# Glycolysis for PET Bottle Recycling

Topic radar · 177 documents · 2021-01-01–2026-11-09 · updated 2026-11-11

Canonical: https://slate.greyb.com/radar/glycolysis-for-pet-bottle-recycling/innovations

## In one minute

### The plant is a loop with three stages, and the reaction is the cheap one
Glycolysis breaks the ester bonds linking PET's repeating units. Ethylene glycol is fed in as both solvent and reactant, and it swaps places with the polymer chain's own diol links. Heat and a catalyst drive the swap. The products are bis(2-hydroxyethyl) terephthalate, BHET, which is the monomer, plus short chains of a few units called oligomers and free ethylene glycol. Everything in the field uses that same reaction at 65 to 255°C.

What differs is where the operator stops, and what happens to the mixture afterwards. The reactor itself converts most of the PET in minutes. The separation of glycol from water, the purification of the monomer, and the repolymerisation back to bottle-grade resin are where the plant's cost and most of its equipment sit. The engineering effort concentrates on those downstream stages, and the reaction conditions are chosen to make them easier.


### The only product with a price on it is the process licence, and the money behind the plants is grant and equity
No glycolysis plant is shown being paid for monomer in volume. What is on sale is a licence to run the reaction. Axens markets Rewind PET as a complete package to buyers worldwide across packaging and textiles, built from a continuous glycolysis step plus proprietary purification, with performance guarantees and equipment inside the offer rather than a monomer specification sold separately.


### Patent output has quadrupled since 2021, and universities and small materials firms are producing most of it
Patent publications in this field rose from five in 2021 to twenty in 2026 through 11 November, a partial year that already exceeds every full year before it. Whatever the plant economics are doing, the rate of invention has not slowed.


### Continuous operation is solved in hardware, and flake size gets a workaround
One design abandons the stirred kettle for a column. An alcoholysis tower takes crushed PET in at the top, contacts it with ethylene glycol vapour and hot liquid in stages, removes residual water, and completes the reaction in the same vessel. It has distinct feed, leaching, plug-flow feed, friction-washing and discharge sections, which is what mass production of bottle-to-bottle product requires.


### The one chain that puts chemically recycled polyester into a bottle is Japanese, and brand owners had to assemble it
In May 2025 nine Japanese companies launched an initiative to make beverage bottles from non-food PET through chemical recycling, described as a first in Japan. The membership is the tell. Kirin Holdings, Kirin Beverage and Asahi Soft Drinks supply the bottle demand. TDK, Murata, Kao and FANCL supply the non-beverage PET. JEPLAN and its plant company PET REFINE TECHNOLOGY run the recycling.


### Half the apparent breadth is one invention published twice
The publication counts overstate how much distinct work exists. Grant publications sit beside their own applications, and a single invention can appear as an application, a granted patent and several national members of one family. Read the applicant list net of duplicates and the field has perhaps thirty distinct inventions rather than sixty publications.


### Heterogeneous catalysts exist mainly to keep the metal out of the product
Almost all filed catalyst work is a solid that can be filtered or packed, and the claims that separate them are recovery and working temperature more than yield. Pellets in a fixed bed. A magnetically recoverable cerium-doped nickel-iron layered double hydroxide, pulled out of the hot mixture with a high-gradient magnetic separator and reused four to six cycles. A zinc-titanium layered double hydroxide that runs at atmospheric pressure. An organic base, DBU, grafted onto porous silica, which keeps metal ions out of the monomer entirely. Carbon-doped boron nitride, filed for the same reason. An ionic liquid used at 0.5 to 3% loading, which needs no post-treatment of the product. A nickel phosphide on silica aimed at coloured and multilayer feed.


### Recycled PET sells below virgin resin wherever a mandate does not force the purchase
In April 2026 recycled polymers traded at a discount to virgin for the first time globally, as oil-driven virgin prices rose faster than recycled grades. The pattern has held in Asia. Regular-grade rPET flake was assessed at 850 a tonne and premium flake at 928.57, with buyers covering only immediate needs because most of the region has no recycled-content law.


## Overview


Glycolysis breaks PET's ester bonds in ethylene glycol and stops where the operator wants it to. The reactor converts most of the PET in minutes; the plant's cost lives in separating glycol from water, purifying the monomer and repolymerising to bottle grade.


## Research

### The reaction temperature spans more than 90 degrees, which sets the plant's steam circuit
The filings report very different operating windows. A composite catalyst claims above 85% yield and above 99% purity between 100 and 130°C, using film, bottle flake or fabric as feed. Zinc acetate with DBU in dimethyl sulfoxide runs at 140 to 155°C in 60 to 100 minutes. The granular fixed bed runs at 150 to 190°C. The magnetic layered double hydroxide route runs at 190 to 220°C. Kumoh's two-step, which dissolves and reprecipitates the PET first to raise its surface area, glycolyses at 170 to 200°C and holds diethylene glycol terephthalate to 0.1 mole percent or less.


### Purifying the monomer is a staged crystallisation with an adsorption polish
The ChemPet family is the most complete published train. Water is added to the hot glycolysate in at least two stages, with the ethylene glycol to water ratio and the temperature both falling. Barrier polymers such as polyamides, EVOH and PVA precipitate at 60 to 90°C while BHET stays dissolved. Cooling to 30 to 60°C drops the oligomers. More water and cooling to 5 to 25°C crystallises the BHET, which is then redissolved in water and treated with an adsorbent. The stated reason for staging is plant cost: keeping the liquid volume small avoids the decantation and filtration load that would make the process impractical at industrial scale.


### One route deletes the glycol-water separation the whole cost argument rests on
Glycolysis leaves a dilute mixture of ethylene glycol and water, and distillation is what makes the line expensive to run. A water-free process replaces it. A micro-sized MgO/SiO2 catalyst gives 95.1% BHET yield, and an organic-solvent nanofiltration step concentrates and purifies the product without phase change. Overall recovery is 93.2%, energy use falls 86% against distillation, and the modelled payback is 2.48 years at 1,000 tonnes of BHET per day.


### One continuous industrial design is running, and it is built to be copied in modules
revalyu's Nashik site runs a patented continuous low-temperature glycolysis on monoethylene glycol, producing monomers it repolymerises into rPET. The design is sold as modules of 50 to 200 tonnes per day, which lets a plant be built at the size of its local bottle supply. Two lines are running and together produce 160 tonnes per day of chips and polymer. A third, fully financed as part of a $100 million investment, adds 120 tonnes and takes the site to a nameplate of 280 tonnes per day.


### The extruder carries a hazard the filings do not address
Ethylene glycol boils at about 196°C. Inside a vented extruder running above that temperature, glycol vapour can escape, which disrupts the glycol to PET ratio and exposes the operator. A study proposes a decreasing temperature profile along the barrel, from PET melt temperature down below 196°C at the die, to protect the die zone and suppress repolymerisation. It tests the idea in a torque rheometer standing in for a twin-screw extruder, not in an extruder. It found 99% of depolymerisation in five minutes at 255°C, with number-average molecular weight falling from 40,000 to 400 g/mol.


### The outlet is chosen upstream of the reactor, and it changes what gets built
Because the reaction can be stopped at any chain length, the same equipment serves three products, and the filings are increasingly explicit about which they target. Stopping at a number-average molecular weight of 5,000 to 7,000 g/mol gives an oligomer that goes directly into polycondensation with no separation step at all. Donghua's midpolymer route controls degree of polymerisation to 3 to 30 with a polydispersity below 3, then feeds 10 to 100 wt% of it into a five-stage virgin polyester line, so the recycled content becomes a blend component instead of a separate product. A one-pot route adds adipic acid to make aromatic polyester polyols for polyurethane instead of bottle resin. A Polish filing goes the other way, acid-hydrolysing the glycolysate to terephthalic acid and glycol, skipping BHET purification entirely.


## Market

### revalyu's Georgia plant now costs four times its 2023 budget for 2.4 times the output
In January 2023 revalyu announced a $50 million project at Statesboro, Georgia, buying 43 acres for a 225,000 square foot plant with initial capacity of up to 225,000 pounds per day. The current plan is 90,000 tonnes a year for $200 million, operational in the middle to end of 2027.


### The one independent throughput figure in the record is small enough to check against a single plant
A vendor analysis put the post-consumer polyester depolymerization technology market at about $1.8 billion in 2024, processing an estimated 420,000 tonnes of feedstock, and forecast it at $6.1 billion by 2032 on 16.4 percent annual growth.


## Strategy

### The companies that actually license glycolysis hold the broadest families, and they file in a different set of offices
The industrial licensors are in the patent record, with wider protection than anything in the academic cluster. IFP Energies Nouvelles with JEPLAN holds a glycolysis-plus-purification family in a dozen or more jurisdictions, covering conditioning, glycol separation under ten minutes, adsorption decolourisation and a diol recycle loop. It adds a thin-film or short-path evaporation step for BHET recovery, with a purge section to stop pigment fouling the evaporator. Eastman holds a dialkyl terephthalate recovery family in seven jurisdictions plus a depolymerisation additive route, and Shintech with JEPLAN covers pigment decomposition and distillation of coloured BHET in five jurisdictions.


### Catalysis is where new organisations keep arriving, and purification is where they stop
The two areas are moving in opposite directions. Catalysis published four patent documents in 2025 and five in the ten months of 2026, and five of its filers are new to the field in that window: Changhua, Dongguan, Hubei, Huiqun and IIT Mandi. Monomer Recovery went from five to three over the same comparison, with no new organisations at all.


### Process engineers are attacking the separation cost, and they are pricing whole plants
The credible attack on the separation cost is coming from chemical engineering groups, and the work is a plant model. A Korean group spanning Chung-Ang University, the Korea Institute of Industrial Technology, the Korea Research Institute of Chemical Technology and Korea University combined a micro-sized MgO/SiO2 catalyst with organic-solvent nanofiltration, reporting 95.1% BHET yield, 93.2% overall recovery, an 86% energy cut against distillation and a 2.48 year payback at 1,000 tonnes of BHET per day.


## Areas

- Depolymerization · 114 documents
- Catalysis · 32 documents
- Monomer Recovery · 15 documents
- Plant Deployment · 10 documents
- Feedstock · 1 documents
- Market Dynamics · 1 documents