Carbon dioxide impregnation of electrospun polycaprolactone fibers

Page 1

A

oati1fs©

K

1

oadgetptcsutetmrc

0d

J. of Supercritical Fluids 41 (2007) 173–178

Carbon dioxide impregnation of electrospun polycaprolactone fibers

Olukemi Ayodeji a, Emily Graham a, Douglas Kniss b, John Lannutti a, David Tomasko c,∗a Department of Materials Science and Engineering, Ohio State University, Columbus, OH 43210, USA

b College of Medicine, Laboratory of Perinatal Research, Ohio State University, Columbus, OH 43210, USAc Department of Chemical and Biomolecular Engineering, Ohio State University, Columbus, OH 43210, USA

Received 1 August 2005; received in revised form 8 September 2006; accepted 10 September 2006

bstract

The electrospinning of polymers has become a potentially important process for the production of tissue engineering scaffolds. CO2 impregnationf these scaffolds may provide a method for tailoring the chemistry of these relatively high surface area scaffolds without altering their biomimeticrchitecture. In pursuing this we found that electrospun polycaprolactone (PCL) fibers melt when exposed to supercritical CO2 even at roomemperature. However, CO2 exposures ranging from 10 to 25 ◦C and 1.0 to 3.44 MPa provided chemical impregnation without apparent changesn physical structure. A test compound, carboxytetramethylrhodamine, was embedded into electrospun PCL using CO at 3.44 MPa and 25 ◦C for

2

0 h. The subsequent release of carboxytetramethylrhodamine into phosphate buffered saline at 37 ◦C was then monitored. Release was observedor 30 days after which the fibers were shown to retain 8.54 �g of carboxytetramethylrhodamine/mg of PCL. Control samples not exposed to CO2

howed no detectable release after 5 days.2006 Elsevier B.V. All rights reserved.

ti

([[pptcrUftvot

eywords: Electrospinning; Polycaprolactone; Carbon dioxide; Polymer

. Introduction

Electrospinning utilizes an electrical field to induce ejectionf a charged jet from the surface of a solution; in biomedicalpplications this is typically a polymer (synthetic or naturallyerived) solution. This jet elongates under the influence of tan-ential stresses and bending instabilities and as it falls the solventvaporates. This can result in fibers having submicron diametershat are then collected on an electrical ground [1,2]. Electrospunoly(s-caprolactone) (PCL) has recently been extensively inves-igated for potential use in tissue engineering and other biomedi-al applications [3–17]. These novel biodegradable nanofibrouscaffolds closely mimic the normal extracellular matrix. Thenique architecture produced by electrospinning can provideemporary structural support and guide tissue regeneration. Sev-ral investigations have involved the use of electrospun scaffoldso develop bone grafts [18,19] in a well-vascularized site using

esenchymal stem cells [16,13]. Other efforts targeted cartilage

eplacement and cardiac grafts [20–22]. Human smooth muscleells [23] and endothelial cells can be seeded on these scaffolds

∗ Corresponding author. Tel.: +1 614 292 4249; fax: +1 614 292 3769.E-mail address: [email protected] (D. Tomasko).

(dts

bm

896-8446/$ – see front matter © 2006 Elsevier B.V. All rights reserved.oi:10.1016/j.supflu.2006.09.011

o form a three-dimensional cellular network suggesting its usen blood vessel substitutes [24,25].

In this context it is well established that supercritical fluidsSCF) can be used to modify a broad range of polymer systems26–33]. Advantages include the absence of organic solvents34,35] and the ability to incorporate delicate biological com-ounds without loss of activity combined with the generation ofossibly desirable porous internal architectures [36]. In contrast,he solution processing of many biodegradable polymers typi-ally involves organic solvents that could leave behind harmfulesidues. Residual CO2 is inert, nontoxic and nonflammable.nder appropriate conditions CO2 is a potent swelling agent

or polymers and facilitates the absorption of desirable addi-ives. The swollen state is characterized by both increased freeolume and greater chain motion enabling more rapid diffusionf large molecules into the polymer. Recent research has shownhat biodegradable polymer matrix, poly-dl-lactide-co-glycolidePLGA) can be impregnated with 5-fluorouracil and �-estradiol,rugs that are used for chemotherapy and estrogen hormoneherapy [37]. The use of CO2 as a means of impregnating other

olutes into various polymers [38–43] has been investigated.

In this work, we extend the use of subcritical levels of car-on dioxide (CO2) previously demonstrated [44] to study theorphological effects on a common tissue engineering scaffold,

Page 2

1 critica

eflinigis

2

2

M(asaal5

2

woCtptc

2

lmOwtgnwtt

2

lamR9

oS

2

cwiwuat

llaw(swSR1f

tesoU

2

ftwba

3

3

IfintFCp

74 O. Ayodeji et al. / J. of Super

lectrospun polycaprolactone. Carboxytetramethlyrhodamine, auorescent molecule, is used to test the concept of impregnation

nto electrospun fibers utilizing subcritical CO2. This becameecessary as our objective was to physically entrap a moleculento a relatively low Tg polymer surface without deforming theeometric dimensions of the surface itself. Since subcritical CO2s of lower density than supercritical CO2, foaming during pres-ure release [36] is not an issue.

. Materials and methods

.1. Electrospinning

A 12 wt.% poly(s-caprolactone) (PCL, mp 60 ◦C; d 1.145;w = 65,000, Aldrich®, Milwaukee, WI) acetone solution

Mallinckrodt® Chemicals, Phillipsburg, NJ) was electrospunt a flow rate of 24 ml/h and an applied voltage of 24 kV. Atainless steel deposition plate was covered with aluminum foilnd positioned 20 cm from the tip of the needle. Immediatelyfter spinning, the electrospun sheet was removed from the col-ection plate, cut into 2.54 cm diameter disks of approximately0 mg and stored in a desiccator until further use.

.2. Morphology versus CO2 exposure

To determine what impregnation conditions could be usedithout destroying the porosity and nanoscalar fibrous naturef these tissue engineering scaffolds, samples were exposed toO2 specifically to examine the effects of carbon dioxide on

heir morphology. CO2 (99.9% purity, Praxair, Columbus, OH)ressures ranged from 1.0 to 4.0 MPa and temperatures from 10o 40 ◦C. Total contact times were 2.5 h in the majority of theonditions investigated.

.3. Impregnation using subcritical CO2

Experiments were performed using a view cell set up simi-ar to that used in previous work [40]. 5-(and-6)-Carboxytetra-ethlyrhodamine (Mw = 466.92, Molecular Probes®, Eugene,R) was dissolved (0.1 mg/ml) in water. A small glass beakeras filled with 3 ml of carboxytetramethlyrhodamine solution;

he electrospun PCL disks were then placed into solution. A sightauge (Jerguson Model RW-20) was used to monitor the impreg-ation process. The beaker was placed in the view cell, whichas then filled with CO2 at the desired conditions. Total contact

ime was 2.5 h and the CO2 was slowly released by reversinghe flow of the syringe pump at a rate of 1.0 ml/min.

.4. Scanning electron microscopy (SEM)

Scanning electron microscopy (SEM, Quanta 300, Nether-ands) was used to examine the surface characteristics of treated

nd non-treated samples. All samples were mounted on an alu-inum stubs covered with carbon adhesive tape (Ted, Pella,eading, CA) and sputter coated (Pelco Model 3 sputter coater1000) with Au-Pd at an emission current of ∼15 mA for a total

sCpb

l Fluids 41 (2007) 173–178

f 70 s under Ar. The coated samples were then analyzed in theEM at an accelerating voltage of 10 kV.

.5. Release studies

Utilizing conditions previously determined to cause nohange in morphology, longer-term impregnation experimentsere carried out to explore both the loading and release behav-

or of these relatively high surface area materials. Five samplesere placed in a stainless steel pressure vessel (Pressure Prod-cts Industries, Inc., Warminster, PA) and carbon dioxide wasdded to the vessel at 3.44 MPa and 25 ◦C. Contact was main-ained for 10 h before the pressure was released.

A UV–vis spectrophotometer (CARY IE, VARIAN®, Sugar-and, TX) was used to evaluate the release of carboxytetrameth-yrhodamine from the treated and non-treated electrospun fibert 550 nm. Carboxytetramethlyrhodamine content, either with orithout CO2 exposure, was based on a linear calibration curve

linear regression coefficient = 0.9993) constructed utilizing aeries of carboxytetramethlyrhodamine standards. Each sampleas immersed in 5 ml of phosphate buffer solution (PBS, 1.0 M,igma–Aldrich, St. Louis, MO) in a water bath shaker (Model76, New Brunswick Scientific, Edison, NJ) and agitated at20 rpm and 37 ◦C for 1–30 days. The PBS was exchanged forresh solution following each sampling interval.

The UV–vis spectrophotometer was also used to quantifyhe total amount of carboxytetramethlyrhodamine present in anntire disk of electrospun fiber. Samples were weighed and dis-olved in acetone. The carboxytetramethlyrhodamine contentf the resulting solution was then analyzed as before using theV–vis spectrophotometer.

.6. Fluorescence microscopy

An epi-fluorescent microscope was used to provide evidenceor the embedding process and characterize the uniformity ofhe carboxytetramethlyrhodamine impregnation both with andithout 10 min of rinsing with PBS. The samples were placedetween glass coverslips and images taken both in phase contrastnd red fluorescence (570 nm) mode.

. Results

.1. CO2 effects on morphology

Fig. 1(a)–(e) shows the effect of CO2 on the electrospun fiber.n Fig. 1(a), a well-defined architecture is visible; the individualbers remain intact and show the distinctive non-woven fibrousetwork. The microstructure of the fibers begins to change whenhey are exposed to specific levels of CO2. Clearly, the fibers inig. 1(c) show signs of melting after exposure to 40 ◦C and aO2 pressure of 3.10 MPa. At the same pressure but a lower tem-erature, 10 ◦C (Fig. 1(b)), the fibers closely resemble the initial

tate and retain a well-defined structure. The apparent effects ofO2 on morphological change under different temperatures andressures have been classified under three categories (Fig. 2)ased on the CO2 temperature and density. The “non-melting

Page 3

O. Ayodeji et al. / J. of Supercritical Fluids 41 (2007) 173–178 175

F expo2

r0tmma1Thtdes

dte

sw

3

tbttsbw

ig. 1. SEM micrographs of electrospun PCL (a) as-fabricated. Following CO2

5 ◦C and 3.44 MPa for 2.5 h and (e) 25 ◦C and 3.44 MPa for 10 h.

egion” is generally below a line stretching from 10 ◦C and.06 g/cm3 through the point at 25 ◦C and 0.035 g/cm3. Underhese conditions electrospun PCL fibers exhibited no detectable

orphological change. “Structural relaxation,” reflecting onlyinimal effects on morphology, was observed at conditions

bove the “non-melting” line and below a line stretching from0 ◦C and 0.09 g/cm3 through the point at 40 ◦C and 0.04 g/cm3.he last region is designated “obvious melting” at conditionsigher than the “structural relaxation” line. The PCL fibers inhe third region were clearly structurally affected and the initialiscrete, well-separated morphology of the fibers was largelyliminated. These results are discussed further in the discussionection.

Based on these studies, exposures at 25 ◦C and 3.44 MPa wereetermined to have no significant effects on morphology andhis condition was chosen for all of the subsequent embeddingxperiments. The contact time was lengthened to 10 h; Fig. 1(e)

vtda

sure at: (b) 10 ◦C and 3.10 MPa for 2.5 h, (c) 40 ◦C and 3.10 MPa for 2.5 h, (d)

hows that, as expected, no apparent changes in morphologyere detected even after this extended exposure.

.2. Carboxytetramethlyrhodamine release

CO2 is known to be effective in impregnating specific func-ionalities into polymer surfaces [40]. Even under relativelyenign conditions, we observed significant, uniform impregna-ion of the electrospun fibers. Fig. 3 shows fluorescence imagesaken before and after rinsing treated PCL with phosphate bufferolution. Rinsing was conducted in an attempt to remove car-oxytetramethlyrhodamine solution that was not truly absorbedithin the fibers. Rinsing is also a component of any in vitro or in

ivo application of this technology so establishing its effects onhese scaffolds following impregnation is critical. After rinsing,etectable levels of carboxytetramethlyrhodamine are presentcross the entire surface of the CO2-treated sample (Fig. 3). The

Page 4

176 O. Ayodeji et al. / J. of Supercritical Fluids 41 (2007) 173–178

Fig. 2. Morphological effects of temperature and CO2 density on electrospunPbc

s(ipttpao

elampeo5lt

FPc

rCb

4

jTp[l

Fd

CL. Generalized regions are defined broadly to suggest expected physicalehavior under these conditions. Diamonds: no effect; squares: limited structuralhanges; triangles: obvious melting.

ample exposed only to carboxytetramethlyrhodamine solutionin the absence of CO2) for 10 h at 25 ◦C showed relatively littlempregnation either before or after rinsing. Additionally, thesearticular fibers were white in color following rinsing and werehus identical in appearance to the starting material. This con-rasted with the CO2-exposed samples that retained a distinctiveinkish color following carboxytetramethlyrhodamine exposurend rinsing; this pink color remained apparent even after a monthf release.

The release profiles (Fig. 4) provide definitive proof of theffect of subcritical CO2 on embedding. Carboxytetrameth-yrhodamine release from the CO2-treated-PCL continued overperiod of 30 days. Even after 30 days of release, fluorescenceicroscopy still showed signs of carboxytetramethlyrhodamine

resence in specific areas of the CO2-treated samples. Samplesxposed to carboxytetramethlyrhodamine solution alone (with-

ut CO2) released only for a very short period (Fig. 4). Afterdays, there was no significant release from these fibers, the

imits of detection being 6.5 �g/ml. The amount of carboxyte-ramethlyrhodamine present in the CO2-treated samples after

tfee

ig. 3. Fluorescence images of CO2 + carboxytetramethlyrhodamine-exposed electristribution of additive changes drastically after rinsing but fluorescence is still detec

ig. 4. Release of carboxytetramethlyrhodamine into PBS from electrospunCL fibers with and without 3.44 MPa of CO2 during 10 h exposure to thearboxytetramethlyrhodamine solution. Inset shows cumulative release vs. time.

insing in PBS was 84.2 �g/mg. The amount remaining in theO2-treated samples following 30 days of release was found toe 8.54 �g/mg.

. Discussion

Electrospinning involves the rapid stretching of an electrifiedet occurring simultaneous with the rapid solvent evaporation.he polymer chains experience strong shear forces during thisrocess and these affect their conformation and crystallinity45]. If conformationally sensitive biomolecules are present thisogically leads to concerns regarding their stability/activity in

he resulting polymer fiber [46,47]. The use of CO2 to add bio-unctionality to electrospun polymer fibers post-formation canliminate these concerns since the biomolecules will not experi-nce these shear forces. In addition, by eliminating contact with

ospun PCL before (A) and after rinsing with PBS (B). An initially uniformtable. The scale bar = 200 �m.

Page 5

critica

tmctt

peefmpvoitadd(rprc

wecrttUc

tpapctltcmtmatsbtpca

ui

prcmdptCa

dtTcaiga

5

p2atdaireCct

A

N0mDPt

R

O. Ayodeji et al. / J. of Super

he sometimes harsh organic solvents used in electrospinning weay also be able to avoid denaturing these biomolecules. The

onditions described in the experimental section are unlikelyo affect a wide range of biological compounds since these areypically stable up to at least 37 ◦C.

It is well established that supercritical CO2 plasticizes manyolymers. However, it is clear that even at low pressures CO2nhances the free volume dispersed among the polymer chainsnabling greater mobility of both the chains and any interdif-using species [39]. With respect to microstructure, polymerorphology must depend on processing parameters such as tem-

erature, pressure, contact time, pressure release rates, and sol-ent density. The solubility, diffusivity and partitioning [48,49]f solutes between the fluid phase and the polymer play anmportant role in impregnation. This can help in determininghe effect of intermolecular interactions of embedded moleculesnd the polymer matrix. Kazarian and Eckert [30] studied theiffusion and solubility of Disperse Orange 25 (DO25) and aye mixture of Disperse Red 1 in the polymethyl methacrylatePMMA)–CO2 system. The effects of CO2 pressures, diffusionate and the partitioning of the dye solute between the fluidhase and PMMA were investigated. Increases in CO2 pressureesulted in higher diffusivity and smaller solute partition coeffi-ients in the PMMA.

In this work, relatively benign subcritical CO2 conditionsere used to examine the resulting morphological changes in

lectrospun PCL. CO2 is considerably less dense under theseonditions and can quickly diffuse out of the polymer uponelease of pressure. Any larger, less mobile solute is typicallyrapped within the polymer network. The polymer relaxes backo its original state without trapping significant amounts of CO2.nder these conditions, foaming, which leads to undesirable

hanges in a number of physical properties, does not occur.Although we sidestep problems with foaming, microstruc-

ural changes do occur as CO2 pressure/density and system tem-erature increase. Within the ‘non-melting’ region of Fig. 2 sub-mbient temperatures and modest (up to approximately 300 psia)ressures (densities up to 0.05 g/cm3) do not result in detectablehanges in nanofiber morphology. Beyond this range “struc-ural relaxation” takes place as swelling of the PCL (a relativelyow-melting (Tm = 60 ◦C) polymer) triggers release of stresseshat likely remain from the relatively rapid electrospinning pro-ess. At higher pressures or densities (Fig. 2) ‘melting’ of theicrostructural components initiates loss of the fiber structure as

he CO2-swollen solid becomes mobile enough to begin activelyinimizing its surface area. The divisions between these regions

re diffuse since they are based on SEM observation of thereated fiber. Therefore, the lines shown in Fig. 2 should onlyerve as a rough guide to the conditions where each type ofehavior is observed. As expected, with increases in tempera-ure and pressure CO2 becomes a better swelling agent for theolymer. While the increases in diffusivity that accompany thesehanges allow greater infusion of the compounds of interest the

ssociated loss of physical structure is biologically undesirable.

Such ‘tuning’ of the CO2 conditions has previously beensed to prevent the occurrence of foamed structures followingmpregnation [36]. It is not necessary to completely saturate the

l Fluids 41 (2007) 173–178 177

olymer to achieve substantial impregnation; for low Tg mate-ials, clearly only moderate pressures are required. Judicioushoice of conditions makes it is possible to avoid changes in theorphology of polymer films and fibers. We believe these con-

itions may be generally related to the distance of the processingressure and temperature from the Tg of the CO2–polymer mix-ure. However, in this work we were unable to measure Tg of theO2–swollen polymer since it was well below ambient temper-tures.

Utilizing these relatively low pressures and temperatures, weetermined that CO2 can be used to enhance the chemical func-ionality of three-dimensional (3D) tissue engineering scaffolds.he ability to maintain polymer morphology while adding spe-ific molecules at depth will allow for follow-on studies of cellttachment, proliferation and differentiation as well as potentialn vivo applications involving a range of highly topical ‘biolo-ies’ – growth factors, cytokines, antiproliferatives – as well asntibiotics, analgesics and other more traditional compounds.

. Conclusions

Under relatively benign conditions, CO2 can alter the mor-hology of electrospun PCL. Temperature ranges from 10 to5 ◦C and pressures of 1.0 to 2.6 MPa (densities up to 0.05 g/cm3

t 10 ◦C) can, however, be used to retain the original form ofhese fibers. The resulting matrix is identical to the original well-efined structure. This subcritical CO2 exposure was then used tollow carboxytetramethlyrhodamine impregnation while retain-ng the original morphology. Carboxytetramethlyrhodamineelease following CO2 + carboxytetramethlyrhodamine solutionxposure occurred for at least 30 days. Samples not treated withO2 showed no detectable release after 5 days. Subcritical CO2onditions can be used to embed chemical functionality intoissue engineering scaffolds without a loss of form.

cknowledgements

The authors gratefully acknowledge the support of theational Science Foundation under Grant Nos. 0221678,425626, the State of Ohio Incentive Fund in Oncological Bio-aterials and the Ohio State University Perinatal Research andevelopment Fund for financial support. We also acknowledgerofessor Darrell Reneker of the University of Akron for his

echnical expertise and excellent advice.

eferences

[1] D.H. Reneker, I. Chun, Nanometer diameter fibres of polymer, producedby electrospinning, Nanotechnology 7 (1996) 216–223.

[2] D.H. Reneker, A.L. Yarin, H. Fong, S. Koombhongse, Bending instabilityof electrically charged liquid jets of polymer solutions in electrospinning,J. Appl. Phys. 87 (2000) 4531–4547.

[3] W.J. Li, R. Tuli, X.X. Huang, P. Laquerriere, R.S. Tuan, Multilineage

differentiation of human mesenchymal stem cells in a three-dimensionalnanofibrous scaffold, Biomaterials 26 (2005) 5158–5166.

[4] K. Fujihara, M. Kotaki, S. Ramakrishna, Guided bone regeneration mem-brane made of polycaprolactone/calcium carbonate composite nano-fibers,Biomaterials 26 (2005) 4139–4147.

Page 6

1 critica

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

78 O. Ayodeji et al. / J. of Super

[5] I.K. Kwon, S. Kidoaki, T. Matsuda, Electrospun nano- to microfiber fabricsmade of biodegradable copolyesters: structural characteristics, mechanicalproperties and cell adhesion potential, Biomaterials 26 (2005) 3929–3939.

[6] J. Zeng, X.S. Chen, Q.Z. Liang, X.L. Xu, X.B. Jing, Enzymatic degrada-tion of poly(l-lactide) and poly(epsilon-caprolactone) electrospun fibers,Macromol. Biosci. 4 (2004) 1118–1125.

[7] M.S. Khil, S.R. Bhattarai, H.Y. Kim, S.Z. Kim, K.H. Lee, Novel fabricatedmatrix via electrospinning for tissue engineering, J. Biomed. Mater. Res.Part B: Appl. Biomater. 72B (2005) 117–124.

[8] Y.Z. Zhang, H.W. Ouyang, C.T. Lim, S. Ramakrishna, Z.M. Huang, Elec-trospinning of gelatin fibers and gelatin/PCL composite fibrous scaffolds,J. Biomed. Mater. Res. Part B: Appl. Biomater. 72B (2005) 156–165.

[9] W.J. Li, R. Tuli, C. Okafor, A. Derfoul, K.G. Danielson, D.J. Hall, R.S.Tuan, A three-dimensional nanofibrous scaffold for cartilage tissue engi-neering using human mesenchymal stem cells, Biomaterials 26 (2005)599–609.

10] C.M. Hsu, S. Shivkumar, Nano-sized beads and porous fiber constructs ofpoly(epsilon-caprolactone) produced by electrospinning, J. Mater. Sci. 39(2004) 3003–3013.

11] C.M. Hsu, S. Shivkumar, N-Dimethylformamide additions to the solutionfor the electrospinning of poly(epsilon-caprolactone) nanofibers, Macro-mol. Mater. Eng. 289 (2004) 334–340.

12] M. Shin, O. Ishii, T. Sueda, J.P. Vacanti, Contractile cardiac grafts using anovel nanofibrous mesh, Biomaterials 25 (2004) 3717–3723.

13] M. Shin, H. Yoshimoto, J.P. Vacanti, In vivo bone tissue engineering usingmesenchymal stem cells on a novel electrospun nanofibrous scaffold, TissueEng. 10 (2004) 33–41.

14] W.J. Li, K.G. Danielson, P.G. Alexander, R.S. Tuan, Biological responseof chondrocytes cultured in three-dimensional nanofibrous poly(epsilon-caprolactone) scaffolds, J. Biomed. Mater. Res. Part A 67A (2003)1105–1114.

15] J. Zeng, X.S. Chen, X.Y. Xu, Q.Z. Liang, X.C. Bian, L.X. Yang, X.B. Jing,Ultrafine fibers electrospun from biodegradable polymers, J. Appl. Polym.Sci. 89 (2003) 1085–1092.

16] H. Yoshimoto, Y.M. Shin, H. Terai, J.P. Vacanti, A biodegradable nanofiberscaffold by electrospinning and its potential for bone tissue engineering,Biomaterials 24 (2003) 2077–2082.

17] K.H. Lee, H.Y. Kim, M.S. Khil, Y.M. Ra, D.R. Lee, Characterization ofnano-structured poly(epsilon-caprolactone) nonwoven mats via electro-spinning, Polymer 44 (2003) 1287–1294.

18] D.W. Hutmacher, Scaffolds in tissue engineering bone and cartilage, Bio-materials 21 (2000) 2529–2543.

19] M.B. Habal, A.H. Reddi, Bone Grafts and Bone Substitutes, W.B. SaundersCo., Philadelphia, 1992.

20] W.-J. Li, R. Tuli, C. Okafor, A. Derfoul, K.G. Danielson, D.J. Hall, R.S.Tuan, A three-dimensional nanofibrous scaffold for cartilage tissue engi-neering using human mesenchymal stem cells, Biomaterials 26 (2004)599–609.

21] W.-J. Li, K.G. Danielson, P.G. Alexander, R.S. Tuan, Biological responseof chondrocytes cultured in three-dimensional nanofibrous poly(e-caprolactone) scaffolds, J. Biomed. Mater. Res. Part A 67A (2003)1105–1114.

22] M. Honda, T. Yada, M. Ueda, K. Kimata, Cartilage formation by cul-tured chondrocytes in a new scaffold made of poly(l-lactide-epsilon-caprolactone) sponge, J. Oral Maxillofacial Surg. 58 (2000) 767–775.

23] B.S. Kim, J. Nikolovski, J. Bonadio, E. Smiley, D.J. Mooney, Engineeredsmooth muscle tissues: regulating cell phenotype with the scaffold, Exp.Cell Res. 251 (1999) 318–328.

24] C.Y. Xu, R. Inai, M. Kotaki, S. Ramakrishna, Aligned biodegradablenanofibrous structure: a potential scaffold for blood vessel engineering,Biomaterials 25 (2004) 877–886.

25] C. Xu, R. Inai, M. Kotaki, S. Ramakrishna, Electrospun nanofiber fabri-

cation as synthetic extracellular matrix and its potential for vascular tissueengineering, Tissue Eng. 10 (2004) 1160–1168.

26] S.G. Kazarian, M.F. Vincent, B.L. West, C.A. Eckert, Partitioning of solutesand cosolvents between supercritical CO2 and polymer phases, J. Supercrit.Fluids 13 (1998) 107–112.

[

l Fluids 41 (2007) 173–178

27] O. Muth, T. Hirth, H. Vogel, Polymer modification by supercritical impreg-nation, J. Supercrit. Fluids 17 (2000) 65–72.

28] S. Sicardi, L. Manna, M. Banchero, Diffusion of disperse dyes in PETfilms during impregnation with a supercritical fluid, J. Supercrit. Fluids 17(2000) 187–194.

29] P. Alessi, K. Ireneo, C. Angelo, F. Alessia, M. Mariarosa, Polydimethyl-siloxanes in supercritical solvent impregnation (SSI) of polymers, J. Super-crit. Fluids 27 (2003) 309–315.

30] T.T. Ngo, C.L. Liotta, C.A. Eckert, S.G. Kazarian, Supercritical fluidimpregnation of different azo-dyes into polymer: in situ UV–vis spectro-scopic study, J. Supercrit. Fluids 27 (2003) 215–221.

31] L.N. Nikitin, M.O. Gallyamov, R.A. Vinokur, A.Y. Nikolaev, E.E. Said-Galiyev, A.R. Khokhlov, H.T. Jespersen, K. Schaumburg, Swelling andimpregnation of polystyrene using supercritical carbon dioxide, J. Super-crit. Fluids 26 (2003) 263–273.

32] S.L. Shenoy, D. Cohen, R.A. Weiss, C. Erkey, Supercritical carbon dioxideaided preparation of conductive polyurethane–polypyrrole composites, J.Supercrit. Fluids 28 (2004) 233–239.

33] M. Tang, T.B. Du, Y.P. Chen, Sorption and diffusion of supercritical carbondioxide in polycarbonate, J. Supercrit. Fluids 28 (2004) 207–218.

34] L.D. Harris, B.-S. Kim, D.J. Mooney, Open pore biodegradable matrixesformed with gas foaming, J. Biomed. Mater. Res. 42 (1998) 396–402.

35] L.A. Stanton, F. Dehghani, N.R. Foster, Improving drug delivery usingpolymers and supercritical fluid technology, Aust. J. Chem. 55 (2002)443–447.

36] R.A. Quirk, R.M. France, K.M. Shakesheff, S.M. Howdle, Supercriticalfluid technologies and tissue engineering scaffolds, Curr. Opin. Solid StateMater. Sci. 8 (2005) 313–321.

37] O. Guney, A. Akgerman, Synthesis of controlled-release products in super-critical medium, AIChE J. 48 (2002) 856–866.

38] S.M. Howdle, M.S. Watson, M.J. Whitaker, M.C. Davies, K.M. Shakesheff,V.K. Popov, F.S. Mandel, J.D. Wang, Supercritical fluid mixing: preparationof thermally sensitive polymer composites containing bioactive materials,Chem. Commun. 1 (2001) 109–110.

39] D.L. Tomasko, H. Li, D. Liu, X. Han, M.J. Wingert, L.J. Lee, K.W. Koelling,A review of CO2 applications in the processing of polymers, Ind. Eng.Chem. Res. 42 (2003) 6431–6456.

40] T.L. Sproule, J.A. Lee, H. Li, J.J. Lannutti, D.L. Tomasko, Bioactive poly-mer surfaces via supercritical fluids, J. Supercrit. Fluids 28 (2004) 241–248.

41] X.B. Yang, M.J. Whitaker, W. Sebald, N. Clarke, S.M. Howdle, K.M.Shakesheff, R.C. Oreffo, Human osteoprogenitor bone formation usingencapsulated bone morphogenetic protein 2 in porous polymer scaffolds,Tissue Eng. 10 (2004) 1037–1045.

42] J. Zhang, A.J. Busby, C.J. Roberts, X. Chen, M.C. Davies, S.J.B. Tendler,S.M. Howdle, Preparation of a poly(methyl methacrylate)/ultrahigh molec-ular weight polyethylene blend using supercritical carbon dioxide and theidentification of a three-phase structure: an atomic force microscopy study,Macromolecules 35 (2002) 8869–8877.

43] I. Kikic, F. Vecchione, Supercritical impregnation of polymers, Curr. Opin.Solid State Mater. Sci. 7 (2003) 399–405.

44] O. Ayodeji, H. Powell, T. Summerfield, D. Powell, D. Kniss, D.L. Tomasko,J.J. Lannutti, Implant-based drug delivery via subcritical CO2 modification,J. Biomed. Mater. Res. Part A, submitted for publication.

45] D. Li, Y.N. Xia, Electrospinning of nanofibers: reinventing the wheel? Adv.Mater. 16 (2004) 1151–1170.

46] D.L. Woerdeman, P. Ye, S. Shenoy, R.S. Parnas, G.E. Wnek, O. Trofi-mova, Electrospun fibers from wheat protein: investigation of the interplaybetween molecular structure and the fluid dynamics of the electrospinningprocess, Biomacromolecules 6 (2005) 707–712.

47] J.B. Xie, Y.L. Hsieh, Ultra-high surface fibrous membranes from electro-spinning of natural proteins: casein and lipase enzyme, J. Mater. Sci. 38(2003) 2125–2133.

48] Y. Sato, T. Takikawa, A. Sorakubo, S. Takishima, H. Masuoka, M.

Imaizumi, Solubility and diffusion coefficient of carbon dioxide inbiodegradable polymers, Ind. Eng. Chem. Res. 39 (2000) 4813–4819.

49] Y. Sato, H. Masuoka, Solubility, diffusivity and related properties for car-bon dioxide + polymer solutions, Nippon Enerugi Gakkaishi 79 (2000)992–997.