Magnetic Memories:

Supported by BMBF grant: FKZ 13N10144

Prof. Dr. Johann W. Bartha (TU Dresden)

Prof. Dr. Karl Leo (TU Dresden)

Dr. Jürgen Rüstig (Namlab GmbH)


Dr. Roland Mattheis (IPHT Jena)

Collaborative Project

In recent years the down scaling of memory cells has proceeded at an enourmos pace due to the ongoing demand of highly integrated volatile and non-volatile random access memory cells. Extrapolating this development to the near future, we face the problem that memories based on charge storage will reach physical limits since either the area for the stored charge (DRAM) or the charge itself (Flash memory) becomes too small.

The aim of the project is to evaluate three new memory concepts:

  • resistive memory (TU Dresden, Namlab GmbH)
  • magnetic memory; MRAM, spintronic (FZ Dresden-Rossendorf)
  • organic memory (TU Dresden)

Magnetic Memory

Compley oxides are playing a fundamental role for the creation of magnetic tunnel junctions, where two magnetic layers are separated by an ultrathin insulating barrier. For typical metal-based memory cells (MRAM – magnetic random access memories) nowadays Al2O3 or MgO are used as a tunnel barrier. The preparation of an ultrathin tunnel barrier with either magnetic semiconducting or organic electrodes we define as a milestone for the realization of new tunnel magnetoresistance cells. In order to optimize the storage capacity a barrier material with a large dielectric constant and a perfect interface quality has to be realized. Since the discovery of the giant magnetoresistance (GMR) effect metal based spintronics has evolved at a rapid pace. End of the 1990th the magnetic RAM (MRAM) was proposed. In 2006 Freescale Semidocutor has introduced MRAM to the market. The major advantages are the non-volatile character of the memory which leads to energy efficient cells and the fast read/write schemes in a simple geometry, which allow for high packing densities. The major drawback is the difficult means of addressing the individual bit using two pulsed magnetic fields. Recently, new concepts based on current-induced switching based on the so-called spin-transfer-torque effect are used to circumvent the problems in bit addressing. Semiconductor spintronics is a relatively young research area with a number of new concepts for memory devices, which are addressed in parallel. Magnetic semiconductors exhibit in comparison to metals extremely long spin coherence times and controllable electrical (charge carrier concentration) and optical (fundamental band-gap) properties. Due to the potentially high spin polarization magnetic, semiconducting oxides, i. e. 3d metall doped oxides, are promising candidates for the application in new tunnel magnetoresistance structures within the area of semiconductor spintronics. If it is possible to create ferromagnetic semiconductors with a robust ferromagnetism and a high Curie temperature, the interaction between charge carriers and ferromagnetism can be controlled via a gate voltage. One goal of the project will be to replace classical ferromagnetic metal electrodes by ferromagnetic semiconducting electrodes, with the advantage of longer spin coherence times.


Kerstin Bernert
Chris Bunce
Jürgen Fassbender
Maciej Oskar Liedke
Kay Potzger
Heidemarie Schmidt
Ilona Skorupa
Sebastian Wintz
Shengqiang Zhou