Novel Inks for Direct-Write Assembly of 3-D Periodic Structures

Introduction

New methods for materials fabrication at the micro- and nanoscale will drive scientific and technological advances in areas of materials science, chemistry, physics, and biology. The broad diversity of potentially relevant materials, length scales, and architectures underscores the need for flexible patterning approaches. One important example is the fabrication of 3-D periodic structures composed of colloidal,1 polymeric,^2–4 or semiconductor⁵ materials. These structures may find potential application as sensors,⁶ microfluidic networks,⁷ self-healing materials,⁸ photonic band gap materials,⁹ and tissue engineering scaffolds.¹⁰ Several strategies have recently emerged for precisely assembling 3-D periodic arrays,^1–5 including colloidal epitaxy,¹ litho-⁵ and holographic,³ and direct-write techniques.^2–4 Of these, only the latter approach offers the materials flexibility, low cost, and ability to construct complex 3-D structures required for advances across multidisciplinary boundaries.

This article focuses on our recent efforts to design novel inks for direct-write assembly. Through careful control of ink composition, rheological behavior, and printing parameters, 3-D structures that consist of continuous solids, high aspect ratio (e.g., parallel walls) or spanning features can be constructed. Of these, 3-D periodic structures offer the greatest challenge for designing inks, because they contain self-supporting features that must span gaps in the underlying layer(s). Below we first describe the direct-write assembly process. We then introduce several ink designs, including concentrated colloidal and nanoparticle,^10–14 polyelectrolyte,^4–15 and sol-gel¹⁶ inks, that enable the directwrite assembly of 3-D periodic architectures of diverse functionality with features ranging from tens of microns to submicron in size. Finally, we highlight the opportunities and challenges associated this approach.

Direct-Write Assembly

The term “direct-write assembly” describes fabrication methods that employ a computer-controlled translation stage, which moves a pattern-generating device, i.e., ink deposition nozzle, to create materials with controlled architecture and composition.¹⁷ Unlike ink-jet printing, our approach relies on extruding a continuous ink filament that is deposited in a layer-by-layer build sequence to generate the desired component of interest. The ink is delivered either under constant displacement or pressure mode through a single or multi-nozzle array. The filament diameter is determined by the nozzle diameter, ink rheology, and deposition speed. The component dimensions, minimum feature size, and build times are dictated in part by the lateral (x-y) and vertical (z) translation distances, resolution, and speed. We have recently implemented two 3-axis, motion-controlled stages in our laboratory, as shown in Figure 1. They range from the highest precision stage, which is mounted on an inverted fluorescence microscope and has maximum x-y-z travel distances of 300 μm with nanometer resolution and travel speeds of ~ 1 mm/sec, to a larger area stage, in which the maximum x-y travel distances exceed several centimeters with a resolution of a tens of nanometers and travel speeds of up to 30 mm/s. These vastly different capabilities allow us to pursue applications that range from photonic crystals to self-healing composites.

Novel Ink Designs

(a) Colloidal and Nanoparticle Inks

Colloidal gels are excellent candidate materials for direct ink writing of complex 3-D structures, because their viscoelastic properties can be tailored over many orders of magnitude to facilitate flow through nozzles and produce patterned filaments that maintain their shape, even as they span gaps in the underlying layers of the printed structure.¹² We designed these inks with two important criteria in mind. First, they must exhibit a well-controlled viscoelastic response, so they flow through the deposition nozzle and then “set” immediately to facilitate shape retention of the deposited features even as they spans gaps in the underlying layer(s). Second, they must contain a high colloid volume fraction to minimize drying-induced shrinkage after assembly is complete, so the particle network is able to resist compressive stresses arising from capillary tension. These criteria required careful control of colloidal forces to first generate a highly concentrated, stable dispersion followed by inducing a system change (e.g., ΔpH, ionic strength, or solvent quality) that promotes a fluid-to-gel transition.

Colloidal gels consist of a percolating network of attractive particles capable of transmitting stress above a critical volume fraction, φ_gel. When stressed beyond their yield point (t_y), they exhibit shear thinning flow behavior due to the attrition of particle-particle bonds within the gel. As the inks flow through the deposition nozzle, they experience a radially varying shear stress. The core of the ink filament remains unyielded and experiences plug-like flow, whereas the outer region of the ink filament exhibits yielding and therefore liquid-like flow.¹² Hence, the ink exits the nozzle as a continuous, rod-like filament with a rigid (gel) core-fluid shell architecture, which simultaneously promotes its shape retention and allows it to fuse with previously patterned features at their contact points. Upon deposition, the fluid shell quickly gels as the attractive particle bonds reform.¹²

We first demonstrated this ink design using a model system consisting of negatively charged, silica microspheres coated with a cationic polyelectrotrolyte, poly(ethylenimine) (PEI) suspended in deionized water.¹³ Concentrated silica suspensions (φ = 0.46) exhibited a fluid-to-gel transition as their pH was adjusted to a value near their point-of-zero charge. A dramatic rise in elastic properties accompanied this phase transition. Both the shear yield stress and elastic modulus increased by orders of magnitude, because of strengthened interparticle attractions near this pH. Using this ink, 3-D periodic structures were assembled (Figure 2).

This ink design can be readily extended to any type of colloidal material provided their interparticle forces can be controlled to produce the desired solids concentration and rheological properties. In addition to changing pH, the requisite ink rheology may be achieved through the addition of salt, oppositely charged polyelectrolyte species, or other coagulants. These strategies have been employed to produce inks from a broad array of colloidal materials, including silica,¹³ lead zirconate titanate,12 barium titanate,¹⁴ alumina,¹⁸ hydroxyapatite (04238),¹⁰ polymer latices,¹⁹ and most recently, metallic nanoparticles. Further reductions in feature size are possible by designing nanoparticle inks, in which the maximum particle diameter is less than 100 nm (Figure 2). The commercial availability of high quality inorganic, polymeric, and metallic nanoparticles, with precise control over particle composition, shape, size, and size distribution, is highly desirable.

(b) Polyelectolyte Inks

It has been a grand challenge to design concentrated inks suitable for direct writing at the microscale. Colloidal inks either experience jamming (or clogging) in the deposition nozzle or require exceedingly large pressures to induce ink flow. To overcome these limitations, we drew inspiration from nature to develop concentrated polyelectrolyte complexes that mimic spider silk in a simplistic way.

This ink design utilizes polyelectrolyte complexes composed of non-stoichiometric mixtures of polyanions and polycations.⁴ We first explored mixtures of poly(acrylic acid), PAA and poly(ethyleneimine),PEI that were nominally 40 wt.% polyelectrolyte in an aqueous solution. By regulating the ratio of anionic (COONa) to cationic (NHx) groups and combining these species under solution conditions that promote polyelectrolyte exchange reactions,¹⁵ we produced homogeneous fluids over a broad compositional range that possessed the requisite viscosities needed for flow through micro-capillary nozzles of varying diameter.

The concentrated polyelectrolyte inks rapidly coagulate to yield self-supporting filaments (or rods) upon deposition into an alcohol/water coagulation reservoir. The exact coagulation mechanism, driven by electrostatics in a water-rich or solvent quality effects in an alcohol-rich reservoir, as well as the magnitude of ink elasticity depend strongly on the alcohol/ water ratio. By carefully tuning this parameter, the deposited ink filament is elastic enough to promote shape retention, while maintaining sufficient flexibility for continuous flow and adherence to the substrate and underlying patterned layers. 3-D micro-periodic scaffolds are created by depositing the PAA-PEI ink into an alcohol-rich reservoir (Figure 3). Such structures may find potential application as sophisticated scaffolds that guide the electrostatic layer-by-layer assembly of materials,⁴ direct cell-scaffold interactions, or interact with other environmental stimuli, or as templates for biomimetic,²⁰ photonic, microfluidic,⁷or low-cost MEMs devices.²¹ This ink design can be readily extended other polyelectrolyte mixtures,^4,2D including those based on biologically, electrically, or optically active species. For example, it should be relatively straightforward to extend our approach to patterning biological materials, such as silk and polypeptides.

(c) Sol-Gel Inks

The ability to pattern oxide structures at the microscale in both planar and three-dimensional forms is important for a broad range of emerging applications, including sensors, micro-fuel cells and batteries, photocatalysts, solar arrays, and photonic band gap (PBG) materials. By designing sol-gel inks based on organometallic precursors, we have recently demonstrated direct ink writing of micro-periodic oxide structures (Figure 2d).¹⁶

Our ink design incorporates a sol-gel precursor solution based on a chelated titanium alkoxide, titanium diisopropoxide bisacetylacetonate (TIA, 325252).¹⁶ An organic polymer, poly(vinyl pyrrolidone) (PVP, e.g. 234257), is also included to mitigate stresses that occur during drying and calcination of the as-patterned structures. Unlike the polyelectrolyte inks described above, these sol-gel inks can be patterned directly in air. Using DIW, we created 3-D micro-periodic structures composed of parallel arrays of orthogonally stacked rods. To convert these structures to the desired oxide phase, in this case TiO₂, they are calcined at elevated temperatures (> 450 °C). Both the anatase and rutile phases can be generated depending on the precise heat treatment conditions. This sol-gel ink design and patterning approach can be readily extended to other organometallic precursors. For example, by simply varying the organometallic precursors used, we have formulated inks for microscale patterning of electrically (e.g., doped-TiO₂), transparent (e.g., indium tin oxide), and ionically (e.g., doped-zirconium oxide) conducting oxides (Table 1). The broad palette of precursor materials available enables a myriad of potential applications to be pursued.

Opportunities and Challenges

Direct ink writing offers the ability to rapidly pattern functional materials in complex 3-D architectures from a diverse array of materials. Using the ink designs highlighted above, the current minimum feature sizes range from approximately 200 μm for colloidal inks to 250 nm for sol-gel inks at characteristic printing speeds of 1–10 mm/s. However, the continual drive towards patterning materials at even finer length scales and faster printing speeds gives rise to many opportunities and challenges. Future advances will require new ink chemistries, better characterization and modeling of ink dynamics during deposition, and enhanced robotic, control, and ink delivery systems to allow three-dimensional writing with greater spatial and composition control. Specific chemistries of interest include semiconducting and metallic nanoparticles as well as novel hydrogel and sol-gel precursors that can be readily formulated with the above considerations in mind.

Acknowledgments

The author gratefully acknowledges the generous support for our work by U.S. Department of Energy through the Frederick Seitz Materials Research Laboratory (Grant# DEFG02- 91ER45439), the Army Research Office through the MURI program (Grant# DAAD19-03-1-0227), the National Science Foundation (Grant# 00-99360) and the Air Force Office of Scientific Research through both the MURI (Grant# F49550-05- 1-0346) and the DURIP program (Grant# FA9550-06-1-0321). This work has benefited from the valuable contributions of J. Cesarano, J. Smay, G. Gratson, M. Xu, R. Shepherd, R. Rao, E. Duoss, D. Lorang, S. White, N. Sottos, D. Therriault, K. Toohey, C. Hansen, W. Wu, and J. Bukowski.