Synthesis of Dual-Size Cellulose ... - ACS Publications

May 30, 2017 - Synthesis of dual-size nanofibers using one-step electrospinning was ... Aerosol filtration measurements show that this multilayer air ...
0 downloads 0 Views 3MB Size
Article pubs.acs.org/Langmuir

Synthesis of Dual-Size Cellulose−Polyvinylpyrrolidone Nanofiber Composites via One-Step Electrospinning Method for HighPerformance Air Filter Ratna Balgis,*,† Hiroyuki Murata,† Yohsuke Goi,‡ Takashi Ogi,*,† Kikuo Okuyama,† and Li Bao§

Downloaded via UNIV OF WINNIPEG on June 22, 2018 at 15:26:40 (UTC). See https://pubs.acs.org/sharingguidelines for options on how to legitimately share published articles.



Department of Chemical Engineering, Graduate School of Engineering, Hiroshima University, 1-4-1 Kagamiyama, Higashi-Hiroshima 739-8527, Japan ‡ DKS Co., Ltd., 5 Ogawara-cho, Kisshoin, Minami-ku, Kyoto 601-8391, Japan § Nippon Muki Co., Ltd., Nisshin Ueno Bldg, 5-1-5 Higashi-Ueno, Tokyo 110-0015, Japan S Supporting Information *

ABSTRACT: Dual-size nanofibers consisting of a random mixture of nano- and submicron-size nanofibers are promising structures for specific applications such as air filters because of their increased specific surface area and low pressure drop. Synthesis of dual-size nanofibers using one-step electrospinning was reported here for the first time. The formation of well-mixed nano- and submicron-size cellulose−polyvinylpyrrolidone nanofiber composites was accomplished utilizing the physical properties of TEMPO-oxidized cellulose nanofibers (i.e., high thixotropy and high magnitude of zeta potential) and tuning the charge of the polymer jet, which influences the formation and shape of Taylor cone, and Coulombic explosion. The dual-size nanofibers were then spun on the surface of a HEPA filter to obtain a multilayer air filter. Aerosol filtration measurements show that this multilayer air filter has an incredibly high performance, shown by the high quality factor (Qf), 0.117 Pa−1, which is 10 times the Qf of commercial HEPA filters.

1. INTRODUCTION Hierarchically structured nanofibers are gaining attention in the fields of energy and the environment, particularly for air filter and fluid membrane applications.1−3 High-durability filters or membranes with high surface area and low pressure drop are essential for such applications. Synthesizing nanofibers with diameters less than 100 nm is one of the methods used to address these issues that is becoming increasingly common. This strategy has been proven to increase the performance of the air filter by allowing a higher surface area to trap more dust particles and introducing the presence of slip flow at the same time.4,5 However, the durability is still questionable because, in general, nanofibers create high-density filters, in which the pressure drop may increase rapidly, even with the presence of slip flow phenomena. The addition of either beads or submicron-size fibers among the nanofibers is a promising way to decrease the density of nanofiber layers.2,6 Synthesis of beaded fibers has been successfully demonstrated via one-step electrospinning by controlling the viscosity of the polymer precursor lowers than critical value, i.e., minimum viscosity that is required to produce straight nanofiber. This critical value is unique for each polymer.6 The presence of beads will create a space or pores on the filter mat, which will maintain the pressure drop of the filter.7 However, the presence of those pores decreases the particle collection ability of the filter, while the addition of submicronsize fibers to the nanofiber mats promises better filter performance and lower pressure drop. Two-step electro© 2017 American Chemical Society

spinning is a common method used to obtain these dual-size fibers. Typically, submicron-size fibers are spun on the surface of previously spun nanofibers.2 Therefore, this process is timeconsuming, and it is also difficult to obtain a good and random mixture of nano- and submicron-size fibers. An appropriate filter design to obtain high particle collection while maintaining the pressure drop at a low level remains a great challenge. Furthermore, a facile method to obtain the optimum morphology is also necessary. To date, most studies of the synthesis of nanofibers have used Newtonian fluid-type polymer sources. Salt addition is one method that has been successfully used to decrease the size of fibers, even though it may not directly affect the rheological properties of the polymer fluid.4,5,7−9 Few studies have investigated the role of the rheological properties of the polymer precursors, and the role of fluid electrics, including the zeta potential, has not been considered. This is also an important factor in the formation of a Taylor cone jet, as it induces tangential stress on the surface of the jet.10 The final morphology, including the size of the fiber, is affected by the shape and the stability of the Taylor cone.11 In previous studies, several types of fibers with various morphologies and diameters have been successfully synthesized by adjusting the viscosity and electrical conductivity of polymer precursors to control the Received: April 7, 2017 Revised: May 15, 2017 Published: May 30, 2017 6127

DOI: 10.1021/acs.langmuir.7b01193 Langmuir 2017, 33, 6127−6134

Article

Langmuir shape of the Taylor cone jet.12,13 The results show that the morphology of the fibers is highly dependent on the rheological properties of the polymer precursor. Therefore, one-step formation of dual-size nano- and submicron-size fiber composites might be accomplished by tuning the rheological properties of the polymer precursor and the formation of the Taylor cone jet. The use of biodegradable materials such as cellulose nanofibers as the polymer precursor has opened many opportunities for green technology and medical applications.14−16 Furthermore, cellulose nanofibers, which are a non-Newtonian fluid, have a high thixotropy and magnitude of the zeta potential. These properties may influence the rheology of the polymer precursor mixture and affect the final morphology of the fiber. In this study, we introduce one-step synthesis of dual-size nanofibers by tuning the rheological properties and utilizing the high difference in the zeta potential of the polymer precursors. We clarified the formation phenomena of dual-size nanofibers during the electrospinning process. Filter performance measurements were also investigated to evaluate the effect of dual-size formation on the particle collection effectivity and pressure drop.

Figure 1. Schematic diagram of the experimental setup. gastight syringe 1000 μm, Hamilton, U.S.A.) with a needle size of 27G, a syringe pump (PHD 2000, Harvard Apparatus, Holliston, MA), two high-voltage generators with positive and negative polarities (HER30P1, Matsusada Precision Inc., Shiga, Japan), a temperature controller (PAU-300S-HC, Apiste Corp., Osaka, Japan), a heater (HLT-61, Hitachi Ltd., Tokyo, Japan), a chamber, and an aluminum collector disk with diameter of 100 mm and thickness of 2.5 mm. Typically, a positive voltage of 4 kV was applied to the syringe needle to obtain a stable liquid jet. A counter voltage of −8 kV was applied to the aluminum collector plate. To evaluate the fibrillation mechanism, experiments with a reverse voltage on the syringe needle and counter voltage were also performed. The syringe position was set at a fixed distance (10 cm) from the collector plate. 2.2. Precursor, Nanofiber, and Aerosol Filtration Characterization. The zeta potentials of aqueous solutions containing CNFs, PVP, or both were determined using a zetasizer (Zetasizer Nano ZSP, Malvern Instruments Ltd., Malvern, U.K.). The viscosity, electrical conductivity, and surface tension of the aqueous solutions were evaluated using a Brookfield DV-III rheometer (Brookfield, Middleboro, MA), an electroconductive meter (WM-50EG, DKK-TOA Corp., Tokyo, Japan), and an automatic surface tensiometer (CBVP-2, Kyowa Interface Science Co., Ltd., Saitama, Japan), respectively. Each measurement was done three times; the average value was taken and shown as the results, and an error bar was also added to show the deviation of those measurements from the average value. The morphology of the spun nanofiber composites was observed using field-emission scanning electron microscopy (SEM; S-5000, 20 kV, Hitachi High-Tech. Corp., Tokyo, Japan). For aerosol filtration measurements, the nanofiber composites were then spun on the surface of a base filter to obtain multilayer filters. A circular microfiber mat [Nippon Muki Co., Ltd., Tokyo, Japan, nonwoven fabric, bicomponent polypropylene with polyester fibers (PP/PET), basis weight of 0.16 mg m−2, and thickness of 0.58 mm] with a diameter of 10 cm was used as the base filter/foundation. The performance of the prepared multilayer filters was evaluated using atmospheric dust particles with sizes of 0.3−0.5 μm. Detailed measurement procedures have been explained elsewhere.5

2. EXPERIMENTAL SECTION 2.1. One-Step Electrospinning Process To Synthesize DualSize Nanofiber Composites. A precursor solution containing 2,2,6,6-tetramethylpiperidine-1-oxyl radical (TEMPO)-oxidized cellulose nanofibers (CNFs; water-based cellulose gel at 2 wt % (Supporting Information, Figure SI1); Dai-ichi Kogyo Seiyaku Co., Ltd., Kyoto, Japan) and polyvinylpyrrolidone (PVP; Mw = 1300 kDa; Sigma-Aldrich, St. Louis, MO) as a polymer source and ethanol−water (30 vol % of EtOH) as a solvent was used to prepare nanofiber composites via electrospinning. Spinning conditions to obtain various nanofiber composites are shown in Table 1. PVP was heated at 80 °C

Table 1. Spinning Conditions for Nanofiber Composites sample CNFPVP 1 CNFPVP 2 CNFPVP 3 CNFPVP 4 CNFPVP 5 CNFPVP 6 CNFPVP 7 CNFPVP 8 CNFPVP 9 CNFPVP10

PVP conc. (wt %)

CNF conc. (wt %)

temp. (oC)

humidity (%)

voltage (kV)

8

0

41

13

10.0

8

0.05

41

13

10.2

8

0.1

41

12

11.5

8

0.15

41

12

12.5

8

0.2

41

13

12.4

0

0.2

40

13

12.0

2

0.2

40

14

12.0

4

0.2

40

14

11.1

6

0.2

40

14

12.4

10

0.2

40

13

12.0

3. RESULTS AND DISCUSSION 3.1. Morphologies of Nanofibers Affected by PVP Concentration. The spinning ability of polymers is affected by the combination of their physical properties. Therefore, it is important to evaluate the physical properties of the polymer precursor in detail to hypothesize the possible formation of the Taylor cone. The electrical conductivity, surface tension, and viscosity of the PVP precursor at various concentrations are shown in Figure 2. Figure 2a shows that the values of surface tension were relatively low compared with that of water-based polymer precursors and stable for various concentrations of PVP at approximately 35 mN m−1. This result occurred because the precursors were made using the same mixture of ethanol− water solvent (30 vol % of EtOH). Meanwhile, the electrical

for 1 h in an oven to reduce the moisture content prior to use. To obtain a well-dispersed precursor, a homogenizer (IKA T10 basic ULTRA-TURRAX homogenizer system, IKA Japan K.K., Osaka, Japan) was used to mix the precursor mixture. The precursor was then pumped with the flow rate set at 4 μL min−1. Figure 1 shows a schematic diagram of the experimental system used to produce the nanofiber composites. The system consists of a syringe (Hamilton 6128

DOI: 10.1021/acs.langmuir.7b01193 Langmuir 2017, 33, 6127−6134

Article

Langmuir

the cone will either form a beaded or a straight nanofiber, after the volatile solvents have evaporated. The increase in electrical force as a result of a higher polymer precursor concentration will finally exceed the surface tension force of the meniscus, creating a more uniform and stable jet, corresponding to the stability of the desired constant current.13,18,19 Furthermore, the viscosity of the polymer precursor is related to the number of polymer entanglements (overlaps between polymer chains) in the precursor solution, which affects the resistance of the jet to breakup in cone-jets of many polymer solutions.10,18 For concentrated precursors, evaporation of a small amount of solvent may lead to immediate skin formation. It is speculated that the solution jet ejected from the Taylor cone undergoes solidification before Rayleigh instability can take effect. Therefore, the fiber morphology was transformed from beaded into smooth and straight proportional to the increase in PVP concentration. The Df value increased to 130, 230, 260, and 450 nm for PVP concentrations of 6, 8, 10, and 12 wt %, respectively. The Lb and Db also increased, from 3.3 and 1.4 μm, respectively, for PVP 6 wt % to 3.8 and 2.3 μm, respectively, for PVP 8 wt %. 3.2. Nanostructure of Cellulose−PVP Nanofiber Composites. The results presented in the previous section show that the final morphology of the fiber is highly influenced by the physical properties of the polymer precursor. Appropriate adjustment of the precursor’s electrical conductivity and viscosity may be the most important key to tuning the morphology of the prepared fibers. Considering the high viscosity of the CNF gel, a medium concentration PVP of 8 wt % was selected for fabricating the nanofiber composite to keep the diameter of the nanofiber composite as small as possible. The physical properties of CNF−PVP dispersion solutions are shown in Figure 4. Similar to the trend of the physical properties of the PVP precursor solution, the addition of CNFs at various concentrations gave relatively similar values of surface tension, and the electrical conductivity values increased proportionally to the concentration of CNFs, as shown in Figure 4a. Interestingly, the viscosity measurement results show that the addition of CNFs into the PVP solution, even in a very small amount, changed the fluid flow properties of the polymer

Figure 2. Physical properties of PVP polymer precursors at various concentrations: (a) electrical conductivity and surface tension; (b) viscosity.

conductivities of the precursors increased proportionally to the PVP concentration, showing that the amount of dissociated ions increased. As expected, the viscosity of the polymer precursor increased proportionally to that of the PVP concentration from 4 to 12 wt %, as shown in Figure 2b. These values show a linear relationship with rotation speed, in which the viscosities of the PVP precursors are not affected by the rotation speed. The tensors that describe the viscous stress and the strain rate are related by a constant viscosity tensor that does not depend on the stress state and velocity of the flow. This result implies that these precursors were a Newtonian fluid. These three physical properties play equally important roles in determining the final morphology of the spun fibers. Figure 3a−e shows that the morphology of the spun PVP precursor was transformed from particles to beaded fibers to straight fibers. The high-magnification SEM image of beaded fiber is shown in Figure 3f. Here, the diameter of the straight fiber is noted as Df, while the diameter and length of the bead are Db and Lb, respectively. At a very low concentration of PVP precursor (4 wt %), the viscosity and electrical conductivity of the precursor were also low. This may cause the formation of an unstable liquid jet, leading to a polymer solution that breaks up into droplets; subsequent solvent evaporation results in the formation of microspherical polymer particles.17 Therefore, a spray process occurred instead of spinning. By increasing the polymer precursor concentration, the liquid jet emitted from

Figure 3. SEM images of spun precursor with PVP concentrations of (a) 4, (b) 6, (c) 8, (d) 10, and (e) 12 wt % and (f) high-resolution SEM image of PVP nanofibers with a concentration of 8 wt %. 6129

DOI: 10.1021/acs.langmuir.7b01193 Langmuir 2017, 33, 6127−6134

Article

Langmuir

stretching during the spinning process. This can be seen from the morphology of the bead, which was transformed into a doughnutlike shape. This may have happened because of the thixotropic effect, which may cause the enlargement of droplets during precursor elongation. Polymer particles tend to go to the droplet surface during solvent evaporation to minimize their surface pressure. The average bead diameters were relatively similar, at approximately 2.1, 2.6, 3.5, and 2.4 μm for samples CNF-PVP 1, 2, 3, 4, and 5, respectively. The low-magnification SEM images indicate that the fiber density decreased as the amount of CNFs increased. The obtained fibers also tended to form curly structures rather than straight at higher concentrations of CNFs because of the electrospinning jet speed mismatch, which leads to bending instability.22 Interestingly, it was also found that the nanofiber composites formed dual-size fibers, as depicted in the HR-SEM images of samples with CNF additions of 0.1 wt % and higher (Figure 5f). The average diameters of fibers fluctuated in a small range at approximately 270, 220, 280, and 230 nm for precursors containing 0.05, 0.1, 0.15, and 0.2 wt % CNFs, respectively, and the average diameters of small fibers were relatively similar at approximately 45 nm. The result that attracted most attention is the ratio of small nanofibers in the composite, which increased proportionally to CNF concentration and were very well-dispersed among the straight fibers. This work is the first report on the synthesis of dual-size fiber composites with nano- and microscale fibers fabricated with a one-step electrospinning process. To confirm the effect of CNFs on the formation of dual-size nanofiber composites, vice versa experiments were performed, where the concentration of CNFs in the precursor was kept at 0.2 wt % and the concentration of PVP was varied from 0 to 10 wt % (Table 1, samples CNF-PVP 5 and 6−10). Analysis of the physical properties of CNF−PVP precursors shows that the surface tensions of precursors containing different amounts of PVP were relatively stable at approximately 37 mN m−1. This result agrees with the values obtained from polymer precursors containing various concentrations of CNFs. The electrical conductivity of precursors linearly increased proportionally to the concentration of PVP, as shown in Figure 6a. Figure 6b

Figure 4. Physical properties of CNF−PVP polymer precursors at PVP concentration of 8 wt % and various concentrations of CNFs: (a) electrical conductivity and surface tension; (b) viscosity.

precursor into that of a non-Newtonian fluid. The CNF gel exhibited a tunable storage modulus G̀ of more than 5 orders of magnitude; thus, it has a high thixotropic property, and its effect on the fluid flow properties of the polymer precursor increased proportionally to the CNF concentration, which can be seen from the slope of the viscosity values taken at various rotation speeds.20,21 A steeper slope can be clearly seen when the CNF concentration increased. The addition of a very small amount of CNFs (0.05 wt %) showed nearly flat viscosity values at various rotation speeds. Figure 5a−e shows the transformation of nanofiber composite morphologies prepared with various concentrations of CNFs. Fibers prepared from only 8 wt % PVP solution showed a smooth straight surface with ellipselike beads in some parts. The average diameter of fibers, diameter of beads, and length of beads were approximately 230 nm, 3.8 μm, and 2.2 μm, respectively. The addition of CNFs affected the polymer

Figure 6. Physical properties of CNF-PVP polymer precursors at CNF concentration of 0.2 wt % and various concentrations of PVP: (a) electrical conductivity and surface tension; (b) viscosity.

shows that the average values of viscosity increased proportionally to the PVP concentration. All of the precursors show a steep slope of viscosity values taken at various rotation speeds, which is the result of the thixotropic effect of 0.2 wt % CNFs. SEM images of the spun CNF precursor solution in the absence of PVP show unstructured flakelike particles (Figure 7a). CNFs, even though they consist of several single nanofibers, have high numbers of hydrogen bonds (arising from surface hydroxyl groups), and the exceptionally high

Figure 5. SEM images of (a) CNF-PVP 1, (b) CNF-PVP 2, (c) CNFPVP 3, (d) CNF-PVP 4, and (e) CNF-PVP 5; (f) HR-SEM image of CNF-PVP 4. 6130

DOI: 10.1021/acs.langmuir.7b01193 Langmuir 2017, 33, 6127−6134

Article

Langmuir

Figure 7. SEM images of (a) CNF-PVP 6, (b) CNF-PVP 7, (c) CNF-PVP 8, (d) CNF-PVP 9, (e) CNF-PVP 5, and (f) CNF-PVP 10.

magnitude of their zeta potential creates a highly entangled cellulose element that results in a gel.23,24 This strong entanglement cannot be easily broken up by the electrospinning method; therefore, the agglomerated CNFs cannot be fibrillated into continuous single nanofibers, as shown in Figure 7a. The addition of a low concentration of PVP influenced the morphology of the sprayed particles, from an unstructured morphology into doughnutlike particles with an average size of 0.9 μm, even though the formation of fibers still cannot be observed, as shown in Figure 7b. This phenomenon shows that self-assembly between PVP and CNF occurred during droplet formation in the electrospinning process. Figure 7c shows that at a higher concentration of PVP (4 wt %), very thin nanofibers with an average size of 34 nm were spread among the 1.5-μm particles, creating a spider web. The stretching process was easier when the added PVP concentration was approximately 6 wt % and up. Figure 7d−f shows that beaded fibers were successfully formed because a sufficient amount of PVP was added to stretch the polymer precursor. The increase in PVP concentration slightly increased the precursor viscosity and created a larger and more unstable cone jet. This can be seen from the morphology of the prepared fibers, with larger fibers of 100, 220, and 310 nm formed for CNF-PVP 9, 5, and 10, respectively. Dual-size nanofibers were clearly observed from high concentration of PVP precursor of CNF-PVP 5 and 10 with small nanofiber sizes of approximately 47 and 62 nm, respectively. The small nanofibers for CNF-PVP 9 cannot be easily seen, even though they existed with average diameters of 44 nm. The extensional viscosity of the jet while in flight to the target is undoubtedly very influential in governing the stretching induced in the jet, which in turn affects the final diameter of the fibers. Hence, the amount of small nanofibers and the size difference were not so high in this sample, and they were therefore well-blended with the larger fibers. The results above show an interesting fact: addition of a nonNewtonian material into the polymer precursor not only leads to the synthesis of interesting morphologies but also affects the size of the main nanofibers. Figure 8 shows a plot of viscosity against average fiber diameter and confirms that the thixotropic effect is higher than that of the polymer viscosity. The error bars in Figure 8 show that the diameter of nanofiber has a slight deviation when sprayed at different batches even though using

Figure 8. Effect of polymer precursor viscosity on the size distribution of (a) large nanofibers and (b) small nanofibers.

the same precursor. Typically, the viscosity depends on the average fiber diameter.25 Linear regression of the data corresponding to log viscosity (η)− log average fiber diameter (Df) gives the following: DfS ∼ η0.59 and DfL ∼ η0.04 for polymer precursors containing various concentrations of PVP (Figure 8a). The nanofiber diameter was strongly dependent on the viscosity, with a power scaling of 0.59 and 0.04 for small and large nanofibers, respectively. An increase in the solution viscosity indicated a larger number of entanglement couplings, thereby generating larger electrospun fibers. Surprisingly, the addition of various concentrations of CNFs into 8 wt % PVP solution resulted in large fibers with nearly the same diameters, as shown in Figure 8b, even though the viscosity of the polymer precursors increased proportionally to that of the concentration of the added CNFs. 3.3. Formation Mechanism of Dual-Size Cellulose− PVP Nanofiber Composites. The formation of dual-size cellulose−PVP nanofiber composites via one-step electrospinning might be influenced by the relationship between the high zeta potential of the CNFs and the voltage applied during the spinning process, as shown in the scheme of Figure 9a. The spun polymer has a negative charge, which is primarily influenced by the zeta potential of cellulose, −90 mV. The application of a certain value and charge of high voltage to the spinning needle will create an electrical force. Here, +4 kV was applied to create a positively charged polymer jet from the syringe needle. During the spinning process, the COO− group of CNF containing a lot of COONa was attracted to the surface 6131

DOI: 10.1021/acs.langmuir.7b01193 Langmuir 2017, 33, 6127−6134

Article

Langmuir

and collector, respectively, the formation of two sizes of fibers was not observed. Only large nanofibers, with average diameters of 247 nm, were obtained. This result shows that the highly negative zeta potential of cellulose nanofibers tended to repulse them from the surface of the negatively charged needle during the spinning process. Even though Coulombic explosion usually happens on polymer jet, the extreme splitting phenomena did not occur. The Taylor cone is more stable than that of the reverse charge conditions. The high-voltage values applied to the needle were varied to confirm that the polymer jet splitting that led to the formation of two sizes of fibers was the effect of the zeta potential of the polymer precursor. Figure 10 shows that the Taylor cone stability was affected by the sign and value of the voltage applied to the needle. The application of a positive voltage to the needle may affect the stability of the Taylor cone, as explained above. The magnitude of the positive voltage further influenced the formation and stability of the Taylor cone. The polymer precursor tended to split in many directions when a high voltage was applied, resulting in less clear dual-size formation, even though the fiber size tended to decrease proportionally to the magnitude of the positive voltage, resulting in relatively monodisperse nanofibers. Conversely, the application of various magnitudes of negative voltage on the needle did not influence the size distribution of the obtained nanofibers. Single-size nanofibers with diameters of approximately 100 nm were obtained with various magnitudes of negative voltage, showing there was no zeta potential effect on polymer jet splitting. In this case, this process occurred mostly as a result of Coulombic explosion. 3.4. Aerosol Filtration Performance of Dual-Size Cellulose−PVP Nanofiber Composites. Here, aerosol particle permeation tests and pressure drop measurements were performed for multilayer air filters composed of a straight microfiber mat and dual-size cellulose−PVP nanofiber composites (Supporting Information, Figure SI2), using atmospheric dust particles with diameters of 0.3−0.5 μm. Because the inhomogeneity of the multilayer filters was quite high, the basis weights of filters were used as a parameter in the measurements and were determined to be 0.16 mg m−2.

Figure 9. (a) Model of dual-size nanofiber composite formation affected by CNF zeta potential; SEM images of spun precursor containing 8 wt % PVP and 0.2 wt % CNFs with various needle and collector voltages: (b) 4 and −8 kV, respectively, and (c) −8 and 4 kV, respectively.

of the positively charged jet and occupied most of the surface content of the composite jet. The high content of cellulose supported by the thixotropic effect on the surface of the jet promotes continuous and unstable polymer jet splitting, leading to the formation of two sizes of fibers. These fibers were than attracted to the surface of the collector, which was negatively charged at −8 kV. On the contrary, when the high-voltage supply was reversed, with −8 and +4 kV applied to the syringe

Figure 10. SEM images of spun precursor containing 8 wt % PVP and 0.2 wt % CNFs with various needle and collector voltages: (a) 4 and −8 kV, (b) 6 and −6 kV, (c) 8 kV and ground, (d) −4 and 8 kV, (e) −6 and 6 kV, and (f) −8 kV and ground. 6132

DOI: 10.1021/acs.langmuir.7b01193 Langmuir 2017, 33, 6127−6134

Article

Langmuir ORCID

Filtration performance is highly affected by the morphology of the filter. In the present work, we found that the presence of dual-size nanofiber composites improved the aerosol filtration performance, with a Qf value of 0.117 Pa−1. This value is several times greater than that of previous results, including commercial HEPA filters (Supporting Information, Table SI1).2,5,7 The collection efficiency of aerosol particles is very high, up to 86.4%, which is probably due to the presence of small nanofibers among the large nanofibers. Nanofiber greatly increased the probability of aerosol particle deposition because of the large surface area-to-volume ratio compared to that of microfibers.26 Interestingly, this packing filter still can maintain a pressure drop with a relatively low value of 17 Pa, probably because of the random distribution of dual-size nanofibers. The presence of large nanofiber helps to collect part of the aerosol particle, thus reducing the loading and curtailing the pressure drop elevation.27 The insertion of polymeric beads into nanofibers pack resulted in a physical separation of the nanofiber layers. The increase in the distance between nanofibers decreases the volume fraction of the structures; thus, filter performance increases as air permeability increases, but the pressure drop is nearly linear.6 Furthermore, the presence of nanofibers which have smaller diameter than the average mean free path show large Knudsen number and transforms the air flow around the fiber to become transitional flow, creating a phenomenon called “slip flow”. The fluid speed on the fiber surface will be increased and resulted a lower pressure drop.7,28 Here, the pressure drop of nanofiber is not only affected by slip flow but also by inhomogeneity of fiber packing because of the dual-size nanofiber composites.29,30

Takashi Ogi: 0000-0002-2026-3743 Notes

The authors declare no competing financial interest.



ACKNOWLEDGMENTS This research was supported by a Grant-in-Aid for Young Scientists B (15K182570A) and Grant-in-Aid for Young Scientists A (26709061) sponsored by the Ministry of Education, Culture, Sports, Science, and Technology of Japan.



(1) Wang, Z.; Zhao, C.; Pan, Z. Porous Bead-on-String Poly(Lactic Acid) Fibrous Membranes for Air Filtration. J. Colloid Interface Sci. 2015, 441, 121−129. (2) Wang, N.; Si, Y. S.; Wang, N.; Sun, G.; El-Newehy, M.; Al-Deyab, S. S.; Ding, B. Multilevel Structured Polyacrylonitrile/Silica Nanofibrous Membranes for High-Performance Air Filtration. Sep. Purif. Technol. 2014, 126, 44−51. (3) Zongjie, L.; Weimin, K.; Huihui, Z.; Hu, M.; Jingge, J.; Deng, N.; Cheng, B. Fabrication of a Polyvinylidene Fluoride Tree-Like Nanofiber Web for Ultra High Performance Air Filtration. RSC Adv. 2016, 6, 91243−91249. (4) Zhao, X.; Wang, S.; Yin, X.; Yu, J.; Ding, B. Slip-Effect Functional Air Filter for Efficient Purification of PM2.5. Sci. Rep. 2016, 6, 35472. (5) Balgis, R.; Kartikowati, C. W.; Ogi, T.; Gradon, L.; Bao, L.; Seki, K.; Okuyama, K. Synthesis and Evaluation of Straight and Bead-Free Nanofibers for Improved Aerosol Filtration. Chem. Eng. Sci. 2015, 137, 947−954. (6) Yun, K. M.; Suryamas, A. B.; Iskandar, F.; Bao, L.; Niinuma, H.; Okuyama, K. Morphology Optimization of Polymer Nanofiber for Applications in Aerosol Particle Filtration. Sep. Purif. Technol. 2010, 75, 340−345. (7) Bao, L.; Seki, K.; Niinuma, H.; Otani, Y.; Balgis, R.; Ogi, T.; Gradon, L.; Okuyama, K. Verification of Slip Flow in Nanofiber Filter Media through Pressure Drop Measurement at Low-Pressure Conditions. Sep. Purif. Technol. 2016, 159, 100−107. (8) Li, N.; Qin, X. H.; Yang, E. L.; Wang, S. Y. Effect on Instability Section of PVA Electrospinning Nanofibers by Adding LiCl. Mater. Lett. 2008, 62, 1354−1384. (9) Moghe, A. K.; Hufenus, R.; Hudson, S. M.; Gupta, B. S. Effect of the Addition of a Fugitive Salt on Electrospinnability of Poly(εcaprolactone). Polymer 2009, 50, 3311−3318. (10) An, S.; Lee, M. W.; Kim, N. Y.; Lee, C.; Al-Deyab, S. S.; James, S. C.; Yoon, S. S. Effect of Viscosity, Electrical Conductivity, and Surface Tension on Direct Current-Pulsed Drop-on-Demand Electrohydrodynamic Printing Frequency. Appl. Phys. Lett. 2014, 105, 214102. (11) Shin, Y. M.; Hohman, M. M.; Brenner, M. P.; Rutledge, G. C. Experimental Characterization of Electrospinning: The Electrically Forced Jet and Instabilities. Polymer 2001, 42, 9955−9967. (12) Deitzel, J. M.; Kleinmeyer, J.; Harris, D.; Beck-Tan, N. C. The Effect of Processing Variables on the Morphology of Electrospun Nanofibers and Textiles. Polymer 2001, 42, 261−272. (13) Subbiah, T.; Bhat, G. S.; Tock, R. W.; Parameswaran, S.; Ramkumar, S. S. Electrospinning of Nanofibers. J. Appl. Polym. Sci. 2005, 96, 557−569. (14) Duan, J.; Gong, S.; Gao, Y.; Xie, X.; Jiang, L.; Cheng, Q. Bioinspired Ternary Artificial Nacre Nanocomposites Based on Reduced Graphene Oxide and Nanofibrillar Cellulose. ACS Appl. Mater. Interfaces 2016, 8, 10545−10550. (15) Nemoto, J.; Saito, T.; Isogai, A. Simple Freeze-Drying Procedure for Producing Nanocellulose Aerogel-Containing, High-Performance Air Filters. ACS Appl. Mater. Interfaces 2015, 7, 19809−19815. (16) Vipin, A. K.; Fugetsu, B.; Sakata, I.; Isogai, A.; Endo, M.; Li, M.; Dresselhaus, M. S. Cellulose Nanofiber Backboned Prussian Blue Nanoparticles as Powerful Adsorbents for the Selective Elimination of Radioactive Cesium. Sci. Rep. 2016, 6, 37009.

4. CONCLUSION The synthesis of dual-size PVP−cellulose composite nanofibers via one-step electrospinning was comprehensively evaluated. The thixotropic property and high-magnitude zeta potential of CNFs determine the final morphology of the prepared CNF− PVP composite nanofibers. Furthermore, the presence of single- or dual-size nanofibers can be tuned by controlling the sign and magnitude of the needle voltage. The filtration performance of the dual-size PVP−cellulose nanofiber composites was very high, as demonstrated by the Qf value of 0.117 Pa−1, which is much better than those of other types and morphologies of filter. This result suggests that the presence of mixed sizes of nanofibers may allow better gas transport through the filter surface and better particle collection.



ASSOCIATED CONTENT

S Supporting Information *

The Supporting Information is available free of charge on the ACS Publications website at DOI: 10.1021/acs.langmuir.7b01193.



REFERENCES

SEM images and table of quality factor values (PDF)

AUTHOR INFORMATION

Corresponding Authors

*Phone: +81-82-424-3765. Fax: +81-82-424-7850. E-mail: [email protected]. *Phone: +81-82-424-7850. E-mail: ratna-balgis@hiroshima-u. ac.jp. 6133

DOI: 10.1021/acs.langmuir.7b01193 Langmuir 2017, 33, 6127−6134

Article

Langmuir (17) Yarin, A. L. Taylor Cone and Jetting from Liquid Droplets in Electrospinning of Nanofibers. J. Appl. Phys. 2001, 90, 4836−4846. (18) Bognitzki, M.; Czado, W.; Frese, T.; Schaper, A.; Hellwig, M.; Steinhart, M.; Greiner, A.; Wendorff, J. H. Nanostructured Fibers via Electrospinning. Adv. Mater. 2001, 13, 70−72. (19) Zander, N. E. Hierarchically Structured Electrospun Fibers. Polymers 2013, 5, 19−44. (20) Isogai, A.; Saito, T.; Fukuzumi, H. TEMPO-Oxidized Cellulose Nanofibers. Nanoscale 2011, 3, 71−85. (21) Tanaka, R.; Saito, T.; Ishii, D.; Isogai, A. Determination of Nanocellulose Fibril Length by Shear Viscosity Measurement. Cellulose 2014, 21, 1581−1589. (22) Cui, X.; Li, L.; Xu, J.; Xu, F. Fabrication of Continuous Aligned Polyvinylpyrrolidone Fibers via Electrospinning by Elimination of the Jet Bending Instability. J. Appl. Polym. Sci. 2010, 116, 3676−3681. (23) Isogai, A. Wood Nanocellulose: Fundamentals and Applications as New Bio-Based Nanomaterials. J. Wood Sci. 2013, 59, 449−459. (24) Missoum, K.; Belgacem, M. N.; Bras, J. Nanofibrillated Cellulose Surface Modification: A Review. Materials 2013, 6, 1745−1766. (25) Munir, M. M.; Suryamas, A. B.; Iskandar, F.; Okuyama, K. Scaling Law on Particle-to-Fiber Formation during Electrospinning. Polymer 2009, 50, 4935−4943. (26) Weiss, D.; Skrybeck, D.; Misslitz, H.; Nardini, D.; Kern, A.; Kreger, K.; Schmidt, H. W. Tailoring Supramolecular Nanofibers for Air Filtration Applications. ACS Appl. Mater. Interfaces 2016, 8, 14885−14892. (27) Leung, W. W. F.; Hung, C. H.; Yuen, P. T. Experimental Investigation on Continuous Filtration of Sub-Micron Aerosol by Filter Composed of Dual-Layers Including a Nanofiber Layer. Aerosol Sci. Technol. 2009, 43, 1174−1183. (28) Hung, C. H.; Leung, W. W. F. Filtration of Nano-Aerosol using Nanofiber Filter under Low Peclet Number and Transitional Flow Regime. Sep. Purif. Technol. 2011, 79, 34−42. (29) Przekop, R.; Gradon, L. Deposition and Filtration of Nanoparticles in the Composites of Nano- and Microsized Fibers. Aerosol Sci. Technol. 2008, 42, 483−493. (30) Yamada, S.; Seto, T.; Otani, Y. Influence of Filter Inhomogeneity on Air Filtration of Nanoparticles. Aerosol Air Qual. Res. 2011, 11, 155−160.

6134

DOI: 10.1021/acs.langmuir.7b01193 Langmuir 2017, 33, 6127−6134