Materials Letters 222 (2018) 156–159

Contents lists available at ScienceDirect

Materials Letters
journal homepage: www.elsevier.com/locate/mlblue

Simple preparation of hydroxyapatite nanostructures derived from fish

scales
Yadong Chai, Motohiro Tagaya ⇑
Department of Materials Science and Technology, Nagaoka University of Technology, Kamitomioka 1603-1, Nagaoka, Niigata 940-2188, Japan

a r t i c l e i n f o a b s t r a c t

Article history: The nanostructured hydroxyapatite (HAp) formation in fish scales were investigated. Calcium and phos-
Received 26 June 2017 phate ions were effectively accumulated in the scales by the immersion processes. The HAp nanoparticles
Received in revised form 6 March 2018 arranged in one direction in the internal fibrillary plate layers were observed, revealing the crystal
Accepted 2 April 2018
growth along with the collagen fibers to resultantly form the porous structures. Thus, the arrangement
Available online 3 April 2018
of collagen fibers inside the scales effectively form the HAp nanostructures.
Ó 2018 Elsevier B.V. All rights reserved.
Keywords:
Biomaterials
Bioceramics
Interfaces
Porous materials
Nanocomposites
Texture

1. Introduction
Hydroxyapatite (HAp, Ca10(PO4)6(OH)2) has attracted attention
because of their potential uses in biomedical fields [1,2]. The bio-
compatible properties can be enhanced by the inorganic/organic
hybrids [3,4]. Accordingly, an important point is to investigate
the HAp formation under the existence of functional organic mole-
cules [5,6]. The HAp forms gradually the charged a- and c- crystal
planes on the organic groups at the nanoscales, and the resultant
crystals can be controlled to have effective interfacial interactions
[7,8], leading to novel synthetic techniques based on the
biomimetics. Although the organic molecule-based approach has
been applied to the preparation of the nano-sized HAp crystals
for the in-growth of natural bones [9–12], the natural products-
based nano-hybrid formation has been unclear.
Scales of teleost fish compose the HAp and collagen [13]. Three
layer types in the scales are distinct; the outer limiting layer, the
external osseous layer and the internal fibrillary plate layer [14].
In the external layer, collagen fibers are randomly oriented. In
the internal layer, the collagen fibers with the diameter of 60–
100 nm are highly-ordered arrangement and are organized as the
lamellae [15]. Although needlelike HAp crystals in random orienta-
tion are observed in the external layer [16], the orientation of the c-
axis of HAp crystals parallel to collagen fibers in the calcification
⇑ Corresponding author.
E-mail address: tagaya@mst.nagaokaut.ac.jp (M. Tagaya).
process of the internal fibrillary plate is observed [17]. The crystal-
lographic orientation of HAp along with the collagen fibers of the
fish scales can be replicated within the calcined products.
In this study, the nanostructural HAp formation templated in
the hierarchical inorganic/organic hybrid structure of the natural
fish scales to form the porous structures by the calcination process
were investigated and carried out to speculate the interfacial inor-
ganic/organic interactions.
2. Materials and methods
Alternate soaking method was used [18] to densely form HAp in
the fish scales. The dry fish scales (Japan Tuna Bait Co., Ltd.) were
transferred in H2O. There were three processes as follows; the fish
scales were washed with H2O (‘‘process 1”), and then immersed in
Na2HPO4 aqueous solution (120 mM, pH = 9.2) at 37 °C for 5 min
(‘‘process 2”). The fish scales were washed with H2O, and was
immersed in the aqueous solution (pH = 7.4) containing CaCl2
(200 mM) and tris(hydroxymethl)aminomethane (Tris, 100 mM)
at 37 °C (‘‘process 3”). The above processes 1–3 were regarded as
1 cycle. The obtained fish scales were dried at 60 °C, and calcined
at 250 °C and then at 550 °C. The fish scales repeated 0 and 5 cycles
were abbreviated as 0cy-bef and 5cy-bef without the calcination,
and the fish scales repeated 0, 5 and 10 cycles and calcined were
as 0cy, 5cy and 10cy.
The characterization was conducted upon energy dispersive X-
ray spectrometer (EDS), wavelength dispersive X-ray fluorescence

https://doi.org/10.1016/j.matlet.2018.04.009
0167-577X/Ó 2018 Elsevier B.V. All rights reserved.
 Y. Chai, M. Tagaya / Materials Letters 222 (2018) 156–159
(XRF) spectrometry, fourier transform infrared (FT-IR) spectrome-
ter, X-ray diffraction (XRD), nitrogen (N2) adsorption and desorp-
tion instrument, and field emission scanning electron microscope
(FE-SEM). In the XRD patterns, according to Scherer’s equation
(K = 0.89), crystalline size (D) were calculated from the half width
of the 002 and 300 diffractions. In the N2 adsorption and desorp-
tion, Brunauer-Emmett-Teller (BET) [19] surface areas and
Barrett-Joyner-Halenda (BJH) [20] pore sizes were calculated.
3. Results and discussion
Fig. 1 shows the FE-SEM cross-sectional images and EDS map-
ping images of 0cy-bef and 5cy-bef. In the secondary electron
images (Fig. 1(a-1) and (b-1)), the internal fibrillary plate layer in
which highly-ordered arrangement of collagen fibers are organized
as the lamella was observed.
In the representative reflection electron images (Fig. 1(a-2) and
(b-2)), the white and black parts distinctly indicated the calcifica-
tion and non-calcification areas, respectively [14]. In the internal
layer, the content of Ca and P in 5cy was higher than that in 0cy.
The results were also observed in the EDS elemental mapping
images (Fig. 1(a-3 and -4) and (b-3 and -4)) and detailed chemical
compositions (Table S1), the observed calcification areas in spot-1
(0cy) and spot-3 (5cy) in Fig. 1(a-1 and b-1) contain Ca of 44.1 and
53.4 mol% and P of 13.1 and 14.7 mol%, and the non-calcification
areas in spot-2 (0cy) and spot-4 (5cy) in Fig. 1(a-1 and b-1) contain
Ca of 0.4 and 2.6 mol% and P of 0.0 and 0.6 mol%, indicating the
increase in the Ca and P elements by the immersion treatment.
Thus, the Ca and P atoms were effectively diffused and accumu-
lated in the internal layers by the immersion.
Fig. 1. Representative FE-SEM cross-sectional (a-1 and b-1) secondary electron and (a-2 and b-2) reflection electron images and EDS (a-3 and b-3) Ca and (a-4 and b-4) P
elemental mapping images of (a-1–a-4) 0cy-bef and (b-1–b-4) 5cy-bef.
Table 1
Resultant Ca/P molar ratios, crystalline seizes and pore properties of the calcined fish scales.
Composition Crystalline sizes
Ca/P molar ratio D300 (nm)
0cy 2.00 4.7
5cy 1.95 5.2
10cy 1.90 5.9
157
From the resultant chemical compositions of the calcined fish
scales (Table S2), the molar ratios of Ca to P of 0cy, 5cy and 10cy
were 2.00, 1.95 and 1.90, respectively (Table 1), which were higher
than 1.67 of stoichiometric HAp, suggesting the existence of car-
bonate ions in the HAp to be substituted with the phosphate ions.
The elements such as Si, Na, Mg and S were also detected
(Table S2), implying that the scales have the incorporation ability
with the mineral components in natural water. In the FT-IR spectra
(Fig. S1), the main adsorption bands appeared at 900–1100 cm 1
due to the characteristic stretching vibrations of phosphate groups
of calcium phosphate compounds. The weak band at 960 cm 1 is
attributed to the symmetric P–O(H) stretching vibrations, indicat-
ing the presence of HPO24 ion. The band at 1033 cm 1 is attributed
to the asymmetric stretching vibration of PO34 . The bands at 876
and 1400–1500 cm 1 are attributed to carbonate ions in the
adsorption or substitution sates in HAp. The absorption band area
ratio of the phosphate to carbonate groups for 0cy, 5cy, and 10cy
were 1.93, 2.34 and 2.99, respectively, indicating that the incorpo-
ration amount of phosphate ions into the fish scales increased with
the treatment.
Fig. 2(a) shows the XRD patterns of 0cy, 5cy and 10cy. The
apparent shapes of the fish scales were preserved with the treat-
ment (Fig. 2(a) Inset). The patterns were attributed to a HAp single
phase (JCPDS 00–009-0432, Ca10(PO4)6(OH)2). The crystalline sizes
(D300 and D002) in Table 1 indicates that the HAp crystal growth
occurred in the fish scale tissue structures with the treatment,
and these sizes were lower than those of the synthesized HAp
nanocrystals (D300 = 15–70 nm and D002 = 20–100 nm) [21]. The
N2 adsorption/desorption isotherms of the 0cy, 5cy and 10cy
(Fig. 2(b)) indicated type II adsorption isotherm shape [22]. The
Pore properties
D002 (nm) SBET (m2 g 1
) rBJH (nm)
14.7 51 2.7, 16, 24
14.7 78 2.7, 16, 24
19.1 55 2.7, 16, 24
158 Y. Chai, M. Tagaya / Materials Letters 222 (2018) 156–159

Fig. 2. (a): XRD patterns of the calcined fish scales with the different treatment cycles. Inset: photograph of 0cy (scale bar: 0.5 cm). (b): Nitrogen adsorption (d) and
desorption (s) isotherms of the fish scales.

Fig. 3. FE-SEM images of (a-1, b-1, c-1) internal and (a-2, b-2, c-2) external layers of (a-1 and a-2) 0cy, (b-1 and b-2) 5cy and (c-1 and c-2) 10cy.

BET surface areas were shown in Table 1, and the BJH pore sizes
were at the mesoscales and the distributions were in the Fig. S2.
The 5cy exhibits the highest BET surface area with the mesopores,
indicating that the porosity was remained by the treatment and
would be affected by the external and the internal layer
nanostructures.
Fig. 3 shows the FE-SEM images of the external and internal lay-
ers of 0cy, 5cy and 10cy. The external and internal layers can be
attributed to the osseous and fibrillary plate parts. In the external
layer, the randomly-distributed HAp particles were observed and
the particle sizes of 0cy, 5cy and 10cy were 16–20, 24–56 and
68–120 nm, respectively. The fine nanoparticles were aggregated
in 0cy and subsequently grew along with the aggregation shape
in 5cy, and ultimately were formed to be the larger plate-like par-
ticles in 10cy. Thus, the existence of randomly-organized collagen
fibers would induce the randomly-distributed particle growth by
the treatment. In the internal layer, the nanoparticles arranged in
one direction were confirmed, revealing the HAp growth along
with the collagen fibers by the treatment. The HAp inside the fish
scale nanostructures was not dense in 5cy and there were voids
between the particles to give the pore nanostructures, and the
voids were filled in 10cy. Therefore, the pores derived from colla-
gen fiber structures were filled with the HAp nanoparticles.
The possible illustration of the cross-sectional fish scale was
suggested as shown in the Scheme S1. Using the distribution of
the HAp and collagen fibers in the external osseous and internal
fibrillary plate layers, the nanostructured HAp was successfully
formed.
4. Conclusions
The nanostructured HAp was templated in the fish scales. The
calcium and phosphate ions were effectively accumulated in the
scales by the immersion process and the HAp grew. The HAp
 Y. Chai, M. Tagaya / Materials Letters 222 (2018) 156–159
nanoparticles arranged in one direction in the internal fibrillary
plate layer were observed, revealing the crystal growth along with
the collagen fibers. The mesopores were formed by the calcination.
Therefore, the arrangement of collagen fibers inside the scales
effectively induced the HAp nanostructures.
Appendix A. Supplementary data
Supplementary data associated with this article can be found, in
the online version, at https://doi.org/10.1016/j.matlet.2018.04.009.
References
[1] K.S. Lew, R. Othman, K. Ishikawa, F.Y. Yeoh, J. Biomater. Appl. 27 (2012) 345–
358.
[2] M. Sadat-Shojai, M.-T. Khorasani, E. Dinpanah-Khoshdargi, A. Jamshidi, Acta
Biomater. 9 (2013) 7591–7621.
[3] S.V. Dorozhkin, M. Epple, Angew. Chem. Int. Ed. 41 (2002) 3130–3146.
[4] S.V. Dorozhkin, Biomaterials 31 (2010) 1465–1485.
[5] M. Kikuchi, S. Itoh, S. Ichinose, K. Shinomiya, Biomaterials 22 (2001) 1705–
1711.
159
[6] A. Letic-Gavrilovic, A. Piattelli, K. Abe, J. Mater. Sci. Mater. Med. 14 (2003) 95–
102.
[7] M. Tagaya, T. Ikoma, M. Takeguchi, N. Hanagata, J. Tanaka, J. Phys. Chem. C. 115
(2011) 22523–22533.
[8] F. Zeng, J. Wang, Y. Wu, Y. Yu, W. Tang, M. Yin, C. Liu, Colloids Surf., A
Physicochem. Eng. Asp. 441 (2014) 737–743.
[9] M. Tagaya, S. Motozuka, T. Kobayashi, T. Ikoma, J. Tanaka, Ind. Eng. Chem. Res.
51 (2012) 11294–11300.
[10] M. Tagaya, S. Motozuka, Opt. Mater. (Amst) 66 (2017) 392–398.
[11] K. Shiba, S. Motozuka, T. Yamaguchi, N. Ogawa, Y. Otsuka, K. Ohnuma, T.
Kataoka, M. Tagaya, Cryst. Growth Des. 16 (2016) 1463–1471.
[12] H. Wang, L. Zhai, Y. Li, T. Shi, Mater. Res. Bull. 43 (2008) 1607–1614.
[13] J. Yamada, N. Watabe, J. Morphol. 159 (1979) 49–65.
[14] A.A. Schönbörner, G. Boivin, C.A. Baud, Cell Tissue Res. 202 (1979) 203–212.
[15] L. Zylberberg, G. Nicolas, Cell Tissue Res. 223 (1982) 349–367.
[16] H. Onozato, N. Watabe, Cell Tissue Res. 201 (1979) 409–422.
[17] A. Bigi, M. Burghammer, R. Falconi, M.H.J. Koch, S. Panzavolta, C. Riekel, J.
Struct. Biol. 136 (2001) 137–143.
[18] T. Taguchi, A. Kishida, M. Akashi, Chem. Lett. 27 (1998) 711–712.
[19] S. Brunauer, P.H. Emmett, E. Teller, J. Am. Chem. Soc. 60 (1938) 309–319.
[20] E.P. Barrett, L.G. Joyner, P.P. Halenda, J. Am. Chem. Soc. 73 (1951) 373–380.
[21] T. Kataoka, K. Shiba, M. Tagaya, Colloids Interface Sci. Commun. 13 (2016) 1–5.
[22] K.S.W. Sing, Pure Appl. Chem. 54 (1982) 2201–2218.