Self-Assembly of Amorphous Calcium Carbonate Microlens Arrays
Self-Assembly of Amorphous Calcium Carbonate Microlens Arrays
Self-Assembly of Amorphous Calcium Carbonate Microlens Arrays
Received 1 Aug 2011 | Accepted 1 Feb 2012 | Published 6 Mar 2012 DOI: 10.1038/ncomms1720
Biological materials are often based on simple constituents and grown by the principle of self-
assembly under ambient conditions. In particular, biomineralization approaches exploit efficient
pathways of inorganic material synthesis. There is still a large gap between the complexity
of natural systems and the practical utilization of bioinspired formation mechanisms. Here
we describe a simple self-assembly route leading to a CaCO3 microlens array, somewhat
reminiscent of the brittlestars’ microlenses, with uniform size and focal length, by using
a minimum number of components and equipment at ambient conditions. The formation
mechanism of the amorphous CaCO3 microlens arrays was elucidated by confocal Raman
spectroscopic imaging to be a two-step growth process mediated by the organic surfactant.
CaCO3 microlens arrays are easy to fabricate, biocompatible and functional in amorphous or
more stable crystalline forms. This shows that advanced optical materials can be generated by
a simple mineral precipitation.
1 Department of Biomaterials, Max Planck Institute of Colloids and Interfaces, Potsdam 14424, Germany. 2 CO Sequestration Research Department,
2
Korea Institute of Geoscience and Mineral Resources, Daejeon 305-350, South Korea. 3 Department of Chemical and Biomolecular Engineering, KAIST,
Daejeon 305-701, South Korea. 4 Physical Chemistry, University of Konstanz, Konstanz D-78457, Germany. 5 Biomedical Engineering Division, Centre
for Biomedical Research, School of Bio-Sciences and Technology, VIT University, Vellore, Tamilnadu 632 014, India. Correspondence and requests for
materials should be addressed to P.F. (fratzl@mpikg.mpg.de).
N
atural materials often grow by biologically controlled self- microlenses with uniform size and focal length by using a saturated
assembly processes under ambient conditions1,2. Despite calcium solution and CO2 in air, together with a widely used sur-
using easily accessible materials only, they often show factant as the base materials. To understand the formation mecha-
remarkable functional properties. For example, biominerals in nism of the microlens structures, confocal Raman spectroscopic
many skeletons are strong and have a crucial role in supporting imaging was carried out to characterize the three main components:
and/or protecting the bodies of living organisms. Some biomin- carbonate, water and organic. The role of the organic molecules, act-
erals also function as optical devices in organisms. This is seen in ing as a surface-active and structure-directing agent, is discussed
glass sponges and brittlestars3,4, for example, where optical fibres and we draw conclusions on how the uniformity of the structures
and microlenses show exceptional optical performances originating is achieved. Furthermore, this article reports on the biocompatibil-
from optimized biomineral shape or crystal orientation. ity as well as the optical properties of the microlens arrays at the
Biomineralization has been widely studied to first understand micro- and macroscopic scale.
how inorganic materials are produced under the control of organic
molecules in nature and second to apply this knowledge in synthetic Results
systems5–9. There have been reports on the synthetic fabrication of Synthesis and formation of the CaCO3 microlens arrays. CaCO3
CaCO3 structures, which mimic biominerals via templating; such precipitates start to form and agglomerate at the interface of air
as the replica of the sea urchin skeletal plate10,11, the micropat- and saturated Ca(OH)2 solution right after the solution reacts
terned single calcite crystal12 and periodic optical nanostructures with CO2 in air. The synthesis process is schematically described
of inverse opal13. These studies have demonstrated that biom- in Fig. 1a. This surface aggregation goes on for 15–20 min with
ineral structures could be reconstructed by using a biomimetic Brownian motion (Supplementary Movie 1). The scanning electron
mineralization approach and showed how to utilize the most repre- microscopy (SEM) images in Fig. 1b,c clearly show the nanometre-
sentative characteristics of biomineralization, namely, amorphous sized CaCO3 precipitates adhered to the agglomerates, as indicated
precursor pathways and templates14–16. These synthetic procedures by white arrows. The agglomerates attain a quasi-hemispherical
remain restricted to small sample sizes, are still very complex and shape and diameter of up to 2.7 µm within 2 min (Fig. 1b,c). The
also require further steps for preparing or removing biological or CaCO3 agglomerates are also formed in the absence of surfactant
artificial templates. or monolayers, although the morphology is not uniform in this case
Various fabrication methods combined with lithographic tech- (Supplementary Fig. S1a). The role of the surfactant, polysorbate
niques have been demonstrated to be practical for the synthesis of 20 (non-ionic, or Tween 20, PS 20), composed of fatty acid esters
microlens arrays17–22. However, these techniques are limited to the of polyoxyethylene sorbitan, is to regulate the growth of CaCO3
fabrication of microlens arrays on flat substrates and require mul- agglomerates into a uniform size and shape (Supplementary Fig. S1).
tiple steps, such as baking, developing and etching. Furthermore, PS 20 molecules have adsorbed onto the CaCO3 precipitates
the aforementioned approaches yield to microlens arrays made of and agglomerates at the early stage of growth based on a time-
organic materials with comparatively low refractive indices. How- dependent surface tension measurement, as depicted in Fig. 1a. In
ever, it is often desirable to have a high refractive index and short these measurements, the surface tension of water containing 10 µM
focal lengths. This can be achieved using inorganic materials. of PS 20 ranged from 40 to 55 mN m − 1 (Supplementary Fig. S2),
Here we show how these microlens arrays can be synthesized by demonstrating that the surfactant was located at the air–water
a simple mineral precipitation process without any template and interface. However, the surface tension of Ca(OH)2 solution with
at ambient conditions. In this work, we prepared arrays of CaCO3 the same concentration of the surfactant was close to that of pure
PS 20 + Ca(OH)2 solution
Self-assembly of
12.52 6
PS 20-adsorbed ACC precipitates
Size (µm)
12.50 4
pH
1–2 min
2
12.48
0
0 20 40 60
Time (min)
Figure 1 | Formation of CaCO3 microlens array. (a) Schematic illustration for the formation of CaCO3 microlens array. (b,c) SEM images of CaCO3
agglomerates after 1–2 min of reaction. The scale bars are 1 µm. The inset of c is a magnified image of the white dotted box. The scale bar is 200 nm.
White arrows indicate the nanometre-sized CaCO3 precipitates. (d) In situ optical microscopy images observed on the surface of Ca(OH)2 solution versus
reaction time. The scale bar is 10 µm. (e) The changes of both pH (red dot) in Ca(OH)2 solution and averaged size (green rectangle) of CaCO3 microlens
with standard deviation plotted versus time (statistics from size measurement of 150 microlenses for each time point, 50 microlenses from each of three
different samples).
water, 70 mN m − 1, indicating that only a few surfactant molecules the focal length of the microlenses, we measured their thickness by
were located at the air–solution interface. This clearly demonstrates atomic force microscopy line scans and the distance from the back
that most of the surfactant must adsorb onto the precipitates and plane of the microlenses and the focused point (Supplementary
the quasi-hemispherical agglomerates. Indeed, considering the Fig. S5). We measured a focal length of 7.2 ± 0.3 µm. The focal length
critical micelle concentration of PS 20 (80 µM at 21 °C), hydrophilic of the microlenses focusing in the glass coverslip was calculated
groups of the non-ionic surfactant are likely to be adsorbed on the using the relation,
hydrophilic surface of CaCO3 (ref. 23).
As the CaCO3 agglomerates grow, the floating agglomerates self- f = nglass R /(nACC − 1) (1)
assemble to form a close packed two-dimensional array and are
linked together by further growth. The growth of these agglomerates
was monitored in real time by optical microscopy in transmission where R = 3.1 ± 0.2 µm is the radius of curvature calculated based on
mode for 1 h, and the images obtained at different reaction times are the geometry of 11 microlenses measured by atomic force micros-
shown in Fig. 1d. The size of the CaCO3 agglomerates and pH change copy line scans, and nglass and nACC are refractive indices of glass
of the solution plotted versus the reaction time is shown in Fig. 1e. (1.50) and ACC (1.58, taken from the measured value of 1.5791–
The growth speed, estimated from the slope of the agglomerate size 1.5830 by Merten and group27), respectively. Entering these param-
as a function of time (Fig. 1e), fits to the corresponding decrease eters into equation (1), we find f = 8.0 ± 0.5 µm.
in pH. As shown below, the proton concentration increases when
carbonate ions are combined with Ca ions to form precipitates24: Structural characterization by Raman spectroscopy. The CaCO3
Ca 2+ + CO 2 + H 2 O→Ca 2+ + H 2 CO 3 →Ca 2+ + H + + HCO 3 − → microlenses show structural complexity as found by depth scans
Ca2+ + 2H + + CO3 − →2H + + CaCO3. with confocal Raman spectroscopic imaging (Fig. 4), revealing
The morphology of CaCO3 microlens arrays after 60 min of the distribution of carbonate, water and organic components.
growth is represented by a hexagonally packed array of hemispheri- A schematic illustration of a depth scan is shown in Fig. 4a.
cal structures of homogeneous size and shape (Fig. 2a,b). The mic-
rolenses are connected to each other forming a film, as shown by the
SEM image of sectionally cut CaCO3 microlens arrays by focused
ion beam milling (Fig. 2b).
Max. Max.
y = 18 µm
Min. Min.
y
z x
Collimated laser
z
y Air
x
z = 16 µm
Focal
point
Glass
Oil
x = 18 µm
Objective lens
Figure 3 | Characterization of the focal length of the microlens array. (a,b) Fluorescence and brightfield confocal images of the microlens array and
(c,d) characterization of the focal length of the microlenses. (a) x–y plane image of the microlens array resulting from the overlay of the brightfield image
(grey scale) showing the microlenses, and the fluorescence image (green) showing the fluorescently labelled chitosan coating of the microscope coverslip.
The brightfield and fluorescence images were recorded simultaneously by illumination/excitation at 488 nm and collection at 488 and 530 nm ( ± 25 nm),
respectively. The circular dashed line shows the localization of a microlens in the array. (b) x–z plane image that represents a cross-section through a stack
of images taken at different depths in the sample. The plane shown corresponds to the position indicated by a solid white line on image a. The thick dashed
line is aligned with the fluorescence from the chitosan coating and corresponds to the glass surface. The thin dashed line (semicircle) indicates the microlens
curved surface. (c) x–y plane image of the light from a collimated laser beam (650 ± 10 nm) focused by the microlenses, at a depth in the sample equal to
their focal length. Image c corresponds to the depth at which a dashed line on image d has been drawn. (d) x–z plane image of the light from a collimated
laser beam (650 ± 10 nm) focused by the microlenses. The image results from a cross-section taken through a stack of images recorded at different depth
in the sample. The plane shown corresponds to the position indicated by a solid white line on image a. The thin dashed vertical lines in a and b (c and
d, respectively) are a guide to the eye to match the images in the x-direction. The diagram in the middle of the figure shows a schematic of the confocal
microscope with one microlens (grey filled semicircle) on a chitosan-coated (green line) cover glass (white box) and the beam path (red dotted lines).
Figure 4b shows confocal Raman images of carbonate (spectral The organic component is mainly distributed inside the microlenses
region 1,040–1,125 cm − 1) at different depths. The Raman spectra at a depth of 1 µm (Fig. 4d). The water peak reduces in intensity at
in this band show the characteristics of ACC (Supplementary Fig. exactly the same position (Fig. 4c). It is noticeable that the organic
S6). In the ACC, the band slightly shifts to lower wavenumbers rings are located inside the microlens structures with diameters of
and is characterized by a significant broadening (full-width at half- 2–3 µm, which means they were formed during the early stages of
maximum 28 cm − 1 in our experimental results), compared with microlens formation (within 2 min, Fig. 1e). This is consistent with
that found in calcite28,29. However, the intensity of carbonate varies the result of the surface tension measurement that most of the sur-
strongly through different z-section planes. In addition, at the factant must be adsorbed on CaCO3 precipitates and agglomerates
depth positions from 0 to 0.5 µm, the Raman scattering of carbon- right after the reaction starts (Supplementary Fig. S2).
ate is higher at the edge than in the inner parts of the microlenses
(Fig. 4b). This could be due to the preferential orientation of the Discussion
carbonate units in the near-edge region of each microlens, as dem- We deduce a two-step growth of the ACC microlenses based on the
onstrated by polarized Raman imaging with different polarization observations of microlens-sizes as a function of time and on molecu-
direction (Fig. 4e)30. The direction of the incident laser polarization lar detection by Raman imaging. The width of the anisotropic edge
was found to have no influence on the intensity of spectra in the in Fig. 4e is 1–2 µm, which is similar to the grown size during the last
inner part of the microlenses (Supplementary Fig. S6). However, the 40–50 min in Fig. 1e. This is a very slow growth in comparison with
upper/lower edges of the microlenses shows higher intensity with the 4–5 µm of growth that occurs in the first 10–20 min. The slow
90° polarization with respect to 0° and vice versa for the left/right growth at the edges of the microlenses could be explained by the for-
edges (Fig. 4e and Supplementary Fig. S6). mation from ion constituents or equivalent small molecules after the
Confocal Raman imaging furthermore shows the molecular dis- fast growth is obtained by aggregation of ACC nanometre-sized pre-
tribution of water and organic components in the microlens arrays cipitates by Brownian collision. The distribution of organic matter,
(Fig. 4c,d). Thermogravimetric analysis gives the quantitative infor- visible by the green ring in Fig. 4d, reflects the fact that the early pre-
mation that 1.46 wt% of organic is included in the hydrated ACC, cipitates agglomerate together with surfactant molecules to form the
which also contains about 1 mol of water (Supplementary Fig. S7). first nucleus of a microlens at the initial stage of growth at 1–2 min.
550
CCD
cts
300
CCD
cts
400
CCD
cts
0° 90°
Figure 4 | Confocal Raman spectroscopic imaging of the CaCO3 microlens array. (a) Schematic illustration of the depth scan. (b–d) Depth scanned
Raman imaging obtained by integrating over the wavenumber ranges of (b) carbonate (1,040–1,125 cm − 1), (c) water (3,000–3,500 cm − 1) and (d) organic
components (2,800–3,000 cm − 1), respectively. The images in the same column indicate the same depth. (e) Raman imaging of carbonate at 0° and 90°
polarization of incident laser light. All the scale bars are 5 µm. CCD cts, charge-coupled device counts.
The adsorbed surfactant is thought to have an essential role in the interaction—in general being the hallmark of biomineralization or
formation of the final CaCO3 structures with uniform shape and size. biomimetic mineralization6,34—remain to be further investigated
Further growth of the microlenses occurs by the accretion of precipi- in our system. One possible interpretation would be that the lip-
tates without organic components, which—at this stage—are already ids of surfactant molecules stabilize the amorphous structure in the
exhausted in the solution. It is notable that the complex inner struc- centre of the microlenses16,32. The carbonate groups are being more
ture of the microlenses (visible in Fig. 4) does not affect the optical oriented in the outer parts of the microlenses, which might also be
images of the ‘A’s, as shown in Fig. 2d. Therefore, we deduce that the related with a directional guidance by the organic molecules form-
inclusion of the organic phase (1.46 wt% as determined by thermo- ing a ring pattern. Indeed, the alkyl chain of self-assembly monol-
gravimetric analysis) does not distort the optical light path signifi- ayers has been reported to be involved in the oriented growth of
cantly. Moreover, the focal distance of the microlenses does not seem calcite35,36.
to be considerably influenced by the organic inclusions (Fig. 3). The size of CaCO3 microlens array films only depends on the
The visualization of the complex structure in which organic com- area of the interface between precursor solution and air, which
ponents, amorphous and/or crystalline CaCO3 phases coexist is not means there is basically no limitation for the fabrication of large film
an easy task8,14,31–33. Aizenberg et al.31,32 have shown the combina- areas. Figure 5 shows that CaCO3 microlens arrays can be attached
tion of the two CaCO3 phases separated by an organic membrane in to gently curved surfaces as well as to flat ones on the centimetre-
an ascidian skeleton by selective etching. In addition, several studies scale by using chitosan as glue. When the film of the CaCO3 mic-
have reported how to visualize the distribution of organic compo- rolens array is coated on a cover glass, the repeated micron-sized
nents in synthetic or biogenic CaCO3 matrices8,33. In the present convex structures prevent incident light from reflecting and keep
work, Raman spectroscopic imaging successfully shows how the the transparency as shown in Fig. 5a,b. This method can be used to
amorphous phase of CaCO3 with preferential orientation of car- coat microlens structures onto curved surfaces similar to that found
bonates and organic matrix is distributed without destroying the on the moth eye (Fig. 5c)37, which is not possible with lithographic
mineral matrix. However, the organic–mineral interface and their techniques17–21. These are effective antireflective lenses over the
Methods
Synthesis of CaCO3 microlens arrays. Ca(OH)2 clear solution was prepared by
dissolving 1 g of Ca(OH)2 powder (Aldrich) in 100 ml millipore water, followed by
complete sedimentation of a white substance after 3 days in a sealed media bottle.
Calculated amounts of PS 20 (Aldrich) were added into the clear Ca(OH)2 solution
and stirred vigorously. A volume of 350 µl of precursor solution was dropped into
each well of a 96 microplate, whose cover was left open in the atmosphere. After 1 h
of reaction, a thin film of CaCO3 microlens array was formed on the surface of the
solution in each well. The film was skimmed off by using a microscope cover glass
after adding 50–80 µl of water in each well and thereby the surface of solution with
the film was raised over the top of the plate. The residual solution on the cover glass
was soaked up by using lint-free tissue paper and dried in air. The convex side of
the CaCO3 microlens array was arranged towards the cover glass at this stage and
the Raman characterization and the ‘A’ projection experiments were carried out
as deposited. For SEM images, the CaCO3 microlens arrays on a cover glass were
transferred to a carbon tape allowing the curved surfaces to be investigated. The films
of a CaCO3 microlens array shown in Figs 3 and 5 were grown for 3 h in each well of
12-well plates filled with 5 ml of precursor solution. After the films were transferred
to a cover glass and dried, they were attached on another cover glass or convex
quartz glass lens (15 mm of focal length) coated with 2 wt% of chitosan solution in
the same way described above, which was followed by drying under nitrogen. For the
characterization of the light path through the microlens array and its focal length, the
fluorescently labelled chitosan solution was used to detect the position of the cover
glass–microlens junction in the z-direction (Fig. 3). Once the CaCO3 microlens
Figure 5 | CaCO3 microlens arrays fixed on chitosan-coated substrates. arrays are transferred from the solution within 1–3 h and dried, the amorphous
(a,b) The pictures of the same CaCO3 microlens array-chitosan composite phase is kept up to several months in ambient atmosphere depending on the humid-
film coated on a cover glass were taken at different angles, which shows ity. Indeed, ACC synthesized in high pH precursor solution (in this study pH 12.5)
the antireflecting effect of the composite film. (a) Blue dashed line and has been reported to be more stable than that formed at a lower pH39. When the
growth of the CaCO3 microlens array proceeds for more than 4 h at the air–Ca(OH)2
red solid line indicate the locations of the cover glass and the CaCO3
solution interface, some structures are protruded on the convex surface of micro-
microlens array, respectively. (b) The picture was taken at the right angle lenses, which further grow into (104) faceted crystals (Supplementary Fig. S8)40.
when the cover glass reflects incident light. Only the area where the
CaCO3 microlens array was coated keeps transparency. (c) The picture Cell culture on microlens array. The microlens array–coated cover slides were
of the composite film coated on a convex quartz glass lens in the form of sterilized under ultraviolet light for 1 h, and NIH3T3 fibroblasts were seeded and
cultured overnight in DMEM medium (Sigma) with 10% calf serum at 37 °C inside
compound eye. The red arrow indicates the uncoated part, which reflects an incubator under humidified atmosphere. Later, microlens arrays were washed
light. Scale bars in a–c are 1 cm. (d) Overview of NIH3T3 fibroblast cell a few times in PBS, and the cells were fixed in 4% paraformaldehyde (Fluka) for
growth on the microlens array. Cells were stained with actin (green), and 5 min at room temperature. Unbound paraformaldehyde was removed by gentle
nucleus (red) with fluorescent dyes (overlay of fluorescent and phase washing in PBS several times. Cells were permeabilized with 0.1% Triton X-100
(Sigma) for 5 min and stained with 1:200 diluted Alexa Fluor 488 (Invitrogen)
contrast images). The scale bar is 100 µm. (e,f) Actin organization within
for 60 min. Nucleus staining was performed with 1:300 diluted TOPRO-3 Iodide
the cell at higher magnification (e, overlay; f, green fluorescence). The (Invitrogen) for 5 min and the microlens array cover slides were washed in PBS,
scale bars are 20 µm. mounted with Vectashield mounting medium and observed under confocal laser
scanning microscope (Leica TCS SP5, Leica).
spectrum of visible light. Moreover, microlens arrays are compat- Characterization of the focal length. The experiment has been performed on
ible for cell attachment and growth as shown in Fig. 5d. Higher a laser scanning confocal microscope (Leica DM IRBE, Leica) in two steps, as
magnification on one individual cell, covering several tens of mic- depicted in Supplementary Fig. S9. First, 488-nm laser light was focused on the
sample using a microscope objective (oil ×100, 1.4 NA, Leica) and scanned across
rolenses, shows the actin organization within the cell in response to the field of view. The fluorescence was collected by the objective (epifluorescence
the arrangement of the microlens array (Fig. 5e,f). Actin fibres are mode) and detected by a photomultiplier tube (PMT) through a dichroic mirror
observed along the cell boundary and also around the microlens (RSP 500, Leica). The signal was spectrally selected by a monochromator set to col-
structures as the cell can feel the topography and adapt its cytoskel- lect wavelengths between 505 and 560 nm. One image contained 1,024×1,024 pixels
eton to the geometry of the microlens array38. This biocompatibility and a final image was created by an average of four recorded images. The sample
was scanned at decreasing height by steps of 0.2 µm. A total of 80 images was
of the CaCO3 material enables the presented microlens arrays to recorded and used to generate a stack of 16 µm in depth. Second, a laser pointer
interact with a biological environment in a similar way as its arche- (640–660 nm, Hama laserpointer LP-18, Hama) was secured above the microscope
type—the brittlestar. stage using a laboratory buret stand and an extension clamp. The collimated light
was directed perpendicular to the coverslip slide supporting the microlens array 24. Gebauer, D., Volkel, A. & Colfen, H. Stable prenucleation calcium carbonate
and collected by the microscope objective. Alignment of the laser beam was ad- clusters. Science 322, 1819–1822 (2008).
justed by reflection of part of the light by the microscope slide surface supporting 25. Lu, Y., Yin, Y. D. & Xia, Y. N. A self-assembly approach to the fabrication
the sample. The monochromator was set to select light between 630 and 670 nm, of patterned, two-dimensional arrays of microlenses of organic polymers.
and an image stack was generated as described above. Adv. Mater. 13, 34–37 (2001).
26. Gu, E. et al. Reflection/transmission confocal microscopy characterization
Characterization methods. The surface tension measurement was carried out us- of single-crystal diamond microlens arrays. Appl. Phys. Lett. 84, 2754–2756
ing a DuNouy Tensionmeter (KRÜSS Tensionmeter K12). The projected images of (2004).
‘A’ through the CaCO3 microlens arrays were obtained by using optical microscopy 27. Helmut, L. & Merten, G. L. B. Stabilized amorphous calcium carbonate.
(Leica DM RXA2, Leica; fitted with a digital colour camera, Leica DFC 480, Leica) US patent no. 4237147 (1980).
set with microlens array and screen ‘A’ on the object stage in transmission mode. 28. Gunther, C., Becker, A., Wolf, G. & Epple, M. In vitro synthesis and structural
For confocal Raman microspectroscopy (WITec alpha300R, WITec), the selected characterization of amorphous calcium carbonate. Z. Anorg. Allg. Chem. 631,
area of the sample was scanned with a continuous green laser beam with lateral 2830–2835 (2005).
and depth resolutions of 250 and 500 nm, respectively. Raman images have been 29. Tlili, M. M. et al. Characterization of CaCO3 hydrates by micro-Raman
generated by integrating the intensity of the signal for the wavenumber ranges spectroscopy. J. Raman Spectrosc. 33, 10–16 (2002).
of carbonate (1,040~1,125 cm − 1, Fig. 3a), water (3,000~3,500 cm − 1, Fig. 3b) and 30. Gierlinger, N. et al. Cellulose microfibril orientation of Picea abies and its
organic (2,800~3,000 cm − 1, Fig. 3c). variability at the micron-level determined by Raman imaging. J. Exp. Bot. 61,
587–595 (2010).
References 31. Aizenberg, J., Lambert, G., Weiner, S. & Addadi, L. Factors involved in the
1. Lakes, R. Materials with structural hierarchy. Nature 361, 511–515 (1993). formation of amorphous and crystalline calcium carbonate: a study of an
2. Fratzl, P. Biomimetic materials research: what can we really learn from nature′s ascidian skeleton. J. Am. Chem. Soc. 124, 32–39 (2002).
structural materials? J. R. Soc. Interface 4, 637–642 (2007). 32. Aizenberg, J., Weiner, S. & Addadi, L. Coexistence of amorphous and
3. Sundar, V. C., Yablon, A. D., Grazul, J. L., Ilan, M. & Aizenberg, J. Fibre-optical crystalline calcium carbonate in skeletal tissues. Connect. Tissue Res. 44, 20–25
features of a glass sponge—some superior technological secrets have come to (2003).
light from a deep-sea organism. Nature 424, 899–900 (2003). 33. Li, H. Y., Xin, H. L., Muller, D. A. & Estroff, L. A. Visualizing the 3D internal
4. Aizenberg, J., Tkachenko, A., Weiner, S., Addadi, L. & Hendler, G. Calcitic structure of calcite single crystals grown in agarose hydrogels. Science 326,
microlenses as part of the photoreceptor system in brittlestars. Nature 412, 1244–1247 (2009).
819–822 (2001). 34. Lowenstam, H. A. & Weiner, S. On Biomineralization (Oxford University Press,
5. Heuer, A. H. et al. Innovative materials processing strategies—a biomimetic 1989).
approach. Science 255, 1098–1105 (1992). 35. Aizenberg, J., Black, A. J. & Whitesides, G. H. Oriented growth of calcite
6. Xu, A. W., Ma, Y. R. & Colfen, H. Biomimetic mineralization. J. Mater. Chem. controlled by self-assembled monolayers of functionalized alkanethiols
17, 415–449 (2007). supported on gold and silver. J. Am. Chem. Soc. 121, 4500–4509 (1999).
7. Meldrum, F. C. & Colfen, H. Controlling mineral morphologies and 36. Han, Y. J. & Aizenberg, J. Face-selective nucleation of calcite on self-assembled
structures in biological and synthetic systems. Chem. Rev. 108, 4332–4432 monolayers of alkanethiols: effect of the parity of the alkyl chain. Angew. Chem.
(2008). Int. Ed. 42, 3668–3670 (2003).
8. Nudelman, F., Chen, H. H., Goldberg, H. A., Weiner, S. & Addadi, L. Spiers 37. Duparre, J. W. & Wippermann, F. C. Micro-optical artificial compound eyes.
memorial lecture: lessons from biomineralization: comparing the growth Bioinspir. Biomim. 1, R1–R16 (2006).
strategies of mollusc shell prismatic and nacreous layers in Atrina rigida. 38. Thery, M. Micropatterning as a tool to decipher cell morphogenesis and
Faraday Discuss. 136, 9–25 (2007). functions. J. Cell Sci. 123, 4201–4213 (2010).
9. Meldrum, F. C. Calcium carbonate in biomineralisation and biomimetic 39. Koga, N., Nakagoe, Y. Z. & Tanaka, H. Crystallization of amorphous calcium
chemistry. Int. Mater. Rev. 48, 187–224 (2003). carbonate. Thermochim. Acta 318, 239–244 (1998).
10. Park, R. J. & Meldrum, F. C. Synthesis of single crystals of calcite with complex 40. Lee, K. B., Park, S. B., Jang, Y. N. & Lee, S. W. Morphological control of
morphologies. Adv. Mater. 14, 1167–1169 (2002). CaCO3 films with large area: effect of additives and self-organization under
11. Cheng, X. G. & Gower, L. B. Molding mineral within microporous hydrogels atmospheric conditions. J. Colloid Interface Sci. 355, 54–60 (2011).
by a polymer-induced liquid-precursor (PILP) process. Biotechnol. Prog. 22,
141–149 (2006).
12. Aizenberg, J., Muller, D. A., Grazul, J. L. & Hamann, D. R. Direct fabrication of Acknowledgements
large micropatterned single crystals. Science 299, 1205–1208 (2003). We thank J. Dunlop, S.-H. Kim, S.-M. Park, L. Bertinetti and P. Zaslansky for fruitful
13. Li, C. & Qi, L. M. Bioinspired fabrication of 3D ordered macroporous single discussions. We also thank A. Verch, C. Pilz and A. Heilig from the Max Planck Institute
crystals of calcite from a transient amorphous phase. Angew. Chem. Int. Ed. 47, of Colloids and Interfaces for technical assistance. The Brain Korea 21 is acknowledged
2388–2393 (2008). for funding. A.M. and P.F. are grateful for support by the Alexander von Humboldt
14. Addadi, L., Raz, S. & Weiner, S. Taking advantage of disorder: amorphous Foundation and the Max Planck Society in the framework of the Max Planck Research
calcium carbonate and its roles in biomineralization. Adv. Mater. 15, 959–970 Award by the Federal Ministry of Education and Research.
(2003).
15. Colfen, H. Single crystals with complex form via amorphous precursors.
Author contributions
Angew. Chem. Int. Ed. 47, 2351–2353 (2008).
S.-W.L. developed the synthetic method. K.L. conceived the idea and performed
16. Gower, L. B. Biomimetic model systems for investigating the amorphous
experiments. I.M., S.B.P., H.C., W.W. and P.F. supervised the study. A.M. performed
precursor pathway and its role in biomineralization. Chem. Rev. 108,
Raman spectroscopic experiment, analysed data and contributed in writing the
4551–4627 (2008).
corresponding part of the article. K.P.K. and M.B. performed experiments of cell culture
17. Wu, H., Odom, T. W. & Whitesides, G. M. Generation of chrome masks
and focal length measurement, respectively, and wrote the regarded parts of article. K.L,
with micrometer-scale features using microlens lithography. Adv. Mater. 14,
W.W. and P.F. wrote the article. All authors discussed the results and commented on the
1213–1216 (2002).
article.
18. Wu, M. H., Park, C. & Whitesides, G. M. Fabrication of arrays of
microlenses with controlled profiles using gray-scale microlens projection
photolithography. Langmuir 18, 9312–9318 (2002). Additional information
19. Yang, S. et al. Functional biomimetic microlens arrays with integrated pores. Supplementary Information accompanies this paper at http://www.nature.com/
Adv. Mater. 17, 435–438 (2005). naturecommunications
20. Yang, S., Ford, J., Ruengruglikit, C., Huang, Q. R. & Aizenberg, J. Synthesis
of photoacid crosslinkable hydrogels for the fabrication of soft, biomimetic Competing financial interests: The authors declare no competing financial interests.
microlens arrays. J. Mater. Chem. 15, 4200–4202 (2005). Reprints and permission information is available online at http://npg.nature.com/
21. Yang, S., Ullal, C. K., Thomas, E. L., Chen, G. & Aizenberg, J. Microlens arrays reprintsandpermissions/
with integrated pores as a multipattern photomask. Appl. Phys. Lett. 86, 201121
(2005). How to cite this article: Lee, K. et al. Self-assembly of amorphous calcium carbonate
22. Aizenberg, J. & Hendler, G. Designing efficient microlens arrays: lessons from microlens arrays. Nat. Commun. 3:725 doi: 10.1038/ncomms1720 (2012).
nature. J. Mater. Chem. 14, 2066–2072 (2004).
23. Paria, S. & Khilar, K. C. A review on experimental studies of surfactant License: This work is licensed under a Creative Commons Attribution-NonCommercial-
adsorption at the hydrophilic solid-water interface. Adv. Colloid Interface Sci. Share Alike 3.0 Unported License. To view a copy of this license, visit http://
110, 75–95 (2004). creativecommons.org/licenses/by-nc-sa/3.0/