Alkaline Thermal Treatment Converts Mixed Plastic Waste Into Hydrogen

Plastic waste to hydrogen conversion has typically hit the same wall regardless of method: either the chemistry only works on oxygenated polymers like PET, or it works on mixed streams but runs hot enough to emit substantial CO2. A study out of UCLA and Ewha Womans University, published in PNAS, reports an alkaline thermal treatment (ATT) process that converts unsorted polyethylene terephthalate (PET), polyethylene (PE), and polypropylene (PP) into hydrogen exceeding 90% purity in a single reactor. The mixed-stream conversion is the headline result, but the more useful detail is in the paper's own mass-correction analysis — the kind of process variable that changes how the yield numbers should actually be read.

Why Presorting Has Been The Real Constraint, Not The Chemistry

Prior low-temperature routes to hydrogen from plastic waste have been structurally limited rather than just inefficient. Solar-driven photoreforming and electrochemical conversion depend on oxygen-containing functional groups to initiate the reaction, which confines them to PET and excludes PE and PP — two of the most abundant polymers in the waste stream. High-temperature gasification sidesteps that limitation by brute force: it handles unsorted mixed plastic, but at 800–1,000°C and megapascal-range pressures, with direct CO2 emissions that scale accordingly. Neither constraint is a matter of degree; each route is built around an assumption — oxygen content or extreme temperature — that the other doesn't share.

ATT doesn't resolve this by being a superior reaction. It resolves it by using a different activation pathway entirely: sodium hydroxide under moderate heat, which lowers the energy barrier for polymer breakdown without requiring the feedstock to already contain oxygen or the reactor to run at gasification temperatures.

Why PET Reacts And PE/PP Don't

PET's ester linkages are the reason it responds to alkaline treatment at all. Hydroxide ions attack the electrophilic carbonyl carbon in the ester bond, cleaving the polymer chain — a reaction the study's density functional theory (DFT) calculations show is thermodynamically favorable, at −1.26 eV. PE and PP offer no equivalent target. Their backbones are saturated C–H and C–C bonds with no electrophilic site for hydroxide to attack, and the calculations confirm that direct nucleophilic attack on those bonds is thermodynamically prohibited, not merely slow.

PE and PP don't fail to react because the chemistry is weak — they fail because there's no electrophilic carbon for hydroxide to attack. A saturated C–H backbone gives the reaction nothing to hold onto.

This is why pristine PE and PP produce almost no gas under ATT conditions on their own. The reaction needs a foothold that these polymers don't naturally have.

The Oxidation Pretreatment: Activation At A Cost

To create that foothold, the researchers introduced a thermal oxidation pretreatment: brief exposure to air at elevated temperature before the plastic ever enters the alkaline reaction. This step installs carbonyl, hydroxyl, and ether groups along the polymer backbone — confirmed via FT-IR, where a distinct carbonyl peak at 1,718 cm⁻¹ emerges post-oxidation, and elemental analysis showing progressive oxygen incorporation with rising treatment temperature. The equipment used for both this pretreatment step and the main ATT reaction is a tube furnace, configured as a three-zone system housing the quartz reaction tube.

DFT calculations confirm the resulting functional groups are reactive: esters, ketones, carboxylic acids, and aldehydes all form stable adducts with hydroxide, with reaction energies between −1.18 and −1.41 eV. Aldehydes show the strongest thermodynamic driving force. Notably, alcohol groups — which can also form during oxidation — are unreactive toward hydroxide addition, since the C–OH bond lacks an electrophilic carbon center. Not every oxidation product is a productive one.

Here's the variable that matters most for anyone evaluating this process: oxidation temperature and mass loss move together. Raw hydrogen yield for PE climbs with oxidation temperature, peaking at 108.9 mmol H2 per gram of oxidized PE at 350°C. But PE also loses roughly 74% of its mass during oxidation at that temperature, and once yield is recalculated per gram of the original plastic rather than the oxidized residue, the effective figure drops to 27.5 mmol/g. The actual mass-corrected optimum sits at 250°C, yielding 44.8 mmol/g. PP shows the same inversion — a raw peak of 60.9 mmol/g at 300°C, but a corrected optimum of 20.7 mmol/g at 200°C, since PP's tertiary carbons make it less thermally stable and more prone to mass loss at higher oxidation temperatures.

The oxidation step that activates PE and PP is also consuming them. The highest raw hydrogen yield for PE looks like it comes from the most aggressive pretreatment — until the math accounts for the 74% of the plastic that pretreatment burns away.

Anyone benchmarking this process against a raw yield table without checking which basis it's reported on is comparing numbers that don't mean what they appear to mean.

Reactor Configuration And Real-Time Monitoring

The reaction itself runs in a quartz tube (2.54 cm outer diameter, 100 cm length) positioned inside the three-zone tubular furnace, with a 3 mL alumina boat holding the plastic and NaOH charge. A thermocouple inside the tube tracks temperature in real time. Water is introduced under controlled flow via a syringe pump fitted with an 18-gauge needle, metered at 23 μL/min through a heated line where it vaporizes before entering the reactor on a nitrogen carrier stream. This is a comparatively modest bench-scale configuration — worth noting given how far it sits from any continuous or pilot-scale process the paper actually demonstrates.

Where The Carbon Actually Ends Up

The carbon accounting is more granular than "no CO2 released." Post-reaction analysis — using elemental content analysis methods including FT-IR for functional group confirmation and total organic carbon (TOC) measurement for solid-residue speciation — shows carbon partitioning into three pools: 32–44% as inorganic carbon (sodium carbonate), 44–62% as liquid-phase carbon (tar and wax residues), and less than 13% as gaseous or residual organic carbon.

More carbon from the plastic ends up as tar and wax than as gas or stable carbonate combined. Calling this a "no-CO2" process is directionally true and mechanistically incomplete.

The sodium carbonate fraction can be converted downstream to calcium carbonate, which is less soluble and better suited to permanent carbon immobilization — and which also has established industrial demand, unlike carbon that's simply been kept out of a gas stream. But the liquid-phase majority is a residue stream, not a captured-carbon stream, and any accounting of this process's climate benefit needs to treat those two fractions differently.

Emissions Comparison Against The Incumbent Route

The paper's life cycle assessment gives a more useful comparison than "ATT beats gasification," which is a low bar. Steam gasification of PET runs 20.67 kg-CO2eq per kg of hydrogen produced, driven mostly by direct emissions. ATT's figures land considerably lower: 10.90 for PET, 8.33 for PP, and 4.83 for PE. The more relevant benchmark is steam-methane reforming — the current commercial standard for hydrogen production — which runs 8–10 kg-CO2eq/kg-H2 without carbon capture. ATT-PE beats that benchmark outright; ATT-PET and ATT-PP land within the same range. That's a meaningfully stronger claim than the emissions framing most coverage of this study settled for, and it's the comparison worth citing if this process is being evaluated against anything other than gasification.

What's Still Unresolved

Reaction selectivity remains moderate — calculated at 32.7% to 49% depending on plastic type — because a substantial share of the sodium carbonate detected after the reaction actually originates from captured CO2 generated during thermal decomposition, not directly from the plastic's own carbon. The mass balance is more complicated than a simple carbon-in, carbon-out accounting suggests. The study also stops at bench scale: no continuous-process engineering data, no cost model, and no scale-up demonstration. The authors are explicit that further work is needed to determine whether the process is economically viable before it moves beyond the lab.

Key Takeaways

ATT's mixed-stream capability comes from an oxidation pretreatment that activates PE and PP by introducing hydroxide-reactive functional groups — but that pretreatment consumes a substantial fraction of the feedstock's own mass, and the optimal operating temperature depends on which basis (oxidized mass vs. original mass) the yield is calculated against.
Carbon fate after the reaction splits three ways — inorganic carbonate, liquid residue, and gas — and the liquid-phase fraction is the largest of the three, not a rounding error against an otherwise clean carbon capture story.
ATT's emissions profile is competitive with steam-methane reforming, the actual commercial hydrogen benchmark, which is a more meaningful comparison than beating gasification alone.
The process remains bench-scale, with open questions on selectivity, continuous operation, and economic viability before any scale-up claim is warranted.

Work like this — mechanism-driven, still short on scale-up data — is exactly the kind of study worth watching rather than acting on prematurely. MSE Supplies supports researchers evaluating early-stage thermal and materials processes with PhD-level technical guidance, and offers customization for labs building non-standard reactor or furnace configurations. Questions on equipment fit can go through contact us, and updates on emerging materials science research are posted regularly on LinkedIn.

Share this post