Balancing Strength and Extreme Thermal Resilience in Lunar Regolith Composites: The Role of Multi‐Walled Carbon Nanotubes

Abstract

The National Aeronautics and Space Administration's (NASA) Artemis program stands at the forefront of commercial space initiatives, aiming to establish sustainable lunar habitats, demanding resilient construction materials with minimal reliance on Earth‐based resources. In response to these demands, this study explores the feasibility of reinforcing lunar regolith with multi‐walled carbon nanotubes (MWCNTs) while relying on minimal water and additives for composite fabrication as a potential solution for building semi‐permanent Moon bases. Composites incorporating 0.00, 0.25, 0.50, and 1.00 wt.% MWCNTs are subjected to freeze‐thaw cycles, vacuum pressures, ambient environments, and oven‐curing methods to emulate the Moon's harsh environment. Results show that ambient‐cured composites containing MWCNTs achieve compressive strengths exceeding 35 MPa, resulting in a 44.44% increase compared to the sample without MWCNTs, highlighting the reinforcing potential of carbon nanotubes (CNTs) for extraterrestrial applications. However, thermal cycling reveals performance limitations due to mismatched coefficients of thermal expansion between MWCNTs and the regolith matrix, causing microcracking. In contrast, vacuum‐cured MWCNT‐free samples surpass 45 MPa, indicating that curing protocols can significantly influence densification and mechanical properties. These findings underscore the trade‐offs between material composition, curing approaches, and thermal stability, offering key insights into designing robust, resource‐efficient lunar construction.

This study investigates the impact of MWCNT reinforcement on the compressive strength of lunar regolith composites under varied curing conditions, including ambient with and without cycling and vacuum with cycling. Utilizing experimental methodologies, it identifies significant relationships between MWCNT content, curing protocols, and thermal cycling, providing insights into optimizing durability and addressing thermal mismatch for resource‐efficient lunar habitat construction.

1. Introduction

Establishing sustainable habitats on the Moon or other celestial bodies requires overcoming numerous challenges, particularly in terms of resource management and material sourcing. Extraterrestrial construction faces significant resource constraints primarily due to the high cost of transporting materials. According to NASA, it costs ≈$10 000 to send one pound of material into orbit,^{[ 1 ]} making sourcing local materials essential for lunar habitats. Due to this high cost, utilizing lunar regolith as a primary construction material becomes an important strategy as it can be excavated directly from the surface of the Moon. Additionally, the discovery of ice at the Moon's south pole further enhances the potential for in situ resource utilization, as water serves multiple life‐support^{[ 2 ]} and various material applications. This has driven efforts to efficiently utilize limited water supplies, including combining regolith with acidic or alkaline solutions to create lunar concrete‐like materials.^{[ 3 ]} Another strategy to lower off‐site material needs is to combine regolith with superplasticizers to lower water needs.^{[ 4 ]} Urea, which serves as an effective superplasticizer, can be locally sourced from a human's urine, sweat, and tears.^{[ 5 ]} Urea, disrupts hydrogen bonds, decreasing the viscosity of composite mixtures and improving their workability while minimizing water requirements. Similarly, attempts to 3D printing habitats using regolith mixtures^{[ 6 ]} heavily depend on the workability of the materials used, as they must flow easily during the printing process while maintaining sufficient stability to form complex structures. Pilehvar et al.^{[ 7 ]} found that the addition of urea in lunar regolith composites reduced the water needed for workability by up to 32%. However, researchers also found a direct relationship between increasing urea content and decreasing compressive strength under both vacuum and ambient pressure conditions. The results showed promise to decrease the water needs but to advance further, it is essential to develop materials that have a higher compressive strength.

Beyond the water usage, the harsh environmental condition on the Moon introduces additional engineering challenges, such as extreme temperature fluctuations ranging from ≈114 to −171 °C, radiation exposure, and micrometeorite impacts.^{[ 8 ]} Therefore, materials must be capable of withstanding harsh conditions with extreme temperature fluctuations, by providing effective thermal insulation, while offering protection against physical and environmental hazards. A study by Zhang et al.^{[ 9 ]} identified a 72‐h curing scheme as suitable for lunar pavement construction, with compressive strengths reaching 31.2 MPa. However, representative cycling tests caused notable deterioration, including flexural strength reductions of up to 70%, microcrack formation, and increased porosity. These findings emphasize the importance of addressing not only vacuum conditions but also the repeated thermal stresses inherent to the lunar surface.

Identifying and implementing strategies to ensure mechanical properties while minimizing the detrimental effects of extreme environmental conditions is essential for space‐based construction. In designing composites for space applications, there is often a balance between structural performance and payload mass. A compressive strength of ≈6 MPa is typically adequate for lunar architecture due to lower gravity (one‐sixth of Earth's),^{[ 10 ]} but achieving higher strength, comparable to the 35 MPa needed for terrestrial structures,^{[ 11 ]} is desirable to ensure safety in the extreme lunar condition. Existing literature reports lunar regolith composite compressive strengths ranging broadly from 12 to 120 MPa, highlighting significant variability of additives and potential for further optimization.^{[ 12 ]} Polymers, for instance, are unsuitable for space due to the cost of transportation, low thermal stability, and radiation sensitivity.^{[ 13 ]} However, magnesium^{[ 14 ]} is beneficial for sintering but is less practical for near‐room‐temperature curing conditions. CNTs on the other hand, offer significant advantages compared to other additives with their exceptional tensile/compressive strength,^{[ 15 ]} superior stiffness, enhanced structural integrity, and compatibility with 3D printing. CNTs have been widely used to augment various composite materials' mechanical properties and environmental resistance, such as aircraft components and concrete composites.^{[ 16} , ¹⁷ , ¹⁸ , ¹⁹ , ^{20 ]} Bodnarova et al.^{[ 21 ]} found that adding CNTs into concrete can effectively improve its mechanical properties and durability. Enhanced CNT dispersion improved mechanical properties by filling internal pores, reducing porosity, and strengthening the interface transition zone, resulting in greater compressive strength by 11.9% and 17.30% flexural strength, frost resistance, and impermeability, even after 100 freeze‐thaw cycles. Since this study uses a very low concentration of CNTs (0.0015 and 0.012 wt.%), the impact on the matrix's overall thermal expansion properties is minimal, avoiding significant internal stresses during temperature changes. Several studies have highlighted the advantages of using CNTs in space applications due to their exceptional strength‐to‐weight ratio.^{[ 22 ]} Notably, Garcia et al.^{[ 23 ]} fabricated highly dense lunar regolith composites by applying 250 MPa of pressure in a pellet die and using CNT additives. The resulting compression properties enhanced from ≈15 MPa using 0.60 wt.% CNT to 100 MPa using 7.00 wt.% CNT. However, since this study did not include lunar cycle temperature cycling, which is crucial for space habitation, the true impact of CNT additives under these conditions remains uncertain. It is also anticipated that applying a pressure of 250 MPa, similar to that used by Garcia et al., would result in a significant increase in the strength observed in this study which only uses a pressure of 8 MPa due to equipment limitations. Despite the potential benefits, such high pressures are highly unfeasible for space applications.

Although prior research has demonstrated impressive compressive strengths using advanced matrices, many of these approaches rely on high water content (up to 30 wt.%),^{[ 24 ]} overlook temperature cycling tests,^{[ 25} , ^{26 ]} or provide limited chemical analysis which are factors that limit their feasibility for lunar construction.^{[ 27 ]} To address these gaps, the present study introduces a novel synthesis method that provides a more sustainable and efficient approach for constructing future lunar habitats. This method integrates MWCNTs into lunar regolith composites while minimizing water usage to 12.48 wt.%. Additionally, it employs a particle size tuning strategy that combines an 80:20 ratio of fine‐to‐coarse regolith fractions to optimize the microstructure. MWCNTs are also paired with urea, and the research uniquely adopts an acidic formulation. Central to this study is its emphasis on the lunar environment, where stability is assessed through the effects of thermal cycling, vacuum pressure curing, and varying MWCNT weight percentages. By examining these factors' influence on microstructure and compressive strength, the study aims to develop regolith‐based composites capable of achieving compressive strengths exceeding 35 MPa while retaining resilience under extreme lunar conditions. This work provides a comprehensive chemical, experimental, and simulation‐based analysis of the benefits of CNTs for regolith composites, while also identifying areas for improvement. The proposed mechanisms offer valuable insights to guide future research directions and advance the development of optimized regolith‐based materials for lunar habitats.

2. Results and Discussion

2.1. Mechanical Strength Enhancement with MWCNT Inclusion without Temperature Cycling

For the purpose of examining the lunar conditions' influence on the regolith composites, this study follows three distinct processing routes that are evaluated with different preconditions (Figure 1a). Route 1 serves as the baseline for comparison consisting of ambient curing without temperature cycling. Route 2 combines ambient curing with four temperature cycles simulating extreme day‐night fluctuations on the lunar surface (Figure 1b). The temperature gradually increases from 80 °C to a peak of 117 °C, then decreases to a minimum of −80 °C before returning to 80 °C, completing one full cycle. Route 3 starts with vacuum curing, followed by the same thermal cycling as in Route 2, thereby enabling an evaluation of how reduced atmospheric pressure and extreme temperature variations jointly influence composite performance. Unlike the Moon's continuous thermal gradient, we cycled between an oven, refrigerator, and ultra‐freezer, which may have caused inconsistencies in thermal exposure. Real lunar temperatures at night were ≈−171 °C, but due to equipment limits, we were only able to reach −80 °C.

Figure 1.

a) Schematic illustration of the three distinct curing routes developed for lunar regolith composites, tailored to address the feasibility of this material for off‐Earth construction, b) Comparison of the experimental temperature cycle with the simulated lunar cycle.

In exploring the different processing routes with varying MWCNT concentrations, there is a focus on a more accessible processing approach using room temperature pre‐curing conditions and low curing conditions (≈100 °C), in contrast to the high‐temperature sintering processes (>1000 °C) typically used in existing research.^{[ 28} , ²⁹ , ^{30 ]} The composition of the composite is illustrated in Figure 2a. A key aspect of the processing is the regolith method of mixing fine (<0.25 µm) and coarser (0.25–250 µm) regolith particles (Figure 2b). Preliminary experiments identified an optimal fine‐to‐coarse particle ratio of 80:20, which achieved the highest compressive strength of 32.57 MPa. In this mixture, fine particles constituted 80% of the total mass, while coarse particles made up the remaining 20%. Alternative ratios of 70:30 and 50:50 yielded lower compressive strengths of 24.07 and 22.65 MPa, respectively. The enhanced performance of the 80:20 ratio is attributed to the fine particles effectively filling the voids between larger particles, thereby reducing porosity and promoting a denser packing structure. This denser matrix improves compressive strength by minimizing internal defects and enhancing load transfer. Simultaneously, the inclusion of larger particles serves to mitigate shrinkage and cracking during curing by acting as a structural skeleton that resists the stresses induced during drying. This synergistic interaction not only minimizes voids and increases packing density but also reduces binder demand while improving strength, impermeability, and volumetric stability.^{[ 31} , ^{32 ]} By leveraging natural particle gradation, this scalable strategy offers a novel contribution to regolith‐based composite research.

Figure 2.

Overview of the 20 mm height and 20 mm diameter composite formulation: a) Chemical composition with weight percentages of each component, b) Particle size tuning strategy highlighting the complementary roles of fine (<25 µm) and coarse (<250 µm) particles in enhancing packing density and mechanical performance, c) Synergistic integration of urea and carbon nanotubes (CNTs) to simultaneously increase compressive strength and reduce water requirements, and d) Use of phosphoric acid as an acidic solution to enhance the chemical stability of the composite.

Beyond the particle size optimization, this research further integrates urea and MWCNTs to reduce water requirements while enhancing the strength of lunar regolith composites (Figure 2c). Unlike prior studies focusing on urea^{[ 33 ]} or CNTs^{[ 34 ]} individually, this approach leverages their synergistic effects. Urea acts as a superplasticizer reducing water content to 12.48 wt.%, lower than typical lunar concrete formulations (5–30 wt.%).^{[ 23} , ^{24 ]} While urea reduces water usage, it causes a trade‐off with strength, which was offset by adding MWCNTs to enhance compressive properties. Additionally, an acidic solution with phosphoric acid (H₃PO₄) was chosen as the pH modifier over the more commonly used sodium hydroxide due to its milder reactivity (Figure 2d). Phosphoric acid contributes hydrogen donor molecules that form strong chemical bonds with regolith compounds, stabilizing the composite structure.^{[ 35 ]} While sodium hydroxide's high pH facilitates the breakdown of silicates and metal oxides to form sodium silicate, phosphoric acid exhibits milder reactivity, dissolving some minerals and forming phosphate salts without overly aggressive reactions. Its stability, particularly at low temperatures, ensures long‐term durability and resistance to extreme lunar temperature fluctuations.

In order to validate the material modification and particle size optimization, the compressive strength, compressive modulus, and chemical characteristics of lunar regolith composites were evaluated under ambient curing conditions without thermal cycling (Route 1), as illustrated in Figure 3 . These experiments establish a baseline understanding of the effects of MWCNT incorporation on the compressive performance of composites. The compressive strength (Figure 3a) and compressive modulus (Figure 3b) results for Route 1 demonstrated a clear enhancement with the inclusion of MWCNTs. The sample with 0.50 wt.% MWCNT exhibited the highest compressive strength, achieving 39 MPa (44.44% increase), and a compressive modulus of 1.98 GPa (41.43% increase) compared to the 27 MPa compressive strength and 1.4 GPa compressive modulus of the 0.00 wt.% MWCNT sample. Additionally, the sample with 0.25 wt.% MWCNT showed a marked increase in compressive strength with 31 MPa (14.81% increase) and 1.8 GPa compressive modulus, indicating the effectiveness of MWCNT additives in reinforcing the composite in comparison with the MWCNT‐free sample. However, when the MWCNT content was increased to 1.00 wt.%, even the best‐performing sample showed a slight decrease in compressive strength to 38 MPa and a lower compressive modulus of 1.5 GPa. In contrast, the best‐performing 0.50 wt.% MWCNT sample achieved a higher compressive strength of 39 MPa and a higher compressive modulus of 1.98 GPa. Additionally, the 1.00 wt.% specimens exhibited larger margins of error in both compressive strength and modulus, as shown in Figure 3b, indicating greater variability in performance. These results suggest that increasing MWCNT content beyond 0.50 wt.% does not improve mechanical properties and may even reduce overall performance. This slight decrease indicates poor interfacial bonding between the CNTs and the matrix material.^{[ 36 ]} When an excessive amount of CNTs disrupts the balance of distribution between CNTs and regolith particles, it negatively impacts and weakens the interfacial bonding between them, ultimately resulting in a decline in compressive strength.

Figure 3.

a) Maximum recorded compressive strength values. b) Average compressive strength and compressive modulus for ambient‐cured composites (Route 1), with error bars representing the standard deviation c) XRD spectroscopy for pure MWCNT, 0.00, and 1.00 wt.% MWCNT composites fabricated under Route 1 d) FTIR spectra for pure MWCNT, 0.00, and 1.00 wt.% MWCNT composites fabricated under Route 1.

To understand the compressive strength improvement achieved by incorporating MWCNTs into the composite under the ambient temperature curing method, an XRD technique was used as illustrated in Figure 3c. The characteristic peaks of MWCNTs observed at 23° and 30° correspond to graphite diffraction,^{[ 37 ]} confirming that the hexagonal graphite structure of the MWCNTs remains intact across all samples.^{[ 38 ]} The broad nature of these peaks at 23° and 30° reflects the layered arrangement of graphitic planes within the MWCNTs, highlighting the absence of significant long‐range crystalline order. When 1.00 wt.% MWCNT is incorporated into the composite, the XRD patterns reveal a noticeable increase in the intensity of the peaks at 23° and 30° compared to the 0.00 wt.% MWCNT sample. This enhancement suggests that the MWCNTs influence the microstructure of the lunar matrix by improving packing density and promoting structural alignment, due to their high surface area and ability to act as nucleation sites during the curing process. In contrast, the 0.00 wt.% MWCNT sample exhibits broader and lower‐intensity peaks in the same region (23°, 30°), indicative of a less ordered and predominantly amorphous structure, which may be due to the absence of the reinforcing effects provided by the MWCNTs. The results show that the sharper, higher‐intensity peaks observed at 23° and 30° in the XRD patterns correlate with improved compressive strength, indicating that the presence of MWCNTs enhances the composite's structural order and compressive performance under the ambient curing method.

Beyond the microstructural alignment revealed by XRD, the chemical interactions between MWCNTs and the lunar matrix were examined using Fourier‐Transform Infrared Spectroscopy (FTIR) (Figure 3d). This analysis highlights the bonding mechanisms and changes in the chemical environment induced by MWCNT integration, further explaining their dual role in reinforcing and modifying the matrix network. The FTIR spectra reveal that incorporating MWCNTs into cementitious composites enhances the out‐of‐plane flexibility of silica tetrahedra, which can improve the material's adaptability, mechanical performance, and overall durability. This is evidenced by the sharper peak observed ≈500 cm⁻¹ in the MWCNT‐containing sample, indicating a structural modification that increases the mobility of silica tetrahedra. Additionally, the reduced intensity observed in samples containing MWCNTs, compared to the 0.00 wt.% MWCNT sample, suggests that MWCNTs help mitigate carbonation within the composite material. Carbonation, a chemical reaction between carbon dioxide and cement components, can degrade the material's structure over time, and the presence of MWCNTs appears to inhibit this process, enhancing the composite's durability. The presence of MWCNTs seems to interfere with this process, likely by altering the chemical environment, making it harder for carbonation to occur.^{[ 39 ]} Peaks corresponding to specific functional groups further elucidate the nature of these interactions and their impact on the composite properties. In the 1700 cm⁻¹ wavelength, associated with C═O stretching,^{[ 40 ]} slight changes in intensity suggest that the MWCNTs interact with carbonyl‐containing groups through secondary bonding mechanisms, such as hydrogen bonding, given their conjugated structure. These interactions improve the matrix cohesion by facilitating additional bonding sites, which enhance load transfer and stability under compressive loading conditions. Furthermore, notable changes in the Si‐O and Si‐O‐Si peaks (1000–500 cm⁻¹)^{[ 41 ]} highlight the influence of MWCNTs on the siloxane network, due to physical and chemical interactions that disrupt the original network while facilitating new interfacial bonds which enhance the structural integrity. MWCNTs not only act as reinforcing agents but also alter the chemical environment of the matrix, as evidenced by reduced transmittance and variations in peak intensities, highlighting their dual role in disrupting and enhancing the matrix network through both chemical and physical mechanisms. These chemical modifications align closely with the observed improvements in mechanical properties, such as the compressive strength and modulus achieved with MWCNT incorporation. The enhanced bonding and structural alignment facilitated by MWCNTs provide a more robust matrix capable of withstanding greater mechanical loads, as evidenced by the 44.44% increase in compressive strength and 41.43% increase in modulus for samples containing 0.50 wt.% MWCNTs. This synergy between chemical and mechanical enhancements underscores the effectiveness of MWCNTs in reinforcing the lunar regolith composite.

2.2. Thermal Cycling Effects on Compressive Properties

To ensure the suitability of lunar regolith composites for space applications, subjecting them to simulated lunar environments is crucial. Thermal cycling, which mimics the extreme temperature fluctuations on the Moon, plays a crucial role in assessing material reliability. The results from mechanical testing demonstrated significant differences between the ambient‐cured samples (Route 1) and those subjected to thermal cycling (Routes 2 and 3). While ambient‐cured samples exhibited predictable trends in compressive strength and modulus with varying MWCNT concentrations, thermal cycling introduced inconsistencies and revealed critical challenges associated with composite behavior under lunar‐like conditions.

In Route 2, which involved thermal cycling after ambient pre‐curing, the compressive strength (Figure 4a) and modulus (Figure 4b) were inconsistent and showed no clear correlation with MWCNT content. While the 1.00 wt.% MWCNT sample achieved the highest compressive strength (24 MPa) and modulus (1 GPa), the overall performance of Route 2 samples was significantly lower than Route 1, with all samples falling below 20 MPa. For example, the 0.00 wt.% MWCNT sample in Route 2, despite its stability, achieved a compressive strength ≈18 MPa with a compressive modulus of 0.98 GPa, falling short of the recorded 27 MPa compressive strength and 1.40 GPa compressive modulus of 0.00 wt.% MWCNT sample in Route 1. In contrast, Route 3 which combined vacuum pre‐curing with thermal cycling, demonstrated better mechanical performance, with the 0.00 wt.% MWCNT sample achieving a compressive strength of 47 MPa (Figure 4c) and modulus of 1.67 GPa (Figure 4d). This improvement is unique, as vacuum pressure typically has a detrimental effect on compressive strength.^{[ 42 ]} However, higher MWCNT content samples in Route 3 performed poorly, with the 1.00 wt.% MWCNT sample showing a compressive strength of only 28 MPa and modulus of 1.36 GPa. This trend suggests that the thermal cycling process interacts with fabrication conditions and MWCNT wt.% to significantly alter mechanical behavior compared to Route 1.

Figure 4.

a–d) Compressive strength and compressive modulus of ambient‐cured (Route 2) and vacuum‐cured (Route 3) specimens subjected to temperature cycling, error bars representing the standard deviation of the average values e–f) remaining water content after pre‐curing and post‐cycling for samples fabricated under Routes 2 and 3, respectively.

The inconsistencies in compressive strength and modulus between Routes 1, 2, and 3 can be explained by the role of water retention during fabrication. In Route 2, samples experienced significant water loss during the ambient exposure phase, leaving lower MWCNT wt.% samples with ≈50% water retention, while higher MWCNT wt.% samples retained less than 20% (Figure 4e). Excessive dryness in these levels of MWCNT wt.% samples contributed to brittleness and poor mechanical performance. Route 3, on the other hand, i.e., Vacuum pre‐curing retains higher water content by creating a localized humid environment that slows moisture loss. The absence of airflow reduces convective drying, while redistributed moisture remains trapped within the material's microstructure. As a result, significant evaporation occurs predominantly during thermal cycling when heat drives moisture out more effectively (Figure 4f). This controlled evaporation process promoted densification and reduced porosity in low MWCNT content samples, resulting in optimal water retention of ≈35%, which correlated with superior compressive strength compared to Routes 1 and 2, respectively. However, higher MWCNT content in Route 3 samples retained ≈50% water after thermal cycling, leading to increased porosity and weaker interfacial bonding, negatively impacting their strength. Under vacuum, the CNT network likely traps water by limiting its escape pathways, creating localized moisture pockets. Additionally, the prevention of rapid surface evaporation in a vacuum, allows the CNT network to retain more water than ambient conditions.

While water retention plays a critical role in determining microstructural integrity, the significant reduction in compressive strength during thermal cycling, particularly in high MWCNT content samples, is largely attributed to the mismatch in thermal expansion coefficients (CTE) between the regolith matrix and MWCNTs. MWCNT exhibited a negative CTE of −1.2 × 10⁻⁵/ °C⁻¹,^{[ 43 ]} MWCNT shrinks when heated, whereas the regolith components, such as SiO₂ (0.24 × 10⁻⁶/ °C⁻¹),^{[ 44 ]} Al₂O₃ (8.1 × 10⁻⁶/ °C⁻¹),^{[ 45 ]} and CaO (13.57 × 10⁻⁶/ °C⁻¹)^{[ 46 ]} expand under the same conditions (Figure S4, Supporting Information). This opposing thermal behavior generates significant internal stresses at the CNT‐matrix interfaces during temperature fluctuations. These stresses disrupt the structural cohesion of the composite, leading to localized microcracking and delamination, which are more pronounced in samples with higher MWCNT concentrations. Consequently, while vacuum pre‐curing improves strength by promoting densification, the addition of MWCNTs aggravates the thermal mismatch issue, reducing mechanical performance during cycling.

To further investigate and quantify the effects of CTE mismatch, thermal cycling simulations were conducted to model the stress and strain distributions in MWCNT‐reinforced composites under lunar‐like temperature fluctuations. The simulations utilized a representative volume element (RVE) approach, with the RVE extracted from a larger computational model populated with homogeneously distributed, randomly oriented MWCNTs (Figure 5a). The MWCNTs were modeled as solid cylinders with diameters of 20 nm and lengths between 1 and 15 µm, ensuring no intersections or overlaps to replicate a realistic microstructure. The chosen RVE size of 5 × 5 × 5 µm³ included a MWCNT volume fraction of 0.77%, closely matching the experimental value of 0.71% (equivalent to 0.5 wt.%). The finite element model, implemented in general finite element analysis software, ABAQUS, with a coupled temperature‐displacement solver, applied a heating‐cooling‐heating cycle ranging from 80 to 117 °C, cooling to −80 °C, and returning to 80 °C. This setup simulated the thermo‐mechanical interactions arising from differences in the CTEs of the MWCNTs and the matrix material.

Figure 5.

a) schematic of the RVE modeling approach incorporating MWCNT distribution. Stress (unit in MPa) and strain contours in the x and y directions within a RVE of the lunar regolith composite reinforced with MWCNTs, evaluated through simulation at b) lunar maximum temperature of 117 °C, c) experimental minimum temperature of −80 °C, d) lunar minimum temperature of −171 °C, and e) after one complete thermal cycle ending at 80 °C.

Incorporating temperature‐dependent properties for both the MWCNTs and the lunar regolith matrix revealed more pronounced effects of thermal mismatch across the full extended cycle. At the highest temperature of 117 °C, the MWCNTs slightly contracted while the matrix expanded, creating localized stress concentrations at their interfaces and increasing principal strain within the adjacent matrix (Figure 5b). As the temperature dropped to −80 °C, the thermal expansion mismatch became more pronounced, with interfacial stresses rising sharply from 33.42 MPa at 117 °C (Figure 5b) to 127.60 MPa (Figure 5c), and the principal strain reaching 5.87 × 10⁻⁴ (Figure 5c). Upon further cooling to −171 °C, the mismatch effects intensified, which the matrix experienced contraction while the MWCNTs continued expanding, causing interfacial stress to exceed 207.20 MPa and principal strain to surpass 1.58 × 10⁻³ (Figure 5d). When reheated to + 80 °C (Figure 5e), the system exhibited partial stress relief; however, residual stress and strain concentrations remained along the interfaces. These persistent stress–strain fields indicate accumulated thermal mismatch and point to the potential for long‐term damage mechanisms such as microcracking and interfacial debonding between the MWCNTs and the regolith matrix.

2.3. Comparative Evaluation of MWCNT‐Incorporated Composites

Achieving consistent compressive strength in composites relies heavily on uniform particle dispersion. To evaluate the quality of dispersion, Scanning Electron Microscopy (SEM) and Energy Dispersive Spectroscopy (EDS) are employed to visualize and analyze the distribution of materials within the composite with 1.00 wt.% MWCNT across the three processing routes (Figure 6 ). The SEM images for all samples across the three synthesis routes reveal minimal porosity, a crucial characteristic for enhancing material strength. The microstructure includes a mix of small and large particles, reflecting a deliberate particle‐size blending strategy designed to optimize packing density and mechanical properties. However, the significant differences in the distribution of MWCNTs are visible among different routes, providing insight into their reinforcement efficiency under different fabrication conditions. In the Route 1 samples (ambient curing without temperature cycling), the MWCNTs appear to wrap around the regolith particles, creating a smoother overall morphology (Figure 6a). This wrapping behavior enhances interfacial bonding and contributes to improved load transfer within the composite, as evidenced by the smoother textures in the corresponding SEM images (detailed further in Figure S1, Supporting Information). The smoother morphology aligns with the minimal thermal stresses present in Route 1, as these samples were not subjected to temperature cycling. The uniformity of the matrix and the close interaction between MWCNTs and particles provide evidence of a cohesive composite structure, reinforcing the mechanical stability, as illustrated in Figure 3a. Conversely, the temperature‐cycled samples (Routes 2 and 3) display a distinct separation between the MWCNTs and the regolith particles, resulting in a rougher morphology (Figures 6b and  7c) (detailed further in Figures S2 and S3, Supporting Information). The separation in the regolith‐MWCNT composite after temperature cycling arises from the mismatch in their coefficients of thermal expansion (CTE). This discrepancy causes stress at the interface, leading to delamination and affecting the composite's structural integrity. This separation was more pronounced in the vacuum‐cured Route 3 samples, suggesting that thermal cycling introduces the CTE mismatch between the MWCNTs and the regolith matrix. To verify the quality of MWCNT dispersion and ensure that the observed carbon distribution was not solely due to the presence of urea, EDS analysis was used to distinguish MWCNT dispersion from other carbon‐containing components. EDS mapping was conducted on a control sample without urea (Figure 6d) and a standard sample containing urea (Figure 6e). Since urea introduces additional carbon into the matrix, this comparative analysis was essential to isolate the contribution of MWCNTs. The results demonstrate a homogeneous distribution of MWCNTs in both cases, confirming that the addition of urea did not adversely affect dispersion. This uniform dispersion, achieved by directly dispersing MWCNTs into the acidic solution, effectively prevented MWCNT conglomeration, which is critical for avoiding stress concentrations and maintaining composite integrity. However, the SEM images indicate that the primary issue lies not in MWCNT dispersion but in the weakened interfacial adhesion caused by temperature cycling. While the uniform distribution of MWCNTs should theoretically enhance load transfer and mechanical performance, the separation between MWCNTs and regolith particles observed in cycled samples undermines this potential. The rougher texture corresponds to reduced interfacial bonding, with MWCNTs failing to form an interconnected network with the surrounding particles. This weak adhesion likely limits the efficiency of load transfer and contributes to stress concentrations at the interfaces between the MWCNT and regolith particles, particularly under temperature cycling conditions. In addition to the adhesion challenges, the separation observed in temperature‐cycled samples highlights the role of CTE mismatch. The contrasting CTEs of the MWCNTs (negative) and the regolith matrix (positive) introduce internal stresses during heating and cooling cycles, further degrading the composite structure. These stresses amplify microcracking risks, particularly in high MWCNT content samples where the MWCNT concentration intensifies these effects. Therefore, while the fabrication process ensures uniform MWCNT dispersion, thermal cycling disrupts interfacial adhesion and matrix cohesion, ultimately reducing the compressive strength of the composite.

Figure 6.

a–c) SEM images of 1.00 wt.% MWCNT samples with callouts d–f) highlighting visible MWCNTs fabricated under ambient‐cured, non‐cycled condition (Route 1), ambient‐cured with temperature cycling condition (Route 2), and vacuum‐cured with temperature cycling condition (Route 3), respectively, g) EDS image showing MWCNT distribution in the control sample without urea to show the quality of dispersion, h) EDS image showing carbon distribution in the Route 1 sample with urea.

Figure 7.

a) Schematic illustrating failure patterns observed during compressive test b–d) post‐compression failure patterns for specimens with varying MWCNT wt.% fabricated under ambient non‐cycled, ambient‐cycled, and vacuum‐cycled routes, respectively e) comparisons of compressive toughness of the specimens fabricated under three distinct routes. Visualization of compressed sample fabricated under f) Ashby plot comparing compressive strength versus density of this study against other literature.^{[ 49} , ⁵⁰ , ⁵¹ , ^{52 ]}

The uniformity of both particle dispersion and degree of dryness is crucial for maintaining a well‐supported composite structure, as evaluated in the SEM and EDS analyses. Both forms of analyses supported that MWCNTs were uniformly dispersed throughout all routes with a good mixing strategy, avoiding conglomerations that weaken the composite. However, the SEM images further demonstrated that temperature cycling (Routes 2 and 3) introduced interfacial adhesion challenges due to the separation of MWCNTs and regolith particles, which intensified the internal stress gradients generated during the curing process. These findings correlate directly with the failure patterns observed in the compressive tests, as illustrated in Figure 7.

Following the compressive tests, failure in the samples consistently began with exterior damage, leading to slight hourglass‐shaped deformation on the interior parts (Figure 7a). These failure patterns were observed across all samples fabricated via different processing methods including ambient curing with no cycling (Figure 7b), ambient curing with cycling (Figure 7c), and vacuum curing with cycling (Figure 7d). This suggests that the exterior parts of the sample cure more efficiently than the interior parts. Cylindrical samples are prone to fracture at any point along the sides or edges, with fractures occurring both in the middle and at the outer parts of the specimens.^{[ 47 ]} This failure, referred to as “hourglass failure,” happens when the load is applied uniformly at the center, resulting in symmetrical failure across all sides due to the absence of loading eccentricity.^{[ 48 ]} Although the pattern of breakage was consistent across all routes, slight differences were observed in the breakage process. Routes 1 and 3 exhibited more visible cracks and breakage, while Route 2 retained its shape without significant break‐off damage. This suggests that the high compressive strength observed in Routes 1 and 3 is primarily due to their hard exterior, even though the interior was not fully dried. In contrast, Route 2 demonstrated lower performance with compressive strengths falling below 20 MPa. This indicates that the outer layer, which is the first to break off, was not as hard in Route 2 compared to the other routes, as evidenced by the lack of pronounced breakage.

The uneven drying process during curing plays a critical role in creating internal stress gradients, which weakens the composite structure. As the material cures, moisture migrates from the interior to the exterior, causing the outer layers to dry faster. This differential drying leads to greater shrinkage on the exterior, generating tensile stresses that can initiate failure on the outer layers. Although the interior is weaker due to these internal stress gradients, the exterior typically experiences higher surface tension during drying, further contributing to localized damage and structural weakening. These stress gradients are less pronounced in Route 1 due to the absence of temperature cycling, resulting in smoother and more cohesive fracture surfaces. In contrast, temperature cycling in Routes 2 and 3 intensified the internal stresses, leading to brittle failure and rougher fracture surfaces due to weakened interfacial adhesion and separation between MWCNTs and regolith particles.

Building on the microstructure characteristics of samples fabricated through different routes, the toughness data provides a comparative measure of the material energy absorption under compressive load (Figure 7e). In Route 1, the toughness increased significantly from 0.64 MPa for 0.00 wt.% specimen to 1.17 MPa for 0.5 wt.% with increasing the MWCNT wt.%, however, a slight decrease to 1.06 MPa was observed at 1.00 wt.% MWCNT content. This increasing trend is attributed to the enhanced strength and ductility imparted by MWCNT reinforcement. This improves the material's capacity to resist deformation and failure under compressive forces. However, the decline at 1.00 wt.% MWCNT is due to poor interfacial bonding, which compromises structural integrity. In contrast, Route 2 exhibits no clear or consistent trend in toughness with results ranging from 0.40 to 0.60 MPa. The lack of a clear pattern suggests that temperature cycling disrupted the interaction between MWCNTs and the composite matrix, potentially due to thermal mismatch stresses or microstructural changes caused by repeated expansion and contraction. These inconsistencies highlight the challenges of maintaining uniform performance when the composite is subjected to thermal fluctuations. For Route 3, the toughness gradually decreases from 1.20 to 0.76 MPa as MWCNT wt.% increases from 0.00 to 1.00 wt.% respectively. While the vacuum curing process positively influences the overall curing quality and minimizes porosity, the higher MWCNT concentrations are adversely affected by the temperature cycling. These results from the amplified CTE mismatch between the MWCNT and the regolith, lead to microcracks and reduced energy absorption capacity. The decreasing trend in Route 3 closely aligns with the observed compressive strength data and the separation of MWCNT and regolith particles observed in SEM images. This also emphasizes the interconnected roles of the MWCNT wt.%, curing conditions, and mechanical properties.

The performance trends observed across the three fabrication routes highlight the critical influence of processing conditions on mechanical properties. To contextualize these results, the Ashby plot of compressive strength versus density highlights the performance of various material systems in comparison with this study (Figure 7f).^{[ 49} , ⁵⁰ , ⁵¹ , ^{52 ]} This study occupies a unique position at the intersection of plastics, metals, composites, concrete, and ceramics. This unique placement underscores their versatility and suitability for a wide range of structural applications, offering an optimal balance of strength and lightweight properties tailored to diverse engineering demands. Our composite exhibits distinct advantages compared to existing lunar regolith‐based materials, including high compressive strength under low‐temperature curing, enhanced performance under vacuum conditions, and a straightforward, energy‐efficient fabrication process tailored for lunar applications. We demonstrated a practical synthesis route that achieved competitive mechanical performance without the need for high‐temperature processing. While higher compaction pressures could potentially further improve strength, our use of relatively low pressure (8 MPa) was intentional, as it reflects the practical limitations of equipment likely to be available for in situ construction on the Moon. This balance between performance and feasibility makes our approach particularly relevant for extraterrestrial infrastructure development.

3. Conclusion

This study presents a detailed exploration of the complex interplay between MWCNT reinforcement and lunar regolith composites under varied curing conditions, highlighting both the potential and limitations of using MWCNTs in extraterrestrial construction. The results of the different processing routes, including ambient cured with no cycling (Route 1), ambient cured with cycling (Route 2), and vacuum cured with cycling (Route 3), demonstrate that incorporating MWCNTs significantly enhances composite strength under specific conditions. In ambient curing without cycling, samples achieved a high compressive strength of 39 MPa at 0.50 wt.% MWCNT, representing a 44.44% improvement over the 0.00 wt.% MWCNT sample. However, a thermal mismatch between the negative CTE of MWCNTs and the positive CTE of the regolith matrix limited durability during lunar temperature cycling, causing microcracking and reduced performance. Under vacuum curing and temperature cycling, the 0.00 wt.% MWCNT composite achieved the highest strength of 47 MPa, attributed to enhanced densification from accelerated moisture evaporation and an optimal water retention level of 35%, while the 1.00 wt.% MWCNT sample performed the worst at 28 MPa. Addressing thermal mismatch challenges requires optimizing CTE compatibility or developing tailored CNT variants to enhance compressive strength and durability under lunar conditions. This study highlights the importance of testing composites under realistic conditions and reveals that the MWCNTs used in this study are incompatible with the regolith matrix during temperature cycling. Improving CTE compatibility is key to mitigating thermal stresses and enhancing material durability for extraterrestrial construction.

4. Experimental Section

Materials

The regolith used in this study was LSP‐1 Lunar South Pole Simulant, sourced from Exolith Lab with a particle size range of <0.04–1000 µm. For the particle size strategy, larger particles were screened out using a sifter, while smaller particles were generated through grinding and further sieving (20 × 60 mesh). To reduce water requirements, urea granules with a 99.0–100.5% assay (anhydrous basis) were from Lab Alley and utilized as a superplasticizer. An acidic solution composed of water and 85% ACS‐grade phosphoric acid was prepared, with the acid sourced from Oakwood Chemical. Multi‐walled carbon nanotubes with a diameter greater than 50 nm and a length of 10–20 µm, purchased from Cheap Tubes Inc., were incorporated to investigate their effects on the mechanical compressive strength of the composite.

Fabrication

The composite system itself comprises five primary components: lunar regolith, urea, distilled water, phosphoric acid, and MWCNT powder (>50 nm), with ≈77.36 wt.% derived from a simulated South Pole regolith. An acidic solution of phosphoric acid mixed with distilled water constitutes ≈19.50 wt.% of the mixture. The MWCNTs were initially dispersed in the alkaline solution for ≈30 min to promote a more uniform distribution, as the liquid medium helps prevent agglomeration and allows the nanotubes to spread more evenly before being introduced to the dry regolith. Other strategies to enhance additive distribution in future work could include milling,^{[ 53 ]} water‐assisted chemical vapor deposition,^{[ 54 ]} and sonication.^{[ 55 ]} After the preparation of all the materials, they were thoroughly mixed manually for 3 min to form a homogeneous slurry. To ensure consistent testing, three replicates of each MWCNT loading were prepared and subjected to uniaxial compression at 8 MPa for 10 min using a 20 mm diameter pellet die (Specac, USA). After compression, the specimens proceeded through one of the three curing pathways. Route 1 involved ambient air curing for four days, followed by 8 h at 70 °C in an oven. Route 2 followed the same ambient and oven curing but introduced four temperature cycles to simulate lunar day‐night extremes, while Route 3 replaced ambient curing with four days in a vacuum chamber and then vacuum oven curing at 70 °C for 8 h, followed by the same thermal cycling. Temperature preconditioning was based on the lunar surface extremes reported by Malla et al.^{[ 8 ]} which was ≈114 °C during daytime to −171 °C at night. Due to equipment constraints, the minimum temperature in this study was adjusted to −80 °C, a maximum temperature of 117 °C was used due to variations in temperature reporting, and the typical 14‐day lunar cycle was compressed into 24 h for practical implementation. Although this work used a maximum achievable vacuum of −0.1 MPa, future studies may aim for conditions nearing 3.04 × 10⁻¹⁵ atm^{[ 56 ]} to replicate the near‐total vacuum on the Moon more closely. Once curing and preconditioning were complete, all specimens underwent chemical analyses and mechanical testing to evaluate their performance under simulated lunar conditions.

Characterization

Chemical analysis was performed to confirm the composition quality of the samples and ensure uniform dispersion of the components in the composite samples. Residual water content was quantified to evaluate the impact of different curing routes on residual water content. For each MWCNT weight percentage under Routes 2 and 3, three samples were weighed before and after thermal curing, and after cycling. Additionally, the density was calculated at these three stages, both with and without assuming complete water evaporation. From these values, the accurate assumption of water remaining was calculated to determine the percentage of residual water content.

The bonding structures of the composites with no cycling were further analyzed using FTIR. FTIR spectroscopy was conducted with a Thermo Nicolet 6700 FT‐IR spectrometer (Thermo Fisher Scientific) in transmittance mode over a wavenumber range of 4000–400 cm⁻¹, focusing on the bottom surface area of each sectioned sample. The crystallographic structures of the composites were analyzed using X‐ray Diffraction (XRD). XRD measurements were performed using a Bruker D2 Phaser with LYKXEYE 1D silicon strip detector, employing Cu K∝ radiation (λ = 1.5406 Å). Data were collected over a 2θ range of 10–90° at a scanning rate of 3°/min, with a step size of 0.05°. Powdered samples were prepared to ensure uniformity and minimize preferred orientation effects, allowing for accurate analysis of the composite's crystalline and amorphous phases. A batch of composites in all routes were sectioned vertically for SEM and EDS analysis. SEM was utilized to examine porosity and surface texture across the 1.00 wt.% CNT concentration for the three curing routes and EDS was performed on the 1.00 wt.% CNT sample to map the distribution of MWCNTs.

Mechanical Analysis

The compressive strength of the lunar regolith composites was evaluated using a Materials Test System (MTS) machine with 80 kN load cell. Compressive strength tests were conducted on samples containing 0, 0.25, 0.50, and 1.00 wt.% CNTs, with a controlled displacement rate of 1 mm min⁻¹. From the compressive strength, the compressive modulus was calculated. Additionally, the toughness factor was determined by calculating the area under the stress–strain curve through numerical integration. The fracture behavior of each sample was analyzed to provide further insight into the mechanical properties and failure mechanisms of the composites.

Computational Analysis

An in‐depth analysis of the CTE was conducted to elucidate the underlying mechanisms driving the results. Notably, this represents the first such investigation for lunar regolith. A simulation basis was developed to investigate thermal mismatch and the resultant thermal stresses in MWCNT‐reinforced nanocomposite matrices based on lunar regolith. The computational approach was integrated with the experimental methodology to validate the behavior of MWCNTs within the composite under lunar temperature fluctuations. A model of the concrete matrix was then created, beginning with generating a larger domain populated with homogeneously distributed, and randomly oriented MWCNTs using a Python script. The MWCNTs were modeled as solid cylinders with a diameter of 20 nm and lengths ranging from 1 to 15 µm, distributed without intersections or overlaps to replicate a realistic microstructure. Subsequently, a smaller sub‐domain rich in MWCNT density was extracted from the larger domain to serve as the 3D RVE for thermal stress analysis. The chosen RVE size was 5 × 5 × 5 µm³, with a fiber volume fraction of 0.77%, closely matching experimental samples (0.71%, equivalent to 0.5 wt.%).

The finite element model for the selected RVE was implemented in Abaqus using a coupled temperature‐displacement solver to analyze thermo‐mechanical interactions. Both the MWCNTs and the lunar regolith matrix were modeled as linear elastic materials with isotropic coefficients of CTE to capture their thermal behavior (Table S1, Supporting Information). A temperature field was applied to all outer surfaces to simulate thermal effects, following a heating‐cooling‐heating cycle replicating one cycle used for this study: the temperature rises from 80 to 117 °C, cooled to −80 °C, and then returned to 80 °C. This cycle was designed to elucidate thermal mismatch effects arising from differences in the CTEs of the MWCNTs and the matrix material.

Boundary conditions of the temperature cycle and displacement were defined to ensure accurate simulation results. Since the sample rests on a surface, penetration was not possible. To account for this, one surface of the RVE was fixed to prevent rigid body motion, while the other surfaces were allowed to deform freely. The interface between the MWCNTs and the lunar regolith was modeled using a “Tie” contact with the nodes‐to‐surface option, ensuring an accurate representation of contact mechanics. A fine mesh with a size of 03 µm was used, employing linear 4‐node thermally coupled tetrahedral elements. This meshing strategy achieves a balance between computational efficiency and the resolution required to capture significant thermal and mechanical stress gradients. Thermal stress evolution during the heating and cooling cycles was then analyzed to understand the impact of CTE mismatches between the MWCNTs and the matrix material. The residual stresses following the cycle provided insights into microstructural stress distributions and their potential effects on the mechanical integrity of the composite under cyclic thermal loading.

Statistical Analysis

All statistical analyses were performed using Microsoft Excel. Figures 3, 4 present the maximum recorded values for compressive strength, while Figures 3, 4 display the average values for both compressive strength and compressive modulus, with error bars representing the standard deviation.

For the density calculations and water retention analysis, the initial weight of each sample was recorded immediately after removal from the pellet die. A theoretical dry weight was then calculated by assuming complete evaporation of all water content, serving as the reference dry value. Samples were subsequently weighed at two key stages: 1) after pre‐curing (prior to oven placement) and 2) after the thermal cycling process. These recorded weights were compared to the reference dry weight to estimate the remaining water content. The percentage of water retention was calculated based on this comparison and plotted for comparative analysis across different conditions.

Conflict of Interest

The authors declare no conflict of interest.

Supporting information

Supporting Information

Hoe A. J., Tarafdar A., Lin W., Fiske M. R., Edmunson J. E., Wang Y., Balancing Strength and Extreme Thermal Resilience in Lunar Regolith Composites: The Role of Multi‐Walled Carbon Nanotubes. Small 2025, 21, 2502220. 10.1002/smll.202502220

Data Availability Statement

The data that support the findings of this study are available from the corresponding author upon reasonable request.