05 — When it breaks

When the material itself changes.

+ The questionHow do water, heat, oxygen and ozone break a polymer, and what comes out?

Two rubbers in hot water

Two polyurethane rubbers can look alike and still last very different times in hot water. In 1967 Hollingsworth and two colleagues aged commercial polyurethane rubbers in water at 70 °C and measured how much strength each kept. They worked at the Explosives Research and Development Establishment, a British government laboratory. Two polyester polyurethanes were "too weak to test" after two and three weeks, while a polyether polyurethane still held about 12 % of its starting strength after a year. The authors blamed "the poor hydrolytic stability of the polyester urethanes". No additive had leached out, as additives do in Chapter 4: the polymer itself had failed, and the authors' word "hydrolytic" names what attacked it, water.

Chapter 4 followed additives out between the chains of Chapter 3 and left the chains themselves in place. Each chain, though, is a row of links, and each kind of link has its own weakness. Water cuts some links, a reaction called hydrolysis. Oxygen and ozone attack others, and heat speeds all of these reactions. In a 1994 bench study, Pegoretti and colleagues kept commercial polyester polyurethanes in water at 70 °C for up to eight weeks, and they state the consequence: "Each cleavage of the backbones leads to the generation of a new molecule". A polymer that breaks down therefore releases substances its recipe never held. A substance that forms only in use is not there to be extracted beforehand, which is one reason Chapter 4 called leachables only "generally" a subset of extractables.

Water cuts the ester link first

A polyurethane chain alternates two kinds of segment, and urethane links tie the segments into the chain. Its soft segments come from the polyol of Chapter 3, and the polyol gives the material its name. A polyester polyurethane has soft segments built with ester links, and a polyether polyurethane has soft segments built with ether links. Water can attack both the ester and the urethane links, but not at the same rate: "the most susceptible to hydrolysis is the ester", Pegoretti and colleagues write, and in their account the urethane link hydrolyzes about ten times more slowly.

Water cuts an ester link at its carbon atom, the one double-bonded to oxygen. The cut leaves two chain ends: one carries a hydroxyl group, as an alcohol does, and the other a carboxylic acid. Brannigan and Dove explain why that carbon is the target in a 2017 review of degradable polyesters. How readily a group hydrolyzes depends on the positive charge of its reacting carbon, and groups whose structure spreads that charge over neighboring atoms hydrolyze more slowly, urethane and carbonate links among them. The acid end then speeds the next cut, because acid catalyzes ester hydrolysis, and in the words of Pegoretti and colleagues "the autocatalytic reaction becomes prevalent". The more links water has cut, the faster it cuts the rest.

Polyether polyurethanes resist hydrolysis. "Generally, the hydrolytic degradation of polyether-urethanes in pure water is minimal", Miri and colleagues write in a 2023 review of polyurethanes in biomedical engineering. The review gives the typical order of hydrolysis for the polyols as polycarbonate slowest, then polyether, then polyester. It states the difference and the 1967 tests measured it, but neither source explains the ether link's resistance at the level of the bond. Carbonate links are slow to hydrolyze but not immune. Breakdown or hydrolysis of polycarbonate after manufacture can occur, SCENIHR's 2015 opinion on BPA in medical devices states, although in a 1996 bench study Hong and colleagues found no hydrolysis in steam cycles of up to 120 minutes (Chapter 7). The study's authors were all staff of Baxter Healthcare.

14 — Where water cuts a polyurethane
POLYESTER POLYURETHANE ESTER LINK ESTER LINK ESTER LINK URETHANE LINK about ten times slower (Pegoretti 1994) HYDROXYL END ACID END the acid speeds the next cut water attacks the carbon double-bonded to oxygen POLYETHER POLYURETHANE ETHER LINK ETHER LINK ETHER LINK O· vulnerable to oxidation instead (Miri 2023) resists hydrolysis TYPICAL ORDER OF HYDROLYSIS, BY POLYOL polyester: fastest polyether polycarbonate: slowest Schematic, simplified. Ester and acid ends: Taourit and colleagues 2022; charge on the reacting carbon: Brannigan and Dove 2017 (review); rates: Pegoretti and colleagues 1994; order: Miri and colleagues 2023 (review).
Device materialReleased substanceGas or fluid pathFocal detail
Plate 14 — Water cuts a polyester polyurethane at its ester links, leaving a hydroxyl end and an acid end, and the acid speeds the next cut; the urethane link hydrolyzes about ten times more slowly. A polyether polyurethane resists water and is vulnerable to oxidation instead.

Oxygen and ozone attack what water spares

The polyether's resistance to water comes with a weakness to oxygen. "PUs made from polyether polyols are vulnerable to oxidation", the review by Miri and colleagues states, while those made from polyester polyols are vulnerable to hydrolysis. Oxidation runs through free radicals, reactive fragments with an unpaired electron, which high temperature generates. The radicals can break chains, which is chain scission, and the short fragments can join neighboring chains together, which is crosslinking. Polyether polyurethanes under strain also crack in an oxidizing environment. The review calls this failure environmental stress cracking and finds it more common in them.

Schubert and colleagues found both chain scission and crosslinking in a 1995 bench test aimed at implants. They exposed strained films of a polyether polyurethane urea to an oxidizing solution of hydrogen peroxide and cobalt chloride. The polyether soft segments and the urethane links were cut, the cutting led to "severe pitting and cracking", and a second route crosslinked the soft segments. Because the test used strained films in an oxidizing solution, not a foam, it does not show which of the two reactions dominates in a foam.

Ozone attacks a narrower target, the carbon–carbon double bond. Natural rubber and polybutadiene carry such bonds in their chains, and ozone adds across each bond to form an unstable ring, the ozonide, which splits. Where the halves recombine the chain survives, and where they separate it is cut, so that "one macromolecule can be degraded to many fragments". Zaikov and colleagues give this route in a 2012 review that includes their own measurements in solution, and they find that polymers without double bonds react with ozone only by a radical route about four to five tenfold steps slower. Ozone also cracks stretched rubber, and in a 1976 report from a US Army laboratory, Bergstrom found that antiozonants and certain waxes delay the cracking.

The running device's cushion is silicone, and silicone polymers generally carry no carbon–carbon double bonds, so ozone cannot attack them by this route, a chemistry-teaching page of the University of Wuppertal states. The page was made in cooperation with WACKER, a silicone maker, and in its experiment silicone rubber showed no change after 70 hours in ozone at 40 °C. Zaikov and colleagues did not measure silicone, but by their finding for chains without double bonds, ozone can react with it only by the slower radical route. In a 2024 notice about products sold to clean CPAP machines, FDA nonetheless warned that "both UV light and ozone can break down CPAP accessories like hoses and masks". The notice names no material and gives no chemistry. Chapter 7 returns to cleaning.

Heat speeds every cut

Heat speeds hydrolysis and oxidation alike. The 1967 polyether rubber reached the same degradation after 2 weeks at 90 °C, about 12 weeks at 70 °C and more than a year at 40 °C. Humid air at 70 °C degraded the rubbers slightly more than water did, and the authors put this down to the air's extra oxygen, so water and oxygen also act together. Water must be inside a polymer to cut it, and in a 2022 bench study of one polyester polyurethane between 40 and 90 °C, Taourit and colleagues scaled the rate of hydrolysis by the water the polymer dissolves, about 2.5 % of its weight at 80 °C.

To read a hot test as a longer time at a cooler temperature, device makers use a convention: each 10 °C rise doubles the rate of aging. The factor for each 10 °C is called Q10. Hemmerich gave its usual value in a 1998 trade-journal protocol for the accelerated aging of medical devices: "The typical relationship selected for commonly used medical polymers is Q10 = 2." He managed a gamma-irradiation facility for Isomedix, a sterilization company. He called the value conservative for a wide range of medical polymers, for use unless a rate had been measured for the material. He also capped the oven at 60 °C, below the material's transitions, warned that heat can produce "unrealistic negative outcomes", and noted that humidity extremes may matter for certain polymers.

The rule is older than the protocol, and it is not a law. In a 2019 analysis, Janting and colleagues trace it to van 't Hoff, who in 1896 allowed that 10 °C "doubles or trebles" a reaction's speed. Two of the three authors worked for companies, ConvaTec and the consultancy MedDevCon. The Arrhenius equation is the usual description of how reaction rates rise with temperature, and the authors show that the doubling cannot be derived from it. They advise testing between about 25 and 60 °C, because a change of mechanism may bend the relation from about 60 to 80 °C. They also advise an interval that holds no transition, such as the glass transition, the temperature below which a polymer turns stiff and glassy. The rule, they write, entered medical-device standards, ASTM F1980 among them, "without a more critical analysis on the accuracy".

F1980 is ASTM's guide to the accelerated aging of sterile barrier systems. ASTM describes its 2016 edition as asking for "a conservative estimate of the aging factor (for example, Q10)", and the description gives no value. FDA recognized the 2021 edition in February 2023, and FDA's record of that edition gives no value either. Chapter 7 follows the guide and the shelf life it sets. In a 2015 bench study of a medical-grade silicone aged in saline at 70 °C, Mahomed and colleagues applied a factor of 2 without discussing the factor's limits, although 70 °C lies above Hemmerich's cap.

Solved example: the doubling convention against one rubber's measured times
  • The data. The 1967 polyether rubber reached the same degradation after 2 weeks at 90 °C, about 12 weeks at 70 °C and more than a year at 40 °C.
  • From 90 to 70 °C. A fall of 20 °C is two steps of 10 °C, and two doublings multiply the time by 4, so the convention predicts 8 weeks at 70 °C. The rubber took about 12 weeks, six times its time at 90 °C, which is a factor of about 2.4 for each 10 °C, since 2.4 × 2.4 is about 6.
  • From 70 to 40 °C. A fall of 30 °C is three steps, a factor of 8, so the convention predicts about 96 weeks from the 12 measured at 70 °C, nearly two years. The report says only "more than a year", which agrees with the prediction but cannot test it.
  • The result. Between 90 and 70 °C this rubber slowed more on cooling than doubling assumes, so the convention would have understated how long it lasts at the cooler temperature, and understating is the safe direction for a claim of how long a part lasts.
  • Assumptions. The same reaction runs at every temperature. Both 90 and 70 °C lie above the protocol's 60 °C cap, the rubbers were 1967 grades in the report's aging conditions, and "about 12 weeks" is the report's own rounding.
15 — The doubling convention against one rubber
above the protocol's 60 °C cap (Hemmerich 1998) 2 4 8 16 32 64 128 30 40 50 60 70 80 90 WEEKS TO THE SAME DEGRADATION AGING TEMPERATURE (°C) doubling per 10 °C (Q10 = 2) 90 °C: 2 WEEKS 70 °C: ABOUT 12 WEEKS 40 °C: MORE THAN A YEAR about 12 weeks where doubling predicts 8 interval Janting and colleagues advise: about 25 to 60 °C, no transition inside Humidity: no factor at source. Humid air at 70 °C degraded the 1967 rubbers slightly more than water did. Q10 = 2: Hemmerich 1998 (trade-journal protocol; author at a sterilization company). Interval: Janting and colleagues 2019. Rubber: Hollingsworth and colleagues 1967; the medium of these three figures is the report's. ASTM describes F1980-16 as asking for a conservative estimate of the aging factor and gives no value.
Device materialLimit or thresholdFocal detail
Plate 15 — The doubling convention predicts 8 weeks at 70 °C for a rubber that took 2 weeks at 90 °C; this 1967 polyether rubber took about 12. Both temperatures lie above the 60 °C cap the convention's protocol sets, and humidity has no factor of its own at source.

What a breaking polyurethane releases

Each cut shortens a chain, and enough cuts change the whole material. In the study by Pegoretti and colleagues, the average molar mass of the polyester polyurethanes fell to 15 % of its starting value after eight weeks in water at 70 °C. The shorter chains also lowered the glass transition. Each cut ester leaves a hydroxyl end and an acid end, and where cuts fall close together they free whole pieces: short chains at first, and in the end the building blocks of the polyol, small enough to leave by the diffusion of Chapter 4.

Philips Respironics, the manufacturer, lists such products for the polyester polyurethane foam of the breathing machines it recalled in 2021 (Chapter 20). In its own 2023 account, the maker names the mechanism "hydrolysis, primarily of the ester groups within the foam". The foam's polyol is "based on diethylene glycol (DEG) and adipic acid (AA)", which hydrolysis can return as a diol and an acid, while the urethane link yields "a toluene diamine containing oligomer" and then toluene diamine, an amine. In the maker's words, the foam "shows measurable degradation with exposure to high temperature and high humidity". The maker adds that its laboratory aging, for example at 90 °C and 95 % relative humidity, is not intended to predict the rate in used devices.

A foam is mostly thin cell walls, so it offers the air a large area for its weight, and in a breathing machine that air goes to the patient. As the foam degrades, it may lose volume and form particles, which "are likely to accumulate and may not be directly emitted by the device", the maker writes. FDA states that the foam "can break down" and that the breakdown "may result from exposure to hot and humid conditions". Some CPAP cleaners use ozone, which the maker calls "a strong oxidant".

16 — A foam cell wall breaking down in the air path
blower air to the patient heat and humidity: hydrolysis, mainly at the ester groups ozone: a strong oxidant MOLECULES, IN THE MAKER'S ACCOUNT diethylene glycol (a diol), from the polyol adipic acid (an acid), from the polyol toluene diamine-containing oligomer, then toluene diamine (an amine), from the urethane link particles: likely to accumulate, may not be directly emitted the wall loses volume and crumbles Schematic, not to scale. Mechanism and products as Philips Respironics (the manufacturer) states them, 2023. FDA: the foam can break down, which may result from hot and humid conditions. The case: Chapter 20.
Device materialReleased substanceGas or fluid pathFocal detail
Plate 16 — In the maker's account, heat and humidity hydrolyze the foam mainly at its ester groups, releasing a diol, an acid and an amine-bearing fragment, while the walls lose volume and crumble into particles that are likely to accumulate in the device.

The chemistry follows the environment of use

A polymer's weak links decide where it can serve, so its chemistry is chosen for the environment it will meet. Chapter 3 left one such choice to this chapter: whether the running device's sound-damping foam is a polyester or a polyether polyurethane. In an air path that can be warm and humid, water would cut a polyester's ester links first, while a polyether or polycarbonate polyurethane resists that attack. This book therefore gives the running device a polyether polyurethane foam, a design choice made for the book and not taken from any maker. The polyether is in turn weak to oxidants, ozone from a cleaner among them, so the device's cleaning must take that weakness into account (Chapter 7). Like any foam, the running device's foam is evaluated as aged and used, not as new (Chapters 12 and 19).

In practice

For each polymer part, list the links in its chains that water, oxygen or ozone can cut, and the water, heat and oxidants it meets in use, cleaning included. Choose the chemistry for that environment and record why in the design file. Age at no more than 60 °C and below the material's transitions, add humidity where the polymer hydrolyzes, and confirm the result with real-time aging (Chapter 7). Analyze the aged material, not only the new one, for breakdown products such as acids, diols and amines (Chapter 12).

+ What this chapter established
  • Water cuts a polyester polyurethane at its ester links first, and the acid each cut leaves speeds the next, while urethane links hydrolyze about ten times more slowly and polyether polyurethanes resist hydrolysis.
  • Oxygen attacks what water spares: polyether polyurethanes oxidize, oxidation cuts and crosslinks chains, and ozone attacks rubbers with carbon–carbon double bonds far faster than chains without them.
  • Heat speeds every cut, and the doubling per 10 °C is a protocol's conservative convention, not a law: it cannot be derived from the Arrhenius equation and is advised only below about 60 °C and the material's transitions.
  • Each cut makes a new molecule, so a degrading polyester polyurethane foam can release short chains, a diol, an acid, an amine and particles that its recipe never held.
Created with Claude (Anthropic).
Educational reference material, not engineering, regulatory or clinical advice. Product and company names referenced are the property of their respective owners.