You are on page 1of 10

Home Search Collections Journals About Contact us My IOPscience

Ultra-fine polyelectrolyte hydrogel fibres from poly(acrylic acid)/poly(vinyl alcohol)

This article has been downloaded from IOPscience. Please scroll down to see the full text article.

2005 Nanotechnology 16 2852

(http://iopscience.iop.org/0957-4484/16/12/020)

View the table of contents for this issue, or go to the journal homepage for more

Download details:
IP Address: 169.237.141.15
The article was downloaded on 07/10/2010 at 20:30

Please note that terms and conditions apply.


INSTITUTE OF PHYSICS PUBLISHING NANOTECHNOLOGY
Nanotechnology 16 (2005) 2852–2860 doi:10.1088/0957-4484/16/12/020

Ultra-fine polyelectrolyte hydrogel fibres


from poly(acrylic acid)/poly(vinyl alcohol)
Lei Li and You-Lo Hsieh1
University of California at Davis, CA 95616, USA

E-mail: ylhsieh@ucdavis.edu

Received 27 June 2005, in final form 8 September 2005


Published 14 October 2005
Online at stacks.iop.org/Nano/16/2852
Abstract
Ultra-fine fibrous membranes were prepared by electrospinning of aqueous
mixtures of poly(acrylic acid) and poly(vinyl alcohol) at 17%–83% PAA or
0.14–3.5 COOH/OH molar ratios and cross-linked by heat-induced (140 ◦ C,
5 min) esterification. The fibre diameters increased from 270 to 450 nm with
increasing PAA contents. The swelling ratio of the fibrous hydrogels
increased up to 31 times their dry weight with increasing pH from 2 to 7,
and most significantly between pH 4 and pH 5. The liquid uptake of the
fibrous hydrogels was attributed to both the liquid diffused into the fibres
and that held in the inter-fibre pores. About 53%, 35%, 43% and 37% of the
swelling was contributed by the liquid in the inter-fibre pores when the
fibrous hydrogels were swollen at pH 2, 4, 5 and 7, respectively. The fully
swollen fibrous membranes could be triggered by an applied electric field to
swell further. Such additional swelling was dependent upon the polymer
compositions, strength of electric field and pH, and was reversible.

1. Introduction decreased to less than an eighth (from 10 to 1.2 mm) [14]. For
thinner rods, the degree of bending in an electric field (30 V in
Hydrogels are three-dimensional hydrophilic polymer net- 0.1 M Na2 SO4 ) increased from 30% to 85% with decreasing
works that can swell to many times their mass and volume in diameters from 1.6 to 0.6 mm [13]. As the diameters of
aqueous environments. Polyelectrolyte hydrogels are known these rod-shaped hydrogels are 30 or more times the 20–
to exhibit additional swelling in response to environmental 50 µm diameters of conventional fibres, hydrogel fibres should
stimuli, such as pH, polarity of solvent, ionic strength, or elec- exhibit faster responses. Further reducing fibre sizes to the sub-
tric field [1–3], and have been studied for applications such as micrometre range would significantly improve the response.
sensors and actuators [4–6], drug delivery [7–9] and entrap- The capability of making nano- and micro-fibres from
ment of biomolecules [10]. water-soluble polymers including polyelectrolytes [15–18]
Hydrogels prepared from mixtures of poly(acrylic acid) by electrospinning offers one way to create hydrogels with
(PAA) and poly(vinyl alcohol) (PVA) are among the most nano-and micro-structures. In our laboratory, we have
investigated polyelectrolyte hydrogels. As in any polymeric successfully generated fibrous hydrogels composed of ultra-
solids and hydrogels, the stimulus-triggered responses are fine fibres with diameters in the range of 500 nm–1.2 µm by
diffusion controlled and the kinetics is proportional to the electrospinning of PAA/β-cyclodextrin [3] and PAA/poly(N-
specific surface. In the case of PAA/PVA hydrogel thin isopropylacrylamide) (PNIPAAm) [19] mixtures followed
films [11, 12] and rods [13, 14], the magnitude and speed of by subsequent heat-induced cross-linking. These fibres
responses to the applied stimuli have been found to be inversely were three orders of magnitude finer than the thinnest rods
related to their thickness and diameters [11, 13, 14]. For fully reported before. They also exhibited pH- and temperature-
swollen PAA/PVA hydrogel rods placed in an electric field responsive behaviour. This article reports our attempt to
(10 V cm−1 ), no deflection was observed when the diameters generate fibrous hydrogels composed of even finer fibres
exceeded 10 mm [14]. The rate in which the rods deflected by electrospinning PAA/PVA aqueous mixtures followed by
increased 25-fold (from 0.04 to 1 mm s−1 ) when the diameter cross-linking via heat-induced esterification. The effects of the
polymer compositions and the cross-linking conditions on the
1 Author to whom any correspondence should be addressed.
morphology, packing and stability of the hydrogel fibres were

0957-4484/05/122852+09$30.00 © 2005 IOP Publishing Ltd Printed in the UK 2852


Ultra-fine polyelectrolyte hydrogel fibres from poly(acrylic acid)/poly(vinyl alcohol)

studied. The swelling behaviours of such ultra-fine fibrous the maximum liquid retention capacity (Cm , µl mg−1 ), which
hydrogels in response to the solution pH, the ionic strength, was calculated by the membrane mass before (W0 ) and after
and the electric field were also investigated and analysed based (Wm ) complete saturation in hexadecane as follows:
on their porous structures.
Cm = [(Wm − W0 )/ρ]/ W0 (1)
2. Experimental details
where ρ is the density of hexadecane. The planar pore volume
2.1. Materials (Cv ) was given by the vertical retention capacity (Cv , µl mg−1 ),
measured using a vertical sample with only its lower edge in
Poly(acrylic acid) (PAA) (average Mv ∼ 450 kDa) and contact with the liquid. The sample masses before (W0 ) and
poly(vinyl alcohol) (PVA) (average Mw 124–186 kDa, 87–89% after (Wv ) reaching steady state are used to calculate Cv :
hydrolysed) were obtained from Aldrich Chemical Company.
pH buffers were purchased from EM Science. Sodium chloride Cv = [(Wv − W0 )/ρ]/ W0 (2)
(ACS reagent grade) was received from Fisher Scientific. All
solutions were prepared with purified water (Millipore Milli-Q where ρ is the density of hexadecane. The planar pore
plus water purification system). connectivity was indicated by the Cv /Cm ratio.
The swelling ratios of the fibrous membranes and cast
films were obtained by weighing the samples before (W0 ) and
2.2. Electrospinning and film casting
after (Ws ) immersion in pH buffers (at 30 ◦ C for 24 h) and
PVA solution was prepared by gently stirring with a magnetic calculating as follows:
bar at 80 ◦ C for 30 min, and then cooled down to the room
temperature. PAA solution was prepared under constant q = (Ws − W0 )/ W0 . (3)
stirring at ambient temperature for 8 h. These two solutions
were mixed to achieve a range of compositions at a constant The Wd was obtained after drying the samples in vacuum at
6 wt% total polymer concentration. The molar ratios between room temperature for 24 h. The mass loss following swelling in
PAA carboxylic acid and PVA hydroxyl (assuming 88% water was obtained by weighing the sample (Wd ) after drying
hydrolysis) were 0.14, 0.35, 0.69, 1.4 and 3.5, corresponding in vacuum at room temperature for 24 h as follows:
to 0.17, 0.33, 0.50, 0.67 and 0.83 PAA weight fractions,
Mass loss = (W0 − Wd )/ W0 × 100%. (4)
respectively. About 5 ml of freshly prepared polymer solution
was placed in a glass capillary with the 0.4 mm inner diameter The average of three measurements was reported.
tip, which was tilted downward at an angle between 0◦ and The swelling behaviour of cross-linked fibrous mem-
30◦ depending on the solution viscosity. A stainless steel branes was also determined under the electric stimulus. The
electrode was immersed in the solution and connected to a 30 mm×5 mm×0.35 mm rectangular shaped membranes were
positive power supply (Gamma high voltage supply, ES 30- swollen in buffer solutions at room temperature to the equilib-
0.1P). The electrospinning voltage used in this research was rium state. Afterwards, the buffer solution was poured into a
about 8 kV for all the polymer solutions. Aluminium foil plastic case equipped with two parallel carbon electrodes. The
was used as the collector and connected to the ground. The swollen membrane was placed in the middle of a plastic case
tip-to-collector distance was 25 cm. All electrospinning was filled with the same buffer and centred between two parallel
conducted at room temperature for about 8 h. electrodes (figure 1). The mass before (Ws ) and after (We )
Films were cast from the same mixture on a Teflon surface the applied electric field was used to calculate the additional
by the evaporation of water in a vacuum oven at 60 ◦ C for 24 h. swelling:

2.3. Measurement and characterization Additional swelling (%) = (We − Ws )/ Ws × 100. (5)
The solution viscosities were measured according to The cyclic swelling behaviour was observed by turning the
ASTM D445 using a Cannon-Fenske viscometer (Cannon electric field on and off while replacing the buffer every hour
instrument company, USA). Conductivities of the solutions to maintain buffer quality since water can be electrolysed.
were determined using a digital conductivity meter (VWR
International). The fibre and fibrous membrane morphology
was observed with a scanning electron microscope (SEM)
3. Results and discussion
(XL30-SFEG, FEI/Philips) at 2 kV accelerating voltage
3.1. Fibre formation
without gold coating. Differential scanning calorimetry (DSC)
analyses were conducted (DSC-60 Shimadzu, Japan) using Both PAA and PVA are readily soluble in water. The viscosities
∼5.0 mg samples at a 10 ◦ C min−1 heating rate in nitrogen. of either aqueous PAA or PVA solutions increased with
The glass transition temperature (Tg ) was recorded as the increasing polymer concentrations and their concentration
temperature at the midpoint of the baseline shift. dependence appeared greater above 7 wt% (figure 2). Aqueous
The overall pore volume and the planar pore connectivity PVA solutions were consistently more viscous than the PAA
of each fibrous membrane were measured by the membrane’s solutions at all concentrations.
capacity to retain a low surface tension and low viscosity liquid, Mixing PAA and PVA solutions produced homogeneous
i.e., hexadecane, using a liquid tensiometer (K14, KRÜSS solutions. Nine mixtures of PAA and PVA with 17–83 wt%
USA) [20, 21]. The overall pore volume was estimated by PAA or at 0.14–3.5 COOH/OH ratios were prepared at a fixed

2853
L Li and Y-L Hsieh

COOH:OH molar ratio 2. 5


V 1800
0.69
1600 0.50
0.35 0.97 2
1400

Conductivity (mS/cm)
0.23
0.14

Viscosity (cp)
1200
0.35mm 1. 5
1000
800
Electrode 1
600 1.4
30mm Buffer solution
400 2.1
Electrode 3.5 0. 5
5mm
200
0 0
0 10 20 30 40 50 60 70 80 90 100
PAA (wt%)
36mm
Figure 3. Viscosities and conductivities of 6 wt% PAA/PVA
Figure 1. Measurement set-up for the electric field effect. aqueous mixtures. The values above the data symbols are the molar
ratios between COOH and OH.

1400
Table 1. Glass transition temperatures (Tg ) and average diameters
1200 of PAA/PVA bicomponent fibres.
PVA
1000 PAA Composition Average
Viscosity (cp)

(weight fraction Experimental Calculated diameter


800 of PAA) ( ◦ C) ( ◦ C) (nm)
600 Pure PAA 129.6
0.83 120.0 116.9 450
400 0.67 110.0 106.7 450
0.58 104.0 101.6 450
200 0.50 95.6 97.5 360
0.33 87.6 89.9 360
0
0 1 2 3 4 5 6 7 8 9 10 11 0.17 84.5 83.8 270
Concentration (wt%) Pure PVA 78.1

Figure 2. Viscosities of PAA and PVA aqueous solutions.


0.69, 0.97 and 3.5 COOH/OH ratios were evaluated for
electrospinning. All six could be electrospun continuously
6 wt% total polymer concentration. At 6 wt%, individual for several hours. The highest fibre collection rate of 80 mg
PVA and PAA solutions had viscosities of 94.7 and 74.6 cp, dry fibres/h was observed with the mixture containing 87 wt%
respectively, whereas all their mixtures had significantly of PAA, the most conductive and least viscous mixture. The
higher viscosities (figure 3). This behaviour is thought to fibre collection rates were lowered to about 70 and 50 mg h−1
be attributed to the strong and abundant hydrogen bonding with mixtures containing 50% and 17% PAA, respectively.
between the PAA carboxylic and PVA hydroxyl groups. The Generally, the rates at which fibres were collected were
viscosities of the mixtures peaked at the composition of positively related to the conductivity of the mixtures. As
50 wt% PAA or 0.69 COOH/OH probably due to the possibly reported by Shin et al, conductivity is one of the most
maximized hydrogen bonding between the two polymers. The important solution properties that affects the fibre formation
composition dependence of the viscosity was not symmetrical during electrospinning [23]. The electric field strength that
(figure 3). Due to the self-suppression effect, PAA molecules is necessary to obtain a stable solution jet is lower for more
ionize to a lower extent at higher concentration. For example, conductive polymer solutions. In this work, since a constant
the ionization degree of PAA in its aqueous solution has been electric field strength was maintained for electrospinning of
reported to be 16% and 6% at 3% and 10% PAA concentration, all PAA/PVA mixtures and the flow rate was not held constant,
respectively [22]. As a result, PAA molecules are expected to the higher conductivity of the PAA-rich solutions may result in
ionize to a larger extent in the PVA-rich mixtures and therefore larger drawing force on the solution jet and thus higher solution
to have a more extended conformation. This may facilitate the flow rate and fibre collection rate.
hydrogen bonding interaction between the two polymers and The average sizes of bicomponent fibres ranged from 270
thus cause the PVA-rich mixtures to be more viscous than the to 450 nm (table 1). The fibres generated from mixtures
PAA-rich mixtures. Although PAA molecules may ionize to a containing equal or higher PAA contents were larger and
lesser extent at higher concentrations, the conductivity of the more homogeneous in size, around 450 nm (figures 4(a)–(c)).
PAA/PVA mixtures still increases from 0.05 to 1.97 mS cm−1 Thinner fibres were produced from the more viscous mixtures
with increasing PAA proportion from 17 to 83 wt% (figure 3). in which PVA was the majority (figures 4(d)–(f)). Although
At 6 wt% total polymer concentration, six mixtures with fibres electrospun from the more viscous solutions that were
17%, 33%, 50%, 58% and 83 wt% PAA or 0.14, 0.35, due to higher concentration were thicker [3, 24], fibre size

2854
Ultra-fine polyelectrolyte hydrogel fibres from poly(acrylic acid)/poly(vinyl alcohol)

a b

Figure 5. Thermal cross-linking mechanism between PAA and PVA


via esterification reaction.

c d
close, indicating that PAA and PVA were highly compatible
solids in the bicomponent fibres.
It has been reported that PAA and PVA are compatible in
the films cast from their aqueous mixtures [28]. Unlike the film
casting method, an electric field is applied in electrospinning.
The evidence of high compatibility between the two polymers
in the bicomponent fibres indicates that the strong voltage
e f
employed in electrospinning did not cause any phase separation
between the partially charged PAA molecules and the neutral
PVA molecules. This could be due to the very strong and
abundant hydrogen bonding between PAA and PVA molecules.

3.3. Cross-linking of PAA/PVA fibres

Figure 4. SEM of PAA/PVA bicomponent fibres (bar = 5 µm)


Heat-induced cross-linking was studied using bicomponent
electrospun from 6 wt% total polymer concentration at varying fibres generated from the mixtures at 3.5 and 0.97 COOH/OH
COOH/OH molar ratio: (a) 3.5 (η = 219.8 cp), (b) 1.4 molar ratios. Upon heating between 120 and 140 ◦ C, all
(η = 531.7 cp), (c) 0.97 (η = 1310.7 cp), (d) 0.69 (η = 1598.3 cp), membranes became insoluble in water, giving evidence of
(e) 0.35 (η = 1380.5 cp) and (f) 0.14 (η = 1131.4 cp). thermally induced esterification reaction between PAA and
PVA (figure 5). In order to keep the fibrous membrane flat
during thermal cross-linking, the edges of each membrane
dependence on solution viscosity here is more complex. In were weighed. The cross-linked fibrous membranes shrank
this work, the PVA-rich mixtures were more viscous than the to about half of their initial dimensions and became more
PAA-rich mixtures at the same 6 wt% polymer concentration. rigid. Although the relaxation times of PAA and PVA
The smaller diameter of fibres electrospun from the PVA- in ethanol/water mixtures were reported to be of the order
rich mixtures could be due to the much shorter PVA chains of milliseconds [29], the shrinkage may indicate that the
as reported by others that lower molecular weight polymers entangled PAA and PVA chains had insufficient time to
favour thinner fibres [25]. Furthermore, the flow rate of the relax completely while solidified into very fine fibres.
mixtures was not held constant so the less viscous and more The Tg values of the cross-linked fibres increased with
conductive PAA-rich mixtures may flow faster and thus give increasing heating time (from 5 to 180 min at 120 ◦ C)
larger diameter fibres. The equal PAA/PVA bicomponent (figure 6(a)) and temperatures (from 120 to 140 ◦ C for 5 min)
fibres were about half of the size of fibres generated from 2:1:1 (figure 6(b)), showing increased cross-linking with both
PNIPAAm/PAA/PVA in DMF [19]. increasing temperature and length of heating. The increase
in Tg with increasing heating time was more significant when
3.2. Compatibility between PAA and PVA in their the reactive carboxyl and hydroxyl groups in the bicomponent
bicomponent fibres fibres are nearly equal (figure 6(a)), which indicated that the
cross-linking reaction was more efficient at this composition.
Both PAA and PVA are glassy polymers. The compatibility All the membranes (COOH/OH molar ratio = 0.97)
between two glassy polymers in a mixed solid can be assessed retained the fibrous structure (figures 7(a) and (b)) after
by whether the glass transition temperature (Tg ) is composition heating. However, only those heated at 140 ◦ C for 5 min
dependent or obeys the Fox equation [26]. The DSC of the could retain their fibrous structure after extended (24 h) water
electrospun PAA, PVA and their bicomponent fibres showed immersion (figure 7(d)). The extended water immersion
a single Tg transition (table 1). The Tg of the PAA and PVA caused those membranes heated at 120 ◦ C for 5 min to partially
fibres was 129.6 and 78.1 ◦ C, respectively (table 1). The Tg lose their fibrous integrity and became film-like at random
of the bicomponent fibres was calculated according to the Fox locations of the membrane (figure 7(c)). The mass losses after
equation [27]. the water immersion were about 5% and 10% for the fibrous
membranes heated at 140 or 120 ◦ C for 5 min, respectively.
1/ Tg = W1 / Tg1 + W2 / Tg2 Similar extents of mass loss were observed on the membranes
with a 3.5 COOH/OH molar ratio. These results showed that
where Tg1 and Tg2 are the glass transition temperatures of PAA the best cross-linked fibrous membrane was the one consisting
and PVA; W1 and W2 are the weight fractions of PAA and PVA, of near equal molar PAA/PVA and optimally heated at 140 ◦ C
respectively. The experimental and calculated values were for 5 min.

2855
L Li and Y-L Hsieh

130 Table 2. Vertical (Cv ) and maximum liquid retention capacity


(a)
(Cm ), of PAA/PVA bicomponent fibrous membranes.
COOH:OH Cured
120 molar 140 ◦ C/ Water Cv b Cm b
ratio 5 min exposurea (µl mg−1 ) (µl mg−1 ) Cv /Cm
Tg ( C)
o

0.97 No No 6.94(0.15) 8.98(0.45) 0.77


COOH:OH = 3.5 Yes No 5.46(0.45) 8.04(0.18) 0.68
110 Yes Yes 2.09(0.30) 5.12(0.12) 0.41
COOH:OH = 0.97
3.5 No No 10.18(1.81) 14.44(0.31) 0.70
Yes No 5.77(0.17) 9.06(0.46) 0.64
100 Yes Yes 1.98(0.09) 4.99(0.36) 0.40
0 50 1 00 150 200 a
Exposed in water for 1 day then dried.
Time (min) b
Standard deviation in parenthesis.
112
(b)
lowing the extended water immersion and drying, fibres were
110 enlarged and most of them merged together (figures 7(c) and
(d)), which significantly decreased the inter-fibre pore sizes.
Tg ( C)

The overall pore volume and planar pore connectivity


o

108
in the membranes were estimated by their liquid retention
capacities, i.e., the maximum liquid retention capacity (Cm )
106 and the vertical (Cv ) to Cm ratio or Cv /Cm , respectively. The
COOH:OH = 0.97 overall (Cm ) and planar (Cv ) pore volumes appeared to be
104 composition dependent (table 2). The as-spun membrane from
115 120 125 130 135 140 1 45 the mixture at a 3.5 COOH/OH molar ratio had significantly
o
Temperature ( C) higher overall (61%) and planar (46%) pore volume. The
planar pore connectivity, or Cv /Cm , of these original fibrous
Figure 6. Glass transition temperatures (Tg ) of bicomponent fibres
cross-linked at (a) 120 ◦ C and (b) for 5 min.
membranes was 77% and 70% for the membranes with 3.5 and
0.97 COOH/OH molar compositions, respectively (table 2).
This indicated that most of the inter-fibre pores were connected
a b
to each other in both membranes and this pore connectivity
a b characteristic appears not to be composition dependent.
The exposure to heat during cross-linking reduced the
overall pore volume by 10% and 37% for the membranes
with 0.97 and 3.5 COOH/OH ratio, respectively, resulting
in only slight difference in their pore volume. Exposure
to water further reduced their overall pore volume by 36%
c
and 45%, yielding insignificant difference between the two
d

d compositions. Both the cross-linking and water exposure


reduced the vertical liquid retention (Cv ) of both membranes
to greater extents than their Cm values. Again, the reduction
in Cv was more significant on the membrane containing
predominantly PAA, i.e., 3.5 COOH:OH, i.e., by 43% and
66%, in comparison to the 21% and 62% for the equal molar
composition. The proportionally higher reduction in Cv than
Figure 7. SEM (bar = 5 µm) of PAA/PVA bicomponent fibres
Cm led to the slightly lowered pore connectivity (Cv /Cm )
(COOH/OH molar ratio = 0.97) cross-linked under different by thermal cross-linking. Far more significant decreases in
conditions, exposed to water and dried in vacuum at ambient Cv /Cm were observed after the extended water treatment of
temperature: (a) 120 ◦ C/5 min, (b) 140 ◦ C/5 min, (c) 120 ◦ C/5 min the fibrous membranes. The final pore characteristics (Cv ,
and (d) 140 ◦ C/5 min, water. Cm , Cv /Cm ) resulting from cross-linking and water exposure
were similar between the two compositions, however. This
3.4. Effect of cross-linking and water immersion on overall suggested that heat and water treatments exhibited greater
pore volume and pore connectivity effects on the inter-fibre pore structures of the membranes
composed of more loosely packed fibres.
Compared with the as-spun membrane (figure 4(c)), the This pore volume and structural information derived
heat-induced cross-linking caused the fibres in membranes from the liquid retention data not only confirmed the SEM
(COOH/OH molar ratio = 0.97) to become more closely observations, but also provided insight into the porous
packed (figures 7(a) and (b)). These changes were more sig- structure. From the SEM images, the fibre packing and inter-
nificant at the higher temperature, where some fibres merged fibre pore sizes of the bicomponent fibrous membranes with
together. The thermal cross-linking caused little change on the different compositions appeared similar (figures 4(a) and (c)).
fibres but appeared to reduce the spaces between fibres. Fol- The overall pore volume measured by liquid retention indicated

2856
Ultra-fine polyelectrolyte hydrogel fibres from poly(acrylic acid)/poly(vinyl alcohol)

35 Fibrous COOH:OH=3.5
Table 3. Swelling ratios of fibrous membranes and cast films
Fibrous COOH:OH=0.97 (COOH:OH = 3.5) before (qbc ) and after (qac ) centrifugation.
30 Fibrous COOH:OH=0.14 Fibrous membrane Cast film
Cast film COOH:OH=3.5

25 pH qbc qac qbc qac


Swelling ratio

2 6.6 1.5 3.1 3.0


20 4 11.9 8.8 7.7 7.5
5 27.9 26.2 15.8 15.5
15 7 31.0 30.0 19.4 18.9

10
The spaces between the fibre layers and in the inter-fibre
5 pores within each layer have been found to be highly
dependent upon the electrospinning process and the cross-
0 linking conditions [31]. Starting with the same original
1 2 3 4 5 6 7 8 dimensions, a fibrous membrane swelled like a flattened
pH
sponge to become several times thicker than the cast film, while
Figure 8. PH-responsive swelling behaviour of hydrogel fibres and expended similarly in lengths and widths. It is this ability
cast film with varying COOH/OH molar ratio. to expand significantly in thickness that makes these fibrous
hydrogels exhibit much higher swelling capacities than the cast
film.
that the fibres were packed differently in the membranes with To estimate the contributions of the liquid held in the
different compositions. The differences appeared to be in the inter-fibre pores and that diffused into the fibres to the overall
thickness direction, which was not observable by the SEM swelling ratios, fully swollen membranes (COOH:OH = 3.5)
due to the limited depth of field. The much larger decrease were centrifuged at 3000 rpm for 10 min to remove the liquid
in the overall pore volume for the membranes with the higher in the inter-fibre pores. The swelling ratios of the fibrous
PAA content from thermal cross-linking could be caused by hydrogels before and after the centrifugation were compared
the more significant thickness reduction. with those of cast films (table 3). As expected, the hydrogels
from the cast films lost a negligible amount of liquid (∼2%)
3.5. pH- and ionic-strength-responsive swelling behaviours from centrifugation. The fibrous membranes fully swollen at
pH 2, 4, 5 and 7, on the other hand, lost 77%, 26%, 6% and 3%
The cross-linked fibrous membranes exhibited pH-responsive of liquids from centrifugation, respectively. This observation
swelling behaviour (figure 8). At pH 2 and 4, the membranes indicated that the inter-fibre pores of swollen membranes were
took up to 6 and 11 times of liquid, respectively. Significantly much smaller at pH 5 and above. This is consistent with higher
more liquid was absorbed at pH 5, up to 29 times, followed swelling of the fibres with increasing pH, reducing the inter-
by a slight increase to 31 times at pH 7. Higher absorption fibre pore sizes of the swollen membranes. Assuming the
was observed with fibrous membranes that contained more amount of liquid diffused into the fibres is the same as that
PAA at all pH values. The compositional effect was also pH in the cast film of the same composition, the proportion of
dependent, i.e., negligible below pH 4 and more distinct at liquid held in the inter-fibre pores could be calculated from the
pH above 4. At a 3.5 COOH/OH molar composition, the difference in their swelling ratios. Based on this assumption,
fibrous membrane absorbed almost twice as much liquid as the proportion of the liquid in the inter-fibre pores was 53%,
the cast film at all pH values between 2 and 7. In fact, the 35%, 44% and 38% for the fibrous membranes fully swollen at
fibrous membrane with less PAA (0.97 COOH/OH) could still pH 2, 4, 5 and 7, respectively. This indicates that the proportion
swell 50% more than the cast films with higher PAA proportion of liquid held in the inter-fibre pores was slightly higher than
(COOH/OH molar ratio = 3.5). that diffused into the fibres at pH 2. With greater swelling of
In a swollen hydrogel, the repulsion among the negatively the fibres at the higher pHs, a minority of liquid absorbed by
charged carboxylate ions on PAA molecules increases the the fibrous membranes was attributed to the liquid diffused into
free volume, thus expanding its capability to take in water. the spaces in between the swollen fibres. The swelling ratios
At the same time, the hydroxide ions in the buffer solutions of the fibrous membranes and cast films increased with pH in
could diffuse into hydrogel, react with the protons dissociated similar patterns (figure 8), indicating that the contribution of
from the carboxyl groups of PAA to form water and swell the fibre swelling to the liquid uptake in the fibrous membranes was
hydrogel [30]. The dissociation of carboxyl groups of PAA is similarly pH dependent to the hydrogel film. However, their
much more significant when pH is higher than its pK a , 4.7, and differences showed that liquid retained in the pores increased
therefore the hydrogel composition showed a greater effect at significantly between pH 4 and 5, but was essentially the same
pH above 4. below and above this range. This major increase of liquid taken
The liquid uptake in fibrous membranes consists of the up by the pores between swollen fibres suggested the enlarged
liquid held in the inter-fibre pores as well as that diffused pore structure in the thickness direction might have occurred
into the fibres. The contribution between the two to the between pH 4 and 5. It should be noted that the centrifugation
total liquid uptake, or the swelling ratio measured, depends force employed may not be strong enough to overcome the
on how the fibres pack and swell. In electrospinning, the capillary force to remove the liquid inside the very small inter-
fibres were deposited layer by layer into a fibrous membrane. fibre pores.

2857
L Li and Y-L Hsieh

10 40
(a)
9
8 COOH:OH = 3.5
7 30 with electric field
Swelling ratio

6 without electric field

Swelling ratio
5
20
4
3 Fibrous membrane
2 Cast film
10
1
0
-3 -2 -1 0
log µ (M) 0
2 4 5 7
Figure 9. Ionic strength responsive swelling behaviour of hydrogel pH
fibres and cast film with the same composition (COOH/OH molar
ratio = 3.5). 100
(b) pH=4, COOH:OH=3.5
pH=5
pH=7
pH=4, COOH:OH=0.97
80
The swelling behaviour in response to ionic strength

Additional Swelling (%)


adjusted by the concentration of sodium chloride (NaCl) was
also evaluated. The swelling ratios did not change at increasing 60
ionic strengths of 0.001–0.01, and 0.1 M. However 44.2%
and 30.9% decreases in swelling were observed at the highest
ionic strength studied, i.e., 1.0 M, for hydrogel fibres and cast 40
films, respectively (figure 9). At this concentration, it appeared
that sufficient Na cations and Cl anions could diffuse into
20
the hydrogel to shield the electrostatic repulsion between the
partially ionized and negatively charged PAA molecules. This
reduced the free volumes inside the hydrogel, thus squeezed 0
out water and made the hydrogel de-swell. Because of the 0 10 20 30 40 50 60
higher specific surface of hydrogel fibres, the ions in the Time (min)
surrounding solution could diffuse into the hydrogel fibres Figure 10. Additional swelling of hydrogel fibres: (a) with and
more easily, thus exhibiting greater effect. without the electric field (10.0 V cm−1 ); (b) kinetics in buffers with
different pH.
3.6. Electric field responsive swelling behaviours
The swelling behaviour of the fully swollen fibrous membranes
pressure difference occurs in the hydrogel, and the hydrogel
under the influence of an electric field was investigated
is bent toward the cathode [14]. The deflection of a fibrous
using the membrane with the highest amount of PAA (3.5
hydrogel (figure 1) was expressed in terms of the distance
COOH:OH). A 10 V cm−1 electric field was shown to trigger
between the ends of the hydrogel before and after bending.
additional swelling at pH 4–7, but not at pH 2 (figure 10(a)).
The magnitude and velocity of the deflection of the hydrogel
Additional swelling was highest, exceeding 80%, at pH 4, and
was much lower (<20%) at pH 5 and 7. It took 5 min to reach with a 3.5 COOH/OH molar ratio (10.0 V cm−1 electric field
a new equilibrium at the low extents of additional swelling (pH and pH 4) was about 12 mm and 0.8 mm s−1 , respectively.
5 and 7) and much longer (30 min) to reach the proportionally With lowering of the COOH/OH molar ratio to 0.97, the
higher additional swelling at pH 4 (figure 10(b)). The deflection magnitude and velocity decreased to about 5 mm
additional swelling (pH 4, and 10 V cm−1 ) was found to be and 0.2 mm s−1 , respectively. The ultra-fine fibrous hydrogels
much lower (10%) for the membrane with nearly equal COOH prepared by electrospinning did not exhibit obvious differences
and OH molar compositions. in the deflection magnitude and velocity when compared with
Under the electric field, the increased swelling ratios of the the hydrogel rod [14]. It may be due to the fact that the ultra-
fibrous membranes at pH between 4 and 7 were accompanied fine hydrogel fibres are entangled together inside the fibrous
with increased width and thickness of the membrane. For hydrogel, which compromises the deflection of the individual
example, about 50% increases were visually observed at pH 4. hydrogel fibres. However, it should also be noted that the
The two freely suspended ends of the fibrous membrane also fibrous hydrogel and the hydrogel rod have different length,
bent toward the cathode of the electric field. When subjected thickness, cross-section shape, swelling media and polymer
to a dc electric field, the cations in the PAA hydrogel such as composition.
the protons dissociated from the carboxyl groups are driven The external electric field may also facilitate the
toward the cathode, while the polyanionic PAA backbones dissociation of the carboxylic acid groups of PAA, further
remain immobile due to the cross-linked polymeric network. expanding the polymer network. The extent of additional
As a result, the osmotic pressure of the anode side increases swelling appeared to be inversely related to the extent of
and becomes larger than that of the cathode side. Therefore the dissociated carboxylic acid in the swollen hydrogel fibres.

2858
Ultra-fine polyelectrolyte hydrogel fibres from poly(acrylic acid)/poly(vinyl alcohol)

100 (a) 90
80

Additional swelling (wt%)


Additional swelling (%)

80 70
60
60 50
40
40 30
10V/cm
4.9V/cm 20
20 3.3V/cm
10
2.0V/cm
0
0 0 2 4 6 8 10 12 14 16 18 20
0 20 40 60 Time (hr)
Time (min)

20
Figure 12. Cyclic swelling (1 h) and de-swelling (1 h) behaviour of
(b) hydrogel fibres (COOH/OH molar ratio = 3.5) in pH 4 buffer
solution and 10.0 V cm−1 electric field.
Additional swelling (%)

15
pH 4, additional swelling between 72% and 83% were reached
(figure 12). The swelling nearly reversed when the electric field
10 was turned off each time. A permanent 8% additional swelling
remained after the first cycle, suggesting certain irreversible
swelling in the hydrogel fibres caused by initial exposure to the
5 10V/cm
4.9V/cm electric field. This swelling and de-swelling cycle could be re-
3.3V/cm peated for at least 20 times while retaining the membrane form.
2.0V/cm
0
0 10 20 30 4. Conclusions
Time (min)
Ultra-fine fibrous membranes were generated by electrospin-
Figure 11. Effect of electric field strength on the additional swelling ning of PAA and PVA aqueous mixtures containing 17%–83%
of hydrogel fibres (COOH/OH molar ratio = 3.5) in (a) pH 4 and (b)
pH 7 buffer, respectively. PAA or COOH/OH molar ratios from 0.14 to 3.5. Generally,
the fibre formation rates were positively related to the PAA
contents. The average fibre diameters also increased from 270
The highly dissociated structures at pH 5 and 7 were less to 450 nm with increasing PAA contents or COOH/OH ra-
affected by the applied electric field than the less dissociated tios. Cross-linking via esterification was best achieved on the
hydrogel fibres at pH 4. The absence of electric field effects fibrous membranes containing near equal COOH/OH molar
at pH 2 suggests a lack of dissociation of the carboxyl groups. composition (58% PAA) and optimally heated at 140 ◦ C for
The additional swelling generally increased with increasing 5 min.
electric field strengths at both pH 4 (figure 11(a)) and pH 7 These hydrogel fibrous membranes took up different
(figure 11(b)). Between 3.3 and 10.0 V cm−1 , the additional extents of aqueous liquids in response to the pH, ionic strength
swelling at pH 4 was consistently higher than that at pH 7. and electric field. The hydrogel fibrous membranes took up
However, under a weaker electric field of 2.0 V cm−1 , no liquid up to 31 times their dry weight and their ability to absorb
further swelling was detected at pH 4 whereas an 8% swelling liquid increased with increasing PAA contents and pH, most
increase was shown at pH 7. At pH 4, it took about 20 significantly between pH 4 and 5. The swelling ratio, or liquid
min for the membranes to reach equilibrium (figure 11(a)). uptake, in the fibrous hydrogel was attributed to the liquid held
Within the first three minutes from exposing to the electric in the inter-fibre pores as well as that diffused into the fibres.
field, the membranes absorbed at 0.6%, 1.4% and 2.4% per The contribution of the two to the total liquid uptake depends
min at 3.3, 4.9 and 10.0 V cm−1 , respectively. Beyond 3 min, on how the fibres layout and swell. About 53%, 35%, 44%
the swelling rates became similar, i.e., 3.5%, 4.0% and 3.3% and 38% of the fibrous hydrogel swelling was contributed by
per min at 3.3, 4.9 and 10.0 V cm−1 , respectively. At pH the liquid held in the inter-fibre pores when swollen at pH
7, equilibrium swelling was reached in about 5 min under 2, 4, 5 and 7, respectively. The fibrous hydrogel reduced
electric fields between 3.3 and 10.0 V cm−1 and at a similar swelling by 44.2% when the ionic strength increased from 0.1
swelling rate of about 3.5% per min (figure 11(b)). However, to 1.0 M (pH 4), but showed no change in swelling at low
at 2.0 V cm−1 , reaching the equilibrium took twice as long and ionic strengths between 0.001 and 0.1 M. This could be the
had a much lower speed of 0.9% per min. At pH 4, the lower critical NaCl concentration where the diffusion of Na cations
initial swelling rate, in particular under weaker electric field, and Cl anions into the hydrogel fibres was maximized to shield
was again attributed to less dissociated carboxylic acid. the electrostatic repulsion among PAA polyanions and reduces
The electric field triggered additional swelling of the free volumes inside the hydrogel.
PAA/PVA hydrogel fibrous membranes was largely reversible. The fully swollen fibrous hydrogel could be triggered
In ten repetitive exposures to the 10.0 V cm−1 electric field at by applied electric field (10 V cm−1 ) to swell further.

2859
L Li and Y-L Hsieh

Under the electric field, the fibrous hydrogel membrane [11] Gerlach G, Guenther M, Suchaneck G, Sorber J, Arndt K-F and
with a 3.5 COOH/OH molar composition showed the largest Richter A 2004 Macromol. Symp. 210403–10
additional swelling of 80% at pH4. The additional swelling [12] Matsuyama H, Teramoto M, Matsui K and Kitamura Y 2001
increased with increasing PAA contents in the membranes and J. Appl. Polym. Sci. 81 936–42
[13] Fei J, Zhang Z and Gu L 2002 Polym. Int. 51 502–9
the strengths of the electric fields, and was reversible with the [14] Shiga T, Hirose Y, Okada A and Kurauchi T 1992 J. Appl.
application and removal of the electric field. Polym. Sci. 44 249–53
Because of their stimulus-responsive swelling behaviour [15] Jin H-J, Fridrikh S V, Rutledge G C and Kaplan D L 2002
and the ultra-high specific surface, these hydrogel fibrous Biomacromolecules 3 1233–9
[16] Ohkawa K, Cha D, Kim H, Nishida A and Yamamoto H 2004
membranes are being further investigated for their potential Macromol. Rapid Commun. 25 1600–5
for controlled release and immobilization of chemical and [17] Yao L, Haas T W, Guiseppi-Elie A, Bowlin G L,
biological significant agents. Simpson D G and Wnek G E 2003 Chem. Mater. 15 1860–4
[18] Reneker D H and Chun I 1996 Nanotechnology 7 216–23
[19] Chen H and Hsieh Y-L 2004 J. Polym. Sci. A 42 6331–9
References [20] Hsieh Y L 1995 Textile Res. J. 65 299–307
[21] Hsieh Y L, Miller A and Thompson J 1996 Textile Res. J. 66
1–10
[1] Seki T and Okahata Y 1984 Macromolecules 17 1880–2 [22] Inzelt G and Grof P 1977 Acta Chim. Acad. Sci. Hung. 93
[2] Cleary J, Bromberg L E and Magner E 2003 Langmuir 19 117–22
9162–72 [23] Shin Y M, Hohman M M, Brenner M P and
[3] Li L and Hsieh Y-L 2005 Polymer 46 5133–9 Rutledge G C 2001 Polymer 42 09955–67
[4] Kim S J, Yoon S G, Lee Y M, Kim H C and Kim S I 2004 [24] Deitzel J M, Kleinmeyer J, Harris D and Beck Tan N C 2000
Biosensors Bioelectron. 19 531–6 Polymer 42 261–72
[5] Fernandez E, Lopez D, Lopez-Cabarcos E and [25] Koski A, Yim K and Shivkumar S 2003 Mater. Lett. 58 493–7
Mijangos C 2005 Polymer 46 2211–7 [26] M’Bareck C O, Nguyen Q T, Metayer M, Saiter J M and
[6] Elliott J E, Macdonald M, Nie J and Bowman C N 2004 Garda M R 2004 Polymer 45 4181–7
Polymer 45 1503–10 [27] Fox T G 1956 Bull. Am. Phys. Soc. 1 123
[7] Mandal T K, Bostanian L A, Graves R A and [28] Vazqueztorres H, Cauichrodriguez J V and
Chapman S R 2002 Pharmaceutical Res. 19 1713–9 Cruzramos C A 1993 J. Appl. Polym. Sci. 50 777–92
[8] Panayiotou M and Freitag R 2005 Polymer 46 615–21 [29] Theron S A, Zussman E and Yarin A L 2004 Polymer 45
[9] Zhang X-Z, Lewis P J and Chu C-C 2005 Biomaterials 26 2017–30
3299–309 [30] De S K, Aluru N R, Johnson B, Crone W C, Beebe D J and
[10] Di Benedetto F, Biasco A, Pisignano D and Cingolani R 2005 Moore J 2002 J. Microelectromech. Syst. 11 544–55
Nanotechnology 16 S165–70 [31] Jin X and Hsieh Y-L 2005 Polymer 46 5149–60

2860

You might also like