Ring-Opening Polymerization

How it works: strain, propagation, and no lost byproduct

ROP is a chain-growth reaction: an initiator generates one active center, and cyclic monomers add one at a time to that single reactive chain end. Because the ring simply unzips into the backbone, ROP is formally an addition polymerization even though the monomers (lactones, lactams, ethers) contain the heteroatom linkages of a condensation polymer. This is the trick that lets chemists make a polyester or polyamide without expelling water — and therefore without the high temperatures and equilibrium limits that plague step-growth routes.

The thermodynamics are governed by ring strain. Three-membered rings such as ethylene oxide carry roughly 27 kcal/mol of strain; four-membered β-lactones and oxetanes are similar; these polymerize readily. Five- and seven-membered rings (γ-butyrolactone's larger cousins, ε-caprolactone, ε-caprolactam) have moderate strain and polymerize under the right catalyst. Nearly strain-free rings are the exception: γ-butyrolactone (five-membered) is famously reluctant to polymerize because ΔG is close to zero. Every ROP has a ceiling temperature (T_c) above which the equilibrium reverses and the polymer depolymerizes back to monomer.

Anionic ROP: alkoxide and amide chain ends

In anionic ROP a nucleophile attacks the monomer and the resulting anion becomes the propagating species. For ethylene oxide, an alkoxide or hydroxide (from KOH, potassium tert-butoxide, or an alkyllithium) attacks the less-hindered ring carbon; the strained C–O bond breaks, and a new alkoxide is generated at the chain end that attacks the next monomer. The reaction is a textbook S_N2 ring opening — backside attack, inversion at carbon — and it proceeds cleanly to give poly(ethylene oxide)/PEG. Because chain transfer and termination are largely absent under anhydrous conditions, ethylene oxide ROP is a classic living polymerization: chains grow uniformly, giving narrow dispersity and predictable molar mass set by the monomer-to-initiator ratio.

For lactones (e.g. ε-caprolactone), the alkoxide attacks the carbonyl carbon and the acyl–oxygen bond cleaves, extending the chain with an ester linkage and regenerating an alkoxide end. For ε-caprolactam → Nylon-6, the industrial anionic route uses a lactam anion (from sodium caprolactamate) plus an N-acyl activator; it runs fast enough for reaction-injection molding of cast nylon parts. A competing hydrolytic route to Nylon-6 opens caprolactam with ~5–10% water at 250–270 °C, then polycondenses — a hybrid of ring-opening and step-growth chemistry.

Cationic ROP: oxocarbenium and activated-monomer routes

Cationic ROP uses electrophilic initiators — Brønsted superacids like triflic acid, Lewis acids like BF₃·OEt₂, or oxocarbenium/oxonium salts such as Meerwein's reagent (Et₃O)BF₄. The classic example is tetrahydrofuran (THF): protonation or alkylation of the ring oxygen gives a cyclic oxonium ion; the next THF's oxygen then does an S_N2 attack on the strained α-carbon, opening the ring and shuttling the positive charge to the new chain end. This gives poly(tetramethylene oxide), the soft-segment glycol used in Spandex/Lycra.

Cationic ROP of epoxides and lactones is notoriously prone to side reactions: back-biting (the chain-end cation attacks an internal oxygen to expel cyclic oligomers such as crown ethers and dioxane) and chain transfer, both of which broaden the dispersity. The activated-monomer mechanism — running the reaction in the presence of an alcohol so that the growing species is a neutral hydroxyl end reacting with a protonated monomer — suppresses cyclic-oligomer formation and gives better molar-mass control. Cationic ROP is essential for monomers that cannot tolerate strong nucleophiles, including cyclic acetals, oxazolines (→ polyethylenimine precursors), and aziridines.

Coordination-insertion ROP: the route to PLA

The most controlled ROP of cyclic esters uses a metal alkoxide catalyst that both activates the carbonyl (by coordinating the Lewis-basic oxygen to the metal) and delivers the growing chain from the metal–alkoxide bond. The monomer inserts into the M–OR bond via a four-membered transition state, cleaving the acyl–oxygen bond and regenerating a metal alkoxide one unit longer. Because the metal shepherds each addition, dispersities can approach 1.0 and end groups are cleanly defined.

Industrially, the workhorse is tin(II) 2-ethylhexanoate, Sn(Oct)₂, used with an alcohol co-initiator to make polylactide (PLA) and polycaprolactone. Sn(Oct)₂ is favored because it is FDA-accepted for food-contact and bioabsorbable applications and works in the melt at 130–180 °C. Aluminum isopropoxide, Al(OiPr)₃, and a family of well-defined zinc, magnesium, and yttrium single-site catalysts (developed heavily by Coates, Feijen, Carpentier and others since the 1990s) give even better stereocontrol. Racemic lactide can be polymerized to isotactic stereoblock or heterotactic PLA depending on the ligand, and the stereoregular stereocomplex of PLLA and PDLA melts ~50 °C higher (~230 °C) than either enantiopure polymer alone.

Ring-opening metathesis polymerization (ROMP)

A distinct and powerful variant is ring-opening metathesis polymerization (ROMP), which opens strained cyclic olefins — norbornene, cyclooctene, dicyclopentadiene — using a metal-carbene (alkylidene) catalyst. The chain grows by repeated [2+2] cycloaddition/retro-[2+2] steps: the catalyst's M=CH forms a metallacyclobutane with the ring double bond, then cleaves to insert the monomer and place the carbene back at the chain end. Every double bond of the monomer is preserved in the backbone, so the polymer is unsaturated and can be further functionalized or crosslinked.

ROMP was transformed by the well-defined catalysts of Richard Schrock (molybdenum/tungsten alkylidenes) and Robert Grubbs (ruthenium alkylidenes) — work recognized with the 2005 Nobel Prize in Chemistry (shared with Yves Chauvin, who proposed the metallacyclobutane mechanism in 1971). Grubbs catalysts tolerate air, moisture, and many functional groups, making ROMP a living polymerization that builds precise block copolymers, brush polymers, and the commercial thermoset polydicyclopentadiene (pDCPD) used in tough molded parts.

Applications, scope, and limitations

ROP underpins an enormous slice of the polymer economy:

• Nylon-6 from ε-caprolactam — fibers, engineering plastics, cast nylon.
• PEG / PEO from ethylene oxide — pharmaceutical excipients, PEGylated drugs, surfactants, lithium-battery electrolytes.
• PLA, PGA, PCL from lactide, glycolide, and caprolactone — compostable packaging, 3D-printing filament, resorbable surgical sutures and drug-delivery scaffolds.
• Silicones from cyclic siloxanes (D₃, D₄) — sealants, medical elastomers, release coatings.
• Polyoxymethylene (POM/Delrin) from trioxane, and Spandex soft segments from THF.

The main limitations are thermodynamic and kinetic. Low-strain six-membered rings often cannot be polymerized at all (the ceiling temperature is too low). Water and other protic impurities act as chain-transfer agents that cap molar mass, so anionic and coordination ROP demand rigorously dry, inert conditions. Cationic ROP fights back-biting and cyclic-oligomer formation. And because most ROP polymers have hydrolyzable backbone linkages, the same feature that makes PLA compostable also limits its long-term stability — a trade-off chemists tune through stereochemistry, comonomers, and end-group design.

The three principal ROP mechanisms

Mechanism	Active chain end	Typical monomers	Representative initiator/catalyst
Anionic	Alkoxide / amide / carbanion	Ethylene oxide, ε-caprolactone, ε-caprolactam	KOtBu, R-Li, NaH, Na-caprolactam
Cationic	Oxocarbenium / acylium / onium	THF, epoxides, oxazolines, cyclic acetals	BF₃·OEt₂, triflic acid, (Et₃O)BF₄
Coordination-insertion	Metal alkoxide (M–OR)	Lactide, ε-caprolactone, glycolide	Sn(Oct)₂/ROH, Al(OiPr)₃, Zn/Mg complexes