HANKUK UNIVERSITY OF FOREIGEN STUDIES
Project-1: Evolution of conducting Filament in SrFeOx using Raman technology and graphene top electrode
In Aug 2019, Our group reported local detection of conducting filament in SrFeOx based ReRAM[Adv. Mat. 1903391 (2019)]. In this report, we revealed the different size of the conducting filament and the composition of the conducting filament. For this purpose, we used a spectrometer with a submicron resolution at a synchrotron facility. We set the device at a high resistance state(HRS) and a low resistance state (LRS). Then we got the spectrum after delaminating the top electrode.
We found that the electric-field-induced phase transition spreads over a large area in (001) oriented SrFeO2.5 devices: (SFO001), where oxygen vacancy channels are ordered along the in-plane direction of the device. In contrast, (111)-grown SrFeO2.5 devices (SFO111) with out-of-plane oriented oxygen vacancy channels, reaching from the bottom to the top electrode, show a localized phase transition.
Two months after the above our report, Oct 2019, Pennycook who is a top group in TEM field reported another paper at Adv. Mat. [Adv. Mater. 1903679 (2019)]. They reported that Perovskite SrFeO3 nanofilaments are formed and extend almost through the Brownmillerite SrFeO2.5 matrix in the ON state and are ruptured in the OFF state, unambiguously revealing a filamentary RS mechanism. However, the nanofilaments weree ≈10 nm in diameter, which is quite inconsistent with our results. While our study showed both SFO (001) and SFO(111) devices, this work covered only SFO(001) devices. Also, their filament at the TEM image did not directly touch the electrode. So the debate needs to be cleared. [Nontheless, the try of 180 nm-size Au nanoparticles as the top electrode was meaningful progress.]
Elías Ferreiro‐Vila et. Al., demonstrated a transformation between Perovskite and Brownmillerite SrFeOx with Sub‐Micrometer Spatial Resolution. [Adv. Funct. Mater. 2019, 29, 1901984]
In the report, the electric field produced by a voltage‐biased atomic force microscopy tip was used to induce such transformation between PV SrFeO3−δ and BM SrFeO2.5 at room temperature and with sub‐micrometer spatial resolution.
The evolution of the room temperature Raman spectra for different voltages is shown in the left Figure. The appearance of the peaks at ≈331, ≈430, and ≈625 cm−1 in the pristine regions of the film are characteristic of a tetragonal or orthorhombic PV, of composition SrFeOx, x ≈ 2.7–2.9.28 The intensity of these peaks decreases gradually and shows a blue‐shift in the regions scanned with an increasing voltage, at the same time that a broad maximum with a shoulder at Raman shift ≈710 cm−1 characteristic of the stretching modes of FeO4 tetrahedra in the BM phase develops. The comparison with the Raman spectrum of a BM SrFeO2.5 film shown in the same figure suggests the full transformation of the PV into BM above ≈−9 V.
Our proposal
-
[Cowork with Prof. Gyu-Chul Yi and Prof. Jae-Geun Park at Physics department of Seoul National University; Confocal Raman Microscopy] We will use a graphene top electrode for SrFeOx device made on LSAT substrate.
-
We can study Raman spectrum at On/Off state of SrFeOx device.The Raman peaks of graphene does not overlap with SrFeO3 and SrFeO2.5.We can see the evolution of the phase below the graphene electrode directly and the switching condition is nearly similar to the real devices.
-
The switching with AFM tip has another artifact which we have demonstrated in our earlier report. [Scientific Report, (2019) 9:118]
Project-2: Nano-scale local control of magnetic/superconducting properties of Top electrode using Oxygen ion injection using conducting filament in SrFeOx layer
Control of Nanoscale magnetic/superconducting properties has long been a really hot topic. [Very famous research scandal on this topic can be found here. J. H. Schön et al. Science 2000;288:656-658]
Our proposal
-
Use of Topotactic ReRAM for control of oxygen injection and dejection into-from the top electrode
-
SFOx layer will give us the merits of reversible and fast oxygen control.
-
Synaptic memory of SFO will give delicate control of oxygen content.
-
We demonstrated the synaptic memory of SFO this year. ACS Applied Materials & Interfaces 2020 12 (37), 41740-41748.
Local measurements of the magnetic properties such as magnetization, susceptibility, and nonlinear response of a sample by moving a scanning squid microscopy (SSM)over the surface of the sample
[scanning squid microscopy (SSM) at X. Renshaw Wang et al. Science 2015;349:716-719]
Spatial resolution: ~ 1 mm
Higher sensitivity: ~ 1x10-10 T/(Hz)1/2 for pickup coil loop dia 7 mm
Non destructive technique.
Project-3: Control of vacancy density and its effect on switching properties.
-
SrFeO2.5 has a brownmillerite structure which consists of a fully oxygenated Octahedral layer and tetrahedral layer with oxygen vacancy sites. The motion of oxygen through this vacancy site is key for the resistance switching.
-
Cowork with Steven May at Drexel university
-
Since Fluorine ion has -1 valence, F doping at O site will control the vacancy density.
-
We will study ReRAM switching will SFOx doped with F ion.
-
We expect to know more about the switching mechanism and get more control of ReRAM switching on this material.
Project-4: Upgrade of ReRAm of SrFeOx.
Though we verified a very high speed and stability in the resistance switching, SrFeO2.5 based ReRAM still has many things to be upgraded for real commercialization. For low power operation, we first tried a lower compliance method. But this resulted in reduced performance. Thus we will try easier supply of oxygen using heterostructure such as BM-SFO/SrFeO3 with some thin buffer layer between them. SrFeO3 can be used as an oxygen source as well as a top electrode. A similar method has been tried for TiO2 based ReRAM. [Nature nanotechnology volume 3, 429 (2008).]
The try of 180 nm-size Au nanoparticles as the top electrode by Pennycook was meaningful progress.[Adv. Mater. 1903679 (2019)]. Here they found that high ON/Off ration in addition to millions of switching with the narrower distribution of switching parameters.
We will try submicron size devices down to about 50 nm to see the stability of ReRAM switching in the nano-deices. [ongoing long-term cowork with Prof. Gyu-Chul Yi of SNU physics department]
Project-5: Compatibility with Si-technology
We will make SrFeOx device on top of SrTiO3 buffered Si, or Pt/SiO2/Si and study the ReRAM switching in more-commercial flatform.