polyacrylonitrile composite and porous nanofibers

IOP PUBLISHING

NANOTECHNOLOGY

Nanotechnology 19 (2008) 085605 (9pp)

doi:10.1088/0957-4484/19/8/085605

Preparation and characterization of silica nanoparticulate–polyacrylonitrile composite and porous nanofibers Liwen Ji1 , Carl Saquing2, Saad A Khan2 and Xiangwu Zhang1,3 1 Fiber and Polymer Science Program, Department of Textile Engineering, Chemistry and Science, North Carolina State University, Raleigh, NC 27695-8301, USA 2 Department of Chemical and Biomolecular Engineering, North Carolina State University, Raleigh, NC 27695-7905, USA

E-mail: xiangwu [email protected]

Received 26 September 2007, in final form 21 December 2007 Published 1 February 2008 Online at stacks.iop.org/Nano/19/085605 Abstract In this study, polyacrylonitrile (PAN) composite nanofibers containing different amounts of silica nanoparticulates have been obtained via electrospinning. The surface morphology, thermal properties and crystal structure of PAN/silica nanofibers are characterized using attenuated total reflection Fourier transform infrared (ATR-FTIR) spectroscopy, wide-angle x-ray diffraction (WAXD), scanning electron microscopy (SEM), transmission electron microscopy (TEM), and differential scanning calorimetry (DSC). The results indicate that the addition of silica nanoparticulates affects the structure and properties of the nanofibers. In addition to PAN/silica composite nanofibers, porous PAN nanofibers have been prepared by selective removal of the silica component from PAN/silica composite nanofibers using hydrofluoric (HF) acid. ATR-FTIR and thermal gravimetric analysis (TGA) experiments validate the removal of silica nanoparticulates by HF acid, whereas SEM and TEM results reveal that the porous nanofibers obtained from composite fibers with higher silica contents exhibited more nonuniform surface morphology. The Brunauer–Emmett–Teller (BET) surface area of porous PAN nanofibers made from PAN/silica (5 wt%) composite precursors is higher than that of pure nonporous PAN nanofibers.

ultra-fine composite nanofibers are notable for their small diameter, large area-to-volume ratio, and small pore size, which not only improve the material properties, but also create new characteristics not observed in bulk materials [2]. Because of these reasons, much progress has recently been made in the production of composite nanofibers through electrospinning inorganic nanoparticulate dispersions in polymer solutions, such as PAN/TiO2 , poly(vinyl alcohol)/Al2 O3 , and PAN/Ag [6–8]. Among various composite nanofibers, silica-filled nanofibers are gaining attention because silica nanoparticulates can provide nanofibers with improved properties such as hydrophilicity, toughness, and permeability [2, 9–11], and these composite nanofibers have potential application in mesoscopic research, nanodevices, optoelectronics devices, and chromatographic supports with high adsorption capacities [12]. Significant efforts have recently

1. Introduction Composite nanofibers, consisting of nanoscale inorganic fillers and polymer matrix, combine the advantages of both the polymer materials, such as light weight, flexibility, and good moldability, and the inorganic materials, such as high strength, heat stability, and chemical resistance [1, 2]. As a result, these composite nanofibers can have enhanced mechanical, electrical, optical, thermal and magnetic properties without losing transparency, and they can be excellent candidates for many multifunctional applications that include membranes, biomedical devices, and energy conversion and storage systems [2–4]. Composite nanofibers can be prepared directly from inorganic nanoparticulate dispersions in polymer solutions via electrospinning [2, 5]. The electrospun 3 Author to whom any correspondence should be addressed.

0957-4484/08/085605+09$30.00

1

© 2008 IOP Publishing Ltd Printed in the UK

Nanotechnology 19 (2008) 085605

L Ji et al

been made to use sol–gel processing and electrospinning techniques to prepare poly(vinyl alcohol)/silica [12] and poly(ethylene oxide) or poly(acrylamide)/silica composite nanofibers [13]. More recently, Sun et al also prepared electrospun poly(2-hydroxyethyl methacrylate) (PHEMA)/silica composite nanofibers, in which nanosized silica particles were covalently embedded inside PHEMA [14]. Instead of the in situ synthesis of silica nanoparticulates, it is more convenient to obtain silica-filled composite nanofibers by directly adding commercially available colloidal silicas into a polymer solution and perform the electrospinning from the silica/polymer dispersion [15, 16]. However, colloidal silicas are typically in the form of water dispersions, and water has to be removed from the electrospinning silica/polymer suspensions. In addition to colloidal silicas, silica-filled composite nanofibers can also be obtained by directly adding fumed silicas into a polymer solution before electrospinning. Unlike colloidal silicas, fumed silicas are prepared by the flame hydrolysis of SiCl4 , resulting in nanosized particles that are in the powder form [17, 18]. These fumed silica nanoparticulates are relatively inexpensive (compared with colloidal silicas) and are also ultra-pure, and as a result, they are excellent candidates for fabricating silicafilled composite nanofibers. In addition to composite nanofibers, porous structure nanofibers have been of significant interest because of their ultrahigh surface areas as compared to nonporous nanofibers with similar diameters. Porous nanofibers with high surface area may have many potential applications in which web (or sheet) morphology and controlled pore structure are strongly required [19, 20]. A number of groups have tried to produce porous nanofibers, with the most common approach involving electrospinning two immiscible polymers into nanofibers, followed by selective removal of one of the polymer components to generate pore structures [21, 22]. Porous nanofibers were also obtained by exploiting the phase separation process of different polymers [23, 24]. Other researchers have reported the production of porous nanofibers through calcination [25, 26]. More recently, Xia et al took advantage of the phase separation between residual solvent and polymer by immersing the collector in a cryogenic liquid and obtaining highly porous fibers with increased surface areas [27]. In our laboratory, we have fabricated porous nanofibers by removing silica nanoparticulates from polymer/silica composite nanofibers. One advantage of this method is that the pore structures can be controlled by selectively adjusting the size and shape of the silica nanoparticulates. In addition, the fast solubilization kinetics of silica nanoparticulates increases the pore generation speed and reduces the cost for fabricating porous nanofibers with controlled structures. In the present study, PAN solutions in N,N -dimethylformamide (DMF) containing silica nanoparticulates were first prepared, and then electrospun into PAN/silica composite nanofibers. We selected PAN as the polymer precursor because of its high dielectric constant, desirable for electrospinning. In addition, this polymer provides well-known routes to carbon fibers, which can be used in many applications such as sensors, catalyst supports, and batteries [28]. In order to create a

larger surface area, we also converted the nanofiber structure from solid to porous form by removing the silica nanoparticulates from the as-spun composite nanofibers using hydrofluoric acid (HF) [29, 30]. The surface morphology, thermal properties, crystal structure and surface area of these composite and porous nanofibers were investigated.

2. Experimental materials and methods 2.1. Materials Polyacrylonitrile (PAN) (Mw = 150 000), N,N -dimethylformamide (DMF) and hydrofluoric (HF) acid were purchased from Aldrich Chemical Company Inc. (Milwaukee, WI). Fumed silica nanoparticulates (Aerosil 380, average diameter = 7 nm; the diameter of the agglomerates or cluster is about 50 nm) were supplied by Degussa (D¨usseldorf, Germany) [17]. All these reagents were used without further purification. 2.2. Preparation of PAN and PAN/silica solutions PAN was dissolved in DMF at 60 ◦ C and the concentration was fixed at 5 wt%. A controlled amount of silica nanoparticulates (0, 1, 2, and 5 wt% to PAN) were dispersed into PAN solution in DMF. Mechanical stirring was applied for at least 24 h at 60 ◦ C in order to obtain homogeneous silicadispersed PAN solutions. These polymer dispersions were used for electrospinning. The silica contents in the asspun nanofibers were confirmed using a TA Hi-Res thermal gravimetric analyzer. 2.3. Electrospinning A variable high voltage power supply (Gamma ES40P20W/DAM) was used to provide high voltages in the range 14–18 kV. Electrospinning solutions were loaded in a 10 ml syringe to which a stainless steel capillary metal-hub needle was attached. The inside diameter of the metal needle was 0.11 mm. The positive electrode of the high voltage power supply was connected to the needle tip. The grounded electrode was connected to a metallic collector wrapped with aluminum foil. Under high voltage, a polymer jet was ejected and accelerated toward the counter electrode. The solvent evaporated rapidly as the high surface area jet traveled to the target. Dry fibers were accumulated on the collection screen and collected as a fibrous mat. The tip–collector distance and flow rate were fixed at 15 cm and 0.5 ml h−1 , respectively. The evolution of the jet during electrospinning was monitored using a CCD camera equipped with a Carl Zeiss monocular lens (Atlanta, USA). 2.4. Preparation of porous PAN nanofibers Porous nanofibers were prepared by placing PAN/silica composite nanofibers in a 35 wt% HF solution for 48 h, followed by washing with deionized water several times and vacuum drying. 2


L Ji et al

(a)

(b)

(c)

(d)

Figure 1. SEM images of PAN/silica composite nanofibers with different silica contents. (a) 0 wt% (pure PAN), (b) 1 wt%, (c) 2 wt%, and (d) 5 wt%. Electrospinning voltage: 14 kV.

2.5. Morphology

2.8. Crystal structure

The morphology and diameter of PAN/silica composite and porous PAN nanofibers were evaluated using scanning electron microscopy (JEOL JSM-6360LV FESEM at 15 kV and FEI XL30 SEM at 5 kV). Electrospun samples were coated with ˚ thickness by a KAu/Pd layers of approximately 100 A 550X sputter coater to reduce charging. Transmission electron microscopy (Hitachi HF-2000 TEM) was also used to study the structure of composite and porous nanofibers placed on 200mesh carbon-coated copper grids with an accelerating voltage of 200 kV.

Wide-angle x-ray diffraction (WAXD) analysis was performed with a Philips XLF ATPS XRD 100 diffractometer using Cu Kα radiation. The operating voltage and current were 40.0 kV and 60.0 mA, respectively. A crystal-monochromated collection system was used to acquire the diffractogram at 0.02◦ intervals at a rate of 1 s/step. Peak positions were determined by the APD 1700 (Version 4) software. 2.9. BET measurements Surface area analysis was carried out using the Brunauer– Emmett–Teller (BET) nitrogen adsorption method. The samples were degassed under flowing UHP grade nitrogen for 2 h at a temperature of 100 ◦ C. Nitrogen gas adsorption measurements were taken at 0.05, 0.1, 0.15, 0.2, and 0.25 of saturation pressure using a Micromeritics Gemini 2360 instrument capable of measuring surface area from 0.01 m2 g−1 and higher. The free space in the analysis tube was measured by the helium method, and five pressure points were used to calculate the BET surface area.

2.6. FTIR spectroscopy Attenuated total reflection Fourier transform infrared (ATRFTIR) spectra were collected from a Nicolet 560 FTIR spectrometer in the range 3700–700 cm−1 at room temperature. At least 32 scans were conducted to achieve an adequate signalto-noise ratio. 2.7. Thermal analysis

3. Results and discussion

The thermal properties of electrospun fibers were measured from 25 to 380 ◦ C at a 10 ◦ C min−1 heating rate using a Perkin Elmer Diamond differential scanning calorimeter with intracooler in a nitrogen environment. A TA Hi-Res 2950 thermal gravimetric analyzer was used to determine the weight loss of PAN composite and porous nanofibers (after solvent evaporation) at a heating rate of 10 ◦ C min−1 between 25 and 800 ◦ C in air environment.

3.1. Morphology of PAN/silica composite nanofibers Figures 1–3 show SEM images of pure PAN and PAN/silica composite nanofibers electrospun from 5 wt% PAN solutions with varying amounts of silica nanoparticulates at three different applied voltages: 14, 16 and 18 kV. The SEM images in figure 1 show that the nanofibers obtained at 14 kV are 3


L Ji et al

(a)

(b)

(c)

(d)

Figure 2. SEM images of PAN/silica composite nanofibers with different silica contents in fibers. (a) 0 wt% (pure PAN), (b) 1 wt%, (c) 2 wt%, and (d) 5 wt%. Electrospinning voltage: 16 kV.

liquid jet is not stable and the resultant nanofibers do not have smooth surface morphology. However, a voltage of 18 kV is too high and the balance between the electrostatic repulsion, surface tension and viscoelastic force also cannot be maintained, thereby resulting in poor surface morphology. Figure 2 demonstrates that a voltage of 16 kV can provide a balance between the electrostatic repulsion, surface tension and viscoelastic force, and hence favors the stabilization of the liquid jet and the formation of smooth nanofiber surface morphology. Therefore, the most suitable voltage for the electrospinning of PAN/silica composite nanofibers is 16 kV, which was used in the succeeding experiments.

not taut. In addition, irregularities and beads are present on the surfaces of the nanofibers, especially those with higher silica contents (such as at 2 and 5 wt%). It can also be observed from figure 1 that the diameters of nanofibers with higher silica contents become nonuniform. The poor morphology of nanofibers with high silica contents are caused by the high solution viscosity, which must be overcome during electrospinning [4, 31]. Figure 3 reveals that the nanofibers produced at 18 kV also have uneven diameter, and serious irregularities, and that non-uniformities are produced at higher silica contents. The PAN/silica nanofibers electrospun at 16 kV are relatively taut and have relatively smooth surface morphology. The diameters of these composite nanofibers range from 200 to 300 nm (with an average diameter of 250 nm). Defects, such as beads or fibers with so-called ‘beads on a string’ morphology [24], can seldom be seen in PAN/silica nanofibers when the silica content is lower than 2 wt%. A small number of irregularities in the form of beaded segments on large fibers still appear in nanofibers with higher silica contents (2 and 5 wt%) and the diameters of these nanofibers also become nonuniform. However, the irregularities of composite nanofibers produced at 16 kV are not as severe as the irregularities of those produced at 14 and 18 kV. During the process of electrospinning, it is very important to achieve a balance between the electrostatic repulsion, surface tension, and viscoelastic force [4, 31]. When the applied voltage is 14 kV, the electric field is not strong enough to provide the necessary electrostatic repulsion to balance the surface tension and viscoelastic force, and hence the

3.2. ATR-FTIR of PAN/silica composite nanofibers ATR-FTIR spectra recorded in the spectral range of 3800– 700 cm−1 are presented in figure 4. It is seen that the relative intensities of the peaks around 1100 cm−1 for PAN/silica composite nanofibers are higher than those for pure PAN nanofibers. The absorption at 1100 cm−1 can be ascribed to the characteristic Si–O–Si asymmetric stretching vibration [32, 33]. The addition of silica nanoparticulates also broadens the peak at 3400 cm−1 , which is the characteristic band of O–H (Si–OH) groups. This may be associated to stretching vibrations involving the hydroxyl groups which result from intramolecular hydrogen bonds of the Si–OH or arising from isolated and hydrogen bonded Si–OH stretching vibrations and hydrogen bonded water [34]. The FTIR spectra also suggest that there may be an interaction between the nitrogen atom in PAN and the silanol groups in silica, which 4


L Ji et al

(a)

(b)

(c)

(d)

Figure 3. SEM images of PAN/silica composite nanofibers with different silica contents. (a) 0 wt% (pure PAN), (b) 1 wt%, (c) 2 wt%, and (d) 5 wt%. Electrospinning voltage: 18 kV.

Table 1. Thermal behavior of PAN/silica composite nanofibers. Silica Tg,PAN (wt%) (◦ C)

Onset of Tcyc,PAN (◦ C)

Tcyc,PAN Hcyc,PAN (◦ C) (J g−1 PAN)

0 1 2 5

287.2 289.7 290.2 293.8

290.1 291.3 293.3 296.4

106.1 107.1 110.3 112.2

a b

352.9 — 378.9 382.1

c d

will be further discussed in the following DSC characterization section.

3700

2700

1700

700

-1

Wavenumbers (cm )

Figure 4. ATR-FTIR spectra of PAN/silica composite nanofibers with different silica contents. (a) 0 wt% (pure PAN), (b) 1 wt%, (c) 2 wt%, and (d) 5 wt%.

3.3. DSC of PAN/silica composite nanofibers DSC thermograms of pure PAN and PAN/silica nanofibers are presented in figure 5. All four nanofiber samples exhibit a relatively large and sharp exothermic peak and a glass transition above 100 ◦ C. Table 1 shows the glass transition temperature (Tg ), exothermic peak temperature, exothermic onset temperature and heat (H ) of the exothermic peak. With increase in silica content, the glass transition and exothermic peak shift to higher temperatures. Adding silica nanoparticulates also increases the H of the exothermic peak. This peak can result from a combination of three principal reactions, namely dehydrogenation, instantaneous cyclization and crosslinking reactions, which are exothermic in nature [20, 35, 36]. Among these three reactions, the predominant process is the instantaneous cyclization of the nitrile groups into an extended conjugated ring system. The fast reaction of PAN nanofibers in nitrogen may be

due to the facile formation of free radicals on the nitrile groups and subsequent recombinations between the radicals intermolecularly or intramolecularly [37]. Increases in peak temperature and activation energy for PAN/silica composite nanofibers (table 1) may be caused by the inhibiting effect of silica nanoparticulates, which hinder the recombinations between the radicals [38]. From figure 5 and table 1, it can also be seen that pure PAN nanofibers have a glass transition temperature of about 106.1 ◦ C, while Tg of PAN/silica composite nanofibers (1, 2, 5 wt% silica) increases to 107.1, 110.3, 112.2 ◦ C, respectively. The formation of intermolecular interaction between PAN and silica nanoparticulates may explain this phenomenon [11, 39]. 5


L Ji et al

a a Intensity (a.u.)

Heat Flow Up (mW)

b

c

b

c d

0 50

100

150 200 250 Temperature (oC)

300

20 30 Degrees (2 theta)

40

50

350

Figure 6. WAXD spectra of PAN/silica composite nanofibers with different silica contents. (a) 0 wt% (pure PAN), (b) 1 wt%, and (c) 5 wt%.

a

Heat Flow up (mW)

10

a

b b c

1600 1500 1400 1300 1200 1100 1000 900 Wavenumbers (cm-1)

d 90

100

110 Temperature (oC)

120

800

700

Figure 7. ATR-FTIR spectra of PAN/silica (5 wt%) composite nanofiber before (a) and after (b) the removal of silica nanoparticulates. (This figure is in colour only in the electronic version)

Figure 5. DSC thermograms of PAN/silica composite nanofibers with different silica contents. (a) 0 wt% (pure PAN), (b) 1 wt%, (c) 2 wt%, and (d) 5 wt%. The bottom figure shows the locations of the glass transition temperatures.

test the validity of the HF treatment, ATR-FTIR spectroscopy and TGA were employed to determine qualitatively and quantitatively the presence of silica nanoparticulates before and after HF treatment. The FTIR spectrum of PAN/silica (5 wt%) composite nanofibers shows a very strong Si–O–Si asymmetric stretching vibrational mode peak at 1100 cm−1 and several symmetric Si–O–Si bond stretching vibrational mode peaks around 790 cm−1 (figure 7(a)). However, HFtreated nanofibers (figure 7(b)) show relatively low peaks at these wavenumbers, indicating the removal of silica nanoparticulates. Meanwhile, the TGA thermograms of PAN/silica (5 wt%) composite nanofibers give about 4.7% residual after reaching 800 ◦ C (figure 8(b)), but HF-treated nanofibers (figure 8(c)) show more than 99.5% weight loss, which is similar with that of pure PAN nanofibers (figure 8(a)). These results further indicate that silica nanoparticulates have been removed from composite nanofibers by using HF. The morphology of porous PAN nanofibers was characterized using SEM and TEM techniques. Figure 9 shows the SEM images of nanofibers treated with HF. It is seen that the surface of HF-treated pure PAN nanofibers is relatively smooth. This

3.4. WAXD of PAN/silica composite nanofibers The WAXD diffraction patterns show two different diffraction peaks at around 2θ = 7◦ and 17◦ for both pure PAN and PAN/silica nanofibers (figure 6). The 2θ = 17◦ peak is characteristic of the (200) crystal planes of PAN [40], and the 2θ = 10◦ peak indicates the formation of large unit cells in the as-spun PAN nanofibers [41–43]. With increase in silica content, the peak intensity at 7◦ becomes stronger, but the intensity at 17◦ decreases. These intensity changes further corroborate the interactions between PAN and silica nanoparticulates observed from the DSC and ATRFTIR analyses. As a result, the crystallinity of PAN is largely influenced by silica nanoparticulates in the PAN/silica composite nanofibers. 3.5. Morphology of porous PAN nanofibers In addition to composite nanofibers, ‘porous’ PAN nanofibers were also prepared by removing the silica component from PAN/silica composite precursor nanofibers using HF acid. To 6


L Ji et al

figure 10. It is seen that, before HF treatment, there are distinct clusters or agglomerates of silica nanoparticulates in composite nanofibers (encircled section in figure 10(a)); however, this kind of morphology disappears after HF treatment, indicating the removal of silica nanoparticulates. It must be mentioned that, since the electrons pass through both the PAN and the pores, the actual nanopores cannot be directly seen from the TEM image. However, comparing these TEM images with BET surface area results (see section 3.6), it can be concluded that porous structures are formed in nanofibers made from PAN/silica (5 wt%) composite precursor. Figure 11 shows the TEM image of a nanofiber obtained by HF-treating PAN/silica (1 wt%) precursor nanofiber. Distinct spheres or clusters on the highly nonuniform fiber surface also indicate the formation of ‘porous’ structures.

100

Weight Loss (%)

80 c

60 a 40

20 b 0

0

200

400 o Temperature ( C)

600

800

Figure 8. TGA thermograms of pure PAN (a), and PAN/silica (5 wt%) composite nanofibers before (b) and after (c) the removal of silica nanoparticulates.

3.6. BET surface area of porous PAN nanofibers

is because no silica nanoparticulates were removed from the precursor nanofibers and the surface morphology remains unchanged. However, with the increase of silica content in precursor nanofibers, the surface morphology of the nanofibers becomes rough and irregular after the HF treatment. Additionally, a number of cavities and valleys can be seen on HFtreated nanofibers obtained from PAN/silica composite precursors, which indicates that the silica nanoparticulates embedded in PAN/silica precursor nanofibers have been removed by HF. TEM images of PAN/silica (5 wt%) composite precursor nanofibers and their HF-treated nanofibers are shown in

Surface area measurements of porous nanofibers were carried out using the BET nitrogen adsorption method. Porous nanofibers obtained by HF-treating PAN/silica (5 wt%) composite nanofibers have a BET surface area of 23.61 m2 g−1 . In comparison, pure PAN nanofibers with similar diameters have a BET surface area of only 19.93 m2 g−1 . Therefore, the surface area of porous nanofibers increased by 20%, complementing the findings obtained from TEM and SEM analyses.

(a)

(b)

(c)

(d)

Figure 9. SEM images of porous PAN nanofibers prepared by removing silica nanoparticulates from PAN/silica precursor nanofibers with different silica contents. (a) 0 wt% (pure PAN), (b) 1 wt%, (c) 2 wt%, and (d) 5 wt%.

7


L Ji et al

(a)

(b)

Figure 10. TEM images of PAN/silica (5 wt%) composite nanofibers before (a) and after (b) the removal of silica nanoparticulates.

results demonstrate the formation of a porous structure in HF-treated nanofibers. The BET surface area of HF-treated PAN/silica nanofibers (with 5 wt% silica) is 20% higher than that of PAN nonporous nanofibers.

Acknowledgments This work was supported by the US National Science Foundation (No. 0555959) and NSF-STC (support grant (CHE-9876674). The authors thank Professor Peter S Fedkiw, Mr Chris Bonino, and Mr Andrew Loebl, of the Department of Chemical and Biomolecular Engineering at NC State University, for their discussions during the preparation of the paper. Liwen Ji would also like to thank Ms Birgit Andersen for her help with sample characterization.

References Figure 11. TEM image of a porous PAN nanofiber prepared by removing silica nanoparticulates from PAN/silica (1 wt%) composite precursor nanofiber.

[1] Dzenis Y 2004 Science 34 1917–9 [2] Sawicka K M and Gouma P 2006 J. Nanopart. Res. 8 769–81 [3] Wang Y Z, Li Y X, Yang S T, Zhang G L, An D M, Wang C, Yang Q B, Chen X S, Jing X B and Wei Y 2006 Nanotechnology 17 3304–7 [4] Li D and Xia Y N 2004 Adv. Mater. 16 1151–70 [5] Reneker D H and Chun I 1996 Nanotechnology 7 216–23 [6] Kedem S, Schmidt J, Paz Y and Cohen Y 2005 Langmuir 21 5600–4 [7] Dai H, Gong J, Kim H and Lee D 2002 Nanotechnology 13 674–7 [8] Wang Y Z, Yang Q B, Shan G Y, Wang C, Du J S, Wang S G, Li Y X, Chen X S, Jing X B and Wei Y 2005 Mater. Lett. 59 3046–9 [9] Gao J, Gao T and Sailor M J 2000 Appl. Phys. Lett. 77 901–3 [10] Nagel H, Metz A and Hezel R 2001 Sol. Energy Mater. Sol. Cells 65 71–7 [11] Rittigstein P, Priestley R D, Broadbelt L J and Torkelson J M 2007 Nat. Mater. 6 278–82 [12] Shao C, Kim H, Gong J and Lee D 2002 Nanotechnology 13 635–7

4. Conclusions PAN/silica composite nanofibers containing different amounts of silica nanoparticulates were prepared by electrospinning. SEM, ATR-FTIR, DSC and WAXD results indicate that the addition of silica nanoparticulates has a significant impact on the surface morphology, thermal properties and crystal structure of the composite nanofibers. Interaction between PAN and silica nanoparticulates also exists. Porous PAN nanofibers were also obtained by removing the silica nanoparticulates in PAN/silica composite nanofibers using HF. ATR-FTIR spectra and TGA thermograms show that the silica component has been removed from PAN/silica composite nanofibers after HF treatment. SEM and TEM 8


L Ji et al

[28] Kim C, Yang K S, Kojima M, Yoshida K, Kim Y J, Kim Y A and Endo M 2006 Adv. Funct. Mater. 16 2393–7 [29] Morinaga T, Ohkura M, Ohno K, Tsujii Y and Fukuda T 2007 Macromolecules 40 1159–64 [30] Fu D G, Zhao J P, Sun Y M, Kang E T and Neoh K G 2007 Macromolecules 40 2271–5 [31] Fong H, Chun I and Reneker D H 1999 Polymer 40 4585–92 [32] Ding B, Kim H, Kim C, Khil M and Park S 2003 Nanotechnology 14 532–7 [33] Lee S W, Kim Y U, Choi S S, Park T Y, Joo Y L and Lee S G 2007 Mater. Lett. 61 889–93 [34] Hajji P, David L, Gerard J F, Pascault J P and Vigier G 1999 J. Polym. Sci. B 37 3172–87 [35] Peebles L H 1994 Carbon Fibers: Formation, Structure, and Properties (Boca Raton, FL: CRC Press) [36] Kim J, Kim Y C, Ahn W and Kim C Y 1994 Polym. Eng. Sci. 33 1452–7 [37] Brandrup J and Peebles L H 1968 Macromolecules 1 59–63 [38] Ko T, Ting H and Lin C 1988 J. Appl. Polym. Sci. 35 631–40 [39] Arrighi V, McEwen I J, Qian H and Serrano Prieto M B 2003 Polymer 44 6259–66 [40] Hou H Q, Ge J J, Zeng J, Li Q, Reneker D H, Greiner A and Cheng S Z D 2005 Chem. Mater. 17 967–73 [41] Calleja B and Vonk C G 1989 X-Ray Scattering of Synthetic Polymers (Amsterdam: Elsevier) [42] Jung B, Ki Yoon J, Kim B and Rhee H W 2005 J. Membr. Sci. 246 67–76 [43] Bashir Z 1994 J. Polym. Sci. B 32 1115–28

[13] Lim J M, Yi G R, Moon J H, Heo C J and Yang S M 2007 Langmuir 23 7981–9 [14] Sun Z F 2005 PhD Dissertation Drexel University Philadelphia [15] Lim J, Moon J, Yi G, Heo C and Yang S 2006 Langmuir 22 3445–9 [16] Kanehata M, Ding B and Shiratori S 2007 Nanotechnology 18 315602 [17] Michael G and Ferch H 1998 Degussa Tech. Bull. Pigment 11 6–50 [18] Zhang X W 2007 J. Electrochem. Soc. 154 B322–6 [19] Bekyarova E, Murata V, Yudasaka M, Kasuya D, Iijima S, Tanaka H and Kaneko K 2003 J. Phys. Chem. B 107 4681–4 [20] Kim C, Jeong Y, Nhu-Ngoc B T, Yang K S, Kojima M, Kim Y A, Endo M and Lee J W 2007 Small 3 91–5 [21] Zhang Y Z, Feng Y, Huang Z M, Ramakrishna S and Lim C T 2006 Nanotechnology 17 901–8 [22] Zhang L and Hsieh Y L 2006 Nanotechnology 17 4416–23 [23] Casper C L, Stephens J S, Tassi N G, Chase D B and Rabolt J F 2004 Macromolecules 37 573–8 [24] Bognitzki M, Czado W, Frese T, Schaper A, Hellwig M, Steinhart M, Greiner A and Wendorff J H 2001 Adv. Mater. 13 70–2 [25] Zussman E, Yarin A L, Brazilevsky A V, Avrahami R and Feldman M 2006 Adv. Mater. 18 348–53 [26] Pang M, Li D, Shen L, Chen Y, Zheng Q and Wang H 2006 Langmuir 22 9368–74 [27] McCann J T, Marquez M and Xia Y N 2006 J. Am. Chem. Soc. 128 1436–7

9