This article appeared in a journal published by Elsevier. The attachedcopy is furnished to the author for internal non-commercial researchand education use, including for instruction at the authors institution

and sharing with colleagues.

Other uses, including reproduction and distribution, or selling orlicensing copies, or posting to personal, institutional or third party

websites are prohibited.

In most cases authors are permitted to post their version of thearticle (e.g. in Word or Tex form) to their personal website orinstitutional repository. Authors requiring further information

regarding Elsevier’s archiving and manuscript policies areencouraged to visit:

http://www.elsevier.com/copyright

Author's personal copy

Electrospun biphasic drug release polyvinylpyrrolidone/ethyl cellulose core/sheathnanofibers

D.G. Yu ⇑, X. Wang ⇑, X.Y. Li, W. Chian, Y. Li, Y.Z. LiaoSchool of Materials Science & Engineering, University of Shanghai for Science and Technology, 516 Jungong Road, Yangpu District, Shanghai 200093, China

a r t i c l e i n f o

Article history:Received 21 July 2012Received in revised form 23 September2012Accepted 16 October 2012Available online 23 October 2012

Keywords:Biphasic releaseCore/sheath nanofibersCoaxial electrospinningNanocomposites

a b s t r a c t

The capability of core/sheath nanofibers prepared using coaxial electrospinning to provide adjustablebiphasic drug release was investigated. Using ketoprofen (KET) as the model drug, polyvinylpyrrolidoneas the sheath polymer, and ethyl cellulose as the core matrix, the coaxial process could be conductedsmoothly and continuously without spinneret clogging. Scanning electron microscopy and transmissionelectron microscopy revealed linear nanofibers with homogeneous and clear core/sheath structures. Dif-ferential scanning calorimetry and X-ray diffraction verified that the core/sheath nanofibers were nano-composites, with the drug present in the polymer matrix in an amorphous state. Attenuated totalreflectance–Fourier transform infrared spectra demonstrated that the sheath polymer and core matrixwere compatible with KET owing to hydrogen bonding. In vitro dissolution tests showed that the core/sheath nanofibers could provide typical biphasic drug release profiles consisting of an immediate andsustained release. The amount of drug released in the first phase was tailored by adjusting the sheathflow rate, and the remaining drug released in the second phase was controlled by a typical diffusionmechanism. The present study shows a simple and useful approach for the systematic design and fabri-cation of novel biomaterials with structural characteristics for providing complicated and programmeddrug release profiles using coaxial electrospinning.

� 2012 Acta Materialia Inc. Published by Elsevier Ltd. All rights reserved.

1. Introduction

In clinical applications, the desired drug release profiles shouldobey biological rhythm for effective and safe drug delivery andconvenient administration. For some pharmaceutical ingredientssuch as non-steroidal anti-inflammatory drugs (NSAID), as wellas antihypertensive, antihistaminic and anti-allergic agents [1,2],an initial release of a fraction of the dose in the shortest time afteradministration is favored for relieving the symptoms of the dis-ease. Meanwhile, sustained release of the remaining dose over adefined period can optimize the therapy and avoid repeatedadministration, for the patients’ convenience [3,4]. During biphasicrelease, a drug is released at two rates or in two periods. A typicalbiphasic release system can provide immediate drug release fol-lowed by a constant release.

In recent decades, different traditional pharmaceutical tech-niques such as tableting, casting and spraying have been investi-gated for preparing biphasic drug delivery systems (DDS). Someof these systems include mixed films, multi-layered films, multipletablets, multi-layered tablets, film-coated tablets and elutingstents [5–8]. Meanwhile, advanced technologies are continuously

exploited in the literature to produce novel materials or DDS forfurnishing biphasic release, taking advantage of more accuratetime-programmed administration of active ingredients and fulfill-ing the specific therapeutic needs of some diseases. These tech-niques include three-dimensional printing, nanotechniques [9–12] and single fluid electrospinning [13].

Electrospinning has attracted considerable attention as a nano-technology for nanofiber production because of its simplicity andcost effectiveness. It can produce nanofibers with unique proper-ties and versatile applications [14–21]. One of the most significantbreakthroughs in this area is coaxial electrospinning, in which aconcentric spinneret can accommodate two liquids [22,23]. Coaxialelectrospinning is widely used in controlling the secondary struc-tures of nanofibers, encapsulating drugs or biological agents intopolymer nanofibers, preparing nanofibers from materials that lackfilament-forming properties, enclosing functional liquids withinthe fiber matrix, manipulating the size of self-assembled nanopar-ticles, preparing ultrafine fibers from concentrated polymer solu-tions, and improving the quality of nanofibers [24–29]. Recently,advanced functional materials fabricated using electrospinninghave attracted considerable attention because of their ability to al-low the controlled release of multiple active ingredients [30,31].Electrospinning could be used to control the microstructure andspatial deposition of components. Thus, with the appropriate

1742-7061/$ - see front matter � 2012 Acta Materialia Inc. Published by Elsevier Ltd. All rights reserved.http://dx.doi.org/10.1016/j.actbio.2012.10.021

⇑ Corresponding authors. Tel.: +86 21 55271687; fax: +86 21 55270632 (D.G. Yu).E-mail addresses: ydg017@gmail.com (D.G. Yu), wangxia@usst.edu.cn (X. Wang).

Acta Biomaterialia 9 (2013) 5665–5672

Contents lists available at SciVerse ScienceDirect

Acta Biomaterialia

journal homepage: www.elsevier .com/locate /actabiomat

Author's personal copy

polymer matrix, it is hypothesized that this technique could also beused to develop nanoproducts with biphasic drug profiles.

Polyvinylpyrrolidone (PVP), a hydrophilic polymer excipientwith a wide variety of applications in medicine, food, pharmacyand cosmetics, was selected as the filament-forming matrix ofthe sheath part for immediate drug release [32–34]. Ethyl cellulose(EC) is an inert, non-toxic and stable hydrophobic polymer suitablefor sustained release matrices [13,35].

Accordingly, the present study investigates the preparation ofcore/sheath nanofibers for providing biphasic drug release usingcoaxial electrospinning. Ketoprofen (KET), a NSAID active ingredi-ent with poor water solubility [36,37], was used as the model drugloaded into the sheath and core parts of the nanofibers.

2. Experimental

2.1. Materials

PVP K60 ðMw ¼ 360;000Þ was purchased from Shanghai Yun-hong Pharmaceutical Aids and Technology Co. Ltd. (Shanghai, Chi-na). KET was purchased from Wuhan Fortuna Chemical Co. Ltd.(Hubei, China). EC (6 mPa s to 9 mPa s) was obtained from AladdinChemistry Co. Ltd. (Shanghai, China). Methylene blue, N,N-dimeth-ylacetamide (DMAc) and anhydrous ethanol were purchased fromSinopharm Chemical Reagent Co. Ltd. (Shanghai, China). All otherchemicals used were analytical grade, and water was doubly dis-tilled before use.

2.2. Electrospinning

The core solutions were prepared by dissolving 24 g EC, 3 g KETand 2 mg methylene blue in 100 ml ethanol. The sheath solutionwas prepared by placing 9 g PVP and 1 g KET in 100 ml of a solventmixture of DMAc and ethanol in a volume ratio of 1:9.

Two syringe pumps (KDS100 and KDS200, Cole-Parmer, IL, USA)and a high-voltage power supply (ZGF 60 kV/2 mA, Shanghai SuteCorp., Shanghai, China) were used for coaxial electrospinning. Allelectrospinning processes were carried out under ambient condi-tions (22 ± 3 �C with relative humidity 58 ± 5%). A homemade con-centric spinneret [38] was used to conduct both single fluid(adjusting the core or sheath fluid flow rate to 0 ml h�1) and coax-ial electrospinning processes. The electrospinning process was re-corded using a digital video recorder (PowerShot A490, Canon,Tokyo, Japan). For optimization, the applied voltage was fixed at15 kV, and the fibers were collected on an aluminum foil at a dis-tance of 20 cm. All other parameters are listed in Table 1.

2.3. Characterization

2.3.1. MorphologyThe morphology of the fiber mats was assessed by field emis-

sion scanning electron microscopy (FESEM) using an S-4800 micro-scope (Hitachi, Tokyo, Japan). Prior to the examination, the samples

were platinum sputter-coated under a nitrogen atmosphere to ren-der them electrically conductive. Images were recorded at an exci-tation voltage of 10 kV. The average fiber diameter was determinedby measuring their diameters in FESEM images at more than 100places, using the NIH Image J software (National Institutes ofHealth, MD, USA). Before platinum coating, the cross sections ofthe fiber mats were prepared by placing them in liquid nitrogenbefore they were manually broken.

Transmission electron microscopy (TEM) images of the sampleswere recorded on a JEM 2100F field emission transmission electronmicroscope (JEOL, Tokyo, Japan). TEM samples of the core/sheathnanofibers were collected by fixing a lacey carbon-coated coppergrid on the collector.

2.3.2. Physical status and compatibilityDifferential scanning calorimetry (DSC) was carried out using

an MDSC 2910 differential scanning calorimeter (TA InstrumentsCo., USA). Sealed samples were heated at 1 at 10 �C min�1 from20 �C to 250 �C. The nitrogen gas flow rate was 40 ml min�1.

The X-ray diffraction (XRD) analysis was conducted using a D/Max-BR diffractometer (RigaKu, Japan) with Cu Ka radiation in a2h range of 5–60� at 40 mV and 300 mA.

Attenuated total reflectance–Fourier transform infrared (ATR-FTIR) spectroscopy was carried out on a Nicolet-Nexus 670 FTIRspectrometer (Nicolet Instrument Corporation, Madison, USA) ata range of 500 cm�1 to 4000 cm�1 and a resolution of 2 cm�1.

2.3.3. In vitro dissolution testsIn vitro dissolution tests were carried out according to the Chi-

nese Pharmacopoeia (2005 edn.) Method II, which is a paddle meth-od using a RCZ-8A dissolution apparatus (Tianjin University RadioFactory, Tianjin, China). Drug-loaded nanofibers (200 mg) wereplaced in 600 ml physiological saline (PS, 0.9 wt.%) at 37 ± 1 �C.The instrument was set to 50 rpm, providing sink conditions withC < 0.2Cs. At predetermined time points, 5.0 ml aliquots of the sam-ples were withdrawn from the dissolution medium and replacedwith fresh medium to maintain a constant volume. After filtrationthrough a 0.22 lm membrane (Millipore, MA, USA) and appropri-ate dilution with PS, the samples were analyzed at 260 nm, usinga UV–vis spectrophotometer (UV-2102PC, Unico Instrument Co.Ltd., Shanghai, China). The concentration of released KET wasback-calculated from the data obtained against a predeterminedcalibration curve.

The actual content of KET in the fibers was quantified by dis-solving each sample in ethanol for extraction and later proceedingto the above-mentioned procedure. The accumulative percentageof drug released from the electrospun fibers was calculated usingthe following equation:

Pð%Þ ¼ qn � V0 þPn�1

i¼1 qi � VQ 0

� 100

where V0 is the volume of the dissolution medium (ml), V is the vol-ume of the withdrawn sample (ml), Q0 is the total amount of KET in

Table 1Experimental parameters for the fabrication of different nanofibers.

No. Process Fluid and flow rate (ml h�1) Fiber morphologyc Diameter (nm)

Sheath fluida Core fluidb

F1 Single – 1.0 Linear 710 ± 130F2 Single 1.0 – Linear 910 ± 240F3 Coaxial 0.5 0.5 Linear 780 ± 90F4 Coaxial 0.8 0.5 Linear 940 ± 80

a Sheath fluid consists of 9% (w/v) PVP K60 and 1% (w/v) KET in a mixture of ethanol and DMAc (1:9, v:v).b Core fluid consists of 24% (w/v) EC and 3% (w/v) KET in ethanol.c In this column, ‘‘Linear’’ morphology refers to nanofibers with few beads or spindles.

5666 D.G. Yu et al. / Acta Biomaterialia 9 (2013) 5665–5672

Author's personal copy

the fibers (mg), qn is the drug concentration determined at No. n(mg l�1), and qi is the drug concentration determined at No. i(mg l�1). The experiments were carried out six times, and thecumulative percentage reported as mean values was plotted as afunction of time (T, h).

3. Results and discussion

3.1. Coaxial electrospinning

A schematic diagram of the modified coaxial electrospinningprocess is shown in Fig. 1a. A homemade concentric spinneretwas employed (Fig. 1b). The critical voltage applied to a fluid to ini-tiate Taylor cone formation and straight thinning jet (Vc) has aclose relationship with the diameter of the sheath part of the con-centric spinneret [39]:

Vc �

ffiffiffiffiffiffiffifficd2

eR

s

where Vc is the critical voltage for a jet emanating from the menis-cus tip, d is the electrode separation, e is the permittivity, c is thesurface tension and R is the principal curvature of the liquid menis-cus. A small diameter spinneret orifice gives a high R value. Thus,only a small Vc is needed to initiate electrospinning. The homemadespinneret used in this work has outside and inner diameters of 1.2and 0.3 mm, respectively (Fig. 1b), facilitating the initiation of coax-ial electrospinning.

A digital image of the apparatus arrangement for the coaxialelectrospinning process is shown in Fig. 2a. Two syringe pumpswere used to drive the sheath and core fluids independently. Analligator clip was used to connect the inner stainless steel capillaryof the spinneret to the high-voltage power supply (Fig. 2b).

The coaxial electrospinning process could be changed to the sin-gle fluid electrospinning process by adjusting the flow rate of oneof the fluids to 0 ml h�1. Both the sheath and core solutions hadgood electrospinnability under the electrical field and the selectedconditions for the preparation of the corresponding drug-loadedcomposite nanofibers F1 and F2. The electrospinning process ofthe core solutions was smoothed by occasionally removing thesemi-solid skin formed around the nozzle of the spinneret to avoidclogging. Clogging is a critical but common problem during elec-trospinning, especially when a highly volatile solvent is used toprepare a polymer solution [40,41]. However, the sheath fluidcould be run smoothly and continuously without user interven-tion. The sheath solvent, consisting of DMAc and ethanol, couldeffectively prevent spinneret clogging during the electrospinningbecause of the presence of DMAc with a high boiling point of165.9 �C.

The coaxial process could be undertaken continuously under avoltage of 12 kV and a flow rate of 0.5 ml h�1 for the sheath andcore fluids. A typical fluid jet trajectory was created. The jet trajec-tory consisted of a straight thinning jet emitted from a compoundTaylor cone, as indicated by the methylene blue marker in Fig. 2c,followed by a bending and whipping instability region with loopsof increasing size (Fig. 2d). The applied voltage was further in-creased to 15 kV, which produced a similar jet trajectory with ashorter straight jet (Fig. 2e). Given that nanofibers become thinnerwith increasing applied voltage, all experiments were conductedunder an applied voltage of 15 kV.

3.2. Morphology and structure of nanofibers

Fig. 3 shows that all four types of nanofibers had smooth sur-faces and uniform structures without any beads-on-a-string mor-phology. No drug particles appeared on the surface of the fibers,indicating good compatibility between the polymers and KET.

The nanofibers F1 and F2 prepared through single fluid electros-pinning had average diameters of 710 ± 130 nm (Table 1; Fig. 3aand b) and 910 ± 240 nm (Table 1; Fig. 3c and d), respectively.The core/sheath nanofibers F3 and F4 had average diameters of780 ± 90 nm (Table 1; Fig. 3e and f) and 940 ± 80 nm (Table 1;Fig. 3g and h), respectively. These results verified that PVP andEC had good electrospinnability in the selected solvent systemsand concentrations, and that the coaxial electrospinning processconsisting of two electrospinnable fluids could be carried outsmoothly.

As expected, the FESEM images of the cross sections of the core/sheath nanofibers F3 (Fig. 4a and b) and F4 (Fig. 4c and d) wereround and smooth. This finding suggests that these nanofibersare nanocomposites with a homogeneous structure. Similar to sin-gle fluid electrospinning, coaxial electrospinning is a one-step‘‘top-down’’ fabrication process, in which electrical energy isexploited to dry and solidify micro-fluid jets. The processes pro-duce nanosized fibers very rapidly, often at �10�2 s [45]. As a re-sult, the physical state of the components in the liquid solutionswas propagated into the solid nanofibers without any discernednanoparticles generated from phase separation.

The nanofibers F3 and F4 had clear core/sheath structures(Fig. 5). Similarly to the FESEM results, no nanoparticles were dis-cerned in the sheath and core parts. This finding suggests thatthese nanofibers have a homogeneous structure. However, thenanofibers of F4 (Fig. 5b) had larger diameters and thicker sheathparts than those of F3 (Fig. 5a). This difference could be attributedto the larger sheath flow rate in F4 than in F3. The fast drying elec-trospinning process not only propagated the physical state of thecomponents in the liquid solutions into the solid nanofibers, butalso duplicated the concentric structure of the spinneret on a

Fig. 1. Coaxial electrospinning: (a) schematic diagram of the process; (b) digital image of the homemade concentric spinneret.

D.G. Yu et al. / Acta Biomaterialia 9 (2013) 5665–5672 5667

Author's personal copy

macroscale to nanoproducts on a nanoscale. As a result, the com-ponents in the sheath and core fluids occurred in the sheath andcore parts of the nanofibers, respectively, with weak diffusion.

3.3. Physical status and compatibility of components

DSC and XRD tests were conducted to determine the physicalstatus of KET in the four nanofibers. DSC thermograms are shownin Fig. 6a. The DSC curve of pure KET exhibited a single endother-mic response corresponding to its melting point of 96.24 �C(DHf = �114.72 J g�1). As amorphous polymers, PVP and EC didnot show any fusion peaks or phase transitions. The DSC thermo-grams of the composite nanofibers F1, F2 and F3 did not showany characteristic peaks of KET. This finding suggests that the drugwas no longer present as a crystalline material, but was convertedinto an amorphous state in all the nanofibers, regardless of thegeneration process.

Numerous distinct reflections were found in the XRD pattern ofpure KET (Fig. 6b), demonstrating that the pure drug is a crystallinematerial. The diffraction patterns of PVP and EC showed a diffusebackground pattern with two diffraction halos, indicating thatthe polymer is amorphous. No characteristic reflections of KET

Fig. 2. Observations of the coaxial electrospinning process: (a) digital image showing the apparatus arrangement; (b) connection between the power supply and thespinneret; (c) typical compound Taylor cone of the core/sheath fluids under a voltage of 12 kV; (d) typical fluid jet trajectory under a voltage of 12 kV; (e) typical fluid jettrajectory under a voltage of 15 kV (the flow rate of the sheath and core fluids is 0.5 ml h�1).

Fig. 3. FESEM images of the electrospun nanofibers and their diameter distributions: (a, b) F2; (c, d) F1; (e, f) F3; (g, h) F4.

Fig. 4. FESEM images of the cross sections of the core/sheath nanofibers: (a, b) F3;(c, d) F4.

5668 D.G. Yu et al. / Acta Biomaterialia 9 (2013) 5665–5672

Author's personal copy

were found in the patterns of nanofibers F1, F2 and F3. This obser-vation indicates that KET was no longer present as a crystallinematerial, but was converted into an amorphous state.

The results of DSC and XRD both confirmed that KET was highlydistributed in the nanofiber matrix in its amorphous state and thatthe original structure of the pure materials was lost. These resultsagree with the FESEM and TEM results, verifying that the electro-spun drug-loaded nanofibers are nanocomposites or drug nanosol-id dispersions that could be used to resolve dissolution problems ofpoorly water-soluble drugs [42–44].

Compatibility among the components, which was investigatedthrough FTIR analysis (Fig. 7), is essential for producing high-qual-ity and stable nanofibers. Second-order interactions such as hydro-gen bonding, electrostatic interactions and hydrophobicinteractions improve compatibility. KET, PVP and EC moleculespossess free hydroxyl (acting as potential proton donors for hydro-gen bonding) and carbonyl (potential proton receptors) groups(Fig. 7b). Therefore, hydrogen-bonding interactions may occurwithin the KET-loaded PVP nanofibers (F1), EC nanofibers (F2)and their core–sheath nanofibers (F3) (Fig. 7c).

The ATR-FTIR spectra of the components and their nanofibersare shown in Fig. 7a. Two well-defined sharp peaks at 1697 and1655 cm�1 were observed for pure crystalline KET. The formerwas assigned to the stretching vibration of the carbonyl group inthe KET dimer (Fig. 7c), whereas the latter was assigned to thestretching of the ketone group. The peak at 1697 cm�1 was ob-served because KET molecules in crystalline form are bound to-gether in dimers. However, the peak at 1696 cm�1 disappeared inthe spectra of F1, F2 and F3, indicating the breakage of the KET di-mers and the formation of hydrogen bonds between the PVP/ECcarbonyl group and the KET hydroxyl group (Fig. 7c). By interactingwith the polymer, KET molecules are less likely to form the dimersthat are essential for the formation of a crystal lattice.

3.4. In vitro drug release profiles

KET has a UV absorbance peak at 260 nm. Thus, the amount ofKET released from the fibers was determined by UV spectroscopy,using a predetermined calibration curve C = 15.27A � 0.0034(R = 0.9996), where C is the concentration of KET (lg ml�1), andA is the solution absorbance at 260 nm (linear range 2–20 lg ml�1).

The actual content of KET in all the fibers was equivalent to thetheoretical calculation value, suggesting no drug loss during theelectrospinning process. The in vitro drug release profiles of thefour types of nanofibers are shown in Fig. 8a and b. The nanofibersof F2 disappeared instantly after they were placed in the dissolu-tion media. In vitro dissolution tests verified that KET was dis-solved completely in the bulk media in the first minutes. Yuet al. [33] reported that PVP nanofiber mats loaded with thepoorly water-soluble drug ibuprofen are good oral fast-integratingmembranes that can release the drug completely in several sec-onds. The reasons can be concluded as follows. First, PVP hashygroscopic and hydrophilic properties, and polymer–solventinteractions are stronger than polymer–polymer attraction forces.Thus, the polymer chain can absorb solvent molecules rapidly,increasing the volume of the polymer matrix and allowing thepolymer chains to loosen out from their coiled shape. Second,the three-dimensional continuous web structure of the membranecan offer a huge surface area for PVP to absorb water molecules,greater porosity for the water molecules to diffuse into the innerpart of the membrane, and void space for the polymer to be swol-len and disentangled and for the dissolved KET molecules to dis-perse into the bulk dissolution medium. Third, the drug and thematrix polymer formed composites at the molecular level. Thus,KET molecules can dissolve simultaneously with PVP molecules.That is, the ability of these nanofibers to improve significantly

Fig. 5. TEM images of the core/sheath nanofibers: (a) F3; (b) F4.

Fig. 6. Physical status characterization: (a) DSC thermograms; (b) XRD patterns of the raw materials (EC, PVP and KET) and the drug-loaded nanofibers: F1 and F2 (preparedby single fluid electrospinning) and F3 (prepared by coaxial electrospinning).

D.G. Yu et al. / Acta Biomaterialia 9 (2013) 5665–5672 5669

Author's personal copy

the dissolution rate of poorly water-soluble drugs is attributableto the synergistic effects of nanosized fibers, the web structure

of the membranes and the drug composites with the filament-forming matrix.

Fig. 7. Compatibility investigation: (a) ATR-FTIR spectra of the components (KET, EC and PVP) and their electrospun nanofibers F1, F2 (prepared by single fluidelectrospinning) and F3 (prepared by coaxial electrospinning); (b) molecular structures of KET, EC and PVP; (c) hydrogen bonding between KET–PVP, KET–EC and KET–KET.

Fig. 8. In vitro dissolution tests: (a, b) in vitro drug release profiles of the KET-loaded EC nanofibers F1, PVP nanofibers F2 (prepared by single fluid electrospinning), F3, and F4(prepared by coaxial electrospinning) at the first 2 h and the full range, respectively; (c)–e) FESEM images and diameter distribution of F3 after the removal of the sheath partin the first phase.

5670 D.G. Yu et al. / Acta Biomaterialia 9 (2013) 5665–5672

Author's personal copy

As expected, the nanofibers of F3 and F4 could provide typicalbiphasic drug release profiles with immediate release percentagesof 30.7% and 41.2%, respectively, after they were placed in the dis-solution media for half an hour. They later released 50.6% and48.5% of the remaining KET sustainably for 24 h (Fig. 8a and b).The core part of the nanofibers of F3 after removal of the sheathKET/PVP had a diameter distribution of 470 ± 50 nm (Fig. 8c–e).The surfaces of the nanofibers remained smooth without discernednanoparticles, suggesting that KET in the sheath part was freed intothe dissolution medium synchronously with the matrix PVPthrough a polymer-controlled erosion mechanism.

The KET release profiles from the KET-loaded EC nanofibers (F1)and the core part of the core/sheath nanofibers (F3 and F4) wereanalyzed using the Peppas equation [46]:

Q ¼ ktn

where Q is the drug release percentage, t is the release time, k is aconstant reflecting the structural and geometric characteristics ofthe fibers, and n is the release exponent that indicates the drug re-lease mechanism.

The regressed result for F2 is Q = 20.8t0.43 (R2 = 0.9947), indicat-ing that the drug release from the KET/EC nanofibers was con-trolled via a typical Fickian diffusion mechanism by a value ofthe release exponent 0.18 (<0.45). For the core parts of nanofibersF3 and F4, the regressed equations are Q = 13.7t0.41 (R2 = 0.9963)and Q = 12.6t0.40 (R2 = 0.9951), respectively. These results demon-strated that the second phasic release of F3 and F4 were still tobe controlled by a typical Fickian diffusion mechanism. Thus, thebiphasic release profiles could be achieved through a core/sheathstructure, wherein the sheath hydrophilic polymer provides fastinitial release and the core matrix furnishes the later sustained re-lease. However, the drug release mechanism was still determinedby the properties of the drug-loaded matrices.

The present study demonstrated that the amount of drug re-leased could be controlled by adjusting the sheath flow rate. Otherparameters such as drug concentrations in the sheath or core fluidscan also be exploited to tailor the amount of drug released at dif-ferent phases. The matrix in the core part can be selected to manip-ulate drug release, e.g., by using a pH-sensitive polymer matrix toproduce a delayed release of the second phase or a colon-targeteddrug delivery. Further in vivo experiments on animals will be con-ducted to investigate the effects of these biphasic drug releasecore/sheath nanofibers.

4. Conclusion

Coaxial electrospinning was carried out to prepare core/sheathnanofibers that could provide biphasic drug release profiles, usingPVP as the sheath polymer and EC as the core matrix. With theselection of suitable sheath and core fluids, the coaxial electrospin-ning process could be conducted smoothly and continuously with-out spinneret clogging, and the generation of clear core/sheathstructures could be achieved. XRD and DSC verified that all thecomponents were present in the core/sheath nanofibers in theiramorphous states. ATR-FTIR spectra demonstrated that the sheathpolymer and core matrix were compatible with KET, owing tohydrogen bonding. In vitro dissolution tests showed that thecore/sheath nanofibers could provide typical biphasic release pro-files consisting of an immediate and sustained release. The amountof drug released in the first phase could be tailored by adjusting thesheath flow rate. The remaining drug released in the second phasewas controlled by a typical diffusion mechanism. The presentstudy showed a simple and useful approach for the systematic de-sign and fabrication of novel biomaterials with structural charac-

teristics for providing complicated and programmed drug releaseprofiles using coaxial electrospinning.

Acknowledgments

This work was supported by the Key project of Shanghai Muni-cipal Education Commission (No. 13ZZ113), the National ScienceFoundation of China (Nos. 51173107 and 51203090) and the Nat-ural Science Foundation of Shanghai (No. 12ZR1446700).

Appendix A. Figures with essential colour discrimination

Certain figures in this article, particularly Figs. 1–3, and 6–8, aredifficult to interpret in black and white. The full colour images canbe found in the on-line version, at http://dx.doi.org/10.1016/j.actbio.2012.10.021.

References

[1] Ofori-Kwakye K, Fell JT. Leaching of pectin from mixed films containing pectin,chitosan and HPMC intended for biphasic drug delivery. Int J Pharm2003;250:251–7.

[2] Loo SCJ, Tan ZYS, Chow YJ, Lin SLI. Drug release from irradiated PLGA and PLLAmulti-layered films. J Pharm Sci 2010;99:3060–71.

[3] Maggi L, Machiste OE, Torre LM, Conte U. Formulation of biphasic releasetablets containing slightly soluble drugs. Eur J Pharm Biopharm1999;48:37–42.

[4] Jha MK, Rahman MH, Rahman MM. Biphasic oral solid drug delivery system: areview. Int J Pharm Sci Res 2011;2:1108–15.

[5] Abdul S, Poddar SS. A flexible technology for modified release of drugs: multilayered tablets. J Control Release 2004;97:393–405.

[6] Bandari S, Eaga CM, Thadishetty A, Yamsani MR. Formulation and evaluation ofmultiple tablets as a biphasic gastroretentive floating drug delivery system forfenoverine. Acta Pharm 2010;60:89–97.

[7] Chen MC, Liang HF, Chiu YL, Chang Y, Wei HJ, Sung HW. A novel drug elutingstent spray-coated with multi-layers of collagen and sirolimus. J ControlRelease 2005;108:178–89.

[8] Ofori-Kwakye K, Fell JT, Sharma HL, Smith AM. Gamma scintigraphicevaluation of film-coated tablets intended for colonic or biphasic release. IntJ Pharm 2004;270:307–13.

[9] Monkhouse DC, Yao J, Sherwood MJ, Cima MJ, Bornancini E. Dosage formsexhibiting multi-phasic release kinetics and methods of manufacture thereof.US Patent No. 6280771; 2001.

[10] Yu DG, Zhu LM, Branford-White C, Yang XL. Three-dimensional printing inpharmaceutics – promises and problems. J Pharm Sci 2008;97:3666–90.

[11] Mora-Huertas CE, Fessi H, Elaissari A. Polymer-based nanocapsules for drugdelivery. Int J Pharm 2010;385:113–42.

[12] Yaghmur A, Glatter O. Characterization and potential applications ofnanostructured aqueous dispersions. Adv Colloid Interf Sci 2009;147–148:333–42.

[13] Huang LY, Branford-White C, Shen XX, Yu DG, Zhu LM. Time-engineeredbiphasic drug release by electrospun nanofiber meshes. Int J Pharm2012;436:88–96.

[14] Huang ZM, Zhang YZ, Kotaki M, Ramakrishna S. A review on polymernanofibers by electrospinning and their applications in nanocomposites.Compos Sci Technol 2003;63:2223–53.

[15] Chung S, Ingle NP, Montero GA, Kim SH, King MW. Bioresorbable elastomericvascular tissue engineering scaffolds via melt spinning and electrospinning.Acta Biomater 2010;6:1958–67.

[16] Im JS, Yun J, Lim YM, Kim HI, Lee YS. Fluorination of electrospun hydrogelfibers for a controlled release drug delivery system. Acta Biomater2010;6:102–9.

[17] Zhang YZ, Venugopal J, Wong SY, Li X, Su B, Ramakrishna S, et al. Enhancedbiomineralization in osteoblasts on a novel electrospun biocompositenanofibrous substrate of hydroxyapatite/collagen/chitosan. Tissue Eng A2010;16:1949–60.

[18] Puppi D, Piras AM, Detta N, Dinucci D, Chiellini F. Poly(lactic-co-glycolic acid)electrospun fibrous meshes for the controlled release of retinoic acid. ActaBiomater 2010;6:1258–68.

[19] Sargeant TD, Aparicio C, Goldberger J, Cui H, Stupp SI. Mineralization ofpeptide amphiphiles nanofibers and its effect on differentiation of humanmesenchymal stem cells. Acta Biomater 2012;8:2456–65.

[20] Liao H, Qi R, Shen M, Cao X, Guo R, Zhang Y, et al. Improved cellular responseon multiwalled carbon nanotube-incorporated electrospun polyvinyl alcohol/chitosan nanofibrous scaffolds. Colloids Surf B: Biointerfaces 2011;84:528–35.

[21] Jiang QR, Reddy N, Yang YQ. Cytocompatible crosslinking of electrospun zeinfibers for the development of waterstable tissue engineering scaffolds. ActaBiomater 2010;6:4042–51.

[22] Moghe AK, Gupta BS. Co-axial electrospinning for nanofiber structures:preparation and applications. Polym Rev 2008;48:353–77.

D.G. Yu et al. / Acta Biomaterialia 9 (2013) 5665–5672 5671

Author's personal copy

[23] Dzenis Y. Spinning continuous fibers for nanotechnology. Science2004;304:1917–9.

[24] Su Y, Su Q, Liu W, Lim M, Venugopal JR, Mo X, et al. Controlled release of bonemorphogenetic protein and dexamethasone loaded in core–shell PLLACL–collagen fibers for use in bone tissue engineering. Acta Biomater2012;8:763–71.

[25] Lu Y, Jiang H, Tu K, Wang L. Mild immobilization of diverse macromolecularbioactive agents onto multifunctional fibrous membranes prepared by coaxialelectrospinning. Acta Biomater 2009;5:1562–74.

[26] Yu DG, Williams GR, Gao LD, Bligh SWA, Yang JH, Wang X. Coaxialelectrospinning with sodium dodecylbenzene sulfonate solution for highquality polyacrylonitrile nanofibers. Colloid Surf A 2012;396:161–8.

[27] Tiwari SK, Tzezana R, Zussman E, Venkatraman SS. Optimizing partition-controlled drug release from electrospun core–shell fibers. Int J Pharm2010;392:209–17.

[28] Yarin AL, Zussman E, Wendorff JH, Greiner A. Material encapsulation andtransport in core–shell micro/nanofibers, polymer and carbon nanotubes andmicro/nanochannels. J Mater Chem 2007;17:2585–99.

[29] Yu DG, Chatterton NP, Yang JH, Wang X, Liao YZ. Coaxial electrospinning withTriton X-100 solutions as sheath fluids for preparing PAN nanofibers.Macromol Mater Eng 2012;297:395–401.

[30] Song B, Wu C, Chang J. Dual drug release from electrospun poly (lactic-co-glycolic acid)/mesoporous silica nanoparticles composite mats with distinctrelease profiles. Acta Biomater 2012;8:1901–7.

[31] SuY SuQ, Liu W, Jin G, Mo X, Ramakrishn S. Dual-drug encapsulation andrelease from core–shell nanofibers. J Biomater Sci Polym Ed 2012;23:861–71.

[32] Yu DG, Branford-White C, White K, Bligh SWA, Williams G, Zhu LM, et al. Self-assembled liposomes from amphiphilic electrospun nanofibers. Soft Matter2011;7:8239–47.

[33] Yu DG, Shen XX, Branford-White C, White K, Zhu LM, Bligh SWA. Oral fast-dissolving drug delivery membranes prepared from electrospun PVP ultrafinefibers. Nanotechnology 2009;20:055104.

[34] Yu DG, White K, Yang JH, Wang X, Qian W, Li Y. PVP nanofibers prepared usingco-axial electrospinning with salt solution as sheath fluid. Mater Lett2012;67:78–80.

[35] Bruno L, Kasapis S, Heng PWS. Effect of polymer molecular weight on thestructural properties of non aqueous ethyl cellulose gels intended for topicaldrug delivery. Carbohydr Polym 2012;88:382–8.

[36] Nyangoga H, Aguado E, Goyenvalle E, Baslé MF, Chappard D. A non-steroidalanti-inflammatory drug (ketoprofen) does not delay b-TCP bone graft healing.Acta Biomater 2010;6:3310–7.

[37] Yu DG, Yu JH, Chen L, Williams GR, Wang X. Modified coaxial electrospinningfor the preparation of high-quality ketoprofen-loaded cellulose acetatenanofibers. Carbohydr Polym 2012;90:1016–23.

[38] Yu DG, Zhu LM, Branford-White C, Yang JH, Wang X, Li Y, et al. Soliddispersions in the form of electrospun core–sheath nanofibers. Int J Nanomed2011;6:3271–80.

[39] Yeo LY, Friend JR. Electrospinning carbon nanotube polymer compositenanofibers. J Exp Nanosci 2006;2:177–209.

[40] Kanjanapongkul K, Wongsasulak S, Yoovidhya T. Prediction of clogging timeduring electrospinning of zein solution: scaling analysis and experimentalverification. Chem Eng Sci 2010;65:5217–25.

[41] Kanjanapongkul K, Wongsasulak S, Yoovidhya T. Investigation and preventionof clogging during electrospinning of zein solution. J Appl Polym Sci2010;118:1821–9.

[42] Bikiaris DN. Solid dispersions, Part II: new strategies in manufacturingmethods for dissolution rate enhancement of poorly water-soluble drugs.Expert Opin Drug Deliv 2011;8:1663–80.

[43] Bikiaris DN. Solid dispersions, Part I: recent evolutions and futureopportunities in manufacturing methods for dissolution rate enhancementof poorly water-soluble drugs. Expert Opin Drug Deliv 2011;8:1501–19.

[44] Dou Y, Lin K, Chang J. Polymer nanocomposites with controllable distributionand arrangement of inorganic nanocomponents. Nanoscale 2011;3:1508–11.

[45] Li D, Xia Y. Electrospinning of nanofibers: reinventing the wheel? Adv Mater2004;16:1151–70.

[46] Peppas NA. Analysis of Fickian and non-Fickian drug release from polymers.Pharm Acta Hel 1985;60:110–1.

5672 D.G. Yu et al. / Acta Biomaterialia 9 (2013) 5665–5672