RAFT Polymerization

The RAFT mechanism, step by step

RAFT rides on top of conventional radical polymerization, so it keeps the same four stages — initiation, propagation, transfer, and termination — but inserts a fast, reversible transfer equilibrium that is the heart of the control.

• Initiation. A thermal initiator such as AIBN (azobisisobutyronitrile) or a redox pair decomposes to give a primary radical, which adds a few monomer units to become a short propagating radical, P_n•.
• Pre-equilibrium (addition–fragmentation). P_n• adds to the C=S double bond of the CTA (structure Z–C(=S)–S–R). This makes an intermediate carbon radical that then fragments, expelling the R group as a new radical, R•, and leaving the polymer capped as a dormant macro-CTA, P_n–S–C(=S)–Z.
• Reinitiation. The released R• adds to fresh monomer and starts its own chain.
• Main equilibrium. Now the process becomes degenerative: a growing radical P_m• adds to a dormant chain P_n–S–C(=S)–Z, forms a symmetric intermediate, and fragments to release P_n• while capping P_m. Because this exchange is far faster than propagation, the radical is passed rapidly among all chains, so each spends most of its life dormant and grows in small, equal increments.

Termination (radical–radical coupling or disproportionation) still happens, but only a tiny fraction of chains are radicals at any instant, so dead-chain formation is minimized. The living character means chains retain the thiocarbonylthio end group and can be re-activated later.

Conditions, reagents, and choosing a CTA

RAFT is prized for being experimentally forgiving. It runs in bulk, in organic solvents (toluene, dioxane, DMF), in water, and even in dispersed media (emulsion, miniemulsion) at temperatures anywhere from ambient (with UV or redox initiation) up to about 70–90 °C for AIBN-driven runs. Radical initiator loading is kept low — often a CTA-to-initiator ratio of 5:1 to 10:1 — so that most chains originate from the CTA, not the initiator.

The target degree of polymerization is set simply by the monomer-to-CTA ratio: [M]₀/[CTA]₀ times fractional conversion gives the number-average degree of polymerization. The CTA must be matched to the monomer:

• More-activated monomers (MAMs) — styrene, methacrylates, acrylates, acrylamides — pair best with dithiobenzoates (Z = phenyl) and trithiocarbonates (Z = alkylthio), which give strong, fast control.
• Less-activated monomers (LAMs) — vinyl acetate, N-vinylpyrrolidone, N-vinylcarbazole — need xanthates (Z = O-alkyl, sometimes called MADIX) or dithiocarbamates, whose weaker C=S stabilization avoids retardation.

The R group must be a good homolytic leaving group and a good reinitiating radical — cumyl, cyanoisopropyl, or a tertiary ester radical are common choices.

Scope, limitations, and the smell problem

RAFT boasts arguably the broadest monomer scope of any controlled radical method, tolerating unprotected functional groups (–OH, –COOH, amines, epoxides) and aqueous conditions — which is why it is a workhorse for water-soluble and bio-relevant polymers. Its main practical drawbacks come from the CTA itself:

• Color and odor. Dithiobenzoate end groups leave products pink-to-red, and sulfur-containing byproducts can be malodorous. End-group removal — by aminolysis to a thiol, radical-induced reduction, or thermolysis — is often a required cleanup step.
• Monomer/CTA mismatch. Using a dithiobenzoate with a LAM, or a xanthate with a MAM, causes retardation or poor control. There is no single universal CTA.
• Hydrolytic sensitivity. Some thiocarbonylthio groups slowly hydrolyze in water at extreme pH, capping the living chain end.

As with all radical processes, no method eliminates termination entirely; RAFT simply keeps its fraction low enough that dispersities near 1.0–1.2 are routine for well-behaved systems.

Architectures: blocks, stars, and beyond

Because every chain keeps a reactivatable end group, an isolated RAFT polymer is itself a macro-CTA. Adding a second monomer restarts polymerization and grows a second block cleanly onto the first — the standard route to well-defined AB and ABA block copolymers. Care is needed with block order: the first block's propagating radical must be a competent leaving group for the second monomer, so MAMs are generally polymerized before LAMs.

• Star polymers come from multifunctional CTAs, attaching several arms to a core (either R-group or Z-group approach).
• Gradient and statistical copolymers emerge from feeding comonomers of different reactivity together.
• Polymer–protein and polymer–surface conjugates use CTAs bearing biotin, NHS esters, or silanes.

A powerful extension is PISA (polymerization-induced self-assembly): growing a solvophobic block onto a soluble macro-CTA in water drives in-situ formation of nanoparticles, worms, or vesicles at high solids content — a scalable route to functional nano-objects.

Why RAFT matters

Controlled radical polymerization let the precision of anionic "living" polymerization escape the glovebox: no rigorously dry, oxygen-free apparatus, no super-reactive organolithium initiators. RAFT specifically brought that precision to water and functional monomers, which is why it dominates in areas anionic methods could never reach:

• Drug delivery and biomaterials — stimuli-responsive block copolymers, well-defined PEG-like and zwitterionic chains, polymer–drug conjugates.
• Coatings, dispersants, and rheology modifiers — where controlled block architecture tunes surface activity.
• Nanotechnology and self-assembly — block copolymer templates for lithography and PISA-made nanoparticles.

Together with ATRP and nitroxide-mediated polymerization, RAFT is one of the three pillars of reversible-deactivation radical polymerization (RDRP), the family that IUPAC now uses as the umbrella term for these controlled methods.

A short history

The controlled-radical revolution accelerated through the 1990s: nitroxide-mediated polymerization was demonstrated in the early 1990s, and ATRP was reported independently by Matyjaszewski and by Sawamoto in 1995. RAFT arrived in 1998 in a CSIRO patent and a landmark Macromolecules paper by Chiefari, Rizzardo, Thang, Moad and coworkers, which showed that a family of thiocarbonylthio compounds could impart living character to conventional radical polymerization. The closely related xanthate-based MADIX process was described by Rhodia around the same time. Over the following two decades RAFT expanded to photoinduced, enzyme-initiated, and oxygen-tolerant variants, and became a standard tool in both academic and industrial polymer laboratories worldwide.

How the three main controlled radical polymerization methods differ

Method	Control element	Metal catalyst	Notable trait
RAFT	Degenerative chain transfer via thiocarbonylthio CTA	None	Widest monomer scope; leaves a colored/odorous end group
ATRP	Reversible halogen abstraction by Cu complex	Cu (or other transition metal)	Halide chain end; needs metal removal
NMP	Reversible capping by a nitroxide (e.g. TEMPO)	None	Simple, but limited monomer range and high temperatures