New Fabrication Method of Plastic Micro-Lens Arrays for

Researching on Compound Eyes of Insects

Toshiyuki Horiuchi and Ryunosuke Sasaki

Tokyo Denki University, 5 Senju-Asahi-cho, Adachi-ku, Tokyo, Japan

Keywords: Compound Eye, Micro-Lens Array, Projection Exposure, Epoxy Resin, Resist Mold.

Abstract: To develop artificial compound eyes, sizes of element lenses of typical insects were actually investigated, and

a new simple and low-cost method for fabricating plastic micro-lens arrays was developed. It was thought

essential to research on artificial compound eyes that lens parameters were freely controllable by our minds.

For this reason, a new easy and low-cost fabrication method had to be developed. In the new method, original

molds of micro-lens arrays with concave profiles were formed lithographically in a thick resist film. The

concave resist patterns were printed using a handmade 1/19 reduction projection exposure system by only one

exposure. Using intentionally defocused exposure, curvature radiuses were controllable in a very wide range

of 21-85 μm for the same transparent hexagon patterns with an inscribed circle diameter of 26.3 μm. It was

also verified that the resist-mold patterns were faithfully replicated to epoxy resin. After pouring the liquid

resin onto the silicon wafer chip with resist-mold patterns, hardened solid resin with micro-lens arrays was

separated from the wafer chip by peeling off the wafer chip mechanically. It is promising to fabricate micro-

lens arrays with aimed lens parameters although some more subjects should be cared from now on.

1 INTRODUCTION

Human beings have been imitating distinguished

functions of other living things, and invented various

apparatuses exercising performances superior to the

original ones. Recently, biomimetic technologies

utilizing exquisite functions or special organs of small

insects or protozoa are drawing attension for

improving our lives or curing some diseases of our

organs. For example, structures of morpho-butterfly

are utilized for obtaining highly reflective surfaces

with beautiful blue colors in wide view angles (Neu

et al., 2015) (Saito et al., 2004). Hydrophobic surfaces

are researched by getting hints from wrinkles of

living things (Bowden et al., 1998) (Genzer and

Groenewold, 2006). Flapping flight technology

reffering to insects are also researched for developing

light and small flying robots (Maet al., 2013).

Organs that the authors are especially interested

are compound eyes of insects. They are excellent

optical systems that humans are not equipped, and a

lot of very interesting papers were published (Chen et

al., 2011) (Ogata et al., 1994) (Sanders and Halford,

1995) (Cao et al., 2015) (Li et al., 2013) (Tanida et

al., 2001) (Duparré et al., 2006) (Jeong et al., 2006)

(Jiang et al., 2015). Compound eyes are highly

precise optics using micro-lens arrays, and new

applications will probably be found in the future

besides ordinary usage as eyes.

To start researches on compound eyes, it was

thought preferable to possess technology for

fabricating micro-lens arrays with various lens

parameters such as diameters and curvature radiuses

by ourselves easily. For this reason, a new fabrication

method of plastic micro-lens arrays is investigated

here.

2 SURVEYS ON SIZES OF LENS

ARRAYS

Compound eyes have been eagerly researched in the

world, and artificial compound eyes have also been

reported (Ogata et al., 1994) (Tanida et al., 2001)

(Duparré et al., 2006) (Li et al., 2013). Accordingly,

typical sizes of micro-lens arrays are known.

However, in the first step, it is preferable that actual

examples are easily obtained, and they are easily

observed for imitating them. For this reason,

compound eyes of cicada, dragonfly and goldbug

were in fact investigated. Outlooks of insects and

magnified photographs of their compound eyes are

shown in Figures 1-3.

Horiuchi, T. and Sasaki, R.

New Fabrication Method of Plastic Micro-Lens Arrays for Researching on Compound Eyes of Insects.

DOI: 10.5220/0005666900400047

In Proceedings of the 9th International Joint Conference on Biomedical Engineering Systems and Technologies (BIOSTEC 2016) - Volume 1: BIODEVICES, pages 40-47

ISBN: 978-989-758-170-0

Figure 1: Insects whose compound eyes were investigated.

Figure 2: Photographs of compound eyes of insects.

All of investigated compound eyes were composed of

micro-lenses arrayed in dense hexagonal arrangement

like a honeycomb. Typical widths of the element

hexagons or the inscribed circle diameters of element

lenses were 33, 53, and 16 μm, respectively. For, this

reason, sizes of fabricating micro-lenses were

Figure 3: Magnified view of compound eyes of insects.

tentatively fixed to around above diameters, and a

new simple and easy method for fabricating plastic

lens arrays was investigated. It was considered that

the lens parameters such as diameters, pitches, and

curvature radiuses should be widely controllable to

increase the facility of optical designs. In addition,

lens arrays should be fabricated by researcher

ourselves according to necessary specifications.

Regrettably, roundness or curvature radiuses of the

micro-lenses were not detectable. However, lens

characteristics are defined by both side curvatures of

the lens. In addition, the medium behind the lens is

not air but organic materials. Accordingly, refraction

characteristics of the lens are not decided only by the

curvature radius of the lens surface. For this reason, it

was thought that measurement of the lens surface

curvature was not necessarily required on this stage.

500 μ

Cicada

Goldbug

Insect

Outlook of compound eye

Dragonfly

Cicada

Dragonfly

Goldbug

Insect

Magnified view of compound eye

50 μm

(b) Dragonfly

(a) Cicada

New Fabrication Method of Plastic Micro-Lens Arrays for Researching on Compound Eyes of Insects

3 NEW FABRICATION METHOD

OF MICRO-LENS ARRAYS

Various fabrication methods of micro-lens arrays

have already been developed in the world.

Representative methods are ink-jetting (Zhu et al.,

2015) (Luoa et al., 2013) (Voigt et al., 2011) (Kim et

al., 2011), etching (Chen et al., 2010) (Deng et al.,

2012), grey scale lithography (Wu et al., 2002)

(Kuang et al., 2009) (Yang et al., 2007), laser tracking

(Wang, 2005) (Chiu and Lee, 2011), heat reflow

(Cheng et al., 2010) (Liu et al., 2010) (Pan and Su,

2007), and others.

However, It was thought that fundamental lens

parameters such as sizes and curvature radiuses

should be in suit with those of actual compound eyes

in the first step. In addition, it was considered that it

was preferable that various lenses were easily

purveyed for the research with a low cost. For this

reason, simple and low-cost fabrication method of

plastic lens arrays was reseached here.

The new method investigated in this research is

shown in Figure 4. At first, original resist-mold

patterns were lithographically formed on silicon

wafers using a handmade 1/19 reduction projection

exposure system (Hirota et al., 2003). Concave resist

patterns were printed by intentionally exposing the

resist under large defocus conditions. By applying

such defocused exposure, smoothly distributed light

intensity profiles were given to the resist films even

using binary reticles with only transparent and opaque

parts and without gray tone parts. After forming the

resist-mold patterns, the wafer was cut in small chips,

and each silicon wafer chip with resist patterns was

fixed at the bottom of a paper cup using a piece of

both-side adhesive tape, as shown in Figure 4(a).

In the next step, a plastic resin was poured on

the resist-mold patterns, as shown in Figure 4(b).

Epoxy resin (Nissin Resin, Crystal Resin NEO) was

used here. After adding a material for hardening the

resin to the main epoxy resin, they were mixed using

a mixing machine with a function of removing air

bubbles (Thinky, AR100), as shown in Figure 4(c).

This time, thicknesses and sizes of the resin blocks

were roughly controlled by the cup diameter and the

poured resin volume. Because the lens characteristics

are influenced by the lens block thickness, however,

the thickness should be strictly controlled in the next

research step.

After the resin block was sufficiently hardened, it

was taken out from the cup, as shown in Figure 4(d),

and the wafer chip with resist patterns was forcibly

peeled off. In concrete, grooves were dug along the

wafer chip edges using a cutter knife, and the wafer

chip was lifted up by climbing a tip of tweezer under

it. As a result, the resin block with profiles inverting

the resist-mold profiles was obtained, as shown in

Figure 4(e).

Convex plastic lens patterns obtained by peeling

the wafer chip off had defects of remained resist

fragments. For this reason, the resist fragments were

removed by dipping the resin block in acetone for a

short while.

Finally, the resin block was thinned by polishing

it. As a result, finished resin plate with micro-lens

arrays were obtained, as shown in Figure 4(f).

Figure 4: New fabrication method of micro-lens arrays.

Because lens sizes and profiles are decided by the

reticle pattern size and the exposure conditions for

forming the original resist patterns, it is expected that

lens parameters are variously and widely changeable

suiting to optical designs.

(b) Pour of epoxy resin.

(a) Wafer chip fixed at the

bottom of a paper cup.

agent and resin. Bubbles

were removed

simultaneously.

(d) Resin block hardened

and taken out of the

paper cup.

(e) Resin block after

removing the wafer

chip.

(f) Finished micro-lens

array plate.

Paper

cup

Liquid

resin

BIODEVICES 2016 - 9th International Conference on Biomedical Electronics and Devices

4 EVALUATION OF RESIST

MOLDS AND RESIN LENSES

4.1 Conditions for Printing Concave

Resist-mold Patterns

Micro-lens arrays were actually fabricated, and cross

sections and bird’s view profiles were evaluated. As

a resist, positive PMER P-LA900PM (Tokyo Ohka

Kogyo) was used, and the resist was coated on silicon

wafers in a thickness of approximately 10 μm.

Because actual compound eyes were hexagonally

arrayed, reticles shown in Figure 5 were used. The

width of transparent hexagon was fixed to 500 μm.

Because the reduction ratio was 1/19, the hexagon

width corresponds to the width of 26.3 μm on the

wafer plane, and this width almost corresponds to the

element lens diameter of cicada’s compound eye. On

the other hand, opaque boundary widths between

transparent hexagons were changed to 400, 300, and

200 μm.

Figure 5: Reticle patterns used for printing resist-mold

patterns of micro-lens arrays.

4.2 Evaluation of Resist-mold Patterns

At first, the best focal position or the focal origin was

decided as the stage position at which the sidewall

profiles of resist patterns became most perpendicular

to the wafer surface. Concave patterns with favorite

circular cross section profiles were obtained at the

defocus of +150 μm, as shown in Figure 6. Here, “+”

means that the exposed wafer was lowered down

from the projection lens. Concave patterns became

rounder and more circular in “+” direction than “-”

direction. Numerical aperture (NA) of the projection

lens was set at 0.12. Photographs of cross sections

were taken using a scanning electron microscope

(JEOL, JSM-5510).

Figure 6: Cross section variations of resist-mold patterns

under various defocus conditions.

Next, exposure time was varied to investigate the

relationship between the element lens profiles and the

exposure dose. In the experiments to decide the focal

origin by observing cross sections, it was very

difficult to break wafer chips at the exact center of the

concave resist patterns. For this reason, cross section

profiles under various exposure conditions were

evaluated using a laser microscope (Keyence, VK-

8510). Using the laser microscope, bird’s-eye views

were also obtained in addition to cross section

profiles.

At first, cross section profiles of resist-mold

patterns were measured. To evaluate cross section

profiles, they were pasted on screens of CAD

(Computer Aided Design) program (Autodesk,

AutoCAD 2014), and circles were delineated

superimposing on the cross section profiles by

defining the circles as they pass the concave bottoms

and both ends of the circular parts, as shown in Figure

7. By these manipulations, curvature radius R,

maximum profile error from the delineated CAD

circle δ, and concave depth d defined in the figure

were investigated.

500

m 200-400

+250 μm

+200 μm

+150 μm +100 μm

+50 μm 0 μm

10 μm

New Fabrication Method of Plastic Micro-Lens Arrays for Researching on Compound Eyes of Insects

Figure 7: Evaluated lens parameters of resist-mold patterns.

Figure 8: Wide variety of concave resist-mold patterns

when opaque pattern widths between transparent hexagonal

patterns and exposure doses are changed.

Dependence of circular radius R on exposure dose

is shown in Figures 8 and 9. Exposure dose of 384

mW/cm

corresponds to exposure time of 10 min. It

was clarified that the curvature radius was not almost

influenced by the opaque pattern width. This

characteristic is very convenient for designing reticles

to be prepared for fabricating micro-lenses with

aimed curvature radius. It was clarified that profiles

of resist-mold patterns were widely controllable by

changing exposure dose and opaque boundary width

between transparent hexagons, as shown in Figures 8

and 9.

Next, dependence of concave depth d is shown in

Fig. 10. In contrast to circular radius R, the concave

depth d varied depending on both exposure dose and

opaque pattern width between transparent hexagons.

The reason why the concave depths decreased for

reticles with narrow opaque parts is probably because

the resist at the opaque positions are exposed doubly

by light rays through transparent hexagons on both

sides, and slightly sensitized by the defocused

exposure. Therefore, the resist thicknesses of opaque

parts are also decreased during the development.

On the other hand, maximum profile error δ from

circular profiles are shown in Fig. 11. Although errors

became somewhat large under relatively large

exposure dose conditions, they were small, and it was

clarified that almost circular resist profiles were

obtained. Errors of 0.1 μm was the read out limit of

the used microscope.

Figure 9: Curvature radius dependence on exposure dose.

Curvature radiuses are not almost influenced by opaque

pattern widths between transparent hexagonal patterns.

R: Curvature radius

δ: Curvature profile error

d: Concave depth

BIODEVICES 2016 - 9th International Conference on Biomedical Electronics and Devices

Figure 10: Concave depth dependence on exposure time

and opaque pattern width between transparent hexagonal

patterns.

Figure 11: Profile error dependence on exposure time and

opaque pattern width between transparent hexagonal

patterns. Detection limit of profile error was approximately

0.1 μm.

4.3 Replication to Plastic Resin

Because technologies for fabricating resist mold

patterns were almost fixed, replication to plastic resin

was investigated next. That is, feasibility of

separating a wafer chip with resist patterns and a resin

block was investigated, and replicated plastic lens

profiles were compared with the original resist-mold

patterns.

Resist-mold patterns were formed by adding

exposure dose of 230 mJ/cm

(exposure time of 6

min). Bird’s-eye views of resist-mold patterns and

replicated plastic lens profiles are shown in Figure

12(a) and Figure 12(b). It was clarified that resist-

mold profiles were almost faithfully replicated to the

resin. In Figure 13, cross sections of resist-mold

patterns and resin lens patterns are compared. It is

known that both cross section profiles are almost the

same though they are symmetrically formed.

Measured curvature radiuses, depths and profile

errors were compared, as shown in Table 1.

Finally, because the hardened resin blocks were

too thick, they were thinned down by polishing.

However, it took very long times to thin the blocks,

and it was very difficult to control the final

thicknesses. It is necessary to develop a better method

to thin the lens plate precisely without spending long

times from now.

(a) Resist-mold patterns replicated by an exposure dose of 230

mJ/cm2 using opaque patterns with a width of 200 μm. The scale

mark of 20 μm is effective only in horizontal direction. Concave

depths are displayed by colors.

(b) Micro-lens pattern finally replicated to epoxy resin.

Figure 12: Comparison of resist mold patterns and faithfully

replicated epoxy micro-lens patterns.

New Fabrication Method of Plastic Micro-Lens Arrays for Researching on Compound Eyes of Insects

Figure 13: Comparison of profile curves between resist-

mold patterns and faithfully replicated epoxy micro-lens

patterns.

Table 1: Size differences between resist mold pattern and

resin lens pattern.

5 RATIONALITY OF CONCAVE

PATTERNING

Good balance of simplicity and accuracy of the new

method depend on the lithography process for

printing concave resist patterns. The key technology

is projection exposure under intentional large defocus

conditions.

In the past research, a similar method was used for

printing SU-8 patterns with vertical side walls and

very high aspect ratios (Hirota et al., 2003). In that

case, the negative resist SU-8 was a highly

transparent material, and the defocus was given for

making the widths of light intensity distribution

curves at the resist bottom equal to those at the resist

surface. However, in this paper, positive resist of

PMER P-LA900PM with large absorption was used,

and the pattern images were made vague even at the

surface. Using such translucent resist materials and

vague and gentle light intensity distributions, smooth

concave patterns with favourable quasi-spherical

profiles were obtained.

6 CONCLUSION

After investigating the typical compound eyes of

insects, plastic micro-lens arrays with similar lens

parameters were actually fabricated by developing a

new method. The aim is to prepare for developing

artificial compound eyes in the future. Because actual

element lens sizes of typical insects were in a range

of 16-53 μm, how to obtain micro-lens arrays with

such sizes simply and easily were investigated.

Various methods for fabricating micro-lens arrays

have already been proposed in the world. However, a

method that was simpler and more inexpensive was

necessary. In addition, it was preferable to make

possible to fabricate lens arrays by ourselves, and

change lens parameters such as diameters, curvature

radiuses and shapes freely.

In the new method, plastic micro-lens arrays were

fabricated by lithographically printing resist-mold

patterns, and faithfully replicating them to epoxy

resin. It was demonstrated that original concave resist

patterns were simply formed by only one lithography

process using intentionally defocused projection

exposure. In spite of the simplicity, curvature radiuses

of resist-mold patterns were widely controllable by

changing exposure dose. In addition, it was also

demonstrated that the resist-mold patterns were

faithfully replicable to epoxy resin by pouring the

resin onto the resist molds, and peeling a wafer chip

with the resist-mold patterns off mechanically after

the epoxy resin was hardened. Thus, convex lens

arrays of epoxy resin arranged in honeycomb styles,

and with an element shape of hexagon were

successfully fabricated.

There are still some subjects. The main subject is

the development of a better method for fabricating

lens arrays with precisely controlled thickness and

designed outline shapes. It is necessary to get down

to work hereafter.

ACKNOWLEDGEMENTS

This work was partially supported by Research

Institute for Science and Technology of Tokyo Denki

University, Grant Number Q15T-03.

REFERENCES

Bowden, N., Brittain, S., Evans, A. G. John et al., 1998.

Spontaneous formation of ordered structures in thin

films of metals supported on an elastomeric polymer,

NATURE 393 (1998) 146-149.

Resist-mold

pattern

Epoxy resin

pattern

Radius

(μm)

Depth

(μm)

Error

(μm)

32.4 4.2 0.1

31.3 4.1 0.1

BIODEVICES 2016 - 9th International Conference on Biomedical Electronics and Devices

Cao, Z., Zhai, C., Wang, K., 2015. Design of artificial

spherical super position compound eye, Optics

Communications 356, 218-222.

Chen, F., Liu, H., Yang, Q. et al., 2010. Maskless

fabrication of concave microlens arrayson silica glasses

by a femtosecond-laser-enhanced local wet etching

method, OPTICS EXPRESS 18, 20334- 20343.

Chen, H., Shen, X., Li, X., Jin, Y., 2011. Bionic Mosaic

Method of Panoramic Image Based on Commpound

Eye of Fly, Journal of Bionic Engineering 8, 440-448.

Cheng, H. C., Huang, C. F., Lin, Y. et al., 2010. Brightness

field distributions of microlens arrays using micro

molding, OPTICS EXPRESS 18, 26888-26904.

Chiu, C. C., Lee, Y. C., 2011. Fabricating of aspheric

micro-lens array by excimer laser micromachining,

Optics and Lasers in Engineering 49, 1232–1237.

Deng, Z., Chen, F., Yang, Q. et al., 2012. A facile method

to fabricate close-packed concave microlens array on

cylindrical glass, J. Micromech. Microeng. 22, 115026

(7pp).

Duparré, J., Dannberg, P., Schreiber, P., Bräuer, A., and

Tünnermann, A., 2005. Thin compound-eye camera,

APPLIED OPTICS 44 , 2949-2956.

Genzer, J. and Groenewold, J., 2006. Soft matter with hard

skin: From skin wrinkles to templating and material

characterization, Soft Matter 2 (2006) 310–323.

Hirota, K., Ozaki, M. and Horiuchi, T., 2003. Low-cost and

High-Performance Micro-Fabrication Method Using

Low Numerical-Aperture Optical Projection

Lithography on Copper-Clad Plastic Boards, Japanese

Journal of Applied Physics 42, 4031-4036.

Jeong, K. H., Kim, J., Lee, L. P., 2006. Biologically

Inspired Artificial Compound eye, Science 312, 557-

561.

Jiang, W., Hu, S., He, Y., Bu, Y., 2015. An artificial

compound eye of photon sieves, Optics & Laser

Technology 74, 93-96.

Kim, J. Y., Brauer, N. B., Fakhfouri, V. et al., 2011. Hybrid

polymer microlens arrays with high numerical

apertures fabricated using simple ink-jet printing

technique, OPTICAL MATERIALS EXPRESS 1, 259-

269.

Kuang, D., Zhang, X., Gui, M., and Fang, Z., 2009.

Hexagonal microlens array fabricated by direct laser

writing and inductively coupled plasma etching on

organic light emitting devices to enhance the

outcoupling efficiency, APPLIED OPTICS 48, 974-

978.

Li, F., Chen, S., Luo, H., Gao, Y., 2013. Curved micro lens

array for bionic compound eye, Optik 124,1346-1349.

Liu, K.H., Chen, M.F., Pan, C.T. et al., 2010. Fabrication of

various dimensions of high fill-factor micro-lens arrays

for OLED package, Sensors and Actuators A 159, 126–

134.

Luoa, Y., Wanga, L., Dinga, Y. et al., 2013. Direct

fabrication of microlens arrays with high numerical

aperture by ink-jetting on nanotextured surface, Applied

Surface Science 279, 36– 40.

Ma, K. Y., Chirarattananon, P., Fuller, S. B., Wood, R. J.,

2013. Controlled Flight of a Biologically Inspired,

Insect-Scale Robot, SCIENCE 340, 603-607.

Neu, S. et al., 2015. Excellent Structure-Based

Multifunction of Morpho Butterfly Wings: A Review,

Journal of Bionic Engineering 12, 170–189.

Niu, S., Li, B., Mu, Z. et al., 2015. Excellent Structure-

Based Multifunction of Morpho Butterfly Wings: A

Review, Journal of Bionic Engineering 12, 170–189.

Ogata, S., Ishida, J., Sasano, T., 1994. Optical sensor array

in an artificial compound eye, Optical Engineering, 33,

3649-3655.

Pan, C.T., Su, C.H., 2007. Fabrication of gapless triangular

micro-lens array, Sensors and Actuators A 134, 631–

640.

Saito, A., Yoshioka, S., Kinoshita, S., 2004. Reproduction

of the Morpho butterfly's blue: arbitration of

contradicting factors, SPIE 5526 (2004) 188-194.

Sanders, J. S., Halford, C. E., 1995. Design and analysis of

apposition compound eye optical sensors, Optical

Engineering, 34, 222-235.

Tanida, J., Kumagai, T., Yamada et al., 2001. Thin

observation module by bound optics (TOMBO):

concept and experimental verification, APPLIED

OPTICS, 40, 1806-1813.

Voigt, A., Ostrzinskia, U., Pfeiffer, K. et al., 2011. New

inks for the direct drop-on-demand fabrication of

polymer lenses, Microelectronic Engineering 88,

2174–2179.

Wang, S. Y., 2005. Computer simulation for the fabrication

of hexagonal micro lens arrays by use of the dragging

process with an excimer laser, OPTICS EXPRESS 13,

5600-5607.

Wu, H., Odom, T. W. and Whitesides, G. M., 2002.

Reduction Photolithography Using Microlens Arrays:

Applications in Gray Scale Photolithography,

Analytical Chemistry, 74, 3267-3273.

Yang, J. J., Liao, Y. S., Chen, C. F., 2007. Fabrication of

long hexagonal micro-lens array by applying gray-scale

lithography in micro-replication process, Optics

Communications 270, 433–440.

Zhu, X., Zhun, L., Chen, H. et al., 2015. Fabrication of high

numerical aperture micro-lens array based on drop-on-

demand generating of water-based molds, Optics &

Laser Technology 68, 23–27.

New Fabrication Method of Plastic Micro-Lens Arrays for Researching on Compound Eyes of Insects