marine foulants ( 65 ). This is typically due to
the increase in heterogeneous foulant nucle-
ation sites, as well as mechanical locking be-
tween the impregnated texture and the solid
foulant. However, for some specific applica-
tions, carefully tailored surface textural fea-
tures have been shown to decrease the contact
area between a surface and a microscopic foul-
ant (such as bacteria or algal spores; Fig. 4, B
and C), reducing their surface accretion ( 65 ).
For example, nanoscale pillars inspired by the
structure of cicada wings have demonstrated
bactericidal properties, as these sharp structures
can penetrate the bacterial cell wall, causing
death ( 66 ). Similarly, sharklet patterns inspired
by shark placoids have been shown to deter the
settlement of different marine foulants such as
the ubiquitous algaeUlva( 67 ).When the porous texture of a lowÐsurface
energy solid is infiltrated with lubricants,^tice
can be lowered by more than an order of mag-
nitude as compared to unmodified control
surfaces (Fig. 5A). In this case, ice no longer
contacts the solid but instead rests on a mo-
lecularly smooth, slippery, liquid layer ( 9 , 35 ).
Lubricated surfaces have shown reduced foul-
ing against a wide range of other hard and softDhyaniet al.,Science 373 , eaba5010 (2021) 16 July 2021 7 of 13
Fig. 5. State-of-the-art ice-shedding and
anti-icing surfaces.(A) Compilation of
apparent ice adhesion values reported in
the literature for different material classes,
testing methodogies, and temperatures.
(B) Compilation of apparent ice adhesion
values reported in the literature plotted
against tested area. Data for (A) and (B) are
compiled from ( 9 , 13 , 15 , 16 , 58 , 59 ,
61 – 63 , 72 – 77 , 79 , 84 – 86 , 89 – 91 , 182 – 184 ).
(C) Collection of individual sessile water
droplet freezing times over surfaces reported
in the literature plotted against water contact
angle, at a given temperature. Circled data
represents surfaces that engender enthalpic
interactions with water. Data compiled
from ( 95 – 105 ). (D) Collection of areal icing
frost coverage on different material categories
reported in the literature plotted against time,
at a given temperature. All data start at
the origin. Timet= 0 may be defined as when
the sample reaches a set temperature
or when the sample starts to cool (this was
variable in each study). Bottom left inset
shows frost-free zones over concave features
of a 3D-printed artificial leaf adapted from
( 115 ); scale bar, 15 mm. Upper right inset
shows frost propagation from interdrop
ice-bridging; adapted with permission from
( 47 ), copyright the American Chemical Society
(2013); scale bar, 40mm. White and red areas
in the image show unfrozen and frozen
droplets, respectively. The numbers depict
the order of ice nucleation, and the yellow
arrows show ice-bridging. Data compiled from
( 9 , 47 , 75 , 78 , 79 , 115 – 118 ).
RESEARCH | REVIEW