Samsung Open Innovation Program For STT-MRAM Technology

Interview conducted by G.P. ThomasSep 17 2013

Alexander Driskill-Smith, Senior Director of Strategic Memory Planning at Samsung Electronics, talks to AZoM about the Samsung open innovation program in STT-MRAM (SGMI), STT-MRAM technology and materials, and the need for university collaborations in this area.

Could you briefly explain the Spin-Transfer Torque (STT) effect that occurs in magnetic tunnel junctions and spin valves, and outline what material properties and conditions are required for this to occur?

When a spin-polarized current of electrons passes through a magnetic layer, it transfers spin angular momentum to that layer which results in a torque on the magnetization. This is known as the spin-transfer torque (STT) effect. Essentially, a torque is generated through the transfer of spin momentum. The torque excites oscillations in the magnetization and if strong enough can cause the magnetization direction to flip.

In the case of a spin valve or magnetic tunnel junction (MTJ), a magnetic layer with a fixed direction of magnetization is used to generate the spin-polarized current: it acts as a spin filter passing only those electrons whose spin is aligned with its magnetization direction. The spin-polarized current then passes through and generates a torque in a thinner magnetic layer whose magnetization is free to rotate: if sufficiently strong, the torque causes the magnetization of this layer to switch to the opposite direction.

Several conditions are required for STT switching to be observed. First, the magnetic layers need to be sufficiently thin, typically a few nanometers, otherwise spin scattering causes the spin polarization to be lost. Second, the lateral dimensions of the magnetic layers need to be sufficiently small, typically less than 100 nanometers, otherwise the magnetic field generated by the high current density passing through the layers interferes with the STT effect.

And third, the damping or rate at which the magnetization relaxes to its equilibrium position needs to be sufficiently small, otherwise the STT effect cannot overcome the energy dissipation due to spin-orbit interactions and other effects.

Spin-polarized current and spin-transfer torque (STT) switching. M1 exerts a torque on incoming electrons that become spin-polarized in the M1 direction. This spin-polarized current in turn exerts a torque on M2, causing M2 precession and switching.

Could you explain how Spin-Transfer Torque is utilised in Magnetic Random Access Memory (STT-MRAM) technology and give a brief history of the research in this area?

The main component of STT-MRAM is the magnetic tunnel junction (MTJ), which is a sub-100 nanometer size magnetic element consisting of two magnetic layers separated by a thin insulating layer or tunnel barrier. The information is stored in the magnetic state of one of the magnetic layers, called the free layer.

A second magnetic layer called the reference layer provides a reference frame required for reading and writing. STT-MRAM functionality is powered by two phenomena discovered within the last two decades: the tunneling magnetoresistance (TMR) effect for reading and the spin-transfer torque (STT) effect for writing.

The TMR effect causes the resistance of the MTJ to depend significantly on the relative orientation of the magnetic layers: the resistance in the antiparallel state can be several times larger than in the parallel state. It enables the magnetic state of the free layer to be sensed and thus stored information to be read. The STT effect, as discussed above, enables the magnetic state of the free layer to be changed if the torque is sufficiently strong, thus information can be written.

The STT effect was theoretically predicted independently by Slonczewski and Berger in 1996 and first observed in all-metallic spin valve structures in 2000. Due to the relatively low degree of spin polarization in such metallic structures, switching currents were initially very high. Subsequent research therefore focused on reducing switching currents, in particular through the development of MTJ structures which have greater spin polarization than metallic spin valves.

The most significant advance was the introduction of magnesium oxide (MgO) tunnel barriers which exhibit a giant TMR arising from a symmetry-based spin filter effect, which was theoretically predicted in 2001 and first observed in MTJs in 2004. Due to the improved polarization, a much reduced STT switching current in such MTJs incorporating MgO tunnel barriers was demonstrated in 2005. MgO-based MTJs are now the standard for the field.

Magnetic Tunnel Junction. Resistance is low when the magnetization of the reference and storage layers are aligned in the same direction (“0” state), and high when the layers are aligned in opposite directions (“1” state).

First-generation MRAM cell. A magnetic field, generated by the bit line and write word line, is used to switch between the “0” and “1” states. (b) STT-MRAM cell. By eliminating the write word line and bypass line, it is considerably smaller than the first-generation MRAM cell.

What are the benefits and potential applications of STT-MRAM technology?

The benefits of STT-MRAM are that it can have the density of DRAM, the speed of SRAM and the non-volatility of Flash, as well as unlimited endurance and moderate to low power consumption. While it is challenging to meet all these requirements at the same time, STT-MRAM can have many potential applications in both the standalone and embedded memory spaces by combining the capability of existing memory technologies with additional functionality.

For example, an STT-MRAM with the density of DRAM and the non-volatility of Flash enables radically better performance and lower power consumption in data intensive applications as well as instant-on capability in mobile applications.

It also offers significant architectural benefits due to the unique combination of high speed, non-volatility, unlimited endurance and random access capability. A further benefit of STT-MRAM is that it provides greater scalability to future technology nodes, particularly when compared to DRAM.

What are the main challenges associated with developing STT-MRAM for commercial use? For instance, what are some of the current limitations in design, manufacturing and materials?

There are challenges in all these areas. On the materials side, the focus has shifted in the past several years from in-plane MTJ development (where the magnetization is in the plane) to perpendicular MTJ development (where the magnetization is perpendicular to the plane, which enables greater scalability) and there is still a need to develop perpendicular MTJ stacks with better properties: for example, higher TMR to enable faster readout, higher spin polarization to enable smaller switching current, and higher perpendicular anisotropy to enable greater thermal stability.

On the design side, there is a need to develop better read sensing circuits so that the difference between the two resistance states can be read faster.

And on the manufacturing side, the biggest challenge is fabricating the MTJs: etching the magnetic materials that make up the MTJs without causing damage or redeposition of unwanted material on adjacent MTJs and doing so with the required uniformity is difficult.

How is STT-MRAM technology unique compared to other memory technologies?

Of all the major existing and prototype memory technologies, STT-MRAM is the only one that has the capacity, endurance and speed of working memory (DRAM and SRAM) as well as the non-volatility of storage memory (Flash and HDD). No other memory technology offers this unique combination of attributes.

As discussed above, it can be challenging to meet all these requirements at the same time and there are some trade-offs involved, but in general STT-MRAM’s attributes can be tuned depending on the requirements of a particular target application.

Could you introduce Samsung’s new open innovation program in MRAM - SGMI?

SGMI stands for Samsung Global MRAM Innovation. It is a new program for funding research into STT-MRAM that is open to the world’s leading universities and research laboratories. SGMI is a special event of the Global Research Outreach (GRO) program which Samsung has conducted since 2009: while the GRO covers a wide range of topics from next-generation computing to medical devices, SGMI is focused purely on STT-MRAM. The SGMI program is designed to create opportunities – at colleges, universities and research laboratories around the world – for exploring breakthrough and innovative STT-MRAM research.

The SGMI call for proposals was issued in June and the proposal submission period runs through September 28. There are 32 research topics of interest grouped into 9 research themes covering all aspects of STT-MRAM technology from magnetic materials, modeling and device characterization to process integration, circuit design and system architecture and applications.

There is also an emerging technology theme which covers new effects such as spin-orbit induced effects and voltage-controlled magnetic anisotropy as well as multi-bit and 3D stacked STT-MRAM technology which may become important in future generations of STT-MRAM.

We invite all interested researchers to submit their best and most novel ideas on any of these compelling research subjects prior to the September 28 deadline. Joint research proposals from multiple universities and research laboratories can also be submitted.

The proposals will be reviewed in the fourth quarter with the goal of announcing the results in December and launching the research collaborations in January 2014. Projects will be funded initially for one year and can be renewed for up to three years.

What does Samsung hope to achieve through this program?

The overall goal is to build mutually beneficial research relationships with the world’s leading universities and research laboratories that will support and sustain the commercialization and long-term future of STT-MRAM. There are challenges that need to be overcome in the near term in order to commercialize high-density STT-MRAM, plus there needs to be a constant pipeline of new magnetic materials, devices and designs that can be incorporated into future generations of STT-MRAM at smaller technology nodes.

In the 21^st century, no single company can do all its research alone and we see it as critically important that we partner with leading universities and research laboratories across the world to build and strengthen a vibrant research community in magnetics and spintronics.

What does it take to bring a new memory technology like STT-MRAM to market?

In addition to the technology research and development which of course is essential, it is also critical to find the right applications and markets for a new technology and to prepare them in advance. There are many examples of technologies being developed in the past which have never found a market. Even a large company like Samsung cannot do all this alone and needs collaboration partners.

It is not just university and research laboratory collaborations that we are addressing with the SGMI program, but also partnerships with other companies like system integrators that will incorporate STT-MRAM into their future products and software providers that will make changes to their code to take advantage of STT-MRAM’s unique capabilities. All these need to come together for a successful market launch.

Why is it so important for universities and research institutes to get involved with the SGMI program and why is it mutually beneficial?

For universities and research institutes, we hope participation in the SGMI program is just the first step in a long and mutually beneficial collaborative relationship with Samsung. And for Samsung, it is not just about the research outcomes but also about taking the worldwide research activity in STT-MRAM and future magnetics and spintronics technologies to a new level.

The more research there is in this area, the more secure STT-MRAM’s future will be, which in turn will lead to greater funding and deeper and longer-term collaborations in the future. It is a virtuous circle in which everyone in the magnetics and spintronics community benefits.

How do you see STT-MRAM technology progressing over the next decade and how does Samsung hope to be involved in this?

With the help of our collaboration partners, we believe that high-density STT-MRAM can be brought to market in the next few years, most likely as a product that combines the capability of existing memory technology like DRAM with additional functionality like non-volatility and low standby power consumption. As discussed earlier, such an STT-MRAM product would enable radically better performance and power consumption in data intensive and mobile applications.

As to how STT-MRAM technology progresses over the decade, it will be interesting to see how some of the emerging spintronics technologies like voltage-controlled magnetic anisotropy and spin-polarized current generation by spin-orbit induced effects develop because if their potential is fully realized they can transform the capability of STT-MRAM in the future.

Perhaps the most exciting thing is bringing magnetism and the spin degree of freedom to conventional semiconductor electronics which opens up a host of new physical phenomena that can be utilized in new STT-MRAM applications.

In the long term, the true potential lies in moving beyond the constraints of conventional memories and introducing new functionality based on these novel spintronics-based phenomena (and others that still await discovery) – an open playground that we have barely started to explore.

Lastly, what are the upcoming important dates for the SGMI program and how can people get involved?

The SGMI proposal submission deadline is September 28. Full details on the research themes and proposal submission process can be found on the SGMI website at www.samsung.com/mram and any questions can be directed to [email protected]. We very much look forward to everyone’s participation!

About Alexander Driskill-Smith

Alexander Driskill-Smith is Senior Director of Strategic Memory Planning at Samsung Electronics. He has more than ten years’ experience in business development, strategic planning and advanced technology research and development in the semiconductor and data storage industries and holds several patents.

He joined Samsung in 2011 through Samsung’s acquisition of Grandis, a start-up company that developed a disruptive non-volatile magnetic memory solution, where he served as Vice President of Business Development. Prior to joining Grandis, he worked at IBM Corporation and Hitachi Global Storage Technologies, where he led the design, fabrication and integration of several new magnetic recording head designs.

He holds M.A. and Ph.D. degrees in Physics from the University of Cambridge, U.K., and an M.B.A. degree from the Wharton School of the University of Pennsylvania. He currently serves on the San Francisco Advisory Board of Junior Achievement of Northern California.