You are on page 1of 7

FULL PAPER

DOI: 10.1002/adfm.200500750

Production and Potential of Bioactive Glass Nanofibers as a


Next-Generation Biomaterial**
By Hae-Won Kim,* Hyoun-Ee Kim, and Jonathan C. Knowles

Over the past decades, bioactive glass has played a central role in the bone regeneration field, due to its excellent bioactivity,
osteoconductivity, and even osteoinductivity. Herein, exploitation of bioactive glass as a one-dimensional nanoscale fiber by
employing an electrospinning process based on a sol–gel precursor is reported for the first time. Under controlled processing
conditions, continuous nanofibers have been generated successfully with variable diameters. The excellent bioactivity of the
nanofiber is confirmed in vitro within a simulated body fluid by the rapid induction of bonelike minerals onto the nanofiber
surface. The bone-marrow-derived cells are observed to attach and proliferate actively on the nanofiber mesh, and differentiate
into osteoblastic cells with excellent osteogenic potential. The bioactive nanofibers have been further exploited in various
forms, such as bundled filament, nanofibrous membrane, 3D macroporous scaffold, and nanocomposite with biopolymer,
suggesting their versatility and potential applications in bone-tissue engineering. Based on this study, the bioactive nanofibrous
matrix is regarded as a promising next-generation biomaterial in the bone-regeneration field.

1. Introduction glasses constitute the essential part of such bioactive materials,


having already been utilized in numerous orthopedic and den-
Materials for biomedical applications have been exploited to tal applications.[5] Most in vivo studies on these bioglasses have
augment and regenerate human tissues that have been sub- confirmed their excellent biocompatibility with hard and even
jected to damage and diseases.[1,2] Over the last decade the de- soft tissues. This is attributed mainly to their ability to form a
mands on synthetic biomaterials have increased significantly, bioactive layer at the interface in contact with living tissues,
to the point where they are now indispensable because autolo- namely the hydroxycarbonate apatite (HCA) layer, which is
gous surgery, regarded as the ‘standard implantation’ tech- equivalent to the mineral phase of human hard tissues.[6] Based
nique, has limited material supplies and requires painful sec- on extensive research conducted in vitro and in vivo, bioactive
ondary operations. Moreover, the alternative allografts or glasses are considered as one of the most-promising biomateri-
xenografts have serious concerns associated with immunogenic als for the “next generation”.[6]
responses.[3] In light of this, significant effort has been devoted Most studies in this field have focused on melt-derived
to the area of biomaterials and tissue engineering, and this has glasses, either in the bulk or granular form. Fiber-type melt-de-
brought about the development of several materials with clini- rived glasses have been produced with diameters of hundreds
cal promise. to tens of micrometers, and these glass fibers reportedly have
Specifically for hard-tissue applications, such as the regen- the potential to act as cell supporters for extracellular matrix
eration and repair of bones and teeth, several bioactive or production and tissue regeneration, with a mechanical strength
bioinert materials have been used clinically.[4] Silica-based bio- superior to that of the equivalent bulk glasses.[7] When formu-
lated with biodegradable polymers, the potential of this bioac-
– tive-glass fiber to act as a tissue-engineering scaffold should be
[*] Prof. H.-W. Kim further increased by adopting the shape flexibility of polymers
Department of Dental Biomaterials, School of Dentistry while retaining optimized mechanical properties (toughness,
Dankook University strength, and elastic modulus) and without sacrificing its excel-
Cheonan 330-714 (Korea)
E-mail: kimhw@dku.edu lent bioactivity.[8] However, with the melt-spinning approach,
Prof. H.-E. Kim the fiber diameter is limited to such micrometer-scale (ca. tens
School of Materials Science and Engineering, to hundreds of micrometers) because of the associated process-
Seoul National University ing restrictions.
Seoul 151-742 (Korea)
More recently, a sol–gel approach was introduced in the pro-
Prof. J. C. Knowles
Division of Biomaterials and Tissue Engineering, duction of bioactive silica glasses.[9,10] Based on the studies un-
UCL Eastman Dental Institute dertaken so far, these sol–gel glasses offer advantages over
London WC1X 8LD (UK) melt-derived glasses in several aspects. Firstly, the bioactivity
[**] The authors greatly appreciate Dr. Y. H. Koh for his constructive dis- of sol–gel glasses is maintained over a wider composition range
cussion and S. Y. Chae for her experimental assistance. This study
was supported by a grant from the Korea Health 21 R & D Project, (i.e., up to a higher silica content) than the melt-derived
Ministry of Health and Welfare, Republic of Korea (A060125). glasses.[10,11] Secondly, and more intriguingly, are the processing

Adv. Funct. Mater. 2006, 16, 1529–1535 © 2006 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim 1529
H.-W. Kim et al./Bioactive Glass Nanofibers as a Next-Generation Biomaterial
FULL PAPER

benefits that the sol–gel approach can


(a) (b) (c)
provide, that is, the accessibility to systems
requiring a scale-reduction for nanobio-
technology. The nanoscale formulation of
the bioactive glasses can also open the
way to the exploitation of new biomedical
areas, such as biocatalysts, biomembranes,
biosensors, and reinforcements. Most of
all, the bioactive glass matrices and scaf-
folds with nanofibrous internal structure
will unquestionably have a significant im- 1 µm 1 µm 1 µm
pact on the bone-tissue regeneration
field.[2] (d) (e)
Herein, we report the production of
sol–gel-derived bioactive glass as a nano-
scale fiber by means of an electrospinning
(ES) technique. The ES process has re-
cently gained significant attention as a
methodology for synthesizing fibrous
structures on the micro-/nanoscale.[12,13]
(f)
Compared to the conventional drawing
techniques used for producing microscale
fibers, the ES method allows the genera-
tion of much smaller diameters, i.e., the
nanoscale fibers. ES is a simple and cost-
effective process, and is thus utilized in di- 100 nm Energy (keV)

verse fields, such as electronics, optics,


Figure 1. Analysis of the glass nanofibers after electrospinning and heat treatment at 700 °C.
and tissue engineering.[12–14] Various types a–c) SEM images of the nanofibers of different average diameters (630 nm, 220 nm, and 84 nm in
of polymeric fibers have been produced sequence) with varying sol concentration (1, 0.5, and 0.25 M in sequence). d) TEM image of the
with scales of 10–1000 nm.[12–14] In more- nanofiber with dave = 84 nm. e) SAED pattern and f) EDS profile of the nanofiber in (d).
recent studies, alkoxide-based sol–gel pre-
cursors were used for the production of
metallic oxide fibers by the ES process.[15,16] As such, the sol– and uniform fibers were created successfully under all condi-
gel glass system described herein can also satisfy the solution tions. At concentrations of over 1 M, micrometer-scale fibers
conditions for the ES process. To the best of our knowledge, were obtained. As the concentration decreased, the diameter
this is the first report on the production of a bioactive glass of the fibers was reduced. However, it was difficult to preserve
nanofiber, and includes examination of its excellent bioactivity the fiber shape at concentrations below approximately 0.25 M,
and osteogenic cell responses. Based on its potential, we also since below this threshold discrete beads were produced exten-
exploited this nanofiber in various structures, such as bundled sively. After the heat-treatment process, the fiber diameters
filaments, fibrous membranes, 3D scaffolds, and in a nanocom- were observed to be reduced by a factor of 2–3, due to the
posite with a biopolymer, showing versatility for practical ap- burn-out of residual polymeric precursors and the consolida-
plications in bone-tissue regeneration. tion of the glass network. The control of the nanofiber diame-
ter made possible by the electrospinning process is effective to
produce tissue-regeneration matrices with a pore size and fiber
2. Results and Discussion network tunable to the demands of specific applications.
Transmission electron microscopy (TEM) observation of a fi-
To obtain bioactive nanofibers, a glass sol of a composition that ber with an average diameter (dave) of 84 nm showed the fiber
retains good bioactivity (here we chose 70 SiO2·25 CaO·5 P2O5) clearly (Fig. 1d). No crystals were formed, as far as could be
was prepared and electrospun under appropriate conditions; confirmed by the TEM selected area electron diffraction
the ES was then followed by a thermal treatment. Depending (SAED) pattern (Fig. 1e). The energy dispersive spectroscopy
on the processing conditions, the electrospun glass fibers pos- (EDS) profile of the fiber revealed the glass composition
sessed a range of diameters. Among the parameters, we ob- (70 SiO2·25 CaO·5 P2O5) quite well (Fig. 1f).
served that the sol concentration was the most dominant factor The heat-treated nanofibers were subsequently incubated in
in controlling the diameter. In Figure 1a–c, scanning electron a simulated body fluid to examine their bioactivity by deter-
microscopy (SEM) showing the morphologies of the nanofibers mining whether they induce the precipitation of bonelike
with different average diameters (630 to 84 nm) at varying ini- minerals on the surface. Data on the fibers obtained with the
tial sol concentrations (1 to 0.25 M) are presented. Continuous 0.25 M sol (dave = 84 nm) are presented as a representative ex-

1530 www.afm-journal.de © 2006 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim Adv. Funct. Mater. 2006, 16, 1529–1535
H.-W. Kim et al./Bioactive Glass Nanofibers as a Next-Generation Biomaterial

FULL PAPER
ample in Figure 2. After incubation of the fiber mesh for 1 day, (a) 3
the surface became quite irregular with some elongated crys-
Ca ref.
tals being detected on the surface by TEM observation
(Fig. 2a). When incubated for 3 days, the fiber morphology

Concentration (mM)
2
was changed more significantly, with numerous elongated crys- Ca
tals being generated over almost the entire surface (Fig. 2b). A
high-magnification image of the fiber incubated for 1 day P ref.
1
clearly shows the formation of crystals (contrast by dark area) Si
on the fiber surface with sizes of a few to tens of nanometers
(Fig. 2c). The SAED pattern of the crystals in Figure 2c con- P
firmed that they consisted of a poorly crystallized apatite with 0
0 1 2 3 4 5 6 7
(112) strong and (002) weak rings (Fig. 2d). The EDS analysis Dissolution time (days)
of the apatite crystal area shows a higher concentration of Ca (b)
and P with respect to that of Si (Fig. 2e, compare data with
Fig. 1f). The Ca/P ratio was approximately 1.56, being close to,

% Reflectance (a.u)
but slightly lower than, the stoichiometry of pure hydroxyapa-
tite [Ca10(PO4)6(OH)2)], 1.67, which has often been reported silica
in the poorly crystallized or carbonated apatites produced by
this kind of biomimetic process, and is more similar to that of 3 days
bonelike minerals.[17]
The ion-concentration change of the medium was monitored 7 days
using inductively coupled plasma atomic emission spectroscopy
(ICP-AES) after incubation of the nanofiber for periods of up 1200 1000 800 600
-1
to 7 days, as shown in Figure 3a. Initially the prepared medium Wave number ( cm )
contained 2.5 and 1 mM of Ca and P, respectively, and no Si.
Within a short period of time, both the Ca and P concentra-
Figure 3. Change in a) ion concentration of medium and b) Fourier trans-
tions decreased abruptly and then stabilized with increasing form infrared (FTIR) spectroscopy of the nanofiber (dave = 84 nm) before
time, while the Si concentration increased initially and then le- and after incubation for 5 and 7 days. In (b), symbols indicate bands
veled off. Only a slight increase in the Ca concentration was related to silicate (䊊), phosphate (䊉), and carbonate (diamond).

(a) (b) (c)

200 nm 200 nm

(d) (e)

(002)
(112) 20 nm
Energy (keV)

Figure 2. Bioactivity analysis of the glass nanofiber (dave = 84 nm) after incubation in a simulated body fluid: TEM image for a) 1 and b) 3 days, c) magni-
fication of (a). d) SAED pattern of the crystal in (c), and e) EDS profile of the crystal.

Adv. Funct. Mater. 2006, 16, 1529–1535 © 2006 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim www.afm-journal.de 1531
H.-W. Kim et al./Bioactive Glass Nanofibers as a Next-Generation Biomaterial
FULL PAPER

observed in the first few hours. These changes in the ionic con- formation.[2] However, this postulation, established from in vi-
centration illustrate the dissolution/precipitation process, i.e., tro conditions, warrants further experiments in vivo when con-
the dissolution of Ca, P, and Si from the nanofiber, and the sub- sidering the dynamic fluidic effect as well as the mediation of
sequent precipitation of Ca–P crystals from the medium, which proteins in the series of reactions.[18]
became supersaturated by the dissolution of Ca and P. In de- The biocompatibility of the glass nanofibers was further
tail, a series of reactions occurred, as proposed by Hench and assessed by their in vitro cellular responses. We used bone-
co-workers: the exchange of alkali ions, such as Ca2+ and Na+ marrow-derived stem cells (BMSCs) to examine the osteogenic
(Ca2+ in this study) with H3O+, the attack of hydroxyl ions potential of our newly developed bioactive nanofiber. The
(OH–) present in the medium through the silica network struc- nanofibrous meshes with an average diameter of 220 nm were
ture to form silanol groups (Si–OH), and through which the used for the cellular test. From electron microscopy images,
precipitation of Ca2+ and PO43– (and mostly CO3– also) occurs, the cells on the nanofiber mesh with 5 days of culturing were
followed by the crystallization of HCA.[6] Of special note, when observed to grow favorably (shown in Fig. 4a). On closer ex-
compared to bulk glasses (melt-derived or sol–gel synthesized) amination, the cells spread actively on the nanofibrous surface
wherein the increase of Ca and P concentrations (in the initial with numerous cytoplasmic extensions, typical of the osteoblas-
dissolution region) usually continues for several days to weeks, tic cellular growth (Fig. 4b). The grown cells were stained to
the nanofibers exhibited a more rapid initial drop (as short as a elicit alkaline phosphatase (ALP) enzyme, which is known to
few days) in the Ca and P concentrations of the medium. This be expressed in the differentiation of osteogenic cells and thus
was mainly attributed to the large surface area afforded by the is regarded as an important marker for osteogenic potential
nanoscale fibers, which resulted in faster dissolution and super- and bone-forming ability. As references, data on the sintered
saturation of the medium with respect to the HCA crystal nu- sol–gel glass disk (same composition as the glass nanofiber)
cleation. Specifically, the sol–gel derived nanofibers developed and the degradable biopolymer (polycaprolactone, PCL) in a
in this study possess around 2–3 orders of magnitude higher nanofibrous form are also provided. The optical image pre-
surface area than the conventional melt-derived glass fibers sented of the glass nanofiber (Fig. 4c1) shows a thicker violet
(diameters of approximately hundreds of micrometers) at an staining of ALP with respect to that on the glass disk (Fig. 4c2),
equivalent volume.[16] To confirm whether these nanoscale- and the difference was clearer when compared to that on the
driven surface properties influence the rate of the CaP induc- PCL nanofiber (Fig. 4c3). The ALP level, as well as the viabili-
tion on the surface, we performed a comparison test using a ty of the cells cultured on the three different samples, was
sintered sol–gel glass disk (with a composition same as the nano- quantified as summarized in Figure 4d. The cells grown on the
fiber) under equivalent medium conditions (see Experimental bioactive glass composition (both nanofiber and disk forms)
section for details). However, the glass disk layer could not in- were more viable than those on the PCL nanofiber, while only
duce the formation of CaP on the surface as rapidly as the nano- a slight difference was observed between the two glass types.
fiber: the Ca–P ionic drop analyzed by ICP-AES was observed However, the expression of ALP on the nanofiber was higher
ca. 7 days after the incubation of the sintered glass disk. than that on the glass disk (particularly significant at day 5).
The crystals formed from the silica fibers were analyzed with Based on these cellular results, we confirm that the nanofibrous
Fourier transform infrared (FTIR) spectroscopy at different glass favors the attachment and growth of the BMSCs and in-
incubation times, as shown in Figure 3b. As the incubation time duces them to differentiate into osteoblastic cells. Moreover,
increased, the bands related to the silica glass (800, 930, 1080, this osteogenic potential of the nanofiber glass was observed to
and 1200 cm–1) were attenuated, while those attributed to be stimulated significantly to levels similar to or even higher
the phosphate groups (570, 605, 960, and 1030–1090 cm–1) than those on the relatively flat surfaces with equivalent com-
increased and carbonate bands (870 and 1200–1300 cm–1) ap- position (glass sintered disk), suggesting the morphologically
peared, suggesting the formation of carbonate-substituted apa- driven (surface associated) properties of the nanofiber, such as
tites (HCA), which is similar to the composition of bone miner- nanoscale-roughened morphology, larger surface area, and
al. higher ionic release, play constructive roles in the osteogenic
These results concerning the crystal formation on the silica- potential. In this study, we used the glass nanofiber with an
glass nanofiber have much in common, for the most part, with average diameter of 220 nm, based on the consideration that it
the previous results obtained for melt-derived or sol–gel pro- is representative of other-sized nanofibers, as we observed sim-
cessed bulk glasses. However, the degree of bioactivity (the ilar in vitro apatite-forming behavior among all the nanofibers.
crystal formation) of the nanofibers in vitro appeared to be sig- However, further study is still warranted regarding the effect
nificantly enhanced by the nanoscale production process. As of diameter of the nanofiber glass on the cellular responses.
suggested above, the extremely large surface area of the nano- Although more extensive works are required as to such ef-
fibers should accelerate the series of reactions taking place on fects on the stimulation of osteogenic potential, the present
the glass surface in contact with the medium. Therefore, when finding drives us to underscore the potential of the nanofiber
these silica glass nanofibers are used as bone substitutes, they when considering the perspectives of manipulation and versa-
could be expected to quickly provide a favorable environment tility of the nanofibrous form, which is useful in a tissue-engi-
owing to the rapid formation of bonelike minerals on the sur- neering matrix (demonstrated in the following Fig. 5). Clearly,
face, thereby exhibiting excellent responses for adhesive pro- the osteogenic potential of our developed nanofiber was
teins and cells and, consequently, resulting in improved bone mostly endowed by the bioactive sol–gel glass composition, as

1532 www.afm-journal.de © 2006 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim Adv. Funct. Mater. 2006, 16, 1529–1535
H.-W. Kim et al./Bioactive Glass Nanofibers as a Next-Generation Biomaterial

FULL PAPER
(a) (b)

50 µm 10 µm

(c1) (c2) (c3)

(d)

Figure 4. Bone-marrow derived osteoblastic cell responses to the bioactive glass nanofiber: Electron microscopy images of the cells on the nanofibrous
mesh at low (a) and high (b) magnification with 5 days of culturing. c1–c3) alkaline phosphatase (ALP) staining of the cells grown on the bioactive glass
nanofiber (c1) and on other comparison samples for 5 days: sol–gel glass sintered disk with the same composition as glass nanofiber (c2) and PCL elec-
trospun nanofiber (c3); ALP was enzymatically stained in violet. d) Quantification of the cellular assays, represented by mean ± one standard deviation
(in brackets), on the samples: cell viability assessed by MTT method at 2 and 5 days and ALP level assessed by an enzymatic reaction at 5 and 10 days.
Statistical significance for the glass nanofiber was considered with respect to the glass disk (+p < 0.05) and PCL nanofibers (*p < 0.05, ** p < 0.01) by
ANOVA.

was demonstrated by the above cellular data in its comparison The usefulness of the bioactive nanofiber was further dem-
with the degradable but bio-inactive polymeric nanofiber PCL. onstrated by utilizing it in 3D structures that could find practi-
This bioactive glass nanofiber preserves the beneficial aspects cal applications as cell-supporting matrices and tissue-engi-
of not only the bioactive composition with high osteogenic po- neering scaffolds. We formulated the nanofibers as various 3D
tential but also the morphological merits permitting use as a matrices using our laboratory-developed techniques, as shown
tissue-regeneration matrix like the degradable polymer nanofi- in Figure 5. As one example, a macroscale filament was pro-
bers. With its excellent properties for the recruitment of stem duced after aligning the nanofibers at a high mandrel winding
cells, the glass nanofiber holds great promise in the tissue-engi- speed (ca. 2000 mm s–1) followed by bundling and compacting
neering field. However, this postulation, established in vitro, them (Fig. 5a). This longitudinally aligned filament with an in-
should be substantiated with further in vivo animal studies on ternal nanofibrous structure can be specifically applied to sup-
the bone-formation ability and the mechanical stability. Never- port the cell ingrowth of aligned tissues such as muscles and
theless, the present in vitro observations, including the rapid nerves. The texture designing of the filament for specific parts
induction of bonelike minerals and the osteogenic cellular re- remains another area for development. A nanofibrous mem-
sponses, suggest great potential in the hard-tissue-regeneration brane (Fig. 5b) was produced by means of stacking the nanofi-
field. brous sheets and warm-pressing followed by heat treatment.

Adv. Funct. Mater. 2006, 16, 1529–1535 © 2006 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim www.afm-journal.de 1533
H.-W. Kim et al./Bioactive Glass Nanofibers as a Next-Generation Biomaterial
FULL PAPER

network (Fig. 5d). The image shows a well-


(a) (b) constructed nanocomposite comprising
the dense nanofibrous network and the
polymeric filler. The glass nanofibers were
1 µm observed to be dispersed well and distrib-
uted uniformly in the PLA matrix.
Further macroporous design of the nano-
composite is expected to be achievable by
utilizing diverse scaffolding techniques.
This nanocomposite approach, employing
the nanofibrous bioactive inorganic in
conjunction with biodegradable polymer,
should have good prospects in the bone-
20 µm 10 µm regeneration field because of the combi-
natorial benefits, such as shape flexibility
(c) (d) and moldability, mechanical properties
tunable to the bone matrix, and the bioac-
tivity and degradability. Further evalua-
tions on the in vivo tissue responses and
mechanical properties of the nanocom-
posite are currently underway.

3. Conclusions

Bioactive nanofibers were produced


successfully by employing an electrospin-
1 µm 1 mm 1 µm 10 µm ning technique using a glass sol–gel pre-
cursor with a bioactive composition. The
Figure 5. 3D structured matrices of the electrospun glass nanofibers for tissue regeneration. a) Fil- biomedical usefulness of bioactive glasses,
ament: electrospun nanofibers were aligned and bundled into a microfilament. b) Membrane: elec-
as has been established over recent de-
trospun sheets were stacked and pressed gently (surface (upper) and cross-section view (lower)).
c) 3D macroporous scaffold: ready-made filaments were architectured using a negative-mold tech- cades, was imparted onto the nanofibrous
nique and then heat-treated; enlarged in inset. d) Polymer-filled nanocomposite: heat-treated structure. We observed that the nanofiber
fibrous mesh was filled with biodegradable polymer PLA; enlarged in inset. possessed excellent bioactivity and osteo-
genic potential in vitro. Moreover, possi-
The nanofibers were observed to maintain their initial fibrous ble formulations of the bioactive nanofibers, such as aligned fil-
structure and form a nanoporous 3D structure. Potential applica- aments, nanofibrous membranes, macroporous scaffolds, and in
tions for this nanofibrous membrane are believed to include nanocomposites with biopolymers, were provided for practical
wound healing and guided bone/cartilage tissue regeneration, applications. Based on the present findings, the bioactive glass
wherein polymeric matrices are still dominantly used. A 3D mac- nanofiber developed herein is proposed as one of the most
roporous scaffold, characterized by a microfilament framework promising next-generation biomaterials.
with a nanofibrous internal structure, was also created to allow
macropores within the nanofibrous structure (Fig. 5c). Here, a
negative-molding technique was used, where the microfilaments 4. Experimental
of nanofibers were organized with carbon filaments and then
heat-treated to produce a 3D open-channeled network with con- Sol Preparation and Electrospinning: The glass composition used in
this study (70 SiO2·25 CaO·5 P2O5) was chosen based on those of pre-
trolled pore size (ca. 500 lm) and porosity (ca. 50 %). This mac- viously developed bulk glasses which exhibit bioactivity in vitro, and
roporous design should facilitate vasculisation and cell ingrowth the sol–gel processing conditions were modified from other reports
more effectively through macroscale open channels. Of particular [10,11]. The glass precursors were added at appropriate ratios (tetra-
interest is the fact that the nanofibrous surface can provide favor- ethyl orthosilicate, calcium nitrate, and triethyl phosphate added in se-
quence with a 6 h interval between each addition, all from Aldrich) in
able surroundings for the initial protein adhesion and cellular re- an ethanol/water solution containing 2 % HCl (1 N) used for catalysis.
sponses, which is typically dissimilar to conventional 3D scaffolds The sol mixture was stirred for 24 h and aged without stirring at 25 °C
with a dense surface.[19] These 3D matrices designed with bioac- for 24 h followed by a further 48 h at 40 °C. The acidic catalyst was nec-
tive nanofibers can have their mechanical properties optimized essary in order to produce a clear sol. Prior to electrospinning (ES),
polyvinylbutyral (PVB, from Aldrich) dissolved in ethanol at 10 % was
when hybridized with biopolymers. We infiltrated a biodegrad- added to the sol at an equivalent volume to adjust the rheological prop-
able polymer (5 % poly(D,L-lactic acid) (PLA) solution dissolved erties of the sol, so that they would be fit for the fiber generation dur-
in tetrahydrofuran) into the interspacings of the glass nanofiber ing ES. 2 mL of the solution was loaded in a syringe and injected into a

1534 www.afm-journal.de © 2006 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim Adv. Funct. Mater. 2006, 16, 1529–1535
H.-W. Kim et al./Bioactive Glass Nanofibers as a Next-Generation Biomaterial

FULL PAPER
rotating mandrel for 10 min at high voltage under controlled conditions genic differentiation marker. The cells cultured on the samples for
(voltage: 12 kV, distance: 8 cm, and injection rate: 0.05 mL min–1). The 5 days were treated enzymatically to stain the ALP and reveal it in vio-
electrospun fibers were subsequently heat-treated at 700 °C for 3 h in let. Moreover, the ALP expression level was quantified following our
air at a heating and cooling rate of 1 and 5 °C min–1, respectively. The previous protocol [21]. At each culturing period, the cell layers were
heat-treatment temperature was determined to be high enough to detached by treatment with 0.1 % Triton X-100, and the cell pellets
eliminate organic sources and nitrates completely but low enough to were disrupted via cyclic freezing/thawing processes. Aliquots of the
avoid crystallization [10,11]. As a reference for the biological tests, samples normalized to the total protein content were enzymatically re-
we also prepared disk-type pallets of the sol–gel composition acted with p-nitrophenyl phosphate (p-NPP) substrate following the
(12 mm × 1 mm thickness) after crushing the gelled product into pow- manufacturer’s instruction (Sigma ALP kit 104). The quantity of p-ni-
ders, molding them in a metal die, and partial sintering at 700 °C for trophenol (p-NP) produced was measured at an absorbance of 410 nm
3 h. As another reference material, a degradable biopolymer, polyca- using a spectrophotometer. The cellular tests were performed on five
prolactone (PCL, Sigma), was prepared in a nanofibrous matrix (aver- replicate samples and data were compared using one-way ANOVA
age fiber diameter ∼ 860 nm) by electrospinning using a solvent (tetra- analysis with statistical significance at p < 0.05 and p < 0.01.
hydrofuran/N,N-dimethylformamide = 3:2) under adjusted conditions 3D Structure Formulation: For the fabrication of the nanofiber in 3D
(voltage: 12 kV, distance: 8 cm, and injection rate: 0.05 mL min–1). structured matrices (bundled filament, nanofibrous membrane, macro-
Bioactivity Test: The heat-treated glass nanofibers were subjected to porous scaffold, and nanocomposite with biopolymer), several labora-
a simulated body fluid (containing similar ion concentrations to a body tory-developed techniques were utilized. A filament was produced by
plasma), in order to form bonelike mineral HCA crystals on the glass bundling the electrospun fiber mesh and compacting to a diameter of
surface due to a dissolution/precipitation process [20]. This final step ∼ 300 nm, and followed by a heat-treatment at 700 °C for 3 h. For the
was carried out in an effort to 1) provide a preliminary assessment of production of fibrous membrane, the electrospun nanofiber mesh was
the bioactivity of the glass nanofibers in vitro, since the HCA forma- stacked layer-by-layer and warm-pressed ∼ 80 °C, and then heat-treated
tion has been regarded as the key phenomenon in explaining the bio- under the same conditions. The 3D macroporous scaffolding was per-
compatibility of bioactive materials [2], and 2) utilize the HCA-sur- formed using the nanofibrous filament (diameter of ∼ 600 nm) and a
face-modified glass nanofibers directly. 10 mg of fiber were incubated designed carbon mold as a support material, by means of aligning and
in 10 mL of medium for up to 7 days without refreshing. The glass disk stacking them, and then followed by a thermal treatment. To produce a
prepared as a reference was also tested under the same conditions. At nanofiber–biopolymer nanocomposite, a biodegradable polymer
predetermined time points, the change in the ion concentration of the poly(D,L-lactic acid) (PLA, Aldrich) solution (5 % in tetrahydrofuran)
medium was monitored with inductively-coupled plasma atomic-emis- was filled within the stacked nanofibrous mesh (heat-treated), homoge-
sion spectroscopy (ICP-AES, Shimadzu), and the change in the chemi- nized by vortexing gently, and then dried and warm-pressed (120 °C).
cal bonding structure of the fiber was analyzed with Fourier transform The ratio of nanofiber to PLA was adjusted to 0.5 by weight.
infrared (FTIR, System 2000, Perkin-Elmer) spectroscopy. The fiber di-
ameter was measured from the SEM image in 20 different sections, and Received: October 28, 2005
averaged. The morphology of the fibers was characterized with field- Final version: December 14, 2005
emission scanning electron microscopy (FESEM, JSM6330F, JEOL), Published online: June 27, 2006
and transmission electron microscopy (TEM, CM20, Philips). The com-
position and crystal pattern were also analyzed from the TEM opera-

[1] R. Langer, D. A. Tirrell, Nature 2004, 428, 487.
tion. [2] L. L. Hench, J. M. Polak, Science 2002, 295, 1014.
Cellular Response Assay: To observe the biocompatibility of the
[3] C. Danien, R. Parsons, J. Appl. Biomater. 1991, 2, 187.
nanofiber, in vitro cellular responses were examined using rat bone-
[4] C. T. Laurencin, A. M. A. Ambrosio, M. D. Borden, J. A. Cooper,
marrow derived mesenchymal stem cells (BMSCs). Bone marrow was
Annu. Rev. Biomed. Eng. 1999, 1, 19.
harvested from the adult rats (age 4–8 weeks, Korean). The whole mar-
[5] J. Wilson, A. Yli-Urpo, R. P. Happonen, in An Introduction to Bio-
row was flushed with a minimum essential medium (a-MEM) from the
excised proximal and distal epiphyses of femora and tibiae using a sy- ceramics (Eds: L. L. Hench, J. Wilson), World Scientific, Singapore
ringe. After centrifugation of the marrow at 1000 g, the supernatant 1993, pp. 63–73.
was collected and suspended again in a-MEM. The homogeneously [6] L. L. Hench, J. Am. Ceram. Soc. 1998, 81, 1785.
suspended cells were maintained in a culture flask containing growth [7] M. A. De Diego, N. J. Coleman, L. L. Hench, J. Biomed. Mater. Res.
medium (a-MEM, 2 mM glutamine, 100 U mL–1 penicillin, and 2000, 53, 199.
100 mg mL–1 streptomycin), supplemented with 15 % fetal calf serum [8] V. Maquet, A. R. Boccaccini, L. Pravata, I. Notingher, R. Jerome,
(FCS) under a humidified atmosphere of 5 % CO2 at 37 °C. After Biomaterials 2004, 25, 4185.
1 day of incubation, the cells were washed with phosphate-buffered sa- [9] R. Li, A. E. Clark, L. L. Hench, J. Appl. Biomater. 1991, 2, 231.
line (PBS) to discard nonadherent cells and then cultured further for [10] J. Zhong, D. C. Greenspan, J. Biomed. Mater. Res. 2000, 53, 694.
up to 7 days to reach near confluence. For accurate measurement of [11] A. J. Salinas, A. I. Martin, M. Vallet-Regi, J. Biomed. Mater. Res.
the cellular tests, the glass nanofiber was prepared by supporting it on a 2002, 61, 524.
bioinert zirconia sintered disk. Glass nanofibers prepared with an aver- [12] M. Bognitzki, W. Czado, T. Frese, A. Schaper, M. Hellwig, M. Stein-
age diameter of 220 nm were chosen as a representative example. For hart, A. Greiner, J. H. Wendorff, Adv. Mater. 2001, 13, 70.
the purpose of comparison, the sol–gel glass disk pallet and PCL nano- [13] Y. Dzenis, Science 2004, 304, 1917.
fibrous matrix (preparation methods described above section in detail) [14] D. Li, Y. Xia, Adv. Mater. 2004, 16, 1151.
were also prepared. The maintained cells were harvested with trypsin–
[15] G. Larsen, R. Velarde-Ortiz, K. Minchow, A. Barrero, I. G. Loscer-
EDTA (EDTA= ethylenediaminetetraacetic acid) and plated onto the
tales, J. Am. Chem. Soc. 2003, 125, 1154.
three types of samples at a seeding density of 5 × 104 cell cm–2. At this
[16] D. Li, Y. Xia, Nano Lett. 2003, 3, 635.
time of culturing, the medium was supplemented with 50 lg mL–1 so-
[17] S. Mann, Nature 1988, 332, 119.
dium ascorbate, 10 mM sodium b-glycerol phosphate, and 10 nM dexa-
methasone in order to induce osteoblastic differentiation. The culture [18] C. Rey, A. Hina, A. Tofighi, M. J. Glimcher, Cells Mater. 1995, 5, 345.
medium was changed every three days. [19] M. M. Stevens, J. H. George, Science 2005, 310, 1135.
At the culturing periods of 2 and 5 days, the cell viability was as- [20] Known as Kokubo’s medium simulated to human body plasma, with
sessed by means of an MTT method [21]. The cell-growth morphology ionic concentrations as follows: Na+ (142.0 mM), K+ (5.0 mM), Ca2+
was observed with SEM after fixing the cells with 2.5 % glutaraldehyde, (2.5 mM), Mg2+ (1.5 mM), Cl– (147.8 mM), HPO42– (1.0 mM), and SO42–
dehydrating them with a graded series of ethanols (70, 90, and 100 %), (0.5 mM).
and gold coating. The osteogenic potential of the cells was assessed by [21] H. W. Kim, J. C. Knowles, V. Saih, H. E. Kim, J. Biomed. Mater. Res.
detecting the alkaline phosphatase (ALP), which is an important osteo- A 2004, 71, 66.

Adv. Funct. Mater. 2006, 16, 1529–1535 © 2006 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim www.afm-journal.de 1535

You might also like