Polymerization Methods — Cross-Cutting Comparison

This note compares every major polymerization mechanism described across the Chemistry library on the axes that actually decide which one a chemist or process engineer picks: chain-growth vs step-growth kinetics, achievable dispersity (Đ), tacticity control, monomer scope, scale-up tractability, and the dominant initiator/catalyst family. Each row maps to a real polymer product line so the table doubles as a polymer-architecture index. Read the dimension tables first, then the decision tree for picking-by-objective.

See also

1. The eight families

All industrial polymerization fits in eight mechanistic boxes. Everything more exotic (RAFT, NMP, ATRP, Grubbs ROMP, thiol-ene click, photoiniferter, ionic liquids, supercritical-CO2 dispersion) is a refinement of one of these eight — usually a cleverer way to control initiation, propagation, or termination.

chain-growth                              step-growth
  |                                          |
  radical    ionic     coordination       polycondensation
  |          |         /insertion           |
  FRP        cationic  Ziegler-Natta      polyester
  ATRP       anionic   metallocene        nylon (PA)
  RAFT       living-A  post-metallocene   polyurethane
  NMP                  ROMP (Grubbs)      epoxy / phenolic
                       ADMET              polycarbonate
                                          silicone
                                          ↓
ring-opening              enzymatic          photopolymerization
  |                          |                    |
  lactones (PLA, PCL)      lipase-CALB         acrylate UV-cure
  lactams (nylon-6)        cutinase            thiol-ene click
  epoxides (PEO)           hydrolase           CuAAC (azide-alkyne)
  cyclic siloxanes (PDMS)                      SuFEx

The Sharpless-Meldal-Bertozzi 2022 Nobel Prize in Chemistry was awarded for click chemistry (CuAAC) and bioorthogonal chemistry; the same year, the Yves Chauvin / Robert Grubbs / Richard Schrock 2005 Nobel for olefin metathesis unlocked ROMP for industrial use. Karl Ziegler and Giulio Natta won in 1963 for the catalysts that enabled isotactic polypropylene.

2. Mechanism, monomer scope, dispersity — the deciding axes

FamilyMechanismMonomer typesĐ (PDI) typicalTacticity controlKey initiator / catalystIndustrial products
Free radical (FRP)chain, radical, irreversible terminationvinyl, acrylate, styrene, vinyl chloride2.0–10none (atactic)BPO, AIBN, persulfate, redox pairLDPE, PS, PMMA, PVAc, PVC, NBR
ATRP (Matyjaszewski 1995)reversible-deactivation radical via Cu–halideacrylates, styrene, MA1.05–1.30none (atactic)CuBr/PMDETA, Cu(0) (ARGET, SARA)precision drug-delivery PEG-acrylate blocks
RAFT (Rizzardo/Moad/Thang 1998)reversible-deactivation radical via chain-transfer agentalmost all vinyl monomers1.05–1.30none (atactic)dithioesters, trithiocarbonates, xanthatesblock copolymers, dispersants, gradient copos
NMP (Hawker, Georges)reversible-deactivation via nitroxide cappingstyrenes, some acrylates1.10–1.40none (atactic)TEMPO, SG1, BlocBuilderbulk-scale block copo coatings (Arkema)
Cationicchain, carbocation, fast irreversibleisobutylene, vinyl ethers, styrene-α-methyl2–5 (uncontrolled), 1.05–1.20 (living)partialBF3, AlCl3, TiCl4/H2O (Kennedy living)butyl rubber (IIR), polyisobutylene (PIB)
Anionic / living anionic (Szwarc 1956)chain, carbanion, no terminationstyrene, butadiene, isoprene, EO, lactams1.02–1.10high (stereo via Li counterion)sec-BuLi, n-BuLi, Na-naphthalideSBR, SBS/SIS Kraton, monodisperse PS standards
Polycondensationstep, condensation (–H2O, –HCl, –ROH)difunctional acid/alcohol, diamine, etc1.5–2.5 (Carothers/Flory most-probable: Đ → 2)n/aacid/base, transesterification cats (Sb, Ti, Ge, Sn)PET, PBT, nylon-6,6, polyurethane, polyester resins, polycarbonate (interfacial / melt)
Polyaddition stepstep, no leaving group (urethane, epoxy)diol+diisocyanate, diamine+epoxide2–4n/atin/Sn catalysts (DBTDL), tertiary aminesPU foams (Bayer 1937), epoxy resins, RTV silicone
Ring-opening (ROP)chain or step from cyclic monomerlactones, lactams, epoxides, cyclic siloxanes, NCAs1.05–1.50 (living) up to 2 (cationic)high (Sn(Oct)2, salen-Al)Sn(Oct)2, salen-Al/Zn, organocatalysts (DBU, TBD, NHC)PLA, PCL, PGA, nylon-6, PEG, PDMS, peptides
Ziegler-Natta (heterogeneous)coordination insertion, multi-siteα-olefins4–20 (multi-site)high (TiCl3/MgCl2 isotactic PP)TiCl4/MgCl2 + AlEt3LLDPE, HDPE, PP (commodity)
Metallocene (single-site)coordination insertion, single-siteα-olefins, polar-fct via post-met2.0–2.5very high (C2-symmetric isotactic, Cs syndio)Cp2ZrCl2/MAO, ansa-bridged systemsmLLDPE (Exxon Exact), mPE (Dow Affinity), iPP (LyondellBasell)
Post-metallocene (Brookhart, Bercaw, Mitsui)coordination, polar-monomer-tolerantethylene, polar comonomers2.0–3.0highbis(imino) Pd/Ni (Brookhart), FI-cat (Mitsui), Pt-diiminebranched PE, EVA via Pd, OBC (Dow INFUSE)
ROMP (Grubbs, Nobel 2005)chain, metathesis, ring-opening of cyclic olefinnorbornenes, cyclooctene, DCPD1.05–1.20 (living Grubbs G2/G3)n/aGrubbs G1/G2/G3 Ru, Schrock Mo/Wpoly-DCPD (Telene/Pentam), Vestenamer (Evonik), pNB optical films
ADMET (acyclic diene metathesis, Wagener)step, metathesis of α,ω-dieneα,ω-dienes1.8–2.2n/aGrubbs Ru, Mo Schrocklinear precision PE w/ defects at chosen spacings
Enzymaticchain or step via enzymehydroxy acids, lactones, polyols/diacids1.2–2.0partiallipase CALB (Novozyme 435), cutinasePCL, PHA-mimics, polyesters in biomedical (limited scale)
Photopolymerization (acrylate UV)chain radical, photoinitiatedacrylate / methacrylate / vinyl ether / thiol-ene2–5 (network, no Đ; uncrosslinked: 2)nonebenzophenone, TPO, Irgacure 184/819, BAPOUV-cure coatings, dental composites, SLA/DLP resins (Formlabs), nail-gel
Thiol-ene click (Sharpless–Meldal–Bertozzi 2022)step, radical chain transfer via thiylthiol + ene (alkene, norbornene, vinyl ether)1.1–2.0n/aphotoinitiator (TPO) or thermal AIBNdental sealants (3M), nail gels, microfluidic seals (Carbon DLS resins)

The single highest-impact axis is dispersity Đ. FRP at Đ = 2–10 gives every commodity plastic in the world. Living/controlled radical (ATRP, RAFT, NMP) drops Đ to ~1.1 and makes block copolymers possible — that is the whole reason RDRP exists. Anionic at Đ ≈ 1.05 is still the gold standard for monodisperse standards in GPC/SEC calibration. Step-growth converges to Đ = 2 (Flory’s most-probable distribution) at full conversion and cannot do better without subtractive fractionation. Ziegler-Natta heterogeneous catalysts give Đ = 4–20 because there are multiple active site types on the catalyst surface, each with its own propagation rate; single-site metallocenes collapse Đ to ~2.0 by definition (one site → Schulz-Flory).

3. Tacticity, branching, architecture

Most chain-growth families have tacticity options; step-growth does not (no stereocenter in the backbone unless the monomer has one).

FamilyIsotacticSyndiotacticAtacticBranchedStar/combBlockGradient
FRP✓ (default)✓ (long-chain via H-abstraction)
ATRP/RAFT/NMP✓ (multifunctional initiator)✓ (sequential monomer addition)✓ (continuous monomer feed)
Anionic✓ (Li in nonpolar)partial✓ (in polar solvent)✓ (canonical Kraton SBS)
Ziegler-Natta✓ (iPP, Natta 1955)✓ (sPP w/ Cs catalyst)✓ (LDPE-like via copolymer)partial (via reactor train)
Metallocene✓ (C2-symmetric)✓ (Cs-symmetric)✓ (C1 / oscillating)✓ (constrained-geometry Dow INSITE)✓ (chain-shuttling Dow INFUSE)
Post-metallocene✓ (Dow INFUSE)
ROMP (Grubbs)n/an/an/a✓ (Grubbs G3)✓ (multifunctional CTA)✓ (sequential monomer)
Polycondensation(cf monomer)✓ (hyperbranched via A2+B3)partial (via interchange)
ROP✓ (PLA, salen-Al)✓ (rac-LA w/ Aida-Al)✓ (multifunctional initiator)

The Dow INFUSE olefin block copolymers (OBC) deserve their own mention — chain-shuttling catalysis uses two catalysts and a chain-shuttling agent (diethyl zinc) to make multiblock PE/octene with sharp blocks, commercialized 2007. Until then, blocks in polyolefins required an anionic-style living polymerization that olefins simply don’t permit.

4. Scale, cost, real production tonnage

FamilyAnnual global tonnage (2024)Capex intensityReactor typeSolventPolymer cost
Ziegler-Natta / metallocene~160 Mt (PE + PP)very high ($1-2B for world-scale unit)slurry, gas-phase fluidized-bed (Unipol, Spheripol), solution (Dow, Nova)hexane / iC4 / none$1-2/kg
Polycondensation (PET)~80 Mthigh ($300-500M)melt polycondensationnone (melt)$1.5/kg
Polycondensation (nylon)~9 Mthighmelt, autoclavenone (melt)$3-5/kg
Polycondensation (PU)~22 MtmediumRIM, slabstock, spraynone (reactive)$2-4/kg
FRP (PS, PVAc, acrylics)~30 Mt PS, 25 Mt PVCmediumbulk, suspension, emulsionwater (emulsion), monomer (bulk), org (sol)$1.5-2/kg
Anionic~5 Mt SBR/SBSmediumsolution, cyclohexanecyclohexane$3-5/kg
ATRP/RAFT/NMP< 0.1 Mtlow (specialty)batchvarious$50-1000/kg (specialty)
ROP (PLA, PCL, PEO)~0.5 Mt PLA, 0.1 Mt PCLmediummelt, solutionnone (melt)$2-5/kg PLA, $10-30/kg PCL
ROMP (DCPD, NB)~0.1 Mt (Telene + Pentam)lowRIMnone$5-20/kg
Enzymatic< 0.001 Mtvery low (specialty)batchaq or org$100-1000/kg
Photopolymerization~3 Mt (UV-cure coatings + nails + AM resins)very low (formulator)static-mix + UV curereactive monomer$10-200/kg formulated
Thiol-ene< 0.05 Mtvery lowstatic-mix + UV curereactive$30-500/kg

For perspective: a single LyondellBasell Spheripol PP line in the Gulf Coast produces ~600 kt/year. The world’s entire RAFT polymer output, by contrast, is well under 5 kt/year. Industrial polymerization is dominated by Ziegler-Natta and polycondensation; controlled-radical is a precision specialty.

5. Reactor mode — bulk, solution, emulsion, suspension, precipitation, dispersion

Orthogonal to the mechanism — every chain-growth method can run in multiple reactor modes, each with trade-offs.

ModeContinuous phaseParticle sizeConversion-limit driverExamples
Bulkmonomern/a (continuous)viscosity, hotspot, gel effectbulk PS, bulk PMMA, bulk LDPE (tubular)
Solutioninert solventn/asolvent recovery costsolution SBR (cyclohexane), Dow solution PE
Emulsionwater + surfactant0.05–0.5 µmnucleation locus + chain transferlatex paint, NBR, SBR (E-SBR), PVC paste
Suspensionwater + suspending agent50–500 µmdroplet shear vs coalescencePVC bottle resin, PS expandable beads (EPS), PMMA cast
Precipitationpoor solventporous solidsprecipitation onsetgas-phase PE (Unipol; ethylene as solvent), PVDF in scCO2
Dispersion (incl. scCO2)non-solvent + stabilizer1–10 µmstabilizer adsorptionscCO2 fluoropolymers (DuPont Teflon-AF), pharma microspheres
Interfacialtwo immiscible liquidsfilmmass transfer across interfacepolycarbonate (BPA + phosgene/CH2Cl2 + NaOH aq)
Reactive extrusion (REX)meltbulkresidence time, screw confignylon, PETG modification, peroxide-initiated PP rheology mod

Emulsion is the safety / heat-management champion (water as continuous phase absorbs ΔH). Bulk is the highest space-time-yield but the worst for heat (PMMA bulk cast sheets need slow cure to avoid runaway). Supercritical CO2 dispersion eliminates VOCs and is the only practical solvent for fluoropolymer chains that swell every conventional solvent.

6. Scale-up tractability

FamilyLab → pilot → world-scaleBottleneckNotable failures
Ziegler-Natta / metallocenewell-trodden; thousands of plantscatalyst poisoning by H2O/O2/CO ppm-levelINEOS Köln 2008 (loss of cooling); Pemex Pajaritos 2016 (VCM not Ziegler but exemplifies the safety frame)
Polycondensationmaturedust + finishing (chip/pellet handling), Sb/Sn residueDowDuPont nylon-6,6 ADN shortage 2018
FRP (bulk + emulsion + suspension)mature; well-understoodgel effect runaway (Trommsdorff–Norrish)T2 Lab 2007 (Jacksonville FL) — sodium-aided MCMT runaway, killed 4
Anionicmature for SBR/SBS; specialty otherwiseabsolute exclusion of H2O/O2/CO2rare incidents — failures are usually capacity scale-up, not safety
ATRP/RAFT/NMPtough; only specialty scalecopper removal (ATRP), CTA odor + color (RAFT), nitroxide cost (NMP)none material
ROP (PLA)mature (NatureWorks Blair NE 150 kt)water sensitivity (Sn(Oct)2 deactivated), monomer purityrecurring water-ingress issues, lost batches
ROMP (DCPD)mature for poly-DCPD via RIMcatalyst cost (Grubbs G2 was $10000/kg in 2010s)none material
Enzymaticnot scaled beyond demoenzyme cost + thermal stability + slow rate (TON limit)many academic demos, no commodity scale
Photopolymerizationscaled but only in thin layers / 3D voxelsUV penetration depth ~100 µm — limits thicknessnot a “plant-scale” mechanism by nature
Thiol-enescaled in coatings / AMthiol odor (alkyl thiols smell vile)none

The two pathological behaviors that have killed people in industry are the Trommsdorff–Norrish gel effect (auto-acceleration in bulk FRP when viscosity rises and termination becomes diffusion-limited — the polymerization rate runs away, generating heat faster than the jacket can remove) and the monomer flash on loss of cooling (the 1989 Phillips Pasadena disaster — ethylene fed into a polymerization reactor with stuck-open valve; the cloud ignited, killed 23). Both are taught in PSM (process safety management) courses and informed the OSHA 1992 Process Safety Management standard, the EPA RMP (40 CFR 68), and Seveso II/III in the EU.

7. Stereochemistry — the catalyst spectrum

Stereoregular polymerization is what makes polypropylene useful. Atactic PP is a tacky, soft, low-MW mess; isotactic PP (Tm 165°C) is a crystalline engineering plastic. The catalyst maps:

  • TiCl3 (heterogeneous), Ziegler 1953, Natta 1955 — isotactic PP. Multi-site, broad Đ.
  • TiCl4/MgCl2/donor (4th gen Z-N) — ~95% isotactic, single product line for all commodity iPP.
  • C2-symmetric ansa-metallocene (Brintzinger, Ewen, Spaleck) — isotactic. Single-site, Đ ~2.0, very high stereo control.
  • Cs-symmetric ansa-metallocene (Ewen 1988) — syndiotactic PP, Tm 137°C, crystalline.
  • C1-symmetric (oscillating) — atactic-isotactic stereoblock (Coates–Waymouth 1995, nice academic curiosity).
  • Salen-Al / salen-Cr / salen-Zn (ROP) — rac-lactide → isotactic PLA via enantiomorphic site control. Coates, Feijen, Spassky.

8. Modern catalyst innovations (2020–2026)

  • Photoiniferter (visible-light) ATRP/RAFT — Hawker, Boyer 2018 onward. Avoids transition metals; ppm-level catalyst with blue/green LED. Now used commercially for high-purity biomedical polymers (Polymer Factory).
  • Aqueous ATRP (Matyjaszewski 2018+) — ARGET/SARA with parts-per-million copper, viable in industrial water-based formulations.
  • Living cationic polymerization at room temperature — Kennedy → Aoshima/Sawamoto/Kamigaito. Polyvinyl ethers at controlled MW.
  • Group transfer polymerization (GTP) — Webster/DuPont 1983, mostly historical, but resurfaced 2020+ for high-throughput MA-derivative libraries.
  • Stereoselective ROMP — Schrock W/Mo Z-selective catalysts (2011+), Grubbs Z-selective (2013+). Make all-cis poly(norbornene).
  • Chain-shuttling olefin block copolymers (OBC) — Dow INFUSE/INTUNE, 2007 introduction, now ~100 kt/year specialty. The single largest mechanism innovation in olefin polymerization since metallocenes.
  • Light-driven step-growth (SuFEx) — Sharpless 2014, click 2.0. Sulfur(VI)-fluoride exchange chemistry. Quietly going through scale-up at multiple specialty firms.
  • Catalytic chain-transfer polymerization (CCTP) — Co(II) catalysts make low-MW macromonomers in FRP, useful for graft architectures.

9. End-of-life and recycling implications

Polymerization mechanism informs depolymerizability:

FamilyDepolymerizable?MechanismNote
Step (PET, nylon)yeshydrolysis, methanolysis, glycolysisLoop Industries, IBM VolCat, Eastman ICI for PET
ROP (PLA, PCL)yeshydrolysis (industrial compost)NatureWorks
ROMP (poly-DCPD)partial (back-biting under heat with Ru cat)retro-metathesisacademic
FRP commodities (PE, PP, PS)noradical depolymerization off-route except PS → styrene at 350°C+mechanical / chemical (pyrolysis) recycling
Ziegler-Natta PEnopyrolysis onlyPlastic Energy, Loop
Anionic SBR/SBSnopyrolysistire recycling
Photopolymer networksnocrosslinked thermosetlandfill / pyrolysis
Polyurethanepartialglycolysis (BASF Loopamid), hydrolysisBASF Schwarzheide demo plant 2024
Epoxyhardvitrimers (Leibler 2011) reversible at high Tnew and growing

Vitrimers (Ludwik Leibler, ESPCI 2011) sit between thermoplastic and thermoset — covalent network with dynamic exchange bonds (transesterification, imine, disulfide). Mallinda’s epoxy vitrimer is shipping for aerospace + wind-blade applications as of 2025.

10. Side-by-side — when to pick each

If your priority is…Pick this mechanism
Cheapest polymer per kg, any propertiesZiegler-Natta PE or polycondensation PET
Isotactic crystalline PPZiegler-Natta 4th gen or C2-metallocene
Block copolymer (well-defined)living anionic > RAFT > ATRP
Block copolymer (commodity scale)chain-shuttling olefin (Dow INFUSE)
Star or comb polymerATRP/RAFT/NMP with multifunctional initiator
Functional-group tolerance + radical chemistryRAFT > ATRP > NMP
Bioresorbable (medical)ROP of lactide / lactone / NCAs
UV-cure coatingacrylate photopolymerization
Dental composite or adhesivethiol-ene click + methacrylate
All-cis polymerSchrock Z-selective metathesis
Aqueous emulsion latexFRP emulsion (BPO/persulfate redox)
RIM thermoset part (auto bumper)poly-DCPD ROMP RIM
Recyclable polyestermelt polycondensation PET (existing infra)
Recyclable polycarbonateBayer CO2-based PC (Covestro 2016+)
Bio-based polyamideROP of lactam (nylon-11 from castor, Arkema)
Network with dynamic bondsvitrimer (epoxy + transesterification)
Living cationic polyvinyl etherAoshima HI/I2-based system

11. Decision tree

What polymer architecture do you need?
├─ Random copolymer, broad Đ, cheap, commodity scale
│    → Ziegler-Natta (LLDPE), polycondensation (PET), FRP emulsion (latex)
├─ Random copolymer, narrow Đ, specialty
│    → RAFT > ATRP > NMP (radical-tolerant)
│    → living anionic (high purity, dry box)
├─ Block copolymer
│    ├─ commodity scale + olefins → chain-shuttling (Dow OBC)
│    ├─ specialty scale + styrenics → anionic (Kraton SBS)
│    ├─ acrylates → RAFT or ATRP
│    └─ polyester/polylactam → sequential ROP (PLA-b-PCL)
├─ Stereoregular (iso/syndio)
│    ├─ α-olefin → Ziegler-Natta or metallocene
│    ├─ polylactide → salen-Al ROP
│    └─ polyvinyl ether → living cationic with stereodirecting cat
├─ Crosslinked network
│    ├─ thermoset rigid → epoxy/phenolic step
│    ├─ elastomer → PU step
│    ├─ optical/dental → thiol-ene or acrylate UV
│    └─ recyclable network → vitrimer (epoxy + transesterification)
├─ Star / comb / hyperbranched
│    ├─ star → multifunctional initiator ATRP/RAFT or anionic
│    ├─ comb → grafting-from ATRP
│    └─ hyperbranched → A2+B3 step
├─ Bioresorbable medical implant
│    └─ ROP (PLA, PCL, PGA, PLGA), Sn(Oct)2 or organocat
└─ Recyclable / depolymerizable
     ├─ step (PET, nylon) → existing
     └─ vitrimer / dynamic covalent → new (Leibler, Du Prez)

Adjacent

When to pick what

The fastest narrowing is: commodity (>1 Mt/yr) → mature step or insertion; specialty (<10 kt/yr) → RDRP or living ionic; biomedical → ROP; photo/AM → acrylate UV or thiol-ene; recyclable network → vitrimer. Within those, the catalyst/initiator system is decided by the monomer’s functional-group compatibility — radical chemistry tolerates almost everything except thiols at scale (chain transfer); anionic chemistry tolerates almost nothing protic; coordination chemistry tolerates almost no polar groups except in post-metallocene Brookhart Pd systems.