Microgels

Microgels were created using an overhead spinning method^{15 (link)}. Briefly, the aqueous pre-gel solution was infused into spinning mineral oil at 2.5 mL/hr with a syringe pump. The oil was supplemented with 1% SPAN-80 and triethylamine (20 µL/mL of gel) and spinning at 350 rpm. Following microgel synthesis, microgels were purified with mineral oil and PBS, sterilized with 70% IPA, and transitioned to sterile PBS in a biosafety cabinet^{15 (link)}. Microgels were synthesized with Alexa Fluor 488 maleimide to allow for fluorescent identification and imaged using a Molecular Devices ImageXpress. Particle sizing was calculated custom thresholding module (ImageXpress)^{15 (link)}.

To reproduce
the behavior of the microgel close to a surface, we model the latter
as two layers of wall particles that are initially located on a compact
square lattice of size σ at the bottom of the simulation box;
the separation between layers is 0.7σ. To avoid crystallization
of the monomers close to the surface, the wall particles are randomly
displaced from the lattice sites, including the direction perpendicular
to the plane, following a Gaussian distribution with standard deviation
σ_sd = 0.2. The obtained layers are then subsequently
fixed throughout the whole simulation runs.
Microgel monomers
interact with wall particles via the WCA potential (eq 1) and the V_αms potential, which is identical with eq 3, but this time replacing
the monomer–monomer attraction α_mm with the
monomer–surface attraction, α_ms, now encoding
the surface hydrophobicity.
To mimic experimental conditions
where the microgel is physically
anchored to the wall, we also consider the case where permanent bonds
between a few monomers and the surface particles are formed. This
is obtained by the following procedure (illustrated in Figure 7 for the hydrophilic scenario
α_ms = 0): (i) a swollen microgel (equilibrated in
bulk at α_mm = 0) is pushed toward the wall; (ii)
when it comes in contact with the surface, monomers with distance
less than dz = 1.5σ from the upper layer of
the wall are considered, and among them; (iii) b monomers
are randomly chosen and anchored to their closest wall-particle via
the harmonic potential V(r) = K(r – r₀)² with K = 15 and r₀ = 2^1/6σ; finally (iv) the microgel
is left to relax to its equilibrium state. The procedure is then repeated
for different surface α_ms conditions. We did it for
both the hydrophilic α_ms = 0 and hydrophobic α_ms = 0.9 conditions, yielding different bonding patterns. Density
profiles calculated with microgels anchored to a hydrophilic surface
are comparable to experiments in all cases, except for the measurements
at T = 35 °C close to a hydrophobic surface,
where we found that the extension of the tail is better captured by
simulations of microgels initially anchored to a hydrophobic surface
(see Figure S15). Another important parameter
to take into account is the number of bonds b that
we should consider in the simulations. As mentioned in Results and shown in Figure S11, for the hydrophilic surface we tried different values of b and found that b = 25 is the most similar
to experimental data. For the hydrophobic case, we expect in experiments
a much larger number of bonds due to the additional attraction to
the surface in the anchoring procedure. For this reason, we performed
all simulations (that take much longer, also due to the long aging
regime) with a fixed value b = 200, roughly 1 order
of magnitude difference with respect to the hydrophilic case. This
value was then found to be in rather good agreement with experiments
in terms of the tail of the density profiles.