MSE’s David Bishop Offers Roadmap in Physics Today Cover Story
By Mark Dwortzan
BU researchers envision mass-producing high-speed nanoscale devices ranging from integrated circuits to biosensors, but their aspirations have been hampered by a persistent inability to precisely manipulate nanomaterials to build reliable, functional products at a reasonable cost. In a nutshell, the key challenge has been to pattern extremely small materials at exact locations in a repeatable manner over relatively large surfaces all within a very short timeframe. Current nanomanufacturing processes are likely to become wasteful and/or expensive if applied on a commercial scale, potentially costing orders of magnitude more than the value of the devices or systems they’re designed to build.
“Absent major breakthroughs in nanomanufacturing, the current trend in smartphones, laptops, PCs and other electronic devices toward smaller, faster, better and cheaper models may grind to a halt,” says MSE Division Head and Professor David Bishop (ECE, Physics, MSE). But Bishop is confident that with sufficient investment, nanomanufacturing can evolve from a laboratory technology to one capable of generating commercial products on a massive scale within the next decade or two. In the cover story of the December 2014 edition of Physics Today, he and coauthor Matthias Imboden, an ECE postdoc, explore three potential pathways—each requiring further research, to overcome current limitations—to the high-speed, low-cost and scalable manufacture of devices with nanoscale features.
The first and primary avenue that nanomanufacturing researchers are pursuing is resist-based nanolithography, a nanoscale version of photolithography, commonly used today in the manufacture of semiconductor devices. In conventional photolithography, manufacturers pass light through a mask, which transfers the pattern of the mask (holes and traces of different shapes and sizes) to an intermediate photoresist layer, which, in turn, transfers it to a target device layer beneath it.
In nanolithography, the features or holes in the mask are on the order of one-billionth of a meter, resulting in similar-sized features on the manufactured device layer. To achieve deep nanoscale features—those sized below 100 nanometers—will require replacing visible light with something that has much shorter wavelengths, such as x-rays, which are hard to produce and focus; or electron beams, which now take a prohibitively long time to get the job done.
The second approach is nanoimprinting or nanostamping, a process akin to using a nanoscale rubber stamp. Capable of making identical copies of nanoscale features on a device layer, nanoimprinting could be ideal for manufacturing memory chips and magnetic displays. It works very well for single-layer systems in the lab, but has challenges ensuring that multiple layers of an integrated circuit or other nanomanufactured device will line up correctly.
The third strategy is to apply a direct writing technique, controlling the placement of atoms with nanoscale precision. An example of this is dip-pen lithography, which deploys atoms at precise locations on a device surface just as a ballpoint pen’s inkball deposits ink on a piece of paper. Dip-pen lithography works with a variety of materials, but it’s a slow, serial process, akin to writing out a newspaper with a pen.
Another direct writing technique that was invented by Bishop at BU, atomic calligraphy, effectively spray paints atoms at desired locations on a surface through precisely positioned holes in a mask. In 2013, in a paper in the journal Nano Letters, Bishop and collaborators at Boston University and Bell Laboratories introduced this new technology: a low-cost, microelectromechanical system (MEMS)-based machine that directs atoms onto a surface through different-sized holes—each one 50 – 200 nanometers across—on silicon plates. These MEMS plates can move with nanometer precision to create exacting patterns over surfaces of more than 400 square microns, roughly the area of several human hairs. Shutters positioned micrometers above each MEMS plate enable high-speed control of where and when atoms are deposited.
The researchers produced lines, bridges, rings, infinity symbols, BU logos and many other nanoscale metal patterns by depositing gold and chromium atoms through the holes while moving the plates. By scaling the number of micromachines and using arrays of apertures, the process could be made parallel, and thus potentially fast enough for commercial-grade nanomanufacturing. Spearheading a project called “Fab on a Chip,” Bishop is exploring ways to incorporate this method in a cost-effective, chip-based fabrication process for atomic-scale materials and devices that are initially designed in digital simulations, making possible everything from downsized electronics to more compact biosensors.
Of the three methods, nanoimprinting has shown the most promise so far for achieving commercial viability, but Bishop cautions against favoring one method over another.
“This is not a one-size-fits-all proposition,” he says. “One technique will probably not end up ‘the winner,’ but they’ll all have regions of optimal applicability. Nanomanufacturers of the future may place different technologies on a chip that can be activated at different times to achieve different purposes.”