The Properties of Magnesium Silicate Hydrate Prepared from the Magnesium Silicate Minerals in the Earth’s Crust

Abstract

In order to explore a wider range and lower cost of raw materials for the preparation of magnesium silicate hydrate (M-S-H), an acid-leaching method was employed to extract and separate high-purity magnesium hydroxide (Mg(OH)₂) with a purity higher than 97% and amorphous silica with a purity higher than 90% from four types of natural silicate minerals (serpentine, peridotite, zeolite, and montmorillonite). These two intermediate products, which are amorphous silica and magnesium hydroxide, were used to prepare M-S-H, and the influence of curing at two temperatures, 50 °C and 80 °C, on the properties of M-S-H was investigated. The results showed that with the increase in curing temperature, the bound water content, tetrahedral polymerization degree, and Mg(OH)₂ content increased. There was a good correlation between the increase in strength and the bound water content of M-S-H. This work provides a possible technological route for expanding the raw materials for preparing magnesium silicate hydrate cementitious materials and utilizing the abundant magnesium silicate minerals in the Earth’s crust.

1. Introduction

Hydrated magnesium silicate (M-S-H) cementitious material is considered to be a low-carbon cementitious material [1] and is typically prepared by reaction of MgO or Mg(OH)₂ with amorphous, reactive silica in the presence of water [2]. Mg(OH)₂ and MgO are usually obtained from carbonate rocks such as magnesite and dolomite or Mg²⁺ in salt lakes and seawater. These methods result in direct or indirect CO₂ emissions. Silica fume is the most common source of Si in the solution reaction method for preparing M-S-H [3,4,5,6]. In addition, amorphous SiO₂ present in rice husk ash [7], fly ash [8], fluidized bed bottom ash [9], finely ground waste glass, and diatomaceous earth [10] can also serve as reactants for the formation of M-S-H. Amorphous SiO₂ and silicic acid [11,12] exhibit higher reactivity compared to crystalline SiO₂. As a supplementary cementitious material, the high cost and specific locality of silica fume limit its use as a cementitious material.

The molar ratio of magnesium to calcium (Mg:Ca) in the continental crust is approximately 1.06, as reported by Hans [13]. This brings to light an intriguing question: is it feasible to utilize naturally occurring magnesium silicate minerals for the synthesis of hydrated magnesium silicate cementitious materials? The potential use of natural magnesium silicate minerals as raw materials in the production of such cementitious materials offers a dual advantage. First, it presents an opportunity to significantly reduce carbon dioxide emissions, as this process circumvents the need for calcining magnesium carbonate. Second, it proposes a cost-effective route for manufacturing hydrated magnesium silicate, thus making it an attractive alternative to traditional cementitious materials.

Ruiter’s research delineates the serpentinization process of olivine-rich rocks, which leads to the formation of brucite. This brucite subsequently dissolves, yielding a magnesium-rich fluid with a notably high pH level (exceeding 9). This alkaline fluid facilitates the dissolution of quartz, culminating in the emergence of a nanocrystalline magnesium silicate hydrate phase [14]. Among magnesium silicate rocks such as serpentine, olivine, sepiolite, and chlorite, there is a notable absence of cementing properties. But, after undergoing dehydration, natural serpentine demonstrates commendable cementitious qualities upon rehydration within autoclave conditions, achieving a compressive strength of up to 17 Mpa [15]. Furthermore, within acidic environments, the trio of minerals—serpentinite, leaf serpentinite, and fiber serpentinite—are capable of dissolving Mg²⁺ and amorphous SiO₂. The purpose of this study is to use natural magnesium silicate minerals as raw materials to prepare hydrated magnesium silicate gelling materials. The characteristics of the magnesium silicate hydrate (M-S-H) derived from isolated Mg(OH)₂ and amorphous SiO₂ were meticulously analyzed through the dissolution of serpentine, peridotite, sepiolite, and chlorite in sulfuric acid. It provides a new path for preparing hydrated magnesium silicate cementing materials.

2. Materials and Methods

2.1. Materials

Serpentine, peridotite, sepiolite, and chlorite are derived from Shanlin Shiyu Mineral Products Co., Ltd., Guzhang County, Xiangxi Prefecture, Hunan Province. The chemical composition is shown in

Table 1. Chemical composition of raw material wt%.

Materials	MgO	SiO2	Al2O3	Fe2O3	CaO	K2O	Na2O	TiO2
Serpentine Olivine	43.48 47.89	45.95 41.77	0.62 0.23	9.11 9.97	0.37 0.08	0.02 0.01	0.11 0.04	0.02 0.01
Sepiolite	28.47	67.45	2.03	0.49	0.44	0.65	0.28	0.09
Chlorite	34.71	42.12	12.70	6.96	3.12	0.12	0.05	0.22

, the X-ray diffraction is shown in Figure 1, and the particle size distribution of the raw materials is shown in Figure 2. Serpentine consists mainly of serpentine group minerals such as Elisserite, fibrous serpentine and leaf serpentine, and plagiochlorite and contains a small amount of magnetite. Peridotite is mainly composed of magnesium olivine. Chlorite is mainly composed of plagiochlorite and brittle chlorite. The silica and magnesium contents of the four minerals are high, which provide rich silicon and magnesium sources for the preparation of hydrated magnesium silicate cementing materials. The particle size distribution analysis of raw materials (as shown in Figure 2) indicates that the mean particle size of serpentine, olivine, sepiolite, and chlorite is 70 μm, 80 μm, 80 μm, and 70 μm, respectively.

2.2. Methods

The schematic diagram of the experiment is shown in Figure 3. The raw materials (serpentine, olivine, sepiolite, and chlorite) were dissolved in a 30% H₂SO₄ solution with a solid–liquid ratio of 1:5 and heated at 90 °C for 3 h. After dissolution, the solution was separated from the residue using a centrifuge. The residue was washed five times with deionized water and then dried in an oven at 50 °C until a constant weight was reached. The dried residue was sealed and stored for later use.

To obtain precipitation, NaOH solution was added dropwise to the separated filtrate using a Pasteur pipette. The resulting precipitation was clarified, separated, and dried to obtain Mg(OH)₂. Since the newly formed Mg(OH)₂ is more conducive to the formation of M-S-H [16], in this experiment, Mg(OH)₂ was calcined to produce MgO, which served as the magnesium source for the preparation of M-S-H. The dried precipitated sample was calcined at 350 °C in a muffle furnace for 1 h, resulting in white MgO powder.

The resulting residue was then mixed evenly with MgO powder in a mass ratio of 6:4, and a paste was prepared using deionized water at a water-to-solid ratio of 1:2. After sealing, the paste was cured in a constant temperature water bath at 50 °C and 80 °C for 3 days, 7 days, 14 days, and 28 days, respectively. Subsequently, the cured samples were dried in a vacuum drying oven at 40 °C for 24 h.

The mineral phase analysis was undertaken using an XRD (D8, Bruker, Berlin, Germany) with CuKα radiation operating at 40 kV/30 mA and scanning the range from 5 to 80 at a speed of 3°/min. Particle size distribution (PSD) was determined by using a laser particle size analyzer. Thermogravimetric analysis (TGA) was performed by using a METTLER (Bern, Switzerland) Toledo TGA/DSC1/1600 thermal analysis system with a temperature accuracy of +/−0.5 °C. Fourier transform infrared (FTIR) spectra were obtained by using a spectrometer (Nicolet IS50, Thermo Fisher, Madison, WI, USA). The FTIR analysis had a resolution of 4 cm⁻¹ and co-addition of 32 scans. The magic-angle rotating solid high-resolution NMR experiments were carried out using an AVANCE400 (SB) fully digital NMR spectrometer of Bruker (Bern, Switzerland), and a solid 15 N~31 P probe of 4 mm/15 kHz.

The strength of the mortar samples was tested according to the GB/T 17671-1999 [17]. The mass ratio of the residue to MgO in the cementing material was 6:4, with 0.2% polycarboxylate superplasticizer. The water–cement ratio for the mortar samples was 0.5. The specimen size was 40 mm × 40 mm × 160 mm, and the specimens were cured in a water bath at 50 °C and 80 °C for the specific age to measure compressive strength.

3. Results and Discussion

3.1. Composition and Particle Size Distribution of the Residue and Mg(OH)₂

The chemical composition of the residue and Mg(OH)₂ are shown in

Table 2. Results of chemical analysis of residue and Mg(OH)₂ wt%.

Materials		SiO2	MgO	SO3	Al2O3	Fe2O3	CaO	K2O	Na2O	TiO2
Residue	Serpentine Olivine	94.68 95.74	0.78 1.35	3.97 1.33	0.09 0.25	0.34 0.92	0.06 0.16	0.01 0.02	0.05 0.08	0.02 0.15
	Sepiolite	95.69	1.72	1.45	0.65	0.20	0.15	/	/	0.14
	Chlorite	90.87	5.44	0.88	/	2.61	/	/	/	0.20
Mg(OH)2	Serpentine Olivine	0.52 2.03	98.22 96.82	0.52 0.32	0.06 0.21	0.64 0.19	/ 0.12	/ 0.01	0.04 0.30	/ 0.01
	Sepiolite	0.38	98.69	0.05	0.60	/	0.08	/	0.20	/
	Chlorite	1.94	97.12	0.24	0.33	0.24	/	0.02	0.11	/

. The XRD patterns of the residue, Mg(OH)₂ and MgO calcined, are shown in Figure 4 and Figure 5. In Table 2, the chemical composition of the residue is mainly SiO₂, a minor of SO₃ introduced from sulfuric acid, and other components. The XRD pattern reveals diffraction peaks of amorphous SiO₂ in the range of 2θ = 15° to 30°, while no distinct crystal diffraction peaks are observed for residue dissolved from serpentine, olivine, and sepiolite.

Based on Figure 4 and Figure 5 and Table 2, the obtained Mg(OH)₂ and the MgO derived from calcination at 350 °C exhibit relatively high purity. The particle size distribution analysis of the residue (amorphous silica) and Mg(OH)₂ precipitation (as shown in Figure 6) indicates that the mean particle size of the amorphous SiO₂ prepared from serpentine is approximately 20 μm, while that from olivine and sepiolite is around 30 μm. For chlorite, the mean particle size is approximately 65 μm. Similarly, the maximum particle size of Mg(OH)₂ precipitated from the four types of minerals is approximately 65 μm.

3.2. XRD and FTIR Analysis

Figure 7, Figure 8, Figure 9 and Figure 10 are XRD patterns of the paste of the residue and MgO in 50 °C and 80 °C water. In the Figure 7, Figure 8, Figure 9 and Figure 10, there is no distinct MgO diffraction peak, indicating that most of MgO has hydrated after 3 d. The characteristic peaks of M-S-H at the 2θ value of 5.0°–10.0°, 35.0°, and 59.9° can be observed [18,19] after 3 d. With the hydration time and the curing temperature, the height of these diffraction peaks gradually increases, while the corresponding Mg(OH)₂ diffraction peaks decrease. The mineral diffraction peaks observed around 2θ at 35° in the samples prepared from chlorite may be attributed to minerals such as augite-Ca (PDF No. 24-0201) present in the chlorite raw materials.

Figure 11, Figure 12, Figure 13 and Figure 14 show the FTIR spectra of the residue and MgO paste at different reaction times under curing at 50 °C and 80 °C. The two absorption bands correspond to the formation of M-S-H. Specifically, the absorption band within the range of 950 to 1100 cm⁻¹ is attributed to the Q³ Si-O vibration, while the absorption band within the range of 870 to 920 cm⁻¹ is attributed to the Q² Si-O vibration [5,19]. Additionally, the absorption peak observed near the wavenumber of 650 cm⁻¹ can be attributed to the bending vibration of Si-O-Si [20]. With an increase in curing time, the intensities of these three absorption bands are significantly enhanced, indicating a gradual increase in the formation of hydration products. The absorption band at the wavenumber of 1080 cm⁻¹ assigned the vibration of Si-O-Mg in M-S-H [7,21,22] also increases with age. These results are consistent with the XRD results.

3.3. TG and DSC Analysis

The mass loss between 50 °C and 300 °C (Δm1) in the TG curve indicates the removal of interlayer water from M-S-H, and the weight loss between 300 °C and 750 °C (Δm3) relates to the dehydroxylation of magnesium hydroxide and the decomposition of M-S-H (Figure 15, Figure 16, Figure 17 and Figure 18,

Table 3. The mass loss between different temperature wt%.

Curing Tem.	Mass Loss	Minerals	3 d	7 d	14 d	28 d
50 °C	Δm1	Serpentine	11.34	14.55	14.07	14.07
		Olivine	13.92	15.25	14.43	15.09
		Sepiolite	15.21	18.02	14.56	17.43
		Chlorite	11.32	9.34	10.61	11.00
	Δm2	Serpentine	13.07	10.70	9.16	9.02
		Olivine	11.75	10.72	8.87	9.33
		Sepiolite	11.41	8.28	9.10	6.02
		Chlorite	12.58	11.56	10.14	10.14
	Δm3	Serpentine	18.44	16.25	15.28	15.35
		Olivine	17.36	16.44	15.43	16.32
		Sepiolite	16.61	13.97	15.18	13.32
		Chlorite	18.28	17.98	16.28	15.94
	Δm3 − Δm2	Serpentine	5.37	5.56	6.13	6.34
		Olivine	5.61	5.72	6.56	6.99
		Sepiolite	5.20	5.69	6.08	7.30
		Chlorite	5.70	6.42	6.14	5.80
80 °C	Δm1	Serpentine	14.63	19.18	18.49	18.17
		Olivine	13.92	14.30	14.15	18.73
		Sepiolite	14.64	19.43	18.70	18.70
		Chlorite	14.48	12.46	12.09	13.99
	Δm2	Serpentine	8.74	6.54	5.29	4.62
		Olivine	11.93	11.23	10.25	5.83
		Sepiolite	10.96	6.16	6.62	6.23
		Chlorite	8.60	7.60	6.60	5.89
	Δm3	Serpentine	15.02	12.96	11.84	11.42
		Olivine	17.36	17.71	17.54	13.08
		Sepiolite	16.60	12.75	13.60	13.12
		Chlorite	15.13	13.60	13.00	12.50
	Δm3 − Δm2	Serpentine	6.28	6.42	6.57	6.80
		Olivine	5.43	6.48	7.29	7.25
		Sepiolite	5.64	6.59	6.98	6.89
		Chlorite	6.53	6.00	6.40	6.61

). Notably, this part of the mass loss includes the removal of the dehydroxylation of brucite at about 400 °C (Δm2) [20,23]. With the increase in curing time, the water content between the layers of M-S-H gradually increases, reaching its maximum value at 7 days, followed by a decrease. The M-S-H phase itself acts as a gel-like material that can retain water within its structure; the continuous formation of it leads the increase in the interlayer water. But as the reaction progresses, the M-S-H phase becomes more densely packed, limiting the amount of water that can be accommodated in the interlayer spaces.

Upon curing at 50 °C for 3 days, the mass loss due to Mg(OH)₂ dehydroxylation (Δm2) reaches maximum value, gradually decreasing thereafter. Furthermore, at 28 days, the Mg(OH)₂ content is only about 70–80% of that of the four samples cured in 50 °C for 3 days. A higher-temperature curing accelerates the hydration reaction. For samples cured at 80 °C, this ratio decreases to 48–68%.

Compared to the 3-day sample, the 28-day sample exhibited an increase in the mass loss (Δm3 − Δm2) of 118% to 140%. This means that the hydration rate decreased over time. A more compact structure, creating barriers for water transportation and decreasing the availability of reactants, hinders further hydration.

The DSC curves of M-S-H pastes prepared with serpentine, olivine, sepiolite, and chlorite are presented in Figure 19, Figure 20, Figure 21 and Figure 22. On the DSC curve, the endothermic valley between 50 and 300 °C is caused by the removal of interlayer water from M-S-H. The second endothermic valley near 300~400 °C corresponds to the dehydroxylation of Mg(OH)₂. The heat release peak at about 850 °C is the heat release during the M-S-H recrystallization to form SiO₂ and enstatite (MgSiO₃) [6,24] or SiO₂ and Mg₂SiO₄ [25]. The heat release peak area of this region is calculated and illustrated in the figures. The heat release of the sample cured at 80 °C is greater than that of the sample cured at 50 °C; that is, the amount of M-S-H formed is higher at 80 °C. Furthermore, with the extension of the curing time, the heat release gradually increases, reaching its maximum at 14 days and slightly decreasing at 28 days (except for the sample prepared with olivine).

In Figure 19, Figure 20, Figure 21 and Figure 22, an interesting observation can be made regarding the heat release peak at 850 °C. Initially, this peak exhibits a single peak or a double peak pattern at early age, but with extended curing time, it transitions into a single peak or a weakening of the double peaks (except for the sample prepared with olivine). This change is considered to be an indicative manifestation of M-S-H maturity [26].

3.4. NMR Analysis

Figure 23 displays the ²⁹Si NMR spectra of the amorphous silica and M-S-H samples cured at 50 °C and 80 °C. The deconvolution results for these spectra at 28 d are summarized in

Table 4. Relative concentrations of different silicon sites of MSH pastes obtained from deconvolution demonstration of ²⁹Si SSNMR spectra.

Curing Tem.	Minerals	I/%					Unreacted Silica I/%		Q3/Q2
		Q1a	Q1b	Q2	Q3a	Q3b	Q3	Q4
SiO2	Serpentine						31.4	68.6
	Olivine						31.8	68.2
	Sepiolite						30.9	69.1
	Chlorite						32.4	67.6
50 °C	Serpentine	4.6	7.8	41.1	32.6	10.1	3.8		1.04
	Olivine	5.8	7.3	43.8	35.6	7.5	0		0.98
	Sepiolite	1.2	4.2	36.1	43.7	5.1	9.7		1.35
	Chlorite	0	23.8	40.7	18.9	14.7	1.9		0.83
80 °C	Serpentine	4.3	7.1	43.6	35.9	7.9	1.2		1.00
	Olivine	4.1	6.1	46.4	37.1	6.3	0		0.94
	Sepiolite	0.7	5.4	38.9	45.9	5.2	3.9		1.31
	Chlorite	0	24.4	43.4	20.2	11.6	0.4		0.73

. According to references [27,28,29], the chemical environment of ²⁹Si is denoted by Qⁿ, where n indicates the number of bridging oxygen atoms shared between a silicon tetrahedron and the surrounding silicon tetrahedra. In the spectrum of amorphous silica, the chemical shift at −112 cm⁻¹ corresponds to the Q⁴ signal, while −102 cm⁻¹ represents the Q³ signal [30]. The ²⁹Si MAS NMR spectra of amorphous silica obtained from the dissolution of four silicate minerals show no significant distinction for the aforementioned Q³ and Q⁴ signals. It is noted that the Q⁴ signal in the M-S-H sample nearly vanishes after 28 days of curing. Furthermore, after 28 days of curing at 50 °C, the remaining Q³ (−102 cm⁻¹) signal is slightly higher than that which underwent curing at 80 °C. From the relative content of the remaining Q³ (−102 cm⁻¹) signal, a higher degree of hydration occurs at elevated curing temperatures. Due to the highest residual Q³ content in amorphous silica, the M-S-H prepared from sepiolite has the lowest degree of hydration. Conversely, the M-S-H prepared using olivine as the raw material has the highest degree of hydration.

The ²⁹Si MAS NMR spectrum of M-S-H is primarily composed of chain end-group Q¹, branched silicon–oxygen tetrahedron Q², and layered silicon–oxygen tetrahedron Q³ [28,31]. According to Bernard [3,32], the chemical shifts at ∼−78.3 ppm, −80 ppm, −85.5 ppm, −92.7 ppm, −94.7 ppm, and −96.7 ppm represent Q¹_a, Q¹_b, Q², Q³_a, Q³_b, and Q³_c, respectively. The relative content of Q¹ in the samples cured at 80 °C is lower compared to that in the samples cured at 50 °C, indicating that high-temperature curing enhances the polymerization of silicate tetrahedra in M-S-H. As can be seen from the spectral peak shapes and Table 4, the formation rate of the Q³ structure in M-S-H prepared with serpentine, olivine, and sepiolite exceeds that of the Q² structure within the first 28 days. Unlike these three minerals, the M-S-H prepared from chlorite has a higher relative content of Q² than of Q³. This indicates that the M-S-H prepared from chlorite has the lowest degree of polymerization of silicate tetrahedra. In all samples, the relative amount of Q³_b, which represents the talc-like T:O:T structure with a chemical shift at −94.7 ppm, is lower than that of Q³_a, which represents the chrysotile-like T:O structure. This indicates that the silicate tetrahedral structure of the M-S-H prepared using these four silicates is closer to the chrysotile-like T:O structure.

Owing to the characteristic layered structure of the M-S-H samples, the M-S-H derived from sepiolite demonstrated a greater relative Q³/Q² concentration compared to the M-S-H derived from other sources. This indicates a higher degree of tetrahedral polymerization when sepiolite is used as the raw material. Compared to other minerals, sepiolite has the highest content of residual unreacted silica, resulting in the highest Mg/Si ratio in the system. A higher Mg/Si ratio leads to an increase in the Q³/Q² of the produced M-S-H.

3.5. Compressive Strength Analysis

The compressive strength of hydrated magnesium silicate paste is shown in Figure 24. The figure clearly demonstrates that the compressive strength of M-S-H gradually increases with the curing time, surpassing 38 MPa at 28 days. The M-S-H specimens cured at 80 °C exhibit higher strength compared to those cured at 50 °C. However, the strength ratio of specimens cured at 80 °C to those at 50 °C gradually decreases, indicating that curing temperature has a more significant impact on the early compressive strength of the specimens. With an increase in curing time, the strength difference caused by environmental temperature gradually diminishes.

Analyzing the relationship between compressive strength and mass loss at different temperature ranges in Figure 25, it is observed that the mass loss (Δm1) between 50 °C and 300 °C shows the weakest correlation with strength. This is because the water evaporated during this temperature range includes not only interlayer water from M-S-H but also a certain amount of free water or adsorbed water. On the other hand, the removal of hydroxyl groups (Δm2) from brucite, which acts as a reactant in M-S-H formation, exhibits a strong negative correlation with M-S-H strength. Subtracting the mass loss of brucite at approximately 400 °C (Δm2) from the mass loss between 300 °C and 750 °C (Δm3) provides the remaining mass loss, which represents the decomposition of M-S-H. The mass loss demonstrates the strongest correlation with strength.

4. Conclusions

In this study, four types of natural magnesium silicate minerals were used as raw materials. Through the acid-leaching method, high-purity amorphous silica and magnesium hydroxide were obtained. Using these intermediates, magnesium silicate hydrate cementitious materials were prepared. The main conclusions drawn are as follows:

(1)

The amorphous silica obtained from the acid leaching of serpentine, olivine, sepiolite, and chlorite reach a purity of over 90%. The silica derived from serpentine, olivine, and sepiolite has an average particle size of about 20 μm, while the silica from acid-leached chlorite has an average particle size of about 65 μm.

(2)

Amorphous silica and magnesium oxide prepared using magnesium hydroxide could form magnesium silicate hydrate within 3 days of hydration. Although there are certain numerical differences between the minerals used as raw materials, in general, the hydration reaction accelerates by 10–20% with increased curing time and temperature, and the degree of polymerization of the silicate tetrahedra of newly formed M-S-H increases.

(3)

The M-S-H prepared using four types of magnesium silicate minerals forms a silicate tetrahedral structure predominantly characterized by the chrysotile-like T:O structure. The resulting magnesium silicate hydrate cementitious materials could achieve a compressive strength of 40 MPa at 28 days, and the compressive strength shows a good correlation with the reduction in magnesium hydroxide content and the increase in magnesium silicate hydrate.