PerspectiveMaterials Science

Improving surface-wetting characterization

See allHide authors and affiliations

Science  15 Mar 2019:
Vol. 363, Issue 6432, pp. 1147-1148
DOI: 10.1126/science.aav5388

Highly hydrophobic surfaces have numerous useful properties; for example, they can shed water, be self-cleaning, and prevent fogging (1, 2). Surface hydrophobicity is generally characterized with contact angle (CA) goniometry. With a history of more than 200 years (3), the measurement of CAs was and still is considered the gold standard in wettability characterization, serving to benchmark surfaces across the entire wettability spectrum from superhydrophilic (CA of 0°) to superhydrophobic (CA of 150° to 180°). However, apart from a few reports [e.g., (48)], the inherent measurement inaccuracy of the CA goniometer has been largely overlooked by its users. The development of next-generation liquid-repellent coatings depends on raising awareness of the limitations of CA measurements and adopting more sensitive methods that measure forces.

CAs reveal the equilibrium states of droplets deposited on surfaces. However, each surface has a range of metastable CAs, and a static CA has a random value in this range. Thus, measuring the minimum and maximum values of the range, termed the receding and advancing contact angles (RCA and ACA), is recommended. This is done by decreasing (RCA) or increasing (ACA) the droplet volume with a needle until the contact line starts to recede or advance on the surface (9). RCA and ACA are often confusingly called “dynamic CAs,” although this terminology should be avoided because the measurements are performed slowly and quasi-statically. Droplet mobility is related to the difference between ACA and RCA, called contact angle hysteresis (CAH). In addition to CAH, mobility is often quantified by tilting the surface below the test droplet until the sliding angle is reached and the droplet begins to move. However, sliding-angle measurements are sensitive to details of the experiment, such as droplet volume and how the droplet was placed on the surface, which can make comparison of results challenging (10).

Tricky contact angle imaging

A simple method for determining surface wetting, measuring contact angles (CAs) of water droplets, can be misleading for superhydrophobic surfaces because of difficulties in positioning the baseline (for more details, see supplementary materials).


Despite being useful quantities in wetting characterization, CAs suffer from practical limitations. The results obtained by independent scientists can vary by up to 10° even with the same setup, especially for CAs exceeding 150° (4, 5), which makes meaningful comparisons almost impossible.

All measurements of CA involve taking a profile image of the droplet followed by image analysis (49). The inaccuracies mainly originate from optical distortions and are affected by experimental parameters such as magnification, lighting, contrast, and camera resolution. The optical distortions are large near the baseline (i.e., the boundary between the solid surface and the liquid droplet in the two-dimensional image; see the figure, top). Not only is the droplet edge diffuse, but it also becomes heavily pixelated, even when a goniometer with a high-resolution camera is used. The diffuse edge and pixelation necessarily introduce a substantial systematic error in CA from about 1° to beyond 10° because of the uncertainty in baseline placement, which becomes subjective. Even the automatic baseline detection feature in goniometer software often fails, likely because of the short baseline length on highly hydrophobic surfaces. Despite the continuous improvement of experimental procedures (9, 11) and analysis methods (68) of CA goniometry, these problems still persist.

The errors in CA resulting from one-pixel displacement of the baseline are shown in the bottom panel of the figure. Simulations (5) and experiments match and demonstrate how the error increases substantially for increasing CA, especially upon reaching the superhydrophobic regime. The uncertainty range in CA corresponds to ∼1° for CAs less than 120°, ∼2° for CAs of ∼150°, and ∼5° for CAs of ∼162°. Propagation of errors in subtraction make the uncertainty of CAH even worse, up to 2 times greater than for single CA values. Droplet reflection (as depicted in the figure, top) can somewhat facilitate the baseline determination, but only for reflective surfaces. Moreover, macroscopically rough surfaces, such as woven textiles, have an irregular baseline and the contact angle is ill-defined.

Droplet shedding and sliding on repellent surfaces are governed by adhesion and friction forces, which are related to contact angles: F ∼ cos RCA – cos ACA. When CAH is low, even a small error in baseline placement causes huge relative errors in CAH and in the calculated adhesion force. For example, for a superhydrophobic surface with RCA of 170°, an error of just one pixel in the baseline height can result in at least 300% error in CAH and adhesion force. Thus, CAH is in practice poorly suited for characterization of highly repellent surfaces. Even with improved optics, such as enhanced camera resolution, the error near the upper CA limit will remain high.

We propose the use of force measurements to characterize hydrophobic surfaces. The classic Wilhelmy plate technique is limited by strict constraints on sample geometry and gives no information about local wetting properties (12). However, it is now possible to detect tiny forces between droplets and surfaces. Measuring the deflection of a thin capillary inserted in a droplet (13) can determine droplet friction. The oscillating droplet tribometer uses back-and-forth motion of a magnetic water droplet to measure droplet friction forces down to 10 nN (14). Scanning droplet adhesion microscopy can measure adhesion forces as small as 5 nN and map wetting properties at microscale spatial resolution (15). These newer methods offer more accurate wetting measurements, especially on highly hydrophobic surfaces.

Supplementary Materials

Supplementary Text

Table S1

Fig. S1

Numerical Modeling (Simulations)

References and Notes

Acknowledgments: Supported by European Research Council grant ERC-2016-CoG (725513-SuperRepel) and the Academy of Finland Centres of Excellence Programme 2014–2019. The source codes and functions used in this work are available at Zenodo (16).

Navigate This Article