Rapid 3D-printing method could be used on space missions

Fresh approaches to 3D printing promise to transform how, where and by whom objects are produced to serve a broad range of industries, ranging from health care to aerospace. A substantial and expanding category of 3D-printing processes — sometimes called digital light manufacturing — uses patterns of light projected onto a liquid resin to trigger the liquid to solidify at defined positions, thereby forming solid structures in desired shapes. The ultimate goal is to develop these processes to rapidly print features at microscopic scales using a range of materials, with excellent surface finishes and at sizes suitable for industrial applications (potentially up to a scale of many centimetres). However, practical issues limit the speeds of current processes, and the quality of the resulting objects is reduced by the scattering, refraction and absorption of the light patterns in the liquid. Writing in Nature, Vidler and colleagues1 report a 3D-printing strategy that could solve these problems.
Read the paper: Dynamic interface printing
Established processes for digital light manufacturing print objects that are made from materials called photopolymers and use one of two main configurations. In the first, light patterns are projected up into a tray of precursor liquid through a transparent window (Fig. 1a), triggering polymerization of a layer of the precursor liquid just above the window2. The pool of liquid in the tray does not need to be deep enough to surround the final manufactured object completely. Instead, as each layer is produced, the printed object is lifted upwards and the liquid precursor is drawn into the zone between the object and the window where the next layer will be added. However, it remains challenging to ensure a steady flow of precursor into this zone, as well as to dissipate heat released by the polymerization reaction and to prevent the printed part from adhering to the window.
The second approach forms layers by illuminating the top surface of a vat of precursor liquid, and the printed object is gradually lowered into the vat3 (Fig. 1b). Here, the challenge is that surface tension drives complex interactions of the liquid with the topmost layer of the solid object, which can sometimes make the liquid surface uneven. A mechanical wiper is therefore generally needed to distribute the liquid evenly across the top layer.
Vidler and colleagues address these challenges by printing at the meniscus — the curve in the surface of a liquid — of a precursor liquid located at the end of a tubular print head (Fig. 1c). In this process, which the authors call dynamic interface printing (DIP), light patterns are projected down onto the meniscus as the print head moves up, out of the vat of precursor. The tube is pressurized to control the shape of the meniscus. The recoating of liquid onto the solid surface of the printed part of the object is driven by the surface tension of the liquid, and is further accelerated by acoustic vibrations (oscillations of pressure of the air inside the tube), which stimulate the liquid in the vat to flow steadily across the meniscus.
3D printing enables mass production of microcomponents
Through these innovations, the researchers report a linear printing speed of up to about 0.7 millimetres per second across centimetre-sized objects. For comparison, another digital light manufacturing process called high-area rapid printing (HARP) was first reported4 with speeds of up to 0.12 mm s−1; and currently available commercial printers that implement HARP have speeds of up to 0.08 mm s−1. However, DIP is currently limited to centimetre-scale objects (the print heads are 5–30 mm in diameter), whereas HARP and related processes have already reached print-tray sizes in the tens of centimetres. It remains to be seen whether larger printing areas can be achieved for DIP. Using multiple print heads packed together in an array might help, if an approach for printing material in the gaps between the tubular heads can be found. Perhaps square or rhombic print heads could be used, to close up the gaps; and perhaps the print heads could be moved sideways to illuminate the areas between the gaps.
Vidler et al. demonstrate that DIP can produce intricate lattice structures with features as small as about 30 micrometres in diameter, intertwined microfluidic channels and centimetre-scale gels in shapes that mimic the heart and kidney. The authors also show that the sedimentation of solid particles, such as biological cells, in the vat can be reduced by the acoustic vibrations. This ability to inhibit sedimentation is promising for applications in which printed structures are embedded with cells by suspending the cells in the precursor liquid. Achieving high cellular density, without printing quality being compromised by light scattering, is a major challenge in this field. The authors show that DIP can achieve a cellular density of 7.2 million cells per millilitre of liquid — comparable with the state of the art in bioprinting, but still well below the cell densities of natural tissue5.
This work amplifies an exciting trend towards blending optical and acoustic manipulation of polymers and of particles in them for 3D printing6,7. Such techniques could enable the fabrication of advanced composite materials, components in electronic circuits and even structures that contain arrays of chiplets (tiny computer chips that perform a subset of a larger system’s functions). An area warranting further thought is the integration of fibres into printed photopolymers to reinforce strength. It would be desirable to align fibres in the directions in which peak strength is needed in an object. However, this might be easier to do using the emerging 3D-printing methods in which the entire volume of an object is solidified at once8, rather than layer-by-layer approaches.
Multi-material 3D printing guided by machine vision
The authors suggest that DIP could produce even smaller microscale features by increasing a property of the system’s optics known as the numerical aperture, which would focus the projected light to a tighter spot. However, increasing numerical aperture would inevitably also reduce the depth of focus, making it harder to deliver a focused pattern across the curved meniscus. It might be possible to overcome this challenge by using customized holographic optics to morph the focal plane of the light to match the shape of the meniscus. Alternatively, engineering the geometry or surface properties of the print head’s rim could make the meniscus flatter, mitigating the focusing challenges.
The DIP print-head design might also provide opportunities to use forced convection in the pressurized air to dissipate the heat produced by polymerization reactions at the meniscus. Moreover, DIP might be attractive for use in space missions or in environments with low or variable gravity, because the meniscus can be held in place by surface tension and by the air pressure in the print head. By contrast, other 3D-printing methods require gravity to hold the liquid in place or to create a stable air–liquid interface.
As more 3D-printing processes enter the field, it is important to consider not only performance metrics such as speed, object size and strength of the printed material, but also the life-cycle impact of choosing a manufacturing process that relies on covalently bonded photopolymers. These materials are generally much harder to recycle than thermoplastic polymers, which can be ground, melted and reprocessed. It will be interesting to see whether DIP and other light-based 3D-printing processes can use the recyclable photopolymers that are now emerging9. DIP could be well placed as a process to scope out the utility of this generation of recyclable materials.