Silicon is liberated from Abbot’s Flatland
The famous physicist Richard Feynman once gave a lecture titled, "There is plenty of room at the bottom." This lecture is nowadays often quoted to highlight the successes of modern micro- and nano-fabrication techniques, and the value of available space that comes with advances in miniaturization. Their continual progress enables new digital technologies that we enjoy daily. In this respect, silicon, the bedrock of modern computers, mobile communications, and photonic devices, has proven to be extremely capable. The material is incorporating more complexity and speed than ever, usually described as Moore's law. Precisely due to these amazing successes, one may be surprised to hear that virtually none of these advances are taking advantage of the vast space available, below the surface, inside the silicon wafers; even the so-called 3D devices are, in essence, stacks of planar structures. In this sense, Si microelectronics and photonics continues to live in Abbot's Flatland. What more could be achieved, if the bulk of silicon was opened to usage?
Now, a diverse team of scientists centered at Bilkent University and Middle East Technical University (both in Ankara, Turkey) have found a way to pack various laser-written structures deep inside silicon chips. In the latest issue of Nature Photonics, the researchers describe their novel approach, where a focused infrared laser beam exploits the inherent optical response of silicon to create 1-μm-resolution (one-hundredth of the human hair) building blocks in a sliver of silicon. For the first time, the researchers demonstrate arbitrary 3D fabrication inside silicon, without structures above or below.
But this was only the first hurdle the researchers had to overcome. Next, they convert these complex 3D architectures into functional optical devices, such as lenses, waveguides, holograms and other optical elements. "We achieve this by exploiting dynamics arising from nonlinear laser-material interactions, leading to controllable building blocks," says Dr. Onur Tokel of the Department of Physics at Bilkent, who is the lead author of the paper. "In any 3D fabrication method, there is a trade-off between speed, resolution, and complexity. With our approach, we are hitting the sweet spot. The critical realization is noticing that most practical components can be made out of rod- or needle-like building blocks. Our method enables creating precisely such blocks, while also preserving a width of about 1 micrometer for each block. Better yet, the rods can be combined to create a 2D layer, or even more complex 3D shapes, which can simply be created by scanning the laser beam over the chip."
A further outcome of the method is related to 3D printing or sculpting. The researchers found that by exposing the laser-modified areas to a specific chemical etchant, it is possible to realize 3D sculpturing of the entire wafer. They demonstrated various microscopic components, such as microchannels, thru-Si vias, cantilevers, and micropillars. Creation of some of these is prohibitively difficult with other methods. "I should note that this is a direct-laser writing approach, without the use of masks, inexpensive compared to reactive ion etching and e-beam lithography," notes Dr. Serim Ilday, of the Department of Physics, one of the coauthors of the paper. The team's approach has the added benefit that all the optical and MEMS devices demonstrated are in principle compatible with the established CMOS fabrication methods.
Inspired by the successes of "on-chip" devices on silicon and other materials, the team coined the term "in-chip" devices, as a shorthand descriptor for this new class of components based on direct 3D laser-fabrication. "The possibilities are endless. It is likely that the method will enable entirely new in-chip devices, such as Si-photonics components that can be used for near- and mid-IR photonics, or meandering microfluidic channels that may be used to efficiently cool electronic chips", observed Prof. Ömer Ilday, another co-author of the paper and member of the Electrical and Electronics Engineering and Physics Departments.
As a matter of fact," he continued, referring to follow-up studies the team is working on, "we have already started to show new in-chip architectures and functionalities, such as developing novel in-chip waveguides, laser-slicing of wafers and exploring expansion to other semiconductors."
Other authors of the paper are Ahmet Turnali, Ghaith Makey, Parviz Elahi, Tahir Çolakoglu, Emre Ergeçen, Özgün Yavuz, René Hübner, Mona Zolfaghari Borra, Ihor Pavlov, Alpan Bek, Rasit Turan, Denizhan Koray Kesim, Serhat Tozburun. The work was supported in part by European Research Council (ERC) Consolidator Grant ERC-379 617521 NLL, EU Marie Curie Fellowship 660769 SMILE, and TÜBITAK under project 113M930.