Professor Bain’s Research, consists of three areas


Near Field Optical Transducers (NFTs) for Heat Assisted Magnetic Recording (HAMR)
Oxide Memory Devices (Memristors)
Phase Change Switches for Reconfigurable Electronics

Near Field Optical Transducers (NFTs) for Heat Assisted Magnetic Recording (HAMR)

PIs: James A. Bain, Yi Luo and T.E. Schlesinger
Students: Yunchuan Kong, Matt Chabalko, and Stephen Powell
Research Sponsors: Data Storage Systems Center(DSSC) and Sponsoring Companies

Heat Assisted Magnetic Recording (HAMR) is a candidate for use in magnetic recording systems operating at densities above 2 Tbits/in2. Today’s hard disk drives (HDD’s) operate at around 0.6 Tbits/in2, and are unlikely to reach 2 Tbits/in2 through extensions of the existing technology. The barrier is rather fundamental and is called the superparamagnetic limit. This limit arises with increasing storage density because shrinking the bit to raise the density also makes the bit more vulnerable to being disturbed by ambient thermal fluctuations. Thermal disturbance means that the recorded data will spontaneously disappear as bits flip randomly under thermal agitation. Making the bits out of magnetically “stiffer” material will suppress this thermally induced flipping. However, it will also make the medium too “stiff” to be able to be switched with the recording head. Thus the medium becomes unwritable, despite the fact that anything successfully written would be stable.

HAMR remedies the writability problem by delivering highly localized thermal doses to the medium to “assist” in the writing process. In this way, the very stiff medium is temporally “softened” during writing. One very important challenge in this approach is to get the heat sufficiently localized. A focused laser beam can make a hot spot of around 500 nm in diameter on the surface of a disk, and, indeed, is the technology used in BlueRay optical disks. However, at 2 Tbits/in2, the bit size is about 35 nm x 8 nm, far below what could be heated with a focused optical laser. Rather the hot spot created on the disk needs to be around 25 nm in diameter, a factor of 20 smaller than focused optics will permit.

The figure on the right shows how this might be accomplished using so-called “plasmonic” structures. These metallic thin film structures of Au and similar metals, when encased in a dielectric material, like SiO2, can guide light at the metal dielectric interfaces, along a highly confined path. If the metallic path comes very close to the medium (as the other parts of the recording head do), the size of the optical spot in the medium is dictated by the size of the metallic structures more than the wavelength of the light. Thus, these metallic structures are deemed “Near Field Transducers” since it is necessary to remain in the optical near field to realize such small spots. Although the metals dissipate some of the optical energy as the light is guided, careful design can yield structures that can deliver sub-25 nm hot spots to a medium with an on-chip power efficiency of greater than 20%. This efficiency is defined as power in medium divided by light launched into thin film optical system input port. These NFT devices must be sized carefully, and must be illuminated with correctly conditioned light, which required a systems oriented optical design.


Selected Relevant Publications:

A few publications of the CMU team are given below for reference. An excellent overview is provided by Professor Mark Kryder and his team when he was Senior VP at Seagate [1]. An early paper by the CMU team was done on novel apertures [2]. More recent papers by the team are attached as well [3][4].

[1] M. H. Kryder, E. C. Gage, T. W. McDaniel, W. A. Challener, R. E. Rottmayer, G. Ju, Y.-T. Hsia, and M. F. Erden, “Heat Assisted Magnetic Recording,” Proceedings of the IEEE, vol. 96, no. 11, pp. 1810–1835, 2008.

[2] A. V. Itagi, D. D. Stancil, J. A. Bain, and T. E. Schlesinger, “Ridge waveguide as a near-field optical source,” Applied Physics Letters, vol. 83, no. 22, p. 4474–6, Dec. 2003.

[3] Yunchuan Kong, M. Chabalko, E. Black, S. Powell, J. A. Bain, T. E. Schlesinger, and Yi Luo, “Evanescent Coupling Between Dielectric and Plasmonic Waveguides for HAMR Applications,” IEEE Transactions on Magnetics, vol. 47, no. 10, p. 2364–7, Oct. 2011.

[4] S. P. Powell, E. J. Black, T. E. Schlesinger, and J. A. Bain, “The influence of media optical properties on the efficiency of optical power delivery for heat assisted magnetic recording,” Journal of Applied Physics, vol. 109, no. 7, p. 07B775–07B775–3, Apr. 2011.

See this HAMR.pdf for papers referenced above.



Oxide Memory Devices (Memristors):

PIs: James A. Bain, Paul Salvador (MSE) and Marek Skowronski (MSE)
Students: Mohammad Noman, Yimeng (Ammon) Lu (MSE), Mohamed Abdelmoula (MSE)
Research Sponsors: AFOSR, DARPA MISCIC Center, NSF

Recently, interest in non-volatile resistive switches constructed from oxide thin films have generated significant interest in industry and the scientific community. These devices, sometimes referred to as “Memristors” are relatively simple in construction: typically a thin oxide film sandwiched between two electrodes. Despite their simple structure, however, they display interesting and complex behavior. Of perhaps the greatest interest is the hysteretic I-V characteristics displayed by these devices. The figure below shows results from a Pt/TiO2/Pt device, but many oxide materials display this type of behavior. Both complex oxides like the perovskites (SrTiO3, SrNbO3, etc.) and simple oxides (TiO2, Ta2O5, HfO2, etc) have been demonstrated to show hysteresis.

This hysteretic behavior means that two-terminal devices can be put into two or more different resistance states, simply through the correct application of voltages. Technological applications for these devices include nonvolatile memory (as a possible replacement for FLASH memory) and reconfigurable electronic switches. The mechanism of the switching is understood broadly to be the motion of oxygen atoms under electric field within these devices. When an oxygen atom is removed from the lattice in these devices, the vacancy that is left behind donates electrons to the conduction band in the oxide. Thus oxygen vacancies act as dopants that can be moved under the influence of electric field. Since dopants control electronic properties, the motion of these dopants is what makes the resistance change.

While the interest in these materials is significant, devices demonstrated to date show a significant amount of variability. This is believed to be due to the fact that the large devices that are often tested can display highly localized switching. This behavior is often terms “filamentary” switching in that the regions of high conduction (red in the figure at right) occur in narrow pathways. The formation of these pathways appears to be initiated by defects in the microstructure and can be complicated by elevated temperature due to resistive heating during their formation. The figure at right also shows a thermal image of a hot spot associated with the filament.

Work in this area is, therefore, focused on fabrication and testing of very small devices, such that the device is entirely “filament” and switches uniformly. Devices of 500 nm in diameter fabricated using electron beam lithography are shown in the above figure, and even smaller devices (less than 100 nm in diameter) are under fabrication and testing. The ultimate goal of this work is to identify films and processing conditions, and stressing conditions that give highly reproducible switching in each device and allow device cycling billions of times.


Selected Relevant Publications:

A few publications of are attached for reference. One of the seminal papers in the area (not by CMU
researchers) [1] is attached to provide an overview of the field. Recent publications by the CMU team
are also attached, focusing on local heating within these devices [2], the mobility of oxygen vacancies
[3]and the required mobility for high speed device functioning [4].

[1] D. B. Strukov, G. S. Snider, D. R. Stewart, and R. S. Williams, “The missing memristor found,” Nature,
vol. 453, no. 7191, pp. 80–83, May 2008.

[2] Yi Meng Lu, Wenkan Jiang, M. Noman, J. A. Bain, P. A. Salvador, and M. Skowronski,
“Thermographic Analysis of Localized Conductive Channels in Bipolar Resistive Switching Devices,”
Journal of Physics D: Applied Physics, vol. 44, no. 18, p. 185103 (6 pp.), May 2011.

[3] W. Jiang, M. Noman, Y. M. Lu, J. A. Bain, P. A. Salvador, and M. Skowronski, “Mobility of oxygen
vacancy in SrTiO3 and its implications for oxygen‐migration‐based resistance switching,” Journal of
Applied Physics, vol. 110, no. 3, p. 034509 (8 pp.), 2011.

[4] M. Noman, Wenkan Jiang, P. A. Salvador, M. Skowronski, and J. A. Bain, “Computational
investigations into the operating window for memristive devices based on homogeneous ionic
motion,” Applied Physics A: Materials Science & Processing, vol. 102, no. 4, p. 877–83, Mar. 2011.

See this Memristors.pdf for papers referenced above.


Phase Change Switches for Reconfigurable Electronics:

PIs: T. E. Schlesinger (ECE), James A. Bain (ECE), Gary K. Fedder (ECE), Jeyanandh Paramesh (ECE), Larry Pileggi (ECE)
Students: Mohammad Noman (ECE), Greg Slovin (ECE), Cheng‐Yuan Wen (ECE)
Research Sponsors: DARPA MISCIC Center, SRC

Reconfigurable electronics are of great interest recently because they allow single hardware platforms to adapt to many different applications, even ones that may not have been foreseen when the hardware was constructed. Central to reconfiguring electronics is the ability to have programmable switches. Ideally these switches are non‐volatile, meaning that they can be switches and then do not require continued electrical stimulus to maintain their open or closed state.

Phase change materials (alloys of GeTe, and others) offer the possibility of such a device, in that these materials can be reversibly switched between high resistance (amorphous) and low resistance (crystalline) states through thermal cycling. Challenges within this work are the development of switches with adequate dynamic range, and low enough on‐state resistance to be useful in circuits.

One application of these switches is reconfigurable RF circuits. RF circuits often employ tuned passive elements like LC oscillators. Being able to change the value of the inductance (L) in an oscillator allows it to resonate at a different frequency. This would allow a radio to switch the bands it operates on – from 4G to wi fi, for example. Phase change switches are being employed within this project to switch section of inductor in and out of RF circuits in a non‐volatile way.

In some cases, MEMS probes are needed to reconfigure the devices because of the very low resistance required in the on‐state. Each PC via might offer an on‐state series resistance of 50 Ohms. In this case 50 vias might be needed in parallel to achieve a one Ohm on‐state. Small dense arrays of vias can be individually accessed with MEMS probes. These probes can navigate mechanically to these vias, deform to make electrical contact to the via and deliver an electrical impulse that can heat the vias such that they transform. Ultimately, these probes are small enough and suitable enough for integration that they could be contained within the reconfigurable chip package itself.


Selected Relevant Publications:

A few publications of are attached for reference. These papers show demonstration of the probes[1],
the inductors [2], and the switches [3], along with a study of the interaction between electrodes and
materials [4].

[1] H. Lo, E. Chua, J. C. Huang, C. C. Tan, C.‐Y. Wen, R. Zhao, L. Shi, C. T. Chong, J. Paramesh, T. E.
Schlesinger, and J. A. Bain, “Three‐Terminal Probe Reconfigurable Phase‐Change Material Switches,”
Electron Devices, IEEE Transactions on, vol. 57, no. 1, pp. 312–320, 2010.

[2] C.‐Y. Wen, E. K. Chua, R. Zhao, T. C. Chong, J. A. Bain, T. E. Schlesinger, L. T. Pileggi, and J. Paramesh,
“A Phase‐change via‐reconfigurable On‐chip Inductor,” in 2010 IEEE International Electron Devices
Meeting (IEDM 2010), 6‐8 Dec. 2010, Piscataway, NJ, USA, 2010, p. 4 pp.

[3] E. K. Chua, L. P. Shi, R. Zhao, K. G. Lim, T. C. Chong, T. E. Schlesinger, and J. A. Bain, “Low resistance,
high dynamic range reconfigurable phase change switch for radio frequency applications,” Applied
Physics Letters, vol. 97, no. 18, 2010.

[4] E. K. Chua, L. P. Shi, M. H. Li, R. Zhao, T. C. Chong, T. E. Schlesinger, and J. A. Bain, “Band alignment
between GeTe and SiO2/metals for characterization of junctions in nonvolatile resistance change
elements,” Applied Physics Letters, vol. 98, no. 23, p. 232104 (3 pp.), Jun. 2011.

See this Phase Change Switches.pdf for papers referenced above.