Karl Fischer Moisture Testing of Recycled Polyethylene Terephthalate

Hydrolysis

Polyethylene terephthalate chains cleave permanently when exposed to trace water at processing temperatures above 260 °C. Free water dissolved in the melt attacks the ester linkages along the polymer backbone. Each consumed water molecule breaks a high molecular weight chain into shorter fragments, forming one terminal carboxylic acid group and one terminal hydroxyl group. That drop in chain length lowers the resin’s intrinsic viscosity, which directly weakens the impact strength, tensile failure strain, and top-load performance of molded parts.

Controlling this degradation in recycled PET streams means holding moisture levels far below what non-hydrolytic polymers tolerate. Amorphous post-consumer flake and re-granulated bottle resin hold both surface moisture and absorbed bulk water. Once the resin enters an extrusion barrel or molding cylinder, heat and shear turn that liquid water into reactive steam.

Chain scission happens fast: at 280 °C, just 100 parts per million of residual water drops the molecular weight measurably within thirty seconds of screw residence time.

Ester Bond Scission under Thermal Stress

In the melt, water acts as an aggressive nucleophile attacking backbone bonds. Scission begins with carbonyl protonation, followed by water addition to form a tetrahedral intermediate that collapses into separate ester and acid end groups. Every split reduces the number-average molecular weight of the chain.

And since intrinsic viscosity links directly to molecular weight through the Mark-Houwink relation, chain cleavage shows up right away as lost melt strength during processing.

As carboxylic acid end groups accumulate, they auto-catalyze further scission. A rising concentration of carboxyl groups lowers the activation energy for subsequent hydrolysis. Because of this loop, early degradation speeds up further breakdown as long as moisture stays in the melt zone.

Controlling incoming resin moisture before processing is the only practical way to stop the cascade.

Moisture Thresholds for Processing Recycled Resin

Bottle preform production requires moisture to stay at or below 0.005 percent by weight to prevent structural collapse. Demanding applications like high-tenacity industrial strapping or monofilament extrusion push drying targets below 0.002 percent ~ 20 parts per million. With so few water molecules per mole of polymer chain at these concentrations, the intrinsic viscosity drop stays under 0.02 deciliters per gram during standard residence times.

Sheet thermoforming tolerates somewhat more residual moisture because of lower structural demands, but running above 0.010 percent by weight causes obvious physical and visual defects. Vaporizing water forms micro-voids in the sheet extruder die land, triggering web tears, gauge variation, and severe melt-curtain instability. The limit cannot be breached without penalty.

Ester scission occurs instantly at 280 degrees Celsius when free water molecules inhabit the amorphous polymer matrix.

Why Does Loss on Drying Fail for Recycled PET?

Gravimetric balance methods simply lack the sensitivity needed for technical polyester processing. Standard loss-on-drying instruments track weight loss while heating sample pellets on an internal balance scale. With standard sample sizes, their resolution floors out around 0.010 percent (100 parts per million).

Since recycled polyester demands moisture verification under 50 parts per million, a loss-on-drying balance is working entirely inside its own margin of error at key production limits.

Post-consumer flake streams also carry volatile contaminants that completely distort gravimetric readings. Wash detergent residues, limonite, acetaldehyde, ethylene glycol, and trace organic volatiles evaporate right along with water when heated on a halogen balance pan. The scale registers all mass loss as water, yielding false positives about resin moisture.

A sample with 30 parts per million of actual water and 150 parts per million of residual solvent looks soaking wet on a loss-on-drying balance, driving operators to overheat and over-dry resin batches for no reason.

Thermal breakdown of low molecular weight recycled fractions skews loss-on-drying balances even further. Amorphous flake fines degrade at lower temperatures than virgin pellets, driving off volatile condensation products that show up as evaporated moisture. In practice, Karl Fischer coulometry replaces thermal balances whenever moisture targets fall below 100 ppm.

Extrusion of polyester flake with 200 ppm moisture at 280 °C drops intrinsic viscosity by 0.12 dL/g, resulting in brittle container walls and instant failure during top-load testing.

Vapor

Thermal desorption ovens drive trapped moisture out of the polymer matrix without melting the solid pellet. PET pellets cannot be injected directly into a Karl Fischer titration cell because the polymer will not dissolve in standard titration solvents at room temperature. Coupling a high-temperature oven to an automated coulometric titrator solves this: the polymer sample heats inside a sealed glass vial, desorbed moisture evaporates, and a dry carrier gas sweeps the vapor straight into the titration vessel.

Running an oven vaporization setup demands tight control over sample temperature, gas flow, and line heating. Water diffuses out of a flake or pellet along a temperature-dependent curve. As heat penetrates the matrix, water molecules gain enough mobility to break hydrogen bonds with ester oxygens and migrate to the surface.

The carrier gas then sweeps these desorbed water molecules down a heated transfer line directly into the reaction cell, where iodine titration measures the total mass.

Thermal Evaporator Chamber Dynamics

Heating vials inside a sealed aluminum block delivers uniform thermal energy throughout the sample. Sample sizes for recycled polyester usually run between 1.0 and 3.0 grams, depending on expected moisture. The oven keeps the vial at a steady isothermal setpoint, avoiding cold spots that would let moisture re-condense on vial walls before reaching the transfer line.

Transfer lines linking the oven to the coulometric cell must be actively heated to between 110 °C and 130 °C. Unheated lines leave cold zones where vaporized water condenses inside the tubing, leading to tailing titration curves, delayed endpoints, and falsely low moisture numbers. Flexible PTFE or PEEK tubing prevents water from adsorbing along the gas path.

Carrier Gas Selection and Flow Calibration

Ultra-high purity nitrogen carries desorbed moisture into the cell. The nitrogen must meet Grade 5.0 specifications (99.999 percent pure) with residual moisture under 2 parts per million. Running lower-grade compressed air or unpurified nitrogen leaks ambient moisture into the carrier stream, driving up background drift and ruining precision at sub-50 ppm moisture levels.

Setting carrier flow between 80 and 120 milliliters per minute with an inline mass flow controller ensures complete reaction in the cell. Too high a flow rate cuts gas residence time in the cell, letting unreacted water vapor slip through the cell exhaust past the iodine reagent. Flow rates under 50 milliliters per minute drag out transport, lengthening analysis times and encouraging moisture to re-condense inside the vial cap.

Thermal Ramp Optimization and Pyrolysis Suppression

Setting the oven to 170 °C separates free water release from organic breakdown. Heating recycled PET above 190 °C starts thermal breakdown of low molecular weight fractions, oligomers, and residual contaminants. That degradation generates secondary reaction water as terminal hydroxyl groups condense, creating falsely high moisture readings that misrepresent the actual free water in the resin.

Thermal breakdown releases troublesome interfering species. Acetaldehyde, diethylene glycol, and volatile ester oligomers evaporate rapidly above 200 °C. Swept into the cell, these organic vapors consume iodine titrant or coat the indicator electrode, destabilizing the baseline and spiking drift rates. Holding the desorption oven at 170 °C yields complete moisture extraction within six minutes while keeping organic matrix pyrolysis at bay.

Karl Fischer coulometry yields an absolute accuracy of 2 micrograms of water when carrier nitrogen flows at 100 milliliters per minute.

Using Grade 5.0 nitrogen fitted with upstream moisture traps keeps background drift down. Daily testing should always begin with leak checks to stop ambient humidity from creeping into manifold fittings. High moisture readings can also stem from volatile organic additives off-gassing in the oven rather than real water trapped in the polymer.

Reagent

Coulometric cells generate iodine quantitatively at an anode inside the electrolyte solution. Unlike volumetric titration, which dispenses a standardized iodine solution from a mechanical burette, coulometric Karl Fischer produces iodine right at a platinum anode in the vessel. Faraday’s law links current consumption directly to the mass of iodine produced, allowing absolute measurement of trace water without manual titrant standardization.

The electrochemistry happens in a generator cell filled with anolyte and catholyte solutions, using either a ceramic diaphragm separator or a diaphragm-free setup. The anolyte supplies iodide ions, sulfur dioxide, an amine buffer, and an alcohol solvent. As water vapor enters, it reacts instantly with iodine, sulfur dioxide, and the buffer, drawing down iodine until the free water is spent and the current reaches its endpoint.

Coulometric Chemistry and Oxidation Stoichiometric Ratios

Under Bunsen stoichiometry, one mole of water consumes exactly one mole of elemental iodine. The full mechanism depends on an alcohol solvent ~ usually methanol or diethylene glycol monoethyl ether ~ to react with sulfur dioxide and form an intermediate alkyl sulfite ester. The amine buffer then neutralizes the resulting alkyl sulfate and hydrogen iodide acids, holding the solution pH in the 5.5 to 7.0 window.

Coulometry measures charge directly: one milligram of water consumes exactly 10.72 Coulombs. The instrument monitors current across the generator electrode, integrating it over time to determine total charge. Because digital titrators integrate current down to sub-microampere levels, the technique detects water amounts down to 1 microgram with high repeatability.

Volumetric Capacity Vs Coulometric Precision

Direct liquid dosing works well for water concentrations above 0.1 percent, but current-based measurement is superior at trace levels. Volumetric Karl Fischer units feed liquid iodine reagent through a mechanical burette with minimum increments around 0.2 microliters. Below 100 parts per million, burette resolution limits introduce noticeable error and scatter.

Coulometric instruments bypass burette hardware altogether, replacing volumetric dosing with digital current integration.

Quality control for recycled polyester relies almost entirely on coulometry because target moisture levels sit below 50 parts per million. A 2.0-gram sample containing 30 ppm moisture yields just 60 micrograms of water. Measuring 60 micrograms volumetrically takes under 0.01 milliliters of reagent ~ well below what a mechanical burette can deliver accurately.

Coulometry measures that 60 micrograms easily by consuming 0.643 Coulombs over a three-minute titration.

Suppression of Side Reactions from Contaminants

Post-consumer flake carries residual glycols and aldehydes that interfere with standard methanol-based buffers. Acetaldehyde reacts with methanol in traditional reagents to produce acetals and secondary water. That side reaction continually feeds water into the titration vessel, keeping the instrument from reaching a stable endpoint and skewing moisture readings upward.

Stopping aldehyde interference takes alcohol-free or keto-formulated anolyte reagents. Swapping methanol for diethylene glycol monoethyl ether suppresses acetal formation and halts secondary water production. Reagent selection must be tailored to the specific contaminants in the recycled stream.

• Methanol-Free Anolyte Formulations replace standard primary alcohols with glycol ethers to prevent acetal formation when testing flake contaminated with residual acetaldehyde.
• Diaphragm-Free Generator Cells reduce vessel maintenance and prevent cathodic side reactions that produce excess hydrogen gas during trace water testing.
• Imidazole-Buffered Electrolytes hold reaction pH within the 5.5 to 6.5 window to prevent acid-catalyzed ester degradation in the titration vessel.
• Sulfur Dioxide Excess Capacity prevents reagent exhaustion during long sample runs containing acidic wash residues.
• Halogenated Solvents Additives boost organic volatile solubility in the anolyte, keeping oligomers from precipitating onto the dual platinum indicator pins.

ISO 15512 Method B2 mandates daily drift determination below 10 micrograms of water per minute before sample introduction.

How much post-consumer wash surfactant residues shift iodine stoichiometry in coulometric cells remains an open question in recycled polymer testing.

Sampling

Regrind particles absorb ambient moisture fast enough to distort lab test results within minutes. Post-consumer rPET arrives at processing plants in various forms: thin amorphous flake, dense hot-washed regrind, and re-pelletized semi-crystalline granules. Each form has a different surface-to-volume ratio, dictating how quickly it takes up or loses moisture to the air.

Amorphous flake takes up moisture remarkably fast when exposed to ambient air. High free volume near the surface of uncrystallized polyester chains lets water enter the matrix in seconds. Sampling in an unconditioned warehouse or scooping loose flake into open plastic buckets causes rapid moisture pickup, making subsequent Karl Fischer results useless for judging the material inside a bulk silo.

Particle Morphology and Atmospheric Water Absorption

Uncrystallized flake offers far more surface area per unit volume than extruded pellets. With flake thickness running between 0.3 mm and 0.8 mm, moisture has a short path to diffuse into the matrix. At 60 percent relative humidity and 22 °C, dry amorphous rPET flake picks up over 50 parts per million of water within five minutes.

Extruded semi-crystalline pellets take up water much slower thanks to lower surface area and dense crystalline regions. Water cannot easily enter PET’s ordered crystal lattice, limiting quick absorption to the amorphous outer layer. Because flake geometry speeds up moisture absorption, testing protocols must strictly enforce fast sample handling.

Sample Packaging and Vial Headspace Management

Laminated aluminum foil pouches protect samples from moisture during transfer to the lab. Standard polyethylene bags allow water vapor transmission right through their walls, making them useless for trace moisture storage. Glass vials with PTFE-lined silicone septa create a tight seal, preserving sample integrity from the sampling point to the titrator.

Headspace air introduces ambient humidity directly into the sample vial. At 50 percent relative humidity and 20 °C, a single milliliter of air holds about 8.6 micrograms of water vapor. In a 20-milliliter vial holding just 1.0 gram of dry resin, the water in the headspace air actually outweighs the water in the polymer, triggering severe false positives.

Vials should be filled completely or purged with dry argon before sealing.

Representative Lot Sampling from Bulk Silos

Sampling thieves extended to three depths catch vertical moisture gradients inside 20-tonne storage vessels. Bulk railcars and silos develop clear internal moisture gradients from thermal cycling and air movement. Material along silo walls cools quickly at night, causing air inside to condense moisture onto outer resin layers while the center stays warm and dry.

Taking a single surface scoop from a gaylord or silo hatch will not catch lot variance. Inspectors need to pull sub-samples from top, middle, and bottom layers using sealed lances. Combining equal parts of those sub-samples in a nitrogen-purged vessel yields a composite sample that reflects the actual moisture of the shipment.

1. Extract polymer sub-samples from bulk storage using a stainless steel thief sealed against air.
2. Transfer sub-samples directly into dry glass vials pre-purged with dry nitrogen gas to stop moisture pickup.
3. Cap vials immediately with aluminum crimp seals fitted with PTFE-lined silicone septa.
4. Weigh the vial on an analytical balance before and after filling to determine net sample mass.
5. Place the sealed vial directly into the thermal desorption block without removing the septum.
6. Puncture the septum with the dual-needle carrier gas manifold to start water vapor transfer into the cell.

Sampling silos requires pulling material from three vertical levels with a closed thief to map moisture variations. Clean tools prevent grease transfer to pellet surfaces. Sealed sample containers left in direct sunlight pull in moisture through temperature-driven pressure differentials faster than unsealed containers sitting in cool shade.

Calibration

Maintaining analytical accuracy requires regular validation with certified water standards. Coulometric Karl Fischer units measure absolutely according to Faraday’s law, but leaks, septum wear, carrier flow drift, and electrode fouling undermine accuracy over time. Routine calibration checks with certified standards verify system response and confirm that desorbed moisture actually reaches the cell.

Validation relies on certified liquid water standards or pure hydrated chemical salts with known thermal decomposition profiles. Injecting a known mass of standard tests the system’s recovery percentage. Dividing measured water mass by expected standard mass gives the recovery ratio, which must hold between 95 percent and 105 percent for data to stand up in contract disputes.

Water Standard Selection and Recovery Verification

Sodium tartrate dihydrate offers a stable crystalline structure containing 15.66 percent water by weight. It releases its water of crystallization cleanly between 150 °C and 180 °C without organic breakdown. Adding 20 to 30 milligrams of sodium tartrate dihydrate into a vial introduces roughly 3.1 to 4.7 milligrams of water, providing a solid benchmark for thermal evaporator delivery.

Liquid water standards in organic solvents provide another option for low-range verification. Supplied in sealed glass ampoules, these standards carry exact water concentrations like 100 ppm or 1000 ppm. Injecting liquid standards directly into the cell checks electrolyte activity, while heating them in capped vials checks transfer line recovery.

Quantifying System Drift and Carrier Blank Values

Background moisture entering the cell establishes the baseline current before testing starts. No titration cell is completely sealed ~ ambient humidity leaks in through glass joints, fittings, and septum pinholes. The titrator measures this background influx as a drift rate, expressed in micrograms of water per minute.

Drift rates determine baseline subtraction. A healthy coulometric system operates with background drift under 5 micrograms per minute. Drift above 15 micrograms per minute signals saturated desiccant traps, loose fittings, or degraded reagents.

The software logs baseline drift right before sample heating starts, subtracting cumulative drift from the total water signal during analysis.

Vessel drift stability governs test accuracy far more than reagent freshness.

Vessel Maintenance and Electrode Cleaning

Platinum generator mesh collects organic film over time, dulling sensor response. Testing recycled PET exposes the cell to volatile oligomers, wash surfactant residues, and solvent vapors. These compounds leave a hydrophobic film on the indicator and generator electrodes, slowing electron transfer and producing erratic endpoints.

Cleaning electrodes requires soaking them periodically in concentrated nitric acid, then rinsing thoroughly with deionized water and electronic-grade methanol. Never scrub electrodes with abrasives ~ doing so ruins the delicate platinum mesh. Clean electrodes give stable baselines.

After acid washing, dry the electrodes in a vacuum oven at 80 °C before putting them back in the cell.

• Degraded Septa Seals allow ambient air to leak into vials during oven heating, causing baseline drift spikes.
• Exhausted Molecular Sieve Traps fail to strip moisture from carrier gas, driving baseline drift past acceptable limits.
• Saturated Anolyte Electrolyte reduces iodine generation efficiency, lengthening titration times and corrupting recovery calculations.
• Hydrophobic Oligomer Film Deposits coat platinum indicator pins, dulling voltage response and delaying titration endpoints.
• Condensation Cold Spots along transfer tubing trap desorbed sample water, causing tailing curves and falsely low moisture totals.

ISO 15512 Method B2 Clause 8.3 requires subtracting background drift from total titrant consumption, which can shift reported moisture figures by up to 15 parts per million on dry samples.

Yield

Loss of intrinsic viscosity leads directly to weakened physical properties and scrapped production. The economics of a recycled PET molding or extrusion line depend on preserving polymer chain length from raw flake to finished part. Every 0.01 dL/g drop in intrinsic viscosity reduces preform impact strength and top-load capacity, raising field failure rates and scrap costs.

Balancing testing accuracy against process profitability comes down to hydrolytic scission arithmetic. Processing wet or poorly tested flake turns resin into low-value scrap. Desiccant drying systems draw significant power, making accurate moisture measurements essential: under-drying produces scrap, while over-drying wastes energy and residence time.

Intrinsic Viscosity Scission Calculations

Chain degradation scales directly with the amount of water present in the extruder throat. The model governing polyester hydrolysis links moisture content to final number-average molecular weight through this relationship:

1 / Mn_final = 1 / Mn_initial + ( w / 18.015 ) f_scission

where Mn is the number-average molecular weight in grams per mole, w represents the resin moisture mass fraction, and f_scission is the fraction of dissolved water consumed during melt residence. In a single-screw preform extruder running at 280 °C with a two-minute residence time, f_scission approaches 0.95 ~ meaning nearly all free water reacts.

Converting number-average molecular weight to intrinsic viscosity relies on the empirical relation:

IV = 0.00047 ( Mn ^ 0.68 )

Take a bottle-grade rPET flake batch with an initial IV of 0.80 dL/g and an Mn_initial of roughly 27,800 g/mol. If processed with 30 ppm water, Mn_final lands at 26,570 g/mol and part IV reaches 0.778 dL/g, easily meeting bottle specs. If that same resin enters the extruder with 250 ppm water because of bad moisture data, Mn_final plummets to 19,200 g/mol.

Part IV drops to 0.632 dL/g, leaving preforms prone to stress cracking and unusable for blow molding.

Drying Energy Economics Vs Scrap Penalties

Running desiccant hoppers at a -40 °C dew point takes serious power. Dry-air hoppers consume 0.06 to 0.10 kilowatt-hours per kilogram of rPET to bring moisture from 0.30 percent down to 0.003 percent. For a preform plant running 2,000 kilograms per hour, annual drying power alone can easily exceed $120,000.

Drying costs offset scrap penalties. Running dryers at max heat all the time guarantees dry resin, but it wastes energy and risks thermal yellowing. Accurate Karl Fischer testing allows closed-loop dryer control, trimming heat once target dryness is reached.

That feedback loop cuts drying energy use by up to 18 percent without risking part performance.

Contractual Specifications for Recycled Pellet Procurement

Supply contracts need explicit test conditions alongside moisture limits to head off disputes. When buying post-consumer rPET flake or pelletized resin, purchase contracts should specify Karl Fischer coulometry under ISO 15512 Method B2 as the sole binding test method. Accepting general datasheet claims without specifying oven temperature, sample size, and test standards leaves buyers unprotected when a wet shipment stops production.

Clear contract limits prevent quality failures downstream. Agreements should include sliding-scale price deductions or rejection rights based on dock moisture testing. A typical commercial clause grants full acceptance under 50 ppm, a $30 per tonne penalty between 51 and 100 ppm, and outright rejection above 100 ppm at the receiving dock.

• Standard Test Method Assignment mandates ISO 15512 Method B2 thermal desorption coulometry as the sole reference for moisture arbitration.
• Oven Temperature Setpoint Binding locks desorption testing strictly at 170 °C to prevent organic pyrolysis interference.
• Maximum Moisture Threshold Limits set upper limits at 30 ppm for technical grades and 50 ppm for bottle preform resins.
• Receiving Dock Sampling Mandates require sealed septum vial sampling within 20 minutes of opening a bulk container.
• Financial Deduction Schedules establish clear per-tonne penalties for off-spec shipments before resin enters processing.

Financial models weigh the cost of desiccant dryer residence time against the price of lost intrinsic viscosity. Setting 30 ppm incoming moisture limits on purchase orders protects finished product IV while keeping annual drying energy costs predictable.