Poly(vinyl alcohol) Molecular Bottlebrushes Nucleate Ice

Ice binding proteins (IBP) have evolved to limit the growth of ice but also to promote ice formation by ice-nucleating proteins (INPs). IBPs, which modulate these seemingly distinct processes, often have high sequence similarities, and molecular size/assembly is hypothesized to be a crucial determinant. There are only a few synthetic materials that reproduce INP function, and rational design of ice nucleators has not been achieved due to outstanding questions about the mechanisms of ice binding. Poly(vinyl alcohol) (PVA) is a water-soluble synthetic polymer well known to effectively block ice recrystallization, by binding to ice. Here, we report the synthesis of a polymeric ice nucleator, which mimics the dense assembly of IBPs, using confined ice-binding polymers in a high-molar-mass molecular bottlebrush. Poly(vinyl alcohol)-based molecular bottlebrushes with different side-chain densities were synthesized via a combination of ring-opening metathesis polymerization (ROMP) and reversible addition–fragmentation chain-transfer (RAFT) polymerization, using “grafting-to” and “grafting-through” approaches. The facile preparation of the PVA bottlebrushes was performed via selective hydrolysis of the acetate of the poly(vinyl acetate) (PVAc) side chains of the PVAc bottlebrush precursors. Ice-binding polymer side-chain density was shown to be crucial for nucleation activity, with less dense brushes resulting in colder nucleation than denser brushes. This bio-inspired approach provides a synthetic framework for probing heterogeneous ice nucleation and a route toward defined synthetic nucleators for biotechnological applications.

Size Exclusion Chromatography in DMF. Size exclusion chromatography (SEC) analysis of graft and bottlebrush polymers was performed on an Agilent Infinity 1260 MDS instrument equipped with differential refractive index (DRI), viscometry (VS), dual angle light scatter (LS) and variable S3 wavelength UV detectors. The system was equipped with 2 x PLgel Mixed D columns (300 x 7.5 mm) and a PLgel 10 µm guard column. The mobile phase used was DMF (HPLC grade) containing 0.1% LiBr at 50 o C at flow rate of 1.0 mL.min -1 . Poly(methyl methacrylate) (PMMA) standards (Agilent EasyVials) were used for calibration between 2,200,000 -550 g.mol -1 . Analyte samples were filtered through a nylon membrane with 0.22 μm pore size before injection. Number average molecular weights (Mn), weight average molecular weights (Mw) and dispersities (ĐM = Mw/Mn) were determined by conventional calibration and universal calibration using Agilent GPC/SEC software.
Size exclusion chromatography (SEC) analysis of poly(vinyl acetate) (PVAc) and poly(norbornene) (PNB-NBoc) homopolymer precursors was performed on an Agilent Infinity II MDS instrument equipped with differential refractive index (DRI), viscometry (VS), dual angle light scatter (LS) and variable wavelength UV detectors. The system was equipped with 2 x PLgel Mixed D columns (300 x 7.5 mm) and a PLgel 5 µm guard column. The mobile phase used was DMF (HPLC grade) containing 5 mM NH4BF4 at 50 o C at flow rate of 1.0 mL.min -1 . Poly(methyl methacrylate) (PMMA) standards (Agilent EasyVials) were used for calibration between 955,000 -550 g.mol -1 . Analyte samples were filtered through a nylon membrane with 0.22 μm pore size before injection. Number average molecular weights (Mn), weight average molecular weights (Mw) and dispersities (ĐM = Mw/Mn) were determined by conventional calibration and universal calibration using Agilent GPC/SEC software.
FTIR Spectroscopy. Fourier Transform-Infrared (FTIR) spectroscopy measurements were carried out using an Agilent Cary 630 FT-IR spectrometer, in the range of 650 to 4000 cm -1 .
Turbidimetry. Turbidimetric analysis was performed an Agilent Cary 60 UV-vis spectrophotometer equipped with a Peltier heating and cooling system. Aqueous solutions of PNBn-g-PVA208 (n = 20, 40) bottlebrush polymer samples were prepared at 1 mg.mL -1 with changes in transmittance monitored at λ = 700 nm by heating each sample from 20 °C to 85 °C at a rate of 5 °C.min -1 . The inflection point of each thermal phase transition curve was used to determine the lower critical solution temperature (LCST) in each case.

S4
Transmission Electron Microscopy. Dry-state stained TEM imaging was performed on either a JEOL JEM-2100 or a JEOL JEM-2100Plus microscope operating at an acceleration voltage of 200 kV. All dry-state samples were diluted with MilliQ water and then deposited onto formvar-coated copper grids.
After roughly 1 min, excess sample was blotted from the grid and the grid was stained with an aqueous 1 wt% uranyl acetate (UA) solution for 1 min prior to blotting, drying and microscopic analysis.
Atomic Force Microscopy. (AFM) imaging was performed on a Bruker Icon AFM microscope. Samples were drop cast from 1 mg.mL -1 aqueous solutions onto a silicon wafer. The substrate was dried under a gentle N2(g) flow prior to imaging. The tips for the AFM analysis (Bruker ScanAsyst-air) were purchased from BRUKER, with resonance frequency in the range of 45-95 kHz and spring constant in the range of 0.2-0.8 N/m. Acquired images taken in peak force tapping mode and were analyzed using the Gwyddion software.

Experimental Procedures
Splat Ice Recrystallization Inhibition Assay. Splat cooling assays were performed as previously described by Tomczak et al. [1] Briefly, a 10 μL sample was dropped 1.40 m onto a chilled glass coverslip, resting on a thin aluminium block cooled to -78 °C placed on dry ice. Upon hitting the coverslip, a wafer with diameter of approximately 10 mm and thickness 10 μm was formed instantaneously. The glass coverslip was transferred onto the Linkam cryostage and held at -8 °C using liquid nitrogen for 30 minutes. Photographs were obtained using an Olympus CX 41 microscope with a UIS-2 20x/0.45/∞/0-2/FN22 lens and crossed polarisers (Olympus Ltd), equipped with a Canon DSLR 500D digital camera. Images were taken of the initial wafer (to ensure that a polycrystalline sample had been obtained) and again after 30 minutes. Image processing was conducted using ImageJ. In brief, the number of ice crystals in the field of view was measured for each photograph. The average (mean) of these three measurements was then calculated to find the mean grain area (MGS). The average value and error were compared to that of PBS solution, as appropriate, as a negative control. Bottlebrush polymer sample solutions were first prepared upon dissolution in PBS on the range of [sample] = 0.5 -0.008 mg.mL -1 .
Modified Sucrose Sandwich Ice Shaping Assay. Briefly, 1 mg.mL -1 of nanoparticle samples were dispersed in 45 wt % sucrose solution and sandwiched between two glass coverslips and sealed with immersion oil. Samples were cooled to −50 °C on a Linkam Biological Cryostage BCS196 with T95-Linkpad system controller equipped with a LNP95-Liquid nitrogen cooling pump, using liquid nitrogen as the coolant (Linkam Scientific Instruments UK). The temperature was then increased to −8 °C and held for 1 h to anneal. The samples were then heated at 0.5 °C.min -1 until few ice crystals remained and then cooled at 0.05 °C.min -1 and the shape of ice crystals observed. Micrographs were obtained every S5 0.1 °C using an Olympus CX41 microscope equipped with a UIS-2 20x/0.45/∞/0−2/FN22 lens (Olympus Ltd.) and a Canon EOS 500D SLR digital. Image processing was conducted using ImageJ.
Ice Shaping using Nanoliter Osmometer. An Otago nanoliter osmometer (Otago Osmometers, Dunedin, New Zealand) was used to measure polymer ice shaping. Briefly, 20 nL droplets of 2 mg.mL -1 samples were suspended in type B immersion oil (Type B, cargille immersion oil) on a 6-well cooling plate using a microcapillary system. [2,3] The samples were rapidly frozen by cooling the osmometer to ~ −40 °C. A rapid temperature increase was then conducted until the melting point was reached, and then the temperature was gradually increased (0.01 °C.min -1 ) until only one ice crystal remained. Just before the ice crystal melted, the temperature was decreased until a discernible growth of the ice crystal was observed. It was observed and photographed with an Olympus CX41 microscope equipped with a UIS 20x/0.45/*/0-2/FN22 lens (Olympus Ltd.) and a Canon EOS 1200D digital SLR camera.

Ice Nucleation Measurements.
A microlitre scale droplet freezing assay was used to establish the ice nucleating effectiveness of the polymer nanoparticles and their precursors. Ice nucleation measurements were performed in essentially the manner previously described by Whale et al. Although using a different temperature control apparatus. [4]  a Sartorius Picus® electronic micropipette. The slides holding droplets were then placed on a temperature-controlled stage and cooled at a rate of 2 °C.min -1 . A video camera was used to monitor droplet freezing. A custom LabView program was used to directly link temperature measurements to video frames, allowing freezing temperatures of droplets to be determined and droplet fraction frozen curves to be constructed. The confidence intervals in Figure 3A were generated using a simple Monte-Carlo simulation. [5] Droplet fraction frozen curves were divided temperature bins with a width of 0.5°C.
The number of events in any bin is expected to follow a Poisson distribution on repeated testing so 5000 S6 Poisson distributed random numbers were generated for each bin, using the observed number of freezing events as the expectation value for the bin. The bars in Figure 3A indicate the interval in faction frozen in which 90% of the generated numbers fall. In experiments of this type, 'pure' water freezes several degrees warmer than would be expected for homogeneous ice nucleation, which occurs in the absence of any heterogeneous ice nucleating particles. Purification to the point that homogeneous nucleation occurs is generally very challenging and the reported background is typical for the size of droplets used. [6] Ice Nucleation Measurements by Differential Scanning Calorimetry (DSC). To assess the ice nucleation effectiveness of the various polymers in small droplets water-in-oil emulsions were frozen in a TA DSC 2500 instrument under nitrogen flow. The method employed is very similar to that described by Marcolli et al. [7] and employed in later studies such as that of Kumar et al. [8] The oil matrix used consisted of 10 wt% lanolin in paraffin oil. To make emulsions a custom-made homogenizer was used to disperse 10 wt% of water or 2 mg.ml -1 polymer solutions in oil matrix. This process resulted in emulsions of the type shown in Figure S38. The size distributions of the emulsions were analyzed using ImageJ. Figure S39 shows a typical size distribution of droplets in an emulsion. The droplet size distribution produced by this method peaked at around 5 µm and was very similar for all emulsions tested. To conduct an experiment 10 mg of emulsion was sealed in a DSC pan and placed in the DSC.
To stabilize the emulsions they were first frozen by cooling from ambient temperature to -50 °C at a rate of 10 °C.min -1 then warmed again to 20 °C to melt the water content of the emulsion. For measurement, the emulsions were cooled at 10 °C.min -1 to -20 °C then at 1 °C.min -1 to -50 °C. The DSC curves reported were generated during the 1 °C.min -1 cooling ramp.
In Figure S39 is can be seen that the DSC curve generated for pure water in our experiments is very similar to that reported by Marcolli et al. for pure water for a similar emulsion. It also can be seen in

Synthesis of exo-norbornene imide N-Boc
The synthesis of the N-Boc protected amino exo-norbornene imide was carried out according to a previously described process. [9,10] First, to a solution of cis-5-norbornene-exo-2,3-dicarboxylic anhydride (2.0 g, 12.1 mmol, 1 eq) in 100 mL of dry toluene was added N-Boc-ethylenediamine (2.34 g, 14.6 mmol, 1.2 eq) in 10 mL of dry toluene. The flask was fitted with a Dean-Stark trap and the reaction mixture was heated at reflux for 18 h. Upon cooling to room temperature, the reaction mixture was washed with 1 M HCl (3 × 100 mL) followed by sat. NaHCO3 (1 × 100 mL). The organic layer was dried over MgSO4, filtered, and concentrated to dryness under reduced pressure to afford a light brown solid as the pure product (2.62 g, 71%). 1
Afterwards, the combined organic phases were washed twice with saturated aq. NaHCO3 (2 × 100 mL), followed by saturated aq. NH4Cl (2 × 100 mL) and brine (100 mL). The organic phase was dried over MgSO4 and the solvent was removed in vacuo to yield an off-white solid. The solid was recrystallized from ethanol to afford the pure product as white crystals (4.21 g, 62%

(III) Synthesis of exo-5-norbornene-2-methylamine
Monomer of exo-5-norbornene-2-methylamine was synthesized according to previously reported process. [12] Exo-5-norbornenecarboxamide (2 g, 14.6 mmol, 1.0 eq) was dissolved in 50 mL of anhydrous THF. A lithium aluminum hydride solution (1.0M in THF, 22 mL, 0.83 g, 21.9 mmol, 1.5 eq) was added dropwise over 5 minutes, and the reaction was heated at reflux for 48 hours. After cooling in an ice water bath, the reaction was quenched by dropwise addition of aqueous 1 M NaOH until there was no more gas evolution upon addition, as determined visually. The aqueous solution was then