Post on 11-Jun-2016
Highly Porous 3D Nanofiber Scaffold Using anElectrospinning Technique
GeunHyung Kim, WanDoo Kim
Bio-mechatronics Laboratory, Department of Future Technology, Korea Institute of Machinery and Materials,Yuseong-Gu, Daejeon, Korea
Received 7 December 2005; revised 23 March 2006; accepted 15 May 2006Published online 21 August 2006 in Wiley InterScience (www.interscience.wiley.com). DOI: 10.1002/jbm.b.30642
Abstract: A successful 3D tissue-engineering scaffold must have a highly porous structure
and good mechanical stability. High porosity and optimally designed pore size provide
structural space for cell accommodation and migration and enable the exchange of nutrients
between the scaffold and environment. Poly(e-carprolactone) fibers were electrospun using an
auxiliary electrode and chemical blowing agent (BA), and characterized according to porosity,
pore size, and their mechanical properties. We also investigated the effect of the BA on the
electrospinning processability. The growth characteristic of human dermal fibroblasts cells
cultured in the webs showed the good adhesion with the blown web relative to a normal
electrospun mat. The blown nanofiber web had good tensile properties and high porosity
compared to a typical electrospun nanofiber scaffold. ' 2006 Wiley Periodicals, Inc. J Biomed Mater
Res Part B: Appl Biomater 81B: 104–110, 2007
Keywords: poly(e-carprolactone); nanofiber; scaffold; chemical blowing agent
INTRODUCTION
One of the important issues in tissue engineering has been
the development of 3D scaffolds, which guide cells to grow
functional tissues and allow the diffusion of nutrients, metabo-
lites, and soluble factors.1,2 Factors governing scaffold de-
sign include considerations of pore size and morphology,
mechanical properties versus porosity, surface properties, and
appropriate biodegradability.3–5 Of these, pore size and po-
rosity are directly related to the success of a scaffold. High
porosity provides structural space to accommodate cells and
serves as the exchange path for nutrients and metabolic waste
between the scaffold and the environment.3 The pore size of
a scaffold should be optimized to allow cell migration. How-
ever, it is difficult to define the optimum pore size accurately;
rather, a range of pore sizes is used. The pore size is also de-
pendent on the tissue type desired. For example, the effective
pore size for cell growth is between 20 and 60 mm in vascular
grafts, while it is 75–150 mm for bone tissue.6
Materials with a 3D micropore structure have been pro-
duced using various manufacturing methods, including a
solid free-form (SSF) fabrication system (or RP process),7,8
fiber bonding (unwoven meshes), solvent casting or particu-
lar leaching, phase separation, high pressure gas expansion,
and emulsion freeze-drying.9–13 Although various techni-
ques have been developed for manufacturing scaffolds, it
is not clear which method would best fabricate a so-called
ideal scaffold.
As other studies have reported, the criteria for an ideal
polymeric scaffold are also dependent on the type of mate-
rial, which can range from natural biodegradable high-
molecular weight materials, such as collagen, agarose, and
alginate,14–16 to synthetic polymers such as poly(glycolic
acid) (PGA), polylactic acid (PLA), poly(e-carprolactone)(PCL), and poly(D,L-lactide-co-glycolide).4,17 An electro-
spun PCL scaffold coated with collagen has been used to
seed smooth muscle cells to construct vascular tissue.18
Moreover, PLA and PGA have also been used for cell trans-
portation and the regeneration of vascular tissue in vivo.19
Recently, the electrospinning process has been explored
widely as a way to fabricate nanofibers for various applica-
tions, such as in biosensors, implantable scaffolds, and ultra-
thin filtration. With increasing demand for a sophisticated
nanosized hybrid and pore structure based on commercial
polymers, various electrospinning processes have been inves-
tigated and developed.20–25 As a good method for fabricating
nanofiber structures, this technique is preferred because of its
simplicity, allowing the manufacture of nanofibers made from
complex ingredients, and high applicability.
The principle of electrospinning is that a Taylor cone is
formed by applying an electric field to polymer solution
hanging from a capillary tip, which causes jets of electrically
charged solution to be emitted when the applied electrostatic
force is stronger than the surface tension of the solution.20–22
Correspondence to: G. H. Kim (e-mail: gkim@kimm.re.kr or xrdghk@yahoo.com)
Contract grant sponsor: Korea Institute Machinery and Materials (KIMM)
' 2006 Wiley Periodicals, Inc.
104
The jets of solution erupt from the apex of the cone at a noz-
zle and travel toward an electrically grounded target on which
they are stacked as a nonwoven mat.
Electrospun nanofibers overcome the limitations of con-
ventionally fabricated scaffolds because of their high surface
area and high porosity, and the easy fabrication process.4,26
However, the pore size of an electrospun nanofiber web is
so small that cultured cells have difficulty penetrating the
scaffold, which is a 3D structure composed of nanofibers.
Consequently, it is not suitable for fabricating 3D nanofiber
scaffolds because of the low rate of cell diffusion in the
thickness direction.
To overcome these problems, we fabricated nanofiber mats
using an electrospinning technique with a chemical blowing
agent (BA), and investigated the influence of the agent on the
spinning process, pore size, and mechanical properties of the
resulting mats. In addition, to control the unstable electro-
spinning process involving a chemical BA, we introduced an
auxiliary cylindrical electrode to this system.
MATERIALS AND EXPERIMENTAL PROCEDURES
Materials
A biodegradable polymer solution was prepared by dissolving
2.4 g of poly(e-carprolactone) (PCL; Sigma–Aldrich, St. Louis,
MO) in 30 g of a solvent mixture consisting of methylene-
chloride (MC) and dimethylformamide (DMF) in a MC/DMF
wt% ratio of 80/20. The PCL solution was prepared with a
fixed 8 wt% concentration. To observe the effect of a chemi-
cal BA on processability, three different weight percents of
BA were mixed with the PCL solution. The BA was azodi-
carbonamide (Cellcom-AC series; KumYang, Seoul, Korea).
The polymer solution was placed in a 20-mL glass syringe
with a G-20 needle. The flow rate of polymer solution was
controlled using a syringe pump system (KDS 230; NanoNc,
Seoul, Korea).
Electrospinning Setup for PCL Nanofibers
PCL nanofibers were fabricated using an electrospinning
technique with a cylindrical auxiliary electrode connected
to a nozzle. The cylindrical electrode and nozzle were con-
nected by a copper wire and could be charged simultaneously
as the fluid passed through the spinneret. The detailed shape
and geometry of the auxiliary electrode are shown in Figure 1.
The cylindrical electrode served to reduce the instability of
the initial jet leaving the apex of the Taylor cone. More details
on the role of the electrode are described later.
The syringe was subjected to the same applied voltage (15–
20 kV) with a distance of 150–300 mm between the target and
needle tip, and a high-voltage power supply (SHV300RD-
50K; Convertech, Korea) was used to control the applied
voltage.
Figure 1. Schematic of the electrospinning setup with an auxiliary
cylindrical electrode. The cylindrical electrode has a radius of 13 mm
and is 0.45 mm thick.
Figure 2. EFCFs for a single nozzle and a nozzle connected to a cylindrical electrode. (a) Schematic
of a nozzle with an auxiliary electrode in the electrospinning setup used in this work, showing thecalculated EFCFs (b) without and (c) with an auxiliary electrode. [Color figure can be viewed in the
online issue, which is available at www.interscience.wiley.com.]
105HIGHLY POROUS 3D NANOFIBER SCAFFOLD
Journal of Biomedical Materials Research Part B: Applied BiomaterialsDOI 10.1002/jbmb
The initial straight jet in the electrospinning process was
visualized using a 532 nm green laser with an expanding op-
tical lens. The initially spun jet was photographed using a
digital camera (E-300; Olympus, Tokyo, Japan).
The spun fibers were deposited on a poly(ethylene tereph-
thalate) (PET) film covering a copper target and the spun
nanofiber web was blown in 2–3 s at 100 8C in a convection
drying oven.
Cell Culture and Experiment
The nanofiber scaffolds were sterilized with 70% ethanol
and exposed to UV light for 1 h They were prewarmed with
Hank’s balanced salt solution for 2 h at 50 8C. Human der-
mal fibroblasts (HDFs) were cultured in Dulbecco’s modi-
fied eagle medium supplemented with 10% fetal bovine
serum and maintained up to passage 7. Isolated fibroblasts
were seeded on each sample (1 3 1 cm2) at a density of 2 3104 cells and cultured for up to 3 days at 37 8C and 5%
CO2. The cells were fixed with 2.5% glutaraldehyde for 1 h
at room temperature, and dehydrated through a series of
ethanol dilutions. The samples were sputter-coated with Pt.
Cell morphology and growth was observed using a scanning
electron microscopy.
Structural Morphology and MechanicalTesting of PCL Scaffolds
The morphologies of the electrospun PCL mats constructed
with and without a BA were observed under an optical micro-
scope (BX FM-32; Olympus) and scanning electron microscopy
(FE–SEM, Sirion; FEI, Hillsboro, OR). Before observation,
the scaffolds were coated with gold using a sputter coater.
The pore size and porosity were measured from the SEM and
optical micrographs. The tensile properties of nanofiber mats
were characterized using a Universal Tensile Machine (Tytronics,
Lowell, MA) at a 10 mm/min crosshead speed.
RESULTS AND DISCUSSION
Effect of the Auxiliary Electrode and Process Parameters
To analyze the effect of the auxiliary electrode in this elec-
trospinning process, ANSYS/Emag-3D was used to conduct
Figure 3. Image showing the stable initial jet of solution (straight line)from a single nozzle during electrospinning to a target plate coved
with PET film with DC ¼ 18 kV and a distance between the nozzle
and ground of 150 mm. The spun fibers were obtained from 8% solu-
tion of PCL and 1 wt% BA in MC/DMF.
Figure 4. Electrical conductivity, k, and surface tension, g, of 8 wt%
PCL solutions in a mixture of MC and DMF containing various con-centrations of BA.
Figure 5. Transmission optical microscope images of electrospun
nanofiber obtained with (a) 0 wt% or (b) 1 wt% BA in an 8 wt% PCLsolution.
Figure 6. Process diagram of electrospinning using PCL solutions
with or without a BA.
106 KIM AND KIM
Journal of Biomedical Materials Research Part B: Applied BiomaterialsDOI 10.1002/jbmb
an electromagnetic analysis using the finite element method.
The numerical models were constructed using three different
elements: 20-node brick elements, tetrahedron elements that
were degenerate brick elements, and infinite elements. The
radius and length of the nozzle were 0.5 and 15 mm, respec-
tively. For the system shown in Figure 2(a), the inner radius,
R, of the auxiliary electrode and the thickness of the auxil-
iary electrode were 13 and 1 mm, respectively. The distance,
H, between the spinneret and the ground electrode was
250 mm. To consider the effects of an applied electric field
in this electrospinning process, the electric field concentra-
tion factor (EFCF) was defined using the equation27:
EFCF ¼ Ep��E
��
rp��rp
��¼ Er
��E
��
where Ep is the projection vector of the electrical field, E,
into the x–y plane, rp is the projection vector, and Er is the
r-directional component of the electrical field in the cylindri-
cal coordinates. Since the e-spinning system is approximately
axisymmetric, we located the origin of the coordinate system
at the center of the ground electrode. In this case, we used the
Cartesian coordinates system and defined the EFCF as the
rightmost term in the aforementioned equation in the cylin-
drical coordinates system. The EFCF is the normalized value
of the electrical field in the r direction. The normalized
r-directional value of the electric field can vary from �1 to
11, where a positive sign means that the divergence of the
electric field at the calculated position tends toward outside
areas, while a negative sign means that it converges on the
spinning axis. To calculate the EFCF for a single nozzle, we
used an applied electric field of 0.8 kV/cm.
The graphs in Figure 2(b,c) show that using the cylindri-
cal electrode results in a stable initial jet emitted straight
from the nozzle tip. The calculated results were compared to
an experimental test. Figure 3 shows that the initial jet from
the nozzle spinneret with a cylindrical electrode is stable.
This experimental observation is in good agreement with the
numerical prediction.
The variable solution parameters, such as the electric con-
ductivity and surface tension, were characterized to determine
the effect of a chemical BA on the formation of nanosize
fiber mats. Figure 4 shows the electrical conductivity and sur-
face tension of PCL solutions containing various amounts of
BA. Increasing the weight percent of BA in the PCL solution
increased the conductivity of the mixture while decreasing
the surface tension. Optical microscope images of the result-
ant submicron nanofibers prepared using different weight per-
centages of BA are shown in Figure 5. From the figure, the
average diameter of spun fibers with 1 wt% BA was 0.74 mmversus 0.53 mm with no BA. This means that the electrical
conductivity and surface tension play an important role in
nanosize filament formation.
Figure 6 shows a process diagram for several flow rates and
applied electric fields. At a constant flow rate, when the solution
contained a BA, a higher electric field was needed to form nano-
fibers compared to a pure PCL solution, and increasing the con-
tent of BA markedly increased the unstable region where the
spinning process produced a mixture of drops and nanofibers.
Effect of a BA on the Porosity of the Nanofiber Mat
A mat composed of nanofibers forms a good scaffold, with
high porosity and a large surface area, and allows good cell
adhesion, propagation, and growth.4,26 Although a nanofiber
Figure 7. SEM images of PCL scaffolds made with and without a BA.
Figure 8. SEM image of a thin cross section of a slightly melted PCLnanofiber mat.
107HIGHLY POROUS 3D NANOFIBER SCAFFOLD
Journal of Biomedical Materials Research Part B: Applied BiomaterialsDOI 10.1002/jbmb
structure has benefits as a biomedical scaffold, several weak-
ness have been reported. These include poor mechanical prop-
erties, too small an interpore size between the electrospun
nanofibers, and difficulty making a 3D pore structure.4,28
In this section, to fabricate a nanofiber scaffold with a
large interpore structure and enhanced mechanical proper-
ties, a chemical BA and heating process were used. This
method is similar to the gas forming technique in that the
microporosity is generated by decomposition of the BA.29–31
Figure 7(a,b) is scanning electron micrographs of a typical
electrospun PCL nanofiber mat and a blown PCL nanofiber
mat. The nanofiber mat in Figure 7(a) was obtained from con-
ventional electrospinning using 8 wt% of PCL solution under
20 kV with 250 mm between the nozzle tip and target. The
distribution of the interpore size of the mat ranged from 10 to
130 mm. The electrospun mat in Figure 7(b) was fabricated
with the assistance of an auxiliary electrode with 1 wt% of a
BA in the solution. The nanofiber mat was made in 2 s at
1008C in a drying oven. As shown in Figure 7(b), the nano-
fiber mat melted slightly due to the low melting temperature
of PCL, and two blown spots are visible. These blown spots
averaged 50–70 mm in size. From this structure, we postulate
that the pores generated using blowing media might provide
good access to deliver seeded cells into a 3D scaffold.
Although the optimal pore size does not need to be
defined because the pore size is dependent on the tissue type
desired, pore size and the porosity of a scaffold clearly influ-
ence cell division and growth. Figure 8 shows a SEM image
of a thin cross section of a nanofiber mat. This slightly melted
nanofiber mat might have sufficient porosity in the thickness
direction, and the structure has many interpore connections
resulting from the electrospun nanofibers. We think although
the pores are too small to allow vascularization, the pore
structure might be enlarged by degrading the PCL scaffold,
which would allow nutrients to reach cells seeded within the
scaffold.
To observe the effect of the blown nanofiber mat to the
cell proliferation, the HDFs cells were embedded in both the
normal electrospun fiber mat and the mat blown with 1 wt%
of BA and maintained in culture for 3 days. As shown in the
Figure 9, cells were spread on the nanofibers matrixes and
show interactions varied in the different scaffold structures.
When the Figure 9(a,b) were compared, the blown PCL mat
had good attachment with the cells compared with a nor-
mally spun fiber mat. As shown in the Figure 9(b), although
the HDFs cells were well spread in the surface of the blown
nanofiber web, the cells were difficult to penetrate to the
thickness direction of the scaffold. We believed that it was
because of the hydrophobic property of the PCL.
The graph in Figure 10 plots pore size and the porosity of
a nanofiber mat against the amount of BA. In Figure 10, the
increase in porosity with a BA was calculated as (porosity
generated with a BA)/(total porosity of the mat). Both pore
size and porosity increased with increasing amounts of BA.
Tensile Properties of Blown Nanofiber Mats
The mechanical properties of a scaffold are an important
design parameter for maintaining mechanical stability when
the scaffold is used to fill a defect in a host and retain struc-
tural intergrity.4,29 The tensile behavior of the electrospun
mat was measured using a Tytron (MTS) at a 10 mm/min
crosshead speed, and is presented in Figure 11. As Figure
11(a) shows the maximum strength and tensile modulus of a
nanofiber mat that was electrospun with a BA, but not blown
in a drying oven, was lower than that of a nanofiber mat
electrospun with a PCL solution without a BA. This differ-
Figure 9. SEM images of human skin fibroblasts on (a) a normal electrospun PCL nanofiber and (b) the blown nanofiber mat.
Figure 10. Effect of BA content on pore size and the increase in po-rosity using a BA for a PCL scaffold fabricated using electrospinning.
108 KIM AND KIM
Journal of Biomedical Materials Research Part B: Applied BiomaterialsDOI 10.1002/jbmb
ence may have resulted from the diameter of the nanofibers
or the interconnectedness and arrangement of randomly
stacked nanofibers. Figure 11(b) shows the stress–strain
curve obtained from samples of electrospun mats. Compared
to Figure 11(a), all the tensile properties increased due to
the enhanced interconnections between slightly melted nano-
fibers. The detailed tensile properties of several nanofiber
mats are compared in Table I.
CONCLUSIONS
We developed a new method of manufacturing microporous
scaffolds for specific tissue engineering applications that
involved electrospinning with an auxiliary cylindrical elec-
trode and a BA. The auxiliary electrode connected to the
spinneret stabilized the unstable initial jet and was influ-
enced by the BA. The effect of the electrode was validated
theoretically and experimentally. High microporosity and
large pores in the nanofiber scaffold were generated via the
decomposition of a chemical BA and this was improved by
melting the nanofibers slightly. The blown nanofiber mat has
good tensile properties compared to a typical nanofiber mat.
The blown nanofiber web provides a favorable environment
for HDFs cell attachment and proliferation comparable to a
normal electrospun web.
Future studies will examine cell culturing using smooth
muscle cells to mimic natural small-diameter blood vessels
and characterize cell proliferation and measure the penetra-
tion of seeded cells into a blown 3D scaffold.
The authors thank Ms. SunA Song, Mr. HyukSub Han, Mr.JongHa Park, Dr. YoungSam Cho, and SuA Park for their experi-mental and calculation assistances.
REFERENCES
1. Alberts B, Bray D, Lewis J, Raff M, Roberts K, Watson JD.Molecular biology of the cell. New York: Garland Publishing;1994. 971p.
2. Langer R, Vacanti JP. Tissue engineering. Science 1993;260:920–926.
3. Hutmacher DW. Scaffold design and fabrication technologiesfor engineering tissue: State of the art and future perspectives.J Biomater Sci 2001;12:107–124.
4. Li W-J, Laurencin CT, Caterson EJ, Tuan RS, Ko FK. Electro-spun nanofibrous structure: A novel scaffold for tissue engi-neering. J Biomed Mater Res 2002;60:613–621.
5. Boyan BD, Hummert TW, Dean DD, Schwartz Z. Role ofmaterial surfaces in regulating bone and cartilage cell response.Biomaterials 1996;17:137–146.
6. von Recum AF, Shannon CE, Cannon CE, Long KJ, van Koo-ten TG. Surface roughness, porosity, and texture as modifiersof cellular adhesion. Tissue Eng 1996;2:241–253.
7. Phan DT, Gault RS. A comparison of technologies. Int J MachTool Manuf 1998;38:1257–1287.
8. Hutmacher DW, Sittinger M, Risbud MV. Scaffold-based tissueengineering: Rationale for computer-aided design and solid free-form fabrication systems. Trends Biotechnol 2004;22:354–362.
9. Mikos AG, Sarakinos AJ, Leite SM, Vacanti JP, Langer R.Laminated three-dimensional biodegradable foams for use intissue engineering. Biomaterials 1993;14:323–330.
10. Mikos AG, Thorsen AJ, Czerwonka LA, Bae Y, Langer R,Winslow DN, Vacanti JP. Preparation and characterization ofpoly(L-lactic Acid) foams. Polymer 1994;35:1068–1077.
11. Freed LE, Marquis JC, Nohria A, Emmanual J, Mikos A,Langer R. Neocartilage formation in vitro and in vivo usingcells cultured on synthetic biodegradable polymers. J BiomedMater Res 1993;27:11–23.
12. Gomes ME, Godinho JS, Tchalamov D, Cunha AM, Reis RL.Alternative tissue engineering scaffolds based on starch: Proc-essing methodologies, morphology, degradation and mechanicalproperties. Mater Sci Eng C 2002;20:19–26.
Figure 11. The tensile properties of a nanofiber PCL scaffold prepared by electrospinning using a chemical BA.
TABLE I. Comparison of the Tensile Properties
Samples
Modulus
(Mpa)
Max. Strength
(Mpa)
Max. Strain
(%)
Before forming
W/O BA 0.25 4.17 37.6
With BA 0.11 3.62 70.9
After forming
W/O BA 0.61 7.42 150.5
With BA 0.41 5.41 97.0
109HIGHLY POROUS 3D NANOFIBER SCAFFOLD
Journal of Biomedical Materials Research Part B: Applied BiomaterialsDOI 10.1002/jbmb
13. Mano JF, Vaz CM, Mendes SC, Reis RL, Cunha AM.Dynamic mechanical properties of hydroxyapatite-reinforcedand porous starch-based degradable biomaterials. J Mater Sci:Mater Med 1999;10:857–862.
14. Dahlberg L, Kreicbergs A. Demineralized allogenic bone matrixfor cartilage repair. J Orthop Res 1991;9:11–19.
15. Watt FM, Dudhia J. Prolonged expression of differentiatedphenotype by chondrocytes cultured at low density on a com-posite substrate of collagen and agarose that restricts cellspreading. Differentiation 1988;38:140–147.
16. Ponticiello MS, Schinagl RM, Kadiyala S, Barry FP. Gelatin-based resorbable sponge as a carrier matrix for human mesen-chymal stem cells in cartilage regeneration therapy. J BiomedMater Res 2000;52:246–255.
17. Agraval CM, Ray RB. Biodegradable polymeric scaffolds formusculoskeletal tissue engineering. J Biomed Mater Res 2001;55:141–150.
18. Venugopal J, Ma LL, Yong T, Ramakrishna S. In vitro studyof smooth muscle cells on polycarprolactone and collagen nano-fibrous matrics. Cell Biol Int 2005;29:861–867.
19. Chu CF, Lu AL, Sipehia R. Enhanced growth of animal andhuman endothelial cells on biodegradable polymer. BiochimBiophys Acta 1999;1472:479–485.
20. Baumgarten PK. Electrostatic spinning of acrylic microfibres.J Colloid Interface Sci 1971;36:71–79.
21. Doshi J, Reneker DH. Electrospinning process and applicationof electrospun fibers. J Electrostatics 1995;35:151–160.
22. Yu JH, Fridrikh SV, Rutledge GC. Production of submicrondiameter fibers from difficult-to-process materials by two-fluidelectrospinning. Adv Mater 2004;16:1562–1566.
23. Kim G-M, Michler GH, Potschke P. Deformation processes ofultrahigh porous multiwalled carbon nanotubes/polycarbonatecomposite fibers prepared by electrospinning. Polymer 2005;46:7346–7351.
24. Han SO, Son WK, Cho D, Youk JH, Park WH. Preparationof porous ultra-fine fibres via selective thermal degradationof electrospun polyetherimide/poly(3-hydroxybutyrate-co-3-hy-droxyvalerate) fibres. Polym Degrad Stab 2004;86:257–262.
25. You Y, Youk JH, Lee SW, Min B-M, Lee SJ, Park WH. Prepara-tion of porous ultrafine PGA fibers via selective dissolution ofelectrospun PGA/PLA blend fibers. Mater Lett 2006;60:757–760.
26. Zong X, Bien H, Chung C-Y, Yin L, Fang D, Hsiao BS, Chu B,Entcheva E. Electrospun fine-textured scaffolds for heart tissueconstructs. Biomaterials 2005;26:5330–5338.
27. Kim GH, Cho YS, Kim WD. Stability analysis for multi-jetselectrospinning process modified with a cylindrical electrode.Eur Polym J. In press.
28. Smith LA, Ma PX. Nano-fibrous scaffolds for tissue engineer-ing. Colloids Surf B: Biointerfaces 2004;39:125–131.
29. Gomes ME, Godinho JS, Tchalamov D, Cunha AM, Reis RL.Alternative tissue engineering scaffolds based on starch: Proc-essing methodologies, morphology, degradation and mechanicalproperties. Mater Sci Eng C 2002;20:19–26.
30. Harris LD, Kim BS, Mooney DJ. Open pore biodegradablematrices formed with gas foaming. J Biomed Mater Res 1998;42:396–402.
31. Gravel M, Gross T, Vago R, Tabrizian M. Responses of mesen-chymal stem cell to chitosan–coralline composites microstruc-tured using coralline as gas forming agent. Biomaterials 2006;27:1899–1906.
110 KIM AND KIM
Journal of Biomedical Materials Research Part B: Applied BiomaterialsDOI 10.1002/jbmb