Dr. Vadym Zayets

Abstract:

The Magnetic Tunnel Junction (MTJ) consists of two ferromagnetic metals, which are are separated by an ultra thin dielectric layer. The dielectric layer is sufficiently thin so there is an electron tunneling between the ferromagnetic layers. Still, the dielectric layer is sufficiently thick to prevent any exchange interaction between the ferromagnetic layers and therefore the magnetization of layers can be changed independently.

The resistance of MTJ substantially depends on mutual magnetization directions of the ferromagnetic layers. It is smallest when magnetization directions are parallel and is largest when magnetizations are opposite.

Magneto-transport though a MTJ is unique and is determined by magnetic and tunneling properties of interfacial states at both side of the MTJ

---

 

---

Content

1.

 

 

 

 

---

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

("free" and "pin" layers) Usually one layer of a MTJ is made a magnetically harder and thicker. It is called the "pin" layer and it is harder to its magnetization by a current and it requires a larger magnetic field for its magnetization reversal. The second layer of a MTJ is made a magnetically softer and thinner. Its magnetization can be reversed by an electrical current.

(data storage) A bit of data is stored in the MTJ by means of two stable magnetization directions of the "free" layer.

(reading) The data reading is done be measuring resistivity (a measurement a small current) of MTJ. The resistivity is smaller when the magnetization of "free" layer is parallel to the magnetization of the "pin" layer. The resistance is larger, when the magnetization direction are opposite.

 

(writing)

 

 

 

 

 

 

---

Puzzle of a magnetic tunnel junction

 

The spin conservation law is a very strict law (nearly the same as the energy conservation law). However, a sufficient electrical current flowing in a MTJ, in which the magnetization of both "pin" and "free" layers e.g. in the up direction, switches the magnetization of the "free" layer into the down direction. It means that a magnetic system, which all region spin- polarized in spin-up direction and there is no region of spin-down direction, the electrical current creates a region ( the "free" layer) of spin-down direction. The spin- down direction is created from nowhere. This behavior cannot be explained by the spin injection between layers and the spin conservation law seems to be broken.

(explanation of the effect) The origin of this effect is the unique features of the interfacial states of the MTJ

The magneto-transport through the is complex interactions between spins of spin- polarized conduction electrons, spins of localized d- electrons and the spins of interfacial states at each sides of the MTJ.

 

 

 

 

---

Interface states of MTJ

(fact) The interfacial states mostly determines the magneto- transport and tunneling properties of the MTJ.

Why the interfacial states are so special?

(reason 1) The spin properties of the interfacial states are between spin properties of the localized d- electrons and conduction electrons. Therefore, their spin features are unique and different from both the conduction and localized electrons.

(reason 2) The interfacial couples conduction and localized d- electrons. As a result, the interaction between them substantially increases due to the interfacial states.

Even though in the ferromagnetic metal the conduction and localized d- electrons interacts with each other due to the sp-d exchange interaction and mutual scatterings, in the bulk of metal such interaction is weak. There are two reasons why it is weak. (reason 1: size difference ): The size of localized d- electrons is very small (about diameter of atomic orbital. In contrast, the length of conduction electrons is longer (~ 1nm and longer) (see here). An interaction between two electrons is proportional to their overlap. (reason 2: running vs. steady ): The localized d- electrons do not move along lattice. In contrast, the contrast the conduction electrons moves along the lattice at substantial speed.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

---

 

 

 

 

 

 

 

 

 

 

I am strongly against a fake and "highlight" research