Nickel Nanoparticles Entangled in Carbon Nanotubes: Novel Ink for

Jan 6, 2016 - We examine the role of sonication in creating an ink that contains the ... diameter and 0.5–2 μm length, US Research Nanomaterials In...
0 downloads 0 Views 3MB Size
Letter www.acsami.org

Nickel Nanoparticles Entangled in Carbon Nanotubes: Novel Ink for Nanotube Printing Abdel Rahman Abdel Fattah,† Tahereh Majdi,‡ Ahmed M. Abdalla,‡ Suvojit Ghosh,‡ and Ishwar K. Puri*,†,‡ †

Department of Mechanical Engineering and ‡Department of Engineering Physics McMaster University, 1280 Main Street West, Hamilton, Ontario L8S 4L7, Canada ABSTRACT: We report the serendipitous discovery of a rapid and inexpensive method to attach nanoscale magnetic chaperones to carbon nanotubes (CNTs). Nickel nanoparticles (NiNPs) become entangled in CNTs after both are dispersed in kerosene by sonication and form conjugates. An externally applied magnetic field manipulates the resulting CNTs-NiNP ink without NiNP separation, allowing us to print an embedded circuit in an elastomeric matrix and fabricate a strain gage and an oil sensor. The new method to print a circuit in a soft material using an NiNPCNT ink is more rapid and inexpensive than the complex physical and chemical means typically used to magnetize CNTs.

KEYWORDS: carbon nanotubes, magnetophoresis, magnetization, polymer nanocomposite, sensors

C

attracted by a permanent magnet (Figure 1a). This was in sharp contrast with our expectations, which were that the NiNPs would be attracted toward the magnet while the CNTs would not. A TEM image of the NiNP-CNT conjugated material, hereinafter referred to as NiCNTs, shows how the NiNPs are enmeshed in CNT bundles (Figure 1b). Although sonication physically entangles NiNPs in CNTs, the high surface energy of NiNPs is likely responsible for NiNPs clustering and πinteractions which bonds them to CNT surfaces and allows the NiNPs to chaperone CNTs by magnetophoresis. Application of a strong magnetic field gradient did not separate these NiNPs from the conjugated material. Although multiwalled CNTs were used in these experiments, we expect similar results with the use of single and double-walled CNTs. A Superconducting Quantum Interference Device (SQUID) magnetometer (Quantum Design) was used to determine the magnetic response of the NiCNTs at room temperature. They were found to possess a strong saturation magnetization Ms = 14.61 emu/g, as shown in Figure 1c. This is substantially higher than that of CNTs (Ms < 0.1 emu/g)23 and is consistent with the 33% weight fraction of NiNPs (Ms = 44.33 emu/g).24 The sample is mildly ferromagnetic, with a remanence of 1.72 emu/ g. Although the NiNPs are individually superparamagnetic, the mild ferromagnetism is attributed to the inter-NiNP magnetostatic interactions that are expected between adjacent particles that are in close proximity, yielding superferromagnetic behavior.25

arbon nanotubes (CNTs) are diamagnetic, i.e., they have very weak magnetic response. Magnetic nanoparticles (MNPs) can be tethered to CNTs1−5 so as to chaperone them with an externally applied magnetic field.6−10 This enables the organization of CNTs into a desired pattern, which leverages the superior mechanical and electrical properties of the nanotubes. The CNTs can be organized in a polymer matrix,5 which enables in situ fabrication of embedded circuitry for sensors, e.g., for use in wearable and soft electronic devices.11−14 There are several methods to magnetize CNTs, e.g., by encapsulating MNPs,2,15,16 plating the nanotubes with a ferromagnetic metal 17 −19 or decorating them with MNPs.20−22 These methods typically require a lengthy and complex sequence of chemical syntheses or physical depositions. Often, expensive reagents and apparatus and special skill sets are also needed. While developing a chemical synthesis method, we serendipitously discovered an inexpensive and rapid alternative that is schematically illustrated in Figure 1. We examine the role of sonication in creating an ink that contains the conjugate material. By encasing the printed ink in PDMS, we demonstrate the versatile functionality of the soft composite material as a strain and oil sensor. We disperse 1 g of nickel nanoparticles (NiNPs, 40 nm diameter, US Research Nanomaterials Inc.) and 2 g of multiwalled CNTs (>95% purity, 20−30 nm outer diameter and 0.5−2 μm length, US Research Nanomaterials Inc.) in 70 mL of kerosene (Fisher Scientific) with probe sonication (QSonica Q500, 15 min, 1 s pule at every 6 s, at 30% power using 1/4″ probe). This magnetized the CNTs, i.e., the entirety of the dispersed phase containing NiNPs and CNTs was © 2016 American Chemical Society

Received: December 1, 2015 Accepted: January 6, 2016 Published: January 6, 2016 1589

DOI: 10.1021/acsami.5b11700 ACS Appl. Mater. Interfaces 2016, 8, 1589−1593

Letter

ACS Applied Materials & Interfaces

When MR ≈ 100%, sonication has effectively allowed sufficient NiNP entanglement to mobilize all CNTs. Figure 2a shows the dependence of MR on NiCNT sonication (at 30%

Figure 2. Effect of sonication time on NiNP-CNT entanglement. (a) Values of the relative mass reduction MR reflect the mass of NiCNTs created with various sonication times ts relative to the maximum reduction in mass MNiCNT observed when all NiCNTs are pulled toward the top surface of the ink. For all cases, MR ≈ 100% when ts > 60 s. One hour of CNT presonication slightly improves NiCNT entanglement for ts < 10 s. (b) TEM image of NiCNTs for ts = 120 s and 1 h of CNT presonication. (c) TEM image of NiCNTs for ts = 120 s and 1 h of Ni presonication. (d) TEM image of NiCNTs for ts = 120 s without presonication. There are no observed differences in the NiCNT conjugates formed for all the cases shown in (b−d).

Figure 1. Nickel nanoparticle entangled carbon nanotubes. (a) Nickel nanoparticles (NiNPs) and carbon nanotubes (CNTs) are dispersed in kerosene by probe sonication. The CNTs entangle the NiNPs, enabling the latter to act as magnetic chaperones without detachment or separation in a strong gradient magnetic field. (b) TEM images show clusters of NiNPs enmeshed in CNTs. (c) SQUID magnetometry shows that the NiCNTs, containing 33% (w/w) NiNPs, have a saturation magnetization Ms = 14.61 emu/g, which is commensurate with the mass fraction of nickel in the sample.

power using a 1/4″ probe) for three samples, i.e., Sample 1, which resulted from 1 h CNT presonication in kerosene, Sample 2, from a similar 1 h NiNP presonication in kerosene, and Sample 3, for which there was no presonication. The NiNPs and CNTs for Samples 1 and 2, respectively, were added after the presonication step so that each of the three samples contained CNTs (0.014 g), NiNPs (0.007 g), and kerosene (1.4 mL). Each sample was created three times to obtain an average MR value for each ts. Figure 2a shows that although CNT presonication increases entanglement for ts = 1 and 10 s, all three samples quickly converge to MR ≈ 100% when ts > 60 s. Further, the TEM images included in Figure 2b−d shows no observable differences between the three samples regarding the NiCNT conjugates that were formed when ts = 120 s.

To investigate the effectiveness of the sonication time ts in creating a robust NiCNT ink, its mass Mo is measured. First, a cuvette of 3.5 cm height is placed on the scale. Then, a magnet (K&J Magnetics, N42, 1/2 in. ×1/2 in. × 2 in.) is suspended at a 5.5 cm height above the scale, i.e., 2 cm above the cuvette. As a result, the NiCNT conjugates present in the ink are pulled upward toward the magnet so that the apparent ink mass M < Mo. The relative mass reduction MR = (Mo − M)/MNiCNT reflects the NiCNT mass created through sonication that is attracted toward the magnet. Here, MNiCNT denotes the maximum mass reduction that can be observed through the measurements, which occurs when all NiCNTs are pulled toward the top surface of the ink. 1590

DOI: 10.1021/acsami.5b11700 ACS Appl. Mater. Interfaces 2016, 8, 1589−1593

Letter

ACS Applied Materials & Interfaces

in.) was placed ∼1.5 cm above the coverslip to produce a magnetic field. The field magnetized the template, which caused the NiCNTs to settle on the glass substrate immediately above it within ∼30s. Thus, the pattern of the template was transferred above the glass in the form of a structure made entirely with NiCNTs. The kerosene was carefully evaporated by heating on a hot plate (100 °C, 5 min). This yielded ink containing a dense network of CNTs with intertwined clusters of NiNPs, as seen in the SEM images (Figure 3a)). The coverslip containing the NiCNT print was placed in a Petri dish, then covered with polydimethylsiloxane (PDMS, Sylgard 184 kit, Dow Corning) to form a ∼1.75 mm film. The PDMS infiltrated the NiCNT network, likely due to its low surface energy, as seen from the SEM images (Figure 3b). This lent an elastomeric matrix to the otherwise powdery network of NiCNTs. The sample was then heated (70 °C, ∼1 h) in an oven (Across International, AccuTemp) to cross-link the PDMS. Once cured, the PDMS film containing the magnetically printed NiCNT structure embedded within it was peeled off. Figure 3c illustrates a U-patterned NiCNT network. Flexible circuits are alternatives to otherwise rigid conductive pathways such as wires, and better suited for wearable electronic devices where flexibility is key.26 Through ambient triggers such as mechanical deformation or chemical change, flexible electronics rely on sensors to convey valuable information such as strain,14 or the presence of specific chemicals27 and biological28 species. The fabrication of flexible sensors can be simplified by implementing a rapid, facile, and inexpensive fabrication method. The NiCNT network forms a ∼2.5 μm thick conductive path, as evidenced when the top thin layer of PDMS is penetrated with two multimeter probes. Further, the matrix housing the NiCNT network, i.e., PDMS, is quite flexible. These features motivated us to evaluate the method for printing a strain gage. A voltage divider was supplied with a steady 5 V DC and the voltage drop, V, across the sensor was measured (Arduino UNO, 10 samples per second) (Figure 4a) when the PDMS block was subjected to bending (Figure 4b) and elongation (Figure 4c). For bending, a 7 mm specimen was cantilevered by fixing one end of the sample and displacing the other by 1.5 mm with a translation stage, and then relaxing the deformation instantaneously. The normalized voltage drop, V/ Vo reaches a maximum of 102.5% ± 0.1% (n = 4), where Vo = V at the sample’s neutral position. For elongation, a 7 mm long specimen was extended incrementally l = 100, 150, 200, and 250 μm, each of which was followed by relaxation. A linear transduction is observed and V/Vo = 2.4ϵ + 1 where ϵ = l/L. In both cases, relaxation is gradual, likely due to the viscoelastic nature of PDMS. However, V/Vo always reached 99.9% ± 0.2% (n = 4), indicating that there is no permanent offset. The sensor response to bending and relaxation mimics the stress−strain response that is expected of a viscoelastic polymer,29 in this case through its electrical response to changes in strain. The correlation between the electrical response of the sensor and the viscoelastic properties of PDMS may be explained through the number of NiCNT interconnections. As the sensor undergoes strain, the PDMS matrix is subjected to deformations, specifically linear elongations, which in turn change the cross-sectional area of the sample. These matrix deformations cause the relatively hard NiCNTs to reorient and move apart so that there are fewer interconnections in the strained sample than in the unstrained configuration, with an increase in the voltage drop across the sample.14,29 As the

The ink used in the experiments contains 2 g of CNTs and 1 g of NiNPs that are probe sonicated (30% power using a 1/4 in. probe) in 70 mL of kerosene for 250 s (15 min, 1 s pulse at every 6 s). After sonication, the resulting ink is black in color. The NiCNTs settle within 1 h, which requires that printing should immediately follow sonication. Next, we examined how the NiCNTs could be used to print circuitry embedded within an elastomeric matrix, as illustrated in Figure 3. A soft iron wire (diameter d = 1.15 mm) was bent into a U-shape to create a magnetic template and then affixed on a glass coverslip (VWR, No. 1). A 100 μL drop of kerosene containing dispersed NiCNTs was placed on the other side of the glass substrate so that it covered the iron template. A permanent magnet (K&J Magnetics, N42, 1/2 in. × 1/2 in. × 2

Figure 3. Magnetic printing with NiCNTs (a) An U-shaped soft magnetic wire is used as a template. A glass coverslip is placed over it, a drop of kerosene containing dispersed NiCNTs is deposited on the coverslip and the wire is then magnetized using a permanent magnet. The wire produces localized gradients in the magnetic field, which settles the dispersed NiCNTs on the coverslip surface immediately adjacent to it. The kerosene is then evaporated by heating, yielding a U-shaped dense network of CNTs intertwined with a few clusters of NiNPs, as observed by SEM. The NiCNT U network is placed inside a Petri dish, covered with liquid PDMS, and heated to cure the PDMS. The PDMS infiltrates the printed structure, and after curing, it lends a polymer matrix to the NiCNTs. (b) When the PDMS is peeled off, it forms a flexible and stretchable membrane with an embedded NiCNT structure. (c) SEM image of a cross-section of such a structure reveals an ∼2.5 μm thick NiCNT-PDMS composite network embedded in pure PDMS. The smooth texture of the top surface suggests full PDMS infiltration of the NiCNT network, resulting in a thin PDMS top layer. 1591

DOI: 10.1021/acsami.5b11700 ACS Appl. Mater. Interfaces 2016, 8, 1589−1593

Letter

ACS Applied Materials & Interfaces

came into contact with the oil, (t ≈ 83 s), V/Vo started to rise. Here Vo = V in the absence of the oil. The normalized voltage drop gradually exceeded 110% (t ≈ 200 s) at which time the experiment was ceased. Overall, the printed circuit produces repeatable strain sensitivity up to 3.6% strain. The use of other silicone rubbers, e.g., Ecoflex, as the matrix supporting the NiCNT network should facilitate the ability to sense up to 500% strain,14 which is useful for high strain wearable electronics applications. While the oil sensing capabilities of CNTs have been explored, e.g., for engine oil monitoring30 and detecting oil-dissolved gases,27 the physical manipulation30 or functionalization of CNTs27 required to assemble CNT-based sensors can be overcome by the rapid and inexpensive self-assembly technique using NiCNT inks. Integrating the NiCNT ink and assembly technique with existing technologies, such as inkjet and 3D printing, can upscale our method. Doing so will provide a simple and direct writing approach to soft circuitry that is capable of fabricating more complex geometries than we have demonstrated. However, rapid settling of NiCNTs would require that the ink be constantly sonicated to ensure nanoparticle homogeneity in the circuit. The ink viscosity would have to be tuned to ensure proper flow through nozzles and tubing. As the nozzle prints a circuit, means to ensure that previously deposited NiCNTs are not distorted would have to be adopted, such as immediate evaporation of the solvent to allow the deposited NiCNTs to adhere to the printing surface. In summary, we have readily magnetized carbon nanotubes with nickel and produced a kerosene-based ink that is patterned into a sensor network. While conventional magnetic functionalization of carbon nanotubes is based on forming strong covalent bonds between carbon nanotubes and magnetic nanoparticles, that method requires considerable volumes of reagents, is time-consuming and includes acid treatment which impairs carbon nanotube properties. Instead, simply entangling nickel nanoparticles with multiwalled carbon nanotube bundles allows the magnetic nanoparticles to behave as mobile chaperons as part of an ink, which can be moved to pattern the entangled carbon nanotubes. The movement of the NiCNT nanomaterial ink through magnetophoresis makes the directed assembly of conductive nanocomposite networks possible, thereby offering a more rapid and inexpensive fabrication process that can be tailored for mass production as compared to the use of stamps and dies. The assembled networks are functional and can be successfully integrated into PDMS to fabricate strain and oil sensors. Readily encasing these networks in a thin polymer layer enables flexibility, which is desirable for wearable and soft electronic devices.

Figure 4. The use of a magnetically printed NiCNT structure as a sensor. (a) A PDMS block containing a printed NiCNT structure is placed in a voltage divider. The voltage drop V across the structure rises when the block is subjected to (b) bending by displacement of the free end by 1.5 mm or (c) elongation of l = 100 μm, 150 μm, 200 and 250 μm. In both configurations, V increases almost instantaneously when a load is applied, but shows a gradual relaxation when the load is removed. (d) When the sensor is held at the air−water interface in a beaker and ∼1 mL of oleic acid is dropped on the water, V rises when sensor contacts the oil.

sample returns to its neutral position after the elongation strain is released, the stress relaxes more slowly due to the viscoelasticity of PDMS. This slow relaxation gradually increases the NiCNT interconnections, which would produce a corresponding slow decrease in V/Vo. Because of the oleophilic nature of PDMS and the porous structure of the PDMS-NiCNT composite, the conductive network is likely to draw oils by capillary action. This is known to cause swelling in PDMS when it is exposed to oils. Such swelling will also disturb the interconnects of the NiCNTs, thus changing the conductivity. Hence, we next examined the use of this method in sensing oil floating on water. A linear structure printed by magnetic patterning of NiCNTs was connected to two leads, then partially immersed in water (Figure 4d) contained in a 250 mL glass beaker. One mL of Oleic acid (Alfa Aesar) was then added to the water. As soon as the network



AUTHOR INFORMATION

Corresponding Author

*E-mail: [email protected]. Tel: 905-525-9140, ext. 24900. Author Contributions

The manuscript was written through contributions of all authors. All authors have given approval to the final version of the manuscript. Notes

The authors declare no competing financial interest. 1592

DOI: 10.1021/acsami.5b11700 ACS Appl. Mater. Interfaces 2016, 8, 1589−1593

Letter

ACS Applied Materials & Interfaces



(17) Li, Q.; Fan, S.; Han, W.; Sun, C.; Liang, W. Coating of carbon nanotube with nickel by electroless plating method. Jpn. J. Appl. Phys., Part 2 1997, 36, L501−L503. (18) Bittencourt, C.; Felten, A.; Ghijsen, J.; Pireaux, J.-J.; Drube, W.; Erni, R.; Van Tendeloo, G. Decorating carbon nanotubes with nickel nanoparticles. Chem. Phys. Lett. 2007, 436, 368−372. (19) Mahanthesha, P.; Srinivasa, C.; Mohankumar, G. Processing and Characterization of Carbon Nanotubes Decorated with Pure Electroless Nickel and their Magnetic Properties. Procedia Mater. Sci. 2014, 5, 883−890. (20) Chin, S. F.; Iyer, K. S.; Raston, C. L. Fabrication of carbon nanotubes decorated with ultra fine superparamagnetic nano-particles under continuous flow conditions. Lab Chip 2008, 8, 439−42. (21) He, H.; Gao, C. Synthesis of Fe3O4/Pt Nanoparticles Decorated Carbon Nanotubes and Their Use as Magnetically Recyclable Catalysts. J. Nanomater. 2011, 2011, 193510. (22) Pistone, A.; Iannazzo, D.; Fazio, M.; Celegato, F.; Barrera, G.; Tiberto, P.; Giordano, A.; Azzerboni, B.; Galvagno, S. Synthesis and magnetic properties of multiwalled carbon nanotubes decorated with magnetite nanoparticles. Phys. B 2014, 435, 88−91. (23) Xu, P.; Han, X. J.; Liu, X. R.; Zhang, B.; Wang, C.; Wang, X. H. A study of the magnetic and electromagnetic properties of γ-Fe2O3− multiwalled carbon nanotubes (MWCNT) and Fe/Fe3C−MWCNT composites. Mater. Chem. Phys. 2009, 114, 556−560. (24) Xuemin, He; Zhong, W.; Chak-Tong, Au; Du, Y. Size dependence of the magnetic properties of Ni nanoparticles prepared by thermal decomposition method. Nanoscale Res. Lett. 2013, 8, 446− 456. (25) Ghosh, S.; Puri, I. K. Changing the magnetic properties of microstructure by directing the self-assembly of superparamagnetic nanoparticles. Faraday Discuss. 2015, 181, 423−35. (26) Zhang, B.; Dong, Q.; Korman, C. E.; Li, Z.; Zaghloul, M. E. Flexible packaging of solid-state integrated circuit chips with elastomeric microfluidics. Sci. Rep. 2013, 3, 3. (27) Lu, J.; Zhang, X.; Wu, X.; Dai, Z.; Zhang, J. A Ni-Doped Carbon Nanotube Sensor for Detecting Oil-Dissolved Gases in Transformers. Sensors 2015, 15, 13522−32. (28) Wang, J. Carbon-Nanotube Based Electrochemical Biosensors: A Review. Electroanalysis 2005, 17 (1), 7−14. (29) Liu, C.-X.; Choi, J.-W. Analyzing resistance response of embedded PDMS and carbon nanotubes composite under tensile strain. Microelectron. Eng. 2014, 117, 1−7. (30) Moon, S.-I.; Paek, K.-K.; Lee, Y.-H.; Kim, J.-K.; Kim, S.-W.; Ju, B.-K. Multiwall Carbon Nanotube Sensor for Monitoring Engine Oil Degradation. Electrochem. Solid-State Lett. 2006, 9, H78−H80.

ACKNOWLEDGMENTS This work is supported by a Natural Sciences and Engineering Research Council of Canada (Grant RGPIN-2014-04066) Discovery Grant (NSERC-DG), and Canada Foundation for Innovation John R. Evans Leaders Fund (CFI-JELF) and Ontario Research Fund Research Infrastructure (ORF-RI) grants (Grant no. 33016). For assistance with measurements, we thank Dr. Carmen Andrei of the Canadian Centre for Electron Microscopy (CCEM) for TEM, and Dr. Paul Dube of the Brockhouse Institute for Materials Research, McMaster University, for SQUID. The CCEM is a national facility supported by NSERC, Canada Foundation for Innovation and McMaster University.



REFERENCES

(1) Fragouli, D.; Das, A.; Innocenti, C.; Guttikonda, Y.; Rahman, S.; Liu, L.; Caramia, V.; Megaridis, C. M.; Athanassiou, A. Polymeric films with electric and magnetic anisotropy due to magnetically assembled functional nanofibers. ACS Appl. Mater. Interfaces 2014, 6, 4535−41. (2) Das, A.; Raffi, M.; Megaridis, C.; Fragouli, D.; Innocenti, C.; Athanassiou, A. Magnetite (Fe3O4)-filled carbon nanofibers as electro-conducting/superparamagnetic nanohybrids and their multifunctional polymer composites. J. Nanopart. Res. 2015, 17, 1−14. (3) Meng, L.; Fu, C.; Lu, Q. Advanced technology for functionalization of carbon nanotubes. Prog. Nat. Sci. 2009, 19, 801−810. (4) Balasubramanian, K.; Burghard, M. Chemically functionalized carbon nanotubes. Small 2005, 1, 180−92. (5) Korneva, G.; Ye, H.; Gogotsi, Y.; Halverson, D.; Friedman, G.; Bradley, J. C.; Kornev, K. G. Carbon nanotubes loaded with magnetic particles. Nano Lett. 2005, 5, 879−84. (6) Ganguly, R.; Puri, I. K. Field-assisted self-assembly of superparamagnetic nanoparticles for biomedical, MEMS and BioMEMS applications. Adv. Appl. Mech. 2007, 41, 293−335. (7) Ganguly, R.; Puri, I. K. Microfluidic transport in magnetic MEMS and bioMEMS. Wiley Interdiscip. Rev.: Nanomed. Nanobiotechnol. 2010, 2, 382−399. (8) Puri, I. K.; Ganguly, R. Particle Transport in Therapeutic Magnetic Fields. Annu. Rev. Fluid Mech. 2014, 46, 407−440. (9) Nandy, K.; Chaudhuri, S.; Ganguly, R.; Puri, I. K. Analytical model for the magnetophoretic capture of magnetic microspheres in microfluidic devices. J. Magn. Magn. Mater. 2008, 320, 1398−1405. (10) Abdel Fattah, A. R.; Ghosh, S.; Puri, I. K. Printing microstructures in a polymer matrix using a ferrofluid droplet. J. Magn. Magn. Mater. 2016, 401, 1054−1059. (11) Li, Y.; Yu, T.; Pui, T.; Chen, P.; Zheng, L.; Liao, K. Fabrication of transparent and conductive carbon nanotube/polyvinyl butyral films by a facile solution surface dip coating method. Nanoscale 2011, 3, 2469−71. (12) Liu, C.-X.; Choi, J.-W. Patterning conductive PDMS nanocomposite in an elastomer using microcontact printing. J. Micromech. Microeng. 2009, 19, 085019. (13) Luo, J.; Fan, F. R.; Zhou, T.; Tang, W.; Xue, F.; Wang, Z. L. Ultrasensitive self-powered pressure sensing system. Extreme Mech. Lett. 2015, 2, 28−36. (14) Amjadi, M.; Yoon, Y. J.; Park, I. Ultra-stretchable and skinmountable strain sensors using carbon nanotubes-Ecoflex nanocomposites. Nanotechnology 2015, 26, 375501. (15) Sameera, I.; Bhatia, R.; Prasad, V. Characterization and magnetic response of multiwall carbon nanotubes filled with iron nanoparticles of different aspect ratios. Phys. E 2013, 52, 1−7. (16) John, A. R.; Arumugam, P. Studies on structural and magnetic properties of pristine and nickel-filled carbon nanotubes synthesized using LaNi 5 alloy particles as a catalyst. Chem. Eng. J. 2014, 243, 436− 447. 1593

DOI: 10.1021/acsami.5b11700 ACS Appl. Mater. Interfaces 2016, 8, 1589−1593