Turning peels into pavers: How Penn designers turn food scraps into biodegradable building materials

Most days during the semester, at nearly every hour, the campus’ dining halls are buzzing with activity. Food may bring people together, but meal preparation often comes with an unavoidable sustainability adversary: waste.

To that end, Laia Mogas-Soldevila, an assistant professor of architecture at the Weitzman School of Design, and research associate Yasaman Amirzehni are turning food scraps into sustainable, biodegradable products.

Their initiative addresses a staggering environmental reality: that unavoidable food waste—what people don’t usually eat such as bones, peels, rinds, and stems—accounts for approximately 14.8% of restaurant food waste and typically ends up in landfills. As it decomposes, it releases potent greenhouse gases, contributing to a cycle of climate degradation. By finding new ways to upcycle, researchers are treating every scrap that enters Mogas-Soldevila’s DumoLab as a crucial ingredient rather than a disposal problem.

“Our lab works with materials and processes that the Earth can digest the same way you can digest food,” explains Mogas-Soldevila. “We are looking at anything that enters the lab as an ingredient that can go back to the Earth’s metabolic system without toxicity or harm to the planet.”

For the past two years, DumoLab researchers and students, led by Amirzehni, have partnered with staff at Penn Dining and the catering company Bon Appétit to collect unavoidable food waste before it reaches students’ plates. In the lab, these ingredients are dried, ground, and mixed with natural, water-based binders to create a new category of “smart biocomposites.”

The scraps? Mostly rinds and peels from melons, pineapples, bananas, and citrus fruits, as well as eggshells. The products? Components designed for urban climate resilience, like thermal insulation blocks, shingles that grow carbon-capturing moss, and tiles that change color to diagnose soil toxicity. Now, Amirzehni is enhancing the compressive strength of biodegradable building materials to develop a biocomposite suitable for indoor and outdoor cladding applications, which could eventually serve as true structural components like load-bearing columns.

“When you enter a space with these biocomposites, the first thing you notice is the sensory experience,” Amirzehni says, noting that people often describe the rooms as “citrusy” or “earthy” in scent and uniquely textured.

Amirzehni believes these much-overlooked features are a crucial part of architectural psychology, stating that sensory details like scent and color deeply impact well-being and “the story that they live in within that building.”

For Mogas-Soldevila, the drive to champion these organic composites is part of her broader interest in building smarter, more functional materials. For example, her team is also working on developing materials that incorporate microorganisms, like highway sound dampeners that use plants and symbiotic fungus to trap tire debris and soil contaminants, as well as biocomposites that change color when exposed to UV light or temperature changes, indicate soil toxicity using pH-sensitive pigments, or trap pollutants.

“This is a no-brainer to me because of all the potential organic materials hold,” she adds, citing how working with living and once-living materials such as structural biopolymers opens doors to sensing properties that inert composites cannot simply be imbued with.

However, before this “positive biomaterial revolution” can fully substitute ubiquitous materials like concrete and plastic on a commercial scale, Mogas-Soldevila acknowledges a critical hurdle: the rigorous work of standardization.

Because organic waste is highly variable—a plant’s fibers change based on climate, water intake, and growth conditions—the lab must do hours of work, recording and analyzing every failure, every success, and every anomaly to categorize these wastes into predictable “families.” Looking ahead, the lab is proving the longevity of these new families by partnering with historic preservation experts—placing their natural blends into accelerated weathering chambers to accurately predict how the materials will behave 70 years into the future.

By establishing a standardized vocabulary and extensively testing these recipes, Mogas-Soldevila’s team, among other international bio-based architectural material researchers, is helping to build an open-source data library of biomaterial performance. Ultimately, this shared database will be used to train AI models, enabling future architects to reliably, consistently predict how different natural blends will behave and making the use of upcycled food waste as standard as mixing concrete.

As labs contribute to this growing open-source knowledge, the researchers see the tipping point of widespread biomaterials adoption rapidly approaching.

“It’s coming. The more we test, the more we know, the more these AI models will be able to predict. I am hopeful,” Mogas-Soldevila says.