With an annual production of hundreds of millions of tons, the few commodity polymers that dominate the plastics market cannot satisfy all the applications and expectations. In this context, the fabrication of thermodynamically stable polymer blends structured on submicrometre scales raises much hope, but poses significant scientific and industrial challenges. Here, we propose and demonstrate for an industrially relevant system, polyethylene and polyamide, that hitherto inaccessible co-continuous morphologies can be produced over a wide range of compositions by reactive blending. Paradoxically, the self-assembled structures are thermodynamically stable because of the molecular polydispersity inherent in the production method. These nanostructured materials present a unique combination of properties impossible to achieve with classical blends. This versatile, low-cost and simple strategy should be widely applicable.

Introduction. Nonaqueous dispersion polymerization 1-4 is a very versatile method to afford polymer dispersions of controlled morphology. It is initiated in a homogenous medium where all components are miscible. When polymer molar mass exceeds a critical limit, polymers phase separate and aggregate. Steric stabilization of polymer particles is achieved when dispersing agents, usually block copolymers, are added. After formation of particles, further polymerization occurs in bulk of the monomer-swollen particles. Provided that no new particle nucleation takes place, it is possible to obtain narrow particle size distributions. Particle sizes depend mainly upon the solubility of the formed polymer in the medium. In case of better solubility, aggregation of polymer chains is delayed, resulting in larger diameters of the formed particles. Low solubility causes quick aggregation and therefore smaller particle sizes.Advanced methods like living anionic polymerization 5,6 and group transfer polymerization 7,8 (GTP) have been performed in nonaqueous dispersions although most research is concerned with free radical polymerization. Living dispersion polymerization gives both control of molar mass and particle morphology.Controlled radical polymerization [9][10][11][12] is a new living polymerization technique using for example 2,2,6,6tetramethyl-1-piperidyloxy radical (TEMPO) combined with azoisobutyronitrile or benzoyl peroxide (BPO) as initiating system. 13,14,15 Preferably controlled radical styrene polymerization is performed in bulk at temperatures of 100-140 °C. This leads to high viscosities and therefore experimental problems related to stirring at high conversions.Combining controlled radical polymerization and nonaqueous dispersion polymerization should provide the above-mentioned advantages of a living polymerization method and lead to a low viscosity product even at high conversions. In this study living radical polymerization in bulk and nonaqueous dispersion are compared with respect to reaction kinetics and control of polystyrene molar mass. Moreover particle formation during controlled radical styrene dispersion polymerization is investigated.Experimental Section. (a) Materials. Styrene obtained from Fluka was stirred over lithium aluminium hydride for one night and distilled under reduced pressure. TEMPO, BPO (Aldrich), and decane (Fluka) were used as received. The dispersant Kraton G1701, supplied by Shell Chemical Co., is a polystyrene-blockpoly(ethene-alt-propene) with a polystyrene content of 34 wt %. The number average molar mass of the PS block is 35 700 g/mol and of the poly(ethene/-alt-propene) is 68 300 g/mol. Prior to use it was dissolved in

Carboxy-terminated polystyrene, poly(styrene-co-acrylonitrile), and polystyreneblock-poly(styrene-co-acrylonitrile) --with ratios of weight-to number-average molar masses MJM, below 1.3 were synthesized via a controlled radical polymerization mechanism. The polymerizations were initiated with 4,4'-azobis(4-~yanopentane-carboxylic acid) and 2,2,6,6-tetramethyl-1 -piperidyloxyl radical and conducted in bulk at elevated temperatures. The polymerization was monitored by nuclear magnetic resonance, size-exclusion chromatography, end-group titration and differential scanning calorimetry.

TEMPO-mediated free radical polymerization was employed for homo-and copolymerization of vinylferrocene (Vfc). Homopolymerization of Vfc resulted in relatively narrow polydispersities (M w /M n = 1.24 -1.8), however, molecular weights were limited to 4 800. Copolymerization with styrene afforded random copolymers with molecular weights (M n ) up to 10 000, narrow polydispersity (1.2 a M w /M n a 1.4) and up to 42 mol-% Vfc. Block copolymers with PS block and P(S-co-Vfc) block with molecular weights (M n ) in the range of 9 000 to 17 600 (M w /M n a 1.3) were also prepared with up to 17 mol-% vinylferrocene. DSC revealed two glass transition temperatures (T g ) evidencing phase separation.

Two novel bis(1,3-oxazoline-2-yl)-functionalized azo initiators were prepared, and their half-lives were determined by UV spectroscopy. These initiators were used in conjunction with 2,2,6,6-tetramethylpiperidin-1-yloxy (TEMPO) to afford controlled-radical styrene polymerization. Polymerization kinetics, molar mass versus conversion dependence and polydispersities with varying [TEMPO]/[initiator] ratios were examined, and the degree of endgroup functionalities was determined by NMR, FT-IR, endgroup titration and size exclusion chromatography (SEC) data. Well-defined narrow-distributed mono(1,3-oxazolin-2-yl)-terminated polystyrenes with molar mass varying between 1000 and 50 000 with polydispersities of 1.2−1.3 were obtained. The reactivity of the oxazoline moiety was proven by reacting monooxazoline-terminated polystyrene with monocarboxy-terminated polystyrene. Increases in molar mass and the formation of ester amide coupling groups, resulting from the reaction of oxazoline with carboxylic acid endgroups of polystyrene were monitored by means of SEC and FT-IR. Oxazoline endgroup conversion increased with decreasing polystyrene molar mass. Kinetics of the ester amide formation showed Arrhenius-type behavior. The activation energy of melt-phase ester amide formation was determined to be 65 kJ/mol.