Polymer composite microtube array produced by meniscus-guided approach

Fabrication of polymeric micro/nanotubes is receiving increasing attention because of their potential applications in microfluidic, catalysis, tissue engineering, drug delivery, photonics, sensing, and so on.^1–6 Main ingredients of polymeric micro/nanotube structures have been mostly limited to single component polymers. Polymer composite that has broad applications as one-dimensional (1-D) fibers,^7,8 wires,⁷ or cables⁹ due to its enhanced mechanical property and thermochemical stability^7–11 is, however, largely unexplored as a form of hollow microtube structure.

Integration of micro/nanotubes with various electrical, chemical or biological properties has been more and more important in bio- or chemical sensors.^5,7,10,12 Here, one of the issues in integration is to regulate tube dimensions (diameters and wall thicknesses) at the desired sizes.^13,14,29 Size regulation during integration is a challenging task with conventional fabrication methods, such as self assembly,^5,15,16 templating,^13,17,18 electrospinning,^19,20 co-electrospinning,²¹ wetting,²² TUFT(tube by fiber template) process,²³ self rolling,¹⁴ co-extrusion method,²⁴ direct laser polymerization,²⁵ and solvent evaporation induced phase separation.²⁶ Size regulation is still a challenge for polymer composite microtubes, even if successfully fabricated.

In this study, to realize single polymer composite microtubes, we propose a meniscus-guided^27–32 approach using a microscale meniscus of a colloidal solution, exploiting the fast evaporation mechanism of solvent.^33–38 We successfully produced single freestanding hollow microtubes of (poly(methyl methacrylate)/polypyrrole (PMMA/PPy)) composite. PMMA/PPy composite is known to show excellent properties of high sensitivity and fast response for chemical sensor^39–41 and of high biocompatibility, electrical conductivity, and chemical surface modifiability for biosensor platform.^42,43 We were able to regulate the outer diameter and wall thickness of PMMA/PPy microtubes from 4.5 to 50 μm and from 0.7 to 9.0 μm, respectively, by systematically controlling the guiding parameters of the meniscus. We in turn produced an array of single PMMA/PPy freestanding microtubes, individually controlled in size at the desired positions. Finally we demonstrated individually addressable gas sensors by integrating PMMA/PPy composite microtubes on a pattern of Pt electrodes.

A colloidal solution containing PMMA/PPy nanoparticles (NPs) in methanol was prepared by two-step microemulsion polymerization¹¹ using pyrrole and MMA, DTAB as a surfactant, iron (III) chloride for oxidation of pyrrole, iodine for doping PPy, and AMPAD as an initiator to PMMA. A Fourier-transform infrared (FT–IR) spectra of the NPs, taken with a Nicolet 6700 FT–IR spectrometer, show that the synthesized polymer NPs consist of PMMA and PPy,¹¹ as demonstrated in Fig. 1(a). The size of the NPs was controlled to a mean size around 40 nm, as measured by a Zetasizer (Melvern Zetasizer Nanoseries) (Data not shown).

Glass micropipettes (tip radius r₀ = 2.2 – 12.5 μm) for guiding the meniscus were prepared using a pipette-puller (P-97, Sutter Instrument). Microtube fabrication procedure was in real-time monitored by using an optical imaging system, which was made of a CCD camera (Infinity1-2, Lumenera) with 1600 × 1200 pixels coupled with an objective lens (×50 objective lens, Mitutoyo, NA 0.42).

Figures 1(b)–1(e) schematically illustrate the fabrication process of a single freestanding PMMA/PPy composite microtube. When a micropipette filled with the NP solution nearly touches a substrate, a microscale meniscus is created outside its opening (Fig. 1(b)).^27–32 Then the evaporating liquid pushes the suspended NPs to the meniscus edge (red arrows in Fig. 1(b)), concentrating them and leaving a ring-like deposit, called the “coffee ring”.^33–36 As the micropipette is slowly pulled up (< 25 μm/s) away from the substrate, the outward liquid flow continually accumulates the NPs on the ring-like deposit. The coffee ring approach is still in play within the meniscus that is guided upward by the micropipette, resulting in the formation of a microtube (Fig. 1(c)). Once exposed to the air by taking away the micropipette from the produced microtube, the remanent solution inside the microtube rapidly evaporates from the top to the bottom of the microtube (Figs. 1(d)), finally leaving a hollow microtube (Fig. 1(e)). The four sequential optical images in Figs. 1(f)–1(i), taken from the real-time MOVIE I (see enhancement) for the whole process, correspond to the sequential steps of Figs. 1(b)–1(e), respectively.

Wall thicknesses (WTs) of the microtubes were uniform along tube height, as demonstrated by three representative freestanding microtubes in Fig. 2. Characterization of WT along tube height was carried out using X-ray microtomography.^44,45 The inset in Fig. 2 demonstrates a 3-D volume-rendering image of a microtube (15 μm outer diameter (OD)) with a WT of ∼3.5 μm.^44,45 X-ray microtomography was performed at 6D XMI beamline (PAL, Korea: http://paleng.postech.ac.kr/). 500 projected images were taken at every 0.36º rotation step per each microtube and then reconstructed by four parallel computers equipped with the Octopus8.5 software^44,45 (inCT, Zwijnaarde, Belgium). Vertically stacked 2D slices were reconstructed as volume-rendered 3D images using the Amira5.2 software^44,45 (Visage Imaging, San Diego, CA, USA).

The OD and WT of PMMA/PPy microtube were dependent on three parameters: 1) spreading time of the meniscus, i.e. the holding time of a micropipette on the substrate, 2) pulling rate of the micropipette, and 3) micropipette radius r₀. The OD and WT increased with spreading time, as shown in Fig. 3(a). The increase was attributed to the increase in meniscus size with spreading time, as seen by the increased bottom diameter of a meniscus (Fig. 4(a)) with spreading time. In fact, the OD and thickness of ring-deposits increased with spreading time (Fig. 4(b)). The decrease (increase) in OD and WT with pulling rate (micropipette radius r₀) in Fig. 3(b) (Fig. 3(c)) was again attributed to the decrease (increase) in meniscus size with pulling rate (micropipette radius r₀).^27,28 Overall, the ODs and the WTs were regulated from 4.5 to 50 μm and from 0.7 to 9.0 μm, respectively, by controlling spreading time, pulling rate, and micropipette radius (Fig. 3(d)).

Three-dimensional meniscus-guiding that was done by controlling the motion of a micropipette using a 3-axis motorized stage²⁷ enabled us to produce an array of single freestanding PMMA/PPy microtubes that were individually regulated in sizes (ODs, WTs, and heights) at the desired positions. Figure 5 shows, for example, a “MSE” pattern of single freestanding PMMA/PPy microtubes, individually regulated not only in sizes (OD, WT, and height), but also in position. Here, the OD and the WT were regulated by spreading time. The height was simply controlled by pulling time.

Finally, we demonstrate a useful application of the microtube produced by the meniscus-guided approach. Individual control of sensing properties of sensor components is potentially important for the integration of multi-function devices. We applied our approach to the integration of individually addressable gas sensors. Figure 6(a) shows an array of three PMMA/PPy microtube arches fabricated by pulling up and then guiding the micropipette from one Pt electrode to another one using a 3-axes motorized stage.²⁷ The real-time gas sensing experiment was carried out for each microtube arch at 1000 ppm of ammonia gas with a voltage of 5V between the Pt-Pt electrodes. The resistance changes of the composite microtube gas sensors under ammonia were monitored in a homemade chamber with gas inlet/outlet using a source-meter (Keithley 2612A) connected to a computer. Gas sensitivity measurements were determined by the resistance change (ΔR = R − R₀), where R is the maximum resistance of the sensing microtube under exposure of ammonia gas and R₀ is the resistance of the sensing microtube before ammonia gas is exposed. As seen in Fig. 6(b), each microtube gas sensor displayed reversible and reproducible response upon on and off of ammonia gas. The sensitivity depending on the geometry of the microtube (OD and WT), as summarized in Fig. 6(c), increases at smaller sizes probably due to the enhanced surface-to-volume (S/V) ratios of the microtubes.⁴⁶ This implies that sensing properties of microtube components can be individually controlled by modulating the size. Such a successful demonstration suggests a possible way to integrated arrays of individually addressable microtube-based microdevices with various functions.¹² Possible applications would range from stretchable electronics and optoelectronics to bio-sensors and functional biocompatible devices.^{7,10,12,42,43}

In summary, we successfully produced single freestanding PMMA/PPy composite microtubes based on guiding a meniscus solution of PMMA/PPy nanoparticles under edgeward flow in a fast evaporating solvent. The ODs and the WTs were well regulated, for instance, from 4.5 to 50 μm and from 0.7 to 9.0 μm, respectively, by controlling spreading time, pulling rate, and micropipette radius. We were able to produce an array of freestanding PMMA/PPy microtubes, individually controlled in size at the desired positions. We finally demonstrated individually addressable gas sensors by integrating PMMA/PPy composite microtubes on a Pt – Pt electrodes. Further study is required to understand detailed growth behaviors of the microtubes. Our approach may be applied to various composite materials systems with complicated geometries.