hueslenfrucht-gele

Gels with a directional structure made of chickpeas, green peas, red and beluga lentils. Credit / Copyright: Andrea Bach / ETH Zurich

Whole Beans Can Get a Meat-Like Texture With a Simple Freezer Trick

In A Nutshell

  • Researchers at ETH Zürich found that freezing whole legume mixtures in one direction, instead of freezing them normally, gives them a firmer, layered texture closer to meat or cheese.
  • The trick works on whole chickpeas, lentils, peas, and beans with no processing into powders or isolates, and no specialized equipment beyond a standard freezer.
  • At higher concentrations, chickpea gels reached a chewiness close to mozzarella, while lower concentrations landed close to silken tofu.
  • A simple insulated container in a regular home freezer reproduced the same layered effect as specialized lab equipment, though a reliable kitchen recipe still needs more development.

Anyone who has bitten into a rubbery veggie burger knows the letdown. Plant-based meat alternatives promise the texture of the real thing and often deliver something closer to wet sawdust. Food companies have spent years and a lot of money trying to fix that problem with pricey machinery and heavily processed protein powders.

A team of researchers at ETH Zürich has shown that a much simpler tool, a freezer, can create surprisingly structured textures from whole legumes, using nothing more than whole chickpeas, lentils, beans, or peas. The findings appear in the journal npj Science of Food.

Freezing Ice in One Direction Builds a Meat-Like Grain

Researchers started with something almost anyone could do in their own kitchen: they soaked chickpeas in water, blended them into a slurry, and heated the mixture until it thickened into a firm, jelly-like block. Heat makes the starch swell and the proteins change shape, helping the mixture set into a gel.

From there, the block was split into three groups. One batch went straight into the refrigerator. A second was frozen the normal way, with cold air hitting it from every side. The third was poured into molds built to freeze from the bottom up only, so that ice crystals grew in a single direction instead of scattering randomly through the mixture. The researchers call this directional freezing.

That design choice mattered a great deal. As ice grows in a straight line, it pushes the solid bits of bean, the starch and protein, out of the way and packs them into narrow channels between the forming ice. When the block later thaws, those packed regions bond together into a structured, layered scaffold instead of collapsing back into a shapeless blob. Under a microscope, the payoff was obvious: refrigerated samples looked like uniform mush, conventionally frozen ones showed patchy alignment pointing every which way, and directionally frozen samples showed clean, organized layers running in one direction, similar to the grain in a piece of wood.

Physical testing backed up what the images showed. Directionally frozen chickpea samples were consistently firmer than the other two groups, roughly one and a half to three times harder, and held together better under repeated pressing and squeezing.

Texture could even be adjusted by changing how much solid bean material went into the mixture. At lower amounts, chewiness was close to silken tofu. At the highest concentration tested, chewiness was close to mozzarella, though still well below firm tofu.

directional freezing
From legumes to a fibrous food: Freeze-structuring can even be carried out in a household kitchen. Credit / Copyright: adapted from Bach A, Perler E et al., NPJ Science of Food 2026, ETH Zurich

Nearly Every Legume Tested Formed the Same Layered Structure

Researchers didn’t stop at chickpeas. They ran the same freezing process on red lentils, green peas, black beans, mung beans, black-eyed peas, and soybeans, all prepared as whole, minimally processed mixtures rather than refined ingredients.

Every legume tested formed an aligned structure at a moderate concentration, though not all of them held together equally well. Red lentils and mung beans produced firm, stable gels with little liquid leaking out. Black beans came out somewhat softer. Soybeans struggled the most at lower amounts, likely because their higher fat content interferes with the starch that holds the structure together. Adding more solid soybean material to the mix improved the texture considerably.

A pattern also emerged linking a bean’s basic makeup to how well it performed. Beans with more starch than protein, like red lentils and mung beans, formed stronger, more self-supporting gels. Beans higher in protein than starch, like soybeans, tended to produce weaker ones, likely because starch is what does most of the structural work.

Researchers also tested whether the method could survive outside a lab, using nothing more than a plastic container lined with foam insulation and a regular household freezer. That simple setup reproduced the same layered effect as the specialized lab equipment, though turning it into a reliable kitchen recipe will take more work.

chickpea purée
Freeze-structured chickpea purée under the scanning electron microscope. The section shown here 1 by 2 millimeters in size. Credit / Copyright: Elin Perler / ETH Zurich

The Method Uses the Whole Bean Instead of Breaking It Apart

Most plant-based meat on the market today relies on breaking legumes down into separate parts, protein powders, starch concentrates, and so on, then rebuilding them into something that resembles meat. That approach often requires ingredient fractionation, specialized equipment, and additional processing before the ingredients are rebuilt into a structured food.

Freeze structuring skips that step entirely, working directly on the whole legume and keeping it largely intact instead of separating it into refined fractions. Making tofu, for instance, separates out a byproduct called okara that carries away much of the soybean’s fiber, along with some of its fat and protein. Because freezing relies on cold-chain technology that already exists throughout the food supply chain, from industrial cold storage to the freezer in an ordinary kitchen, the method may be adaptable to equipment many kitchens and food-processing facilities already have.

Legumes have long been called an overlooked answer to global nutrition problems: they’re cheap, packed with nutrients, and grow in a wide range of conditions. What has held them back isn’t the ingredient itself but the gap between raw beans and food people actually want to eat. This research suggests that producing more structured whole-legume foods may require less processing than some current methods. Closing that gap may not require a factory at all. A carefully set up freezer could be part of the answer.


Disclaimer: This article is based on peer-reviewed research. The findings reflect laboratory testing of texture and structure and do not represent a finished commercial product, a sensory taste test, or a nutritional comparison with existing meat alternatives. Readers should not attempt to replicate specialized laboratory procedures without following validated safety and food-handling practices.


Paper Notes

Limitations

The study notes that its data were generated using commercially sourced raw materials that can vary from batch to batch, which limited how broadly the specific raw data could be shared, though the researchers state the method itself remained robust across different legumes and batches. The paper also points out that future work is needed to systematically study how factors like protein gelling behavior, starch behavior, and formulation conditions such as acidity influence the final structure and texture, as well as how the process might affect protein digestibility and nutrient absorption. The researchers additionally note that translating the method into consumer and home kitchen use will require developing simple, repeatable preparation steps covering soaking, milling, heating, and freezing with standard kitchen equipment.

Funding and Disclosures

The Chair of Food Structure Engineering acknowledged financial support from Nestlé, Bühler, and Givaudan through the ETH Zürich Foundation. The paper states the funders played no role in the study’s design, data collection, analysis, interpretation, or the writing of the manuscript. Open access funding was provided by the Swiss Federal Institute of Technology Zürich. The authors declared no competing interests.

Publication Details

Title: “Freeze structuring unlocks minimal-processing strategies for legume texturization” Authors: Andrea Bach, Elin Perler, Lenja S. Lemcke, and Patrick A. Rühs, Institute of Food, Nutrition, and Health, Department of Health Sciences and Technology, ETH Zürich, Zürich, Switzerland Journal: npj Science of Food, published in partnership with Beijing Technology and Business University and the International Union of Food Science and Technology Citation: npj Science of Food (2026) 10:258 DOI: https://doi.org/10.1038/s41538-026-00991-5

About StudyFinds Analysis

Called "brilliant," "fantastic," and "spot on" by scientists and researchers, our acclaimed StudyFinds Analysis articles are created using an exclusive AI-based model with complete human oversight by the StudyFinds Editorial Team. For these articles, we use an unparalleled LLM process across multiple systems to analyze entire journal papers, extract data, and create accurate, accessible content. Our writing and editing team proofreads and polishes each and every article before publishing. With recent studies showing that artificial intelligence can interpret scientific research as well as (or even better) than field experts and specialists, StudyFinds was among the earliest to adopt and test this technology before approving its widespread use on our site. We stand by our practice and continuously update our processes to ensure the very highest level of accuracy. Read our AI Policy (link below) for more information.

Our Editorial Process

StudyFinds publishes digestible, agenda-free, transparent research summaries that are intended to inform the reader as well as stir civil, educated debate. We do not agree nor disagree with any of the studies we post, rather, we encourage our readers to debate the veracity of the findings themselves. All articles published on StudyFinds are vetted by our editors prior to publication and include links back to the source or corresponding journal article, if possible.

Our Editorial Team

Steve Fink

Editor-in-Chief

John Anderer

Associate Editor