But a new study suggests that the materials do not have to come from Earth at all. In research publishing on September 10 in the journal Chem Circularity, a team of scientists has developed a recipe that could one day be used to 3D print houses on Mars. The formula combines one ingredient that Mars has in abundance—Martian rock—with a pair of surprisingly humble Earthly imports: gelatin and yeast. Once the printed material dries and hardens under Martian conditions, it becomes a living building material with a compressive strength comparable to low-grade concrete. And because the material is alive, it can be broken down, recycled, and brewed again for new construction, offering something that no conventional building product on Earth or Mars can match: a genuinely circular building economy on another world.

The inspiration for the project came, perhaps unexpectedly, from the kitchen. My inspiration came from freeze-dried fruits that become harder, says civil engineer and senior author Jishen Qiu of The Hong Kong University of Science and Technology. Mars’s extremely low temperature and pressure create conditions that are strikingly similar to the freeze-drying process used to preserve food, in which moisture is drawn out of a material while it is frozen. So I asked myself if we can take advantage of that and make some materials, Qiu explains. Rather than fighting the planet’s punishing environment, the team decided to exploit it, turning the very conditions that make Mars hostile into a natural manufacturing step that hardens and stabilizes their printed structures.

To conjure the printable material, Qiu and her colleagues needed the right mixture of sand and glue. In this case, the sand is crushed Martian rock, which provides mass and structural bulk, while the glue is a carefully engineered blend of gelatin and a specialized strain of yeast. The team engineered yeast cells coated with highly adhesive proteins, including the same kinds of molecules that mussels use to cling tenaciously to rocks in the surf zone. The gelatin knits the ingredients together into a coherent paste and provides a hospitable scaffold in which the yeast cells can live and grow, while the adhesive proteins bind the mineral particles to one another and to the organic matrix. The result is a slurry that behaves like an ink, thick enough to hold its shape as it emerges from a printing nozzle yet fluid enough to be extruded layer by layer into a designed form.

The critical test came when the researchers exposed their printed material to simulated Martian conditions. Once the mixture pushes through the nozzle, the extreme cold and low pressure of the chamber freeze-dry it almost immediately. Under these conditions, water does not simply melt and evaporate; instead, it freezes and then sublimates, turning directly from ice into vapor without ever passing through a liquid phase. As the ice crystals vanish, they leave behind a network of microscopic pores throughout the material. The final product resembles a foam—light, porous, and rigid—whose internal architecture is sculpted by the sublimation process itself. In effect, the Martian atmosphere performs the final curing step for free, requiring no kilns, no furnaces, and no external energy input beyond the cold and vacuum that the planet already provides.

For now, the printed structures are modest in scale: small domes roughly the size of a wine cork, standing 45 millimeters tall and 30 millimeters wide. But the material itself is far more impressive than its dimensions suggest. The team measured a compressive strength of 10 to 12 megapascals, a figure comparable to low-grade concrete used in everyday construction on Earth. So this is actually strong enough to build a one- or two-story building on Earth whose gravity is three times that of Mars, Qiu says. So, you can probably easily build a multistory building on Mars with the material. Because Martian gravity is only about a third as strong as Earth’s, structures printed from this material would experience far less compressive load than their terrestrial counterparts, meaning the strength demonstrated in the laboratory translates into even greater structural headroom on the Red Planet itself.

The approach stands in sharp contrast to many other proposals for building extraterrestrial cities, most of which involve heating and melting Martian rocks or lunar dust into bricks and beams. Sintering or smelting regolith requires substantial energy, typically delivered by kilns, lasers, or concentrated sunlight, all of which demand heavy equipment and a robust power supply that settlers would need to establish before construction could even begin. The biology-based material takes an entirely different manufacturing approach, saving the enormous energy cost of heating by letting freeze-drying do the hardening instead. Just as importantly, it could support a circular economy on Mars. Settlers could recover the yeast from dismantled or damaged structures and regrow it in bioreactors, turning old buildings into feedstock for new ones. As long as there’s one yeast that’s still alive, you can grow them again, Qiu says—a resilience that ordinary concrete, mortar, and steel can never offer.

Significant challenges remain before anyone could print a habitat on the Martian surface. So far, the living building material has only been tested on Earth, under chambers that simulate Martian temperature and pressure but cannot capture every nuance of the planet’s environment, including its radiation environment and the chemistry of its dust. The team does not yet know whether the engineered yeasts can survive actual Martian conditions over the long term, where cosmic radiation and desiccation would test any living system. The material still relies, albeit to a lesser degree than conventional approaches, on Earthly ingredients: the gelatin, the engineered yeast, and the adhesive proteins would all need to be shipped from home, at least initially. Its ultimate success will therefore depend in part on advances in rocket technology and launch economics. Qiu estimates that meaningful on-site engineering could require hundreds of tons of cargo from Earth, a figure that underscores how much progress is still needed before biology can truly carry the weight of extraterrestrial construction.

Even so, the researchers are optimistic that the fundamental physics and biology are on their side. I always ask myself: Is there any physical law or fundamental mechanism that prevents us from doing this? Qiu says. I can’t see any at this point in time. We are confident in scaling it up. The team’s confidence rests on the fact that every step of the process—mixing, extrusion, freeze-drying, and biological recovery—has already been demonstrated, at least at laboratory scale, under conditions that mimic Mars. What remains is engineering: scaling up the printers, hardening the yeasts against radiation, and closing the loop on ingredient recycling so that each mission carries less and reuses more. It would surprise me if materials for future Martian engineering will not be as diverse as those used in Earth engineering—and biology will certainly contribute, Qiu says. If she is right, the first houses on Mars may not be built at all in the traditional sense. They may be grown, printed, and brewed into existence by an army of microscopic builders, one freeze-dried layer at a time, turning the planet’s own hostile climate into the most reliable construction worker on site.

Subject of Research: Engineered living building material made from Martian rock, gelatin, and yeast for 3D-printed low-energy construction on Mars

Article Title: Scientists want to 3D print houses on Mars with the help of yeast

Article References: Scientists want to 3D print houses on Mars with the help of yeast. (n.d.). Original publication

Image Credits: AI Generated

DOI: Not provided

Keywords: Mars, 3D printing, living building material, engineered yeast, gelatin, freeze-drying, compressive strength, circular economy, space construction, regolith, bioreactors, Cell Press

The choice of test conditions in this study reflects the extreme environment of the lunar polar regions. Permanently shadowed regions at the poles never receive direct sunlight, and surface temperatures there remain low enough that water-ice can survive for billions of years even under ultrahigh vacuum, sublimating only very slowly. To replicate these conditions, samples were held at approximately 93 K while the probe itself was varied across a range from 83 K to 163 K. Maintaining such temperatures during mechanical testing is nontrivial, and the experimental design addressed this with careful thermal management, including thermally isolating components so that sensitive instrumentation such as the load cell remained near room temperature even while the probe tip was immersed in cryogenic conditions. This attention to thermal boundary conditions matters because the mechanical behavior of ice is strongly temperature dependent, and uncontrolled warming during a test could substantially distort the measured strength.

A central contribution of the work is its systematic comparison of four distinct icy regolith morphologies prepared from the same lunar highlands simulant. Ice-cemented samples are made by freezing liquid water into a granular mineral matrix, producing a material in which ice acts as a cement between particles, analogous to the way cement binds concrete. Pressure sintered samples instead combine granular ice and dry regolith compacted together, yielding a more granular, less cemented texture. Unsintered samples represent loosely mixed material, while vapor deposited samples simulate ice that has condensed directly from vapor onto cold regolith surfaces, a process thought to be relevant in permanently shadowed regions where ice may be deposited from transient atmospheres or impact-generated vapor. Because no in-situ observations have yet been made within a lunar PSR, the actual form of lunar ice remains unknown, and each of these morphologies represents a plausible candidate for what future missions may encounter.

The finding that microstructure dominates geotechnical behavior has significant implications for how the community interprets prior simulant studies. Previous research had already hinted at enormous variability: ice-cemented simulants showed sharp strength increases at critical ice contents of roughly 1 and 3 weight percent, and ice-enriched material above about 12 weight percent was reported to behave like hard rock or concrete with strengths exceeding 100 megapascals. In contrast, pressure sintered mixtures at 10 weight percent ice were far weaker, below 14 megapascals, and showed no abrupt strength transitions. The new controlled measurements confirm that even under identical ice contents, densities, and temperatures, different preparation methods can produce materials whose strengths differ by orders of magnitude. This means that the preparation recipe, not merely the ice content, controls mechanical response.

The quantitative trafficability results are particularly striking. When cone index gradients were corrected for lunar gravity and normalized to a common penetration depth, a vapor deposited sample containing 6.88 weight percent ice proved 3.92 times stronger than a pressure sintered sample with 8 weight percent ice, and even 1.23 times stronger than the upper bound of in-situ cone index values measured in the Descartes Highlands during Apollo 16. This comparison anchors the laboratory measurements to the only direct penetrometer data ever collected on the lunar surface, providing a useful reality check. It suggests that some icy morphologies could produce surfaces that are at least as firm as the dry regolith traversed by Apollo astronauts, which is encouraging for rover mobility planning in polar terrain.

The excavation energy results reveal a more nuanced picture, with opposing trends depending on morphology. For pressure sintered samples, the energy required per unit mass of excavated water fell from 0.35 joules per gram at 1 weight percent ice to 0.11 joules per gram at 30 weight percent, a 3.2-fold improvement in efficiency. This counterintuitive result arises because richer ice deposits contain proportionally more water per unit of material excavated, so the energy cost attributable to each gram of recovered water declines even if the material itself is somewhat harder to dig. For ice-cemented samples the trend runs in the opposite direction, with excavation energy climbing from 2.74 joules per gram at 1 weight percent to 143.03 joules per gram at 12 weight percent, a 52-fold increase. The cementing action of ice between mineral grains makes the material progressively harder to break apart, overwhelming the benefit of higher water content.

These opposing trends carry direct consequences for the economics of lunar water extraction. In-situ resource utilization concepts that envision mining polar ice to produce propellant, drinking water, and oxygen depend on favorable energy budgets for excavation and processing. If lunar ice occurs predominantly in sintered, granular, or vapor deposited forms, the study’s results suggest that excavation may be considerably more economical than many hardware development programs have assumed, since tests based on ice-cemented simulants would have systematically overestimated the resistance that mining equipment must overcome. Conversely, if substantial fractions of polar ice are genuinely cement-like, excavation energy demands could be far higher than optimistic projections. Because the true morphology is unknown, mission planners face a wide envelope of possible outcomes, and this work helps bound that envelope with quantitative data.

The implications extend to the validation of flight hardware. Robotic excavators, drills, and mobility systems destined for the lunar poles are typically qualified against simulant beds in terrestrial laboratories. If those beds are made with ice-cemented recipes, the hardware will be tested against a material that may be far stronger than what it actually encounters, leading to over-engineered systems with unnecessary mass and cost. On the other hand, hardware validated only against weak sintered simulants could fail catastrophically if it encounters cemented ice. The authors’ observation that ice-cemented simulants, rocks, and concrete analogues may be unsuitable as general-purpose proxies for all icy regolith morphologies therefore challenges a widespread practice in the field and argues for morphology-specific testing campaigns matched to the best available scientific estimates of how lunar ice actually forms.

The study also addresses strategic knowledge gaps formally identified by the International Space Exploration Coordination Group, including the geotechnical characteristics and accessibility of permanently shadowed regions, the form and distribution of frozen polar volatiles, and trafficability across the lunar surface. Missions such as the Volatiles Investigating Polar Exploration Rover and the crewed Artemis program will operate in precisely the environments where these uncertainties matter most, traversing and sampling terrain whose mechanical properties cannot yet be observed directly. Quantitative laboratory measurements of the kind reported here provide a bridge between remote sensing indications of ice and the engineering parameters, such as bearing capacity and excavation resistance, that determine whether wheels will sink, slip, or find firm purchase.

Methodologically, the work demonstrates the value of miniature cone penetration testing as a compact, information-rich technique for cryogenic simulant characterization. The apparatus described, combining a small stainless steel cone with a low-capacity load cell mounted on a universal test machine, achieves sufficient precision to resolve differences between morphologies while remaining simple enough for routine laboratory use. The inclusion of a thermocouple within the hollow cone tip allows probe temperature to be monitored and controlled independently of sample temperature, an important capability given that probe-sample friction and ice deformation behavior both depend on temperature. Calibration against a certified reference load cell further strengthens confidence in the absolute values reported.

Looking forward, the results suggest several productive avenues for the field. First, confirming the morphology of lunar ice through in-situ measurements, whether by penetrometers on future rovers or by sample return from polar regions, would dramatically narrow the uncertainty in excavation and mobility predictions. Second, extending this experimental framework to other volatiles, such as carbon dioxide ice, and to mixtures of multiple ices, would broaden its applicability to other cold solar system bodies including Mars, comets, and the moons of the outer planets. Third, relating cone index measurements to full-scale excavation forces and wheel-soil interaction models would help translate these laboratory parameters into mission-level engineering requirements. In the meantime, the demonstration that icy lunar regolith may be easier to excavate and traverse than previously thought offers a cautiously optimistic outlook for the emerging economy built around lunar polar resources, while reminding the community that the ground truth beneath the shadows remains one of the most important open questions in planetary exploration.

Subject of Research: Trafficability and excavatability of icy lunar regolith simulants quantified using cone penetration

Article Title: Trafficability and excavatability of icy lunar regolith simulants quantified using cone penetration

Article References: Ricardo, D., Hodgkinson, J., Rhamdhani, M. A., & Brooks, G. (2026). Trafficability and excavatability of icy lunar regolith simulants quantified using cone penetration. Space and Planetary Resources, 2(1), Article 9. https://doi.org/10.1007/s44461-026-00015-w

Image Credits: AI Generated

DOI: 10.1007/s44461-026-00015-w

Keywords: Trafficability, excavatability, lunar, regolith, simulants, quantified, cone, penetration, scientific research