Poly(2-isopropyl-2-oxazoline)-b-poly(lactide) (PiPOx-b-PLA) Nanoparticles in Water: Interblock van der Waals Attraction Opposes Amphiphilic Phase Separation

Poly(2-isopropyl-2-oxazoline)-b-poly(lactide) (PiPOx-b-PLA) diblock copolymers comprise two miscible blocks: the hydrophilic and thermosensitive PiPOx and the hydrophobic PLA, a biocompatible and biodegradable polyester. They self-assemble in water, forming stable dispersions of nanoparticles with hydrodynamic radii (Rh) ranging from ∼18 to 60 nm, depending on their molar mass, the relative size of the two blocks, and the configuration of the lactide unit. Evidence from 1H nuclear magnetic resonance spectroscopy, light scattering, small-angle neutron scattering, and cryo-transmission electron microscopy indicates that the nanoparticles do not adopt the typical core–shell morphology. Aqueous nanoparticle dispersions heated from 20 to 80 °C were monitored by turbidimetry and microcalorimetry. Nanoparticles of copolymers containing a poly(dl-lactide) block coagulated irreversibly upon heating to 50 °C, forming particles of various shapes (Rh ∼ 200–500 nm). Dispersions of PiPOx-b-poly(l-lactide) coagulated to a lesser extent or remained stable upon heating. From the entire experimental evidence, we conclude that PiPOx-b-PLA nanoparticles consist of a core of PLA/PiPOx chains associated via dipole–dipole interactions of the PLA and PiPOx carbonyl groups. The core is surrounded by tethered PiPOx loops and tails responsible for the colloidal stability of the nanoparticles in water. While the core of all nanoparticles studied contains associated PiPOx and PLA blocks, fine details of the nanoparticles morphology vary predictably with the size and composition of the copolymers, yielding particles of distinctive thermosensitivity in aqueous dispersions.

where n 0 is the refractive index of the solvent, λ the wavelength of the laser, and θ the scattering angle. Γ is obtained from a second order cumulant expansion of the first-order electric field correlation function of the scattered electric field g 1 (t). Particle size and dispersity were analyzed also by the CONTIN algorithm (see SI4).
The averaged intensity of scattered light I sample at θ was measured by SLS over 3 min, the accumulation time in the DLS experiment, and used to calculate the Rayleigh ratio of the sample R Θ sample . 32 (Equation 4) = ( -) where I solvent and I tol are the scattered light intensities of the pure solvent and toluene at 90°, respectively, and R tol the Rayleigh ratio of toluene. For a precise extrapolation of R Θ sample to Θ = 0°, measurements were conducted at 11 angles between 50° and 150°. The resulting form factor P(Θ) was fitted to a Guinier model to obtain the radius of gyration R g .
Small angle neutron scattering: The absolute value of the wave-vector, is given by: where θ is the scattering angle. The samples were kept in 2 mm quartz cuvettes, equipped with stoppers. The measuring cells were placed onto a copper base for good thermal contact before mounting in the sample chamber. Samples were kept at a temperature of 25 °C. The transmission was measured separately. The absolute scattering cross section (cm -1 ) was calculated by taking into account the contribution from the empty cell and the general background. The samples were prepared in D 2 O instead of H 2 O to enhance contrast and reduce incoherent background.
The size of non-interacting moieties was determined in terms of the radius of gyration (R g ) at small q-values through the Guinier expression for the scattered intensity (I(q)): This approximation is valid only for qRg < 1. Since the Guinier approximation yielded quite coarse fits of the data, we also made full model fits of the scattering curve: where N is the number of particles inside a sample volume V, P(q) the particle form factor, and S(q) the particle structure factor. Since the particle dispersions used here are quite dilute we can assume non-interacting particles and for this case S(q)=1. The form factor of homogenous spherical objects with a radius R can be written as: where  =  p - 0 is the difference in scattering length density (SLD) between particle and solvent (scattering contrast) and Vp is the particle volume. In the fitting of the SANS data, we generally found this model adequate for the samples measured before heating. However, in one case it was necessary to employ the spherical core-shell model given by where V s is the total particle volume (including the shell), V c is the volume of the core, R s is the radius of the shell, and R c is the radius of the core, thus Rs = Rc + d, where d is the thickness of the shell. The parameter  c is the scattering length density of the core and  s is the scattering length density of the shell. j 1 (x) = (sin x -x cos x)/ x 2 , and scale is a scale factor proportional to the sample concentration.
Some of the samples were best described by a cylinder model after heating. The core-shell cylinder model for overall random particle orientation has a form factor P(q) described by: Here, α is the angle between the axis of the cylinder and the q-vector, V c and V s are the core volume and total volume, as before. Thus, for the cylinder V c =  R c 2 L and V s =  R s 2 L, where L is the length of the core, R c is the radius of the core, and t is the thickness of the shell (R s = R c + t). The total length of the outer shell is given by L+2t. J 1 is the first order Bessel function, and The scattering length densities employed in this study are 1.73e-06 Å -2 for PLA 33 and 0.7e-06 Å -2 for PiPOx. The model fitting was implemented via the SasView analysis package (http://www.sasview.org/).

S2: Solubility of PLA in H 2 O/THF mixtures
In a series of vials, PLA homopolymers were dissolved in THF (1 wt%, 1 mL) and different amounts of water were added under stirring. The onset of turbidity was detected visually (Table S2-1).
Two criteria define the hydrophobicity of each PLA. The first is the minimal water volume fraction at which each PLA instantaneously turns turbid (marked red). According to this PLLA3 is the most and PDLLA1 the least hydrophobic polymer of the selection. The second criterion is the duration until precipitation occurs at a constant water volume fraction. For example, at 20 vol% water, all PDLLA homopolymers remain soluble for 40 days but all PLLA homopolymers precipitate in finite time. Both criteria taken together, the hydrophobicity increases in the order: PDLLA1 < PDLLA2 < PDLLA3 < PLLA1 < PLLA2 < PLLA3.

S3: Particle stability at room temperature
The particle stability at room temperature (c BCP : 0.5 g/L) was investigated by light scattering. The samples (0.5 g/L) were placed in cuvettes and not agitated for the duration of the stability test. Figure   S2-1 shows size information (z-average) and derived count rate normalized to the respective quantity at day one. Particles of 2L3 were stable for 20 days, neither size nor count rate changed considerably.  q 2 / 10 -10 cm -1 Figure S4-14 3DL3, at 20 °C, before annealing. Selected normalized autocorrelation function of scattered light intensity |g 2 (t)| vs time t, ln[g 1 (t)] vs tq 2 , and CONTIN distributions of 3DL3 at scattering angles of 60 ° (squares, black line), 90 ° (circles, red line), and 120 ° (triangles, blue line). Natural logarithm of the form factor P(q) vs q 2 and second order fit. q 2 / 10 -10 cm -1 Figure S4-15 3DL3, at 20 °C, after annealing 50 °C for 2 h. Selected normalized autocorrelation function of scattered light intensity |g 2 (t)| vs time t, ln[g 1 (t)] vs tq 2 , and CONTIN distributions at scattering angles of 60 ° (squares, black line), 90 ° (circles, red line), and 120 ° (triangles, blue line). Natural logarithm of the form factor P(q) vs q 2 . Guinier fit not appropriate due to bimodality.