April 13, 2012 — Blogger Michael A. Fury, Techcet Group, reports from the MRS Spring 2012 meeting in San Francisco. Highlights from the fourth day: nanowire FETs, laminate MEMS, nanoparticles in security printing, graphene nanoribbons, Ta2O5 memristors, redox flow batteries, graphene, and more.
Day 4 of the MRS Spring 2012 meeting opened Thursday at Moscone West in San Francisco under sunny skies and a brief rain shower, after a stormy night of heavy rain. No scientists were lost in landslides in the surrounding hills.
AA7.1 Charles Lieber at Harvard University took us on a tour of the frontier between biology and nanotechnology. The field effect transistor (FET) structure is readily adaptable to monitoring basic biological signals. Nanowire FETs can bring us into the regime of 10nm active length detectors with diameters of 3-5nm. Devices have been demonstrated with femtomolar sensitivity to specific cancer markers, and 100-billion-fold discrimination in blood serum separation. Nanowire growth can be periodically slowed to introduce dopants with sharp interfaces, enabling pn junctions along the length of a single wire. These nano devices allow recording in a regime where the active device area is much smaller than the membrane area per ion channel. The wires can be shaped by altering the crystal growth, like a plumber bending pipes at the atomic level. Measurements were shown of electrical activity in a live beating-heart cell. Nanowires can be attached to probe arms and integrated into measurement circuitry for routine measurements, to the extent that probing a living brain or heart will ever be routine. The nanotube probe tip can also be joined directly as the gate of an FET, resulting in a “branched intracellular nanotube transistor” or BIT-FET. Nanotubes 2-3nm long achieve sufficient bandwidth to measure fast action potentials in cells. Extending these concepts to “cyborg tissue” will require fabrication of silicon and nanowire electronics on flexible, macro-porous substrates resembling biological scaffolds so that cells can interpenetrate with the addressable electronic matrix positions. Resistance is futile. Prepare to be assimilated. His stated objective is to blur the distinction between electronic devices, circuits, living cells and tissues.
B3.1 Guann-Pyng Li of University of California Irvine (UC Irvine) described several implementations of laminate micro electro mechanical systems (MEMS) for heterogeneous integrated systems. Mechanical frequency switching is likely to replace integrated circuitry as the band cutoff can be much more sharply defined. Microfluidic systems can be built up and integrated with conventional electronics dropped into place. Organic transistors can be fabricated in place on the laminated PCB substrates for on-board electronics. Vapor deposited pentacene for organic transistors costs $4000/g at the required 99.99% purity for an electron mobility ~3cm2/Vs. Solution deposition of pentacene uses 98% pure material at $20/g and gets you an electron mobility ~3cm2/Vs because solution processing has intrinsic purification properties. The choice is yours…
YY9.3 William Cross of the South Dakota School of Mines described the use of rare-earth-doped nanoparticles in security printing applications. The level of sophistication in covert authentication makes me wonder how counterfeiters can survive any more. Elements mentioned include Er, Yb, Tm and Y, so this technology poses no threat to the chemical mechanical polishing (CMP) CeO2 supply chain.
DD12.4 Iñigo Martin-Fernandez of Lawrence Berkeley National Laboratory (Berkeley Labs or LBNL) developed a method for direct growth of graphene nanoribbons (GNR) for device fabrication using 20nm-wide Ni catalyst patterned with an Al2O3 mask. Ribbons longer than 40µm were demonstrated, and selectivity to growth only on the catalyst was excellent.
E7.5 Antonio Torrezan of HP Labs shared his work on sub-nanosecond switching and energy efficiency in Ta2O5 memristors. A custom 20GHz testing apparatus was used to test the fast switching dynamics of memristors integrated into a coplanar waveguide. Reproducible resistance switching speeds ~100ps were demonstrated for both ON and OFF switching.
O7.6 Giu Yang of PNNL described a strategy for electrochemical energy storage and integration of renewable resources into grid applications. Electric cars address one mode of transportation energy, but only add to the grid capacity problem. Fortunately, an analysis of the usage modes and quality requirements shows that storage technologies can be optimized to address only one mode at a time, resulting in relatively bite-sized capacity to deal with. Redox flow batteries can be developed with a range of chemical components that can be stored separately in huge quantities until the power is needed. Such systems are typically operated at 1V or less to avoid hydrogen generation. Nafion is typically used as a separation membrane, but it allows too much cross-contamination between electrolytes as it is actually designed more for fuel cells. The many oxidation states of vanadium make it a most interesting system for designing redox systems that can be regenerated without requiring chemical separation. There is a long way to go, but the direction is promising.
EE7.1 Taishi Takenobu of Waseda U (Japan) showed some recent work on inkjet printing of single-wall carbon nanotubes (SWCNT) thin-film transistors (TFT). Separation of metallic from semiconducting CNT is required to avoid being trapped by a tradeoff between carrier mobility and on/off current ratio. Earlier work, which concluded that conductivity in high-density semiconducting CNT films was due to residual metallic CNT contamination, is disputed; rather, it is here proposed that it is due to unintentional doping of the CNT during the TFT fabrication process. A flexible ion gel CNT TFT built on polyimide underwent bend testing to a radius of 0.27mm at 178° (basically folded in half) suffered no degradation of on/off current ratio. Another TFT was built on SiO2 using S-CNT for gate and M-CNT for S/D. Performance with SiO2 gate dielectric was poor, while ion gel dielectric was good, indicating once again that CNT density itself is not the controlling factor in inkjet printed devices.
B5.2 Clint Landrock of Simon Fraser U (Canada) developed some autonomous function applications that integrate MEMS, ICs and organic photovoltaics (OPV) cells for low cost, flexible devices. These are fabricated and stored in air with no encapsulation and have shown stability beyond 2 years owing to a novel indium metal OPV cathode. Another component is a fast charging Na+ polymer supercapacitor ~100F/g developed in this group. The current energy cost for the system is estimated at $0.24/kWh, with a roadmap addressing known issues to take it to $0.01/kWh in 3-5 years.
II11.4 Daniel Collins of U Victoria (Canada) walked us through a focused ion beam (FIB) method for fabricating 2D and 3D graphene junctions with graphite that is non-destructive to the extremely delicate monolayer graphene. Fundamental studies will commence now that the fabrication method has been developed.
C10.5 Max Gage of Applied Materials described some modeling work on Cu through silicon via (TSV) CMP via reveal planarization. The model is designed for a single dielectric and circular via, but can be extended to other configurations. Separate equations are developed for each of three polishing regimes: pillar dielectric, Cu not exposed; pillar removal with Cu exposed but field not yet in contact with pad; and final planarization with field in full contact with pad. The model matches well with experimental data while allowing only the effective pattern density as an adjustable parameter.
EE7.6 Andrea Ferrari of U Cambridge spoke on graphene optoelectronics in applications ranging from ultrafast lasers to flexible displays. “Graphene will never replace silicon in our lifetime; that’s stupid.” Good thing. “Graphene Valley” lacks the cachet of “Silicon Valley.” But seriously… Graphene’s properties open up optical detection into the THz range (far IR). Graphene sheets have been fabricated into log normal antennae for effective capture of THz RF radiation, which may be applicable to security applications. The fast laser concept is based on the electroluminescence of graphene oxide, but the device itself remains hypothetical.
F12.3 Matthias Stender of Cabot Microelectronics described an enabling CMP process for GeSbTe (GST) phase-change random access memory (PRAM). GST has a Young’s modulus and shear modulus comparable to but slightly softer than Cu, and it does not form a protective oxide layer upon oxidation. An acceptable balance was found between oxidation and corrosion inhibition to remove the GST film uniformly without changing the GST composition remaining in the cells. Test structures were 30nm wide and 100nm deep in nitride. The slurry was based on colloidal SiO2 with H2O2 oxidizer; other additives are proprietary, though publication of a pending patent application is imminent. A high degree of customization is required for different GST compositions.
J13.10 Sangchul Lee of Gwangju Institute of Science & Technology (Korea) used patterned graphene as a transparent conducting electrode for organic TFT and OPV applications. Ni is used as an etch mask for O2 etch patterning, then the graphene electrodes are integrated with pentacene for an OTFT with reasonable performance, with higher mobility, lower contact resistance and higher injection efficiency than a comparable Au electrode device. Various surface pretreatments had a significant effect on the pentacene morphology and the associated OTFT performance. Fortunately, the highest mobility correlated with the lowest contact resistance. OPV devices benefitted from slight doping of the graphene layer and performed well under bending tests to 5mm radius.
Michael A. Fury, Ph.D. is director and senior technology analyst of Techcet Group.