The core of the technique is a natural phenomenon that scientists call microbiologically induced calcite precipitation, or MICP, and which engineers more plainly refer to as biocementation. When certain types of bacteria are combined with sand or other solid grains under the right chemical conditions, they trigger the formation of calcium carbonate, the same mineral that makes up limestone, chalk and mussel shells. As Antorveza Paez puts it, the team is using bacteria to produce a kind of mineral glue. The glue is not a sticky organic substance in the conventional sense. Rather, microscopic calcite crystals precipitate in the gaps between individual sand grains and interlock with one another, mechanically locking the particles into a rigid, coherent structure. The result looks and behaves, at least in its early forms, much more like a masonry solid than like anything biologists typically grow in a petri dish.

Biocementation itself is far from new. The workhorse bacterium, Sporosarcina pasteurii, has been studied for years, notably as a way of stabilising loose subsoil beneath roads and foundations, and it has also found a role in so-called self-healing concrete, where bacterial spores embedded in the mix germinate when cracks admit moisture and seal the damage with freshly precipitated mineral. What Antorveza Paez recognised was that the field had remained, in a sense, a servant to conventional materials. Bacteria were added to existing building products or injected into the ground; nobody had yet treated the biological process itself as the primary manufacturing route for architecture. Her innovation was to fuse biocementation with digital manufacturing, two research fields that had until then developed largely in isolation, and to ask what was missing. In her own words, what was lacking were manufacturing processes by which these biological processes could be controlled as required and used to make architectural elements.

Her doctoral work culminated in three scientific studies describing two distinct fabrication methods. The first is 3D biocement printing. The team formulated a bio-ink composed of a gel based on water and two natural polymers, alginate and xanthan, mixed with sand and live bacteria. Like a traditional 3D printing paste, this bio-ink can be extruded layer by layer into any digitally defined geometry. At this stage the printed object is fragile, held together only by the gel matrix, and it would not survive handling. The decisive transformation happens next, when the printed structure is immersed in a mineralisation bath containing urea and calcium chloride. There the bacteria get to work, using an enzyme called urease to break down the urea. This reaction releases carbonate ions into solution, along with ammonium as a by-product, and the carbonate ions combine with the calcium from the calcium chloride to precipitate calcium carbonate, the mineral of limestone, precisely wherever the bacteria are located inside the printed object.

The mineralisation is slow and deliberately controlled. Depending on the strength required of the finished component, the entire process takes between two and five days, and each day the object receives a fresh mineralisation bath so that the bacteria do not exhaust the chemical precursors needed to keep forming limestone. Once the bath cycles are complete, the structure is left to dry. At that point the transformation is finished: the object has become a solid mineral body, the bacteria are no longer active, and no heat source has been used at any stage. Antorveza Paez compares the sequence to baking a cake. The dough, she notes, is malleable and delicate, and only the oven can turn it into something solid. In her process the bacteria play the role of the oven, slowly setting the printed structure until it is bonded, solid and dry at ambient temperature. Among the demonstrator objects produced this way is a facade element that could serve as sun protection or as a partition wall.

The second method takes a radically different route to the same mineral, and it dispenses with the immersion bath altogether. In robotic biocementation spraying, a robotic arm applies bacteria, mineralisation substances and sand to a surface in repeated, alternating passes, building the material up layer by layer directly in place. The inspiration here is explicitly biological and geological: stromatolites, the layered sedimentary rock formations that arise in nature when microorganisms deposit minerals over immense stretches of time. The robotic process compresses that ancient rhythm into weeks. At the 2025 Biennale Architettura in Venice, the team let the spraying process run uninterrupted for six months, allowing visitors to watch a biologically mineralised structure grow in real time. As Antorveza Paez observes, the demonstration proved that the entire biocementation workflow can be automated.

The two processes are complementary rather than competing, and they were developed with different priorities in mind. Printing, Antorveza Paez explains, is geared more towards discrete architectural elements, while the spray process is designed for surfaces. The printing research concentrated on materials science: which bio-ink formulation works, and how it can be reliably extruded and mineralised. The spraying research was chiefly concerned with the machine itself, and with the considerable control problem of orchestrating three separate streams of material delivered in precise alternation. Reflecting those different emphases, the two processes also stand at different stages of maturity. The spray method in particular still requires substantial research into its mechanical properties, including how the grains and the bonding medium are arranged within the built-up object, questions that the printing studies have already begun to address for their own process.

For Antorveza Paez, the deeper significance of the work lies in the relationship it forges between architects and their materials. Architects, she points out, normally know nothing about what went into making a material, which is manufactured somewhere else entirely. She describes her fascination with the fact that nature has found ways of producing very sophisticated materials, and with the possibility of making those biological strategies applicable to the problems the built environment faces today. That orientation also carries a sobering awareness of the process’s limitations. A third study, currently underway, examines the environmental impact of biocementation and searches for solutions to its most significant drawback: the ammonium released during the urease reaction is damaging to the environment. If the disposal of this chemical by-product is not factored into the design of biocement systems from the start, Antorveza Paez warns, using biocement for construction will not be the sustainable practice it promises to be.

Equally clear-eyed is her assessment of where the material stands relative to the industrial benchmark. Biocementation, she admits, currently resembles mortar more than concrete. The comparison is instructive, because the recipe for concrete, specifying how many grains of each size are combined and in what proportions, has been refined for over a century and continues to be fine-tuned to this day. For biocementation, that systematic optimisation is only just beginning. Researchers know that the bacteria can in principle be combined with certain grain sizes, but the question of how to translate that into a long-lasting material that is as stable as possible remains open. The pivotal point, as Antorveza Paez concludes, is no longer whether biocement works, but rather how well its properties can be managed. In other words, the bacterial masons have proved they can lay the stones; the century-long work of engineering the mix now begins.

Subject of Research: Biocementation: using bacteria to bind sand into solid architectural materials

Article Title: When bacteria become masons

Article References: When bacteria become masons. (n.d.). Original publication

Image Credits: AI Generated

DOI: Not provided

Keywords: biocementation, MICP, Sporosarcina pasteurii, 3D printing, robotic spraying, calcium carbonate, ETH Zurich, sustainable construction, digital fabrication, cement emissions, bio-ink, architecture

The motivation behind the work lies in one of the most pressing challenges of the coming decades. The world’s population, which surpassed 7.6 billion in 2018, is projected to reach 9.2 billion by 2050, and food demand is expected to rise by anywhere from 59 to 102 percent. Global meat consumption is anticipated to climb in parallel, reaching an estimated 4.55 million tons by 2050, with developing countries accounting for 56 percent of feed grain consumption. Against this backdrop, cultured meat, produced by cultivating livestock cells rather than slaughtering animals, has attracted intense interest as a way to conserve finite resources, address animal welfare concerns, and reduce the environmental footprint of protein production. Yet conventional cultured meat approaches, which rely on cells alone, have largely been limited to simple forms such as patties.

This is where 3D bioprinting enters the picture. By depositing cell-laden materials layer by layer, bioprinting can produce constructs in a wide variety of shapes that more closely mimic traditional meat products. It also allows researchers to optimize the cell growth environment, shortening production time, minimizing raw material use, and improving cost-effectiveness, while opening the door to customized meat tailored to individual dietary preferences or health needs. The critical enabling ingredient in this process is the bio-ink itself, typically built from materials such as alginate and gelatin, whose properties ultimately determine the quality of the final product. Gelatin, a degradation product of collagen, is prized for its excellent biocompatibility and its ability to support cell attachment, proliferation, growth, and differentiation, all of which are essential for natural and efficient tissue formation.

What has been missing, the authors argue, is a food-focused perspective. Most prior studies of gelatin-based bio-inks have been rooted in tissue engineering and biomedical applications, prioritizing biocompatibility and mechanical performance without adequately considering edibility, food-grade applicability, and source-dependent functionality within a food system. For cultured meat, bio-ink materials must simultaneously provide structural stability and printability while remaining suitable for consumption and scalable for food production. Systematic comparisons of edible animal-derived gelatin sources under standardized formulation conditions have been scarce, so the Korean team set out to fabricate bio-inks from bovine, porcine, and fish gelatin powders under unified conditions and to evaluate how source and concentration independently and interactively influence rheological behavior, printability, and structural properties.

The experimental design was straightforward but rigorous. Beef gelatin from Holstein cattle, pork gelatin from Landrace-Yorkshire-Duroc pigs, and fish gelatin from salmon, all with molecular weights ranging from roughly 10 to 100 kilodaltons depending on source, were dissolved in distilled water at concentrations of 4, 12, 20, 32, and 40 percent by weight at 45 degrees Celsius. Because gelatin-only formulations showed insufficient thermal stability at the 37 degrees Celsius incubation temperature used in cell culture, each gelatin solution was mixed 1:1 with a 6 percent sodium alginate solution, a supporting polymer that improves structural stability during printing and post-printing incubation. The blends were loaded into printing cartridges and stored at 4 degrees Celsius before testing. Printability trials used a Cellink BIO-X bioprinter with 27-gauge conical nozzles, a print bed held at 10 degrees Celsius, a print speed of 5 millimeters per second, and a pressure of 150 kilopascals, followed by ionic cross-linking in a 100 millimolar calcium chloride bath.

The results revealed systematic patterns across nearly every measured property. As gelatin concentration increased, pH decreased in all three sources, a trend the researchers attribute to the rising content of acidic amino acids such as aspartic and glutamic acid, which elevate hydrogen ion concentration as collagen breaks down into gelatin. Notably, fish gelatin samples consistently showed the highest pH values, a difference thought to reflect the vitamins, minerals, and amino acid profiles specific to each livestock species. Because higher pH levels are associated with increased metabolic activity in cells, the authors suggest that bio-inks with elevated pH may be particularly suitable as scaffold materials for cultured meat production. Color measurements told a complementary story: lightness declined with increasing gelatin concentration while redness and yellowness rose, and beef gelatin samples were significantly more yellow than the others, likely because bovine gelatin contains more lysine and other amino acids susceptible to the Maillard reaction, the non-enzymatic browning that occurs when the powder is dried in the presence of atmospheric oxygen.

Viscosity, arguably the most important rheological property for extrusion printing, showed the sharpest source-dependent differences. Initial viscosity increased with gelatin concentration in all formulations, but pork gelatin exhibited the highest values across every concentration tested. The explanation lies in molecular architecture: pork gelatin possesses greater structural integrity of its beta and gamma chains, creating physical entanglements that impede polymer flow, whereas fish gelatin has a looser structure with weaker intermolecular bonds and lower levels of proline and hydroxyproline, the amino acids that govern gelation and melting point, making it heat-sensitive and unable to build high viscosity even at elevated concentrations. Intriguingly, the most concentrated pork gelatin formulations, at 32 and 40 percent, showed viscosity that declined over time, apparently because excessive intermolecular interactions at very high concentrations prevent the formation of a stable gel network and increase internal stress. This observation led the team to recommend keeping porcine gelatin below 32 percent when manufacturing bio-ink.

Printability tests drove the study’s central conclusion home. At 4 percent, all three gelatin types extruded as a liquid with no visible lattice formation whatsoever. Beef and fish gelatin at 12 and 20 percent could be extruded but collapsed rapidly, retaining lattice structures only up to 0.3 centimeters in height. Pork gelatin, by contrast, formed partial lattices at 12 and 20 percent and was the only formulation to maintain scaffold integrity up to 0.5 centimeters. At 32 and 40 percent, pork gelatin became so solidified that it could not be extruded at all, a consequence of hydrogen bonding and van der Waals forces increasing the material’s elastic restoring force until it behaved more like a solid than a printable viscoelastic fluid. The richer proline and hydroxyproline content of porcine gelatin appears to foster stronger hydrogen bonding and a more stable three-dimensional network, providing the internal binding strength needed to resist gravitational collapse after deposition.

Structural analysis and statistical correlation rounded out the picture. Under the microscope, pork gelatin scaffolds displayed rougher surfaces than the other sources, a feature that may actually benefit cultured meat production by increasing surface irregularities and contact area for cell attachment. The 20 percent pork gelatin sample showed significantly higher grid size and gel strength than all other formulations, a combination the researchers highlight as a favorable balance: larger grid sizes facilitate nutrient diffusion and material exchange within the printed structure, while greater gel strength preserves scaffold integrity during handling and culture. Pearson correlation analysis quantified these relationships, revealing a strong positive correlation between gel strength and viscosity with a coefficient of 0.88, and between grid size and gel strength with a coefficient of 0.89. Even color proved tightly linked, with redness and yellowness showing a near-perfect correlation of 0.99, suggesting that bio-ink color tone can be controlled predictably, an attribute the authors note matters for the appearance quality control of commercial cultured meat.

Taken together, the findings offer a practical recipe for the emerging cultured meat industry: choose pork gelatin, keep the concentration near 20 percent, and pair it with alginate for thermal stability. The work also underscores a broader lesson in biomaterials science, that seemingly minor differences in molecular composition, down to the amino acid content of a protein extracted from a particular animal’s skin, can cascade into dramatic differences in how a material flows, gels, and holds its shape under a printing nozzle. As the demand for alternative proteins accelerates, studies like this one, which treat food-grade applicability as seriously as mechanical performance, will help determine whether 3D-printed cultured meat can move from laboratory demonstration to dinner plate.

Subject of Research: Physicochemical and printability properties of animal-derived gelatin bio-inks for 3D-printed cultured meat scaffolds

Article Title: Physicochemical properties of 3D bio-ink prepared with beef, pork, and fish gelatin

Article References: Kang, K.-M., Lee, S.-H., & Kim, H.-Y. (2026). Physicochemical properties of 3D bio-ink prepared with beef, pork, and fish gelatin. Food Science of Animal Resources, 46(1), Article 66. https://doi.org/10.1007/s44463-026-00071-y

Image Credits: AI Generated

DOI: 10.1007/s44463-026-00071-y

Keywords: gelatin, bio-ink, 3D bioprinting, cultured meat, pork gelatin, printability, viscosity, gel strength, scaffolds, sodium alginate, food science, collagen