Critical Review of Nanomechanical Properties of Calcium (Alumino) Silicate Hydrates: Test Methods, Influencing Factors, and Enhancing Strategies

Jiawei Wang , Stuart McElhany , Zhangli Hu , Jiaping Liu , Carlo Carraro , Paulo J.M. Monteiro , Roya Maboudian

Engineering ›› 2026, Vol. 64 ›› Issue (9) : 224 -243.

PDF (5536KB)
Engineering ›› 2026, Vol. 64 ›› Issue (9) :224 -243. DOI: 10.1016/j.eng.2025.11.015
Research
review-article
Critical Review of Nanomechanical Properties of Calcium (Alumino) Silicate Hydrates: Test Methods, Influencing Factors, and Enhancing Strategies
Author information +
History +
PDF (5536KB)

Abstract

The mechanical properties of calcium (alumino) silicate hydrates (C-(A)-S-H) represent a critical focus within the cement and concrete industry. This review begins by summarizing particle- and subparticle-scale models of C-(A)-S-H. Building on these models, the effects of chemical composition and microstructure on the intrinsic mechanical properties of C-(A)-S-H, as determined by high-pressure X-ray diffraction, are described. Existing studies have demonstrated that increasing the Ca/Si ratio and Al incorporation enhances the intrinsic mechanical properties of C-(A)-S-H. Advanced techniques, such as high-pressure Raman spectroscopy and synchrotron radiation-based techniques, have been employed to elucidate the origins of intralayer sliding and preferred intragranular orientation, offering insights into the fundamental mechanisms of creep. Compared with intralayer sliding, the preferred intragranular orientation of cement-based materials significantly contributes to creep in C-(A)-S-H, establishing a direct link to macroscopic creep behavior. Based on these findings, this review summarizes several “bottom-up” strategies for strengthening and toughening C-(A)-S-H.

Graphical abstract

Keywords

Nanomechanical properties / Preferred orientation / Sliding / Enhancing strategies

Cite this article

Download citation ▾
Jiawei Wang, Stuart McElhany, Zhangli Hu, Jiaping Liu, Carlo Carraro, Paulo J.M. Monteiro, Roya Maboudian. Critical Review of Nanomechanical Properties of Calcium (Alumino) Silicate Hydrates: Test Methods, Influencing Factors, and Enhancing Strategies. Engineering, 2026, 64 (9) : 224-243 DOI:10.1016/j.eng.2025.11.015

登录浏览全文

4963

注册一个新账户 忘记密码

1. Introduction

Calcium (alumino) silicate hydrate (C-(A)-S-H) is among the most important hydration products of cement-based materials, accounting for approximately 70% of hydration products by volume. It is essential in enhancing various properties, including strength and deformation, of cement-based materials [[1], [2], [3], [4]]. Therefore, elucidating the relationship between its physicochemical characteristics and intrinsic mechanical performance, defined as mechanical properties measured independently of porosity, at the fundamental level remains both highly significant and challenging [[5], [6], [7]]. Achieving this understanding offers significant benefits for advancing cement and concrete science. However, owing to the porous and poorly crystalline, colloidal nature of C-(A)-S-H, the applicability of conventional methods for evaluating its mechanical properties is limited. Nanoindentation and atomic force microscopy (AFM) are essential techniques for evaluating the micromechanical properties of C-(A)-S-H, each with distinct strengths and limitations [8]. Nanoindentation offers reliable and reproducible measurements of modulus, hardness, and creep; however, it is influenced by phase interactions and subjective deconvolution processes [[9], [10], [11], [12]]. AFM excels at nanoscale precision by capturing particle interactions with minimal phase interference; however, its results are sensitive to environmental factors, such as humidity, surface roughness, and carbonation [[13], [14], [15], [16]]. Although nanoindentation is ideal for bulk assessments, AFM provides detailed insights at the nanoscale, with each technique complementing the other for comprehensive material characterization. However, these commercial test technologies are sensitive to pores and local packing density, limiting their ability to reveal the intrinsic mechanical properties of C-(A)-S-H. With the development and feasibility of synchrotron radiation technology and high-pressure (HP) science, significant advances have been made in elucidating the intrinsic mechanical properties of C-(A)-S-H [[17], [18], [19], [20], [21], [22], [23]]. Although inherently based on ensemble-averaged measurements, HP diffraction techniques can statistically capture the intrinsic mechanical response of billions of unit cells across a representative volume of nanocrystallites. This approach differs fundamentally from macroscale mechanical tests, wherein localized stress variations occur and replicate measurements are often required [[23], [24], [25], [26]]. In recent years, the “genetic” factors of C-(A)-S-H, such as the Ca/Si and Al/Si ratios, and their influence on its intrinsic mechanical properties have been extensively investigated [[27], [28], [29], [30], [31], [32], [33], [34], [35], [36]]. The key structural parameters, including main chain length, basal spacing, and interlayer and intralayer structures, of C-(A)-S-H have been explored to understand their impact on mechanical performance [[37], [38], [39], [40]]. However, the absence of a systematic organization and synthesis of these advanced findings limits the understanding of the underlying multiscale mechanisms.

Most of the findings summarized in this review are based on synthetic C-(A)-S-H, rather than on C-(A)-S-H directly extracted from hydrated cement paste. Although synthetic and hydrated C-(A)-S-H have similar chemical compositions and atomic structures, their morphologies are notably different. For example, synthetic C-(A)-S-H typically exhibits a foil-like structure, whereas hydrated C-(A)-S-H exhibits both foil-like and fibrillar morphologies [41]. However, the key compositional parameters, such as Ca/Si and Al/Si ratios, of synthetic C-(A)-S-H closely resemble those of hydrated C-(A)-S-H. For instance, synthetic C-(A)-S-H can achieve Ca/Si ratios up to 1.6, which are comparable to those observed in hydrated systems [42]. Similarly, the Al/Si ratios in synthetic samples can reach approximately 0.2 [43], even when incorporating the formation of other unavoidable Al-bearing phases, such as katoite and strätlingite [44]. Furthermore, the interlayer spacing of synthetic C-(A)-S-H is consistent with naturally occurring tobermorite phases (e.g., 14 and 11 Å) [45]. Employing elevated-temperature synthesis (e.g., 80 °C), cross-linked structures involving Q3 Al species (referring to aluminum atoms coordinated to three bridging oxygen atoms) that serve as “columns” between layers can be induced, enhancing structural complexity and mechanical performance. Considering the challenge in isolating pure C-(A)-S-H from hydrated cement paste, synthetic C-(A)-S-H remains the most representative and practical model available for mechanistic studies. Nevertheless, particular attention should be paid to the Ca/Si and Al/Si ratios in C-(A)-S-H solids; these ratios are influenced by both the solution environment and equilibrium phase assemblage. Accurate interpretation of composition often requires back-calculation based on these factors. In addition, the drying method used during the synthesis can lead to variations in the initial water content, subsequently affecting its microstructure and mechanical properties. These considerations are critical when translating findings from synthetic models to real hydrated systems.

Creep in C-(A)-S-H is a critical factor influencing the deformation and long-term durability of concrete structures. However, the exact location and mechanism of creep within C-(A)-S-H remains debatable [46]. Some studies have suggested that creep occurs via interlayer sliding, where adjacent C-(A)-S-H sheets shift relative to each other, with the interlayer water acting as a lubricant [47,48]. Others have reported that creep occurs in intergranular regions, with stress concentrated in areas of higher solubility, leading to dissolution at particle boundaries, followed by diffusion and reprecipitation in lower-stress areas [49]. Another theory suggested that creep results from the reorganization of gel-like C-(A)-S-H particles, potentially triggered by drying and stress-induced recrystallization [50]. Additionally, some studies argue that intraparticle water redistribution contributes to deformation, as water moves between different pore spaces within the C-(A)-S-H matrix to relieve stress [51,52]. Although these theories highlight the complexity of the behavior of C-(A)-S-H, no consensus has been reached regarding the exact location of creep, necessitating further research using advanced characterization techniques.

A fundamental understanding of C-(A)-S-H deformation is essential for strengthening and toughening cement-based materials using a “bottom-up” approach [[53], [54], [55], [56], [57], [58], [59], [60]]. Although numerous studies have employed advanced materials such as polyvinyl alcohol [61,62], graphene [63,64], and carbon nanotubes [65,66] to enhance the mechanical properties of C-(A)-S-H, the lack of insight into the underlying mechanisms has hindered a comprehensive interpretation of these results. However, the insufficient knowledge of the fundamental mechanisms remains a major barrier to unlocking the full potential of these innovations.

Herein, we systematically review the multiscale models of C-(A)-S-H. This review summarizes recent advancements in the studies on C-(A)-S-H, focusing on its intrinsic mechanical properties, as influenced by varying chemical compositions and microstructures, via synchrotron radiation and HP experimental techniques. The key findings of these studies reveal the deformation mechanisms of C-(A)-S-H across multiple scales, offering a comprehensive understanding of its nanoscale behavior and its connection to macroscopic creep performance. Finally, we propose practical strategies for strengthening and toughening cement-based materials, focusing on reinforcement along different axes.

2. C-(A)-S-H structure

As shown in Fig. 1 [67], concrete is a multiscale material composed of approximately 10%-15% binder, 15%-20% water, and 65%-75% aggregates by volume. Upon mixing with water, the binder, typically a combination of Portland cement and supplementary cementitious materials (SCMs), undergoes hydration to form a cohesive gel-like cement paste that binds the aggregates and contributes to mechanical strength. Within this paste, the poorly crystalline C-(A)-S-H is the primary hydration product, accounting for approximately 70% of the total volume. At the nanoscale, C-(A)-S-H exhibits a layered sandwich structure comprising calcium oxide sheets interleaved with silicate chains. Between the atomic-level structure and micrometer-scale morphology, the Jennings colloidal model introduces an intermediate framework, where C-(A)-S-H forms 5-6 nm diameter globules that aggregate into a porous network [[68], [69], [70]]. These globules yield two distinct domains: low-density C-S-H (LD-C-S-H), characterized by loosely packed particles with higher porosity and water content, and high-density C-S-H (HD-C-S-H), featuring densely packed regions with lower porosity.

2.1. Atomic structure

Extracting the poorly crystalline gel structure of C-(A)-S-H from cement-based materials has proven difficult, posing challenges for in-depth studies. Tobermorite and jennite, which share structural similarities with C-S-H but exhibit more ordered arrangements, are often used as model compounds. Tobermorite exists in three distinct basal spacings: 0.9, 1.1, and 1.4 nm [71]. Among these, the 1.4 nm form of tobermorite closely resembles the C-S-H formed during cement hydration (left panel of Fig. 2 [72]). Tobermorite has a Ca/Si ratio of approximately 0.83, and its layered structure comprises a primary layer and an interlayer. The primary layer exhibits a typical sandwiched structure, with a distorted Ca-O sheet at the center, flanked by silicate chains composed of repetitive “Dreierkette units,” each comprising three silicate tetrahedra. The silicate tetrahedra that share two oxygen atoms with the Ca-O layer are termed paired tetrahedra (Q2P). By contrast, tetrahedra that share only one oxygen atom are known as bridging tetrahedra (Q2B). The interlayer region is filled with calcium ions and water molecules to maintain charge neutrality. Unlike macro-crystalline tobermorite, C-(A)-S-H is often nanocrystalline and contains structural defects, such as vacancies and Al-for-Si substitution within the dreierkette chains (Fig. 2 [72]). Jennite shares significant structural similarities with tobermorite; however, it exhibits a higher Ca/Si ratio of 1.5. The primary distinction between jennite and tobermorite lies in the primary layer; in jennite, only half of the oxygen atoms are bonded to the silicate chains, and the remaining oxygen atoms form hydroxyl (-OH) groups, whereas all the oxygen atoms in the primary layer of tobermorite are connected to the silicate chains [73]. Notably, a high Ca/Si ratio of C-S-H cannot be built by these two models and exhibits structural variability that can be modeled using the following primary approaches: the tobermorite-calcium hydroxide (T/CH) and tobermorite-jennite (T/J) models. In the T/CH model, the high Ca/Si ratio results from the interlayering of tobermorite-like structural units with calcium hydroxide (CH), forming a solid solution. Alternatively, the T/J model attributes the elevated Ca/Si ratio to a combination of tobermorite and jennite units, with the latter incorporating hydroxyl groups and additional calcium within its structure [74].

With the increasing use of Al-rich SCMs to reduce carbon emissions, silicon in the bridging tetrahedral sites can be substituted by aluminum, forming C-A-S-H (right panel of Fig. 2 [72]). This substitution increases the main chain length and reduces the Ca/Si ratio of C-A-S-H. The exact location of aluminum within C-A-S-H remains debatable; however, some consensus suggests that Al(IV) (tetrahedral coordination) is located in the dreierketten chain, and Al(V) (pentahedral coordination) and Al(VI) (octahedral coordination) are associated with strätlingite, a crystalline calcium aluminate hydrate with a layered structure, and a third aluminate hydrate, respectively. However, Al(VI) has also been proposed as an inherent component of C-A-S-H (Ref. [75]) or, controversially, as nonexistent [76]. In addition, crosslinking between aluminosilicate chains has been observed in low-Ca (Ca/Si ratio < 1) cement [77,78] and at elevated synthesis temperatures [19,21,79,80]. Al-induced crosslinking improves the structural order. However, aluminum incorporation typically reduces the crystallinity of C-A-S-H at room temperature, where “crystallinity” refers to the extent of structural order within C-A-S-H, particularly the regularity and coherence of its atomic arrangement at the nanoscale [81,82].

2.2. Bulk structure

Particle-scale description focuses on the structural characteristics of C-S-H with dimensions exceeding 10 nm. The first C-S-H gel model, known as the P-B model, was proposed by Powers and Brownyard in 1947 [83]. This model distinguishes between capillary, physically adsorbed, and non-evaporable water. However, it lacks information on the porosity and density of C-S-H and cannot explain the irreversible behavior observed during isothermal adsorption. Subsequently, Feldman and Sereda revised the P-B model and introduced a layered-structure model [84]. This model describes C-S-H as disordered aggregates of layers bound together by van der Waals forces. They further proposed that water could be an integral part of C-S-H, existing within the interlayer spaces. Building on this, Wittmann developed the Munich model [85], which suggests that C-S-H gel particles form a three-dimensional poorly crystalline network owing to chemical bonds and van der Waals forces. According to this model, the pores in C-S-H result from the packing of individual particles rather than from disordered layer stacking. Recent advances using 1H nuclear magnetic resonance relaxometry have provided direct experimental evidence supporting the presence of a distinct interlayer and gel water in C-S-H without the need for destructive drying procedures [86]. These studies revealed that C-S-H evolves from a loosely packed structure with significant gel porosity at the early hydration stages to a denser morphology with reduced gel pore volume, consistent with a densification mechanism. Furthermore, the average C-S-H domain may comprise only two to three stacked layers, offering new insights into the nanocrystalline organization within the gel.

With advancements in testing techniques and a deeper understanding of cement-based materials, the colloid model (CM-I and CM-II), as shown in Fig. 3, was proposed by Jennings et al. [69,70]. This widely accepted model provides critical parameters, including the specific surface area, density, fundamental particle size, and water content at varying relative humidities, of C-S-H. According to the colloid model, C-S-H comprises quasi-spherical particles with diameters of approximately 5 nm, and the internal structure of these particles resembles that of layered tobermorite. In this model of the C-S-H microstructure, the fundamental structural units are globules, typically approximately 4.2 nm in size, comprising tightly packed brick-like subunits arranged in a layered, sheet-like manner. These globules aggregate into globule flocs, forming a hierarchical structure. Globules exhibit irregular geometries, and their sizes and shapes contribute to the formation of a porous network at the nanoscale level. The globule flocs are fractal-like clusters in which individual globules are partially aligned. The spaces between these flocs, termed small gel pores, range from 1 to 3 nm in diameter and significantly determine the porosity of the system. The arrangement within each floc is not densely packed, resulting in interstitial spaces between the globules, affecting the mechanical and physical properties of the material. Additionally, the flocs, which are approximately 30-60 nm in size, overlap with neighboring flocs, forming a loosely packed network. This overlap leads to a wide distribution of pore sizes, including large gel pores, and influences the overall structure and performance of C-S-H. Interestingly, this globule-floc structure is closely linked to the intrinsic chemical composition of C-(A)-S-H formed during Portland cement hydration. Extensive experimental data, including those reported in Ref. [87], have shown that C-(A)-S-H consistently forms with Ca/Si ratios approximately 1.7 and H/S ratios near 4.0. As discussed in Ref. [88], such compositions result from the charge-balancing requirement between the silicate chains and interlayer calcium ions, resulting in energetically favorable and structurally stable configurations. This preferred stoichiometry inherently supports the formation of loosely packed disordered aggregates, reinforcing the globule-floc model. This model demonstrates the effect of both the packing arrangement and underlying composition of globules and flocs across different length scales on the porosity and mechanical behavior of C-S-H.

3. Intrinsic mechanical properties

C-(A)-S-H is a porous and poorly crystalline material with a porosity of approximately 30%. Standard methods, such as macroscopic uniaxial compression, micro/nano-indentation, and AFM, for determining mechanical properties can only measure its mechanical properties, including pore influence. The influence of pores must be excluded to sufficiently understand the relationship between its physical structure and mechanical properties. The mechanical properties of pore-free materials are defined by their intrinsic mechanical properties. This approach enables a deeper understanding of the Ca/Si ratio, the influence of Al uptake on the C-A-S-H structure, and its effects on the intrinsic mechanical properties.

3.1. High-pressure X-ray diffraction

As shown in Fig. 4, the test setup includes synchrotron X-rays and a diamond anvil cell (DAC). To maintain hydrostatic pressure, the sample was surrounded by a liquid (methanol and ethanol in a ratio of 4:1). Several high-pressure X-ray diffraction (HP-XRD) experiments on C-(A)-S-H samples have utilized this DAC, which is a stable and versatile device for HP-XRD [[89], [90], [91]]. The DAC features a 90° symmetrical axial opening, which enhances the range of accessible diffraction angles and renders it ideal for high-resolution diffraction measurements [92]. It utilizes a piston-cylinder design that ensures mechanical stability under high loads without requiring a complex angular alignment of the diamond anvils. In this experiment, a 250 μm thick steel gasket is pre-indented to 80 μm, with a 150 μm diameter hole drilled by laser milling. C-A-S-H sample powders, along with a ruby sphere (3-5 μm in diameter) for pressure calibration, are inserted into the hole, filling approximately half of the available volume. The remaining space is filled with a pressure medium comprising methanol and ethanol in a volume ratio of 4:1. The pressure-transmitting medium eliminates the influence of mesoscale pores (i.e., gel pores, interhydrate pores, and capillaries) on the measured mechanical response, thereby enabling a direct assessment of the intrinsic compressibility of the material. The BX90 design facilitates precise pressure control, with pressures measured via the fluorescence signal of the ruby, ensuring hydrostatic conditions of up to several GPa [25]. The large axial and side openings of the cell and its compatibility with high-energy synchrotron X-rays enable the efficient collection of diffraction patterns. For each C-A-S-H sample, 6-9 diffraction patterns were recorded at pressures ranging from ambient to 9-12 GPa. Synchrotron X-rays with a wavelength of ∼0.49755 Å are used as the incident beam, enabling detailed structural analysis of the samples under varying pressure conditions [91]. Notably, the broadening of diffraction peaks under a high applied hydrostatic pressure is attributed to the development of microstrain within the sample instead of any alteration of its intrinsic structure [93]. The axial incompressibility along each crystallographic direction is typically determined by jointly refining multiple diffraction peaks that contribute to the corresponding lattice parameters rather than relying on a single reflection.

The two-dimensional (2D) diffraction patterns were processed into diffractograms using Dioptas software [94]. Strong and isolated diffraction peaks were tracked across different pressure levels using the XFIT code [95], and the resulting data were used to refine the lattice parameters at each pressure using CELREF [96]. Volumetric changes under pressure were analyzed to calculate the bulk modulus using the second-order Birch-Murnaghan equation of state [97].

$P=\frac{3}{2}{K}_{0}\left[{\left(\frac{V}{{V}_{0}}\right)}^{-\frac{7}{3}}-{\left(\frac{V}{{V}_{0}}\right)}^{-\frac{5}{3}}\right]$

where V0 is the initial unit cell volume at ambient pressure, V is the unit cell volume at pressure P, and K0 is the bulk modulus at ambient pressure. The volumetric strain is defined as εV = 1−V/V0; therefore, Eq. (1) can be revised as

$P=\frac{3}{2}{K}_{0}\left[{\left(1-{\epsilon }_{V}\right)}^{-\frac{7}{3}}-{\left(1-{\epsilon }_{V}\right)}^{-\frac{5}{3}}\right]$

The HP-XRD results indicated that the incompressibility of CaO-SiO2-reacted C-S-H along the c-axis is significantly dependent on its Ca/Si ratio, substantially influenced by variations in the interlayer density, which in turn is governed by differences in basal spacings and interlayer Ca contents. The results showed that, in C-S-H synthesized via the CaO-SiO2 reaction, decreasing the bulk Ca/Si ratio led to an increase in basal spacing [20]. This trend aligns with previous observations in dried low-Ca systems; however, it must not be generalized, as the relationship between the Ca/Si ratio and basal spacing depends on factors such as the hydration state, Al content, and structural heterogeneity [88]. However, despite these findings, the study did not fully decouple the effects of basal spacing and interlayer Ca content on the nanomechanical properties of C-S-H. Both factors appear to contribute to the mechanical response; however, their roles remain unclear. Although increasing the Ca content leads to densification of the interlayer, enhancing mechanical stiffness, the simultaneous changes in basal spacing complicate the direct understanding of how each variable independently affects the compressibility and stiffness of the material [20].

Furthermore, C-S-H samples with three different Ca/Si ratios (0.8, 1.3, and 1.5) but comparable basal spacings were used to investigate the influence of the Ca/Si ratio on intrinsic mechanical properties [18]. As presented in Table 1 [18], the lattice parameters of the three C-S-H samples are largely consistent, with differences in basal spacing within approximately 0.2 Å. This system provides a solid foundation for investigating the effects of varying Ca/Si ratios on intrinsic mechanical performance by decoupling the effects of Ca content and basal spacing. The X-ray absorption spectroscopy results show that a higher Ca/Si ratio leads to an increased concentration of interlayer calcium, owing to several factors (Fig. 5(a) [18]). First, the increased presence of Ca2+ ions in the second coordination shell of Si enhances the electron shielding effect, reducing the electron density around Si and facilitating the incorporation of more Ca2+ into the interlayer. Second, as the Ca/Si ratio increases, silicate polymerization decreases, resulting in more fragmented and discontinuous silicate chains, creating additional space to accommodate interlayer calcium. Finally, interlayer calcium is essential in balancing the negative charges on silicate chains. Fewer silicate chains are present at higher Ca/Si ratios, requiring additional Ca2+ ions in the interlayer to maintain charge neutrality.

As illustrated in Fig. 5(b), the basal spacing of the three C-S-H samples with different Ca/Si ratios was approximately 12.75 Å. Figs. 5(c) and (d) illustrate incompressibility along different axes. The incompressibility along the a- and b-axes of co-precipitated C-S-H is independent of the Ca/Si ratio, indicating that silicate polymerization and the Ca/Si ratio do not significantly affect the deformation in these directions. By contrast, the incompressibility along the c-axis is positively correlated with the Ca/Si ratio, implying that as the Ca/Si ratio increases, the material becomes less compressible along the c-axis. This indicates that the interlayer calcium density, which increases with higher Ca/Si ratios, primarily governs the compressibility along the c-axis.

In another study, two types of non-crosslinked C-(A)-S-H with varying aluminum coordination were investigated alongside an Al-free sample [45]. The lattice parameters are listed in Table 2 [45], and the basal spacings are illustrated in Figs. 6(a)-(c) [45]. The Al-free C-S-H sample served as a baseline and is characterized by a simpler structure with minimal interlayer disruption, resulting in a smaller basal spacing than those of Al-containing samples. In the Al-containing samples, the stable phase (Al/Si ratio = 0.10, 546 d of curing) predominantly exhibited tetrahedrally coordinated Al(IV), which was incorporated into the silicate chains and slightly expanded the basal spacing. The sample with Al/Si ratio = 0.05, cured for 182 d, demonstrated a mix of Al(IV) and pentahedrally coordinated Al(V), contributing to further basal spacing expansion and increased structural disorder. Additionally, trace amounts of octahedrally coordinated Al(VI) were detected in this metastable sample, indicating incomplete reaction and structural development.

The study [45] demonstrated that the nanomechanical properties of non-crosslinked C-(A)-S-H along the a- and b-axes exhibit no consistent trend related to the varying aluminum coordination environments (Al(IV), Al(V), and Al(VI)), as shown in Fig. 6(d) [45]. Despite structural changes, such as differences in basal spacing and interlayer calcium content, in aluminum incorporation, the mechanical behavior in the ab-plane does not exhibit a noticeable pattern or dependence on these variations [45]. However, significant differences in the incompressibility along the c-axis were observed (Fig. 6(e) [45]). C-A-S-H, with an Al/Si ratio of 0.10 exhibits the highest incompressibility, whereas metastable C-A-S-H with an Al/Si ratio of 0.05 exhibits the lowest. This variation is attributed to the influence of interlayer ion concentration. During the substitution of silicon by aluminum, an additional proton or calcium ion is required to maintain electrical neutrality (Si4+ → Al3+ + H+ or CaOH+). The basal spacing of C-A-S-H with an Al/Si ratio of 0.10 increases by only 1 Å. However, owing to the higher aluminum content and extensive Si-to-Al substitution, the interlayer ion concentration is higher, suppressing deformation along the c-axis.

By contrast, the metastable C-A-S-H with an Al/Si ratio of 0.05 exhibits fewer Si-to-Al substitutions; however, it has a larger basal spacing, exhibiting the lowest interlayer ion concentration. Consequently, the c-axis in this sample is more prone to deformation under stress. Because the incompressibility along the a- and b-axes is negligible, that along the c-axis is dominant, leading to the observed positive trend in the bulk modulus. For instance, C-A-S-H with an Al/Si ratio of 0.05 has the highest bulk modulus, whereas metastable C-A-S-H with the same Al/Si ratio has the lowest (Fig. 6(f)).

Cross-linked (Al-) tobermorite structures have been observed with increasing synthesis temperatures and in practical scenarios, such as Roman concrete, oil wells, steam curing, and certain geological conditions [76,80,98]. As shown in Fig. 7, four types of C-A-S-H samples were investigated: cross-linked tobermorite with Al/(Al + Si) molar ratios of 0, 1/12, 1/6, and non-cross-linked tobermorite (single-chain) [19]. The single-chain structure refers to C-S-H without interlayer cross-links; double-chain C-S-H involves interlayer Q3Si bridging, whereas Al double-chain C-A-S-H results from cross-linking through Q3Al. The cross-linked tobermorite samples exhibit a more interconnected silicate network, with increasing aluminum content leading to greater incorporation of Al into the silicate chains, resulting in expanded basal spacing owing to the charge-balancing role of interlayer cations. As the Al content increases, additional Al is substituted for Si at the tetrahedral sites, enhancing structural complexity and introducing more interlayer Ca. By contrast, the non-cross-linked tobermorite (single-chain) has a simpler silicate structure with fewer cross-links between chains, resulting in a lower interlayer density and a more open structure, influencing its overall porosity and water retention capacity. The structural differences between the cross-linked and non-cross-linked C-A-S-H significantly affect their chemical and physical properties.

Fig. 8 shows the lowest incompressibility [19]. The mechanical properties along the a- and b-axes of C-A-S-H exhibit a noticeable trend that correlates with aluminum incorporation. As the Al content increases, particularly in the cross-linked samples, fewer defects were present in the silicate structure, enhancing the uniformity of the material. This reduction in the number of structural defects following the substitution of Si by Al leads to improved mechanical performance along the ab-plane, as the cross-links introduced by Al stabilize the silicate chains and reduce structural variability. Therefore, the presence of Al contributes to a more ordered structure, resulting in a consistent improvement in the mechanical properties along the ab-plane. As shown in Fig. 8(b), the double-chain cross-linked tobermorite exhibits the highest incompressibility along the c-axis, whereas the single-chain tobermorite shows the lowest. The mechanical differences along the c-axis between the cross-linked (double-chain) and non-cross-linked tobermorite (single-chain) result from the structural cross-links in the silicate chains. In the cross-linked tobermorite, these cross-links provide additional structural support, enhancing resistance to compression along the c-axis. The increased interlayer stability owing to these cross-links leads to higher incompressibility and stiffness. By contrast, these supporting connections are missing in the non-cross-linked tobermorite, thereby rendering the interlayer region more susceptible to compression and reducing the mechanical strength along the c-axis. Therefore, the absence of cross-linking in the silicate structure is a key factor contributing to the reduced rigidity observed in the non-cross-linked tobermorite.

The nanomechanical properties along the ab-plane and c-axis can be visually interpreted, as shown in Fig. 9. In Fig. 9(a), the double-chain tobermorite forms a hinge-like structure, making compression along the c-axis difficult [31]. However, because of the lack of support in the ab-plane (resulting from the low interlayer ion density), deformation occurs more easily in this direction. As shown in Fig. 9(b), because of the absence of support in the single-chain tobermorite along the c-axis, incompressibility is weakest in this direction. Because energy is dissipated through deformation along the c-axis, the incompressibility along the ab-plane remains unaffected. As shown in Fig. 9(c), cross-linked aluminum forms pillar-like structures between the layers, significantly enhancing the incompressibility along the c-axis. Additionally, during the substitution of silicon for aluminum, the resulting increase in interlayer ion strength (Si4+ → Al3+ + H+ or CaOH+) further suppresses deformation along the ab-plane. Owing to the cross-linked structure and Al uptake, the Al-cross-linked double-chain tobermorite exhibits the highest bulk modulus (Fig. 8(c) [19]). Although molecular dynamics simulations are effective for capturing atomic-scale mechanisms like chain polymerization, water reactivity, and tensile failure, they are inherently limited in describing volumetric lattice changes under hydrostatic pressure because of their small model size, periodic boundaries, and focus on anisotropic loading [99,100]. By contrast, experimental HP-XRD enables direct measurement of unit cell deformation and bulk modulus. To address this issue, coarse-grained or multiscale simulations with isotropic compression protocols and higher crystallinity models are required to quantitatively correlate structural evolution with mechanical stiffness in a manner consistent with experimental observations. Notably, the lattice response measured in the HP-XRD experiments are likely affected by non-hydrostatic pressure conditions, grain rearrangement, or the preferred orientation induced during loading. Additionally, sample compaction under high pressure can alter porosity or induce irreversible structural changes that deviate from ambient-state behavior. The presence of diamond-related background signals may also interfere with diffraction peak assignments, particularly in low-intensity regions, requiring careful spectral interpretation to avoid artifacts. These factors necessitate cautious analysis when evaluating the intrinsic mechanical properties of C-(A)-S-H under HP conditions.

3.2. HP-Raman spectroscopy

HP-Raman spectroscopy is an advanced technique employed to investigate the microstructural and mechanical properties of C-S-H under varying pressures [48,101,102]. In this method, as illustrated in Fig. 10, a DAC is utilized to apply controlled pressures up to several gigapascals to the sample, enabling the analysis of bond behavior under compression. The DAC compresses the sample, and rubies in the sample chamber provide in-situ pressure calibration through ruby fluorescence [103]. Raman spectroscopy captures vibrational frequencies, particularly those of Si-O bonds, by focusing a laser beam through a high-resolution objective onto the sample and analyzing the scattered light. A key feature of this technique is its ability to measure shifts in vibrational frequencies of chemical bonds as pressure is applied, providing insight into bond strength and anharmonicity.

The Grüneisen parameter, which describes the linear relationship between bond vibrational frequency and volume change under stress, was determined using the Raman spectra obtained at different pressures. This parameter describes the manner in which the vibrational frequencies of a material respond to pressure-induced volume changes. In Eq. (3), the potential energy function U(l) incorporates the balance between attractive and repulsive forces as the interatomic distances, l, change. The mode Grüneisen parameter γi, shown in Eq. (4), expresses the sensitivity of the vibrational frequency ωi to volume changes, with larger γi values indicating softer bonds that are more sensitive to pressure. The bulk Grüneisen parameter γ, as expressed in Eq. (5), is the weighted average of individual mode Grüneisen parameters, reflecting the collective behavior of the vibrational modes of the material. The heat capacity contribution of each mode is outlined in Eq. (6), providing a detailed calculation of the contributions of different vibrational modes to the overall behavior of the system. The dimensionless parameter ${X}_{i}=\frac{\hslash {\omega }_{i}}{{k}_{\text{B}}T}$, expressed in Eq. (7), further characterizes the vibrational modes. These equations explain the effect of pressure on the mechanical properties of a material through changes in vibrational frequencies. HP-Raman spectroscopy provides direct experimental access to these parameters, which are essential for improving computational models, such as C-S-H-force field (CSH-FF), and designing C-S-H-based materials with enhanced mechanical performance [104]. This technique has been applied to C-S-H and model minerals such as tobermorite and xonotlite, highlighting its utility in characterizing complex bond interactions in cementitious materials [48,102]. Notably, the shift and broadening of spectral bands under pressure are primarily attributed to stress-induced microstrain and local plastic deformation, rather than any permanent disruption of the atomic structure.

$U(l)={U}_{0}-a{l}^{-A}+r{l}^{-R}$

where U0 is the reference energy, a is the coefficient of the attractive term, A is the power exponent of the attractive potential, r is the coefficient of the repulsive term, and R is the power exponent of the repulsive potential. The frequency of a given vibration, ωi, varies with pressure (P) as well as with the corresponding volume. The ratio of the relative change in frequency to the relative change in the associated volume is defined as the mode Grüneisen parameter, γi. Under small strain conditions, the mode Grüneisen parameter of a particular oscillator is related to the exponents in Eq. (1) as follows:

${\gamma }_{i}=-\frac{{V}_{i}}{{\omega }_{i}}\frac{\text{d}{\omega }_{i}}{\text{d}V}=-\frac{{V}_{i}}{{\omega }_{i}}\frac{\text{d}{\omega }_{i}/\text{d}P}{\text{d}V/\text{d}P}=\frac{A+R+3}{6}$

where Vi represents the volume that varies during vibration rather than the unit cell volume. Larger values of γi correspond to softer and more anharmonic interactions. Typically, the mode Grüneisen parameter is about 3.5 for 6-12 Lennard-Jones interactions, approximately 2.2 for ionic interactions, and around 1 for covalent interactions.

The bulk Grüneisen parameter 𝛾 is related to the mode Grüneisen parameters of individual oscillators described in Eq. (4) through Eqs. (5), (6), (7).

$\gamma =\frac{{{\displaystyle \sum }}_{i}{N}_{i}{c}_{Vi}{\gamma }_{i}}{{{\displaystyle \sum }}_{i}{N}_{i}{c}_{Vi}}=\frac{\alpha {K}_{0}}{\rho {C}_{V}}$
$\frac{{c}_{Vi}}{{k}_{\text{B}}}={X}_{i}^{2}\frac{\mathrm{exp}\left({X}_{i}\right)}{{\left(\mathrm{exp}\left({X}_{i}\right)-1\right)}^{2}}$
${X}_{i}=\frac{\hslash {\omega }_{i}}{{k}_{\text{B}}T}$

where cVi is the constant-volume heat capacity of a given oscillator, Ni is the number of such oscillators, α is the constant-pressure thermal expansivity, K0 is the bulk modulus, ρ is the density, CV is the bulk constant-volume heat capacity, kB is the Boltzmann constant, Xi is the ratio of vibrational to thermal energies, T is the absolute temperature, and $\hslash $ is the reduced Planck constant.

Based on the aforementioned Grüneisen parameters and Raman shifts of C-(A)-S-H presented in Table 3 [105,106], structural changes in C-(A)-S-H under pressure can be inferred. The peak at approximately 667-695 cm−1 is attributed to Si-O-Si symmetric bending. The peak near 1010 cm−1 is assigned to the symmetric stretch of all Q2 Si-O groups rather than the bridging or paired ones.

In C-(A)-S-H, the Grüneisen parameter values are in the range 0.3-1.4 (Table 4 [102]), depending on the specific vibrational mode and chemical composition of the sample. For instance, the Q2 Si-O stretching mode, which is significant for understanding the structural behavior under pressure, has a Grüneisen parameter within the range of 0.4-1.4 for C-(A)-S-H with a Ca/Si ratio of 0.8-1.3. The pressure-induced shift of this mode in vibrational frequencies (dω/dP) is approximately within the range of 5.5-8.1 cm−1·GPa−1. This indicates that, as pressure increases, the silica tetrahedra in C-(A)-S-H experience significant anharmonic distortions, particularly in the interlayer regions where the calcium ions and water molecules are located. The interlayer water content is essential, as an increased water content correlates with higher anharmonicity. For example, C-(A)-S-H with a higher water content demonstrates a more pronounced Grüneisen parameter, typically approximately 1.1, owing to the greater flexibility of the silica tetrahedra and the ability of the interlayer to accommodate structural shifts under pressure. These changes contribute to a reduction in interlayer spacing, thereby influencing the overall compressibility and mechanical behavior of the material under HP conditions.

However, a comparison of the Grüneisen parameters from the HP-Raman experiments and CSH-FF molecular dynamics simulations showed notable differences. Experimentally, the mode Grüneisen parameter for Si-O symmetric stretching in C-A-S-H ranges from 0.9 to 1.4, depending on the sample composition and water content. By contrast, simulations with the CSH-FF force field tend to overestimate these values, particularly for 14 Å tobermorite, where the calculated parameter reaches an unusually high value of 12.1. This discrepancy indicates that the CSH-FF force field inadequately models the anharmonicity in the silicate structure, primarily because it considers Si-O interactions as predominantly electrostatic. However, the experimental results indicated that the covalent character of these bonds was more significant, particularly in the cross-linked structures. This highlights the need for the reparameterization of CSH-FF, particularly for Si-O bonds, to more accurately capture the behavior of the material under pressure [22]. Although HP-Raman spectroscopy effectively probes the pressure response of silicate-bonding environments, it is sensitive to local thermal effects, microstrain, and variations in water content. Furthermore, Raman signals are likely influenced by diamond anvil background peaks, particularly under high magnification or laser intensity, which requires careful spectral deconvolution to avoid misinterpretation.

4. C-(A)-S-H deformation mechanism

4.1. C-(A)-S-H sliding

The distinct deformation mechanisms of cross-linked and non-cross-linked C-(A)-S-H under hydrostatic stress enable to understand their bulk behavior. Extending this analysis to deviatoric stress reveals differences in their sliding mechanisms, as shown in Fig. 11 [48]. In the cross-linked C-(A)-S-H (Fig. 11(a)), the plot demonstrates that the Q2 Si-O symmetric stretching strain diverges from the Q2 Si-O-Si bending strain under deviatoric conditions, indicating relaxation near the Si-O(-Ca) bond. This is attributed to the relaxation of intralayer bonds, where the structure absorbs stress through intralayer sliding without causing significant bond strain. Fig. 11(c) shows a plateau in the strain response, indicating that the intralayer Si-O-Ca bonds accommodate stress by redistributing it within the material, thereby preventing further strain accumulation. This mechanism facilitates in maintaining the structural integrity of C-(A)-S-H and delays the onset of plastic deformation because crosslinking provides additional resistance to sliding.

By contrast, the non-crosslinked C-S-H (Fig. 11(b)), which lacks aluminum-induced crosslinks, exhibits interlayer sliding under similar stresses. The Q2 symmetric stretching and bending modes exhibited consistent strain, indicating that deformation occurs between rather than within layers. This leads to sliding along the interlayer plane, where the weaker bonds between calcium, water molecules, and silicate chains cause significant plastic deformation. As presented in Table 5, this results in a reduction in grain thickness and formation of defects, as observed via small-angle X-ray scattering.

As shown in Fig. 12, the underlying mechanism for these differences is the structural composition of the material. In C-A-S-H, Al incorporation can lead to partial cross-linking between adjacent layers through charge-balancing and bridging sites, resulting in a large rigid interlayer region. This structural constraint inhibits interlayer sliding under deviatoric stress. Consequently, deformation tends to localize within the intralayer domain, particularly through shear displacements within the CaO7 polyhedral layers. Such intralayer sliding may lead to stacking disorder, particularly in high-Al phases. By contrast, the absence of cross-linking in C-A-S-H results in weaker interlayer cohesion, rendering the interlayer a primary deformation site. This mechanistic distinction incorporates the observed differences in the plastic deformation behaviors of C-A-S-H.

4.2. C-(A)-S-H preferred orientation

In micromechanical studies on C-A-S-H, deformation under hydrostatic and deviatoric stresses must be distinguished. Without a pressure-transmitting fluid, a mixed stress field develops, inducing anisotropic nanocrystallite orientation, as reported in Refs. [17,107]. This highlights the need to incorporate nonuniform stress distributions when analyzing the micromechanical behavior and creep mechanisms of C-A-S-H at the microscale. As shown in Fig. 13 [17], the C-A-S-H specimens with and without load were tested using synchrotron-based X-ray diffraction, as mentioned in Section 3.1. The beam-sample interaction volume is approximately 30 μm3. Fig. 13 [17] illustrates the diffraction patterns of C-A-S-H under both loaded and unloaded conditions, showing the differences in the 2D X-ray diffraction patterns. The unloaded sample exhibits random orientation of nanocrystals, indicating isotropy. By contrast, loading induces a strongly preferred orientation, aligning the c-axis of the C-A-S-H nanocrystals along the compression direction. This alignment causes significant anisotropy in the diffraction pattern, particularly in the pronounced intensity of specific diffraction rings corresponding to the (002) plane. This indicates that deviatoric stress under compression facilitates the reorientation of nanocrystals, contributing to the anisotropy observed in the material.

Fig. 14 [17] illustrates the integration of the diffraction patterns at different azimuthal angles. A raw 2D X-ray diffraction image was divided into 72 segments, each representing a 5° azimuthal slice. The diffraction intensity was calculated for each segment as a function of the azimuthal angle. The specific diffraction peaks in Fig. 14 [17], including the (002) plane, were measured and labeled at selected azimuthal angles. The diffraction patterns at 0°, 180°, and 360° correspond to the directions parallel to the loading axis. This indicates that the C-A-S-H nanocrystals were preferentially aligned along the compression direction in these orientations. By contrast, at 90° and 270°, the patterns reflect directions perpendicular to the loading axis, providing insights into the anisotropic structure of the material under loading. The intensity variation across different azimuthal angles indicates the pronounced preferred orientation induced by deviatoric stress applied during compression, with the strongest alignment observed along the c-axis of the nanocrystals.

When analyzing preferred orientation, the degree of crystallographic alignment must be quantified to better understand the evolution of anisotropy in C-(A)-S-H under applied loads. An effective approach for this quantification is texture analysis, which enables the orientation distribution of crystallites to be measured. An effective indicator of this alignment is the concept of multiples of random distribution (m.r.d.), which serves as a reliable metric for quantifying the preferred orientation of crystallites [17,107]. The m.r.d. values were calculated by integrating the intensity of diffraction peaks over azimuthal angles and comparing the results to a reference state in which crystallites are randomly oriented.

As illustrated in Fig. 15 [17], the m.r.d. represents the ratio of the observed intensity distribution along a given crystallographic direction to that of a random orientation. An m.r.d. value greater than 1 indicates a preferred orientation, implying that the crystallites are aligned more strongly along certain directions than along a random distribution. Conversely, m.r.d. values less than 1 indicate a weaker alignment or that crystallites are less frequently oriented in a particular direction than would occur randomly. An m.r.d. value of 1 represents a truly random orientation, where no preferred orientation is observed.

Additionally, relative humidity (RH) is crucial for enhancing the preferred orientation of C-A-S-H nanocrystals, as indicated by the high m.r.d. values under moist conditions. For example, under a load of 200 MPa, the m.r.d. value for the [002] direction increases from 6.1 in dry conditions to 7.6 at RH 100%, indicating that moisture significantly strengthens the alignment of nanocrystals [17], as illustrated in Fig. 16. This effect is primarily attributed to the lubricating properties of water in gel pores. Water reduces inter-crystallite friction, enabling nanocrystals to rotate and reorient more easily in response to applied stress. The resistance to this reorientation is higher in the absence of water, leading to a less pronounced preferred orientation. Therefore, water in the matrix promotes more effective crystallite realignment along the compression direction.

Furthermore, humidity amplifies the time-dependence of this process. When a 200 MPa load is applied for 300 s, the m.r.d. under RH 100% reaches 15.4. The presence of water facilitates continuous nanocrystal movement and alignment over time, enabling a stronger preferred orientation as the applied time of the load increases.

Several studies have quantitatively demonstrated the influence of load on the preferred orientation of C-A-S-H nanocrystals through m.r.d. analysis. Under compressive loads, the m.r.d. values for the [002] crystallographic direction increase significantly, reflecting a stronger c-axis alignment with the loading direction. For example, at a stress of 200 MPa, the m.r.d. reaches approximately 7.6; at 500 MPa, it increases further to 8.3. This illustrates a direct relationship between applied load and degree of preferred orientation. The mechanism behind this load-induced preferred orientation can be attributed to the rotation and realignment of C-A-S-H nanocrystals under deviatoric stress. As the load increases, the energy supplied to the system enhances the ability to address inter-crystallite friction, promoting reorientation. This realignment causes nanocrystals to stack with their c-axes aligned along the compression direction. Furthermore, maintaining the stress for 300 s at 200 MPa significantly increases the m.r.d. from 7.6 to 15.4, indicating a time-dependent process in which prolonged pressure facilitates further reorientation of nanocrystals. This process is time-dependent because the sustained stress provides additional opportunities for crystallite rotation and realignment over time, leading to a more pronounced preferred orientation. The stronger texture formation at higher loads can be attributed to higher energy input, which addresses the barriers to reorientation more frequently, enabling a more complete alignment of crystallites.

Nanoindentation has been widely applied to test the micromechanical properties of synthesized C-A-S-H [[108], [109], [110], [111]]. However, the preferred orientation occurs during the initial compaction of the powder into pellets and during the indentation process [22], as shown in Fig. 17. In the compaction stage, compressive stresses (often reaching 200 MPa or more) first cause C-A-S-H nanocrystals to align, particularly with their c-axes parallel to the compression direction, leading to significant anisotropy in the microstructure of the material. The indentation process further enhances this preferred orientation, particularly in the region directly beneath the indenter. The localized stress, which is typically substantially higher than the initial compaction pressure, applied by the indenter promotes additional reorientation of the nanocrystals. Indentation depths are commonly in the range of 100-300 nm, producing a probe volume ranging from 1 to 10 μm3 [8]. The effect of the probe volume is crucial because it encompasses a localized region that may predominantly reflect the softer c-axis of the nanocrystals, which tends to align with the compression direction. This can lead to an underestimation of the elastic modulus and hardness, as the measurements may overrepresent the mechanical response along the c-axis. Previous studies have observed that the c-axis typically exhibits lower stiffness than the a- and b-axes, leading to biased results when the probe volume is dominated by these aligned regions [19,91]. Therefore, special care is required when using nanoindentation to evaluate the micromechanical properties of C-A-S-H. The preferred orientation induced by the compaction and indentation processes, combined with the influence of probe volume, can skew the results and lead to an inaccurate representation of the bulk mechanical properties. To mitigate this issue, careful experimental design and consideration of orientation effects are crucial to ensure reliable and representative measurements of the mechanical behavior of C-A-S-H.

The evolution of the C-(A)-S-H nanocrystallite orientation under deviatoric stress provides insights into the nanoscale mechanisms driving creep in cement-based materials. Initially, when subjected to a lower compressive stress (e.g., less than 100 MPa), the C-S-H nanocrystallites, starting with a random orientation, develop a preferred orientation, as shown in Fig. 18 [107]. The c-axes of the nanocrystallites, representing the softest direction, progressively align with the direction of the applied compressive stress. This initial reorientation is primarily attributed to the removal of intergranular pores and closer packing of the nanocrystallites, facilitating a more ordered structure. The intensification of this preferred orientation is a critical factor that contributes to the time-dependent creep behavior of the material under service conditions. However, as the applied stress continues to increase beyond ∼100 MPa, the deformation mechanism shifts, weakening the previously established texture. Two primary processes contribute to this change: interlayer sliding and nanocrystallite breakage. Interlayer sliding occurs when the layers within the C-(A)-S-H nanocrystallites begin to slip past each other under higher compressive stresses. This sliding is most noticeable in nanocrystallites whose c-axes are already well-aligned with the compressive direction. As sliding disrupts the alignment of these crystallites, the overall texture weakens and the intensity of the (002) diffraction decreases. In addition to interlayer sliding, an increase in stress induces breakage within the nanocrystallites, specifically along the silicate chains. After stress release, the preferred orientation briefly strengthens as the nanocrystallites experience reduced friction and can realign. This short-term recovery reflects the elastic nature of the material under a lower stress. However, as the unloading continues, this recovery diminishes, and the texture weakens, indicating that damage caused by interlayer sliding and chain breakage is not completely reversible (as shown in the right part of Fig. 18 [107]). This breakage, which has been observed in other studies [48] at stress levels of approximately 200 MPa, further disrupts the ordered structure of C-(A)-S-H. The breakdown of the silicate network reduces the cohesion of the nanocrystallites and contributes to reducing the preferred orientation at higher stress levels. These transformations within the C-S-H structure are significant because they introduce plastic deformation mechanisms that are less relevant under typical service conditions but become prominent under higher experimental stress levels.

The combination of these two mechanisms (interlayer sliding and nanocrystallite breakage) highlights the complexity of the response of C-(A)-S-H to stress. Although the preferential reorientation of nanocrystallites under lower stresses contributes to the creep behavior of concrete in real-world applications, additional plastic deformation mechanisms at higher stress levels, such as those encountered in nanoindentation, may not accurately represent the behavior of the material under normal service conditions. This shows the importance of considering the stress environment when interpreting experimental results and modeling the mechanical behavior of cement-based materials.

4.3. Interpretation of cement paste creep

Many studies have investigated the effects of the Ca/Si and Al/Si ratios on the micromechanical properties of C-A-S-H. Furthermore, efforts have been made to establish the relationships between the chemical composition of C-A-S-H and the micromechanical properties of cement-based materials, particularly creep. These studies aimed to use the micromechanical properties of C-A-S-H to elucidate the creep mechanisms in cement-based materials. However, most studies remain limited to the time and length scales of C-A-S-H creep, rendering it challenging to establish a direct connection with the creep behavior of cement-based materials. The preferred aforementioned orientation offers the potential to directly reveal the relationship between the deformation of C-A-S-H and the creep of cement-based materials. The creep process in cement-based materials is believed to involve the rearrangement of C-A-S-H, including aging and sliding, where hydration products tend to achieve optimal packing. Preferred orientation is essential in the motion of C-A-S-H. As shown in Fig. 19, the left side illustrates the distribution of hydration products in the cement paste before compression. These products, including fibrillar C-S-H, CH, and ettringite, exhibit random distribution. As indicated on the left side of Fig. 19, C-S-H, CH, and ettringite also display a preferred orientation [17]. The preferred orientations of the CH and ettringite crystals under compressive stress are driven by their distinct crystal morphologies and mechanical responses. For CH (portlandite), which has a plate-like structure, the [001] direction, corresponding to its short dimension, aligns with the compression axis during compaction. This alignment minimizes inter-crystalline friction and optimizes packing efficiency, facilitating stress-induced reorientation. The degree of orientation is evidenced by the full width at half maximum (FWHM) value of 121.5° for the (001) plane and a moderate m.r.d. range of 0.74-1.51 under a load of 500 MPa without holding time. By contrast, ettringite, which has a needle-like morphology, tends to align in the [002] direction perpendicular to the compression axis to reduce mechanical resistance along its stiffest axis. This behavior is supported by the FWHM value of 60.6° for the (002) plane and an m.r.d. range of 0.61-1.21 under a load of 200 MPa without holding time. These results [17] demonstrate that the morphological and mechanical characteristics of the CH and ettringite crystals govern their preferential reorientations, although to a lesser extent than the strong texture observed for C-A-S-H under similar conditions. The preferred orientations of C-S-H, CH, and ettringite collectively reduce the cement paste volume, contributing to creep.

The effect of adding SCMs on creep in blended cement paste remains controversial. This preferred orientation provides new insights. As the Ca/Si ratio decreases, the morphology of C-A-S-H shifts from fibrillar to foil-like [41]. Because of the secondary hydration reactions of SCMs, the CH content is significantly lower than that of the cement paste. Although this renders microscopic creep analysis complex, it still provides a credible mechanism. However, substituting Al in the interlayers enhances the interlayer Ca2+ density, reducing the compressibility of C-A-S-H. Furthermore, cross-linked C-A-S-H forms at elevated temperatures, significantly improving mechanical performance along the c-axis. Consequently, the creep of C-A-S-H decreases. By contrast, CH exhibits a weaker preferred orientation than C-A-S-H, with m.r.d. values ranging from 0.74 to 1.51 and a broad FWHM of 121.5°. Further, C-A-S-H exhibits stronger alignment, with m.r.d. values exceeding 15.4 and a narrowed FWHM of 24.4°, aided by its nanocrystalline structure and the lubrication provided by gel-pore water. Therefore, reducing the CH content and increasing the C-A-S-H content can lead to increased creep. Moreover, CH, owing to its high mechanical performance, is often considered a key factor in suppressing creep, and its reduction could contribute to increased creep. However, the varying degrees of the preferred orientations of C-A-S-H, CH, and ettringite are the primary factors driving the increased creep in blended cement paste (Fig. 20). The contribution of each phase to creep must be carefully evaluated based on specific conditions.

Recent studies have investigated the creep behavior of alkali-activated materials using both experimental and multiscale modeling approaches. In particular, one-part alkali-activated slag-fly ash pastes exhibit a lower specific basic creep strain than ordinary Portland cement (OPC) paste [112]. This behavior is attributed to the denser pore structure and cross-linked bonding configuration of the C-(N)-A-S-H gel, as evidenced by 29Si nuclei magnetic resonance (NMR) and nitrogen adsorption analyses. However, estimating gel-level creep properties involves assumptions regarding the agglomeration state and unreacted phase content, and the separation of shrinkage and creep effects remains challenging in real paste systems. However, this confirms the function of C-(N)-A-S-H in creep. In addition to the aforementioned effects, the creep behavior of alkali-activated systems can be further explained through preferred orientation, as shown in Fig. 21. Unlike (blended) cement paste, the hydration products of alkali-activated cement-based materials differ significantly, causing the formation of Mg-Al layered double hydroxide (LDH). The intrinsic mechanical properties of Mg-Al hydrotalcite are influenced by its layered structure and the interactions between brucite-like layers and interlayer ions. Molecular dynamics simulations reveal that the material exhibits anisotropic behavior with different mechanical responses along the c-axis compared with the ab-plane. The ab-plane, with Mg-O bond lengths of approximately 0.207 nm, is relatively stable. In comparison, the c-axis exhibits greater flexibility under hydration and pressure conditions [113]. HP-XRD experiments show interlayer spacings around 0.76 nm, with a contraction of 0.305 nm under certain conditions [114,115]. Apparently, Mg-Al LDH also develops a significantly preferred orientation under load, contributing to creep.

Nevertheless, creep in cement pastes is not governed by a single dominant mechanism. In addition to anisotropic deformation associated with preferred orientation, nanoscale sliding along the lamellar structure of C-(A)-S-H and intergranular debonding-rebonding processes are essential in the overall time-dependent response. These deformation modes are sensitive to the atomic packing density and local stiffness contrast [2], both of which vary with the Ca/Si ratio [46]. The relative contributions of these mechanisms may vary depending on the intrinsic composition and morphology of the C-(A)-S-H gel. Therefore, a comprehensive interpretation of cementitious creep must integrate the effects of both intrinsic structural anisotropy and competing nanoscale deformation processes, highlighting the necessity of combined experimental and modeling approaches. Although the development of a unified framework reconciling mechanisms such as interlayer sliding and dissolution-precipitation remains an important long-term goal, it is challenged by the complexity and multiscale nature of creep phenomena. Instead, this review focuses on the intrinsic mechanical behavior of C-(A)-S-H from a foundational perspective, aiming to provide mechanistic insights that may inform and support future efforts toward more integrated theoretical models.

5. Improving the stiffness of C-(A)-S-H

As discussed previously, the preferred orientation of C-A-S-H under loading is one of the primary mechanisms driving creep in cement-based materials. As shown in Fig. 22, modern concrete structures, particularly long-span bridges and skyscrapers, experience sustained loads that cause C-A-S-H to undergo rearrangement along the softer [002] direction, resulting in volumetric deformation and increased creep. This review proposes a bottom-up strategy to address this issue by enhancing the mechanical properties of C-A-S-H along the [002] direction. This approach is expected to significantly reduce deformation during the service life of cement-based materials. Subsequently, several feasible practical strategies are proposed.

As discussed in Section 3, optimizing the Ca/Si ratio within calcium silicate C-S-H is an effective strategy for improving its mechanical performance. Studies have demonstrated that an increase in the Ca/Si ratio results in a higher interlayer calcium content, which densifies the interlayer region and contributes to enhanced structural integrity, as shown in the top row of Fig. 23. This densification leads to a notable increase in both the stiffness along the layering (c-axis) direction and bulk modulus, which are essential for improving the mechanical resilience of C-S-H. The stability of basal spacing across varying Ca/Si ratios ensures that the observed improvements are primarily caused by the interlayer calcium density rather than by changes in structural dimensions.

Incorporating aluminum into C-S-H structures forms cross-links that enhance mechanical properties by reinforcing the material along the c-axis, as shown in the middle row of Fig. 23. Aluminum substitutes silicon at bridging sites, creating Al-O-Si linkages that serve as structural supports and increase interlayer stability and stiffness under mechanical stress. These cross-links also promote stronger hydrogen bonding and enhance the concentration of charge-balancing calcium ions, further contributing to the durability and bulk modulus of the material. Several practical methods can be used to facilitate the formation of these cross-linked structures. Steam curing at elevated temperatures (> 80 °C) is widely used to accelerate the development of Al-tobermorite phases, contributing to early strength gain and long-term stability. Nevertheless, carefully controlling the temperature and curing time can mitigate the risk of delayed ettringite formation, which is a potential durability concern. Hydrothermal treatment offers another pathway that uses controlled pressure and temperature to synthesize aluminum-rich C-S-H phases that resist carbonation and provide superior mechanical performance. Incorporating SCMs, such as fly ash or slag, during hydration encourages aluminum uptake, forming robust cross-linked structures, while reducing the carbon footprint of cement production. These approaches optimize material performance and durability and align with sustainable construction goals by enabling the production of high-performance concrete with enhanced load-bearing capacity and longevity.

Incorporating alkali metals, particularly sodium, into C-A-S-H forms C-N-A-S-H, enhancing mechanical properties through the synergistic interaction between sodium and aluminum. As an interlayer counterion, Na balances the negative charges introduced by Al substitutions, thereby stabilizing the structure. Furthermore, aluminum cross-links within the silicate chains increase stiffness along the c-axis, increasing the bulk modulus to over 100 GPa, compared with the typical 20-30 GPa of unmodified C-A-S-H [116]. These enhanced properties have practical applications for alkali-activated concrete. Alkali activation promotes rapid polymerization of aluminosilicate networks, yielding both early-age strength and long-term durability, as shown at the bottom of Fig. 23. Elevated-temperature curing, such as at 80 °C, accelerates cross-linked phase formation, achieving compressive strengths of over 60 MPa within 24 h. The presence of Na+ also mitigates creep deformation by stabilizing interlayer spacing and reducing microcracking under load. This approach offers a practical pathway for producing sustainable high-performance concrete that combines superior stiffness, durability, and stability to satisfy modern construction requirements.

As another approach, the synergistic interaction between graphene oxide (GO) and polycarboxylate ether (PCE) enhances the nanomechanical properties of C-S-H by improving dispersion and reinforcing the basal planes [117,118], as shown in Fig. 24. GO alone tends to aggregate in alkaline environments, thereby reducing its interaction with C-S-H. However, PCE aids in dispersing GO nanoplatelets, preventing agglomeration and aligning them with the basal planes. This alignment strengthens the intralayer Ca-O bonds, contributing to improved stiffness along the a- and b-axes, increasing the bulk modulus from 43.2 GPa for pure C-S-H to 82.2 GPa, indicating a 90% increase. Mechanistically, reinforcement occurs as the functional groups on GO interact with the CaO sheets of C-S-H, while the steric stabilization of PCE ensures uniform nanoplatelet distribution. This interaction minimizes lattice distortions and prevents microcracking, enhancing stiffness along the basal planes and overall structural integrity. The bulk modulus improvement is primarily attributed to enhanced in-plane reinforcement because the c-axis remains relatively unaffected. In engineering applications, the GO-PCE synergy provides a pathway for developing high-performance concrete with superior mechanical properties. This is particularly useful for precast concrete, bridge components, and alkali-activated materials, where enhanced stiffness and reduced deformation are critical. To maximize these benefits in practice, optimizing the dosage and mixing protocols to maintain GO dispersion throughout the cement matrix is essential to ensure long-term durability and resilience in demanding environments. Although the synergistic interaction between GO and PCE has demonstrated significant improvements in nanomechanical performance at the laboratory scale, practical implementation at larger scales requires careful consideration of dispersion stability and dosage optimization. In field applications, achieving a uniform distribution of GO remains a major challenge because it tends to aggregate in highly alkaline environments. Therefore, scalable mixing protocols and the development of compatible admixture systems are essential to translate these enhancements into real-world concrete production. Continued efforts are needed to balance performance benefits with economic and processing feasibility. Although this review focuses on the intrinsic mechanical enhancement of C-(A)-S-H, future studies must evaluate the long-term durability of these strategies under aggressive environmental conditions.

6. Conclusions

This study reviews the particle- and sub-particle-scale models of C-A-S-H, starting with their atomic structures. Advanced characterization techniques, particularly HP testing methods, were utilized to investigate the intrinsic mechanical properties of C-A-S-H and its deformation under load using HP-XRD and HP-Raman spectroscopy. The sliding and preferred orientation of C-A-S-H explain the underlying mechanism of creep in cement-based materials. Strategies for enhancing the performance of cement-based materials were proposed based on the bottom-up theory. The key conclusions are as follows:

(1) As the Ca/Si ratio is increased, the nanomechanical properties of C-S-H are enhanced, particularly with greater incompressibility along the c-axis. This is attributed to the reduced interlayer spacing, which increases the interlayer ion strength and suppresses deformation along the c-axis. The incorporation of aluminum also enhances the mechanical performance of C-A-S-H along the c-axis. This is attributed to the increase in ion strength required for charge balance in the interlayer when silicon is substituted for aluminum, further suppressing deformation.

(2) HP-Raman spectroscopy is a reliable method for investigating the deformation of C-(A)-S-H. The shift in the Q2 wavenumbers enables the quantitative investigation of C-A-S-H deformation at the atomic scale. Under loading, both interlayer and intralayer sliding occur in C-(A)-S-H. In the absence of cross-linked structures, sliding occurs in the interlayer; otherwise, it occurs in the intralayer. The application of HP-Raman spectroscopy to cement-based materials provides a reliable reference for molecular dynamics simulations.

(3) Under loading, preferred orientation is a major factor influencing the deformation of cement-based materials, particularly creep. Preferred orientation can occur even under a deviatoric stress of 40 MPa, which is comparable to the stress that causes macroscopic creep. The creep behavior of both (blended) cement paste and alkali-activated concrete can be explained by the preferred orientation of different phases, primarily C-(A)-S-H. Therefore, the gap between the nano-mechanical and micro-mechanical properties of C-A-S-H was bridged.

(4) To enhance the strength of C-(A)-S-H, the primary focus is to improve the mechanical performance along the c-axis to enhance the durability of concrete structures during their service life. Steam-curing or alkali-activated reactions are effective approaches to generate cross-linked structures. In the ab-plane, the synergistic effect of GO and PCE has also been shown to significantly improve nanomechanical properties.

References

[1]

Hu Z, Hilaire A, Wyrzykowski M, Lura P, Scrivener K. Visco-elastic behavior of blended cement pastes at early ages. Cem Concr Compos 2020; 107:103497.

[2]

Vandamme M, Ulm FJ. Nanogranular origin of concrete creep. Proc Natl Acad Sci USA 2009; 106(26):10552-7.

[3]

Hu Z, Wyrzykowski M, Griffa M, Scrivener K, Lura P. Young’s modulus and creep of calcium-silicate-hydrate compacts measured by microindentation. Cem Concr Res 2020; 134:106104.

[4]

Yang G, Wang J, Li H, Yao T, Wang Y, Hu Z, et al. Creep behavior of selfcompacting rubberized concrete at early age. J Mater Civ Eng 2022; 34 (3):04021474.

[5]

Wang J, Hu Z, Chen Y, Huang J, Ma Y, Zhu W, et al. Effect of Ca/Si and Al/Si on micromechanical properties of C(-A)-S-H. Cem Concr Res 2022; 157:106811.

[6]

Wang J, Ding S, Du F, Tang J, Liao J, Deng L, et al. Interaction between calcium hydroxide and calcium-alumino-silicate-hydrate on micromechanical properties. Constr Build Mater 2024; 449:138321.

[7]

Zhang Z, Yan Y, Qu Z, Geng G. Endowing strength to calcium silicate hydrate (C-S-H) powder by high pressure mechanical compaction. Cem Concr Res 2022; 159:106858.

[8]

Wang J, Gao C, Tang J, Hu Z, Liu J. The multi-scale mechanical properties of calcium-silicate-hydrate. Cement Concr Compos 2023; 140:105097.

[9]

Chen JJ, Sorelli L, Vandamme M, Ulm FJ, Chanvillard G. A coupled Nanoindentation/SEM-EDS study on low water/cement ratio Portland cement paste: evidence for C-S-H/Ca(OH)2Nanocomposites. J Am Ceram Soc 2010; 93(5):1484-93.

[10]

Davydov D, Jirásek M, Kopecky´ L. Critical aspects of nano-indentation technique in application to hardened cement paste. Cem Concr Res 2011; 41 (1):20-9.

[11]

Frech-Baronet J, Sorelli L, Chen Z. Effect of plasticity on the experimental characterization of the creep property of a cement paste: spherical vs. conical microindentation. Cem Concr Res 2021; 141:106290.

[12]

Hay R, Li J, Celik K. Influencing factors on micromechanical properties of calcium (alumino) silicate hydrate C-(A-)S-H under nanoindentation experiment. Cem Concr Res 2020; 134:106088.

[13]

Trtik P, Kaufmann J, Volz U. On the use of peak-force tapping atomic force microscopy for quantification of the local elastic modulus in hardened cement paste. Cem Concr Res 2012; 42(1):215-21.

[14]

Asgari H, Ramezanianpour AA, Butt HJ. Nano-mechanical behavior of calcium silicate hydrate and calcium hydroxide in cement paste: elevated peak-force study. Int J Civ Eng 2018; 16(3):273-80.

[15]

Ruiz-Cabello FJM, Trefalt G, Csendes Z, Sinha P, Oncsik T, Szilagyi I, et al. Predicting aggregation rates of colloidal particles from direct force measurements. J Phys Chem B 2013; 117(39):11853-62.

[16]

Yang T, Holzer L, Kagi R, Winnefeld F, Keller B. In situ nanomanipulators as a tool to separate individual tobermorite crystals for AFM studies. Ultramicroscopy 2007; 107(10-11):1068-77.

[17]

Li J, Zhang W, Monteiro PJM. Preferred orientation of calcium aluminosilicate hydrate compacts: implications for creep and indentation. Cem Concr Res 2021; 143:106371.

[18]

Li J, Zhang W, Monteiro PJM. Structure and intrinsic mechanical properties of nanocrystalline calcium silicate hydrate. ACS Sustain Chem Eng 2020; 8 (33):12453-61.

[19]

Li J, Zhang W, Garbev K, Beuchle G, Monteiro PJM. Influences of cross-linking and Al incorporation on the intrinsic mechanical properties of tobermorite. Cem Concr Res 2020; 136:106170.

[20]

Geng G, Myers RJ, Qomi MJA, Monteiro PJM. Densification of the interlayer spacing governs the nanomechanical properties of calcium-silicate-hydrate. Sci Rep 2017; 7(1):10986.

[21]

Geng G, Myers RJ, Li J, Maboudian R, Carraro C, Shapiro DA, et al. Aluminuminduced dreierketten chain cross-links increase the mechanical properties of nanocrystalline calcium aluminosilicate hydrate. Sci Rep 2017; 7(1):44032.

[22]

Geng G, Vasin RN, Li J, Qomi MJA, Yan J, Wenk HR, et al. Preferred orientation of calcium aluminosilicate hydrate induced by confined compression. Cem Concr Res 2018; 113:186-96.

[23]

Monteiro PJM, Geng G, Marchon D, Li J, Alapati P, Kurtis KE, et al. Advances in characterizing and understanding the microstructure of cementitious materials. Cem Concr Res 2019; 124:105806.

[24]

Yang R, Li J. Probing the intrinsic mechanical properties of artinite using highpressure X-ray diffraction for reactive MgO cement. Cem Concr Res 2025; 194:107894.

[25]

Li J, Zhang W, Monteiro PJM. Mechanical properties of struvite-K: a highpressure X-ray diffraction study. Cem Concr Res 2020; 136:106171.

[26]

Qoku E, Xu K, Li J, Monteiro PJM, Kurtis KE. Advances in imaging, scattering, spectroscopy, and machine learning-aided approaches for multiscale characterization of cementitious systems. Cem Concr Res 2023; 174:107335.

[27]

Hu Z, Hilaire A, Ston J, Wyrzykowski M, Lura P, Scrivener K. Intrinsic viscoelasticity of C-S-H assessed from basic creep of cement pastes. Cem Concr Res 2019; 121:11-20.

[28]

Nguyen DT, Alizadeh R, Beaudoin JJ, Raki L. Microindentation creep of secondary hydrated cement phases and C-S-H. Mater Struct 2012; 46 (9):1519-25.

[29]

Plassard C, Lesniewska E, Pochard I, Nonat A. Investigation of the surface structure and elastic properties of calcium silicate hydrates at the nanoscale. Ultramicroscopy 2004; 100(3-4):331-8.

[30]

Thomas JJ, Jennings HM, Allen AJ. Relationships between composition and density of tobermorite, jennite, and nanoscale CaO-SiO2-H2O. J Phys Chem C Nanomater Interfaces 2010; 114(17):7594-601.

[31]

Shahsavari R, Buehler MJ, Pellenq RJM, Ulm FJ. First-principles study of elastic constants and interlayer interactions of complex hydrated oxides: case study of tobermorite and jennite. J Am Ceram Soc 2009; 92(10):2323-30.

[32]

Pelisser F, Gleize PJP, Mikowski A. Effect of the Ca/Si molar ratio on the micro/nanomechanical properties of synthetic C-S-H measured by nanoindentation. J Phys Chem C Nanomater Interfaces 2012; 116 (32):17219-27.

[33]

Alizadeh R, Beaudoin JJ, Raki L. Mechanical properties of calcium silicate hydrates. Mater Struct 2011; 44(1):13-28.

[34]

Pourbeik P, Beaudoin JJ, Alizadeh R, Raki L. Mechanical property-porosity relationships of layered calcium silicate hydrate phases. Mater Struct 2013; 46(9):1489-95.

[35]

Abdolhosseini Qomi MJ, Krakowiak KJ, Bauchy M, Stewart KL, Shahsavari R, Jagannathan D, et al. Combinatorial molecular optimization of cement hydrates. Nat Commun 2014; 5(1):4960.

[36]

Manzano H, Dolado J, Ayuela A. Elastic properties of the main species present in Portland cement pastes. Acta Mater 2009; 57(5):1666-74.

[37]

Dharmawardhana CC, Misra A, Aryal S, Rulis P, Ching WY. Role of interatomic bonding in the mechanical anisotropy and interlayer cohesion of CSH crystals. Cem Concr Res 2013; 52:123-30.

[38]

Zhang Y, Guo L, Shi J, Luo Q, Jiang J, Hou D. Full process of calcium silicate hydrate decalcification: molecular structure, dynamics, and mechanical properties. Cem Concr Res 2022; 161:106964.

[39]

Yaphary YL, Lau D, Sanchez F, Poon CS. Effects of sodium/calcium cation exchange on the mechanical properties of calcium silicate hydrate (C-S-H). Constr Build Mater 2020; 243:118283.

[40]

Liu L, Sun C, Geng G, Feng P, Li J, Dähn R. Influence of decalcification on structural and mechanical properties of synthetic calcium silicate hydrate (CS- H). Cem Concr Res 2019; 123:105793.

[41]

Zhu X, Richardson IG. Morphology-structural change of C-A-S-H gel in blended cements. Cem Concr Res 2023; 168:107156.

[42]

Harris M, Simpson G, Scrivener K, Bowen P. A method for the reliable and reproducible precipitation of phase pure high Ca/Si ratio (> 1.5) synthetic calcium silicate hydrates (C-S-H). Cem Concr Res 2022; 151:106623.

[43]

L′Hôpital E, Lothenbach B, Kulik D, Scrivener K. Influence of calcium to silica ratio on aluminium uptake in calcium silicate hydrate. Cem Concr Res 2016; 85:111-21.

[44]

Wilson W, Sorelli L, Tagnit-Hamou A. Unveiling micro-chemo-mechanical properties of C-(A)-S-H and other phases in blended-cement pastes. Cem Concr Res 2018; 107:317-36.

[45]

Li J, Zhang W, Sanz-Camacho P, Duttine M, Gardner D, Carraro C, et al. The nanomechanical properties of non-crosslinked calcium aluminosilicate hydrate: the influences of tetrahedral Al and curing age. Cem Concr Res 2022; 159:106900.

[46]

Ye H. Creep mechanisms of calcium-silicate-hydrate: an overview of recent advances and challenges. Int J Concr Struct Mater 2015; 9(4):453-62.

[47]

Alizadeh R, Beaudoin JJ, Raki L. Viscoelastic nature of calcium silicate hydrate. Cem Concr Compos 2010; 32(5):369-76.

[48]

Gardner DW, Li J, Kunz M, Zhu C, Monteiro PJM, Maboudian R, et al. Plastic deformation mechanism of calcium-silicate hydrates determined by deviatoric-stress Raman spectroscopy. Cem Concr Res 2021; 146:106476.

[49]

Pignatelli I, Kumar A, Alizadeh R, Le Pape Y, Bauchy M, Sant G. A dissolutionprecipitation mechanism is at the origin of concrete creep in moist environments. J Chem Phys 2016; 145(5):054701.

[50]

Wyrzykowski M, Scrivener K, Lura P. Basic creep of cement paste at early agethe role of cement hydration. Cem Concr Res 2019; 116:191-201.

[51]

Tamtsia BT, Beaudoin JJ, Marchand J. The early age short-term creep of hardening cement paste: load-induced hydration effects. Cem Concr Compos 2004; 26(5):481-9.

[52]

Bisschop J. Does applied stress affect Portland cement hydration? In: Proceedings of the 13th International Congress Chemistry Cement; 2011 Jul 3-8; Madrid, Spain. Duesseldorf: ICCC Permanent Secretariat; 2011. p. 3-8.

[53]

Tang J, Gao C, Li Y, Xu J, Huang J, Xu D, et al. A review on multi-scale toughening and regulation methods for modern concrete: from theory to practical engineering application. Research 2024; 7:0518.

[54]

de Souza FB, Shamsaei E, Chen SJ, Sagoe-Crentsil K, Duan WH. Controlled growth and ordering of poorly-crystalline calcium-silicate-hydrate nanosheets. Commun Mater 2021; 2:84.

[55]

Hartgerink JD, Beniash E, Stupp SI. Self-assembly and mineralization of peptide-amphiphile nanofibers. Science 2001; 294(5547):1684-8.

[56]

Gao HL, Chen SM, Mao LB, Song ZQ, Yao HB, Cölfen H, et al. Mass production of bulk artificial nacre with excellent mechanical properties. Nat Commun 2017; 8:287.

[57]

Picker A, Nicoleau L, Burghard Z, Bill J, Zlotnikov I, Labbez C, et al. Mesocrystalline calcium silicate hydrate: a bioinspired route toward elastic concrete materials. Sci Adv 2017; 3(4):e1701216.

[58]

Han B, Zhang L, Zeng S, Dong S, Yu X, Yang R, et al. Nano-core effect in nanoengineered cementitious composites. Compos A Appl Sci Manuf 2017; 95:100-9.

[59]

Wang X, Dong S, Li Z, Han B, Ou J. Nanomechanical characteristics of interfacial transition zone in nano-engineered concrete. Engineering 2022; 17:99-109.

[60]

Wang J, Han B, Li Z, Yu X, Dong X. Effect investigation of nanofillers on C-S-H gel structure with Si NMR. J Mater Civ Eng 2019; 31(8):04018352.

[61]

Zhou Y, Orozco CA, Duque-Redondo E, Manzano H, Geng G, Feng P, et al. Modification of poly(ethylene glycol) on the microstructure and mechanical properties of calcium silicate hydrates. Cem Concr Res 2019; 115:20-30.

[62]

Zhou Y, Tang L, Liu J, Miao C. Interaction mechanisms between organic and inorganic phases in calcium silicate hydrates/poly(vinyl alcohol) composites. Cem Concr Res 2019; 125:105891.

[63]

Wang P, Qiao G, Guo Y, Zhang Y, Hou D, Jin Z, et al. Molecular dynamics simulation of the interfacial bonding properties between graphene oxide and calcium silicate hydrate. Constr Build Mater 2020; 260:119927.

[64]

Gao Y, Jing HW, Wu JY, Fu GP, Feng CD, Chen WQ. Molecular dynamics study on the influence of graphene oxide on the tensile behavior of calcium silicate hydrate composites. Mater Chem Phys 2022; 292:126793.

[65]

Li Y, Li H, Wang Z, Jin C. Effect and mechanism analysis of functionalized multi-walled carbon nanotubes (MWCNTs) on C-S-H gel. Cem Concr Res 2020; 128:105955.

[66]

Qin R, Zhou A, Yu Z, Wang Q, Lau D. Role of carbon nanotube in reinforcing cementitious materials: an experimental and coarse-grained molecular dynamics study. Cem Concr Res 2021; 147:106517.

[67]

Li J. Municipal solid waste incineration ash-incorporated concrete: one step towards environmental justice. Buildings 2021; 11(11):495.

[68]

Jennings HM. Colloid model of C-S-H and implications to the problem of creep and shrinkage. Mater Struct 2004; 37(265):59-70.

[69]

Jennings HM. Refinements to colloid model of C-S-H in cement: CM-II. Cem Concr Res 2008; 38(3):275-89.

[70]

Jennings HM. A model for the microstructure of calcium silicate hydrate in cement paste. Cem Concr Res 2000; 30(1):101-16.

[71]

Richardson IG. The calcium silicate hydrates. Cem Concr Res 2008; 38:137-58.

[72]

Kunhi Mohamed A, Moutzouri P, Berruyer P, Walder BJ, Siramanont J, Harris M, et al. The atomic-level structure of cementitious calcium aluminate silicate hydrate. J Am Chem Soc 2020; 142(25):11060-71.

[73]

Taylor HFW. Cement chemistry. 2nd ed. London: Thomas Telford Publishing; 1997.

[74]

Richardson IG. Tobermorite/jennite- and tobermorite/calcium hydroxidebased models for the structure of C-S-H: applicability to hardened pastes of tricalcium silicate, b-dicalcium silicate, Portland cement, and blends of Portland cement with blast-furnace slag, metakaolin, or silica fume. Cem Concr Res 2004; 34(9):1733-77.

[75]

Pardal X, Brunet F, Charpentier T, Pochard I, Nonat A. 27Al and 29Si solid-state NMR characterization of calcium-aluminosilicate-hydrate. Inorg Chem 2012; 51(3):1827-36.

[76]

Kumar A, Walder BJ, Kunhi Mohamed A, Hofstetter A, Srinivasan B, Rossini AJ, et al. The atomic-level structure of cementitious calcium silicate hydrate. J Phys Chem C 2017; 121(32):17188-96.

[77]

Myers RJ, Bernal SA, San Nicolas R, Provis JL. Generalized structural description of calcium-sodium aluminosilicate hydrate gels: the crosslinked substituted tobermorite model. Langmuir 2013; 29(17):5294-306.

[78]

Myers RJ, Bernal SA, Gehman JD, van Deventer JSJ, Provis JL, Struble L. The role of Al in cross-linking of alkali-activated slag cements. J Am Ceram Soc 2015; 98(4):996-1004.

[79]

Ortaboy S, Li J, Geng G, Myers RJ, Monteiro PJM, Maboudian R, et al. Effects of CO2 and temperature on the structure and chemistry of C-(A-)S-H investigated by Raman spectroscopy. RSC Adv 2017; 7(77):48925-33.

[80]

Myers RJ, L′Hôpital E, Provis JL, Lothenbach B. Effect of temperature and aluminium on calcium (alumino) silicate hydrate chemistry under equilibrium conditions. Cem Concr Res 2015; 68:83-93.

[81]

L’Hôpital E, Lothenbach B, Le Saout G, Kulik D, Scrivener K. Incorporation of aluminium in calcium-silicate-hydrates. Cem Concr Res 2015; 75:91-103.

[82]

Renaudin G, Russias J, Leroux F, Frizon F, Cau-dit-Coumes C. Structural characterization of C-S-H and C-A-S-H samples-part I: long-range order investigated by Rietveld analyses. J Solid State Chem 2009; 182(12):3312-9.

[83]

Powers TC, Brownyard TL. Studies of the physical properties of hardened Portland cement paste. ACI J Proc 1946; 43(9):101-32.

[84]

Feldman RF, Sereda PJ. A new model for hydrated Portland cement and its practical implications. Eng J 1970; 53(8):53-9.

[85]

Wittmann F. The structure of hardened cement paste-a basis for a better understanding of the materials properties. In:Proceedings of the Conference on Hydraulic Cement Pastes: Their Structure and Properties; 1976 Apr 8-9; Sheffield, UK. Slough: Cement and Concrete Association; 1976. p. 96-117.

[86]

Muller ACA, Scrivener KL, Gajewicz AM, McDonald PJ. Densification of C-S-H measured by 1H NMR relaxometry. J Phys Chem C 2012; 117(1):403-12.

[87]

Matsuyama H, Young JF. Effects of pH on precipitation of quasi-crystalline calcium silicate hydrate in aqueous solution. Adv Cem Res 2000; 12(2):29-33.

[88]

Richardson IG. Model structures for C-(A)-S-H(I). Acta Crystallogr B Struct Sci Cryst Eng Mater 2014; 70(Pt 6):903-23.

[89]

Geng G, Myers RJ, Yu YS, Shapiro DA, Winarski R, Levitz PE, et al. Synchrotron Xray nanotomographic and spectromicroscopic study of the tricalciumaluminate hydration in the presence of gypsum. Cem Concr Res 2018; 111:130-7.

[90]

Li J, Geng G, Myers R, Yu YS, Shapiro D, Carraro C, et al. The chemistry and structure of calcium (alumino) silicate hydrate: a study by XANES, ptychographic imaging, and wide- and small-angle scattering. Cem Concr Res 2019; 115:367-78.

[91]

Geng G, Li J, Zhou Y, Liu L, Yan J, Kunz M, et al. A high-pressure X-ray diffraction study of the crystalline phases in calcium aluminate cement paste. Cem Concr Res 2018; 108:38-45.

[92]

Kantor I, Prakapenka V, Kantor A, Dera P, Kurnosov A, Sinogeikin S, et al. BX90: a new diamond anvil cell design for X-ray diffraction and optical measurements. Rev Sci Instrum 2012; 83(12):125102.

[93]

Zhang W, Li J. The intrinsic mechanical properties of hydromagnesite, Mg5(CO3)4(OH)2·4H2O, a key phase of reactive MgO carbonate cement. Cem Concr Res 2024; 177:107410.

[94]

Prescher C, Prakapenka VB. DIOPTAS: a program for reduction of twodimensional X-ray diffraction data and data exploration. High Press Res 2015; 35(3):223-30.

[95]

Cheary RW, Coelho AA. Programs XFIT and FOURYA, deposited in CCP14 powder diffraction library. Warrington: Engineering and Physical Sciences Research Council, Daresbury Laboratory; 1996.

[96]

Laugier J, Bochu B. Computer code CELREF. Version 3. Cell parameter refinement program from powder diffraction diagram. Paris: Laboratoire des Matériaux et du Génie Physique (LMGP), Ecole Nationale Supérieure de Physique de Grenoble (INPG); 2002.

[97]

Birch F. Elasticity and constitution of the Earth’s interior. J Geophys Res 1952; 57(2):227-86.

[98]

Kalousek GL, Prebus AF. Crystal chemistry of hydrous calcium silicates: III, morphology and other properties of tobermorite and related phases. J Am Ceram Soc 1958; 41(4):124-32.

[99]

Yang J, Hou D, Ding Q. Structure, dynamics, and mechanical properties of cross-linked calcium aluminosilicate hydrate: a molecular dynamics study. ACS Sustain Chem Eng 2018; 6(8):9403-17.

[100]

Zhang G, Li Y, Yang J, Ding Q, Sun D. Insight into the strengthening mechanism of the Al-induced cross-linked calcium aluminosilicate hydrate gel: a molecular dynamics study. Front Mater 2021; 7:484.

[101]

Gardner DW, Li J, Kunz M, Liao WC, Zhu C, Carraro C, et al. Uniaxial stress increases layer stacking disorder in calcium silicate hydrates with low calcium content. Acta Mater 2023; 246:118726.

[102]

Gardner DW, Li J, Morshedifard A, Masoumi S, Abdolhosseini Qomi MJ, Monteiro PJM, et al. Silicate bond characteristics in calcium-silicate-hydrates determined by high pressure Raman spectroscopy. J Phys Chem C 2020; 124 (33):18335-45.

[103]

Chai M, Brown JM. Effects of static non-hydrostatic stress on the R lines of ruby single crystals. Geophys Res Lett 1996; 23(23):3539-42.

[104]

Wang J, Yang R, Sanz-Camacho P, Duttine M, Huthwelker T, Li J. Aluminum incorporation lowers the intrinsic mechanical properties of single-chain calcium silicate hydrate. Cem Concr Res 2026; 199:108026.

[105]

Yan Y, Yang SY, Miron GD, Collings IE, L′Hôpital E, Skibsted J, et al. Effect of alkali hydroxide on calcium silicate hydrate (C-S-H). Cem Concr Res 2022; 151:106636.

[106]

Kirkpatrick RJ, Yarger JL, McMillan PF, Ping Y, Cong X. Raman spectroscopy of C-S-H, tobermorite, and jennite. Adv Cem Bas Mat 1997; 5(3-4):93-9.

[107]

Li J, Zhang W. Preferred orientation of calcium silicate hydrate and its implication to concrete creep. Compos B Eng 2022; 247:110297.

[108]

Georget F, Wilson W, Scrivener KL. edxia: Microstructure characterisation from quantified SEM-EDS hypermaps. Cem Concr Res 2021; 141:106327.

[109]

Krakowiak KJ, Wilson W, James S, Musso S, Ulm FJ. Inference of the phase-tomechanical property link via coupled X-ray spectrometry and indentation analysis: application to cement-based materials. Cem Concr Res 2015; 67:271-85.

[110]

Mejdi M, Wilson W, Saillio M, Chaussadent T, Divet L, Tagnit-Hamou A. Quantifying glass powder reaction in blended-cement pastes with the Rietveld-PONKCS method. Cem Concr Res 2020; 130:105999.

[111]

Wilson W, Rivera-Torres JM, Sorelli L, Durán-Herrera A, Tagnit-Hamou A. The micromechanical signature of high-volume natural pozzolan concrete by combined statistical nanoindentation and SEM-EDS analyses. Cem Concr Res 2017; 91:1-12.

[112]

Gao H, Shikhov I, Hamed E, Hajimohammadi A, Al-Damad I, Arns C, et al. New insights on the basic creep mechanism of one-part alkali activated slag and fly ash paste. Cem Concr Res 2024; 186:107691.

[113]

Wang JW, Kalinichev AG, Kirkpatrick RJ, Hou XQ. Molecular modeling of the structure and energetics of hydrotalcite hydration. Chem Mater 2001; 13 (1):145-50.

[114]

Basahel SN, Al-Thabaiti SA, Narasimharao K, Ahmed NS, Mokhtar M. Nanostructured Mg-Al hydrotalcite as catalyst for fine chemical synthesis. J Nanosci Nanotechnol 2014; 14(2):1931-46.

[115]

López-Muñoz BE, Iturbe-García JL, Olguin MT. Physicochemical properties of layered double hydroxides by-product obtained from Al-MgO mechanochemical process to produce hydrogen. Chem Pap 2018; 73 (2):415-24.

[116]

Li J. Characterization of calcium (alumino) silicate hydrate phases with synchrotron-based techniques [dissertation]. Berkeley: University of California, Berkeley; 2019.

[117]

Li J, Zheng Q. The first experimental evidence for improved nanomechanical properties of calcium silicate hydrate by polycarboxylate ether and graphene oxide. Cem Concr Res 2022; 156:106787.

[118]

Han B, Ding S, Wang J, Ou J. Nano-engineered cementitious composites: principles and practices. Berlin: Springer; 2019.

PDF (5536KB)

1601

Accesses

0

Citation

Detail

Sections
Recommended

/

〈 〉