ARTICLE IN PRESS
Biomaterials 29 (2008) 955 962
www.elsevier.com/locate/biomaterials
Mechanical properties of single electrospun collagen type I fibers
a
Lanti Yanga, Carel F.C. Fitie´ , Kees O. van der Werfb, Martin L. Benninkb,
Pieter J. Dijkstraa, Jan Feijena,
a
Polymer Chemistry and Biomaterials, Faculty of Science and Technology and Institute for Biomedical Technology (BMTI),
University of Twente, P.O. Box 217, 7500 AE, Enschede, The Netherlands
b
Biophysical Engineering, Faculty of Science and Technology and MESA+ Institute for Nanotechnology,
University of Twente, P.O. Box 217, 7500 AE, Enschede, The Netherlands
Received 14 August 2007; accepted 26 October 2007
Abstract
The mechanical properties of single electrospun collagen fibers were investigated using scanning mode bending tests performed with an
AFM. Electrospun collagen fibers with diameters ranging from 100 to 600 nm were successfully produced by electrospinning of an 8% w/
v solution of acid soluble collagen in 1,1,1,3,3,3-hexafluoro-2-propanol (HFP). Circular dichroism (CD) spectroscopy showed that 45%
of the triple helical structure of collagen molecules was denatured in the electrospun fibers. The electrospun fibers were water soluble and
became insoluble after cross-linking with glutaraldehyde vapor for 24 h. The bending moduli and shear moduli of both non- and cross-
linked single electrospun collagen fibers were determined by scanning mode bending tests after depositing the fibers on glass substrates
containing micro-channels. The bending moduli of the electrospun fibers ranged from 1.3 to 7.8 GPa at ambient conditions and ranged
from 0.07 to 0.26 MPa when immersed in PBS buffer. As the diameter of the fibrils increased, a decrease in bending modulus was
measured clearly indicating mechanical anisotropy of the fiber. Cross-linking of the electrospun fibers with glutaraldehyde vapor
increased the shear modulus of the fiber from 30 to 50 MPa at ambient conditions.
r 2007 Elsevier Ltd. All rights reserved.
Keywords: Electrospun fibers; Collagen type I; Cross-linking; Atomic force microscopy; Bending modulus; Shear modulus
1. Introduction polymer solution, polymer fibers can be produced in a
continuous and relatively easy way. Also, electrospinning
Fibers with diameters ranging from a few to hundreds of allows control over the fiber diameter and orientation of
nanometers have found a broad range of applications in the fibers in a mesh [8]. Many biodegradable synthetic
different areas, such as reinforcement in composite polymers have been electrospun into fibrous meshes and
materials, highly porous membrane supports and electro- successfully used in cell culture systems for tissue engineer-
nics [1 3]. Both synthetic and natural polymer fibers have ing [9 11]. Recently, several studies have been performed
been extensively studied for the preparation of scaffolds for using natural polymer fibers as scaffolding materials
tissue engineering applications [4 6]. It is envisaged that because of their biocompatibility and resorbability
the high surface to volume ratio of fiber meshes enhances [12 14]. Electrospinning was used to prepare tubular
the efficiency of mass transport and cell attachment to scaffolds of soluble collagen type I for tissue engineering
scaffolds [5,7]. of blood vessels [15]. These electrospun collagen fibers had
Electrospinning is a suitable technique for the produc- diameters ranging from 100 to 700 nm. Preliminary cell
tion of fibers with diameters smaller than 1 mm and has a culture experiments have been performed successfully with
number of advantages. By applying a high voltage to a the obtained collagen tubular scaffolds.
In order to mimic the physical and chemical structure of
the native extracellular matrix, understanding of the
Corresponding author. Tel.: [phone]; fax: [phone].
E-mail address: [e-mail] (J. Feijen). material properties is important for an optimal scaffold
0142-9612/$ - see front matter r 2007 Elsevier Ltd. All rights reserved.
doi:10.1016/j.biomaterials.2007.10.058
ARTICLE IN PRESS
956 L. Yang et al. / Biomaterials 29 (2008) 955 962
glutaraldehyde vapor [15,20,21]. Crosslinking was carried out by placing
design. The mechanical properties of the scaffold must be
the substrate with the electrospun collagen fibers above 20 ml of 25 wt%
taken into account because the scaffold must be able to
glutaraldehyde solution (grade I, Sigma, Zwijndrecht, the Netherlands) in
withstand the forces exerted by, e.g. pulsed blood flow [7].
a sealed beaker for 24 h at room temperature. After the cross-linking
Also, different types of cells may require different
reaction, the samples were left to dry at ambient conditions for at least
mechanical properties of the scaffolds. Therefore, under- 24 h under a glass cover.
The cross-linked and non-crosslinked electrospun collagen fibers were
standing the mechanical behavior of the single fibers in a
immersed in MilliQ water for 4 h to test their solubility.
scaffold is essential in optimization of the scaffold design.
Although polymer fibers have been successfully used in
2.3. Determination of the primary amino group content
the preparation of scaffolds, only recently a few studies
have focused on the mechanical properties of single fibers
The primary amino group content of non- and cross-linked electrospun
in the sub-micron range [7,16,17]. By performing tensile
collagen fibers was determined using the 2,4,6-trinitrobenzenesulfonic acid
tests with single polymer fibers, the Young s modulus,
(TNBS) assay. Electrospun collagen fibers (3 5 mg) cut from the wire
tensile strength and the strain at break of the tested fibers
frame were incubated for 30 min in 1 ml of a 4 wt% solution of NaHCO3.
To this mixture 1 ml of a freshly prepared solution of TNBS (0.5 wt%) in
were determined. However, up to now studies of the
4 wt% NaHCO3 was added. The resulting mixture was left for 2 h at 40 1C.
mechanical properties are limited in the fiber direction.
After the addition of HCl (3 ml, 6 M), the temperature was raised to 60 1C.
Alignment and reorientation of molecules is known to
Hydrolyzation of collagen was achieved within 90 min. The resulting
occur during electrospinning [18,19]. Knowing the shear
solution was diluted with 5 ml MilliQ water and cooled to room
modulus of electrospun fibers is also important in
temperature. The absorbance at 420 nm was measured using a Varian
Cary 300 Bio spectrophotometer (Middelburg, The Netherlands). A blank
investigating their effect on cellular activities.
was prepared applying the same procedure, except that HCl was added
In this paper, we present a study on the mechanical
before the addition of TNBS. The absorbance was correlated to the
properties of single electrospun collagen fibers. Micromecha-
concentration of free amino groups using a calibration curve obtained
nical bending tests were performed on native and glutar-
with glycine. The primary amino group content was expressed as the
aldehyde cross-linked single electrospun fibers using atomic
number of primary amino groups per 1000 amino acids (n/1000).
force microscopy (AFM). Both bending moduli and shear
moduli of the electrospun collagen fibers were determined.
2.4. Circular dichroism
The mechanical anisotropy of the electrospun fibers was
investigated and related to the structure of the fibers.
Solutions of untreated acid soluble collagen and electrospun collagen
fibers were prepared by dissolving 5 mg dry material in 5 ml acetic acid
(0.1 M) and stirring for 3 h at room temperature. Subsequently, the clear
2. Materials and methods
solutions were equilibrated overnight at 4 1C. A solution prepared from
untreated acid soluble collagen was placed in an oven at 95 1C for 20 min
2.1. Electrospinning of collagen
to obtain thermally denatured acid soluble collagen [22,23]. CD spectra
(Jasco J-715 spectropolarimeter) were recorded at 25 1C over the
wavelength interval 180 260 nm at a scan speed of 20 nm/min and data
An 8% w/v solution of acid soluble collagen type I from calf skin (Elastin
Products Company Inc., Owensville, USA) in 1,1,1,3,3,3-hexafluoro- were averaged from 5 scans. A bandwidth of 1 nm and a time constant of
1 s were used in the measurements. The path length of the cylindrical
2-propanol (HFP) was prepared for electrospinning [15]. The spinning
sample cell was 0.1 mm. A reference spectrum of 0.1 M acetic acid was
solution was transferred to a 50 ml syringe (Perfusor-Spritze OPS, B-Brain,
recorded, and subtracted from the sample spectra. The circular dichroism
Melsungen, Switzerland) which was connected to a syringe pump (Perfusor
(CD) of the collagen is expressed as mean molar residual ellipticity.
E secura, B-Braun, Melsungen, Switzerland). The spinning solution was
The fraction of triple helical collagen in the samples was calculated
pushed through a silicon tube (Nalgene 50, 1.6 mm ID 3.1 mm OD, New
using Eq. (1) assuming a linear dependence of the individual data [23,24]:
York, USA) into a blunt steel needle (22G 11  , 0.7 mm ID 30 mm length)
4
at a constant speed of 80 ml/min. The needle was attached to the electrode of
½yŠE ½yŠDN
a high voltage supply (Bertan Associates Inc., 230R, Hicksville, NY, USA). f ź (1)
100,
TH
½yŠC ½yŠDN
A stainless steel ring was connected to the same electrode and placed just
under the needle tip to stabilize the jet and direct it downwards. A grounded where fTH is the fraction triple helices in the electrospun fibers determined
piece of aluminum foil (12 12 cm2) was placed at a 15 20 cm distance from at wavelength l (%) [y]E the ellipticity of electrospun collagen at
the needle tip to collect the fibers. A continuous production of collagen wavelength l (deg cm2 d mol 1) [y]DN the ellipticity of denatured soluble
electrospun fibers was obtained at 19 21 kV. collagen at wavelength l (deg cm2 d mol 1) [y]C the ellipticity of soluble
During the electrospinning quartz glass substrates containing micro- collagen at wavelength l (deg cm2 d mol 1).
channels with a depth of 600 nm and width of 3 mm were placed on the
aluminum foil for 30 s to collect a suitable amount of electrospun fibers for
2.5. Micromechanical tests of electrospun collagen fibers using
the micromechanical bending tests. Electrospun fibers were also collected
AFM
on a rotating rectangular stainless steel wire frame (wire diameter 0.5 mm,
12 3cm2) placed just above the collecting aluminum plate in order to
Micromechanical bending tests were performed as previously described
collect samples for determining the primary amino group content and the
[25]. In brief, triangular silicon nitride cantilevers (coated sharp micro-
denaturation temperature of the collagen fibers. The collecting time was
levers MSCT-AUHW, type F, spring constant kź0.5 N/m, Veeco,
varied between 30 and 180 min.
Cambridge, UK) of which the tips were removed were used in the bending
tests. The spring constant of each tip-less cantilever was calibrated by
2.2. Cross-linking of the electrospun fibers
pushing on a pre-calibrated cantilever as described elsewhere [26]. Single
electrospun collagen fibers freely and perpendicularly spanning multiple
Electrospun collagen fibers cut from the wire frame and deposited on channels of the glass substrate were selected for the bending tests in
the glass substrate containing microchannels were cross-linked using scanning mode. The electrospun collagen fiber was bent by pushing with
ARTICLE IN PRESS
L. Yang et al. / Biomaterials 29 (2008) 955 962 957
the AFM cantilever in the z-direction. A typical piezo movement of
2.7. Morphology
1.5 3.0 mm with a frequency of 1.3 Hz in the z-direction was applied. In the
scanning mode bending test, after every step of bending, the tip is moved
The morphology and diameter of all micromechanically tested
one step further along the fiber. A complete scan consists of 256 steps, the
electrospun collagen fibers was determined from SEM measurements
total measuring time is 200 s, and the total scanning distance was chosen to
(Leo Gemini 1550 FEG-SEM). No sputter coating was applied on the
be slightly larger than the channel width. The mechanical properties of the
sample for imaging.
elecrospun collagen fibers were investigated both at ambient conditions
and in PBS buffer (pHź7.4).
3. Results and discussion
3.1. Electrospun collagen fibers
2.6. Data analysis
Electrospinning of soluble collagen into meshes that can
A custom computer program, written in Labview (version 6.1, National
Instruments, Austin, TX, USA) was used to analyze the data. A force
be applied for tissue engineering applications is not
displacement curve of every 256 bending measurements was obtained
possible using aqueous solutions but was recently accom-
using the following equations:
plished by Matthews et al. [15] using HFP. The viscosity
and conductivity of the solution, the spinning voltage, the
zźA D, (2)
flow rate and the collecting distance were adjusted in
preliminary experiments to optimize the electrospinning
FźDk. (3)
conditions. During electrospinning of an 8% w/v collagen
In which z is the displacement of the fiber in the z-direction during
solution in HFP, jet formation was first observed at 12 kV.
bending, A is the piezo movement in the z-direction and D is the deflection
Increasing the voltage to values between 19 and 21 kV
of the cantilever in nanometer. F is the force applied to the fiber and k is
offered stable operating conditions to electrospin continu-
the calibrated spring constant of the cantilever.
The bending modulus (Ebending) of the single electrospun fibers was ously for more than 3 h. Due to evaporation of HFP, dry
deduced by fitting the slope of the force displacement curve at different
collagen fibers could be collected at a distance of 15 20 cm
positions (x) of the suspended fiber to the following equation [27]:
on a grounded plate. A deposition rate of 22 mg/h was
obtained using a wire frame collector placed just above the
dF 3l3EbendingI
ź . (4)
grounded plate. It is noted that the collagen fibers
dz
ðl xÞ3x3
produced by this method using the mentioned parameters
In which x is the relative position along the fiber (0pxpl/2), l is the
have similar dimensions as native collagen fibrils and range
width of the channel, I is the moment of inertia (Iź1pR4), dF/dz is
4
from 100 to 600 nm.
the slope of the force displacement curve obtained during bending of the
In the electrospinning process, the morphology of
electrospun fibers. The fiber is considered to be a rod with a circular cross-
section with radius R. the resulting electrospun fiber is influenced by many
Fig. 1. (A) SEM image of collagen fibers electrospun from an 8% w/v solution of acid soluble collagen type I from calf skin in HFP. (B) High-resolution
SEM image of electrospun collagen fibers showing a smooth surface. (C) SEM image of an electrospun collagen fiber spanning a microchannel in the glass
substrate. The channel width is 4 mm and the diameter of this electrospun fiber is 190 nm.
ARTICLE IN PRESS
958 L. Yang et al. / Biomaterials 29 (2008) 955 962
parameters [2,28,29], and was analyzed by SEM. As shown reported that HFP as a highly fluorinated alcohol
in Fig. 1(A), most fibers have a smooth surface. Some destabilizes the native triple helical structure of the protein
split and ribbon-type fibers were also observed. High- although the exact mechanism of the process is not
resolution SEM measurements showed that the character- completely understood yet [34,35].
istic D-period found in native collagen fibrils is not present Electrospun collagen fibers were collected on glass
(Fig. 1(B)). It is known that soluble collagen molecules substrates containing micro-channels during electrospin-
assemble in a structure similar to native collagen fibrils ning for the micromechanical tests. The diameter and
depending on the pH and electrolytes [30,31]. The absence suspended length of every tested fiber was determined from
of a D-period indicates that the structure is different from a SEM image (Fig. 1(C)) after the bending measurements.
that of native collagen. Moreover, it is known that
aliphatic alcohols affect the secondary structure of proteins 3.2. Collagen CD spectral analysis
[32,33]. Also, in the specific case of collagen, it has been
To gain information on the fraction of the triple helical
secondary structure present in the electrospun collagen
10000
fibers, CD measurements were performed. The triple
helical secondary structure found in native collagen gives
0
rise to a characteristic positive peak at around 220 nm in
the CD spectrum which decreases upon denaturation. It
 10000
was reported that complete denaturation results in the total
 20000
disappearance of the positive peak at 220 nm [22,23]. The
CD spectra of the electrospun collagen fiber, untreated acid
 30000
soluble collagen and thermally denatured acid soluble
collagen have been measured and are shown in Fig. 2.
Soluble collagen
 40000
Electrospun collagen
The general shape and peaks of the spectrum change when
Denatured soluble collagen
the collagen is denatured and are in agreement with
 50000
literature data. The spectrum of the resolubilized electro-
180 200 220 240 260 spun collagen fibers displays positive and negative peaks
wavelength (nm)
with lower intensity compared to the native soluble
collagen, which implies that the electrospun collagen was
Fig. 2. Circular dichroism (CD) spectra of solutions prepared from
partly denatured. As mentioned above, using HFP as a
electrospun collagen and soluble collagen in 0.1 M acetic acid at 25 1C. All
solvent may destabilize the native helical structure of the
the spectra shown were obtained after subtraction of baseline spectrum for
the acetic acid. collagen. The triple helical fraction of the partly denatured
Fig. 3. SEM images of electrospun collagen fibers: (A) without cross-linking before immersion in Milli Q water; (B) without cross-linking after immersion
in Milli Q water at room temperature for 4 h; (C) cross-linked with glutaraldehyde before immersion in Milli Q water; and (D) cross-linked with
glutaraldehyde after immersion in Milli Q water at room temperature for 4 h.
2
-1
Mean residual elipticity (deg.cm dmol )
ARTICLE IN PRESS
L. Yang et al. / Biomaterials 29 (2008) 955 962 959
electrospun collagen fibers is approximately 45% calcu- explained by the alignment of individual molecules
lated from the positive peak using Eq. (1). The lower triple preferably in the axial direction during the spinning process
helical content might alter or disturb the hierarchical order [18,19]. The alignment of the collagen molecules may
as seen in native soluble collagen, and explain the absence enhance intermolecular interactions and result in a bending
of any periodicity in the structure. modulus of the electrospun collagen fibers, similar to that
of native collagen fibrils.
3.3. Fiber stability in aqueous solution before and after For glutaraldehyde cross-linked electrospun collagen
cross-linking fibers the mechanical properties were measured both for
fibers at ambient conditions and for fibers in PBS buffer.
Non-cross-linked electrospun fibers (Fig. 3(A)) were The bending modulus of the fibers is not affected by the
soluble in water. After 4 h immersion in Milli Q water, the cross-linking process. The same values and dependency on
fibers appear to flatten (Fig. 3(B)). Thus, to potentially use the fiber diameter were observed (Fig. 5(A)). When
electrospun collagen fibers for biomedical applications, like immersed in PBS buffer lower Ebending values were
a scaffold material, cross-linking of the collagen is obtained but the moduli show a similar dependency on
necessary. Cross-linking was performed by placing the the fiber diameter (Fig. 5(C)). The dependence of the
electrospun fibers in a sealed beaker filled with glutaralde- bending modulus on the diameter indicates the presence of
hyde (GA) vapor. The stability of the cross-linked fibers shearing between segments in the electrospun collagen
was subsequently tested by incubating the samples in water
for 4 h. SEM pictures as represented in Fig. 3(C) and (D)
shows that these fibers cross-linked with GA preserve their
fibrous structure on the supporting glass surface.
200
The degree of cross-linking was estimated from the a b
c
decrease in primary amino groups when compared to non-
cross-linked collagen. The primary amino group content
150
decreased from 36 to 4 per 1000 amino acids and is
comparable to previously reported data on GA cross-
100
linking [36].
3.4. Micromechanical properties of electrospun collagen
50
fibers
0
After a fiber that perpendicularly crossed a channel in
the glass substrate was selected, the cantilever was
0 50 100 150 200
positioned above the fiber at the edge of the channel.
Displacement (nm)
Bending experiments were then performed by recording
256 force distance curves along the suspended fiber. The
vertical piezo movement was between 1.5 and 3.0 mm and
3.5
adjusted in such a way that the maximum displacement of
the fiber was 200 nm. As shown in Fig. 4(A), a linear
3.0
increase of force was found from bending the electrospun
fibers. The slope (dF/dz) of each force displacement curve
2.5
of the tested electrospun fibers was determined by a linear
fit. Fitting the data as a function of the scanning position to
2.0
Eq. (4), afforded the Ebending of the single electrospun
collagen fiber. As shown by the curve fit (Eq. (4)) in
1.5
Fig. 4(B), the experimental data do fit to Eq. (4) with a
standard error of 2 6% in the least-squares fit parameter.
1.0
The bending moduli of several electrospun fibers
edge
middle
determined at ambient conditions with diameters ranging
0.5
from 150 to 450 nm are presented in Fig. 5(A). The bending
0.6 0.8 1.0 1.2 1.4 1.6 1.8 2.0 2.2
modulus decreases from 7.571.6 to 1.470.2 GPa with
increasing fiber diameter up to 250 nm and remains at a
Fig. 4. (A) Three force displacement curves obtained from bending the
constant value of 1.4 GPa for fibers with larger diameters.
electrospun collagen fiber at different positions: (a) at the edge of
It is interesting to note that the bending modulus of the
the channel; (b) between the edge and middle of the channel; and (c) at the
fibers is comparable to the modulus of native collagen
middle of the channel. (B) Graph showing the slope (dF/dz) of the
fibrils with comparable diameters [37] although the
force displacement curves as a function of the scanning position. A curve
electrospun fibers are partly denatured. This can be fit using Eq. (4) is shown as a solid line.
Force (nN)
dF/dz
(nN/nm)
ARTICLE IN PRESS
960 L. Yang et al. / Biomaterials 29 (2008) 955 962
1.2
10 Non-crosslinked electrospun
collagen fibers
Cross-linked electrospun 1.0
8
collagen fibers
0.8
6
0.6
4
0.4
2
0.2 G=29.4Ä…2.6 MPa
G=48.0Ä…6.0 MPa
0.0
0
150 200 250 300 350 400 450 500 0.5 1.0 1.5 2.0 2.5
Fiber diameter (nm) R2 / I2 ( x10-3 )
0.35
Cross-linked electrospun 18
collagen fibers in PBS buffer
0.30
16
0.25 14
12
0.20
10
0.15
8
0.10
6
G=5.2Ä…0.9 MPa
4
0.05
2
200 250 300 350 400 450 0 1 2 3 4 5 6
Fiber diameter (nm) R2 / I2 ( x10-3 )
Fig. 5. (A) Bending modulus as a function of the diameter of non-cross-linked electrospun collagen fibers ( ) and cross-linked electrospun fibers (&) at
ambient conditions. (B) Reciprocal bending modulus (Ebending) as a function of R2/l2 of non-cross-linked electrospun collagen fibers ( ) and cross-linked
electrospun fibers (&) at ambient conditions. The slope of the linear fit (solid line: non-cross-linked electrospun fibers; dotted line: cross-linked electrospun
fibers) equals to 120/(9G), from which the shear modulus (G) was calculated. (C) Bending modulus as a function of the diameter of cross-linked
electrospun fibers (K) in PBS buffer. (D) Reciprocal bending modulus (Ebending) as a function of R2/l2 of cross-linked electrospun fibers (K) in PBS buffer.
The shear modulus (G) was calculated as in (B).
fibers [38,39]. Using the applied force and the displacement magnitudes lower than the bending modulus and increases
of the fiber at the middle of the channel, the shear modulus to 48.076.0 MPa for cross-linked fibers. As shown above,
can be calculated from the following equation [27]: the primary amino group content decreased upon glutar-
aldehyde cross-linking from 35 to 4. The intra- and inter-
zźzBþzSźFl3=192EIþf Fl=4GA
s
molecular covalent bonds formed between glutaraldehyde
and the amine groups of collagen are responsible for the
źFl3=192EbendingI ð5Þ
improved mechanical strength in the lateral direction. The
shear modulus of the cross-linked electrospun collagen
and therefore
fibers in PBS buffer was found to be 5 MPa (Fig. 5(D))
1 1 12f R2
s which is 20 50 times lower than the bending modulus.
ź þ . (6)
Ebending E G The fact that the shear modulus of the fiber is one to two
l2
magnitudes lower than the bending modulus indicates that
In which, z is the total displacement of the fiber in the the single electrospun fibers are mechanically anisotropic.
z-direction, zB is the deflection from bending, zS is the From the CD results, it was shown that a high percentage
deflection from shearing, E is the Young s modulus, G is of the collagen triple helical structure has been changed
the shear modulus, fs is the form factor of shear which to random coil in the electrospun fibers. Therefore,
equals 10/9 for a cylinder and A is the cross-sectional area the shearing could result from displacement between
of the fiber. collagen molecules and individual a-chains with respect
Plotting 1/Ebending versus (R2/l2), the shear modulus (G) to each other.
of the (non-) cross-linked fibers at ambient conditions was From the micromechanical tests, the bending and
determined from the slope using a linear fit as shown in shear moduli of single electrospun collagen fibers
Fig. 5(B). The shear moduli of the electrospun collagen were calculated (results summarized in Table 1). The
fibers are 29.472.6 MPa, which are approximately two results indicate that the mechanical properties of single
-1
bending
bending
E
(GPa)
1/E
(GPa )
-1
bending
bending
E
(GPa)
1/E
(GPa )
ARTICLE IN PRESS
L. Yang et al. / Biomaterials 29 (2008) 955 962 961
Table 1
Bending and shear moduli of electrospun collagen fibers obtained from scanning bending measurements
Electrospun collagen fibers Conditions Range of diameters (nm)a Bending moduli (GPa)b Shear modulusc (MPa)
Non-cross-linked Dry 179 356 7.5 1.4 29.472.6
Cross-linkedd Dry 191 450 7.8 1.3 48.076.0
Cross-linkedd PBS buffere 210 425 0.26 0.07 5.270.9
a
Range of diameters: the ranges of the diameter from different electrospun fibers used in the mechanical tests. The error (SEM) of the diameter of
individual fiber is 3% calculated from multiple measurements on the same fiber.
b
Ranges of bending moduli: the range of bending moduli determined from electrospun fibers with different diameters. A 23% relative error is estimated
for the bending modulus of individual fibers [25].
c
The error in the shear modulus is the standard error of the weighted least-squares fit parameter.
d
Crosslinking was performed with glutaraldehyde vapor.
e
PBS: phosphate buffered saline, pHź7.4.
electrospun collagen fibers are comparable to the mechan- Acknowledgment
ical properties of native collagen fibrils, a principal
structural element of fiber forming collagens. The resem- This research is financially supported by the Softlink
blance of the mechanical properties makes electrospun program of ZonMw. Project no. 01SL056.
collagen fibers and also collagen polymer composites
promising materials to be applied as scaffolds materials
References
in tissue engineering.
[1] Frenot A, Chronakis IS. Polymer nanofibers assembled by electro-
spinning. Curr Opin Colloid Interface Sci 2003;8:64 75.
4. Conclusions
[2] Li D, Xia Y. Electrospinning of nanofibers: reinventing the wheel?
Adv Mater 2004;16:1151 70.
[3] Bognitzki M, Czado W, Frese T, Schaper A, Hellwig M, Steinhart M,
By electrospinning soluble collagen in HFP, it was
et al. Nanostructured fibers via electrospinning. Adv Mater 2001;
possible to obtain collagen fibers with diameters between
13:70 2.
100 and 600 nm. Cross-linking in glutaraldehyde vapor for
[4] Xu CY, Inai R, Kotaki M, Ramakrishna S. Aligned biodegradable
24 h introduced a high degree of cross-linking in the
nanofibrous structure: a potential scaffold for blood vessel engineer-
electrospun fibers and afforded water-resistant fibers. The ing. Biomaterials 2004;25:877 86.
[5] Buttafoco L, Kolkman N, Engbers-Buijtenhuijs P, Poot A,
CD spectrum of electrospun collagen fibers indicates that
Dijkstra PJ, Vermes I, et al. Electrospinning of collagen and
the fibers were partly denatured.
elastin for tissue engineering applications. Biomaterials 2006;27:
The mechanical properties of single electrospun collagen
724 34.
type I fibers were investigated by micromechanical bending
[6] Khil MS, Kim HY, Kim MS, Park SY, Lee DR. Nanofibrous mats of
tests in scanning mode. The bending moduli were derived poly(trimethylene terephthalate) via electrospinning. Polymer 2004;
45:295 301.
from the slopes of force displacement curves. The bending
[7] Chew SY, Hufnagel TC, Lim CT, Leong KW. Mechanical properties
moduli of the cross-linked electrospun collagen fibers at
of single electrospun drug-encapsulated nanofibres. Nanotechnology
ambient conditions range from 1.3 to 7.8 GPa with
2006;17:3880 91.
diameters decreasing from 450 to 190 nm and decrease
[8] Teo WE, Ramakrishna S. A review on electrospinning design and
20 times when the fibers are immersed in PBS buffer. A nanofibre assemblies. Nanotechnology 2006;17:R89 R106.
[9] Zhang YZ, Venugopal J, Huang ZM, Lim CT, Ramakrishna S.
high degree of alignment of the collagen molecules in the
Characterization of the surface biocompatibility of the electrospun
electrospun collagen fibers can explain the bending moduli
PCL collagen nanofibers using fibroblasts. Biomacromolecules 2005;
observed similar to that of native collagen fibrils. By testing
6:2583 9.
fibers with different diameters, the shear modulus of the
[10] Zhong S, Teo WE, Zhu X, Beuerman R, Ramakrishna S, Yung LYL.
tested collagen fibers was estimated from the relation- Formation of collagen glycosaminoglycan blended nanofibrous
scaffolds and their biological properties. Biomacromolecules
ship between bending modulus and different length to
2005;6:2998 3004.
diameter ratios. The electrospun fibers showed anisotropic
[11] He W, Yong T, Ma ZW, Inai R, Teo WE, Ramakrishna S.
mechanical properties with a two orders of magnitude
Biodegradable polymer nanofiber mesh to maintain functions of
lower shear modulus compared to the bending modulus.
endothelial cells. Tissue Eng 2006;12:2457 66.
Cross-linking with glutaraldehyde increases the shear [12] Zhang YZ, Venugopal J, Huang ZM, Lim CT, Ramakrishna S.
Crosslinking of the electrospun gelatin nanofibers. Polymer 2006;47:
modulus and the stability of the electrospun fibers.
2911 7.
Evaluation of the mechanical properties of single electro-
[13] Ayres C, Bowlin GL, Henderson SC, Taylor L, Shultz J, Alexander J,
spun collagen fibers results in a better understanding of the
et al. Modulation of anisotropy in electrospun tissue-engineering
potential use of the electrospun fibers as scaffolds for, e.g.
scaffolds: analysis of fiber alignment by the fast Fourier transform.
tissue engineering. Biomaterials 2006;27:5524 34.
ARTICLE IN PRESS
962 L. Yang et al. / Biomaterials 29 (2008) 955 962
[14] Bhattarai N, Li Z, Edmondson D, Zhang M. Alginate-based [28] Huang ZM, Zhang YZ, Kotaki M, Ramakrishna S. A review on
nanofibrous scaffolds: structural, mechanical, and biological proper- polymer nanofibers by electrospinning and their applications in
ties. Adv Mater 2006;18:1463 7. nanocomposites. Compos Sci Technol 2003;63:2223 53.
[15] Matthews JA, Wnek GE, Simpson DG, Bowlin GL. Electrospinning [29] Doshi J, Reneker DH. Electrospinning process and applications of
of collagen nanofibers. Biomacromolecules 2002;3:232 8. electrospun fibers. J Electrostat 1995;35:151 60.
[16] Inai R, Kotaki M, Ramakrishna S. Structure and properties of [30] Goh MC, Paige MF, Gale MA, Yadegari I, Edirisinghe M,
electrospun PLLA single nanofibres. Nanotechnology 2005;16: Strzelczyk J. Fibril formation in collagen. Physica A 1997;239:
208 13. 95 102.
[17] Tan EPS, Goh CN, Sow CH, Lim CT. Tensile test of a single [31] Jiang F, Horber H, Howard J, Muller DJ. Assembly of collagen into
¨ ¨
nanofiber using an atomic force microscope tip. Appl Phys Lett microribbons: effects of pH and electrolytes. J Struct Biol 2004;148:
2005;86, Art. No. 073115. 268 78.
[18] Gu SY, Wu QL, Ren J, Vancso GJ. Mechanical properties of a single [32] Konno T, Iwashita J, Nagayama K. Fluorinated alcohol, the third
electrospun fiber and its structures. Macromol Rapid Commun 2005; group of cosolvents that stabilize the molten-globule state relative to
26:716 20. a highly denatured state of cytochrome c. Protein Sci 2000;9:564 9.
[19] Reneker DH, Chun I. Nanometre diameter fibers of polymer, [33] Hirota N, Mizuno K, Goto Y. Cooperative a-helix formation of
produced by electrospinning. Nanotechnology 1996;7:216 23. b-lactoglobulin and melittin induced by hexafluoroisopropanol.
[20] Ramires PA, Milella E. Biocompatibility of poly(vinyl alcohol) hya- Protein Sci 1997;6:416 21.
luronic acid and poly(vinyl alcohol) gellan membranes crosslinked [34] Bianchi E, Rampone R, Tealdi A, Ciferri A. The role of aliphatic
by glutaraldehyde vapors. J Mater Sci Mater Med 2002;13:119 23. alcohols on the stability of collagen and tropocollagen. J Biol Chem
[21] Chen GP, Ushida T, Tateishi T. A hybrid network of synthetic 1970;245:3341 5.
polymer mesh and collagen sponge. Chem Commun 2000;16:1505 6. [35] Doillon CJ, Drouin R, Cote MF, Dallaire N, Pageau JF, Laroche G.
Ć
[22] Usha R, Ramasami T. The effects of urea and n-propanol on collagen Chemical inactivators as sterilization agents for bovine collagen
denaturation: using DSC, circular dicroism and viscosity. Thermo- materials. J Biomed Mater Res 1997;37:212 21.
chim Acta 2004;409:201 6. [36] Olde Damink LHH, Dijkstra PJ, van Luyn MJA, van Wachem PB,
[23] Long CG, Braswell E, Zhu D, Apigo J, Baum J, Brodsky B. Nieuwenhuis P, Feijen J. Glutaraldehyde as a cross-linking agent
Characterization of collagen-like peptides containing interruptions in for collagen-based biomaterials. J Mater Sci Mater Med 1995;6:
the repeating Gly-X-Y sequence. Biochemistry 1993;32:11688 95. 460 72.
[24] Holmgren SK, Bretscher LE, Taylor KM, Raines RT. A hyperstable [37] Yang L, van der Werf KO, Koopman BFJM, Subramaniam V,
collagen mimic. Chem Biol 1999;6:63 70. Bennink ML, Dijkstra PJ, et al. Micromechanical bending of single
[25] Yang L, van der Werf KO, Fitie´ CFC, Bennink ML, Dijkstra PJ, collagen fibrils using atomic force microscopy. J Biomed Mater Res
Feijen J. Mechanical properties of native and cross-linked type I Part A 2007;82:160 8.
collagen fibrils. Biophys J 2007, Online. [38] Kis A, Kasas S, Babic´ B, Kulik AJ, Beno1
Ć t W, Briggs GAD, et al.
[26] Torii A, Sasaki M, Hane K, Okuma S. A method for determining the Nanomechanics of microtubules. Phys Rev Lett 2002;89, Art.
spring constant of cantilevers for atomic force microscopy. Meas Sci No.248101.
Technol 1996;7:179 84. [39] Salvetat JP, Briggs GAD, Bonard JM, Bacsa RR, Kulik AJ, Stockli
¨
[27] Gere J, Timoshenko SP. Mechanics of materials. London: Chapman T, et al. Elastic and shear moduli of single-walled carbon nanotube
& Hall; 1991. p. 545 51, 691 4. ropes. Phys Rev Lett 1999;82:944 7.