How to flex, bend and stretch

As microprocessors get smaller only nanotechnology can further the miniaturisation of technology. Here we learn how a nanotech approach is helping to bring flexible electronic devices to market   

Today’s information era has been possible through the exceptional advancement of silicon-based technologies. Its excellent semiconducting properties and abundance make silicon the best possible option to produce fast, cheap and long-battery-lasting electronic devices. The incredible processing capabilities of current silicon-based microprocessors are only possible thanks to the extremely high level of integration (ultra large scale integration, ULSI) that has been enabled through mature photolithographic techniques.

Moore’s law has motivated engineers and technologists over almost 50 years, to keep shrinking the transistors inside of microprocessors to the point that today’s fastest microprocessor exhibit devices as small as 22 nm¹, thus falling under the definition of nanotechnology. As the miniaturisation continues, only new and exciting developments in nanotechnology will allow the current technological progress rate.

Recently, a new scientific goal for many has been to expand the rigid nature of modern electronics into a flexible, bendable and stretchable form, having the capability of adjusting its shape for novel applications. Being able to adopt curvilinear shapes will allow integration of electronics into biology, and that will represent a new milestone for medicine to enter into a new era of assisted intervention mediated by flexible electronics. Organic materials represent an interesting development, especially for display applications, though their performance is still far from their inorganic counterpart. Professor John Rogers’ team, from University of Illinois at Urbana-Champaign, has been actively contributing to this recent field taking the concepts of bio-inspired and bio-integration into a reality^2-3. His vision is to combine the outstanding electrical properties of silicon and the flexibility and stretch-ability of polymer-based materials. The key to achieving this is to reduce the dimensions of regular bulk substrates into the nanosize regime, where their bending stiffness allows them to easily bend. Nanotechnology plays again an important role here. Silicon nanoribbons or micro/nanomembranes can be released from silicon-on-insulator (SOI) or Si (111) or (110) substrates (at the expense of a higher material cost compared to the standard Si (100)). A great variety of photonic and electronic devices can be fabricated on top of these nanostructures that are subsequently transferred and “printed” onto polymeric materials such as polymide (PI), or polydimethylsiloxane (PDMS), which serve as flexible mechanical support for the transferred devices^4-6.

The opportunities for bio-integrated and bio-inspired devices are countless, however  the attainable complexity through this approach is still far from the extremely high density, resolution and nanometric alignment, required by electronic gadgets to process the ultra-high data traffic in today’s cloud computing. It is now common to see companies develops new conceptual ideas for flexible, foldable devices, easier than ever to carry around or ultra-portability⁷. Consequently a real opportunity is in flexible high performance electronics for the mobile electronic gadgets industry.

The question is how to bring the nanometric precision of the state-of-the-art silicon-based industry to the flexible electronics paradigm. Many examples can be found that use other types of nanotechnology, such as transistors based on arrays of carbon nanotubes (CNT) or graphene^8,9, nevertheless the density level used by the high performance processing in mobile devices, with billions of cutting-edge, silicon-based, nano-sized transistors, is still not achievable with this approach. Furthermore, since graphene is a semi-metallic and low band-gap carbon nano-sheet, it does not show on/off behavior, which limits its applications to the radiofrequency (RF) regime.

The quest continues and our Integrated Nanotechnology Lab at King Abdullah University of Science and Technology has recently proposed and demonstrated an alternative; transforming the current rigid and opaque silicon substrate into a thin, flexible and semitransparent substrate while keeping the outstanding electrical behaviour of silicon. The process starts with the regular choice for substrate in semiconductor industry, a cheap Si (100) wafer, which then goes through a series of standard complementary metal oxide semiconductor (CMOS) processes to fabricate nanoscaled transistors with ultra-high density and performance. A final step is to peel off a thin layer from the top of the wafer where the devices were fabricated. This peel off process relies on standard microfabrication techniques to first create holes into the substrate by deep reactive ion etching (DRIE), then deposit a resilient protective layer, one atomic layer at a time, through atomic layer deposition (ALD).

Next, selective etching exposes the silicon at the bottom of the deep trenches, so it can be removed by isotropic XeF₂-based etching. Once enough material is removed to form a continuous, buried void, the top portion of the wafer can be released from the substrate and become a flexible host for high performance electronics¹⁰. This method also has the advantage of leaving the remaining of the wafer to be reused after a chemical mechanical polishing (CMP) process, thus appealing from a cost perspective¹¹.

The first logic demonstration consisted on the fabrication of metal oxide semiconductor capacitors (MOSCAP)^12,13, fundamental building blocks for CMOS technology, which help us show that even large-sized devices can be released through our process. Our results showed not only remarkable flexibility but also practically unchanged electrical behavior after releasing of devices. Next, we attempted a full electrical reliability study on metal insulator metal capacitors (MIMCAP), an important step for the development of flexible memories¹⁴. Again, we did not observe any meaningful variation in electrical performance regardless of bending radius, and the reliability study suggested no significant difference in performance of flexible versus nonflexible MIMCAPs and actually shows an interesting improvement in lifetime projections.

The next rational step and with the motto “can we build a truly high performance computer which is flexible and transparent?” we fabricated fast and reliable metal oxide semiconductor field effect transistors (MOSFET)¹⁵. We achieved a remarkable sub-threshold swing of 80 mV/dec (indicator of how quickly a transistor can switch between ON and OFF state), ON/OFF ratio of near 10⁴, a minimum bending radius of 5 mm and an average light transmittance of ~7% in the visible spectrum. Moreover, the transistors feature the most advanced set of materials, high-k/metal gate, currently used in commercial devices. Again, atomic layer deposition (ALD) helps to deposit extremely thin and high quality films to achieve a dielectric (Al₂O₃) thickness of only 10 nm and a metal gate (TaN) of 20 nm. The selection of these materials is critical to address the nanoscaling issues related to unwanted leakage-induced excessive power dissipation, introduced due to short-channel and quantum effects.

What is even more noteworthy is that our procedure can be extended not only to logic devices but also applications in micro electro mechanical systems (MEMS) and energy harvester and storage devices. We were able to fabricate movable thermal actuators exhibiting similar performance before and after bending¹⁶. Additionally we have demonstrated a mechanically ?exible thermoelectric energy generator (TEG) that shows nearly 30 times more power generation than a TEG located on a solid silicon substrate¹⁷. The main reason is the presence of holes so the heat can conduct between two temperature zones using only a fraction of the original fully solid substrate and thus maintaining higher temperature difference in the TEGs. Finally, a thin film based lithium ion battery was fabricated on the flexible silicon platform showing a final storing capacitance of ~1 µAh/cm² and a bending radius that can go down to 0.84 mm (1.18% nominal strain) ¹¹.

These exciting developments are only the tip of the iceberg. Our coming work includes exploring how to expand our methods to include new materials such III-V semiconductors and also produce a fully stretchable silicon substrate by using innovative shapes¹⁸. Furthermore, we are developing a flexible platform for devices with the most advanced 3D architecture and sub-100nm features.  We strongly believe this will open up nanotechnology-enabled, high performance, cost-effective and simple, yet innovative implementations and will help expand the market of transparent, flexible and ultra-mobile electronics.

References

1. http://newsroom.intel.com/docs/DOC-2032
2. J. A. Rogers, T. Someya, Y. Huang, Science 327, 2010, pp. 1603
3. J. A. Rogers, Y. Huang, PNAS 106, 2009, pp. 10875
4. J. Yoon, et al, Nat. Mat. 7, 2008, pp. 907
5. H.-S. Kim, et al, Appl. Phys. Lett. 95, 2009, pp. 183504
6. K. Tae-il, et al, App. Phys. Lett. 102, 2013, pp. 182104
7. http://www.myrolltop.com/what-is-rolltop.html
8. S. J. Kang, et al, Nature Nanotech. 2, 2007, pp. 230
9. S.-K. Lee, et al, Nano Lett. 11, 2011, pp. 4642
10. J. P. Rojas, A. Syed, , 25th IEEE Intl. Conf. MEMS, 2012, pp. 281
11. G. A. Torres Sevilla, et al., The 17^th Intl. Conf. Solid-State Sensors, Actuators and Microsystems, 2013 (accepted)
12. J. P. Rojas, M. M. Hussain, Phys. Status Solidi RRL 7(3), 2013, pp. 187
13. J. P. Rojas, G. A. Torres Sevilla, M. M. Hussain, Appl. Phys. Lett. 102, 2013, pp. 064102
14. J. P. Rojas, et al, IEEE Trans. Elect. Dev. 99, 2013
15. J. P. Rojas, G. A. Torres Sevilla, M. M. Hussain, Scientific Reports 3, 2013, pp. 2609
16. S. Ahmed, et al, 27th IEEE Intl. Conf. MEMS, 2014 (submitted)
17. G. A. Torres Sevilla, et al., Small, 2013 (doi: 10.1002/smll.201301025)
18. A. M. Hussain, et al, MRS Fall Meeting & Exhibit. 2013 (accepted)

Author: Dr Muhammad Mustafa Hussain is Associate Professor of Electrical Engineering at King Abdullah University of Science & Technology (KAUST) in Saudi Arabia