Ultra-thin self-healing vitrimer coatings for durable hydrophobicity

Durable hydrophobic materials have attracted considerable interest in the last century. Currently, the most popular strategy to achieve hydrophobic coating durability is through the combination of a perfluoro-compound with a mechanically robust matrix to form a composite for coating protection. The matrix structure is typically large (thicker than 10 μm), difficult to scale to arbitrary materials, and incompatible with applications requiring nanoscale thickness such as heat transfer, water harvesting, and desalination. Here, we demonstrate durable hydrophobicity and superhydrophobicity with nanoscale-thick, perfluorinated compound-free polydimethylsiloxane vitrimers that are self-healing due to the exchange of network strands. The polydimethylsiloxane vitrimer thin film maintains excellent hydrophobicity and optical transparency after scratching, cutting, and indenting. We show that the polydimethylsiloxane vitrimer thin film can be deposited through scalable dip-coating on a variety of substrates. In contrast to previous work achieving thick durable hydrophobic coatings by passively stacking protective structures, this work presents a pathway to achieving ultra-thin (thinner than 100 nm) durable hydrophobic films.


Supplementary Figures
Supplementary Figure 1. Apparent advancing and receding contact angles of deionized water droplets deposited on dyn-PDMS films. The films were deposited on a polished silicon wafer, aluminum tab, copper tab, and glass slide using spin-coating and dip-coating. The uncertainty used for each value represents the standard deviation of three spatially distinct measurements on each individual sample.
Supplementary Figure 2. Bright field optical microscopy of water vapor condensation on 1.5:1 dyn-PDMS coatings. The films were fabricated using (a) spin-coating and (b) dip-coating. The substrates were a (i) polished Si wafer, (ii) aluminum tab, (iii) copper tab, and (iv) glass slide. Gravity points into the page for each image. Supplementary Figure 5. Atomic force microscopy (AFM) images of different films deposited on different substrates using spin-coating or dip-coating. The sample information is labeled at the lower-left corner of each image. The scale bars in all figures represents 1 μm. Inset color bar: surface height from low (black) to high (white). Dip-coating samples are generally thicker than spin-coated samples. This is because that for dip coating, the amount of materials on the substrate prior to thermal annealing is more than the ones of spin-coated samples, hence the thermal annealing recipe that can remove all excessive material for the spin-coated samples cannot guarantee the removal of all excessive materials for the dip-coated samples.
Supplementary Figure 6. Surface energy measurements. Side-view optical microscopy images of a deionized water droplet and a diiodomethane droplet residing in the apparent advancing state on the (a) dyn-PDMS (1.5:1), and (b) dyn-PDMS (1.7:1) coated Si-wafer samples, respectively. Each image is labeled with its apparent advancing contact angle, a. All scale bars represent 1 mm. The uncertainty of the contact angle measurements are determined as the standard deviation of 3 independent measurements at 3 different locations on one single sample. Figure 7. Chemical robustness of (a) 10 nm 1.5:1 dyn-PDMS and (b) 10 nm 1.7:1 dyn-PDMS samples. Samples were immersed in different liquid environment including IPA, 0.6 mol·L -1 NaCl solution, 0.1 mol·L -1 H2SO4 solution, and distilled water. The uncertainty of the contact angle measurements are determined as the standard deviation of 3 independent measurements at 3 different locations on one sample.  Supplementary Figure 4 and 5. The roughness here is defined as the maximum peak-to-peak RMS roughness of the AFM scan. Roughness ratio is defined as the total AFM scan surface area normalized by the AFM scan projected area multiplied by 100%.

Supplementary Discussions
Supplementary Discussion 1. CFx film degradation The irregular objects observed in Figure 3a of the manuscript are 'water blisters'. These blisters also look like condensate films forming on a hydrophilic surface. To show that the irregular objects are indeed water blisters instead of condensate films, we coupled our DSLR camera with a 10X objective (CFI Plan 10X, Nikon). We show the microscopic images of the water blisters in

Supplementary Method 1. Sample thickness measurement by IR-AFM
To ensure the existence of the 10-nm 1.5:1 dyn-PDMS on the polished silicon wafer, we performed an AFM step measurement. The substrate we prepared was a 3 cm × 3 cm polished silicon wafer.
The wafer was first cleaned by rinsing in acetone, IPA, water and IPA in sequence, then dried with a clean N2 gas stream. Then, Kapton tape (Purchased from Uline) was partially stuck on to the edge of the silicon wafer (as illustrated in Supplemental Figure 8a). The sample was then further purified through air plasma cleaning (Harrick Plasma, PDC-32G) for ten minutes at high power (RF, 18 W). The 10 nm dyn-PDMS film was then deposited on the cleaned substrate by the spincoating and annealing process. which yielded a coating thickness of 4 ± 3 nm. It should be noted that this thickness measurement method has a relatively high noise/signal ratio (80%) compared to ellipsometeric measurements (1%) due to the difficulty in obtain clean cuts, which can be affected by the flow of the materials and residuals after Kapton tape peeling.

Supplementary Method 2. Surface energy measurements
The surface energies of a flat surface (s) can be measured by the contact angle approach based on the Fowkes model. The model assumes that the surface energy has two components, one is dispersive surface energy γ , and another is polar component γ , . 1 The two components of surface energy can be determined by measuring the intrinsic advancing contact angle of water (θ ) and diiodomethane (θ ). All apparent contact angles were measured on at least 3 spatially distinct spots on the surface using a piezoelectric micro-goniometer (MCA-3, Kyowa), 2 with measurement result summarized in Supplementary Figure 1, along with the uncertainty as defined by the standard deviation of the measurement results. The maximum base radius of the sessile droplets was controlled such that it is smaller than the capillary length (approximately 3 mm) of both water and diiodomethane to ensure the contact angle is not affected by gravity. 2 The dispersive surface energy, , , is measured using diiodomethane (Sigma Aldrich, ReagentPlus, 99%. , = 50.8 mJ·m -2 , , = 0 mJ·m -2 at 20 ℃) as the probing liquid: Then, di-ionized (DI) water ( , = 21.8 mJ·m -2 , , = 51.0 mJ·m -2 at 20 ℃) was used to obtain , : Contact angle measurements were performed on at least three different spots on each surface, and the uncertainties of contact angle ∆ and ∆ are defined as the standard deviation of the different measurements.
The uncertainty of , is calculated by: And the uncertainty of , is calculated by:

Supplementary Method 3. Evaluating the chemical robustness of dyn-PDMS thin films
The chemical robustness of the 10 nm thick dyn-PDMS films (both 1.5:1 and 1.7:1) were empirically tested by measuring the change in apparent advancing contact angle and contact angle hysteresis of deionized-water droplets deposited on samples that have been immerses into different liquid environments (100 mL