Science - USA (2021-07-16)

(Antfer) #1

cells, and bacteria (Fig. 4C) ( 110 , 111 ). These
hygro-responsive surfaces include hydrogels
(such as polyethylene glycol), brushes, mono-
layers, and zwitterionic materials, all of which
rely on the formation of a surface hydration
layer to reduce the adhesion of different foul-
ants ( 112 , 113 ).
There are several differences between liter-
ature reports on anti-icing surfaces in terms of
testing temperatures, droplet sizes, criteria for
identifying start and complete freezing times,
and cooling modes—anisotropic versus iso-
thermal. Broadly, Fig. 5C shows that for most
surfaces, loweringgSVincreases the droplet
freezing time, whereas the freezing time de-
creases with decreasing testing temperature.
Dynamic freezing experiments, such as imping-
ing supercooled droplets on a surface, highlight
the effectiveness of water-repellent hydropho-
bic or superhydrophobic surfaces ( 14 , 114 ).
However, such tests are likely not representa-
tive of many real-world icing conditions where
frost formation can occur ( 57 , 89 ). Thus, as for
low-ice adhesion surfaces, there is a need for
standardization of testing methodology used
to evaluate anti-icing surfaces.
Whereas freezing time and nucleation tem-
peratures are useful for characterizing the
freezing transition of isolated droplets, areal
coverage and propagation rates are perhaps a
better measure for comparing the overall frost-
ing or ice accretion rates on surfaces (Fig. 5D).
Frost can propagate from one condensed water
droplet to the next through interdroplet ice-
bridging. When the condensed droplets are
readily removed, or spatially constrained, frost
propagation can be slowed down substantially
( 11 , 47 , 115 ).
Recently, micro- and macroscale patterned
surfaces have been utilized for the spatial con-
trol of frosting ( 115 , 116 ). Yaoet al.( 115 ) showed
that frosting on natural leaves is spatially dis-
continuous, with frosting enhanced on the peaks
and suppressed in the valleys (Fig. 5D, inset).
This is due to the enhanced evaporation rate
for droplets that condense within the valleys.
Condensate droplet mobility can be increased
through the use of lubricated surfaces ( 9 , 75 ).
However, the easy shedding of droplets on such
surfaces requires an inclined surface, which
may not always be possible ( 9 , 117 ). In addi-
tion, as before, subjecting lubricated surfaces
to multiple frosting-defrosting cycles depletes
the lubricant, diminishing liquid and solid re-
pellency ( 75 , 78 ). In one study, the total frosting
time was reduced by almost 300% after only
10 frosting cycles, and another study showed
the complete depletion of the lubricant after
the second frosting cycle (Fig. 5D) ( 75 , 78 ).
Surfaces utilizing phase-change liquids (solid
at subzero temperatures) that undergo local-
ized melting from the latent heat released
during water condensation and freezing have
also been developed for retarding ice and frost


propagation ( 117 ). The use of amphiphilic and
charged surfaces for impeding frost propaga-
tion also offers promise ( 79 , 118 ).
Overall, analysis of the anti-solid fouling
landscape (Fig. 4, B and C) demonstrates that
large strides have been made toward combat-
ting the full range of solid foulants, but also
that major gaps remain. For example, no single
surface design strategy has yet been shown
to resist fouling by all the different possible
forms of ice and associated accretion length
scales. Additionally, practical application fre-
quently demands that a surface resist fouling
by multiple solid contaminants simultane-
ously. Consider that ships traveling the recent-
ly opened arctic routes need to resist fouling
by micro and macroscopic marine species ( 119 ),
as well as ice. However, most surface modifi-
cation technologies typically focus on reducing
the adhesion of only a single foulant, over a
narrow foulant accreting length scale.

Overlap between solid and liquid
fouling resistance
Real-world fouling environments are complex
and often involve multiple foulant phases that
must be simultaneously removed. There is
limited overlap between liquid-repellent and
solid-repellent surface design strategies. One
key element of this divide is surface texture.
Control over surface texture is a fundamental
tool in liquid-repellent surface design. How-
ever, the length scale of solid fouling initia-
tion and nucleation is typically much smaller
than the texture length scale utilized for liq-
uid repellency. As a result, the solid foulant
penetrates the solid texture, increasing the
adhesive bonding for a range of different solid
foulants including ice ( 78 ), scale ( 120 ), and dif-
ferent marine foulants ( 65 ).
The development of surfaces with ultralow
liquid contact angle hysteresis (Dq), as well as
interfacial slippage to facilitate the release of
solid foulants, shows promise in this regard.
The few surfaces that have been demonstrated
to repel both liquid and solid foulants have all
taken this approach and rely upon a molec-
ularly smooth, highly mobile interface. The
largest subset of these surfaces is made up of
various liquid-infused surface designs. How-
ever, such coatings are susceptible to mechani-
cal damage, and the lubricant may be readily
depleted, especially by abrasion ( 121 ) or con-
densation of low–surface tension liquids ( 37 , 49 ),
or under high shear flows ( 40 ). Alternatively,
as discussed above, some work has also been
reported in the area of covalently tethering
flexible molecular chains to a surface to mimic
the highly mobile interface of a liquid lubricant
with greater stability ( 39 , 122 ). These surfaces
demonstrate low–contact angle hysteresis with
a wide range of liquids ( 39 , 122 ), and lower solid
adhesion ( 38 ). However, much work still re-
mains to be done to develop effective surface

design strategies for simultaneously prevent-
ing solid and liquid accretion.

Challenges and outlook
Mechanical durability
The issue of mechanical durability has received
considerable attention in recent literature on
solid- and liquid-repellent surfaces ( 3 ). How-
ever, different studies typically use widely dif-
fering testing methodologies for evaluating
a surface’s durability, making it difficult to
compare performance between surfaces. Popu-
lar methods to evaluate mechanical durability
include linear and circular abrasion ( 63 , 123 ),
tape peeling ( 124 ), falling sand abrasion ( 125 ),
blade scratching ( 3 ), and water-jet impact
tests ( 124 ). Although some studies evaluate a
surface’s mechanical durability using ASTM
standardized Taber tests (Fig. 6A) ( 123 ), most
studies use a custom-made abrasion test setup
( 126 )wheretheabrasiveisasandpaperofa
specific grit. Overall, the differences in the
choice of abrader material, technique, abrasion
speed, applied load, and abrasion duration can
prevent a direct comparison between different
studies ( 127 , 128 ).
We attempt to provide a more universal ap-
proach to evaluate and compare the mechani-
cal durability of different surface coatings.
Archard’swearequation( 129 , 130 )iswidely
used to characterize wear resistance of mate-
rials and is given asQ¼KWS=Hs. Here,Qis
the wear volume,Wis the normal load applied,
Sis the total sliding distance of the abradant
over the softer test surface with hardnessHs,
andKis a dimensionless constant known as
the wear coefficient ( 129 , 130 ). The abradants
used in various reports selected for Fig. 6G,
whether sandpaper or Taber abrasers, are typ-
ically composed of abrasive materials like
SiC or Al 2 O 3 with Vickers hardness values of
~1800 to 2600 ( 130 ) (also see Fig. 6, B to F).
The constant,K, can be correlated with the size
of the wear particle produced by every asperity
contact ( 130 ). Additionally, the abrasive wear
rate is dependent on the size of the abrasive
particles for particle sizes <100mm( 130 ) and
consequently,Qºr, whereris the mean par-
ticle size of the hard abradant. As the data re-
ported in Fig. 6G use abrasive particles <100mm,
QºWSr, and thus, we useWSras our measure
of abrasion severity.
Surfaces that rely on topographical textures
to provide water repellency, such as several
different superhydrophobic surfaces, are often
susceptible to mechanical wear (Fig. 6G) ( 127 ).
In addition, poor chemical bonding and the
formation of wear debris can cause a loss of
water repellency. This can be observed by look-
ing at the data for textured metallic and steel
surfaces functionalized with small-molecule
silanes, long-chain fatty acids, or hydrophobized
nanoparticles on the surface (Fig. 6G) ( 131 ).
However, when a much harder material forms

Dhyaniet al.,Science 373 , eaba5010 (2021) 16 July 2021 9 of 13


RESEARCH | REVIEW

Free download pdf