104
S. Krishnia and W. S. Lew
Understanding the interactions between electric current and the DWs, a
phenomenon that was first investigated more than 30 years ago in magnetic thin
films [11, 12], has become of enormous interest over the past decade. Although,
there are several possible ways in which the current can interact with the DWs,
however, the interactions those can drive the DWs, are of the most interest to develop
future spintronic devices. Amongst the key emerging memory technologies, spintransfer torque magnetic random access memory (STT-MRAM) is the predominant
non-volatile memory, expected to have infinite endurance along with the ability to
scale down below 20 nm.
1
Though there is no inherent magnetic degradation in STT-MRAM, there is
however an electrical breakdown of the MgO tunnel barrier that causes rapid aging
of the device. Also, the read and write paths are identical in the STT-MRAM devices
that may cause the unintentional writing by read current. To avoid this issues, the
write voltage must be kept low, below 400 mV [13]. A solution for this is the SOTbased MRAM—a three terminal device that separates the read and write path [14,
15]. However, SOT-MRAM poses scaling issue to low dimensions. Moreover, the
fabrication and process steps in MRAM are complex.
A possible memory alternative which combines the positive attributes of MRAM
along with having simplified fabrication steps and avoiding the usage of tunnel barrier
is DW based memory devices. The first prototype of the DW based memory device
was proposed by IBM scientists [16]. The information in DW devices are stored in
the form of magnetic domains that are separated by DWs [7, 10, 17]. The DW devises
also offer as a higher speed and reliable alternative of hard-disk drives (HDD) since
it does not have moving parts. On one end the bits may be written by changing the
magnetization directions in the nanowire. Operation of the DW devices is analogous
to a non-volatile shift register. The bits can be pushed by using electric current
pulses and read using a magnetic tunnel junction or Hall probe as the sensor. An
operational mechanism of DW memory device is illustrated in Fig. 1. The device
speed and reliability depend on how fast the DWs can be shifted in the nanowire i.e.
the DW velocity, while the storage density is determined by how closely the DWs can
be placed within a nanowire as well as how closely the nanowires can be placed. The
DW devices offer a possibility to re-orient the structure in the vertical direction and
store the bits in a three-dimensional geometry, thereby increasing storage density.
Therefore, controlled nucleation, high speed motion and detection of the DWs are
the key prerequisites for making the memory devices feasible.
2
1 Handbook of Magnetism and Advanced Magnetic Materials. Edited by Helmut Kronmüller and
Stuart Parkin. Volume 5: Spintronics and Magnetoelectronics.2007, John Wiley & Sons, ISBN:
978–0-470–02,217-7.
2 Introduction to Magnetic Random-Access Memory. Edited by Bernard Dieny, Ronald B. Goldfarb,
and Kyung-Jin Lee. 2016, Wiley-IEEE Press, ISBN: 978–1-119–00,974-0.
S. Krishnia and W. S. Lew
Understanding the interactions between electric current and the DWs, a
phenomenon that was first investigated more than 30 years ago in magnetic thin
films [11, 12], has become of enormous interest over the past decade. Although,
there are several possible ways in which the current can interact with the DWs,
however, the interactions those can drive the DWs, are of the most interest to develop
future spintronic devices. Amongst the key emerging memory technologies, spintransfer torque magnetic random access memory (STT-MRAM) is the predominant
non-volatile memory, expected to have infinite endurance along with the ability to
scale down below 20 nm.
1
Though there is no inherent magnetic degradation in STT-MRAM, there is
however an electrical breakdown of the MgO tunnel barrier that causes rapid aging
of the device. Also, the read and write paths are identical in the STT-MRAM devices
that may cause the unintentional writing by read current. To avoid this issues, the
write voltage must be kept low, below 400 mV [13]. A solution for this is the SOTbased MRAM—a three terminal device that separates the read and write path [14,
15]. However, SOT-MRAM poses scaling issue to low dimensions. Moreover, the
fabrication and process steps in MRAM are complex.
A possible memory alternative which combines the positive attributes of MRAM
along with having simplified fabrication steps and avoiding the usage of tunnel barrier
is DW based memory devices. The first prototype of the DW based memory device
was proposed by IBM scientists [16]. The information in DW devices are stored in
the form of magnetic domains that are separated by DWs [7, 10, 17]. The DW devises
also offer as a higher speed and reliable alternative of hard-disk drives (HDD) since
it does not have moving parts. On one end the bits may be written by changing the
magnetization directions in the nanowire. Operation of the DW devices is analogous
to a non-volatile shift register. The bits can be pushed by using electric current
pulses and read using a magnetic tunnel junction or Hall probe as the sensor. An
operational mechanism of DW memory device is illustrated in Fig. 1. The device
speed and reliability depend on how fast the DWs can be shifted in the nanowire i.e.
the DW velocity, while the storage density is determined by how closely the DWs can
be placed within a nanowire as well as how closely the nanowires can be placed. The
DW devices offer a possibility to re-orient the structure in the vertical direction and
store the bits in a three-dimensional geometry, thereby increasing storage density.
Therefore, controlled nucleation, high speed motion and detection of the DWs are
the key prerequisites for making the memory devices feasible.
2
1 Handbook of Magnetism and Advanced Magnetic Materials. Edited by Helmut Kronmüller and
Stuart Parkin. Volume 5: Spintronics and Magnetoelectronics.2007, John Wiley & Sons, ISBN:
978–0-470–02,217-7.
2 Introduction to Magnetic Random-Access Memory. Edited by Bernard Dieny, Ronald B. Goldfarb,
and Kyung-Jin Lee. 2016, Wiley-IEEE Press, ISBN: 978–1-119–00,974-0.