Our coffee pods have to survive water at around 200 degrees Fahrenheit under pressure, keep coffee fresh on a shelf for months, and then dissolve or biodegrade. Very few materials do all three. We tested nine and published the results.
The study combined laboratory testing of biopolymer film samples we made ourselves with a review of the published literature on each candidate. Nine materials were scored against six criteria. Pullulan, polyvinyl alcohol and sodium alginate all dissolved and broke down well. Polylactic acid and cellulose were strongest under pressure. Only pullulan scored well on every criterion at once, which is why it became the pod.
The material brief for Naturista was written by surgery, not by packaging. Bioabsorbable implants are devices designed to be absorbed by the body over time: they provide temporary scaffolding while tissue heals, break down into byproducts the body metabolises, and remove the need for a second operation to retrieve them. Read as a specification rather than as medicine, that is exactly what a coffee pod should do.
Modern bioabsorbable devices work by controlled degradation. Polymers such as polylactic acid, polyglycolic acid and poly-4-hydroxybutyrate break down into natural byproducts the body eliminates; magnesium and zinc alloys corrode in a controlled way and are absorbed or excreted. Absorption is tuned by material and application, anywhere from several months to about two years, so support is present through the critical healing window and load transfers back to tissue as the implant disappears.
The clinical advantages are the same ones we wanted: no removal step, no permanent foreign object, and a force profile that becomes more natural as the material degrades. The known limitations are also familiar — bioabsorbable materials are generally weaker than permanent ones and cost more — which is precisely the trade we spent two hundred prototypes negotiating.
Note the overlap in the chemistry: polylactic acid appears both in the implant literature and as candidate four in our screening matrix. It is strong, and it only fully breaks down under industrial conditions. In a body that is acceptable, because the body is the processing environment. In a kitchen bin it is not, which is why the medical shortlist had to be re-scored against soil and a home compost heap instead of tissue.
An implant holds structure only while it is needed. A pod shell holds geometry only while water passes through the grounds. Both are disposable structures pretending to be permanent ones.
In bioabsorbable design the breakdown rate is engineered, not hoped for. We carried that over: dissolution time was measured against wall thickness rather than assumed.
Avoiding a removal surgery is the headline benefit of a dissolving implant, and avoiding the sorting, peeling and industrial composting step is ours. In both cases the byproducts have to be handled safely: medical polymers degrade into metabolites the body can process, and pullulan degrades into glucose and is treated as a dietary fibre.
Background reading on bioabsorbable implants: Lifetime Surgical, Bioabsorbable Materials: The Future of Surgical Implants, June 2025. Naturista is not a medical device and makes no clinical claim; the borrowing is conceptual.
Pullulan is a natural, water soluble polysaccharide biopolymer made from starch and agricultural plant waste. It is tasteless, it dissolves completely in hot water, and it biodegrades on its own. It is already used in food and medicine, and it is the whole pod.
Pullulan is a chain of glucose units. Every three form a maltotriose block joined by alpha 1,4 bonds, and those blocks link to each other by alpha 1,6 bonds. That alternating pattern is the whole trick. The regular 1,4 runs give the film its strength, and the 1,6 links between them keep the chain flexible and highly soluble, which is how it holds a shape in hot water and then disappears.
Glca1,4Glca1,4Glca1,6Glca1,4Glca1,4Glca1,6n
Not softens, not fragments. It goes into solution in hot water and leaves nothing behind.
Starch and agricultural plant waste are the inputs. There is no petroleum anywhere in the pod.
It does not change how the coffee tastes, which is why it beat the other candidate materials.
Already used in breath strips and pill capsules. The FDA reviewed it for food use and responded with no questions.
Pullulan is a prebiotic fibre. Swallowed, it is treated as fibre rather than as a foreign material.
Low oxygen permeability keeps the roast fresh on the shelf without an aluminium lid.
A pod has to do two opposite things. It must survive water at roughly 200 degrees Fahrenheit under pressure for the length of a brew, then dissolve or biodegrade once it is thrown away. Most materials can do one or the other.
| Stage | What the pod is subjected to | What the material has to do |
|---|---|---|
| Shelf | Months of oxygen exposure at room temperature | Keep oxygen out so the roast does not stale |
| Puncture | Needles pierce the lid and base | Hold together rather than shatter or tear |
| Extraction | Water near 200 °F, forced through under pressure, for 20 to 30 seconds | Keep its geometry so the water passes through the grounds, not around them |
| Disposal | Soil or a home compost heap, no industrial processing | Dissolve or biodegrade in days, leaving no fragments |
Solubility and strength usually move in opposite directions. A polymer chain that water can get into quickly is a chain that comes apart under stress. A chain packed tightly enough to resist hot water and pressure is one that soil bacteria struggle to digest.
Pullulan sidesteps that because of how it is built. Runs of three glucose units joined by alpha 1,4 bonds give the film order and tensile strength. The alpha 1,6 bond between each block interrupts that order, keeping the chain flexible and leaving room for water to penetrate. Strength from one bond type, solubility from the other, in the same molecule.
Glca1,4Glca1,4Glca1,6Glca1,4Glca1,4Glca1,6n
Before any material was scored, each performance band was tied to a measurable threshold drawn from our own experiments and from published benchmarks, so a rating means the same thing in every row.
| Criterion | High | Moderate | Low |
|---|---|---|---|
| Biodegradability | 85% or more mass loss after 30 days in soil | 50 to 84% mass loss | Under 50% mass loss |
| Solubility | Complete dissolution within 60 seconds at about 200 °F | Dissolution within 1 to 5 minutes | Partial or no dissolution after 5 minutes |
| Thermal stability | No visible deformation under water at about 200 °F | Minor softening or deformation without structural failure | Significant deformation or structural collapse |
| Formability | Easily moulded into capsule shape with strong structural retention | Mouldable with some deformation or fragility | Poor shape formation, or unable to retain structure |
| Food contact safety | FDA approved for direct food contact (GRAS) | Limited or indirect regulatory approval | Not approved for food contact |
| Scalability | Established large scale production and commercial feasibility | Emerging production with scalability challenges | Limited production, or experimental stage material |
Table 1 from the paper: operational definitions and classification thresholds for biopolymer performance evaluation.
Nine candidates against six requirements. A material had to clear all six to stay in. Reading across a row shows why each one dropped out, and reading down a column shows how few materials clear any single requirement.
| Candidate | Biodegrades without industrial processing |
Dissolves completely in water |
Forms a shell castable and mouldable |
Thermally stable at 200 °F under pressure |
Food safe and allergen free |
Scalable at pod volumes |
|---|---|---|---|---|---|---|
| Pullulan | ||||||
| Polyvinyl alcohol | ||||||
| Sodium alginate | ||||||
| Polylactic acid | ||||||
| Cellulose | ||||||
| Hydroxypropyl methylcellulose | ||||||
| Gelatin | ||||||
| Seaweed extract | ||||||
| Chitosan |
Pullulan is the only row without a disqualifying mark. Its one compromise is supply: it is produced commercially today, but by relatively few manufacturers, which is a cost and concentration risk rather than a technical one.
Every candidate was prepared the same way so the comparison held. Films were cast to a controlled thickness, cured, then put through the same sequence of tests.
Each biopolymer was dissolved to a working solution, cast onto a flat mould and dried to a target wall thickness. Thickness was the controlled variable, stepped from 0.2 mm to 1.0 mm, because it governs both structural strength and dissolution time and is the parameter a manufacturer would actually tune.
Samples of equal starting mass, 2 grams, were placed in soil and recovered at intervals across 90 days. Remaining dry mass was recorded at each interval, which produces the decay curves in Figure 1. Reporting mass rather than visual breakdown matters, because a film can fragment while losing very little mass.
Films at each thickness were immersed and timed to complete dissolution, giving the curves below. The useful result is not a single number but the slope: how sharply dissolution time rises as the wall gets thicker determines whether a wall strong enough to brew through is still a wall that disappears.
Shells were exposed to water at brewing temperature, around 200 °F, under pressure for the length of a full extraction. The pass condition was geometric: the shell had to keep its shape so water passes through the grounds rather than around a collapsed wall.
Results across all six criteria were normalised and plotted together, producing the radar comparison in Figure 3. This is the step that separates a material that is excellent at one thing from a material that is adequate at everything, which is what a pod actually needs.
Eight of them failed on at least one criterion. Here is where each one fell down.
| Material | What it did well | Why it was not used |
|---|---|---|
| Pullulan | Dissolved completely, held its shape hot, food safe, broke down at home | Selected |
| Polyvinyl alcohol | Dissolved readily and broke down well | A synthetic polymer. We did not want manufactured plastic in a product sold as plastic free |
| Sodium alginate | Dissolved well, seaweed derived | The film was too weak to hold shape under brewing pressure |
| Polylactic acid | Strongest structural performance of the group | Only composts in an industrial facility, which is the problem we set out to solve |
| Cellulose | Strong and widely available | Does not dissolve in water, so it leaves a capsule behind |
| Hydroxypropyl methylcellulose | Good film former, used in capsules | Too little margin between brewing temperature and the point where the film softens |
| Gelatin | Forms films easily and is well understood | Animal derived, which rules it out for a large share of coffee drinkers |
| Seaweed extract | Natural and biodegradable | Allergy risk, and inconsistent behaviour at brewing temperature |
| Chitosan | Natural, antimicrobial | Derived from shellfish, so it carries a serious allergy risk |
The measurements told us which material to use. Getting from that material to a shell that actually brews took another eighteen months.
The pattern that kept repeating was a trade off. Materials that dissolved quickly collapsed under pressure. Materials that held up under pressure would not biodegrade outside an industrial composter. Getting past that meant changing one variable at a time: film thickness, casting method, drying time, mould geometry.
We cast, moulded, brewed and destroyed more than two hundred shells over eighteen months on kitchen counters in New Jersey and Texas. In January 2025 a pullulan formulation held its shape through a full brew cycle and then dissolved on demand.
Some problems, poor shelf life above all, could not be solved on a kitchen counter, so both founders took them into university laboratories. Karan did materials science research at UT Austin and Aditya researched shelf life extension at the Rutgers food science lab. What they learned came back to Naturista and made the pods better.
Every work cited in Comparative Analysis of Biopolymer Candidates for Sustainable Coffee Pod Manufacturing, in the order they appear in the published reference list.
The claims on this site come from these. Where a figure is an industry estimate rather than a study, we say so.