Periodic Reporting for period 1 - LUCIOLE (Layering, Understanding, Controlling and Integrating Ferroelectric Polar Textures on Silicon)
Période du rapport: 2023-09-01 au 2026-02-28
Ferroelectric materials have been known, and their fascinating properties exploited for more than 100 years. Still, in recent years, exotic polar textures resembling those found in magnetic materials have been unraveled
in nanoscale ferroelectric perovskite oxides, such as flux closure domains, curling polarization, vortices, and
the groundbreaking discoveries of polar skyrmions and merons in 2019/2020. Such topological polar textures (small swirling domains of typically few nanometers) - potentially protected from being unwound - hold huge promise for novel robust topological electronic devices such as ultra-compact (> Tbit/in2) memories. However, so far, nanoscale polar domains have only been studied on special oxide substrates (e.g. SrTiO3), which precludes practical applications.
LUCIOLE takes a revolutionary approach to the field of topological polar textures by bringing them to the realm of the Si platform. Not only is this necessary to implement energy efficient nanodevices, but Si material and advanced CMOS semiconductor technologies offer multiple unexplored ways to engineer mechanical and electrical boundary conditions in ultrathin films and nanostructures of transition metal oxides.
With LUCIOLE, we push the limits in the exploration and future harnessing of emergent states in ferroics
with the following objectives:
• Objective 1: Create monolithically integrated topological polar domains on silicon by molecular beam epitaxy and atomic layer deposition of epitaxial nanoscale ferroelectrics and glassy/nanocrystalline composites using strain, interface, curvature and defect engineering.
• Objective 2: Understand the topological polar textures of proper and improper ferroelectrics created on Si by combined advanced correlative microscopies and spectroscopies at the nanoscale including operando time-resolved measurements under an applied electric field.
• Objective 3: Integrate topological polar textures at the front-end- and back-end-of-line of Si chips in sub- 500 nm two- & three terminal devices with tunable stress /geometry and study their manipulation under electric field by statistically screening at wafer scale their properties to unveil potentially rich behavioral
patterns.
in nanoscale ferroelectric perovskite oxides, such as flux closure domains, curling polarization, vortices, and
the groundbreaking discoveries of polar skyrmions and merons in 2019/2020. Such topological polar textures (small swirling domains of typically few nanometers) - potentially protected from being unwound - hold huge promise for novel robust topological electronic devices such as ultra-compact (> Tbit/in2) memories. However, so far, nanoscale polar domains have only been studied on special oxide substrates (e.g. SrTiO3), which precludes practical applications.
LUCIOLE takes a revolutionary approach to the field of topological polar textures by bringing them to the realm of the Si platform. Not only is this necessary to implement energy efficient nanodevices, but Si material and advanced CMOS semiconductor technologies offer multiple unexplored ways to engineer mechanical and electrical boundary conditions in ultrathin films and nanostructures of transition metal oxides.
With LUCIOLE, we push the limits in the exploration and future harnessing of emergent states in ferroics
with the following objectives:
• Objective 1: Create monolithically integrated topological polar domains on silicon by molecular beam epitaxy and atomic layer deposition of epitaxial nanoscale ferroelectrics and glassy/nanocrystalline composites using strain, interface, curvature and defect engineering.
• Objective 2: Understand the topological polar textures of proper and improper ferroelectrics created on Si by combined advanced correlative microscopies and spectroscopies at the nanoscale including operando time-resolved measurements under an applied electric field.
• Objective 3: Integrate topological polar textures at the front-end- and back-end-of-line of Si chips in sub- 500 nm two- & three terminal devices with tunable stress /geometry and study their manipulation under electric field by statistically screening at wafer scale their properties to unveil potentially rich behavioral
patterns.
In the past two years, we focused on the stabilization and advanced characterization of polar textures in BaTiO3-based systems and on their integration into devices.
We achieved the realization of nanostructures or nanoscale heterostructures on silicon using three different approaches:
(1) Nanostructures (nanodisks) fabricated by a top-down approach
(2) Nanostructures (nanoislands) fabricated by a bottom-up approaches
(3) Superlattices
Nanodisks fabricated by Ne ion milling of epitaxial BaTiO3 thin films on Si: Nanodisks of diameter ~100 nm are shown to host non-trivial polar patterns . Three distinct patterns are observed for the vertical component of the polarization, which correspond to those predicted by theory. Moreover, controllable multilevel polarization states are achieved, which is of particular interest for analogue memory applications.
Nanoislands, with a lateral size as small as of 30-60 nm, embedded in a BaTiO3 continuous film, were synthesized by a modified molecular beam epitaxy (MBE) route. The domains are of center-down convergent type.Compared to a cylindrical shape, the conical shape of the nanoislands (like a funnel) gives rise to a lateral swirling polarization component, which confers chirality. Reversible electrical switching of these topological polar domains is demonstrated. One future challenge is to succeed in stabilizing nanoislands with a single chirality and to “switch” the chirality.
We achieved the first realization of periodic ordered vortices and of polar skyrmion-like bubbles in BaTiO3/SrTiO3 superlattices on silicon. The polar domains are ordered along the <110>BaTiO3 directions in contrast to PbTiO3/SrTiO3 systems, for which the polar states are order along <100>PbTiO3 directions. Strain plays a fundamental role in the stabilization of polar textures in in BaTiO3/SrTiO3 superlattices on silicon.
As part of the tasks for achieving objective 3, we developed the epitaxial growth on SrTiO3-buffered Si of epitaxial La0.7Sr0.3MnO₃ by MBE and epitaxial SrRuO3 by sputtering. LSMO/BaTiO3/LSMO/SrTiO3/Si and LSMO/(BaTiO3/SrTiO3)n/LSMO/ SrTiO3/Si devices exhibit well-saturated ferroelectric hysteresis loops with a coercive voltage Vc of 0.5 V and a remanent polarization Pr of 12 μC cm-2. The large leakage currents in BaTiO3 appear to be a major obstacle for device development. We therefore focused on how to reduce the leakage currents. We demonstrated a huge reduction in leakage currents, by up to seven orders of magnitude, for doping with 5-8 mol% Mn. With this breakthrough, we then used this doping for the growth of Ba(Ti, Mn)O3/SrTiO3 superlattices fabricated on p++ doped Si substrates and achieved closed and fully saturated P-V hysteresis loops (without the need for a conducting oxide bottom electrode!).
We collaborated with the group of Prof. Christophe Gatel / Dr. Martin Hÿtch at CEMES-CNRS in Toulouse, France, for operando (in situ biased) dark field electron holography to study Hf0.5Zr0.5O2 (HZO) ferroelectric tunnel junctions. We have achieved the direct measurement of local electric fields, the visualization of ferroelectric domain switching (on a large field of view of ~ 300 nm) and the quantification of interfacial charges. These results have been obtained thanks to the development of a novel methodology by this group. For the very first time, it has been possible to visualize unambigusouly ferroelectric switching in a ferroelectric film by dark field electron holography.
We achieved the realization of nanostructures or nanoscale heterostructures on silicon using three different approaches:
(1) Nanostructures (nanodisks) fabricated by a top-down approach
(2) Nanostructures (nanoislands) fabricated by a bottom-up approaches
(3) Superlattices
Nanodisks fabricated by Ne ion milling of epitaxial BaTiO3 thin films on Si: Nanodisks of diameter ~100 nm are shown to host non-trivial polar patterns . Three distinct patterns are observed for the vertical component of the polarization, which correspond to those predicted by theory. Moreover, controllable multilevel polarization states are achieved, which is of particular interest for analogue memory applications.
Nanoislands, with a lateral size as small as of 30-60 nm, embedded in a BaTiO3 continuous film, were synthesized by a modified molecular beam epitaxy (MBE) route. The domains are of center-down convergent type.Compared to a cylindrical shape, the conical shape of the nanoislands (like a funnel) gives rise to a lateral swirling polarization component, which confers chirality. Reversible electrical switching of these topological polar domains is demonstrated. One future challenge is to succeed in stabilizing nanoislands with a single chirality and to “switch” the chirality.
We achieved the first realization of periodic ordered vortices and of polar skyrmion-like bubbles in BaTiO3/SrTiO3 superlattices on silicon. The polar domains are ordered along the <110>BaTiO3 directions in contrast to PbTiO3/SrTiO3 systems, for which the polar states are order along <100>PbTiO3 directions. Strain plays a fundamental role in the stabilization of polar textures in in BaTiO3/SrTiO3 superlattices on silicon.
As part of the tasks for achieving objective 3, we developed the epitaxial growth on SrTiO3-buffered Si of epitaxial La0.7Sr0.3MnO₃ by MBE and epitaxial SrRuO3 by sputtering. LSMO/BaTiO3/LSMO/SrTiO3/Si and LSMO/(BaTiO3/SrTiO3)n/LSMO/ SrTiO3/Si devices exhibit well-saturated ferroelectric hysteresis loops with a coercive voltage Vc of 0.5 V and a remanent polarization Pr of 12 μC cm-2. The large leakage currents in BaTiO3 appear to be a major obstacle for device development. We therefore focused on how to reduce the leakage currents. We demonstrated a huge reduction in leakage currents, by up to seven orders of magnitude, for doping with 5-8 mol% Mn. With this breakthrough, we then used this doping for the growth of Ba(Ti, Mn)O3/SrTiO3 superlattices fabricated on p++ doped Si substrates and achieved closed and fully saturated P-V hysteresis loops (without the need for a conducting oxide bottom electrode!).
We collaborated with the group of Prof. Christophe Gatel / Dr. Martin Hÿtch at CEMES-CNRS in Toulouse, France, for operando (in situ biased) dark field electron holography to study Hf0.5Zr0.5O2 (HZO) ferroelectric tunnel junctions. We have achieved the direct measurement of local electric fields, the visualization of ferroelectric domain switching (on a large field of view of ~ 300 nm) and the quantification of interfacial charges. These results have been obtained thanks to the development of a novel methodology by this group. For the very first time, it has been possible to visualize unambigusouly ferroelectric switching in a ferroelectric film by dark field electron holography.
- The achievement of polar vortices and skyrmion-like bubbles periodically ordered in BaTiO3/SrTiO3 superlattices on silicon had never been reported so far, nor on silicon, nor on oxide substrates. This result significantly advances the state of the art in polar textures on silicon.
- The realization of chiral polar textures in epitaxial nano-islands on silicon (as small as 30-60 nm in lateral dimension), represents another breakthrough. This result has been the result of a modified growth route by molecular beam epitaxy where the first stage of the growth (passivation of the Si substrate by Sr) has been tweaked. A press release has been published on this work by HZB (https://www.helmholtz-berlin.de/pubbin/news_seite?nid=28946&sprache=en(s’ouvre dans une nouvelle fenêtre)). We draw a parallel between the center-down-convergent polarization field in the truncated nanoislands, whose shape resembles an inverted narrowing funnel and the flow of water in such a funnel. The image provided below (designed by Laura Canil) illustrates the analogy of the texture in the nanoislands resembling a swirling vortex of liquid flowing into a narrowing funnel.
- To the best of our knowledge, evidencing a ferroelectric polarization in thin films by electron holography has never been achieved, and our work with CEMES-CNRS in France, where a new methodology was developed, puts an end to 30 years of unsuccessful attempts.
- The realization of chiral polar textures in epitaxial nano-islands on silicon (as small as 30-60 nm in lateral dimension), represents another breakthrough. This result has been the result of a modified growth route by molecular beam epitaxy where the first stage of the growth (passivation of the Si substrate by Sr) has been tweaked. A press release has been published on this work by HZB (https://www.helmholtz-berlin.de/pubbin/news_seite?nid=28946&sprache=en(s’ouvre dans une nouvelle fenêtre)). We draw a parallel between the center-down-convergent polarization field in the truncated nanoislands, whose shape resembles an inverted narrowing funnel and the flow of water in such a funnel. The image provided below (designed by Laura Canil) illustrates the analogy of the texture in the nanoislands resembling a swirling vortex of liquid flowing into a narrowing funnel.
- To the best of our knowledge, evidencing a ferroelectric polarization in thin films by electron holography has never been achieved, and our work with CEMES-CNRS in France, where a new methodology was developed, puts an end to 30 years of unsuccessful attempts.