Article

pubs.acs.org/Langmuir

Hierarchical Surface Architecture of Plants as an Inspiration for

Biomimetic Fog Collectors
M. A. K. Azad,*,† W. Barthlott,† and K. Koch‡
†
Nees Institute for Biodiversity of Plants, Rheinische Friedrich-Wilhelms-University of Bonn, Venusbergweg 22, 53115 Bonn,
Germany
‡
Faculty of Life Sciences, Rhine-Waal University of Applied Sciences, Marie-Curie Straße 1, 47533 Kleve, Germany
*
S Supporting Information

ABSTRACT: Fog collectors can enable us to alleviate the water crisis in certain arid regions of
the world. A continuous fog-collection cycle consisting of a persistent capture of fog droplets and
their fast transport to the target is a prerequisite for developing an efficient fog collector. In regard
to this topic, a biological superior design has been found in the hierarchical surface architecture of
barley (Hordeum vulgare) awns. We demonstrate here the highly wettable (advancing contact
angle 16° ± 2.7 and receding contact angle 9° ± 2.6) barbed (barb = conical structure) awn as a
model to develop optimized fog collectors with a high fog-capturing capability, an effective water
transport, and above all an efficient fog collection. We compare the fog-collection efficiency of the
model sample with other plant samples naturally grown in foggy habitats that are supposed to be
very efficient fog collectors. The model sample, consisting of dry hydrophilized awns (DH awns),
is found to be about twice as efficient (fog-collection rate 563.7 ± 23.2 μg/cm2 over 10 min) as
any other samples investigated under controlled experimental conditions. Finally, a design based
on the hierarchical surface architecture of the model sample is proposed for the development of
optimized biomimetic fog collectors.

■ INTRODUCTION
Shortages of fresh water are a major problem affecting around 1
billion people worldwide;1 the problem is most acute in arid,
semiarid tropical, and subtropical climates with limited or no
rainfall at all. Fog is an important source of water that is often
ignored. However, fog collection can be an elegant solution to
meet the demand of water for afforestation, for irrigation, and
above all as a drinking water source for humans and animals in
these areas.2−4 In the foggy regions, the dripping of water from
the foliage is one of most often seen phenomena. The
arrangement, density, type, and size of the foliage play an
important role in the amount of fog drip. Plants having “needle-
like” structures, e.g., pine, redwood, and fir, are reported to
collect a good amount of fog whereas the “leaflike”
impermeable structures (no regular void space on the surface)
allow the moisture to flow around the surface of the leaf.5−7
Increased branchiness of plant species, such as narrow leaf
syndrome, is reported to be an important trait to collect
fog.8−10 Tillandsia usneoides, Opuntia microdasys,11 and Cotula
fallax,12 all having hierarchical structures, have been demon-
strated to collect fog.9,13 Vogel and Müller-Doblies14 did an
extensive study on eight monocot families and some Oxalis
species in some semidesert regions (Namaqualand and adjacent
regions) of South Africa, where they demonstrated that these
plants did have “special morphological adaptations” of their
aerial parts that facilitate fog and dew collection. A comparison
of fog collection between a microstructured Namib grass
(Stipagrostis sabulicola) and a Namib beetle (Onymacris
unguicularis) demonstrated a better efficiency of the plant
species.15
In the last few decades the only technology for collecting fog
has been large fog collectors (LFCs) made of polyolefin mesh;
it has been used in 40 countries in South America, Europe,
Africa, and Asia where suitable regular fog is present.2,3,16 With
the aim to increase the amount of fog collection, a multifunnel
fog collector has been proposed.17 Although the simulations
showed that the multifunnel fog collector is more efficient than
the old design of the LFC, the design can be further improved
by using elements based on the hierarchical architectures of
plants. Some studies have shown the potential of artificial silk
fiber,18−24 electrospun nanofibers,25−27 and hydrophilic−
hydrophobic cooperative surfaces/systems28−30 to collect fog.
Inspired by the cactus spine,11 conical copper31 and zinc-
oxide32 wire or artificial spines27 and fog-collecting imperme-
able surfaces with cone arrays11,33−35 have been demonstrated.
Although the cone arrays were proved to be highly efficient for
a continuous fog collection, there were not enough indications
in these publications of how the structures can be implemented
to build a large fog collector, with one exception,11 where the
proposed prototype surface lacks void spaces that may result in
the hindrance to flow of the fog-carrying wind. The fog-carrying
wind has the tendency to deviate the direction while
approaching an impermeable obstacle. This is why fog
Received: July 1, 2015
Revised: November 11, 2015

© XXXX American Chemical Society A DOI: 10.1021/acs.langmuir.5b02430

Langmuir XXXX, XXX, XXX−XXX
Langmuir Article

collectors, or effective fog-collecting trees, have an open
structure with collecting surfaces a millimeter or less wide.
Therefore, we propose a meshlike structure as a viable solution.
In our previous study36 we demonstrated the positive influence
of microstructure and superhydrophilicity on fog collection. We
demonstrate here the barbed DH awns of Hordeum vulgare as a
model to develop optimized fog collectors based on a high fog-
capturing capability and an effective water transport. The
hierarchical architecture of the model sample also provides a
solution for the problem of the clogging of fog droplets
observed in an earlier study.37 To be noted, when the droplets,
attached to the fiber surfaces of a mesh or net, block the void
space between the fibers and consequently hinder the flow of
the fog-carrying wind, the phenomenon is defined as
“clogging”. A comparison of fog-collection efficiency of the
samples with and without barbs has been conducted here for
the first time. We also propose a possible design to develop
optimized fog collectors.
needle-like structures (NLSs) were arranged vertically in such a way
that there were gaps (width + 5−10% of the width of the awns/Ns/
NLSs) between the needles. The samples were prepared in such a way
that their effective surface areas remained close to each other. About
50−55% of the surface area of each prepared 2 × 2 cm2 sample was
occupied by the Ns/NLSs and the rest was free to get rid of the
impedance of the fog flow. Therefore, the samples have equivalent
effective surface areas for fog collection. The measurement of the
surface area coverage by the Ns/NLSs was calculated from the images
of the samples by Photoshop CS3. For the first set of samples, we used
intact Ns/NLSs, whereas for the second set, the needles of Abies and
Sequoia were cut to reach as close to the diameter of the Hordeum awn
as possible. The Tillandsia leaves (NLSs) of smaller diameter close to
Hordeum awn were also selected for the second set as well. Samples of
2 × 2 cm2 were prepared as previously described.
Hydrophilization. Dry Hordeum awns were dip-coated with a
water-based polymer solution containing TiO2 (titanium dioxide)
nanoparticles (TA 2202, Nadico Technologie GmbH, Langenfeld,
Germany), followed by drying at room temperature for ∼48 h. After

■
EXPERIMENTAL SECTION
Plant Material. Plant material includes fresh and dry awns of
Hordeum vulgare L. (the rougher surface of the awn was directed
toward the fog source), leaves (needle-like structures (NLSs)) of
Tillandsia usneoides (L.) L., and needles (Ns) of Pinus canariensis C.
Sm. (the round shape of the needle was directed toward the fog
source), Abies bracteata (D. Don) Poit. (the upper surface was directed
toward the fog source), Sequoia sempervirens (D. Don) Endl. (the
upper surface was directed toward the fog source), and Sequoiadendron
giganteum (Lindl.) J. Buchholz. A fresh and a DH awn (model species)
and other plant materials (Ns and NLSs) are shown in Figure S1a−g
in the Supporting Information. All plants were cultivated in the
Botanical Gardens of University of Bonn (BG-Bonn).
Sample Preparation. Samples (2 × 2 cm2) were prepared for
each of the model plant species (Figure 1a). The needles (Ns) or
Figure 1. Schematic of the (a) sample and (b) setup for the fog-
collection experiment. Adapted with permission from ref 36.
Copyright 2015 IOP Publishing.
evaporation of the solvent, the particles are bonded into a remaining
matrix that prevents the release of free TiO2 into the environment.
The average particle size of TiO2 is given as <8 nm (manufacturer’s
information). The thickness of the coating, measured by an atomic
force microscope (NanoWizard II, JPK Instruments, Germany), was
∼300 nm. In this article, we refer to hydrophilized dry Hordeum awns
as “DH awns” (D for dry and H for hydrophilized).
Contact Angle Measurement. Dynamic contact angle measure-
ments were carried out by a goniometer (DataPhysics OCA 20,
Filderstadt, Germany) at room temperature (21 °C). To perform the
measurements, the volume of a demineralized water droplet was
increased from 1 to 2 μL, followed by a decrease from 2 to 1 μL at a
rate of 0.1 μL/s with a delay time of 5 s. Because of the very low
surface area of the samples, we could only use 2 μL of water. We
calculated the contact angle hysteresis from the advancing and
receding contact angle values. The average and standard deviation of
the values were calculated from 5 measurements taken on each kind of
sample surface.
Light and Scanning Electron Microscopy. The analysis of the
fresh and/or dry sample surfaces was done by a Keyence VHX-1000
digital microscope (Keyence Corporation, Japan) and a Stereoscan
200 SEM (scanning electron microscope) (Cambridge, U.K.). Prior to
SEM analysis, the specimens were coated with gold for 30 s at 60 mA
(sputter coater, Balzers Union SCD040, BAL-TEC AG, Liechten-
stein).
Experimental Setup. Droplet deposition behavior on the surfaces
of the samples and their fog-collection efficiency were analyzed with a
setup previously used by Azad et al.36 and its extension. The samples,
which were mounted vertically in an artificial fog stream, had a
distance of 17 cm (Figure 1b) from the outlet of an ultrasonic
humidifier (Honeywell, BH-860E; fog output maximum 0.4 L/h). The
velocity of the fog flow was measured (by a Testo 416 anemometer,

Table 1. Structural characterization of the samples

sample structural characteristics

barbed Hordeum vulgare width of the awns: 708 ± 113 μm, barb (Figure 2a, b); length: 173 ± 26 μm with an apex angle (2α) of 22 ± 3°; tilted angle of the
awns (DH awn) barbs with the main axis of the awns: 37 ± 4°; distance between two barbs: 370 ± 106 μm on both sides and approximately twice in
the middle of the awn in some areas and irregular in some other areas
Hordeum vulgare width of the awns: 706 ± 108 μm; barb length: 163 ± 21 μm with an apex angle (2α) of 19.4 ± 2.9°; tilted angle of the barbs with the
(fresh awn) main axis of the awns: 31 ± 4°; distance between two barbs: 349 ± 107 μm on both sides and approximately twice in the middle of
the awn in some areas and irregular in some other areas
Ns Pinus canariensis width of the Ns: 913 ± 87 μm
Abies bracteata width of the Ns (intact): 1374 ± 100 μm; for the second set of samples, Ns were cut to a width of 717 ± 75 μm
Sequoia width of the Ns (intact): 2645 ± 175 μm; for the second set of samples, Ns were cut to a width of 814 ± 54 μm
sempervirens
Sequoiadendron Ns have conelike structures on the main axis; diameter of the Ns: 1074 ± 91 μm (see Figure S2 for details)
giganteum
NLSs Tillandsia diameter of the NLSs: 600−1800 μm; for the second set of samples, NLSs with diameter of 600−800 μm were used
usneoides

B DOI: 10.1021/acs.langmuir.5b02430
Langmuir XXXX, XXX, XXX−XXX
Langmuir Article

Figure 2. SEM of (a) a DH awn of Hordeum vulgare shows hierarchical surface architecture resulting from a combination of the microstructures (e.g.,
barbs and microgrooves) of different dimensions; (b) a single barb attached to the microgrooved main axis of the awn; (c) microgrooves of the awn
surface; waxy surfaces of the needles of (d) Abies bracteata and (e) Sequoiadendron giganteum.

Lenzkirch, Germany) to be ∼1.6 m/s at 17 cm from the outlet. The Surface Wettability. The presence of wax crystals on the
setup was in a chamber (length/width/height =100 × 90 × 180 cm), surfaces of Pinus, Abies, Sequoiadendron, etc. (Figure 2d−e) and
which is called a fog chamber. During the experiments, the a thin wax film39−41 on the surface of fresh Hordeum awn leads
temperature and the relative humidity inside the chamber were 19− to their hydrophobicity,42 as exhibited by the droplet behavior
20 °C and 75−85%, respectively. on the surface during the fog-collection experiment (see Figure
A D90 Nikon camera with a Medical-Nikkor 120 mm objective and
S3). The degree of wettability of the samples is shown in Table
a charge-coupled device (CCD) camera attached to the goniometer
were used for imaging.
2. The barbs were too tiny to place droplets on them to
Measurement of Fog-Collection Efficiency. The amount of
water collected by the samples over 10 min was measured to Table 2. Surface Wettability of All the Samples
determine their efficiency. Although the samples have an equivalent receding
effective surface area (50−55% of 2 × 2 cm2), a slight difference in advancing contact contact angle contact angle
their areas where the fog was coming into contact with them was also sample angle (θadv) (θrec) hysteresis (CAH)
taken into consideration during the calculation. A control experiment Hordeum vulgare 16 ± 2.7 9 ± 2.6 7 ± 1.8
(5 times each of 10 min) was also conducted without mounting the (DH awn)
samples. Hordeum vulgare 112 ± 2.9 84 ± 3.6 28 ± 1.2

■
(fresh awn)
Tillandsia 0 0
RESULTS usneoides
Structural Characterization of the Samples. Structural Pinus canariensis 114 ± 4 78 ± 5.6 31 ± 6.4
characteristics of the samples are presented in Table 1. The Abies bracteata 120 ± 2.8 94 ± 2.6 25.8 ± 2.4
awns of Hordeum vulgare (both DH and fresh) have barbs Sequoiadendron 115 ± 7 85 ± 3.9 29.5 ± 5.3
giganteum
(conical structure), the needles of Sequoiadendron giganteum Sequoia 94 ± 2.7 60 ± 4.5 33.5 ± 4.2
have barblike conical structures, and the other samples do not sempervirens
have barbs.
Surface Microstructures. The awns of Hordeum, fresh or measure the contact angle. Therefore, the contact angles for the
dry, have similar surface microstructures. A combination of the awn samples shown here were measured only on the surface of
microstructures of different dimensions result in the the main axis of the awns. However, we assume that barb
hierarchical architecture.38 They have barbs (Figure 2a, b) on surfaces will have a similar wettability to those on awn surfaces
both sides as well as along the midsection of the awns and have because of similar surface chemistry.
microgrooves on the surfaces of the awns (Figure 2b, c). Comparison of Fog-Collection Efficiency. Barbed
Moreover, SEMs in Figure 2b, c show that the thin coating samples were found to be more effective than the samples
(∼300 nm) of TiO2 does not fill the microgrooves (groove without barbs (Figure 3). Highly wettable (θadv = 16° ± 2.7 and
height = 3−3.6 μm). Wax crystals are visible on the θrec = 9° ± 2.6) DH awns showed a higher efficiency than
hydrophobic surfaces of the needles of Abis and Sequoiadendron hydrophobic fresh awns (θadv = 112° ± 2.9 and θrec = 84° ±
(Figure 2d, e). 3.6). The efficiency of superhydrophilic (contact angle = 0°)
C DOI: 10.1021/acs.langmuir.5b02430
Langmuir XXXX, XXX, XXX−XXX
Langmuir Article

Figure 3. Amount of water collected by 2 × 2 cm2 samples over 10

min (S* = Sequoiadendron). (In the first set of samples, intact Ns/
NLSs were used but the diameters/widths of Tillandsia, Abies, and
Sequoia were too large (Table 1) compared to that of the Hordeum
awns. In the second set the widths of the Ns of Abies and Sequoia were
reduced by cutting, and Tillandsia NLSs of smaller diameter were used.
Ns/NLSs in each 2 × 2 cm2 sample in both sets occupied similar
percentage of area and therefore an equivalent effective surface area for
fog collection.)

Tillandsia was higher than that of any other hydrophobic

samples (see contact angle in Table 2) and hydrophilic Sequoia.
Needles of Sequoiadendron and Sequoia showed the least
efficiency among the samples. DH awns of Hordeum were
found to be the best among the samples analyzed here. The
control experiment (without mounting the samples) showed
that 10 ± 4 μL of fog settled down in the container, which is
negligible.
Fog Collection by the Samples with Barbs. The
mechanism of fog collection on DH awns is shown cyclewise
in Figures 4 and 5; parts a−g in Figure 4 represent the first
cycle and parts h−l and m−o represent the next cycles of fog
droplet capture and transport, respectively. The droplets that
hit the surface, with the exception of the barbs of the awn
(DH), spread and/or followed the grooves. A second
mechanism occurred when droplets hit the barbs; in this
case, the droplets captured on the barbs moved from the tip to
the base. As a direct consequence of the process, an
enlargement of a droplet encompassing a small area of the
awn was observed (Figure 4c−f). The droplet pulled other
adjacent water droplets accumulated at the base of the barbs
toward itself.
With the increase of the amount of water in the droplet, it
moved downward (shown by the horizontal dotted lines in
Figure 4c−g, j−l, and n−o). The droplets moved to the bottom
of the awn, when large enough (shown by the vertical dotted
arrows in Figure 4g, l, o) after a time frame of between 4 and 25
s. The first cycle of fog collection on the DH awn, starting with
drop formation to falling downward (Figure 4a−g), took 15−
25 s depending on the width of the awns. The subsequent
cycles (shown in Figure 4h−l and m−o) took 4−10 s
depending on the width of the awns. The cycle continues.
Over a length of 2 cm, mostly in one place, rarely in two, the
droplet(s) was seen to form and there was obviously a droplet
at the tip (Figure 4p). A magnified image of a portion of Figure
4p is shown in Figure 4q to demonstrate the movement of fog
droplets from the tip to the base of the barbs.
Figure 5 describes the final collection of water from the
bottom of the awns. The droplet moved to the bottom of the
awn, and there the droplet increased in size (shown in the
D DOI: 10.1021/acs.langmuir.5b02430
Figure 4. Mechanism of fog droplet capture and transport by the DH
awns: capture, then coalescence, followed by the transport to the
target. The first cycle (a−g) takes about 15−25 s, and afterward every
cycle (h−l or m−o) takes 4−10 s, depending on the width of the awns.
From (c−f) the droplet is getting larger, and (g) shows the downward
movement of the droplet. (h−l) or (m−o) show the next similar
cycles; (p) two droplets increasing in size, one anywhere in the middle
and another at the bottom of the awn; the gap between two droplets
forming on the DH awn is also shown. (q) Magnified image of a
portion of (p) shows the movement of fog droplets from the tip to the
base of the barbs. Scale bar: (a−p) 650 μm, (q) 300 μm.
Figure 5. Size of the droplet at the bottom is increasing over time (a−
d). It pulls the water film on the awn. Hence, the fog droplets captured
by all the barbs above are moving downward with the film (shown by
the white arrows in b−d). Another drop anywhere from the middle of
the awn moving downward (shadow can be seen) shown by the arrow
(e); droplet about to start dripping shown by the arrow (f); scale bar 1
mm.
direction from Figure 5a−d) by pulling the droplets from the
nearby barbs and the surface of the awn, resulting in the
formation of a film on the DH awn. Simultaneously, other
larger droplets forming continuously anywhere in the middle of
the awn were also combining with the droplet at the bottom
(the shadow and the arrow in Figure 5e). Therefore, both the
water film flow and the droplets falling from anywhere in the
middle of the awn resulted in the enlargement of the droplet at
the bottom, as well as continuous dripping after about every
27−58 s from the bottom (Figure 5f).
Langmuir XXXX, XXX, XXX−XXX
Langmuir
The hydrophobic awns did not have an efficient capture of
fog droplets because of the saturation of the barbs due to the
accumulation of water at their bases (Figure 6a). However, in
Figure 6. Fog droplet behavior on a fresh hydrophobic awn of
Hordeum vulgare (a); hydrophobic needle of Abies bracteata (b). Fog
droplet behavior on the DH awn at different moments during fog
collection; after (c) 30 s, (d) 60 s, and (e) 90 s free barbs are seen on
both sides of the awns. Scale bar = 1 mm.
some instances it was observed that when a large droplet from
the top fell down it hit other droplets on the way, leading to a
cascade where the droplets further down the awn fell down
together. Only then can a new cycle start on the awns
(especially on the barbs). The accumulated water at the bases
of the barbs covered the barbs (Figure 6a) and reduced the fog
droplet capture by those barbs. For convenience, we name the
four larger droplets on the awn/needle (Figure 6a, b) d1, d2, d3,
and d4 from the top to the bottom, respectively. If d1 falls first it
will collect other droplets available on its way down. Similar
results were seen for d2 and d3; logically d4 cannot influence the
downward movement of d1, d2, and d3.
In contrast, DH barbs followed a continuous fog-collection
cycle as they were never saturated (Figures 4, 5, and 6c−e) at
their bases. Due to a film formation, there was an overall
influence of the fog droplets on the DH awns. Droplets
captured on the barbs (DH awn) were subsequently trans-
ported to the base of the barbs followed by the transport
downward (observed and documented by the CCD camera).
No droplets were seen attached on the sides of the awns
(observed and documented by the CCD camera and a camera
with Nikor medical lens). Even the film on the surface was not
thick enough to reduce the void space between two awns
(Figures 4, 5, and 6c−e) because the film was continuously
flowing downward due to the grooves directed from the top to
the bottom and due to the gravitational force as well. Other
orientations of the awns, e.g., horizontal, upward orientation,
etc., resulted in a lower transport of water. Thus, a reduced
efficiency was observed.
For 50% coverage of the sample of DH awns, no clogging
was observed except at the bottom for a short time due to the
fast transport of the tiny droplets from the tips of the barbs to
their bases, followed by the downward movement. Even if we
increased the coverage by DH awns to ∼65%, we did not
observe any clogging, with the exception of at the bottom for a
short period of time.
Fog Collection by the Samples without Barbs. After
2−5 min of effective fog capture by Tillandsia NLSs, the
trichomes were saturated (Figure S4a, b in Supporting
Information) and remained attached to the NLSs. Con-
sequently, during the last half (5 min) of fog-flow duration,
the dripping of water from the NLSs of the sample reduced.
Nevertheless, due to their superhydrophilic trichomes43 on the
E DOI: 10.1021/acs.langmuir.5b02430
Article
NLSs, they showed moderate fog-collection efficiency (Figure
3). It is to be noted that the amount of water remaining on the
surface was less than those of the hydrophobic samples, e.g.,
Hordeum vulgare (fresh awn), Abies bracteata, etc., but one of
the highest among all the samples due to their absorption of
water in the NLSs.
Droplets remained pinned to the surfaces of Abies bracteata,
Pinus canariensis, and Sequoiadendron giganteum (Figures 6b and
S3a, b in Supporting Information) until they were large enough
to be dominated by the gravitational force, which is consistent
with the contact angle and CAH (Table 2) of the surfaces. The
illustration of the droplet behavior on a needle of Abies
bracteata is shown in Figure 6b. It showed similar droplet
behavior to that of the hydrophobic Hordeum awn. Pinus
canariensis also showed similar behavior (Figure S3a), but the
droplets on the needles were more elongated than those of
Abies and Sequoiadendron due to higher CAH and the presence
of microgrooves. The droplets attached elongated after getting
larger and fused with similar droplets, forming a thick film on
the hydrophilic needles (see contact angle and CAH in Table
2) of Sequoia sempervirens.
In the second set of samples, the diameter of Sequoia and
Abies needles was reduced by cutting them. This resulted in the
increase of the sharpness and surface area of the edges of the
needles. We also used Tillandsia NLSs of smaller diameter (see
Table 1). Although the effective surface area occupied by the
Ns/NLSs remained the same (50−55% coverage), their fog
collection increased by 7−13% (Tillandsia 7.3%, Abies 9%, and
Sequoia 12.6%). Despite the increase, only Tillandsia from them
could reach the efficiency comparable to the barbed Hordeum
awn (fresh), but it was still only ∼60% of the efficiency of DH
awns. However, the efficiency of Tillandsia decreased over time
due to the saturation of the whole NLSs, and therefore there
were no free trichomes to catch fog droplets after a few
minutes. Moreover, a drastic decrease (Table 1) of the width of
Sequoia Ns is a possible cause for a total of 12.6% increase of
fog collection (Figure 3).
■ DISCUSSION
The size, form, and surface structure of the samples where fog
droplets can be captured are very important factors in fog
collection. The fog droplet carrying wind, while approaching an
obstacle in front of it, tends to deviate its direction to the right
or left according to the space available. If the obstacle is a
meshlike structure, the edges of the fibers (i.e., the awns or
needles of the samples in this study) play a crucial role in
capturing fog droplets. We not only checked the influence of
edges of the Ns/NLSs but also controlled their diameter, as
closely as was possible, so that they were comparable in size.
The increase in fog collection by the modified samples is
believed to be caused by the reduced diameter, increase of edge
for higher impingement, and increase of sharpness and
wettability of the edges due to cutting. Despite these measures
being taken, the amount of water collected by Hordeum (DH)
was almost twice as high as any other samples. Here, an
effective transport of the water captured played the major role
in keeping the barbs free to continuously capture more fog
droplets. The efficiency of the DH awns was attributed mainly
to the conical shape of the barbs; this generates a Laplace
pressure gradient between the two opposite sides of the drop
on the barb.44 It can be described as follows,
Langmuir XXXX, XXX, XXX−XXX
Langmuir
R2 2γ
ΔPbarb = − ∫R 1 (R + R 0)2
sin αdz
where γ is the surface tension of water, R is the barb’s local
radius (Figure 7 shows two local radii (R1 and R2) of the barb at
Figure 7. Schematic of the movement of a water droplet (in blue)
from the tip to the base of a barb. The region near the tip (radius R1)
of the barb has larger pressure than the base region (radius R2); thus
water droplet moves from higher to lower pressure. The direction of
droplet movement is marked by an arrow.
the two opposite sides of the drop), R0 is the radius of the drop,
α is the half apex angle of the barb (Figure 7), and dz is the
integration variable. The region near the barb’s tip (region with
a smaller radius) has a larger Laplace pressure than that near
the base (region with a larger radius). The energy generated
from the pressure difference (ΔPbarb) drives the droplet to
move from the tip to the base of the barb.11
The combination of higher wettability and barbs on both
sides proved to be the major factors for a high capture rate. In
our previous study,36 we showed that both microstructured and
superhydrophilic surfaces have higher capability to capture fog
droplets than those of flat and/or hydrophilic or hydrophobic.
Because of the influence of morphological differences in the
samples, the influence of their wettability on fog-collection
efficiency cannot be expected to be exactly similar to the earlier
study. However, a higher efficiency of superhydrophilic (θ =
0°) T. usneoides than of the hydrophobic samples (see contact
angles in Table 2) tends to show a similarity. In the case of
hydrophobic surfaces, such as fresh Hordeum awns and needles
of Abies, Pinus, and Sequoiadendron, water droplets remained
attached to their surfaces (Figures 6a, b and S3a, b), resulting in
a discontinuous fog-collection cycle.28
If we deduct the amount of water remaining on the
hydrophobic samples, their efficiency would be much lower,
i.e., 10−20% lower. It is to be noted that even after 30 min the
amount of water that remained on the surface would be the
same as it was after 10 min, but ultimately the presence of these
droplets influence the total amount of fog collection. If only the
amount of water collected in the container under the sample is
considered, excluding the water remaining on the sample after
10 or 30 min observation, again the DH awn is expected to be
the most efficient among the samples investigated, because it
retained the lowest amount of water on the sample.
The hydrophilization of the hierarchical architecture of the
barbed awn attributed higher wettability of the entire surfaces.
Higher wettability of the conical structures, as well as other
parts of the awns allowed the droplets to spread. Consequently,
the presence of a precursor liquid film on the surface provided a
F DOI: 10.1021/acs.langmuir.5b02430
Article
lubrication effect that increased the velocity of the drop motion
on the surfaces.31,44 Nature also provides logic to support it.
Many lower plants, such as lichens, mosses, etc., and higher
plants, such as Bromeliads, epiphytic orchids, etc., have evolved
superhydrophilic surface property to uptake water.43,45,46
Moreover, the presence of microgrooves on the surface of
the awns improved the downward movement of water. In other
words, a faster transport on the surface has been achieved by
higher wettability along with the microgrooves.36,47,48 The
droplets that hit the grooved surface of the awns tend to fuse
with neighboring droplets and grow larger. Capillary forces
cause the droplets that accumulate at the base of the barbs, to
be pulled into the grooves. Then, their movement is influenced
by the energy barrier in different directions. The first thing that
helps downward movement is the vertical orientation of the
awns. The second thing is the zero-energy barrier parallel to the
groove direction.49 Therefore, the droplets form an elongated
filament shape in the aforementioned direction, where it finds
another similar source to follow the path until the end and a
spontaneous spreading of water along the direction occurs as it
gathers sufficient momentum to do so.50−53 Furthermore,
during the accumulation of the fog drops on the surfaces
(actually condensation), heat transfer is an important issue.
Hydrophilic surfaces could be an advantage compared to the
hydrophobic surfaces in this respect.48 Therefore, directional
microgrooved surfaces with higher wettability enhance the
drainage performance along with a high rate of droplet
condensation and/or accumulation. To reduce re-entrainment
in fog collection, highly adhesive superhydrophobic surfaces,
e.g., peanut leaf surface, have been studied by a research
group.54 The surfaces remain occupied with the droplets
attached on them unless the droplets have been influenced by
the gravitational force. Moreover, this effect would have a
negative influence for the hierarchically barbed samples by
blocking the barbs.
The meshes with 50% coverage theoretically remove 50% of
the droplets approaching them.55 Although barbs on the edges
of the fibers (as seen on the awn sample) only occupy a low
space, they can capture a high percentage of the fog droplets
that would otherwise be lost if there were no barbs. Park et al.37
showed that clogging would be a problem in hydrophilic
meshes. Provided that their samples contained no barbs, the
droplets remained attached at the edges of the fibers, resulting
in clogging. The clogging observed only at the bottom of the
DH awn for a short period of time would not be a problem as it
did not hamper the flow of the fog. It should be noted that it is
also possible to overcome this problem with design
optimization (Figure 8). Therefore, even superhydrophilic
meshes of barbed fibers could solve the clogging problem.
Figure 8. Schematic of the proposed design of a part of a fog collector
(a); barbed vertical fibers having microgrooves from top to bottom
(b); a single barb with an apex angle 2α (c).
Langmuir XXXX, XXX, XXX−XXX
Langmuir
Moreover, because of the higher wettability, the droplets
captured on the surface of the awns could not be blown away
by the wind, which has been shown to be a major problem for
the superhydrophobic surface of the mesh during fog
collection.37 Thus, the superhydrophilic surface property also
would nullify the possibility of re-entrainment (loss of captured
droplets by the wind). Due to adequate adhesion at
hydrophobic surfaces (a function of CA and CAH; see Table
2), there is no possibility of re-entrainment from the
hydrophobic samples (not superhydrophobic) used here.54
Proposed Design of Fog Collectors. A design for highly
efficient fog collectors has been proposed in Figure 8. Barbed
fibers are vertically oriented with a constant gap in between.
The horizontal fibers are only for support, and they are placed
as far apart as possible so that the transport is not hampered. It
should be noted that it would be better if the horizontal fibers
can be placed on the backsides of the vertical barbed fibers,
resulting in a challenge of engineering optimization. The
dimension can be chosen as discussed for the model sample
and/or optimized by further research.
■
CONCLUSIONS
We have discussed an approach to improve the efficiency of
biomimetic fog collectors in this study. We have also proposed
a possible design to fabricate meshes with hierarchical
architecture by using inexpensive materials for the target user
of people in poor communities in water-deficient dry areas.
With this new design, fog droplet capturing could be optimized,
clogging and re-entrainment problems could be solved, and
transport of water along the fog collector can be greatly
improved. Therefore, the investigation of fog collection on the
hierarchical, barbed surface architecture may open new
possibilities to develop highly efficient fog collectors.
■
ASSOCIATED CONTENT
*
S Supporting Information
This material is available free of charge via the Internet at
http://pub.acs.org. The Supporting Information is available
free of charge on the ACS Publications website at DOI:
10.1021/acs.langmuir.5b02430.
Awns/needles/needle-like structures used in the fog-
collection experiment; structural characterization of a
needle of Sequoiadendron giganteum; a dry and a wet leaf
of Tillandsia usneoides before and during fog-collection
experiment, respectively; fog droplet behavior on the
surfaces of the needles of Pinus canariensis and
Sequoiadendron giganteum (PDF)
■ AUTHOR INFORMATION
Corresponding Author
*E-mail: azad@uni-bonn.de.
Notes
The authors declare no competing financial interest.
■
ACKNOWLEDGMENTS
The research was supported by the Graduate School of Bionics
under the Ph.D. program (DFG Graduiertenkolleg GRK1572):
“BionicsInteraction across boundaries to the environment”,
in University of Bonn, Germany, financed by the German
Research Foundation (DFG). We thank the staff of the
Botanical Gardens and Institute of Crop Science and Resource
Conservation, Crop Genetics and Biotechnology Unit,
G DOI: 10.1021/acs.langmuir.5b02430
Article
University of Bonn, for providing plant samples and Axel
Hinnemann (Rhine-Waal University of Applied Sciences) for
assistance in atomic force microscopy. We also thank Danica
Christensen (Nees Institute, University of Bonn), a native
English speaker, for proofreading the manuscript.
■
REFERENCES
(1) World Health Organization and UNICEF, Joint Monitoring
Program. Progress on sanitation and drinking-water, 2013 update; World
Health Organization: Geneva, 2013.
(2) Klemm, O.; Schemenauer, R. S.; Lummerich, A.; Cereceda, P.;
Marzol, V.; Corell, D.; van Heerden, J.; Reinhard, D.; Gherezghiher,
T.; Olivier, J.; Osses, P.; Sarsour, J.; Frost, E.; Estrela, M. J.; Valiente, J.
A.; Fessehaye, G. M. Fog as a Fresh-Water Resource: Overview and
Perspectives. Ambio 2012, 41 (3), 221−234.
(3) Schemenauer, R. S.; Cereceda, P. Fog-Water Collection in Arid
Coastal Locations. Ambio 1991, 20 (7), 303−308.
(4) Fessehaye, M.; Abdul-Wahab, S. A.; Savage, M. J.; Kohler, T.;
Gherezghiher, T.; Hurni, H. Fog-water collection for community use.
Renewable Sustainable Energy Rev. 2014, 29 (0), 52−62.
(5) Goodman, J. The Collection of Fog Drip. Water Resour. Res.
1985, 21 (3), 392−394.
(6) Limm, E. B.; Simonin, K. A.; Bothman, A. G.; Dawson, T. E.
Foliar water uptake: a common water acquisition strategy for plants of
the redwood forest. Oecologia 2009, 161 (3), 449−459.
(7) Simonin, K. A.; Santiago, L. S.; Dawson, T. E. Fog interception by
Sequoia sempervirens (D. Don) crowns decouples physiology from
soil water deficit. Plant, Cell Environ. 2009, 32 (7), 882−892.
(8) Stanton, D. E.; Horn, H. S. Epiphytes as “filter-drinkers”: life-
form changes across a fog gradient. Bryologist 2013, 116 (1), 34−42.
(9) Martorell, C.; Ezcurra, E. The narrow-leaf syndrome: a functional
and evolutionary approach to the form of fog-harvesting rosette plants.
Oecologia 2007, 151 (4), 561−573.
(10) Martorell, C.; Ezcurra, E. Rosette scrub occurrence and fog
availability in arid mountains of Mexico. J. Veg. Sci. 2002, 13 (5), 651−
662.
(11) Ju, J.; Bai, H.; Zheng, Y.; Zhao, T.; Fang, R.; Jiang, L. A multi-
structural and multi-functional integrated fog collection system in
cactus. Nat. Commun. 2012, 3, 1247.
(12) Andrews, H. G.; Eccles, E. A.; Schofield, W. C. E.; Badyal, J. P. S.
Three-Dimensional Hierarchical Structures for Fog Harvesting.
Langmuir 2011, 27 (7), 3798−3802.
(13) Westbeld, A.; Klemm, O.; Griessbaum, F.; Strater, E.; Larrain,
H.; Osses, P.; Cereceda, P. Fog deposition to a Tillandsia carpet in the
Atacama Desert. Ann. Geophys. 2009, 27 (9), 3571−3576.
(14) Vogel, S.; Müller-Doblies, U. Desert geophytes under dew and
fog: The ″curly-whirlies″ of Namaqualand (South Africa). Flora 2011,
206 (1), 3−31.
(15) Nørgaard, T.; Ebner, M.; Dacke, M. Animal or Plant: Which Is
the Better Fog Water Collector? PLoS One 2012, 7 (4), e34603.
(16) Schemenauer, R. S.; Cereceda, P. A Proposed Standard Fog
Collector for Use in High-Elevation Regions. J. Appl. Meteorol. 1994,
33 (11), 1313−1322.
(17) Holmes, R.; Rivera, J. d. D.; de la Jara, E. Large fog collectors:
New strategies for collection efficiency and structural response to wind
pressure. Atmos. Res. 2015, 151, 236−249.
(18) Bai, H.; Ju, J.; Sun, R.; Chen, Y.; Zheng, Y.; Jiang, L. Controlled
Fabrication and Water Collection Ability of Bioinspired Artificial
Spider Silks. Adv. Mater. 2011, 23 (32), 3708−3711.
(19) Bai, H.; Ju, J.; Zheng, Y.; Jiang, L. Functional Fibers with Unique
Wettability Inspired by Spider Silks. Adv. Mater. 2012, 24 (20), 2786−
2791.
(20) Bai, H.; Tian, X.; Zheng, Y.; Ju, J.; Zhao, Y.; Jiang, L. Direction
controlled driving of tiny water drops on bioinspired artificial spider
silks. Adv. Mater. 2010, 22 (48), 5521−5525.
(21) Hou, Y.; Chen, Y.; Xue, Y.; Zheng, Y.; Jiang, L. Water collection
behavior and hanging ability of bioinspired fiber. Langmuir 2012, 28
(10), 4737−4743.
Langmuir XXXX, XXX, XXX−XXX
Langmuir Article

(22) Chen, Y.; Wang, L.; Xue, Y.; Jiang, L.; Zheng, Y. Bioinspired tilt- (43) Koch, K.; Barthlott, W. Superhydrophobic and superhydrophilic
angle fabricated structure gradient fibers: micro-drops fast transport in plant surfaces: an inspiration for biomimetic materials. Philos. Trans. R.
a long-distance. Sci. Rep. 2013, 3, 2927. Soc., A 2009, 367 (1893), 1487−1509.
(23) Hou, Y.; Gao, L.; Feng, S.; Chen, Y.; Xue, Y.; Jiang, L.; Zheng, Y. (44) Lorenceau, E.; Quéré, D. Drops on a conical wire. J. Fluid Mech.
Temperature-triggered directional motion of tiny water droplets on 2004, 510, 29−45.
bioinspired fibers in humidity. Chem. Commun. 2013, 49 (46), 5253− (45) Koch, K.; Bhushan, B.; Barthlott, W. Diversity of structure,
5255. morphology and wetting of plant surfaces. Soft Matter 2008, 4 (10),
(24) He, X.-H.; Wang, W.; Liu, Y.-M.; Jiang, M.-Y.; Wu, F.; Deng, K.; 1943−1963.
Liu, Z.; Ju, X.-J.; Xie, R.; Chu, L.-Y. Microfluidic fabrication of bio- (46) Rauh, W.; Schill, R.; Ehler, N.; Barthlott, W. Some remarks on
Inspired microfibers with controllable magnetic spindle-Knots for 3D the water supply of bromeliads. J. Bromeli. Soc. 1973, 23, 89−109.
assembly and water collection. ACS Appl. Mater. Interfaces 2015, 7, (47) Roth-Nebelsick, A.; Ebner, M.; Miranda, T.; Gottschalk, V.;
17471−17481. Voigt, D.; Gorb, S.; Stegmaier, T.; Sarsour, J.; Linke, M.; Konrad, W.
(25) Dong, H.; Wang, N.; Wang, L.; Bai, H.; Wu, J.; Zheng, Y.; Zhao, Leaf surface structures enable the endemic Namib desert grass
Y.; Jiang, L. Bioinspired electrospun knotted microfibers for fog Stipagrostis sabulicola to irrigate itself with fog water. J. R. Soc.,
harvesting. ChemPhysChem 2012, 13 (5), 1153−1156. Interface 2012, 9 (73), 1965−1974.
(26) Birajdar, M. S.; Lee, J. Nanoscale bumps and dents on (48) Liu, L.; Jacobi, A. M.; Chvedov, D. A surface embossing
nanofibers enabling sonication-responsive wetting and improved technique to create micro-grooves on an aluminum fin stock for
moisture collection. Macromol. Mater. Eng. 2015, 300, 1108−1115. drainage enhancement. J. Micromech. Microeng. 2009, 19 (3), 035026.
(27) Bai, F.; Wu, J.; Gong, G.; Guo, L. Biomimetic “Cactus Spine” (49) Kannan, R.; Sivakumar, D. Drop impact process on a
hydrophobic grooved surface. Colloids Surf., A 2008, 317 (1−3),
with Hierarchical Groove Structure for Efficient Fog Collection. Adv.
694−704.
Sci. 2015, 2 (7), 1500047.
(50) Chen, Y.; He, B.; Lee, J.; Patankar, N. A. Anisotropy in the
(28) Cao, M.; Xiao, J.; Yu, C.; Li, K.; Jiang, L. Hydrophobic/
wetting of rough surfaces. J. Colloid Interface Sci. 2005, 281 (2), 458−
hydrophilic cooperative Janus system for enhancement of fog
464.
collection. Small 2015, 11 (34), 4379−4384. (51) Kusumaatmaja, H.; Vrancken, R. J.; Bastiaansen, C. W. M.;
(29) Garrod, R. P.; Harris, L. G.; Schofield, W. C. E.; McGettrick, J.; Yeomans, J. M. Anisotropic Drop Morphologies on Corrugated
Ward, L. J.; Teare, D. O. H.; Badyal, J. P. S. Mimicking a Stenocara Surfaces. Langmuir 2008, 24 (14), 7299−7308.
Beetle’s Back for Microcondensation Using Plasmachemical Patterned (52) Kannan, R.; Sivakumar, D. Impact of liquid drops on a rough
Superhydrophobic−Superhydrophilic Surfaces. Langmuir 2007, 23 surface comprising microgrooves. Exp. Fluids 2008, 44 (6), 927−938.
(2), 689−693. (53) Kannan, R.; Vaikuntanathan, V.; Sivakumar, D. Dynamic contact
(30) White, B.; Sarkar, A.; Kietzig, A.-M. Fog-harvesting inspired by angle beating from drops impacting onto solid surfaces exhibiting
the Stenocara beetleAn analysis of drop collection and removal from anisotropic wetting. Colloids Surf., A 2011, 386 (1−3), 36−44.
biomimetic samples with wetting contrast. Appl. Surf. Sci. 2013, 284 (54) Yang, S.; Ju, J.; Qiu, Y.; He, Y.; Wang, X.; Dou, S.; Liu, K.; Jiang,
(0), 826−836. L. Peanut Leaf Inspired Multifunctional Surfaces. Small 2014, 10 (2),
(31) Ju, J.; Xiao, K.; Yao, X.; Bai, H.; Jiang, L. Bioinspired Conical 294−299.
Copper Wire with Gradient Wettability for Continuous and Efficient (55) Schemenauer, R. S.; Joe, P. I. The collection efficiency of a
Fog Collection. Adv. Mater. 2013, 25 (41), 5937−5942. massive fog collector. Atmos. Res. 1989, 24 (1), 53−69.
(32) Heng, X.; Xiang, M.; Lu, Z.; Luo, C. Branched ZnO Wire
Structures for Water Collection Inspired by Cacti. ACS Appl. Mater.
Interfaces 2014, 6 (11), 8032−8041.
(33) Cao, M.; Ju, J.; Li, K.; Dou, S.; Liu, K.; Jiang, L. Facile and
Large-Scale Fabrication of a Cactus-Inspired Continuous Fog
Collector. Adv. Funct. Mater. 2014, 24 (21), 3235−3240.
(34) Ju, J.; Yao, X.; Yang, S.; Wang, L.; Sun, R.; He, Y.; Jiang, L.
Cactus Stem Inspired Cone-Arrayed Surfaces for Efficient Fog
Collection. Adv. Funct. Mater. 2014, 24 (44), 6933−6938.
(35) Peng, Y.; He, Y.; Yang, S.; Ben, S.; Cao, M.; Li, K.; Liu, K.; Jiang,
L. Magnetically induced fog harvesting via flexible conical arrays. Adv.
Funct. Mater. 2015, 25, 5967−5971.
(36) Azad, M. A. K.; Ellerbrok, D.; Barthlott, W.; Koch, K. Fog
collecting biomimetic surfaces: Influence of microstructure and
wettability. Bioinspiration Biomimetics 2015, 10 (1), 016004.
(37) Park, K.-C.; Chhatre, S. S.; Srinivasan, S.; Cohen, R. E.;
McKinley, G. H. Optimal Design of Permeable Fiber Network
Structures for Fog Harvesting. Langmuir 2013, 29 (43), 13269−13277.
(38) Lakes, R. Materials with structural hierarchy. Nature 1993, 361
(6412), 511−515.
(39) Barthlott, W.; Theisen, I. Epicuticular Wax Ultrastructure. In
Flowering PlantsMonocotyledons; Kubitzki, K., Ed.; Springer: Berlin/
Heidelberg, 1998; Vol. 3, pp 20−22.
(40) Wilhelmi, H.; Barthlott, W. Mikromorphologie der Epicuticu-
larwachse und die Systematik der Gymnospermen. Trop. subtrop.
Pflanzenwelt 1997, 96, 1−49.
(41) Stabentheiner, E.; Pfeifhofer, H. W.; Peters, J.; Jiménez, M. S.;
Morales, D.; Grill, D. Different surface characteristics of primary and
secondary needles of Pinus canariensis. Flora 2004, 199 (2), 90−99.
(42) Neinhuis, C.; Barthlott, W. Characterization and Distribution of
Water-repellent, Self-cleaning Plant Surfaces. Ann. Bot. 1997, 79 (6),
667−677.

H DOI: 10.1021/acs.langmuir.5b02430
Langmuir XXXX, XXX, XXX−XXX