Unusual Lower Critical Solution Temperature Phase Behavior of Poly(benzyl methacrylate) in a Pyrrolidinium-Based Ionic Liquid

Polymer/ionic liquid systems are being increasingly explored, yet those exhibiting lower critical solution temperature (LCST) phase behavior remain poorly understood. Poly(benzyl methacrylate) in certain ionic liquids constitute unusual LCST systems, in that the second virial coefficient (A₂) in dilute solutions has recently been shown to be positive, indicative of good solvent behavior, even above phase separation temperatures, where A₂ < 0 is expected. In this work, we describe the LCST phase behavior of poly(benzyl methacrylate) in 1-butyl-1-methylpyrrolidinium bis(trifluoromethylsulfonyl)imide for three different molecular weights (32, 63, and 76 kg/mol) in concentrated solutions (5–40% by weight). Turbidimetry measurements reveal a strong concentration dependence to the phase boundaries, yet the molecular weight is shown to have no influence. The critical compositions of these systems are not accessed, and must therefore lie above 40 wt% polymer, far from the values (ca. 10%) anticipated by Flory-Huggins theory. The proximity of the experimental cloud point to the coexistence curve (binodal) and the thermo-reversibility of the phase transitions, are also confirmed at various heating and cooling rates.     

Effect of dilution on a block copolymer in the complex phase window

The phase behavior of a styrene–isoprene (SI) diblock copolymer, with block molecular weights of 1.1 × 10⁴ and 2.1 × 10⁴ g/mol, respectively, is examined in the neutral solvent bis(2-ethylhexyl) phthalate (DOP) and the styrene-selective solvent di-n-butyl phthalate (DBP). DBP is a good solvent for PS, but is near a theta solvent for PI at approximately 90°C. Small-angle X-ray scattering (SAXS), rheology, and static birefringence are used to locate and identify order–order (OOT) and order–disorder transitions (ODT); all three techniques gave consistent results. The neat polymer adopts the gyroid (G) phase at low temperatures, with an OOT to hexagonally-packed cylinders (C) at 185°C, and the ODT at 238°C. Upon dilution with the neutral solvent DOP, the C window is diminished, until for a polymer concentration ϕ = 0.65, a direct G to disorder (D) ODT is observed. These results reflect increased stability of the disordered state, based on the different concentration scalings of the interaction parameter, χ, at the OOT and ODT. The OOT follows the dilution approximation, i.e., χ_OOT ∼ ϕ⁻¹, but the ODT is found to follow a stronger concentration dependence, i.e., χ_ODT ∼ ϕ^−1.4, similar to the scaling of ϕ^−1.6 found previously for lamellar SI diblocks in toluene and DOP. Addition of the selective solvent DBP produces dramatic changes in the phase behavior relative to DOP and the melt state; these include transitions to lamellar (L) and perforated layer (PL) structures. The observed phase sequences can be understood in terms of trajectories across the SI melt phase map (temperature vs. composition): addition of a neutral solvent or increasing temperature corresponds to a “vertical” trajectory, whereas adding a selective solvent amounts to a “horizontal” trajectory. When the solvent selectivity depends on temperature, as it does for the SI/DBP system, increasing temperature results in a diagonal trajectory. For both neutral and selective solvents the domain spacing, d*, scales with ϕ and χ as anticipated by self-consistent mean-field theory. © 1998 John Wiley & Sons, Inc. J Polym Sci B: Polym Phys 36: 3101–3113, 1998     

Unexpected Intermediate State for the Cylinder‐to‐Gyroid Transition in a Block Copolymer Solution

The cylinder‐to‐gyroid transition in a concentrated solution of polystyrene‐block‐polyisoprene in dibutyl phthalate has been studied using rheology and small angle X‐ray scattering. Following an appropriate temperature quench, the oriented cylinder phase transforms to the gyroid structure epitaxially. Remarkably, an intermediate state appears for a deep quench, whereas for a shallow quench the transition proceeds directly; the intermediate state exhibits scattering signatures consistent with a hexagonally perforated layer structure.     

Dynamics of Bond‐Fluctuation Model Chains in Good and Theta Solvents

The dynamic Monte Carlo algorithm is employed to explore the dynamics of flexible linear chains. The chains are represented by the bond‐fluctuation model with and without attractions between non‐bonded units placed at close distances. This mimics the behavior of real chains in the good and poorer solvents. We obtain the chain sizes, diffusion coefficients, Rouse modes, and their relaxation times. We also evaluate the time correlation function of the end‐to‐end vector at different concentrations. Subsequently, we compare the dependence of the simulation results on chain length, solvent quality, concentration, and mode order with the corresponding theoretical predictions. We observe a retardation of diffusion for non‐dilute systems close to the theta state. This retardation is too high to be exclusively attributed to the increase of global friction and can be caused by temporary adherence of the chains to transient clusters.     

Entropically driven phase separation of highly branched/linear polyolefin blends

Small‐angle neutron scattering (SANS) was used to examine the melt phase behavior of a heavily branched comb PEE polymer blended separately with two linear PEE copolymers. In this case, PEE refers to poly(ethylene‐r‐ethylethylene) with 10% ethylene units; therefore, the molecular architecture was the only difference between the two components of the blends. The molecular weights of the two linear random copolymers were 60 and 220 kg/mol, respectively. The comb polymer contained an average of 54 long branches, with a molecular weight of 13.7 kg/mol, attached to a backbone with a molecular weight of 10 kg/mol. Three different volume compositions (25/75, 50/50, and 75/25) were investigated for both types of blends. SANS results indicate that all the blends containing the lower molecular weight linear polymer formed single‐phase mixtures, whereas all the blends containing the high molecular weight linear polymer phase‐separated. These results are discussed in the context of current theories for polymer blend miscibility. © 2000 John Wiley & Sons, Inc. J Polym Sci B: Polym Phys 38: 2965–2975, 2000     

Influence of long‐chain branching on the miscibility of poly(ethylene‐r‐ethylethylene) blends with different microstructures

The melt miscibility of two series of poly(ethylene‐r‐ethylethylene) (PE_Exx) polymers with different percentages (xx) of ethylethylene (EE) repeat units was examined with small‐angle neutron scattering (SANS). The first series consisted of comb/linear blends in which the first component is a heavily branched comb polymer (B90) containing 90% EE and an average of 62 long branches with a weight‐average molecular weight (M_W) of 5.5 kg/mol attached to a backbone with M_W = 10.0 kg/mol. The comb polymer was blended with six linear PE_Exx copolymers, all of which had M_W ≈ 60 kg/mol and EE percentages ranging from 55 to 90%; they were denoted L55 to L90, with the number referring to the percentage of EE repeat units. The second series consisted of linear/linear blends; the first component, with M_W = 220 kg/mol and 90% EE, was denoted L90A, and the second components were the same series of linear polymers, with M_W ≈ 60 kg/mol and various EE compositions. The concentrations investigated were 50/50 w/w, except for the blend of branched B90 and linear L90 (both components were 90% EE), for which 25/75 and 75/25 concentrations were also examined. The SANS results indicated that for the comb/linear blends, only the dB90/L90 blends were miscible, whereas the other five blends phase‐separated; for the linear/linear blends, dL90A/L83 and dL90A/L78 were miscible, whereas the other three blends were immiscible. These results indicate that long‐chain branching significantly narrowed the miscibility window of these polyolefin blends. © 2002 Wiley Periodicals, Inc. J Polym Sci Part B: Polym Phys 40: 466–477, 2002; DOI 10.1002/polb.10102     

Star-shaped polymers by living cationic polymerization. VIII. Size and shape of star poly(vinyl ether)s determined by dynamic light scattering and computer simulation

The sizes and shapes of star-shaped poly(vinyl ether)s, prepared by living cationic polymerization, were studied by dynamic light scattering and molecular mechanics-based computer simulation. The hydrodynamic radii (R_h) of star poly(isobutyl vinyl ether)s (4a; M?_w = 2.2 × 10⁴ ? 1.7 × 10⁵) determined by dynamic light scattering were in the range from 30 to 90 Å in tetrahydrofuran or ethyl acetate. Consistent with the expected multiarmed architecture of 4a, the radius for a given number (f) of arms per molecule increased with the degree of polymerization [DP(arm)] of the arms, and for a fixed DP(arm), the radius increased with f. The relationship between arm number f and the “shrinking” factor h [R_h(star)/R_h(linear)] was consistent with multibranched structures for the star polymers. These results are supported by those for the molecular weight itself; the apparent weight-average molecular weights by size-exclusion chromatography are less than the corresponding absolute values by static light scattering. The dependence of h on f suggests some degree of asymmetry in the star shape. Similar results were also obtained by the computer simulation of potential energy-minimized conformations of the arms, which implied almost spherical but slightly asymmetric shapes. The computer simulation also demonstrated that the star polymer (4b) with pendant hydroxyl groups in the arms is smaller in size than the corresponding alkyl (isobutyl) (4a) with the identical arm number and arm degree of polymerization. © 1995 John Wiley & Sons, Inc.     

Evolution of multicompartment micelles to mixed corona micelles using solvent mixtures

Miktoarm star triblock copolymers mu-[poly(ethylethylene)][poly(ethylene oxide)][poly(perfluoropropylene oxide)] self-assemble in dilute aqueous solution to give multicompartment micelles with the cores consisting of discrete poly(ethylethylene) and poly(perfluoropropylene oxide) domains. Tetrahydrofuran is a selective solvent for both the poly(ethylethylene) and poly(ethylene oxide) blocks, and thus in tetrahydrofuran mixed corona micelles are favored with poly(perfluoropropylene oxide) cores. The introduction of tetrahydrofuran into water induces an evolution from multicompartment micelles to mixed corona [poly(ethylethylene) + poly(ethylene oxide)] micelles, as verified by dynamic light scattering and nuclear magnetic resonance spectroscopy. A mixed solvent containing 60 wt % tetrahydrofuran corresponds to the transition point, as verified by analysis of a poly(ethylethylene)-poly(ethylene oxide) diblock copolymer in the same solvent mixtures. Furthermore, cryogenic transmission electron microscopy suggests that, as the poly(ethylethylene) block transitions from the core to the corona, the micelle morphologies evolve from disks to oblate ellipsoid micelles (with some vesicles), with worms and spheres evident at intermediate compositions.     

Electron tunneling in lithium-ammonia solutions probed by frequency-dependent electron spin relaxation studies

Electron transfer or quantum tunneling dynamics for excess or solvated electrons in dilute lithium-ammonia solutions have been studied by pulse electron paramagnetic resonance (EPR) spectroscopy at both X- (9.7 GHz) and W-band (94 GHz) frequencies. The electron spin-lattice (T(1)) and spin-spin (T(2)) relaxation data indicate an extremely fast transfer or quantum tunneling rate of the solvated electron in these solutions which serves to modulate the hyperfine (Fermi-contact) interaction with nitrogen nuclei in the solvation shells of ammonia molecules surrounding the localized, solvated electron. The donor and acceptor states of the solvated electron in these solutions are the initial and final electron solvation sites found before, and after, the transfer or tunneling process. To interpret and model our electron spin relaxation data from the two observation EPR frequencies requires a consideration of a multiexponential correlation function. The electron transfer or tunneling process that we monitor through the correlation time of the nitrogen Fermi-contact interaction has a time scale of (1-10) × 10(-12) s over a temperature range 230-290 K in our most dilute solution of lithium in ammonia. Two types of electron-solvent interaction mechanisms are proposed to account for our experimental findings. The dominant electron spin relaxation mechanism results from an electron tunneling process characterized by a variable donor-acceptor distance or range (consistent with such a rapidly fluctuating liquid structure) in which the solvent shell that ultimately accepts the transferring electron is formed from random, thermal fluctuations of the liquid structure in, and around, a natural hole or Bjerrum-like defect vacancy in the liquid. Following transfer and capture of the tunneling electron, further solvent-cage relaxation with a time scale of ～10(-13) s results in a minor contribution to the electron spin relaxation times. This investigation illustrates the great potential of multifrequency EPR measurements to interrogate the microscopic nature and dynamics of ultrafast electron transfer or quantum-tunneling processes in liquids. Our results also impact on the universal issue of the role of a host solvent (or host matrix, e.g. a semiconductor) in mediating long-range electron transfer processes and we discuss the implications of our results with a range of other materials and systems exhibiting the phenomenon of electron transfer.