricated into blocks approximately 50 mm × 50 mm × 5 mm in size with a certain amount of water. After drying for 3 days, the mixed samples were heated at a rate of 5 °C per minute up to 1000 °C, and were then maintained at 1000 °C for 3 h [22,23]. The samples were quickly taken out of an electric muffle furnace and cooled to room temperature. After cooling, the mixed samples were slaked and made into powder (Figure 1). The calcined mixtures in mass ratios of 1:9, 2:8, 3:7, and 4:6 were numbered K1, K2, K3, and K4, respectively.

2.2.2. Specimens’ Preparations

HL was slaked for 10 days and then crushed into powders, which had a particle size of approximately 80 µm. Due to chemical stability and physical performance, quartz sand was used as an admixture in NHL lime mortars [24]. HL and NHL were mixed with China ISO (International Organization for Standardization) standard sand (a particle size range of 0.08–2 mm) at a mass ratio of 1:1, and the specimens were fabricated with a water–binder ratio of 0.33. The prismatic specimens with a size of 40 mm × 40 mm × 160 mm (width × depth × length) were subjected to the center-point loading flexural test. The specimens 70 mm × 70 mm × 70 mm in size were fabricated to test wave velocities. Mortar specimens were cured for 28 days or 56 days at 20 °C and 70% relative humidity (RH). NHL2 and NHL5 were produced by Saint Astier Company, France.

2.3. Test Methods

2.3.1. Microstructure Analysis

The compositions of the different samples were tested by an X-ray diffraction instrument (XRD) (Japanese Science RINT2000), with Cu as the target, an operating voltage of 40 kV, a working current of 100 mA, a 2θ angle scan range of 5–70°, and a step rate of 0.02° per s. In addition, Multiple Document Interface (MDI) Jade software was used to calculate the quantification of each phase of the samples. Additionally, the micro-morphology and energy spectrum of the different samples were analyzed by a low-vacuum scanning electron microscope (SEM) and an Energy-Dispersive Spectrometer (EDS).

2.3.2. Mechanical Properties Test

The flexural and compressive strength testing of specimens was conducted in accordance with CEN. EN 1015-11:1999/A1:2006 [25]. The shrinkage was determined using the retractometer by ASTMC 1148-92a:2008 [26]. The elastic wave velocities were tested by using the nonmetal sonic apparatus (mode of RSM-SY5) with a frequency of 50 kHz.

2.3.3. Durability Test

Accelerated aging tests can better reflect the durability of the samples, therefore, all durability tests were conducted using lime mortar specimen curing for 56 d.

(1) Water stability test

The specimens aged 56 days were immersed in water at 20 °C for 24 h. The specimens were taken out of the water, and their surface was wiped dry. The compressive and flexural strength was tested immediately.

(2) Soundness test

The specimens aged 56 days were immersed in the saturated solution of Na₂SO₄ for 20 h. Then, the specimens were taken out of the solution and baked at 105 °C for 4 h. After five cycle tests, the strength of the specimens was tested.

(3) Alkali resistivity test

The specimens aged 56 days were immersed in a 2% solution of NaOH for 12 h. Then, the specimens were taken out of the solution and baked at 105 °C for 4 h. Afterward, the strength of the specimens was tested.

(4) Frost–thaw cycle test

The specimens aged 56 days were put in the fridge and frozen at under −30 °C for 12 h. Then, the specimens were cured at 20 °C and RH 70% for 12 h. After eighteen cycles, the strength of the specimens was tested.

(5) Temperature and humidity cycle test

The specimens aged 56 days were heated for 12 h at 100 °C, and then the specimens were cured at 20 °C and RH 70% for 12 h. After eighteen cycles, the strength of the specimens was tested.

3. Results and Discussions

3.1. Changes in Hydraulic Lime Composition with Mixing Percentage

Figure 2 and

Table 2. Mass percentage of the air-hardening component and hydraulic component for K1, K2, K3, K4, NHL2, and NHL5.

Components and Content	Mass Percentage (%)
	Air-Hardening Component			Hydraulic Component
No.	CaO	Ca(OH)2	Total	C2S	C2AS	Total
K1	66.5	3.5	70.0	17.6	3.3	20.9
K2	40.9	10.2	51.1	28.4	10.1	38.5
K3	33.1	5.1	38.2	31.0	12.9	43.9
K4	29.6	5.1	34.7	27.4	19.3	46.7
NHL2	23.9	56.8	80.7	18.3	0.0	18.3
NHL5	45.0	20.2	65.2	33.8	0.0	33.8

show the XRD results of the mixed potassium feldspar and limestone powder, with mass percentage ratios of 1:9, 2:8, 3:7, and 4:6 after burning for 3 h at 1000 °C. MDI Jade 9 software was used to calculate the mass percentage of the air-hardening component and hydraulic component of the samples via XRD. Jade software is able to conduct qualitative and semi-quantitative analyses of the samples, and the semi-quantitative analysis results are calculated primarily based on the peak value. The semi-quantitative analysis has a certain error in the analysis of the phase with relatively simple ingredients, but this error is within an acceptable range. Therefore, the results of this experiment are relatively reliable.

The XRD characterization was performed after burning but before slaking. According to the XRD results, the resulting products after high-temperature calcination were primarily composed of CaO, 2CaO•SiO₂ (C₂S), and 2CaO•Al₂O₃•SiO₂ (C₂AS). Among those products, CaO was the air-hardening cementing material, while C₂S and C₂AS were the hydraulic-cementing materials [3]. After potassium feldspar and limestone were calcined at a mass ratio of 1:9, K1 generated 3.5% Ca(OH)₂ and 66.5% CaO, 17.6% C₂S, and 3.3% C₂AS.

With an increase in the potassium feldspar proportion, the content of Al₂O₃ also gradually increased. Al₂O₃ and CaCO₃ continued to generate C₂AS. The C₂AS content increased from 3.3% to 19.3%, while C₂S slowly increased from K1 to K3, and K4 decreased slightly. With increasing potassium feldspar content, more C₂AS is generated, and the C₂S content reaches its maximum when the potassium feldspar content is 30% at 1000 °C after burning for 3 h. Due to the production of C₂S and C₂AS, the content of CaO in the sample gradually decreases. According to European Standards [20], the high content of CaO and Ca(OH)₂ must be greater than or equal to 35%. The contents of CaO and Ca(OH)₂ are higher than 35% in K1, K2, and K3, while K4 is less than 35%.

Although K4 has more hydraulic components, its air-hardening phase is less than 35%. Therefore, K3 has more hydraulic phases than K1 and K2 and enhances the early strength of the test blocks even more. K3 is the best choice for comparison with NHL2 and NHL5.

3.2. Physical and Mechanical Performance

3.2.1. Flexural and Compressive Strength of Mortars

SK, SL2, and SL5 are prismatic or square blocks made of K3, NHL2, NHL5, and standard sand with a mass ratio of 1:1 and a water-binder ratio of 0.33, respectively. Both the flexural and compressive strength of the mortar specimens were tested after 3, 7, 14, 28, and 56 days with curing (Figure 3 and Figure 4). Under the same curing conditions, the flexural strength and compressive strength of the three types of hydraulic lime mortar specimens increased with aging.

The early strength growth of the test blocks primarily comes from the hydraulic phase. If there are more hydraulic components, the early strength growth is fast, and if there are fewer hydraulic components, the early strength growth is slow. C₂AS also takes place in hydration reactions, but its early hydration activity is relatively poor [27]. It is important that the C₂S phase determines the early strength of the specimens. K3 and NHL5 have similar C₂S phases, and therefore SK and SL5 have similar compressive strength. However, there is no evidence that higher compressive strength means higher flexural strength during aging [28]. The flexural strength of SK is higher than SL5. The higher flexural strength is good for improving the toughness of the protected materials. NHL2 has the smallest hydraulic phase and the lowest strength.

3.2.2. Shrinkage Rate

The smaller shrinkage rate is conducive to the grouting and reinforcement of stone relics so that the slurry and the rock can be closely conglutinated to achieve the ideal reinforcement effect [29]. Figure 5 shows the shrinkage rates of three hydraulic lime mortar blocks at 3, 7, 14, 21, and 28 days.

The shrinkage rates of the specimens are small for all hydraulic lime types. The shrinkage rates of SK and SL2 are between 0.15 and 0.25%, and that of SL5 is 0.11% at 28 days. Therefore, SL5 has a smaller shrinkage rate than SK and SL2. Although SK has a longer hydraulic phase, there is no calcium oxide after slaking. In addition to the hydration reaction in SK, moisture can volatilize and cause shrinkage of the sample. There is moisture consumption via calcium oxide and hydration reactions in SL2 and SL5, which prevents the shrinkage of the sample.

3.2.3. Relationship between the Elastic Wave Velocity and Age

From the elastic wave velocity age curve for three samples (Figure 6), it can be observed that the general tendency of the entire curves is primarily coincident for different specimens. The general tendency is that there is a decline, and afterward, continuous growth emerges. In detail, for SK, declines occur during 1–4 d, while SL2 and SL5 increase. This was attributed to SK interiors, which still contained liquid water during the initial stages and caused absorption and scattering attenuation of the elastic waves. With volatilization and water hydration in the specimens, the solid phases and crystal particles continued to grow. Afterward, the waves began to propagate through the solid-phase path, and the wave velocities continued to increase. SL2 and SL5 contain CaO and consume more water. Therefore, there will be no large volatilization of water in the early stages of the specimens, and there will be no decrease in wave velocity. It can also be observed that the wave velocity of SL2 and SL5 increased rapidly starting at 2 days.

3.3. SEM-EDS Analysis

Figure 7, Figure 8 and Figure 9 demonstrate the fracture micromorphology and EDS spot analysis of the hydraulic lime K3 after curing for 3, 7, and 28 days. At 3 days, flaky Ca(OH)₂ was still present, with some CaCO₃ and C-S-H forming. After curing for 7 days, Ca(OH)₂ reacted with H₂O and CO₂ to form CaCO₃, while the hydration reactions produced certain quantities of C-S-H. When cured for 28 days, more CaCO₃ heavy crystals were produced, and larger quantities of C-S-H formed inside the hydraulic lime. Thus, the mesh structures that formed between the particles became interwoven. The EDS pattern of the marked area is shown in Figure 7a, Figure 8a and Figure 9a and the corresponding data are shown in Figure 7b, Figure 8b and Figure 9b and

Table 3. The EDS data of red areas in SEM images.

Area	Element, Atomic (%)
	C	O	Al	Si	K	Ca
1	2.89	60.54	0.00	2.51	0.00	34.06
2	2.43	26.59	2.72	9.06	1.22	57.99
3	16.55	43.99	2.66	7.56	0.73	28.51

, respectively. From the analysis of the first spot, there is 60.54% O and 34.06% Ca, with very small amounts of C and Si, and the composition of the spot is Ca(OH)₂. The second spot contains 26.59% O, 9.06% Si, and 57.99% Ca, which, combined with the microstructure, shows that the spot is C-S-H. The third spot contains 16.55% C, 43.99% O, and 28.51% Ca, which is known to be CaCO₃ when combined with the microstructure. With prolonged curing time, the reactions continued, and CaCO₃ and C-S-H continued to grow; thus, the mechanical strength continued to increase.

3.4. Durability of the Specimens

3.4.1. Water Stability Test

Figure 10 shows the compressive and flexural strengths of three specimens before and after tests. It can be evidenced from the graph that for SK, SL2, and SL5, the compressive strength and flexural strength decreased after the test. The compressive strengths of the three smaller samples declined before and after immersion. The compressive strengths of SK, SL2, and SL5 were reduced by 10.55%, 19.67%, and 8.73%, respectively. The flexural strengths of SK and SL2 have a larger loss than SL5, but the flexural strength value of SK is larger than SL5. Of the three specimens, SK and SL5 possess the maximum resistivity to water compared to SL2. Additionally, the surfaces of the three specimens did not have any cracks after water immersion.

3.4.2. Soundness Test

Salt crystallization of sodium sulfate is an important factor in the deterioration of stone relics, which undoubtedly damages reinforcement material [30]. Figure 11 shows the compressive and flexural strengths before and after the test for SK, SL2, and SL5. The compressive strengths of SK and SL2 increased by 31.64% and 21.05%, respectively, after the test, while that of SL5 decreased by 2.06%. The flexural strength dramatically increased by 89.52% for SL2 and 53.48% for SL5 and decreased by 34.14% for SK. In addition, the surfaces of the three specimens did not have any cracks after the soundness test. The significant improvement of the compressive strength of SK is conducive to the protection and reinforcement of stone cultural relics. Overall, the three samples have good resistivity to sodium sulfate.

3.4.3. Alkali Resistivity Test

Figure 12 shows that the compressive and flexural strengths of SK and SL5 did not change much after the alkali test, whilst that of SL2 decreased slightly. In addition, the surfaces of the three specimens did not have any cracks before and after the test. As a result, the three samples have good resistance to alkali.

3.4.4. Frost–Thaw Cycle Test

A large number of stone relics are preserved in northwest China, a region that is colder in winter. Therefore, the frost–thaw cycle is a necessary test to detect the cold resistance of protective materials. Figure 13 shows the comparison of compressive and flexural strengths before and after tests for SK, SL2, and SL5. Results reveal that the compressive strengths of SK, SL2, and SL5 decreased by 1.86, 11.08, and 0.97%, respectively, after the frost–thaw cycle. Meanwhile, the reduction in the flexural strength of SK is 8.00%, and those of SL2 and SL5 did not change. In addition, the surfaces of the three specimens did not have any cracks. It was noticed that the compressive strengths of SK and SL5 were reduced slightly, which can guarantee reinforcement performance. Overall, the three samples have good resistance to the frost–thaw cycle.

3.4.5. Temperature and Humidity Cycle Test

The temperature and humidity changes in the environment can have an important impact on the preservation of cultural relics. Therefore, an accelerated aging test temperature and humidity cycle can effectively evaluate HL and NHL durability. Figure 14 shows the changes in strengths between before and after tests for SK, SL2 and SL5. Results reveal that the compressive strengths of SK, SL2, and SL5 decreased by 3.60, 25.21, and 3.27% after the test, respectively, meanwhile, flexural strengths decrease respectively by 13.54, 43.55, and 30.43%. SK and SL5 had smaller strength loss than SL2. Furthermore, the surfaces of the three specimens did not have any cracks. It should be mentioned that despite the strength reduction, the loss rate of compressive strength is relatively small, which can still guarantee reinforcement performance. In all, SK and SL5 have good resistance to temperature and humidity cycles.

3.4.6. Discussion

Hydraulic lime contains air-hardening and hydraulic phases. When hydraulic lime is used as lime mortar blocks, the initial reaction that occurs is the hydration of C₂S to enhance the early strength of the material. As the hydration reaction continues, the products of the hydration reaction, C-S-H and Ca(OH)₂, undergo a carbonation reaction to produce CaCO₃, which further enhances the strength of the lime mortar. Its formula was shown in Equations (1)–(3):

$$2\mathrm{CaO}·{\mathrm{SiO}}_2+{\mathrm{nH}}_2\mathrm{O}=\mathrm{xCaO}·{\mathrm{SiO}}_2·{\mathrm{yH}}_2\mathrm{O}+\left(2-\mathrm{x}\right)\mathrm{Ca}{(\mathrm{OH})}_2$$

(1)

$$\mathrm{CaO}·{\mathrm{SiO}}_2·{\mathrm{yH}}_2\mathrm{O}+{\mathrm{CO}}_2={\mathrm{CaCO}}_3+{\mathrm{SiO}}_2·{\mathrm{yH}}_2\mathrm{O}$$

(2)

$$\mathrm{Ca}{\left(\mathrm{OH}\right)}_2+{\mathrm{CO}}_2={\mathrm{CaCO}}_3+{\mathrm{H}}_2\mathrm{O}$$

(3)

According to the above analysis, NHL2 contains more lime and NHL5 and K3 contain more hydraulic phases. When the specimens were cured for 56d, a large amount of C-S-H and calcium carbonate formed inside SL5 and SK, which had higher strength and could better resist the erosion of the external environment. Due to the higher lime content in NHL2 and the lack of a substantial amount of CO₂ in the curing environment, there is still more Ca(OH)₂ present inside the NHL2 mortar. During water stability, temperature and humidity cycle, and frost–thaw cycle tests, the strength loss of NHL2 mortar is significantly higher than that of NHL5 and K3 mortars due to water erosion. In soundness and alkali resistivity tests, Na₂SO₄ and NaOH crystals may fill the inside of the lime mortar to increase the strength [31]. However, the three mortars NHL2, NHL5, and K3 did not show obvious regularity, which may also be related to the pore space of each mortar and needs further study.

4. Conclusions

Hydraulic lime was produced with limestone and potassium feldspar at 1000 °C and calcined for 3 h. The main components of hydraulic lime contain the hydraulic phases C₂S and C₂AS and the air-hardening phases CaO and Ca(OH)₂. As a compromise, when potassium feldspar accounted for 30% of the total mass, HL contained a more suitable air-hardening component and hydraulic component. SEM-EDS analyses show that the phases of calcium carbonate and hydrated calcium silicate gradually increased with prolonged curing time for HL.

The shrinkage of the K3 mortar is larger than that of NHL, which is related to its premature slaking. The strength of HL gradually increased with prolonged curing time, and on the 56th day, the compressive strength value of the K3 mortar reached the NHL5 mortar, and the flexural strength was higher than that of NHL5.

Accelerated aging tests indicated that K3 mortar and NHL5 mortar led to higher resistance to fluctuation in the temperature and humidity, frost–thaw cycles, and the water environment than the NHL2 mortar. Meanwhile, the K3 mortar also showed good salt resistance and sulfate resistance. In addition, the surfaces of all specimens did not have any cracks between before and after tests. Therefore, it can partly be concluded that the K3 mortar can resist harsh conditions.

In this paper, the strength, shrinkage, and durability of HL were studied without discussing the compatibility between HL and stone relics. The compatibility of conservation materials will be investigated further in future studies.