The biological surface has developed functional structures during long-term evolution, which inspires the development of biomimic materials for optical and optoelectronic applications. For example, the micropapillae and nanofolding structures of rose petals could enhance light absorption and color saturation. Here, the authors report a successful replication of rose hierarchical surface structures by simple and cost-effective processes. A variety of rose structured surfaces were investigated, which confirmed the diversity of functional surface architecture. The polydimethylsiloxane (PDMS) negative replica was formed by casting PDMS solution on top of a rose petal followed by a temperature-assisted curing process. The hierarchical structure was further transferred into photoresist films by ultraviolet nanoimprint using the PDMS replica as molds. The imprinted photoresist films demonstrated uniform replications of rose microconvex cells with nanofolding details in the scale of a square centimeter. Super-hydrophobicity was demonstrated on both PDMS negative replica and photoresist positive replica. The incorporation of photoresist replica on the surface of photodetectors improved the responsivity by 35% to 42% due to enhanced light management effect. This bio-inspired transfer imprint process with PDMS provided a high-fidelity and cost-effective method to reproduce functional structures from biological surfaces. This study also demonstrated the potential of utilizing rose structures in photovoltaic and optoelectronic applications.

Nature has offered us a lot of inspirations for designing and fabricating new functional materials and structures.1,2 The biological surface, as the first interface between the nature creatures and surrounding environment, has developed multiple functions during uninterrupted evolutions.3–6 The moth eyes, for example, have subwavelength cornel nipple array structures, which act as an effective medium with a gradually changed refractive index.7 The butterfly wing scales have multilayer ridge array structures and show beautiful iridescence due to the photonic crystal effects.8 Both moth eyes and butterfly wings have been extensively studied and imitated to form antireflection and light trapping layers.6–9 Besides animal surfaces, the surface of plants including, flowers10–12 and leaves,3,13 also feature intelligent and interesting structures and surface architectures. While the color of plants is normally pigment-based, the micro- and nanohierarchical surface structures strengthen the color saturation by reducing light reflection and redirecting photons to strike pigments.14 With such light trapping mechanism, flowers appear more attractive for pollinators, while leaves could sustain metabolism in low-light environment via photosynthesis.15,16 Reproducing such functional surface structures would enable a lot of applications such as forming super-hydrophobic surface for water repellent or self-cleaning,3 and light harvesting element for photovoltaic and optoelectronic devices.10 However, the research on plant surfaces and their inspired microfabrication are still limited to date compared with animal counterparts.

Rose, as one of the most popular flowers over the world, is known for its beauty and fragrance. Previously, Feng et al.12 discovered rose structure-based super-hydrophobic property with high water-adhesive force called “petal effect.” Direct mimic of the three-dimensional hierarchical structures is challenging and relies on an expensive microfabrication technique such as interference lithography or electron-beam lithography.17,18 Replication via nanoimprint19–21 or molding,11,22 on the other hand, provides advantages of low-cost and high-throughput. A two-step transfer process by first generating a negative replica and then forming a positive replica is usually involved. Successful replication of rose surface structures have been reported with polystyrene,7 epoxy resin,11 and UV-curing resist.10 Hünig et al.10 also demonstrated reduced broadband reflection and improved solar cells performance by coating resist replica of the Rosa “El Toro.” Inspired by their work, here we studied the diverse surface structures of rose petals, and successfully replicated the hierarchical structures by a simple and low-cost process. The polydimethylsiloxane (PDMS) negative replica was first synthesized via solution mixing and curing. The photoresist positive replica was then formed by the UV-imprint with the PDMS mold. The imprinted rose structures were further incorporated on photodetectors, which demonstrate improved sensitivity.

1. Rose petals preparation

Roses with a variety of species and colors were collected to study their relation with surface structures. Fresh roses without pretreatment were used to avoid collapse of surface structures due to dehydration. Rose samples were prepared by cutting off from the flat part of the rose petals, which is usually in the size of a few square centimeters.

2. PDMS negative replicas

In order to replicate the rose hierarchical structures, the PDMS mold with negative replica was first prepared as shown in the schematic of Fig. 1. First, the PDMS solution was synthesized from the prepolymer (Momentive RTV 615A) and cross-linking agent (Momentive RTV 615B) in a 10:1 volume ratio in a glass beaker. Then, the glass beaker was placed in a low-vacuum chamber for 1 h to evacuate the air bubbles that may cause defects in replication. After that, the rose samples were placed in a petri-dish with front-side facing up and the PDMS solution was transferred into the petri-dish. The petri-dish was then placed back in the low-vacuum chamber for another 20 min, which further evacuate air bubbles that introduced during PDMS solution transfer. The rose hierarchical structures were conserved in the low-vacuum environment under shelter of PDMS solutions. After that, the PDMS/rose assembly was heated for 1 h at 60 °C. The temperature was chosen to facilitate the PDMS curing process while reserving the shape and structures of the rose petals. Finally, the hardened PDMS/rose assembly was carefully removed from the petri-dish with razor blade after cooling down to room temperature. In order to separate the PDMS mold from rose petals, the PDMS/rose assembly was soaked in heated piranha solution (mixture of sulfuric acid and hydrogen peroxide in 2:1 volume ratio) at 110 °C for 10 min. The PDMS negative replica was completed after rinsing with deionized water and blow dried with nitrogen.

Fig. 1.

(Color online) Schematic of process to create PDMS mold with negative replication of rose petal structures.

Fig. 1.

(Color online) Schematic of process to create PDMS mold with negative replication of rose petal structures.

Close modal

3. Photoresist positive replicas

Positive replica of rose structures in photoresist films were created by UV-imprint with the PDMS molds, as shown in Fig. 2. A droplet of SU-8 2002 from MicroChem was placed on the substrate, and carefully pressed with the PDMS mold to ensure contact without air bubbles. The photoresist was then cured with 350 nm UV at an irradiance of 30 mW/cm2 for 10 min in an Electro-Lite Electro-Cure 500 light exposure system. Finally, the PDMS mold was carefully separated from the SU-8 2002 coated substrate, leaving positive replication of rose structures.

Fig. 2.

(Color online) Schematic of UV-nanoimprint process to create photoresist positive replica of rose structures.

Fig. 2.

(Color online) Schematic of UV-nanoimprint process to create photoresist positive replica of rose structures.

Close modal

1. Surface characterization

The surface morphology of the rose petals, PDMS negative replicas or molds, and photoresist positive replicas were investigated with scanning electron microscopy (SEM). The SEM measurements were conducted with Zeiss Supra–40 SEM systems with a chamber vacuum below 1 × 10−4 Torr. All samples were sputtered with ∼10 nm gold layer by Hummer VII sputter deposition system prior to SEM characterization.

The contact angles of negative replicated PDMS and flat pristine PDMS were measured using Ramé-Hart Contact Angle Goniometer with 20 μl deionized water droplet on the sample surface.

2. Optoelectronic characterization

Optoelectronic performance of two sets of Si photodiodes was characterized. One set of photodiodes was coated with photoresist films of positive rose replica while the other set was coated with flat photoresist films. A Keithley 4200 SCS and a Cascade probe station were used for photodetector's I-V measurements. The devices under dark environment and different illumination conditions (0.05 to 530 μW/cm2) were tested. A 150 W halogen lamp was used for illumination for all devices.

The rose hierarchical structures were reported to be capable of reducing light reflection and increasing light absorption for color strengthen.23 The morphology of a variety of rose petals was investigated with SEM. Figure 3(a) shows the rose featuring uniform and closely packed microconvex cells, which is referred as micropapillae,4 with diameter around 20–30 μm. The individual micropapillae also demonstrated cuticular foldings with irregular ridges in a few hundred nanometers to a few micrometers [Fig. 3(a), inset]. More folding details were observed on the top of micropapillae compared with the sidewalls. Figure 3(b) shows the rose with only microconvex structures but without nanofoldings. The aspect ratio of its micropapillae, defined as the ratio of vertical depth to base diameter, is much higher than that of most flower petals reported by the literature11,23 [Fig. 3(b), inset]. The two different rose morphologies represent the diversity of functional surface developed during evolution. It was reported that both micropapillae with high aspect ratio and nanofoldings would benefit the light trapping.23 In particular, specular reflection is largely reduced ascribed to multiple reflections between micropapillaes. Figures 3(c) and 3(d) show the hierarchical structures of same rose Rosoideae species with yellow and pink colors, respectively. No obvious micropapillae and cuticular folding structure difference were observed, which indicated the colors are chemical pigment-based. However, the color intensity could be enhanced with surface hierarchical structures as more light could be absorbed by the pigments ascribed to reduced reflection and photon redirection. Therefore, the roses appear more optically attractive for the pollinators. Note that the micropapillaes and nanofolding structure shrinkage were observed in all SEM characterized rose petals, with Rosoideae rosa showing most severe shrinkage or collapse, ascribing to the dehydration of cells in high-vacuum chamber.4 

Fig. 3.

SEM micrographs of rose petal surface morphology. (a) Rose hierarchical structure with micropapillae and cuticular nanofoldings; inset shows nanofolding structures with higher magnification. (b) Rose with only micropapillae structures; inset shows high aspect ratio micropapillae with tilted view at 45°. (c) and (d) Rosoideae rosa hierarchical structure with (c) yellow and (d) pink colors, respectively.

Fig. 3.

SEM micrographs of rose petal surface morphology. (a) Rose hierarchical structure with micropapillae and cuticular nanofoldings; inset shows nanofolding structures with higher magnification. (b) Rose with only micropapillae structures; inset shows high aspect ratio micropapillae with tilted view at 45°. (c) and (d) Rosoideae rosa hierarchical structure with (c) yellow and (d) pink colors, respectively.

Close modal

The PDMS negative replica was characterized with SEM and a typical micrograph of the surface morphology was shown in Fig. 4(a). The homogeneous and uniform microconcave cells are negative replication of the rose micropapillaes. The nanofolding structures are also conserved on each concave cell. The ridges on the bottom and sidewall are correlated with the grooves structures on rose micropapillae cuticle [Fig. 4(a), inset]. Using fresh rose petals as well as optimizing PDMS synthesis is important for successful replication. A major difficulty for PDMS replication is to achieve macroscopically surface flatness, as flower petal has a natural shape. Preprocessing the rose petals by depressing or sandwiching between glasses slides, although improving the sample flatness, could simultaneously damages the micropapillae structures. Nevertheless, replicating flat and high-fidelity rose structures are generally much easier than that based on animals such as moth eyes and butterfly wings, as their structures are only homogeneously distributed in scale of square millimeter without discontinuity. Figure 4(b) shows a picture of PDMS/rose assembly in petri-dish before curing. Reasonable flatness in ∼1 in.2 is achieved by cutting rose sample from the center of petal. The optimum curing temperature is found at 60 °C, which conserves the shape and structures of rose petals while facilitating the PDMS cross-linking process. An example view of the cured PDMS/rose assembly is shown in Fig. 4(c). The shape and color of the yellow rose petal are conserved in the solidified PDMS. The appearance of the final PDMS mold after removing rose residues by piranha cleaning is demonstrated in Fig. 4(d). The structure-replicated PDMS possess white color with reduced transparency. The appearing color of PDMS replica does not correlate with rose color, as rose color is chemical pigment-based. The PDMS mold retains a reasonable flatness which is beneficial for transfer imprint to create rose positive replicas. The contact angle was measured on both the PDMS negative replicas and the pristine PDMS without structures. The pristine PDMS demonstrated hydrophobic property with a contact angle of 112° [Fig. 4(e)], which agrees with the literature.24,25 The structured PDMS replica, on the other hand, shows super-hydrophobicity with a contact angle of 150° [Fig. 4(f)], demonstrating the impact of micro- and nanostructures on macroscopic surface hydrophobicity. Such a super-hydrophobic surface could be used for various applications such as water-repelling and self-cleaning.

Fig. 4.

(Color online) SEM and photographs of the synthesized PDMS replicas. (a) SEM micrographs of PDMS negative replica showing microconcave cells with nanofoldings; inset shows nanofolding structures at higher magnification. Photographs of (b) a PDMS/rose assembly before curing, (c) a cured PDMS/rose assembly, and (d) a final PDMS mold after rose petals removed by piranha clean. Photographs of water droplets on a PDMS surface (e) without structures, showing a contact angle of 112° and (f) with negative replicated rose structures, showing a contact angle of 150°.

Fig. 4.

(Color online) SEM and photographs of the synthesized PDMS replicas. (a) SEM micrographs of PDMS negative replica showing microconcave cells with nanofoldings; inset shows nanofolding structures at higher magnification. Photographs of (b) a PDMS/rose assembly before curing, (c) a cured PDMS/rose assembly, and (d) a final PDMS mold after rose petals removed by piranha clean. Photographs of water droplets on a PDMS surface (e) without structures, showing a contact angle of 112° and (f) with negative replicated rose structures, showing a contact angle of 150°.

Close modal

The photoresist positive replica was created by imprinting PDMS negative replica into UV-curing photoresist and curing with UV. We found at that no external pressure is required between PDMS mold and the substrate during the curing step, which simplifies the process and does not require any instrument. However, the contact of the PDMS molds with the photoresist films plays an important role for defect control. Before imprinting, it is typical to pretreat the PDMS with oxygen plasma to reduce the mold surface hydrophobicity (e.g., 150° before treatment to 86° after treatment) for better contact with resist. Almost perfect contact of the PDMS mold with the photoresist films in large scale without any air gap or bubbles could be achieved with the treated PDMS surface. However, such treatment introduced difficulty during demolding process, even when PDMS surface recovers the hydrophobic property after a few hours of imprinting. Therefore, an optimum process was developed with untreated flat PDMS molds, by gently pressing into sufficient liquid resist droplet on the substrate to remove air bubbles before curing, and the PDMS super-hydrophobicity yields easy mold release. SEM was used to characterize the morphology of imprinted SU-8 2002 thin films, and the hierarchical structures in Figs. 5(a) and 5(b) were corresponded with the rose structures shown in Figs. 3(a) and 3(c), respectively. The uniform and plump microconvex cells with an average diameter around 20 to 30 μm were observed without visible defects or structure shrinkage, suggesting high reliability of replication process. The insets of Figs. 5(a) and 5(b) illustrate the zoom-in view of microconvex cells, which demonstrated detailed cuticular folding structures on top and sidewall with excellent reproduction of original rose structures. The SU-8 2002 thin films imprinted with PDMS negative replica also demonstrated hydrophobicity. The water contact angle of 150° [Fig. 5(c)] and 120° [Fig. 5(d)] were correlated with photoresist surface shown in Figs. 5(a) and 5(b), respectively, while a reference SU-8 2002 film imprinted with a flat pristine PDMS mold shows a contact angle of 90°. The higher contact angle of Fig. 5(a) compared with that of Fig. 5(b) was possibly ascribed to more nanofolding structures. The results demonstrated that large-scale, high-fidelity positive replication of rose hierarchical structures including micropapillae and nanofoldings could be successfully created on the photoresist thin-film by this simple imprint process. In addition, such functional structures could be imprinted repeatedly with the same PDMS mold, given the reliable demolding process due to its super-hydrophobicity.

Fig. 5.

(a) and (b) SEM micrographs of imprinted photoresist positive replicas of rose hierarchical structures in Figs. 3(a) and 3(c), respectively. The uniform micropapillae with nanofoldings are positive replication of rose surface structures. The inset shows nanofolding structures at higher magnification. (c) and (d) Photographs of water droplets on photoresist positive replica surface of (a) showing a contact angle of 150°, and (b) showing a contact angle of 120°.

Fig. 5.

(a) and (b) SEM micrographs of imprinted photoresist positive replicas of rose hierarchical structures in Figs. 3(a) and 3(c), respectively. The uniform micropapillae with nanofoldings are positive replication of rose surface structures. The inset shows nanofolding structures at higher magnification. (c) and (d) Photographs of water droplets on photoresist positive replica surface of (a) showing a contact angle of 150°, and (b) showing a contact angle of 120°.

Close modal

The micro- and nanohierarchical structures of rose petals can improve light absorption and color saturation.23 Successful replication of rose functional structures has potential application on photovoltaic or optoelectronic devices for performance enhancement. Here, photodetectors with incorporated rose functional structures were studied. Two sets of identical Si photodiodes were coated with 15 μl of SU-8 2002. One was imprinted by a PDMS mold with negative rose structure replications while the other was imprinted by a flat pristine PDMS mold without structures for reference. The photoelectrical performance of these two photodiodes was then characterized in the dark and under halogen light illumination. Their corresponding I-V characteristics under reverse junction bias were illustrated in Fig. 6(a). In dark state, both photodiodes show comparable currents at the level of tens of nanoampere. Under illumination, the currents significantly increased due to the generation of electron–hole pairs upon photon absorption, and charge carrier extraction under applied electrical field. The illuminated currents increased gradually with increasing irradiance. Under the same irradiance, the photodiode with replicated rose structures demonstrated higher current compared with the one without rose structures. Such behavior improvement was also illustrated in the plot of photodiode current versus irradiance at a 10 V reverse bias in Fig. 6(b). The responsivity commonly used to evaluate the efficiency of a photodetector responding to an optical signal is defined by

R=IilluminatedIdarkLlight×S,
(1)

where Llight is the irradiance, S is the effective area, and Iilluminated and Idark are the current with and without illumination, respectively. Here, the relative responsivity is calculated by the ratio of responsivity with rose structure to that with referenced device of flat surface. The plot of relative responsivity versus irradiance, as expressed in Fig. 6(b), demonstrated ∼35% to 42% improvement under 0.05 to 530 μW/cm2 illumination by incorporating replicated rose structures on the photodetectors. Similar behaviors are observed on multiple devices. The improved performance is ascribed to the light management effect of replicated rose hierarchical structures. The high-density and high-aspect ratio micropapillae function similar to the microlens array which offers antireflection and photon redistribution.26 Such light trapping effect is further enhanced with the irregular cuticular nanofolding structures.10 

Fig. 6.

(Color online) Optoelectronic characteristics of Si photodetectors with replicated rose structures and reference flat device. (a) Logarithm I-V plot of Si photodiodes under reverse junction bias tested in the dark and under halogen light illumination conditions (0.05–530 μW/cm2). The solid line represents the photodiode with photoresist rose replica while the dashed line represents the reference device without structures. (b) Plot of photodetector illuminated current vs irradiance (left Y-axis), and calculated photodetector relative responsivity vs irradiance (right Y-axis). The photodetector responsivity improved ∼35% to 42% by incorporating replicated rose structures. The error bars represent standard deviation in multiple tests.

Fig. 6.

(Color online) Optoelectronic characteristics of Si photodetectors with replicated rose structures and reference flat device. (a) Logarithm I-V plot of Si photodiodes under reverse junction bias tested in the dark and under halogen light illumination conditions (0.05–530 μW/cm2). The solid line represents the photodiode with photoresist rose replica while the dashed line represents the reference device without structures. (b) Plot of photodetector illuminated current vs irradiance (left Y-axis), and calculated photodetector relative responsivity vs irradiance (right Y-axis). The photodetector responsivity improved ∼35% to 42% by incorporating replicated rose structures. The error bars represent standard deviation in multiple tests.

Close modal

In summary, the diverse surface hierarchical structures of roses were investigated. The micropapillae and nanofoldings were observed on the rose petals, which improve light absorption and color saturation. Inspired by the rose functional structures, the PDMS molds with negative structure replication were synthesized in a simple solution casting and molding process. Uniform photoresist positive replicas of rose structures were formed successfully in the scale of square centimeter by UV-nanoimprint with the PDMS molds without an instrument. Super-hydrophobicity was demonstrated on both PDMS negative replicas and photoresist positive replicas, which could be used for water-repelling and self-cleaning applications. The super-hydrophobicity of PDMS replica also improved mold releasing and repeated patterning. The replicated rose structures were incorporated on the photodetectors, contributing to a 35% to 42% improvement on the device responsivity due to light management effect. The results demonstrated the potential of utilizing plant surface structures in photovoltaic and optoelectronic applications by a simple and cost-effective replication process.

This work was partially supported by the National Science Foundation (Grant Nos. CBET-1606141 and ECCS-0955027). W. Hu thanks support from the 1000 Talent Program of Shanghai, China. S. Cheng, B. Li, A. Li, and G. Du thank the CAST-TX's STEM Bridge program for supporting their high school summer camp.

1.
K.
Koch
and
W.
Barthlott
,
Philos. Trans. R. Soc., A
367
,
1487
(
2009
).
2.
B.
Bhushan
,
Beilstein J. Nanotechnol.
2
,
66
(
2011
).
3.
W.
Barthlott
and
C.
Neinhuis
,
Planta
202
,
1
(
1997
).
4.
K.
Koch
,
B.
Bhushan
, and
W.
Barthlott
,
Soft Matter
4
,
1943
(
2008
).
5.
L.
Feng
et al.,
Adv. Mater.
14
,
1857
(
2002
).
6.
Q.
Chen
,
G.
Hubbard
,
P. A.
Shields
,
C.
Liu
,
D. W. E.
Allsopp
,
W. N.
Wang
, and
S.
Abbott
,
Appl. Phys. Lett.
94
,
263118
(
2009
).
7.
Y.-F.
Huang
et al.,
Nat. Nanotechnol.
2
,
770
(
2007
).
8.
P.
Vukusic
and
J. R.
Sambles
,
Nature
424
,
852
(
2003
).
9.
S.
Niu
,
B.
Li
,
Z.
Mu
,
M.
Yang
,
J.
Zhang
,
Z.
Han
, and
L.
Ren
,
J. Bionic Eng.
12
,
170
(
2015
).
10.
R.
Hünig
et al.,
Adv. Opt. Mater.
4
,
1487
(
2016
).
11.
A. J.
Schulte
,
D. M.
Droste
,
K.
Koch
, and
W.
Barthlott
,
Beilstein J. Nanotechnol.
2
,
228
(
2011
).
12.
L.
Feng
,
Y.
Zhang
,
J.
Xi
,
Y.
Zhu
,
N.
Wang
,
F.
Xia
, and
L.
Jiang
,
Langmuir
24
,
4114
(
2008
).
13.
K.
Koch
,
B.
Bhushan
,
Y. C.
Jung
, and
W.
Barthlott
,
Soft Matter
5
,
1386
(
2009
).
14.
H. L.
Gorton
and
T. C.
Vogelmann
,
Plant Physiol.
112
,
879
(
1996
).
15.
H.
Bargel
,
K.
Koch
,
Z.
Cerman
, and
C.
Neinhuis
,
Funct. Plant Biol.
33
,
893
(
2006
).
16.
K.
Koch
,
B.
Bhushan
, and
W.
Barthlott
,
Springer Handbook of Nanotechnology
, edited by
B.
Bhushan
(
Springer
,
Berlin
,
2010
), pp.
1399
1436
.
17.
Y.
Hu
,
Z.
Wang
,
Z.
Weng
,
M.
Yu
, and
D.
Wang
,
Appl. Opt.
55
,
3226
(
2016
).
18.
R.
Kirchner
,
V. A.
Guzenko
,
M.
Rohn
,
E.
Sonntag
,
M.
Mühlberger
,
I.
Bergmair
, and
H.
Schift
,
Microelectron. Eng.
141
,
107
(
2015
).
19.
S. Y.
Chou
,
P. R.
Krauss
, and
P. J.
Renstrom
,
J. Vac. Sci. Technol., B
14
,
4129
(
1996
).
20.
Y.
Yang
,
K.
Mielczarek
,
M.
Aryal
,
A.
Zakhidov
, and
W.
Hu
,
ACS Nano
6
,
2877
(
2012
).
21.
H.
Wang
,
R.
Haroldson
,
B.
Balachandran
,
A.
Zakhidov
,
S.
Sohal
,
J. Y.
Chan
,
A.
Zakhidov
, and
W.
Hu
,
ACS Nano
10
,
10921
(
2016
).
22.
K.
Kerstin
,
S.
Anna Julia
,
F.
Angelika
,
N. G.
Stanislav
, and
B.
Wilhelm
,
Bioinspiration Biomimetics
3
,
046002
(
2008
).
23.
A. J.
Schulte
, “Light-trapping and superhydrophobic plant surfaces - optimized multifunctional biomimetic surfaces for solar cells,” Ph.D. thesis (
Rheinische Friedrich Wilhelms Universität
,
Bonn, Germany
,
2012
).
24.
D.
Bodas
and
C.
Khan-Malek
,
Sens. Actuators, B
123
,
368
(
2007
).
25.
S. H.
Tan
,
N.-T.
Nguyen
,
Y. C.
Chua
, and
T. G.
Kang
,
Biomicrofluidics
4
,
032204
(
2010
).
26.
J. D.
Myers
,
W.
Cao
,
V.
Cassidy
,
S.-H.
Eom
,
R.
Zhou
,
L.
Yang
,
W.
You
, and
J.
Xue
,
Energy Environ. Sci.
5
,
6900
(
2012
).