Final Report Summary - IFOX (Interfacing Oxides)
Executive Summary:
The goal of IFOX was to explore, create and control novel electronic and magnetic functionalities, with focus on interfaces, in complex transition metal oxide heterostructures to develop the material platform for ‘More than Moore’ (MtM) and ‘beyond CMOS’ electronics, VLSI integratable with performance and functionality far beyond the state-of-the-art.
In order to reach these goals an integrated effort was started which included theory, growth, processing, analysis and testing whereas at the same time the potential impact and IP was addressed. Theoretical modelling was developed and used to identify new oxide materials and interfaces with enhanced functionality to inspire the growth of new structures. Growth processes for these materials and heterostructures were developed and optimized on academic and industrial scale, to test growth of and ultimately provide growth processes for oxides which can be applied in industry. The processes were demonstrated to be able to control the growth with single atomic layer precision. In parallel, growth on silicon substrates using suitable buffer systems was investigated for future CMOS compatibility. As a result several major advances were reported including for example metal organic aerosol deposition of oxides with in situ ellipsometry and large area pulsed laser deposition and pulsed plasma deposition systems for four and three inch wafers, respectively, the latter ready for sale by the developing company. This was complemented by the parallel development of processing technologies that up to now were unavailable for these materials or unsuitable for industrial application. Characterization of the heterostructures was a major effort, which encompassed electrical, magnetic, optical and structural characterization down to the atomic scale. In addition, major breakthroughs in characterization techniques were achieved: for example ultra-high resolution transition electron microscopy and high speed magneto-optics have made a leap which now allows much deeper insight in oxide structures and magnetic properties on the sub nm and ps length and time scale, respectively. Characterization and growth both provided feedback to theory, showing which model systems or functionalities can be realized, information which again could be used by theory to direct the effort towards the most promising and most supporting systems and tasks at hand.
Because IFOX targeted industrial applications, the industry partners supervised the work and evaluated the results. Based on these evaluations, after three years a list of the seven most promising oxide heterostructures with functionalities targeting the research areas of IFOX was established and put into a new description of work. The final period of IFOX was completely dedicated to the realization of these heterostructures, including to grow them on large area silicon substrates. For two of the seven heterostructures functionality could be demonstrated indeed on large area silicon substrates. For the others (except one) growth on silicon was performed and at least partial functionality could be demonstrated. This effort was supported by a new partner (Protec Surface Technology) who specializes in large area industrial deposition systems and supported the effort of upscaling to large area substrates. All seven heterostructures were finally assessed in terms of maturity down to the technology readiness level of growth and processing and competitiveness in a targeted field of application.
As an important result, all data on growth and processing were collected in several web-based interactive databases and evaluated by the industrial partners. From these data three catalogues were established, one on growth processes, one on processing and one on possible environmental impact of chemistry which allow for immediate assessment of needs and risks when transferring these heterostructures to industrial application.
It should finally be mentioned that all deliverables of the project were achieved.
Project Context and Objectives:
The goal of IFOX was to explore, create and control novel electronic and magnetic functionalities that result from the rich interplay of charge, spin and orbital degrees of freedom in transition metal oxide heterostructures and their interfaces. The desired outcome was the establishment of a material platform for novel ‘More than Moore’ and ‘beyond CMOS’ electronics, compatible with VLSI and which delivers performance and functionality far beyond the state-of-the-art. The consortium should:
(1) establish a theoretical basis to identify the most promising materials/heterostructures and to understand the new functionalities relevant for applications;
(2) grow oxide films on commercial substrates with a quality comparable to state-of-the-art semiconductor growth;
(3) establish oxide patterning and processing conditions within and beyond the constraints of current fabrication technologies; and
(4) characterize structural, electronic and magnetic properties as a prerequisite to delivering novel device concepts for memory, programmable logic and sensors.
A multidisciplinary consortium was formed to address these challenges and build on complementary expertise in theory and modelling, epitaxial growth, processing, characterization, device fabrication and knowledge transfer. Crucially, technological relevance was guaranteed by the active participation of a world-leading company: Fiat and two SMEs, namely TSST and Organic Spintronics. A third SME, Protec Surface Technology, later joined the consortium to strengthen the effort of upscaling the oxide technology towards large area VLSI compatible substrates. The project was central to section NMP-2009-2.2-1 of Theme 4 of the NMP Call (Oxide materials for electronics applications).
The project was based on the following assessment: Transition metal oxides offer a wide variety of exploitable properties. Depending on composition, they can be insulating, semiconducting or metallic, ferromagnetic, ferroelectric or even multiferroic. Interfaces between different oxides often show electronic reconstruction. They can generate new functionalities and be utilised, for example, in high performance tunnelling magnetoresistance (TMR) junctions or spin-polarized interfacial electron gases, as recently demonstrated by members of IFOX. By combining ferroelectric and ferromagnetic oxides, artificial multiferroics with very strong coupling constants can be formed. The advantages of transition metal oxides and oxide heterostructures over other materials are: scalability – because of the stronger interactions and therefore, typically, more localised electronic states; versatility – because of the very rich interplay of different phenomena in oxides; stability – because of the robustness with respect to oxidation or diffusion; and non-volatility and low energy consumption - because of the intrinsic behaviour of ferromagnetic and ferroelectric ordering. Despite this rich potential for use in electronics, actual applications were relatively few in the beginning of IFOX as they rely on atomic scale control and understanding of these materials and their interfaces. The main ambition in IFOX was to improve the structural quality of interfaces by developing and selecting optimal growth and processing protocols and to realise tailored magnetic and electronic properties by achieving a deeper understanding of interface-related phenomena. This should lead to the required high degree of control for novel spin- and/or charge-based electronics in the framework of CMOS fabrication technology, but with next-generation performance and functionality. Applications were defined in the beginning and during the project by the needs of our major industrial beneficiary; technology transfer was facilitated by two SMEs specialized in industrial-scale oxide deposition systems, and the third SME, Protec Surface technology, who specializes in large area industry industrial sputtering systems.
To understand the workplan of IFOX it is necessary to understand that certain key ingredients necessary to access the entire potential of oxides for electronic integrated circuits were missing when IFOX started:
(1) intimate understanding of novel device-relevant functionalities that originate from the electronic reconstructions at interfaces between complex oxide layers, such as (spin-polarized) two-dimensional electron gases, multiferroic attributes, highly efficient spin filters, etc;
(2) full control of thin film growth and interface matching with substrate materials at sizes suitable for industrial applications; and
(3) comprehensive processing parameters, ranging from layer deposition to complete device fabrication, all compatible with mainstream CMOS technology, either in back end or front end processes, and any emergent successor technology.
IFOX was targeting the following four major areas of oxide heterostructures, all of which closely linked to technological applications to open up new opportunities for our industrial beneficiaries.
(1) Ferromagnetic oxide heterostructures controlled by charge or spin-polarized currents in advanced TMR structures for memory, programmable logic, sensors, and oxide/semiconductor spintronics.
(2) Ferroelectric oxide heterostructures controlled by charge and electric fields for sensors, nanoionic switches and fused memory-logic elements in which ferroelectric polarization and/or redox processes are used to create hysteretic resistance changes.
(3) Multifunctional multiferroic oxide heterostructures controlled either by charge or by spin and ultimately by ultrashort light pulses and capable of supplying both electric and magnetic fields on a very short timescale. These will be integrated into concepts for novel magneto-electronics and opto-magneto-electronics devices with unprecedented functionality.
(4) Growth of high quality oxide layers on large silicon substrates, essential for the implementation of all three material classes into electronic applications.
The major technological challenges to be mastered in order to realise advanced industrial applications were defined as:
• grow the oxide films of interest on Si and SiOx with a quality comparable to state-of-the-art semiconductor growth;
• increase the substrate size for typical oxide deposition methods to 200 mm and beyond;
• establish processing conditions for the relevant oxide heterostructures to allow nanoscale control and reproducibility of material parameters whilst maintaining CMOS compatibility;
• control the electronic reconstruction at the interfaces;
• understand and control the influence of interfacial strain.
The above are underpinned by fundamental challenges in understanding and predicting:
• electronic reconstructions at the interfaces;
• spin tunnelling and transport across oxide interfaces;
• fundamental properties of nanoscale oxide elements with respect to the ionic and electronic contributions to the screening of polarization charges at the electrode -oxide interfaces;
• the interaction of redox-centers and extended defects with ferroelectric polarization charges;
• interface coupling between ferromagnetic and ferroelectric oxides in artificial multiferroics.
Each challenge relates to interfacial properties, starting from the substrate/oxide interface that determines the quality of epitaxial growth, the internal oxide/oxide interfaces that confer functionality, the oxide/electrical contact interface at which charge redistribution and intermixing can influence device performance, to the oxide/air interface at the surface. The IFOX methodology was to achieve full control and understanding of each interface through extensive and imaginative investigations. The wide range of methods available in IFOX and their application to charge and spin controlled oxides should give rise to new insights and lead to novel device concepts. Finally, unprecedentedly fast optical control of certain oxides - particularly artificial multiferroics - will be explored, leading to new concepts for magneto-optic electronics.
Four target research areas were defined for which suitable heterostructures and concepts should be developed:
Magnetoresistive tunnel junctions
Magnetoresistive tunnel junctions are still the major candidates for magnetic random access memory (MRAM). At the beginning of IFOX the state of the art MRAM using MgO barriers and spin transfer torque for programming yielded prototypes almost ready for the market. However, the parameters in terms of the necessary currents for switching were too high. IFOX was to investigate and develop novel oxide materials which, when integrated in heterostructures, should reduce the switching current due to their intrinsic high spin polarization. This included also a major effort in the optimization of novel ferromagnetic oxides which are conducting and ferromagnetic well beyond room temperature.
Ferroelectric and valence-change effects in oxides
Ferroelectrics (FEs) and related oxides have been subject to intensive studies because of possible applications for FE memories. Redox-based resistance switching effects, however, exhibit much better scalability but at the potential expense of retention time. A resistance-based memory effect in which the FE polarization charge interacts with the space-charge region of a redox-controlled interface might lead to a novel memory device offering both high scalability and retention time. This concept has the potential to open a completely new road towards ultra-dense, energy efficient non-volatile memory devices as well as the fusion of memory and logic within new architectures.
Artificial multiferroics for new multifunctional electronics
Intrinsic multiferroic (MF) materials, which display both magnetic and electronic order, are of particular technological interest since they offer the prospect of multi-state logic and memory. However, the operation of today’s IMF materials was and still is limited to temperatures much below room temperature, largely due to the mutually-exclusive nature of the ferroic orderings. In contrast, IFOX has chosen artificial MF materials which hold much more promise. IFOX aimed to advance the state-of-the-art on several levels. It should develop the still-incomplete understanding of the interface coupling and improve its control. Moreover, IFOX should focus on the preparation of new magnetoelectric oxide film materials with advanced characteristics (e.g. enhanced Curie temperature) by means of interface and atomic sequence engineering within the SL growth approach. Multiferroic tunnel junctions were another promising approach, enabling control of spin-dependent carrier flow by electric field, and thereby the construction of four-state memories. Hence IFOX should finally target the integration of artificial MFs into completely new magneto-electronic multifunctional demonstrators, controlled by electric and/or magnetic fields and, ultimately, by light.
Oxide growth on silicon
Silicon is the most desirable platform for the integration of epitaxial multifunctional oxide devices with current semiconductor electronic devices. A capability to grow high quality layers on large area silicon substrates is therefore a major prerequisite for the integration of the novel oxide materials into mainstream electronics. Difficulties in coherent oxide growth are one of the biggest technological hurdles to oxide exploitation. IFOX was to extend the already existing research to relevant FEs, FMs and MF heterostructures. The ultimate goal was to achieve controlled epitaxial growth of the most promising oxides on 150 mm Si substrates and to define the necessary boundary conditions to access wafer sizes of 300 mm, which are currently used in production semiconductor facilities.
Furthermore the IFOX consortium was not only to use PLD for oxide growth but also pulsed plasma deposition or Aerosol deposition and sputtering. In the case that PLD would be unable to achieve the necessary buffer quality, both these processes can be used for alternative approaches. Especially Aerosol deposition with subsequent annealing may present a promising alternative because of its completely different technology and growth physics.
The above mentioned goals were defined at the beginning of the project. Because IFOX was planned for four years an assessment of the results and a revision of the workplan was foreseen after year three in which the goals should be defined in more detail, now focusing on well-defined heterostructures with novel functionalities belonging to the objectives and target research areas of IFOX.
This assessment was done and resulted in the following definition seven heterostructures:
1) Manganite/TiO based heterostructures which exhibit resistive switching suitable for novel storage and possibly logic applications
2) Heterostructures based on LaAlO3/SrTiO3 with additional ferromagnetic properties to combine transistor functionality with non-volatile storage for example for logic applications
3) Heterostructures based on BaTiO3 and NiFeO3 with inherent tunneling magnetoresistive functionality which may allow for MRAM cells which are switched by the magnetoelectric coupling between the BTO and the NFO.
4) Heterostructures of (La,Sr)MnO3 (LSMO) and ZnO which combine magnetoresistive sensing and gas sensing properties for automotive applications
5) Heterostructures using a bilayer of (La,Sr)MnO3 and BiFeO3 in which the magnetization of the LSMO can be switched via magnetoelectric coupling to the BFO which again can be switched by an electrical pulse. This structure would also address the area of magnetic memory with low switching power
6) Superlattices of LaMnO3 and SrMnO3 in which the magnetic properties compared to LSMO are much improved at room temperature and very high sensitivity in sensor applications (MR-sensors and Bolometers are to be expected).
7) Heterostructures using ferromagnetic (La,Sr)MnO3, a barrier of SrTiO3 and a ferromagnetic metal counter electrode which realise a combined element for resistive switching and magnetoresistance which can serve as a stateful logic device to combine logic and memory in a single cell.
Under the guidance of one lead partner per structure, each heterostructure was to be fabricated and investigated taking care of the following aspects:
- The structures should be optimized for maximum performance
- Growth on silicon substrates, if possible on large scale, was to be demonstrated and the results compared to those on matching oxide substrates
- Theory support was to be given wherever necessary to help design the structures and understand the results of characterization
- All results should be assessed with the help of the industrial partners in terms of maturity, technology readiness level of growth and processing, target application and competitiveness.
- For all heterostructures growth and industry compatible processing should be documented and made available as catalogues of
o Growth processes with detailed parameters
o Processing tools and recipes with detailed parameters
o Chemicals used and possible environmental impact.
An additional task of upscaling was added which was to be addressed by a third SME that joined the project at this stage with an expertise in industrial thin film deposition tools. This task should concentrate on the development of suitable buffer layers on large area silicon.
By this newly defined workplan the more general optimization and identification of promising heterostructures was guided towards concrete plans on heterostructures with defined functionality for which all technology should be ready for transfer to industrial application. The final results of IFOX can be summarized as (see for details the rest of the report):
1) All seven heterostructures were further developed including growth on silicon substrates and a thorough assessment was performed.
2) For two of the seven heterostructures (ZnO/LSMO-based gas sensor and PCMO/TiO based ReRAM) all important factors namely functionality at room temperature or above, fabrication on large area silicon wafers and industry compatible processing were demonstrated.
3) three catalogues were established, one on growth processes, one on processing and one on possible environmental impact of chemistry which allow for immediate assessment of needs and risks when transferring these heterostructures to industrial applications.
Project Results:
Description of the main S & T results/foreground from the IFOX Project
The S&T results of IFOX can be divided in several classes
- Growth of heterostructures by various technologies and development of growth technologies for oxides, including growth on large area silicon substrates
- Development of processing technologies for oxide heterostructures using industry compatible processes and equipment
- Characterization of the resulting structures and further development of characterization techniques for the requirements in IFOX
- Guidance and feedback by theory, including the development of theorical methods adequate for the goals of IFOX
- Industrial assessment of the results obtained
- Collection and cataloguing of growth and processing data for industrial use
- Oxide heterostructures with novel or improved functionalities
In the following an overview of these results is given, following the order of the workpackages in IFOX. Because these descriptions often refer to the seven heterostructures (HS) that the IFOX partners were targeting during the final period they are again listed here for further reference:
HS#1: Si/YSZ/SRO/PCMO/TiO: Resistive Random Access Memory (ReRAM) based on resistive switching.
HS#2: Si/STO/LaAlO3/SrTiOs: Field effect transistor based on ferromagnetic 2D electron gas in LaAlO3 (LAO) /SrTiO3 (STO) heterostructures.
HS#3: Si/STO/SRO/BTO/LSMO: Magneto-electric coupled memory device for room-temperature operation
HS#4: Si/YSZ/LSMO/ZnO: Multifunctional automotive sensor.
HS#5: Si/STO/BFO/LSMO: Multiferroic memory element.
HS#6: Si/STO/LMO/SMO: AMR field sensors and bolometers.
HS#7: Si/STO/LSMO/STO/FM1 (FM1=Co, Fe3O4): Resistive switching device (ReRAM) enhanced with magnetic functionality.
Main results of activities within IFOX
WP1: Theory and Modelling
The theory activity within IFOX aimed at the description of heterostructures both in terms of model parameters derived from the bulk as well as by calculating properties from setups modelling of realistic junctions and interfaces. The main tool used is the Density Functional Theory (DFT) that has transformed material physics and likewise computational chemistry, surface science and nanoscience. This framework, which provides the capability to describe the electronic structure, interatomic forces and in part also the electronic excitations of oxide materials containing hundreds of atoms in a computational volume, has developed into the standard method for realistic descriptions of materials from a quantum mechanical point of view.
At first, the calculations performed by the theory groups within IFOX concentrated on calculating the properties of bulk oxide materials. In particular a strong focus was on the determination of the complex bandstructure of several oxides (BaTiO3/PbTiO3/SrTiO3/KaNbO3/ZnO/CFO/NFO/PZT/...) which all could be used as possible materials in tunnelbarriers. These results provided the key input for the correct interpretation of tunnelling experiments and can additionally be used in the design process of heterostructures. In this first stage of the project the computational results were collected in the materials data-base developed within the consortium.
After these fundamental studies on bulk materials theory concentrated on all-oxide tunnel-junctions. In such setups we demonstrated the mechanism through which the TER (tunnelling electro-resistance) effect and the TMR (tunnelling magnetoresistance) effect can coexist in a tunnel junction. We presented the ab initio demonstrations of the mechanisms with our calculations of the SRO/STO/BTO/SRO junction. Using these results we proposed and demonstrated that the Fe/BTO/MgO/Fe junction could also function in this manner.
The study of interfaces between different oxides as well as the closely related field of domain walls in ferroelectrics was another field of active research. In particular, the influence of additional defects in such interfaces was studied. We found that the presence, concentration and nature of such defects play important roles in the stability of the interfaces and their electronic configuration.
The most intriguing effect of magneto-electric coupling at interfaces of oxides was investigated theoretically in different situations:
a) For the ferrites (CFO and NFO) we showed that it is the band alignment with the electrodes which determines the dominant spin character of the current, highlighting the importance not only of the tunnelling material but also the electronic structure of the electrodes and the interface. This understanding has led to the concept that by using a ferroelectric material the band alignment to CFO or NFO can be altered, this in turn changing the sign of the dominant tunnelling electrons resulting in a massive coupling between the magnetic spin of the tunnelling state and the electrical polarization of the ferroelectric.
b) Also the two dimensional electron gas (2DEG) formed at the interface between oxides of different polar character can be manipulated. In a BiAlO3/Sr2NiWO6 interface we showed that one can realize a magneto-electric coupling across the all-oxide interface, if we consider the effect of an external (exchange) magnetic field on the Rashba-type spin-orbit split 2DEG. Unfortunately growth efforts showed that this interface was not to be realized within the scope of IFOX.
In the final phase of the project, in which the work was concentrated on several well-defined heterostructures the theory activities also focused accordingly. For example, there were multiple simulations performed to clarify the nature and the properties of the 2D electron gas formed at the LAO/STO interface. A particular focus has been the magnetic properties of this interface and the question how the magnetism can be stabilized and manipulated. Our calculations indicate that magnetism can indeed be expected at this interface due to the localization of the electrons and the formation of Ti3+ ions, but that at the same time the ordering temperatures will be rather low. Our estimates give a critical temperature of about 35K for the ideal interface. This can be improved by doping with magnetic Mn ions and further enhanced by gating the 2D electron gas appropriately, but our results do not indicate that these effects can lead to effects observable at room temperatures.
Additionally, in this last part of the project, substantial work on the physics of oxygen defects in ZnO has been performed. Such vacancies are believed to be the key players in effects like gas sensing devices and resistive memories and the understanding of details of the formation and the diffusion of such vacancies thus is of significant practical importance. Our simulations were performed by first studying the different interfaces and surfaces and by investigating diffusion barrier heights with the nudged elastic band method. To a large extend these diffusion processes studied proved to be independent of external influences and hence the optimization of heterojunctions including ZnO are believed to by achievable by independent optimization of the constituents.
These results nicely showed that theory was able to supply new concepts to the experimentalists. Some of these were shown to be possibly realized, some were later shown to be inaccessible to current growth techniques. Those where realization seemed possible made their way into the finally chosen heterostructures. Also the simulations for example of the diffusion of vacancies are a good example how theory can and did help to interpret results of characterization.
WP2 Growth
The main objectives of this workpackage were associated to the development of state of the art depositions of functional complex oxides stacks for integration with existing CMOS technology (i.e. deposition on large area Silicon substrates). In details, the WP2 objectives were:
• To deliver high class transition oxide FM, FE and MF materials for all workpackages
• To establish a new state-of-the-art for the deposition of novel materials such as double perovskites and oxide super-lattices (SL)
• To assess the efficacy and industrial scalability of a range of oxide deposition processes for the first time
• To develop an all-oxide TMR stack
• To perfect the deposition of a range of oxides onto industry-standard Si wafers
• To progress in the areas of equipment and transfer of knowledge to the market
The above objectives have been achieved, with partial amendment for what concerns double perovskite growth, and the main results are described in the following.
2.1 Growth of high quality transition oxides
A large number of high quality thin films and heterostructures have been produced, and the feedback with other WPs (WP1 for theory, WP4 for characterization and WP3 for patterning) allowed the optimization process. The overall high quality growth obtained by IFOX enabled the functionalities claimed for heterostructures (HS) 1-7. In particular the growth of functional interfaces (between them STO/LAO for application in FET; integration of Ferrolectric oxides (BTO, BFO) with conducting perovskitic oxide (SRO and LSMO) for TER; LSMO/STO for MTJ, LSMO and ZnO for sensing) has been obtained through the control of structure and composition down to the interatomic scale as confirmed by high resolution TEM images provided within the experimental activity carried out in WP4.
2.2 New state-of-the-art for the deposition of novel materials
WP2 produced high quality functional superlattices with control down to the atomic scale. Some of the growth methods used in the project were designed as particularly useful for the purpose: a powerful example is the use of metal-organic aerosol deposition (MAD) for LMO/SMO superlattices with novel ferromagnetic high temperature phase (see for more details HS6).
2.3 Efficacy and industrial scalability of a range of oxide deposition processes
WP2 activities allowed to set up key-parameters for deposition processes depending on the specific technique (sputtering, PLD, PPD, CSA, MAD, ...). In collaboration with WP6, the scalability of the lab process to industrial one has been evaluated to define the main constraints (see the process database available on the web). A more industrial friendly technique (sputtering) has been successfully used for a full oxide HS growth (HS3) and high quality interfaces have been obtained.
2.4 All-oxide TMR stack
A full oxide TMR device has been obtained by TCD. The growth parameters of device constituents were first optimized on individually grown layers, followed by the incorporation of these layers into a full stack. The SRO/NFO/BTO/LSMO structure has been demonstrated to act as a TMR element.
2.5 Deposition of a range of oxides onto industry-standard Si wafers
A consistent effort of the IFOX project has been devoted to the development of appropriate buffer layers for the growth of complex functional oxide on Si substrates. A number of buffer layers (STO, YSZ/CeO2) have been obtained by varying the deposition conditions and the selection of the optimal process sequence has been set, allowing the deposition of epitaxial layer and heterostructures on Silicon substrates. Compatibility up to 4” wafers has been attained by PLD and PPD techniques, allowing to reach the project goal of industrial scalability of HSs.
2.6 Equipment and transfer of knowledge to the market
The active involvement of private companies (OS, CRF, TSST and PST) in the development of functional HSs has boosted the transfer of developed technologies to possible market applications. The optimization of large area deposition of oxides on Si opens then the possibility of implementing standard CMOS technology with new and functional materials. OS as well as TSST have developed new prototypes for large area deposition during the course of IFOX.
In relation to the functional Heterostructures (HSs 1-7) identified in the second part of the project, we summarize in the following table the main results achieved.
The integration onto Silicon substrates have been obtained for all the HSs, with the exception of HS3, indicating the successful approach of the activity. It should be noted that the extension of growth of HSs to large area substrates is mainly related to the processing steps: this is the main reason for having two HSs growth processes tested on 4’’ substrates.
HS
Growth on single crystal substrates Integration with Si Compatibility with Large Area Comments
HS1 X X X Leakage depends on crystallinity of buffer layer on Si
HS2 X X Different parameters for Si and single crystal substrates
HS3 X - Criticalities on the deposition systems
HS4 - X X Growth parameters allow independent functionalities
HS5 X X Working temperature below RT
HS6 X X Growth on Si attained but some criticalities on functionality
HS7 X X Role of STO crystallinity in functionality
WP3 Processing
In order to functionalise the heterostructures created for the IFOX project, various processing steps are typically required. Such steps often require a variety of technologies, from the methods used to clean substrates before deposition to device fabrication using UV and electron-beam lithography to pattern devices of micron and sub-micron dimensions, combined with various etching techniques to create the final devices. Multiple steps are usually required in order to create sensors which can be measured and integrated into a testing rig. Such processing often creates difficulties due to tight tolerances of the materials and equipment used when working on such small dimensions. Each process must also be fully compatible with the oxide materials used in the heterostructures.
During the course of the project, various factors were considered for processing of device structures, namely compatibility of technology, ease of integration into existing manufacturing, as well as the environmental impact of each step. The scale of the project meant that many partners had to work on various aspects of the processing in parallel. Partners involved in processing are Halle (MLU), Gottingen (OGOE), Twente (TSST), Glasgow (UGla), Dublin (TCD), Mainz (JGUM), CRFiat (Turin), PST (Beddizzole) and PSI (Villigen).
Work carried out during the project involved the deposition of thin films of oxide materials by a number of different physical and chemical vapour deposition techniques. This involves the use of various materials as substrates. Before deposition can occur, the proper steps must be taken to ensure that the surface is clean and properly terminated for the desired purpose. To this end, standard methods have been developed or modified to prepare Si, SrTiO3 and GGG substrates.
Cleaning of Si and GGG substrates was done by immersion in acetone and sonication to remove any organic surface contaminants, followed by rinsing in isopropanol. GGG was annealed in a furnace at 950°C for 30 minutes. Si was treated with RCA solution, followed by buffered HF to remove any surface oxide. SrTiO3 was heated in a furnace for 30 minutes at 950°C followed by HF treatment to produce a TiO2 terminated surface.
Following deposition of the heterostructure, the fabrication of devices is completed in a series of etching and deposition steps, depending on the device structure. Of the seven device heterostructures investigated during the project, three used a hall bar structure, two used small diameter metallic pillars as contact pads, one used large area metallic pads for contact, and one used a current-perpendicular-to-plane tunnel junction design.
In the first instance, hall bars were constructed by two different methods – by lift-off or by etching. HS2 used the lift-off method to form an open window to the substrate surrounded by an insulating layer of amorphous LAO. This window was defined in PMMA photoresist using e-beam lithography, with critical dimensions as low as 100nm. By depositing amorphous LAO, and removing the PMMA, the defined shape is left open for subsequent steps. Crystalline LAO was then deposited into the window to form a 2DEG, followed by deposition of a metal gating layer.
This method was employed due to reduced likelihood of issues arising from etching of the STO substrate, known to cause conduction in the surface, as well as improved performance of the 2DEG interface between the STO and LAO.
The hall bars in HS5 and HS6 were constructed by defining a hall pattern in photoresist on the top surface of the finished heterostructure stack, and etching using Ar+ ion milling or inductively-couple-plasma reactive ion etching (ICP-RIE). In this case, the etching rate of each material in the stack must be well calibrated, or an endpoint detection system must be in place, in order to ensure that etching beyond the desired layer does not occur. HS5 requires one lithographic step, which incorporates contact pads into the shape of the hall bar. HS6 requires a two step process, by etching a segment of the top surface followed by deposition and lift-off of Cr/Au contact pads. Wet etching of the manganite surface of HS6 was also tested with KI buffered HCl solution. It was found that, while effective, the reproducibility and quality was much more varied as compared to RIE, and so it is considered to be less desireable.
Both HS1 and HS7 use small diameter pillars for testing. The heterostructures include a metal as the top layer to ensure good contact. In both cases, pillars are patterned in photoresist and the surrounding material is etched down to the oxide layers. Thus, the processing of HS7 is completed and testing is carried out via probe station contacting of pillars. HS1, on the other hand, goes on to deposit a hard insulating AlOx layer, followed by deposition and lift-off to define larger contact pads.
It is also possible to pattern devices without using photoresist, as in the case of HS4. Shadow masks in close contact to the substrate have been used during deposition to produce the desired structure. This process is significantly cheaper and more straightforward than the use of photoresist, however it is also suitable only for larger structures that do not require precise placement of features.
HS3 proved difficult to process due to the complexity of its structure, as well as the incompatibility of the substrate and etching process. The steps required to form a tunnel junction device involve the patterning of a bottom electrode, followed by small mesa structures and finally top contact pads, while ensuring that there is no shorting between top and bottom electrodes. Normally Ar+ ion milling would be used to etch the material, however it has been seen that this type of etching method causes oxygen vacancies in SrTiO3, the substrate material, leading to conductive behaviour. This can cause issues with shunting of current and significantly increases noise during measurements. As such, ICP-RIE was employed as an alternative method. RIE uses chemical and physical means to achieve etching of materials, and so has a selectivity towards certain materials, allowing it to be used in situations where purely physical etching methods are not possible. BCl3 was used as the etchant in this case, and proved highly suited to etching of complex oxides. It provided demonstrably improved etching of SrTiO3 substrates without the issue of conductive pathways, and is thus considered a vital technique developed during the project for processing in future works.
In the course of the project we investigated the quality and magnetic properties of SrTiO3 substrates of different cuts from various suppliers. Only (100) substrates from just two of the suppliers were free of ferromagnetic Fe-Ni inclusions in the surface. These inclusions, present at the 1 ppm level, give spurious hysteretic ferromagnetic signals. A new type of non-hysteretic surface magnetism of up to 10 Bohr magnetons per square nanometre, associated with oxygen vacancies, was discovered in vacuum annealed crystals.
One of the major aims of the project was integration of complex oxide heterostructures onto Si substrates to establish their feasibility in future large scale manufacturing alongside other CMOS features and devices. In order to do this, a method was developed to create STO seed layers on buffered Si substrates. Using a combination of CeO2 and yttrium-stablised zirconia as buffer layers, highly textured complex oxides were able to grow. Experiments were then carried out to elucidate how these seed layers would react to the various etching techniques employed throughout the project. It was found that both Ar+ ion milling and ICP-RIE were suitable techniques, however wet chemical etching with buffered HF showed significant damage to the layer edges after etching.
A number of etching techniques have been identified during the project, which are easily integrated into modern silicon and wafer manufacturing setups. The materials used in the project need special consideration when considering the etching method, and ICP-RIE using BCl3 has been identified as a flexible and powerful tool for processing of complex oxides and sensitive substrates.
WP4 – characterization
The successful fabrication and application of novel transition metal oxide heterostructures depends on atomic scale control of the structure and composition of the produced materials. The goal of WP4 was to analyze the structure, chemical composition and electronic, magnetic, ferroelectric, multiferroic, opto-electronic and opto-magnetic properties of the as-grown and processed materials. During the course of the IFOX project close to 100 different material systems have been characterized with a variety of techniques, and the knowledge for an extensive materials database was created. The extensive knowledge about the characteristics of the heterostructures served as the basis for selecting the seven most promising functional oxide heterostructures to work on in year 4. Measurements directly related with the sample growth, the interfacial characterization and measurements related with the functionality of the heterostructure have been performed.
Due to the scientific complexity and the challenging measurement conditions of oxide heterostructures several instrumentation developments were needed. As a result, major achievements were obtained in the structural characterization of interfaces. For example, high-resolution transmission microscopy combined with electron energy-loss spectroscopy enabled the study of strain and oxygen vacancies effects at interfaces. The local structure with atomic resolution was studied by means of high-angle annular dark-field (HAADF) scanning transmission electron microscopy (STEM), energy-loss-spectroscopy (EELS) and energy-dispersive X-ray spectroscopy (EDX), performed using a FEI Titan microscope, equipped with aberration correctors for both image and probe forming lenses. This was used for a variety of scientific questions. For example transition metal L2;3 electron energy-loss spectra for a wide range of V-, Mn- and Fe- based oxides were recorded and carefully analyzed for their correlation with the formal oxidation states of the transition metal ions. Also the structural phase transition and spontaneous interface reconstruction in superlattices was studied on an atomic level. The combination of these atomic level characterization techniques with advanced sample growth enabled to create enhanced local magnetization by interface engineering in perovskite-type correlated oxide heterostructures, to discover the extreme mobility enhancement of two dimensional electron gases at oxide interfaces by charge-transfer-induced modulation doping, or control the lateral anisotropy in correlated manganite heterostructures by oxygen octahedral coupling. These techniques also enabled to study the oxygen vacancy transport mechanism and to make detailed comparison with predicted interface structures.
Impressive developments have been achieved in the nanoscale imaging of the magnetization states of oxide materials. Using scanning transmission electron microscopy (STEM) and Lorentz DPC microscopy it was possible to correlate the structural and functional properties of ferromagnetic oxide thin films. The novel capability of collecting magnetic, structural, and spectroscopic information with nanometre spatial resolution in a single experiment was demonstrated, complementing recent advances in analytical electron microscopy and spectroscopy with aberration corrected Lorentz DPC imaging. Variations in Curie temperatures were mapped and the role of defects and damage in modulating local magnetic properties on the nanoscale were assessed, including the pinning of magnetic domain walls.
Another important aspect was the development of the in-situ control of parameters such as electric fields for different types of imaging experiments. This enabled for example investigating the role of the electroforming process in the establishment of resistive switching behaviour for layered heterostructures acting as nonvolatile Resistance Random Access Memories (RRAMs). Another example is the study of interfacial magnetoelectric coupling in artificial multiferroics, which are meta-materials that show the existence of a magneto-electric coupling, obtained by the interfacial coupling of a ferroelectric and a ferromagnetic constituent. This enabled to investigate the ultimate mechanisms responsible for artificial magnetoelectric coupling in different interfaces and possible routes to optimize the coupling and extend it to the nanoscale. A 90 degree electric field-induced uniform magnetization rotation in single domain 100*200 nanometer ferromagnetic islands grown on a ferroelectric single crystal was demonstrated using x-ray photoemission electron microscopy.
The experimental findings are well correlated with micromagnetic simulations, showing that the reorientation occurs by the strain-induced magnetoelectric interaction between the ferromagnetic nanostructures and the ferroelectric crystal. Specifically, the ferroelectric domain structure plays a key role in determining the response of the structure to the applied electric field, resulting in three strain-induced regimes of magnetization behavior for the single domain islands. Furthermore, it was demonstrated that such correlations can be extended to oxide heterostructures deposited on a silicon substrate, making it particularly interesting for applications. This paves the way towards magnetoelectric-MRAM devices containing an artificial multiferroic film stack with low power consumption and high switching reliability. Understanding the dynamics of the interfacial coupling in heterostructures is important to optimize the samples for best performance and the imaging of magnetization dynamics in oxide systems with sub-ns time resolution has been developed in the IFOX network using X-ray microscopy and magneto-optical imaging techniques. For example optical control of a magnetic tunnel junction has been tested with a specially designed set-up that allows optical access and pump-probe experiments on a magnetic tunnel junction, while applying bias voltages.
Another example is the imaging of the propagation dynamics of spinwaves (SWs) excited by a femtosecond pump laser studied with a new developed Time-Resolved Magneto-Optical (TRMO) imaging system.
A large fraction of the work in WP4 in the last year was focused towards characterizing and testing the functional properties of the seven selected heterostructures for a range of applications going from sensing of gases, or as bolometers to novel ways of data storage. For this, the IFOX network could make full use of the variety of standard characterizations techniques available and the new advanced techniques developed during the project, which were pushing the limits. The IFOX partners have made significant, internationally recognized contributions to the advancement of state of the art characterization methods of oxide materials, which enabled to generate systematic data sets. The large size of the WP allowed characterizing the samples with many techniques providing deep insights into the materials and in particular into their interfaces. The information about the sample quality was reported back to WP1 (theory) which enabled modelling inspired growth and improved understanding of interfacial effects and WP2 (growth) and WP3 (processing) which helped to improve the sample quality. The functionality of the materials originating from the interfacial coupling has been investigated and was provided as input for WP5 (Industrial supervision) and WP6 (Exploration of new functionalities).
WP 5 Industrial control and supervision concerned the industrial aspects involved in the production of finalized heterostuctures selected within the project and the corresponding achieved functionalities for the appropriate applications. Also collection and management of related information were carried out. Main activities can be summarized in the following categories:
• Procedure and standardization, related to: deposition or growth techniques, fabrication processes, interface functionality and potential hazard involved.
• Prototyping, related to machinery and large area issues.
• Techical Data Collection and Data Bases Management, related to: chemicals, materials, growth and fabrication processes.
Procedure and standardization
Concerning Growth processes procedures, the involved partners have described in details the adopted techniques in order to make comparison between them related to working conditions parameters, process timing and potential hazard involved. The specific goals of depositions on wide area substrates, Silicon platform and Growth technique efficiency are clearly indicated. Sputtering, which is already an industrial technique, has been used to prove heterostructures viability and reached required functionality. Pulsed technologies such as PLD and PPD have shown a high grade of versatility and more over offer the advantage of a common deposition platform. Those last two techniques seem to offer special features that enable industrial production: the most important one is the capability in transferring complex oxides stoichiometry and crystalline structure, from bulk material to thin film form.
All the adapted techniques have assured a good interface quality with a high control of space localization of chemical compounds which have guaranteed final functionality. Another important achievements obtained by IFOX partners was that depositions were performed on Silicon substrates enabling in this way integration in Semiconductor Industry.
Regarding adopted fabrication processes most of them are already developed at industrial level (Etching, Photolithograpy, E-beam Lithography-EBL). For this reason their compatibility with previous thin film formation processes was verified in order to guarantee the final functionality.
For H#1 fabrication process has been demonstrated for samples of any area dimension, demonstrating a full technological compatibility with previous processes.
For H#2 different patterning techniques were developed and tests were performed on obtained heterostructures, demonstrating that the interface quality is not affected at all. Also a new dry etching technique was applied which allows the patterning of the material on lateral sizes while maintaining the metallic behavior. This patterning technique is compatible to standard industrial etching. Also for the device containing
H#3 a lithographic process was applied, patterning it for vertical transport measurement. It has been demonstrated that the metal contacts can be achieved without damaging the interface properties.
Substrates of H#4 were specifically prepared to host gas sensing material and were patterned by chemical etching and UV photolithography.
Hs#5 were patterned by electron beam lithography followed by ion milling down to 250nm scale. This final device processing does not have particular restrictions and, as it was verified, does not alter the layers properties and interface’ ones as well.
The device containing H#6 was fabbricated trough E-beam nanopatterning followed by dry etching for different oxide films including magnetite. Also a direct patterning processes by using lift off techniques with sacrificial layers were applied. No particular boundary conditions arise from such processing techniques.
For H#7 patterning and nano-patterning processing were successfully applied.
Concerning interface functionality of H#1, the required characteristic of “ON-OFF” switching was very near to the goal over the entire device area of this specific application.
The H#2 required functionality was obtained through the control of the two dimensions electron gas that can be used to form a field effect transistor.
Concerning H#3, the coexistence of electronic polarisation and magnetisation in adjacent thin films has been shown during the IFOX project. The coexistence of TER and TMR has not been proven due to the absence of TMR in the measured devices.
H#4 has shown both: magneto-resistance effect when subjected to a variable magnetic field and change in the conductivity when exposed to proper gas mixtures. This combination generates a multifunctional device for which the magneto-resistive behavior is a function of the gas adsorption.
For the H#5 device concept two different spin configurations were coupled due to exchange bias at the interface. Therefore, by applying an electric field it is possible to alter the magnetization in one layer and vice versa, obtaining the required final functionality for low energy consumption memory devices.
Concerning H#6, interface functionality was based on the enhanced magnetism and anisotropic magneto-resistance in a Digitalsuperlattice (DSL) composed of different oxides layers. For this heterostructure, application as a bolometer and AMR at room temperature must require additional optimization steps.
Related to H#7, the final device functionality, switching between two or more distinctive resistive states, has been demonstrated for growth on a single crystal. TMR behavior is still too low and at low temperature.
Concerning potential hazard involved in materials handling and processes set-up, all partners contributed in collecting info on chemicals used in their premises and in the cleaning procedure standards for industrial environment. The research done had led to the identifications of a number of dangerous substances used within IFOX, which probably, in the future, will be banned. In particular two of them have to be replaced with lower risks ones and, because they involve rare earth elements, maybe could give rise to problems concerning their availability. Related to deposition processes, most of them are performed in high vacuum conditions preventing the dispersions of nanoparticles.
Prototyping
WP5 partners have faced the up-scaling process for large area deposition, developing industrial prototype and production procedures. Another important achievement obtained by IFOX partners was that depositions were performed on Silicon substrates enabling in this way integration in the Semiconductor Industry. Related to deposition techniques up-scaling process, magnetron sputtered (MS) buffer layers deposited on 4“ silicon with suitable homogeneity and crystal quality under standard industrial conditions has been reviewed and its procedure has been defined.
Once moved away from early stage DC to RF driven sputter deposition, process initial parameters and hardware were refined to improve homogeneity and, at the same time, feasibility tests, quality control and production of HSs on 4” full scale wafer were carried out. With such technique, in order to further improve crystal quality to a level similar to those achievable by other more sophisticated techniques (such as PLD and PPD), a large optimization work has been attempted on the pre-processing cleaning conditions and on the early stage temperature of the deposition processes.
Since magnetron sputtering is not the most suitable technique to achieve this kind of growth, the lower level of crystallinity respect to more sophisticated techniques is not a surprisingly negative result.
PLD has successfully applied to obtain very complex oxides heterostructures. Within IFOX projects PLD main limits have been partially overcome: although its maximum repetition rate of 20Hz is still too low, depositions were performed on 4” wide substrates overcoming the small area limitation. This was one of the most important goals for the IFOX project, a common platform technique which enables depositions on wide areas.
Also with PPD systems depositions can be performed on large area substrates, furthermore an industrial prototype system has been set-up which is already working in a continuous way on (infinite) strips of 360 mm width. Related to the desired thin film properties, it is possible to adjust different PPD parameters and select, in a short time and with few attempts (about 5 samples), the optimal values to perform growth processes. In order to improve deposition quality, in particular the distance between target material and the substrate was modified, reaching, a result of 6% of homogeneity at 3 σ. An “in-situ” RF Cleaner system was realized to improve oxides layers adhesion. For adopted fabrication processes most of them are already developed at industrial level (Etching, Photolithograpy, EBL) and there’s no particular need in developing prototyping models.
Techical Data Collection and Data Bases Management
WP5 work was also related to collect information from all partners on all chemicals, materials, growth processes, fabrication procedures and interfaces validation for its technological success. Those fundamental data were elaborated and organized in on-line data bases used during the project development. Technological databases are a powerful instrument to sustain the driving of the innovative aspects of reached technological results close to industrial organizations which can cover the final steps to the market.
Main databases content are related, but not limited, to the seven (or more) “concept devices”, based on seven different oxide heterostructures, which applications are clearly indicated and, in most cases, their functionality completely described and obtained. For such cases where the results do not match the expected behaviour, they however represent a resource of information for not to “waste time and money”: this is very useful from an industrial point of view. Data were also used for the transfer on Silicon platform of the oxide heterostructures, which was successful in most cases, with the help of a buffer layer, allowing the integration with the semiconductor industry. WP5 Partners also elaborated an evaluation of TRL (Technology Readiness Levels) in order to give an idea of the global results reached in a common used language.
WP6 Exploration of new functionalities
WP6 was dedicated to the exploitation of the results of IFOX, and this exploitation happened in three different ways:
1) Assessment of results obtained in WP1 through WP4 with focus on the identification of promising materials as well as processing technologies.
2) Identification of new functionalities and possible emerging applications
3) Deliver the necessary boundary conditions for industrial application of oxides.
Furthermore, within WP6 the goals and deliverables of IFOX have been checked throughout the project to guide necessary revisions of the work and tasks. WP6 created a platform for the use of new functionalities for new device concepts, with the output of other work packages as input.
Most important deliverables (next to intermediate deliverables) are:
D6.1.1 Report on promising materials for use inside IFOX
D6.2.2 Final catalogue of processing technologies necessary for the industrial application of oxides
D6.4.1 Definition of new concepts for novel functionalities as deliverables for month 48
D6.5.2 Report on assessment of SRO/PCMO/TiO based ReRAM devices
D6.6.2 Report on assessment of LAO/STO based field effect transistors
D6.7.2 Report on assessment Si/STO/SRO/NFO/BTO/LSMO based device structure
D6.8.2 Report on assessment of gas sensing reactivity of devices based on Si/YSZ/LSMO/ZnO and Si/YSZ/FO/ZnO
D6.9.2 Report on assessment Si/STO/BFO/LSMO based memory device
D6.10.2 Report on assessment of Si/STO/LMO/SMO based demonstrators for AMR and bolometers
D6.11.2 Report on the assessment of Si/STO/LSMO/STO/FM1 (FM1=Co, Fe3O4) based ReRAM with magnetic functionality
Within the first years of IFOX the project has focused on the deliverables on most promising materials and the definition of new concepts for novel functionalities as deliverables. An important result was the definition of the most promising heterostructures, which were the basis for the device concepts for novel functionalities. The report on promising materials in IFOX lists all oxides that are grown by the partners and the reasons why they are deemed promising and for what reason. It also gives a short overview of the properties of the different materials. The materials database contains many different compounds out of which those have been selected which were investigated in the heterostructures.
In the final year within IFOX 7 heterostructures have been explored, and these are:
HS#1: Si/YSZ/SRO/PCMO/TiO: Resistive Random Access Memory (ReRAM) based on resistive switching.
HS#2: Si/STO/LaAlO3/SrTiOs: Field effect transistor based on ferromagnetic 2D electron gas in LaAlO3 (LAO) /SrTiO3 (STO) heterostructures.
HS#3: Si/STO/SRO/BTO/LSMO: Magneto-electric coupled memory device for room-temperature operation
HS#4: Si/YSZ/LSMO/ZnO: Multifunctional automotive sensor.
HS#5: Si/STO/BFO/LSMO: Multiferroic memory element.
HS#6: Si/STO/LMO/SMO: AMR field sensors and bolometers.
HS#7: Si/STO/LSMO/STO/FM1 (FM1=Co, Fe3O4): Resistive switching device (ReRAM) enhanced with magnetic functionality.
Achieved results
HS#1: Functional epitaxial heterostructures (down to 200 nm) have been fabricated for the first time using heterostructures grown on 4” Si. The IFOX partners have actively worked together and this resulted in functional resistive switching devices within epitaxial heterostructures grown on 4” Si wafers. The large area epitaxial integration on Si has not been shown before. It was found that the interface between the TiOx and the PCMO is important in the switching characteristics. Most important, epitaxial integration of the devices in 4” Si wafers has shown that high crystalline quality (in terms of epitaxy and crystallographic purity) is very required. PLD has shown to be capable of achieving such quality. Furthermore, sub-micron devices down to 200 nm have been processed from 4” heterostructures, showing resistive switching.
HS#2: IFOX showed that integration of LAO/STO with Si is possible, and STO films of high quality can be epitaxially grown on Si wafer by molecular beam epitaxy. LAO films were then grown on STO/Si by Pulsed laser deposition (PLD). A two dimensional electron gas was revealed at both amorphous LAO and crystalline LAO and STO interfaces with suitable mobility at room temperature. Furthermore, nano-patterning down to 100 nm of LAO/STO interface by E-beam lithography has been achieved, specifically to test short scale lateral electrical insulation.
HS#3: Tunnel junction devices incorporating a synthetic multiferroic tunnel barrier have been fabricated and tested. This is the first known attempt to create a voltage controlled TMR/TER device. IFOX has identified an all-transition-metal-oxide structure that shows promise for future storage device applications. Growth and fabrication of highly crystalline and epitaxial device stacks has been demonstrated.
HS#4: LSMO/ZnO heterostructures exhibit both magneto-resistance, when subjected to a variable magnetic field, and a sensible change of conductivity when exposed to proper gas mixtures. The two effects were expected to be interwoven, but from modelling and physical experiments it can be stated that the interplay is limited and not affecting significantly the HS4 functional behaviour, which is dominated by the intrinsic properties of the materials employed, in such a way that the interfacial properties are reflecting only slight changes at the atomic level due to gas absorption. The present achievement is totally new and represents a step forward of the state of the art.
HS#5: Devices prepared on both STO and Si substrates have been fabricated and characterized. IFOX showed that the coupling between the ferroelectric BFO layer and the ferromagnetic LSMO layer occurs preferentially for rough interfaces. This indicates that device fabrication can be implemented even for the crystallinity that can be achieved on Si, which is necessary for realistic device applications.
HS#6: Applications as AMR sensor or bolometer have been completed with success. The interfacial nature of the high temperature (TC ≈ 350 K) ferromagnetic phase (HTP) confines the ferromagnetic (FM) state to two dimensions. Since the two-dimensional FM phase is expected to strongly enhance spin-orbit coupling via the reduced symmetry of the MnO6 octahedra, the anisotropic magnetoresistance (AMR) within the interface is highly increased. The AMR- effect directly links the sample’s resistance to changes in the magnetization direction with respect to the probe current. External magnetic fields affect the magnetization direction and the FM domain landscape and hence can be sensed by a simple resistance 4-point-measurement method. Within IFOX, two different approaches to make functional use of the material inherent AMR were tested.
HS#7: The heterostructures showed for the first time the intercorrelation between magnetoresistance and bistability. The magnetic functionality added to a resistive switching system enables the latter with an additional degree of freedom and opens a series of various advantages considering device applications. Firstly, standard resistive switching elements, working in either memristor or different regime, represent a two terminal device with all its advantages and disadvantages. The magnetic functionality, that is, the possibility to modulate with magnetic field the resistive states established by non-volatile switching mechanisms, can be considered as a third terminal (external magnetic field) promoting the utilization of such devices in more complex electronic circuits and systems.
Conclusion:
The results described above show that IFOX has achieved a lot of results not only on the academic side but also on the side of industrial application or in paving the way there to. For two of the seven heterostructures (ZnO/LSMO-based gas sensor and PCMO/TiO based ReRAM) all important factors namely functionality at room temperature or above, fabrication on large area silicon wafers and industry compatible processing were demonstrated
As a further important result, all data on growth and processing were collected in several web-based interactive databases and evaluated by the industrial partners. From these data three catalogues were established:
- A materials database listing possible candidate oxide materials including relevant physical properties
- A chemicals database where all chemicals are listed which are used in one of the growth or processing steps investigated
- A process database where for each process an infinite number of process steps can be entered, and each process step includes the equipment used, the environmental conditions, materials involved, process times and other details if necessary. The tree format allows for an easy overview, however, with fast access to the details and allows for immediate assessment of needs and risks when transferring these heterostrctures to industrial application.
In addition, a questionnaire to evaluate necessary costs and equipment for oxide electronics was created.
Furthermore, the relevant results have been published in peer reviewed international journals (91 manuscripts) and disseminated to the major inter- national conferences (almost the same number of talks). During the 2015 E-MRS Fall Meeting, held in Warsaw University of Technology, from September 15 to 18 2015, the IFOX consortium presented a session within symposium L "Materials for electronics and optoelectronic applications away from silicon”. IFOX also setup a booth at this E-MRS meeting with posters on all heterostructures in which discussions with possibly interested industry partners were held. In addition, a massive search in patent databases was done in order to evaluate possible restrictions and possible areas for protection of IP and to identify competing technologies.
January 2016
Potential Impact:
Potential impacts
Socio-economic impact
At present oxides are important for semiconductor technology mainly in the context of improving the dielectric/insulating and interface properties of gate oxides for use in MOSFET devices. However, their real promise for novel future applications is that they can be engineered with additional functionality, e.g. ferroelectric, multiferroic, and magnetic properties, by tailoring their chemical composition and making multilayer stacks of a range of different oxides.
The aim of IFOX was to explore, create and control novel electronic and magnetic functionalities that result from the rich interplay of charge, spin and orbital degrees of freedom in transition metal oxide heterostructures and their interfaces. The outcome aimed at the establishment of a material platform for novel ‘More than Moore’ and ‘beyond CMOS’ electronics, compatible with VLSI and which delivers performance and functionality far beyond the state-of-the-art.
To achieve these goals we carried out a research programme with workpackages on theory and modelling, epitaxial growth, processing, characterization, device fabrication and knowledge transfer. Crucially, technological relevance and impact was guaranteed by the active participation of a world-leading company, Fiat and three SMEs, namely TSST and Organic Spintronics, with expertise and infrastructure for epitaxial oxide growth on Si and Protec Surface Technology, with expertise on engineering and production of PVD and PECVD machinery for thin films and coatings deposition and customised deposition process developments. Focusing the research on industry needs ensured that IFOX worked on relevant problems with tangible results. Success for the project means delivering useable new technology and improved understanding to industry beneficiaries which will allow them to make significant progress on the roadmap for future generations of semiconductor devices.
While the first three periods of IFOX were dedicated to the growth, processing, and investigation of various oxide thin films and their heterostructures, the final period focused on seven different heterostructures, chosen for their largest potential for possible future applications in the target research areas:
1) Manganite/TiO based heterostructures which exhibit resistive switching suitable for novel storage and possibly logic applications
2) Heterostructures based on LaAlO3/SrTiO3 with additional ferromagnetic properties to combine transistor functionality with non-volatile storage for example for logic applications
3) Heterostructures based on BaTiO3 and NiFeO3 with inherent tunneling magnetoresistive functionality which may allow for MRAM cells which are switched by the magnetoelectric coupling between the BTO and the NFO.
4) Heterostructures of (La,Sr)MnO3 (LSMO) and ZnO which combine magnetoresistive sensing and gas sensing properties for automotive applications
5) Heterostructures using a bilayer of (La,Sr)MnO3 and BiFeO3 in which the magnetization of the LSMO can be switched via magnetoelectric coupling to the BFO which again can be switched by an electrical pulse. This structure would also address the area of magnetic memory with low switching power
6) Superlattices of LaMnO3 and SrMnO3 in which the magnetic properties compared to LSMO are much improved at room temperature and very high sensitivity in sensor applications (MR-sensors and Bolometers are to be expected.
7) Heterostructures using ferromagnetic (La,Sr)MnO3, a barrier of SrTiO3 and a ferromagnetic metal counter electrode which realise a combined element for resistive switching and magnetoresistance which can serve as a stateful logic device to combine logic and memory in a single cell.
In addition the project foresaw an effort to attempt and assess the realization of all these heterostructures on large area silicon wafers dramatically reducing possible production costs compared to heterostructures deposited on single crystalline oxide substrates.
All these seven heterostructures were further developed including growth on silicon substrates and a thorough assessment was performed. For two of the seven heterostructures (ZnO/LSMO-based gas sensor and PCMO/TiO based ReRAM) all important factors namely functionality at room temperature or above, fabrication on large area silicon wafers and industry compatible processing were demonstrated. As follow up, CRF will continue the study to increase the TRL of the HS4 up to 6. Besides, to go deep inside into the measurement repeatibility of the gas sensitivity, and the life cycle costing of the production will be part of the future investigation to evaluate the value for money in the replacement of the standard gas sensor today employed. The evaluation will take into account the improved absolute sensitivity of HS4 demonstrated in IFOX. The exploitation of the coupled magneto resistance funtion of HS4 will be requiring further developments and analysis to gauge at what extent the airgap variation sensitivity of the MR response will affect the precision of the absolute position of a mangnetized valve. It is expected to go deeper inside these issue by a second ICT project employing novel advance designed smart systems.
To facilitate the transfer of results to industry all partners contributed to a huge effort, which was guided by the industry partners, to collect the knowledge on growth and processing in a suitable format. This resulted in three major deliverables
1) A catalogue of growth techniques and growth parameters for all heterostructures
2) A catalogue of processing techniques and parameters for all heterostructures
3) A catalogue of all chemicals involved in growth and processing, assessing the environmental impact of the chemical compounds
In addition a tentative assessment of growth and processing in terms of technology readiness level (TRL) is provided, resulting in typical values ranging from 3 to 5 for all steps (see WP5 for details).
Exploration of new functionalities
For all heterostructures an overall assessment is provided which not only addresses the results of characterization. It also provides a placement of the heterostructure within the targeted research area and an evaluation of the maturity of the necessary technology and a competitive comparison to possible alternative devices on the market or in the development chain.
The data of the three catalogues allow for immediate assessment of needs and risks when transferring these heterostrctures to industrial application.
After completion of the programme, we can state that IFOX has delivered results that are highly relevant for our industrial partners in terms of novel sensors, growth and processing and integration to Si.
Societal implications
The importance of the ICT sector from an economic, societal, and (increasingly) environmental viewpoint is obvious: Europe needs a strategic and innovative semiconductor foundation underlying the ICT portfolio to bolster industry, increase innovation and take a lead in economic, societal and environmental terms. The semiconductor market is currently worth 200 billion Euro (256 billion USD) globally on an annual basis. There is a multiplier effect of approximately 25 times that amount in terms of equipment and services ranging from television sets to IT. In other words, semiconductor technology provides the knowledge and technologies that generate some 10% of global GDP. The European semiconductor market attained a value of approximately Euro 30 billion in 2007, representing just 16% of the worldwide total .
IFOX has improved the competitiveness of the EU semiconductor industry by its development of novel device technologies based on oxide materials with novel device functionalities beyond the current state-of-the-art. This will support the need for industry linked technological innovation in the EU as described in the 2008 Europe Innova Innovation watch report: "A prospective innovation challenge for the ICT sector is the insufficient collaboration between public sector and business sector researchers".
The results obtained in IFOX is multi-disciplinary and beyond what could have been accomplished by one or two individual research groups as they do not have the expertise nor infrastructure for all the various aspects of theory, growth, structuring and characterization necessary for the achieved functional structures. In addition, without the critical evaluation and feedback of the industrial partners, the focusing on the most relevant and prospective structures and their compatibility with Si technology would not have been accomplished.
Regarding impact of research, the EU still ranks as the world’s largest producer of scientific knowledge (measured by publications), but contributes less than the US to high impact publications . These data provide the backdrop against which the consortium has tackled its research goals. The best research conducted within IFOX has been published in high-profile peer reviewed journals and the researchers have presented their findings at leading international conferences.
IFOX has provided the groundwork and knowledge necessary for magnetic oxide inclusion in future generations of ICT products, critically, by engaging a key end-user company Fiat. The new knowledge generated by IFOX is targeted in such a critical area that it will be truly groundbreaking in worldwide terms and put Europe ahead of other regions, safeguarding existing ICT jobs in Europe and working to create new markets for job increases. This should help the stagnation of the EU-27's R&D investment (R&D expenditure as % of GDP) which is currently at 1.84% vs. the Lisbon target of 3%.
Dissemination and/or exploitation of project results, and management of intellectual property
The IFOX programme is centred on increasing the knowledge and understanding about transition metal oxides and their properties and also on applying this knowledge to create new oxide functionalities for potential device application. A key component of this effort is to generate and protect intellectual property (IP) with a view towards exploitation and delivering novel device functionalities in industry.
Dissemination
IFOX contributed during its whole duration to create and disseminate knowledge on complex oxides and interfaces to the overall community. In particular, the consortium published on peer reviewed international journals 91 manuscripts, and dissemination to the major inter- national conferences almost the same number of talks.
IFOX also disseminated results in the frame of WP9 by organizing a Summer school which took place in Bildungszentrum Hesselberg from July 29° to August the 3° 2014 that count- ed 35 students recruited within IFOX and 10 external students recruited by other projects attending. Nine speakers were experts of IFOX and one was invited from the ORAMA project.
Towards the end of the project the dissemination became more and more important. During the 2015 E-MRS Fall Meeting, held in Warsaw University of Technology, from September 15 to 18 2015, the IFOX consortium was presented within symposium L "Materials for electronics and optoelectronic applications away from silicon”. The IFOX session had the following talks:
11:00 Gertjan Koster, University of Twente, Enschede, the Netherlands, "Integrating epi- taxial oxides with Si"
11:15 Rik Groenen, TSST, Enschede, the Netherlands, "large area epitaxial oxide growth using PLD"
11:30 Gwenael Atcheson, TCD, Dublin, Ireland, "All-oxide multiferroic tunnel junctions"
11:45 Patrizio Graziosi CNR-ISMN, "Memristive and magnetoresistive properties of
SrTiO3 based junctions" 12:00 Nicolas Gauquelin, EMAT, University of Antwerp, "Application of advanced mi- croscopy techniques to the study of oxide-based electronic devices"
12:15 Mehran Vafaee, Johannes-Gutenberg-Universität Mainz, Mainz, Germany "The cor- relation between the interface roughness and exchange bias coupling strength in BiFeO3/La0.7Sr0.3MnO3 heterostructures"
IFOX also setup a booth at this E-MRS meeting with posters on all heterostructures in which discussions with possibly interested industry partners were held. In addition, a massive search in patent databases was done in order to evaluate possible restrictions and possible areas for protection of IP and to identify competing technologies.
Exploitation of results
With respect to exploitation, the industrial partners were mostly responsible for the analysis according to the respective man power allocated. Magnetron sputtering is acknowledged as being one of the most promising techniques for industrial use perspective, and it was used here by PST to exploit the production of complex oxide buffer layers for many industrial sectors (both for low- and high-end prod- ucts) where thin films are a mass-production major issue. On the other hand, TSST and OS are actively involved in the production of deposition equipment for these applications. Finally, CRF is one of the major end users for the novel sensors developed within IFOX.
PST, besides analysing more than 1000 valid patents to preview possible technological bottlenecks possibly preventing use of the know-how developed (buffer layers), also found through the analysis the possible market incubators to place their machinery and processes. The analysis carried out to foresee the use potential of the complex oxide buffer materials developed in the project took into account the material quality level in terms of (i) crystallinity (H/E: highly crystalline/Epitaxial, HT: highly tex- tured, LT: low texturing, P: polycrystalline), (ii) adhesion, and (iii) homogeneity over area as large as 4” to address that single applications according to the material required need. This analysis is essential to tailor the material production according to the chosen use and properly penetrate the markets addressed.
From this analysis the exploitation strategy of PST is to offer to the market buffer materials of tailored crystallinity to allow significantly decrease their cost for unit and let novel product penetrate the market.
TSST has built a prototype for large area deposition PLD and developed the necessary processing technology in the course of IFOX. Upscaling is demonstrated of the synthesis of complex oxide films with perfect control over the structure and composition of the material on large wafer size Si up to 4”. This has led to new standards for the precision synthesis of oxide films, and this removed in a radical manner the limitations on the availability of such films on industrial scale Si wafers. Important technological steps, including technologies to uniformly deposit the material on a uniformly heated substrate, have been made. These steps have enabled the epitaxial growth of crystalline complex oxide thin films on Si. The PLD system is at present ready to enter the market, enabling TSST to increase the market share. TSST hopes to expand their sales by 30% in the next coming years.
OS has in the course of IFOX built a prototype for a multisource pulsed plasma deposition system, (PPD - 5 parallel sources), able to deposit oxides with 6% of homogeneity on 3‘‘ wafers and on an ‘‘infinite‘‘ strips of 300mm width. This industrial prototype is already working in an Industrial Research Centre and almost ready to enter the market in 2017. At the same time, OS has started to develop an enlargement of the area covered by a single source. The result is that the last prototype system uses only two sources to deposit on substrate areas of 300mm width instead of the previous 5 sources.
OS hopes to expand their sales by 2018 by dealing this type of system specifically adapted for customer applications.
Another important business strategy that OS will follow is the possibility to insert a PPD source in an already used industrials system since PPD deposition rate has become compatible with similar deposition techniques.
For CRF, exhaust gas sensors is the main application of the IFOX complex oxides developed, namely: YSZ, CeO2, ZnO, and LSMO; oxides which are all sensitive to the gas exhausted by internal combustion engine powered cars, which do guarantee improved sensitivity and selectivity in the shape of sandwiches as e.g. HS4 addressed by IFOX. A plurality of sensor types and electrochemical detection systems are in use for the polluting gaseous compounds exhausted by the current production internal combustion engines. In particular, the focus is placed on the detection systems dedicated to the detection of carbon mono-oxide and di-oxide (CO, CO2), and of nitrogen oxides (NOx), used to limit noxious emissions, to control the EGR flux, and to detect the combustion misfire. Nitrogen oxide gases, commonly referred to collectively as NOX, are common pollutants formed in internal combustion engines and industrial combustion systems by thermal fixation and oxidation of atmospheric nitrogen1. NOX, which includes NO2 and NO gases, have adverse effects on the environment, and are the leading cause of green-house effect, acid rain, and photochemical smog. In humans, exposure to more than 3 ppm of NO2 gas for periods longer than 8 hours can cause respiratory and cardiovascular diseases. Hence, better detection of such gases is of utmost importance, and that calls for better understanding of detection mechanism to facilitate development and optimization of sensing devices. Semiconducting metal oxides based thin films have traditionally been used in detection of NOX gases, with the common measuring arrangement relying on changes in electrical conductivity upon interaction with the gas. In particular, ZnO based gas sensors have been employed extensively in the detection of both reducing and oxidizing gases due to the superior properties of this material such as enhanced sensitivity, stability, and low cost. A widely adopted route to reduce NOx emissions is Exhaust Gas Recirculation (EGR). The EGR system has a valve which cannot respond instantly to changes in demand, and the exhaust gas takes time to flow around the EGR circuit. This makes the calibration of transient EGR behaviour particularly complex. To have a better control of EGR it would be very important to be capable of sensing the NOx content, or relatively knowing the actual O2 concentration in the exhaust air to be recirculated, and at the same time controlling precisely the quantity of gas recirculated. This in principle can be done if a gas sensor to probe precisely the quantity of O2 or the NOx within the exhausted mixture, and at the same time, a contactless magnetic sensor to instantly know the throttle-valve position which adjust the quantity of EGR. For all these reasons to have access to a multifuctional sensor capable to probe, also non simultaneously, a recirculated-gas concentration and the position of a magnetized throttle valve on move would allow to close-loop drive the EGR precisely. The IFOX HS4 is the bases of a multifunctional sensor conceived of probing precisely the quantity of NOx within the exhausted mixture, and at the same time, measuring the position of the EGR valve which adjusts the quantity of EGR. With the presence in the consortium of PST, and TSST, the supply chain of the manufacturing of the HS4 multifunctional sensor is complete, thus paving the way for a promising exploitation strategy of the result which can lead to a start of a new product to serve the worldwide automotive market as a whole if quality and cost will be sufficient over the full life cycle, thus to include to fuel consumption and emission reduction of the future vehicle employing the HS4 assisted combustion control. Although to give a honest assessment on the potential use of HS4 solution other development and analysis including the optimized models and engine control algorithms, as well as a real testing on a running vehicle are required, as well as the evaluation of pilot production of the device to allow the calculation of cost of production. It is a serious intention to plan the further understanding of these open points in a future collaborative research project, possibly in the context of the ICT theme.
Management of intellectual property
PST, has analysed more than 1000 valid patents to preview possible technological bottlenecks possibly preventing use of the know-how developed in the area of buffer layers. To this end, patents today still valid have been matter of intensive investigation, and the solutions protected compared with the new publications in order to evaluate beforehand the possible bottlenecks of the technological choices undertaken in the project, and assess better the interim results of the project against the prior-art. Particular attention was paid to the protected use of the addressed buffer layers to see the perspective of exploitation and screen possible clients of customized materials for post-project co-designing of new potential productions.
From this patent analysis the complex oxide buffer layer developed in the project is very often a prerequisite to conceive a product rather than the core of the idea. This issue opens up the perspective for the buffer material developers (in particular PST, but also TSST more specifically for oxide electronics) in the project to be part of new product supply chains. As one can see from this report, most of products employ today oxide buffer layer of extremely controlled crystallinity and quality standards, while, aside from microelectron- ics, most of them would find suitable, at a lower cost, less quality materials (sensors, dis- crete devices, coatings and electrolyte, fuel cells, etc.). From this assumption the prior ex- ploitation strategy of PST is to offer to the market buffer materials of tailored crystallinity to allow significantly decrease their cost per unit and let novel products penetrate the market.
CRF contributed 6PM on studying the IPR, and the current technologies of sensing to gauge the applicability perspective of the IFOX achievements, and in particular for the heterostructures more related to sensing functionalities, namely HS4 and HS6. A report on sensors for automotive was delivered to the commission at month 24.
CRF expects to generate better and solid IPR on the novel design based on HS4. Right now no patent has been filed yet as more development is needed to increase confidence.
Potential impacts
Socio-economic impact
At present oxides are important for semiconductor technology mainly in the context of improving the dielectric/insulating and interface properties of gate oxides for use in MOSFET devices. However, their real promise for novel future applications is that they can be engineered with additional functionality, e.g. ferroelectric, multiferroic, and magnetic properties, by tailoring their chemical composition and making multilayer stacks of a range of different oxides.
The aim of IFOX was to explore, create and control novel electronic and magnetic functionalities that result from the rich interplay of charge, spin and orbital degrees of freedom in transition metal oxide heterostructures and their interfaces. The outcome aimed at the establishment of a material platform for novel ‘More than Moore’ and ‘beyond CMOS’ electronics, compatible with VLSI and which delivers performance and functionality far beyond the state-of-the-art.
To achieve these goals we carried out a research programme with workpackages on theory and modelling, epitaxial growth, processing, characterization, device fabrication and knowledge transfer. Crucially, technological relevance and impact was guaranteed by the active participation of a world-leading company, Fiat and three SMEs, namely TSST and Organic Spintronics, with expertise and infrastructure for epitaxial oxide growth on Si and Protec Surface Technology, with expertise on engineering and production of PVD and PECVD machinery for thin films and coatings deposition and customised deposition process developments. Focusing the research on industry needs ensured that IFOX worked on relevant problems with tangible results. Success for the project means delivering useable new technology and improved understanding to industry beneficiaries which will allow them to make significant progress on the roadmap for future generations of semiconductor devices.
While the first three periods of IFOX were dedicated to the growth, processing, and investigation of various oxide thin films and their heterostructures, the final period focused on seven different heterostructures, chosen for their largest potential for possible future applications in the target research areas:
1) Manganite/TiO based heterostructures which exhibit resistive switching suitable for novel storage and possibly logic applications
2) Heterostructures based on LaAlO3/SrTiO3 with additional ferromagnetic properties to combine transistor functionality with non-volatile storage for example for logic applications
3) Heterostructures based on BaTiO3 and NiFeO3 with inherent tunneling magnetoresistive functionality which may allow for MRAM cells which are switched by the magnetoelectric coupling between the BTO and the NFO.
4) Heterostructures of (La,Sr)MnO3 (LSMO) and ZnO which combine magnetoresistive sensing and gas sensing properties for automotive applications
5) Heterostructures using a bilayer of (La,Sr)MnO3 and BiFeO3 in which the magnetization of the LSMO can be switched via magnetoelectric coupling to the BFO which again can be switched by an electrical pulse. This structure would also address the area of magnetic memory with low switching power
6) Superlattices of LaMnO3 and SrMnO3 in which the magnetic properties compared to LSMO are much improved at room temperature and very high sensitivity in sensor applications (MR-sensors and Bolometers are to be expected.
7) Heterostructures using ferromagnetic (La,Sr)MnO3, a barrier of SrTiO3 and a ferromagnetic metal counter electrode which realise a combined element for resistive switching and magnetoresistance which can serve as a stateful logic device to combine logic and memory in a single cell.
In addition the project foresaw an effort to attempt and assess the realization of all these heterostructures on large area silicon wafers dramatically reducing possible production costs compared to heterostructures deposited on single crystalline oxide substrates.
All these seven heterostructures were further developed including growth on silicon substrates and a thorough assessment was performed. For two of the seven heterostructures (ZnO/LSMO-based gas sensor and PCMO/TiO based ReRAM) all important factors namely functionality at room temperature or above, fabrication on large area silicon wafers and industry compatible processing were demonstrated. As follow up, CRF will continue the study to increase the TRL of the HS4 up to 6. Besides, to go deep inside into the measurement repeatibility of the gas sensitivity, and the life cycle costing of the production will be part of the future investigation to evaluate the value for money in the replacement of the standard gas sensor today employed. The evaluation will take into account the improved absolute sensitivity of HS4 demonstrated in IFOX. The exploitation of the coupled magneto resistance funtion of HS4 will be requiring further developments and analysis to gauge at what extent the airgap variation sensitivity of the MR response will affect the precision of the absolute position of a mangnetized valve. It is expected to go deeper inside these issue by a second ICT project employing novel advance designed smart systems.
To facilitate the transfer of results to industry all partners contributed to a huge effort, which was guided by the industry partners, to collect the knowledge on growth and processing in a suitable format. This resulted in three major deliverables
1) A catalogue of growth techniques and growth parameters for all heterostructures
2) A catalogue of processing techniques and parameters for all heterostructures
3) A catalogue of all chemicals involved in growth and processing, assessing the environmental impact of the chemical compounds
In addition a tentative assessment of growth and processing in terms of technology readiness level (TRL) is provided, resulting in typical values ranging from 3 to 5 for all steps (see WP5 for details).
Exploration of new functionalities
For all heterostructures an overall assessment is provided which not only addresses the results of characterization. It also provides a placement of the heterostructure within the targeted research area and an evaluation of the maturity of the necessary technology and a competitive comparison to possible alternative devices on the market or in the development chain.
The data of the three catalogues allow for immediate assessment of needs and risks when transferring these heterostrctures to industrial application.
After completion of the programme, we can state that IFOX has delivered results that are highly relevant for our industrial partners in terms of novel sensors, growth and processing and integration to Si.
Societal implications
The importance of the ICT sector from an economic, societal, and (increasingly) environmental viewpoint is obvious: Europe needs a strategic and innovative semiconductor foundation underlying the ICT portfolio to bolster industry, increase innovation and take a lead in economic, societal and environmental terms. The semiconductor market is currently worth 200 billion Euro (256 billion USD) globally on an annual basis. There is a multiplier effect of approximately 25 times that amount in terms of equipment and services ranging from television sets to IT. In other words, semiconductor technology provides the knowledge and technologies that generate some 10% of global GDP. The European semiconductor market attained a value of approximately Euro 30 billion in 2007, representing just 16% of the worldwide total .
IFOX has improved the competitiveness of the EU semiconductor industry by its development of novel device technologies based on oxide materials with novel device functionalities beyond the current state-of-the-art. This will support the need for industry linked technological innovation in the EU as described in the 2008 Europe Innova Innovation watch report: "A prospective innovation challenge for the ICT sector is the insufficient collaboration between public sector and business sector researchers".
The results obtained in IFOX is multi-disciplinary and beyond what could have been accomplished by one or two individual research groups as they do not have the expertise nor infrastructure for all the various aspects of theory, growth, structuring and characterization necessary for the achieved functional structures. In addition, without the critical evaluation and feedback of the industrial partners, the focusing on the most relevant and prospective structures and their compatibility with Si technology would not have been accomplished.
Regarding impact of research, the EU still ranks as the world’s largest producer of scientific knowledge (measured by publications), but contributes less than the US to high impact publications . These data provide the backdrop against which the consortium has tackled its research goals. The best research conducted within IFOX has been published in high-profile peer reviewed journals and the researchers have presented their findings at leading international conferences.
IFOX has provided the groundwork and knowledge necessary for magnetic oxide inclusion in future generations of ICT products, critically, by engaging a key end-user company Fiat. The new knowledge generated by IFOX is targeted in such a critical area that it will be truly groundbreaking in worldwide terms and put Europe ahead of other regions, safeguarding existing ICT jobs in Europe and working to create new markets for job increases. This should help the stagnation of the EU-27's R&D investment (R&D expenditure as % of GDP) which is currently at 1.84% vs. the Lisbon target of 3%.
Dissemination and/or exploitation of project results, and management of intellectual property
The IFOX programme is centred on increasing the knowledge and understanding about transition metal oxides and their properties and also on applying this knowledge to create new oxide functionalities for potential device application. A key component of this effort is to generate and protect intellectual property (IP) with a view towards exploitation and delivering novel device functionalities in industry.
Dissemination
IFOX contributed during its whole duration to create and disseminate knowledge on complex oxides and interfaces to the overall community. In particular, the consortium published on peer reviewed international journals 91 manuscripts, and dissemination to the major inter- national conferences almost the same number of talks.
IFOX also disseminated results in the frame of WP9 by organizing a Summer school which took place in Bildungszentrum Hesselberg from July 29° to August the 3° 2014 that count- ed 35 students recruited within IFOX and 10 external students recruited by other projects attending. Nine speakers were experts of IFOX and one was invited from the ORAMA project.
Towards the end of the project the dissemination became more and more important. During the 2015 E-MRS Fall Meeting, held in Warsaw University of Technology, from September 15 to 18 2015, the IFOX consortium was presented within symposium L "Materials for electronics and optoelectronic applications away from silicon”. The IFOX session had the following talks:
11:00 Gertjan Koster, University of Twente, Enschede, the Netherlands, "Integrating epi- taxial oxides with Si"
11:15 Rik Groenen, TSST, Enschede, the Netherlands, "large area epitaxial oxide growth using PLD"
11:30 Gwenael Atcheson, TCD, Dublin, Ireland, "All-oxide multiferroic tunnel junctions"
11:45 Patrizio Graziosi CNR-ISMN, "Memristive and magnetoresistive properties of
SrTiO3 based junctions" 12:00 Nicolas Gauquelin, EMAT, University of Antwerp, "Application of advanced mi- croscopy techniques to the study of oxide-based electronic devices"
12:15 Mehran Vafaee, Johannes-Gutenberg-Universität Mainz, Mainz, Germany "The cor- relation between the interface roughness and exchange bias coupling strength in BiFeO3/La0.7Sr0.3MnO3 heterostructures"
IFOX also setup a booth at this E-MRS meeting with posters on all heterostructures in which discussions with possibly interested industry partners were held. In addition, a massive search in patent databases was done in order to evaluate possible restrictions and possible areas for protection of IP and to identify competing technologies.
Exploitation of results
With respect to exploitation, the industrial partners were mostly responsible for the analysis according to the respective man power allocated. Magnetron sputtering is acknowledged as being one of the most promising techniques for industrial use perspective, and it was used here by PST to exploit the production of complex oxide buffer layers for many industrial sectors (both for low- and high-end prod- ucts) where thin films are a mass-production major issue. On the other hand, TSST and OS are actively involved in the production of deposition equipment for these applications. Finally, CRF is one of the major end users for the novel sensors developed within IFOX.
PST, besides analysing more than 1000 valid patents to preview possible technological bottlenecks possibly preventing use of the know-how developed (buffer layers), also found through the analysis the possible market incubators to place their machinery and processes. The analysis carried out to foresee the use potential of the complex oxide buffer materials developed in the project took into account the material quality level in terms of (i) crystallinity (H/E: highly crystalline/Epitaxial, HT: highly tex- tured, LT: low texturing, P: polycrystalline), (ii) adhesion, and (iii) homogeneity over area as large as 4” to address that single applications according to the material required need. This analysis is essential to tailor the material production according to the chosen use and properly penetrate the markets addressed.
From this analysis the exploitation strategy of PST is to offer to the market buffer materials of tailored crystallinity to allow significantly decrease their cost for unit and let novel product penetrate the market.
TSST has built a prototype for large area deposition PLD and developed the necessary processing technology in the course of IFOX. Upscaling is demonstrated of the synthesis of complex oxide films with perfect control over the structure and composition of the material on large wafer size Si up to 4”. This has led to new standards for the precision synthesis of oxide films, and this removed in a radical manner the limitations on the availability of such films on industrial scale Si wafers. Important technological steps, including technologies to uniformly deposit the material on a uniformly heated substrate, have been made. These steps have enabled the epitaxial growth of crystalline complex oxide thin films on Si. The PLD system is at present ready to enter the market, enabling TSST to increase the market share. TSST hopes to expand their sales by 30% in the next coming years.
OS has in the course of IFOX built a prototype for a multisource pulsed plasma deposition system, (PPD - 5 parallel sources), able to deposit oxides with 6% of homogeneity on 3‘‘ wafers and on an ‘‘infinite‘‘ strips of 300mm width. This industrial prototype is already working in an Industrial Research Centre and almost ready to enter the market in 2017. At the same time, OS has started to develop an enlargement of the area covered by a single source. The result is that the last prototype system uses only two sources to deposit on substrate areas of 300mm width instead of the previous 5 sources.
OS hopes to expand their sales by 2018 by dealing this type of system specifically adapted for customer applications.
Another important business strategy that OS will follow is the possibility to insert a PPD source in an already used industrials system since PPD deposition rate has become compatible with similar deposition techniques.
For CRF, exhaust gas sensors is the main application of the IFOX complex oxides developed, namely: YSZ, CeO2, ZnO, and LSMO; oxides which are all sensitive to the gas exhausted by internal combustion engine powered cars, which do guarantee improved sensitivity and selectivity in the shape of sandwiches as e.g. HS4 addressed by IFOX. A plurality of sensor types and electrochemical detection systems are in use for the polluting gaseous compounds exhausted by the current production internal combustion engines. In particular, the focus is placed on the detection systems dedicated to the detection of carbon mono-oxide and di-oxide (CO, CO2), and of nitrogen oxides (NOx), used to limit noxious emissions, to control the EGR flux, and to detect the combustion misfire. Nitrogen oxide gases, commonly referred to collectively as NOX, are common pollutants formed in internal combustion engines and industrial combustion systems by thermal fixation and oxidation of atmospheric nitrogen1. NOX, which includes NO2 and NO gases, have adverse effects on the environment, and are the leading cause of green-house effect, acid rain, and photochemical smog. In humans, exposure to more than 3 ppm of NO2 gas for periods longer than 8 hours can cause respiratory and cardiovascular diseases. Hence, better detection of such gases is of utmost importance, and that calls for better understanding of detection mechanism to facilitate development and optimization of sensing devices. Semiconducting metal oxides based thin films have traditionally been used in detection of NOX gases, with the common measuring arrangement relying on changes in electrical conductivity upon interaction with the gas. In particular, ZnO based gas sensors have been employed extensively in the detection of both reducing and oxidizing gases due to the superior properties of this material such as enhanced sensitivity, stability, and low cost. A widely adopted route to reduce NOx emissions is Exhaust Gas Recirculation (EGR). The EGR system has a valve which cannot respond instantly to changes in demand, and the exhaust gas takes time to flow around the EGR circuit. This makes the calibration of transient EGR behaviour particularly complex. To have a better control of EGR it would be very important to be capable of sensing the NOx content, or relatively knowing the actual O2 concentration in the exhaust air to be recirculated, and at the same time controlling precisely the quantity of gas recirculated. This in principle can be done if a gas sensor to probe precisely the quantity of O2 or the NOx within the exhausted mixture, and at the same time, a contactless magnetic sensor to instantly know the throttle-valve position which adjust the quantity of EGR. For all these reasons to have access to a multifuctional sensor capable to probe, also non simultaneously, a recirculated-gas concentration and the position of a magnetized throttle valve on move would allow to close-loop drive the EGR precisely. The IFOX HS4 is the bases of a multifunctional sensor conceived of probing precisely the quantity of NOx within the exhausted mixture, and at the same time, measuring the position of the EGR valve which adjusts the quantity of EGR. With the presence in the consortium of PST, and TSST, the supply chain of the manufacturing of the HS4 multifunctional sensor is complete, thus paving the way for a promising exploitation strategy of the result which can lead to a start of a new product to serve the worldwide automotive market as a whole if quality and cost will be sufficient over the full life cycle, thus to include to fuel consumption and emission reduction of the future vehicle employing the HS4 assisted combustion control. Although to give a honest assessment on the potential use of HS4 solution other development and analysis including the optimized models and engine control algorithms, as well as a real testing on a running vehicle are required, as well as the evaluation of pilot production of the device to allow the calculation of cost of production. It is a serious intention to plan the further understanding of these open points in a future collaborative research project, possibly in the context of the ICT theme.
Management of intellectual property
PST, has analysed more than 1000 valid patents to preview possible technological bottlenecks possibly preventing use of the know-how developed in the area of buffer layers. To this end, patents today still valid have been matter of intensive investigation, and the solutions protected compared with the new publications in order to evaluate beforehand the possible bottlenecks of the technological choices undertaken in the project, and assess better the interim results of the project against the prior-art. Particular attention was paid to the protected use of the addressed buffer layers to see the perspective of exploitation and screen possible clients of customized materials for post-project co-designing of new potential productions.
From this patent analysis the complex oxide buffer layer developed in the project is very often a prerequisite to conceive a product rather than the core of the idea. This issue opens up the perspective for the buffer material developers (in particular PST, but also TSST more specifically for oxide electronics) in the project to be part of new product supply chains. As one can see from this report, most of products employ today oxide buffer layer of extremely controlled crystallinity and quality standards, while, aside from microelectron- ics, most of them would find suitable, at a lower cost, less quality materials (sensors, dis- crete devices, coatings and electrolyte, fuel cells, etc.). From this assumption the prior ex- ploitation strategy of PST is to offer to the market buffer materials of tailored crystallinity to allow significantly decrease their cost per unit and let novel products penetrate the market.
CRF contributed 6PM on studying the IPR, and the current technologies of sensing to gauge the applicability perspective of the IFOX achievements, and in particular for the heterostructures more related to sensing functionalities, namely HS4 and HS6. A report on sensors for automotive was delivered to the commission at month 24.
CRF expects to generate better and solid IPR on the novel design based on HS4. Right now no patent has been filed yet as more development is needed to increase confidence.
Potential impacts
Socio-economic impact
At present oxides are important for semiconductor technology mainly in the context of improving the dielectric/insulating and interface properties of gate oxides for use in MOSFET devices. However, their real promise for novel future applications is that they can be engineered with additional functionality, e.g. ferroelectric, multiferroic, and magnetic properties, by tailoring their chemical composition and making multilayer stacks of a range of different oxides.
The aim of IFOX was to explore, create and control novel electronic and magnetic functionalities that result from the rich interplay of charge, spin and orbital degrees of freedom in transition metal oxide heterostructures and their interfaces. The outcome aimed at the establishment of a material platform for novel ‘More than Moore’ and ‘beyond CMOS’ electronics, compatible with VLSI and which delivers performance and functionality far beyond the state-of-the-art.
To achieve these goals we carried out a research programme with workpackages on theory and modelling, epitaxial growth, processing, characterization, device fabrication and knowledge transfer. Crucially, technological relevance and impact was guaranteed by the active participation of a world-leading company, Fiat and three SMEs, namely TSST and Organic Spintronics, with expertise and infrastructure for epitaxial oxide growth on Si and Protec Surface Technology, with expertise on engineering and production of PVD and PECVD machinery for thin films and coatings deposition and customised deposition process developments. Focusing the research on industry needs ensured that IFOX worked on relevant problems with tangible results. Success for the project means delivering useable new technology and improved understanding to industry beneficiaries which will allow them to make significant progress on the roadmap for future generations of semiconductor devices.
While the first three periods of IFOX were dedicated to the growth, processing, and investigation of various oxide thin films and their heterostructures, the final period focused on seven different heterostructures, chosen for their largest potential for possible future applications in the target research areas:
1) Manganite/TiO based heterostructures which exhibit resistive switching suitable for novel storage and possibly logic applications
2) Heterostructures based on LaAlO3/SrTiO3 with additional ferromagnetic properties to combine transistor functionality with non-volatile storage for example for logic applications
3) Heterostructures based on BaTiO3 and NiFeO3 with inherent tunneling magnetoresistive functionality which may allow for MRAM cells which are switched by the magnetoelectric coupling between the BTO and the NFO.
4) Heterostructures of (La,Sr)MnO3 (LSMO) and ZnO which combine magnetoresistive sensing and gas sensing properties for automotive applications
5) Heterostructures using a bilayer of (La,Sr)MnO3 and BiFeO3 in which the magnetization of the LSMO can be switched via magnetoelectric coupling to the BFO which again can be switched by an electrical pulse. This structure would also address the area of magnetic memory with low switching power
6) Superlattices of LaMnO3 and SrMnO3 in which the magnetic properties compared to LSMO are much improved at room temperature and very high sensitivity in sensor applications (MR-sensors and Bolometers are to be expected.
7) Heterostructures using ferromagnetic (La,Sr)MnO3, a barrier of SrTiO3 and a ferromagnetic metal counter electrode which realise a combined element for resistive switching and magnetoresistance which can serve as a stateful logic device to combine logic and memory in a single cell.
In addition the project foresaw an effort to attempt and assess the realization of all these heterostructures on large area silicon wafers dramatically reducing possible production costs compared to heterostructures deposited on single crystalline oxide substrates.
All these seven heterostructures were further developed including growth on silicon substrates and a thorough assessment was performed. For two of the seven heterostructures (ZnO/LSMO-based gas sensor and PCMO/TiO based ReRAM) all important factors namely functionality at room temperature or above, fabrication on large area silicon wafers and industry compatible processing were demonstrated. As follow up, CRF will continue the study to increase the TRL of the HS4 up to 6. Besides, to go deep inside into the measurement repeatibility of the gas sensitivity, and the life cycle costing of the production will be part of the future investigation to evaluate the value for money in the replacement of the standard gas sensor today employed. The evaluation will take into account the improved absolute sensitivity of HS4 demonstrated in IFOX. The exploitation of the coupled magneto resistance funtion of HS4 will be requiring further developments and analysis to gauge at what extent the airgap variation sensitivity of the MR response will affect the precision of the absolute position of a mangnetized valve. It is expected to go deeper inside these issue by a second ICT project employing novel advance designed smart systems.
To facilitate the transfer of results to industry all partners contributed to a huge effort, which was guided by the industry partners, to collect the knowledge on growth and processing in a suitable format. This resulted in three major deliverables
1) A catalogue of growth techniques and growth parameters for all heterostructures
2) A catalogue of processing techniques and parameters for all heterostructures
3) A catalogue of all chemicals involved in growth and processing, assessing the environmental impact of the chemical compounds
In addition a tentative assessment of growth and processing in terms of technology readiness level (TRL) is provided, resulting in typical values ranging from 3 to 5 for all steps (see WP5 for details).
Exploration of new functionalities
For all heterostructures an overall assessment is provided which not only addresses the results of characterization. It also provides a placement of the heterostructure within the targeted research area and an evaluation of the maturity of the necessary technology and a competitive comparison to possible alternative devices on the market or in the development chain.
The data of the three catalogues allow for immediate assessment of needs and risks when transferring these heterostrctures to industrial application.
After completion of the programme, we can state that IFOX has delivered results that are highly relevant for our industrial partners in terms of novel sensors, growth and processing and integration to Si.
Societal implications
The importance of the ICT sector from an economic, societal, and (increasingly) environmental viewpoint is obvious: Europe needs a strategic and innovative semiconductor foundation underlying the ICT portfolio to bolster industry, increase innovation and take a lead in economic, societal and environmental terms. The semiconductor market is currently worth 200 billion Euro (256 billion USD) globally on an annual basis. There is a multiplier effect of approximately 25 times that amount in terms of equipment and services ranging from television sets to IT. In other words, semiconductor technology provides the knowledge and technologies that generate some 10% of global GDP. The European semiconductor market attained a value of approximately Euro 30 billion in 2007, representing just 16% of the worldwide total .
IFOX has improved the competitiveness of the EU semiconductor industry by its development of novel device technologies based on oxide materials with novel device functionalities beyond the current state-of-the-art. This will support the need for industry linked technological innovation in the EU as described in the 2008 Europe Innova Innovation watch report: "A prospective innovation challenge for the ICT sector is the insufficient collaboration between public sector and business sector researchers".
The results obtained in IFOX is multi-disciplinary and beyond what could have been accomplished by one or two individual research groups as they do not have the expertise nor infrastructure for all the various aspects of theory, growth, structuring and characterization necessary for the achieved functional structures. In addition, without the critical evaluation and feedback of the industrial partners, the focusing on the most relevant and prospective structures and their compatibility with Si technology would not have been accomplished.
Regarding impact of research, the EU still ranks as the world’s largest producer of scientific knowledge (measured by publications), but contributes less than the US to high impact publications . These data provide the backdrop against which the consortium has tackled its research goals. The best research conducted within IFOX has been published in high-profile peer reviewed journals and the researchers have presented their findings at leading international conferences.
IFOX has provided the groundwork and knowledge necessary for magnetic oxide inclusion in future generations of ICT products, critically, by engaging a key end-user company Fiat. The new knowledge generated by IFOX is targeted in such a critical area that it will be truly groundbreaking in worldwide terms and put Europe ahead of other regions, safeguarding existing ICT jobs in Europe and working to create new markets for job increases. This should help the stagnation of the EU-27's R&D investment (R&D expenditure as % of GDP) which is currently at 1.84% vs. the Lisbon target of 3%.
Dissemination and/or exploitation of project results, and management of intellectual property
The IFOX programme is centred on increasing the knowledge and understanding about transition metal oxides and their properties and also on applying this knowledge to create new oxide functionalities for potential device application. A key component of this effort is to generate and protect intellectual property (IP) with a view towards exploitation and delivering novel device functionalities in industry.
Dissemination
IFOX contributed during its whole duration to create and disseminate knowledge on complex oxides and interfaces to the overall community. In particular, the consortium published on peer reviewed international journals 91 manuscripts, and dissemination to the major inter- national conferences almost the same number of talks.
IFOX also disseminated results in the frame of WP9 by organizing a Summer school which took place in Bildungszentrum Hesselberg from July 29° to August the 3° 2014 that count- ed 35 students recruited within IFOX and 10 external students recruited by other projects attending. Nine speakers were experts of IFOX and one was invited from the ORAMA project.
Towards the end of the project the dissemination became more and more important. During the 2015 E-MRS Fall Meeting, held in Warsaw University of Technology, from September 15 to 18 2015, the IFOX consortium was presented within symposium L "Materials for electronics and optoelectronic applications away from silicon”. The IFOX session had the following talks:
11:00 Gertjan Koster, University of Twente, Enschede, the Netherlands, "Integrating epi- taxial oxides with Si"
11:15 Rik Groenen, TSST, Enschede, the Netherlands, "large area epitaxial oxide growth using PLD"
11:30 Gwenael Atcheson, TCD, Dublin, Ireland, "All-oxide multiferroic tunnel junctions"
11:45 Patrizio Graziosi CNR-ISMN, "Memristive and magnetoresistive properties of
SrTiO3 based junctions" 12:00 Nicolas Gauquelin, EMAT, University of Antwerp, "Application of advanced mi- croscopy techniques to the study of oxide-based electronic devices"
12:15 Mehran Vafaee, Johannes-Gutenberg-Universität Mainz, Mainz, Germany "The cor- relation between the interface roughness and exchange bias coupling strength in BiFeO3/La0.7Sr0.3MnO3 heterostructures"
IFOX also setup a booth at this E-MRS meeting with posters on all heterostructures in which discussions with possibly interested industry partners were held. In addition, a massive search in patent databases was done in order to evaluate possible restrictions and possible areas for protection of IP and to identify competing technologies.
Exploitation of results
With respect to exploitation, the industrial partners were mostly responsible for the analysis according to the respective man power allocated. Magnetron sputtering is acknowledged as being one of the most promising techniques for industrial use perspective, and it was used here by PST to exploit the production of complex oxide buffer layers for many industrial sectors (both for low- and high-end prod- ucts) where thin films are a mass-production major issue. On the other hand, TSST and OS are actively involved in the production of deposition equipment for these applications. Finally, CRF is one of the major end users for the novel sensors developed within IFOX.
PST, besides analysing more than 1000 valid patents to preview possible technological bottlenecks possibly preventing use of the know-how developed (buffer layers), also found through the analysis the possible market incubators to place their machinery and processes. The analysis carried out to foresee the use potential of the complex oxide buffer materials developed in the project took into account the material quality level in terms of (i) crystallinity (H/E: highly crystalline/Epitaxial, HT: highly tex- tured, LT: low texturing, P: polycrystalline), (ii) adhesion, and (iii) homogeneity over area as large as 4” to address that single applications according to the material required need. This analysis is essential to tailor the material production according to the chosen use and properly penetrate the markets addressed.
From this analysis the exploitation strategy of PST is to offer to the market buffer materials of tailored crystallinity to allow significantly decrease their cost for unit and let novel product penetrate the market.
TSST has built a prototype for large area deposition PLD and developed the necessary processing technology in the course of IFOX. Upscaling is demonstrated of the synthesis of complex oxide films with perfect control over the structure and composition of the material on large wafer size Si up to 4”. This has led to new standards for the precision synthesis of oxide films, and this removed in a radical manner the limitations on the availability of such films on industrial scale Si wafers. Important technological steps, including technologies to uniformly deposit the material on a uniformly heated substrate, have been made. These steps have enabled the epitaxial growth of crystalline complex oxide thin films on Si. The PLD system is at present ready to enter the market, enabling TSST to increase the market share. TSST hopes to expand their sales by 30% in the next coming years.
OS has in the course of IFOX built a prototype for a multisource pulsed plasma deposition system, (PPD - 5 parallel sources), able to deposit oxides with 6% of homogeneity on 3‘‘ wafers and on an ‘‘infinite‘‘ strips of 300mm width. This industrial prototype is already working in an Industrial Research Centre and almost ready to enter the market in 2017. At the same time, OS has started to develop an enlargement of the area covered by a single source. The result is that the last prototype system uses only two sources to deposit on substrate areas of 300mm width instead of the previous 5 sources.
OS hopes to expand their sales by 2018 by dealing this type of system specifically adapted for customer applications.
Another important business strategy that OS will follow is the possibility to insert a PPD source in an already used industrials system since PPD deposition rate has become compatible with similar deposition techniques.
For CRF, exhaust gas sensors is the main application of the IFOX complex oxides developed, namely: YSZ, CeO2, ZnO, and LSMO; oxides which are all sensitive to the gas exhausted by internal combustion engine powered cars, which do guarantee improved sensitivity and selectivity in the shape of sandwiches as e.g. HS4 addressed by IFOX. A plurality of sensor types and electrochemical detection systems are in use for the polluting gaseous compounds exhausted by the current production internal combustion engines. In particular, the focus is placed on the detection systems dedicated to the detection of carbon mono-oxide and di-oxide (CO, CO2), and of nitrogen oxides (NOx), used to limit noxious emissions, to control the EGR flux, and to detect the combustion misfire. Nitrogen oxide gases, commonly referred to collectively as NOX, are common pollutants formed in internal combustion engines and industrial combustion systems by thermal fixation and oxidation of atmospheric nitrogen1. NOX, which includes NO2 and NO gases, have adverse effects on the environment, and are the leading cause of green-house effect, acid rain, and photochemical smog. In humans, exposure to more than 3 ppm of NO2 gas for periods longer than 8 hours can cause respiratory and cardiovascular diseases. Hence, better detection of such gases is of utmost importance, and that calls for better understanding of detection mechanism to facilitate development and optimization of sensing devices. Semiconducting metal oxides based thin films have traditionally been used in detection of NOX gases, with the common measuring arrangement relying on changes in electrical conductivity upon interaction with the gas. In particular, ZnO based gas sensors have been employed extensively in the detection of both reducing and oxidizing gases due to the superior properties of this material such as enhanced sensitivity, stability, and low cost. A widely adopted route to reduce NOx emissions is Exhaust Gas Recirculation (EGR). The EGR system has a valve which cannot respond instantly to changes in demand, and the exhaust gas takes time to flow around the EGR circuit. This makes the calibration of transient EGR behaviour particularly complex. To have a better control of EGR it would be very important to be capable of sensing the NOx content, or relatively knowing the actual O2 concentration in the exhaust air to be recirculated, and at the same time controlling precisely the quantity of gas recirculated. This in principle can be done if a gas sensor to probe precisely the quantity of O2 or the NOx within the exhausted mixture, and at the same time, a contactless magnetic sensor to instantly know the throttle-valve position which adjust the quantity of EGR. For all these reasons to have access to a multifuctional sensor capable to probe, also non simultaneously, a recirculated-gas concentration and the position of a magnetized throttle valve on move would allow to close-loop drive the EGR precisely. The IFOX HS4 is the bases of a multifunctional sensor conceived of probing precisely the quantity of NOx within the exhausted mixture, and at the same time, measuring the position of the EGR valve which adjusts the quantity of EGR. With the presence in the consortium of PST, and TSST, the supply chain of the manufacturing of the HS4 multifunctional sensor is complete, thus paving the way for a promising exploitation strategy of the result which can lead to a start of a new product to serve the worldwide automotive market as a whole if quality and cost will be sufficient over the full life cycle, thus to include to fuel consumption and emission reduction of the future vehicle employing the HS4 assisted combustion control. Although to give a honest assessment on the potential use of HS4 solution other development and analysis including the optimized models and engine control algorithms, as well as a real testing on a running vehicle are required, as well as the evaluation of pilot production of the device to allow the calculation of cost of production. It is a serious intention to plan the further understanding of these open points in a future collaborative research project, possibly in the context of the ICT theme.
Management of intellectual property
PST, has analysed more than 1000 valid patents to preview possible technological bottlenecks possibly preventing use of the know-how developed in the area of buffer layers. To this end, patents today still valid have been matter of intensive investigation, and the solutions protected compared with the new publications in order to evaluate beforehand the possible bottlenecks of the technological choices undertaken in the project, and assess better the interim results of the project against the prior-art. Particular attention was paid to the protected use of the addressed buffer layers to see the perspective of exploitation and screen possible clients of customized materials for post-project co-designing of new potential productions.
From this patent analysis the complex oxide buffer layer developed in the project is very often a prerequisite to conceive a product rather than the core of the idea. This issue opens up the perspective for the buffer material developers (in particular PST, but also TSST more specifically for oxide electronics) in the project to be part of new product supply chains. As one can see from this report, most of products employ today oxide buffer layer of extremely controlled crystallinity and quality standards, while, aside from microelectron- ics, most of them would find suitable, at a lower cost, less quality materials (sensors, dis- crete devices, coatings and electrolyte, fuel cells, etc.). From this assumption the prior ex- ploitation strategy of PST is to offer to the market buffer materials of tailored crystallinity to allow significantly decrease their cost per unit and let novel products penetrate the market.
CRF contributed 6PM on studying the IPR, and the current technologies of sensing to gauge the applicability perspective of the IFOX achievements, and in particular for the heterostructures more related to sensing functionalities, namely HS4 and HS6. A report on sensors for automotive was delivered to the commission at month 24.
CRF expects to generate better and solid IPR on the novel design based on HS4. Right now no patent has been filed yet as more development is needed to increase confidence.
“Mastering Innovation – Shaping the Future”, ESIA 2008 Competitiveness Report
ii Technology and Competitiveness (STandC) key figures 2008 (http://ec.europa.eu/research/era/)
List of Websites:
www.ifox-project.eu
The goal of IFOX was to explore, create and control novel electronic and magnetic functionalities, with focus on interfaces, in complex transition metal oxide heterostructures to develop the material platform for ‘More than Moore’ (MtM) and ‘beyond CMOS’ electronics, VLSI integratable with performance and functionality far beyond the state-of-the-art.
In order to reach these goals an integrated effort was started which included theory, growth, processing, analysis and testing whereas at the same time the potential impact and IP was addressed. Theoretical modelling was developed and used to identify new oxide materials and interfaces with enhanced functionality to inspire the growth of new structures. Growth processes for these materials and heterostructures were developed and optimized on academic and industrial scale, to test growth of and ultimately provide growth processes for oxides which can be applied in industry. The processes were demonstrated to be able to control the growth with single atomic layer precision. In parallel, growth on silicon substrates using suitable buffer systems was investigated for future CMOS compatibility. As a result several major advances were reported including for example metal organic aerosol deposition of oxides with in situ ellipsometry and large area pulsed laser deposition and pulsed plasma deposition systems for four and three inch wafers, respectively, the latter ready for sale by the developing company. This was complemented by the parallel development of processing technologies that up to now were unavailable for these materials or unsuitable for industrial application. Characterization of the heterostructures was a major effort, which encompassed electrical, magnetic, optical and structural characterization down to the atomic scale. In addition, major breakthroughs in characterization techniques were achieved: for example ultra-high resolution transition electron microscopy and high speed magneto-optics have made a leap which now allows much deeper insight in oxide structures and magnetic properties on the sub nm and ps length and time scale, respectively. Characterization and growth both provided feedback to theory, showing which model systems or functionalities can be realized, information which again could be used by theory to direct the effort towards the most promising and most supporting systems and tasks at hand.
Because IFOX targeted industrial applications, the industry partners supervised the work and evaluated the results. Based on these evaluations, after three years a list of the seven most promising oxide heterostructures with functionalities targeting the research areas of IFOX was established and put into a new description of work. The final period of IFOX was completely dedicated to the realization of these heterostructures, including to grow them on large area silicon substrates. For two of the seven heterostructures functionality could be demonstrated indeed on large area silicon substrates. For the others (except one) growth on silicon was performed and at least partial functionality could be demonstrated. This effort was supported by a new partner (Protec Surface Technology) who specializes in large area industrial deposition systems and supported the effort of upscaling to large area substrates. All seven heterostructures were finally assessed in terms of maturity down to the technology readiness level of growth and processing and competitiveness in a targeted field of application.
As an important result, all data on growth and processing were collected in several web-based interactive databases and evaluated by the industrial partners. From these data three catalogues were established, one on growth processes, one on processing and one on possible environmental impact of chemistry which allow for immediate assessment of needs and risks when transferring these heterostructures to industrial application.
It should finally be mentioned that all deliverables of the project were achieved.
Project Context and Objectives:
The goal of IFOX was to explore, create and control novel electronic and magnetic functionalities that result from the rich interplay of charge, spin and orbital degrees of freedom in transition metal oxide heterostructures and their interfaces. The desired outcome was the establishment of a material platform for novel ‘More than Moore’ and ‘beyond CMOS’ electronics, compatible with VLSI and which delivers performance and functionality far beyond the state-of-the-art. The consortium should:
(1) establish a theoretical basis to identify the most promising materials/heterostructures and to understand the new functionalities relevant for applications;
(2) grow oxide films on commercial substrates with a quality comparable to state-of-the-art semiconductor growth;
(3) establish oxide patterning and processing conditions within and beyond the constraints of current fabrication technologies; and
(4) characterize structural, electronic and magnetic properties as a prerequisite to delivering novel device concepts for memory, programmable logic and sensors.
A multidisciplinary consortium was formed to address these challenges and build on complementary expertise in theory and modelling, epitaxial growth, processing, characterization, device fabrication and knowledge transfer. Crucially, technological relevance was guaranteed by the active participation of a world-leading company: Fiat and two SMEs, namely TSST and Organic Spintronics. A third SME, Protec Surface Technology, later joined the consortium to strengthen the effort of upscaling the oxide technology towards large area VLSI compatible substrates. The project was central to section NMP-2009-2.2-1 of Theme 4 of the NMP Call (Oxide materials for electronics applications).
The project was based on the following assessment: Transition metal oxides offer a wide variety of exploitable properties. Depending on composition, they can be insulating, semiconducting or metallic, ferromagnetic, ferroelectric or even multiferroic. Interfaces between different oxides often show electronic reconstruction. They can generate new functionalities and be utilised, for example, in high performance tunnelling magnetoresistance (TMR) junctions or spin-polarized interfacial electron gases, as recently demonstrated by members of IFOX. By combining ferroelectric and ferromagnetic oxides, artificial multiferroics with very strong coupling constants can be formed. The advantages of transition metal oxides and oxide heterostructures over other materials are: scalability – because of the stronger interactions and therefore, typically, more localised electronic states; versatility – because of the very rich interplay of different phenomena in oxides; stability – because of the robustness with respect to oxidation or diffusion; and non-volatility and low energy consumption - because of the intrinsic behaviour of ferromagnetic and ferroelectric ordering. Despite this rich potential for use in electronics, actual applications were relatively few in the beginning of IFOX as they rely on atomic scale control and understanding of these materials and their interfaces. The main ambition in IFOX was to improve the structural quality of interfaces by developing and selecting optimal growth and processing protocols and to realise tailored magnetic and electronic properties by achieving a deeper understanding of interface-related phenomena. This should lead to the required high degree of control for novel spin- and/or charge-based electronics in the framework of CMOS fabrication technology, but with next-generation performance and functionality. Applications were defined in the beginning and during the project by the needs of our major industrial beneficiary; technology transfer was facilitated by two SMEs specialized in industrial-scale oxide deposition systems, and the third SME, Protec Surface technology, who specializes in large area industry industrial sputtering systems.
To understand the workplan of IFOX it is necessary to understand that certain key ingredients necessary to access the entire potential of oxides for electronic integrated circuits were missing when IFOX started:
(1) intimate understanding of novel device-relevant functionalities that originate from the electronic reconstructions at interfaces between complex oxide layers, such as (spin-polarized) two-dimensional electron gases, multiferroic attributes, highly efficient spin filters, etc;
(2) full control of thin film growth and interface matching with substrate materials at sizes suitable for industrial applications; and
(3) comprehensive processing parameters, ranging from layer deposition to complete device fabrication, all compatible with mainstream CMOS technology, either in back end or front end processes, and any emergent successor technology.
IFOX was targeting the following four major areas of oxide heterostructures, all of which closely linked to technological applications to open up new opportunities for our industrial beneficiaries.
(1) Ferromagnetic oxide heterostructures controlled by charge or spin-polarized currents in advanced TMR structures for memory, programmable logic, sensors, and oxide/semiconductor spintronics.
(2) Ferroelectric oxide heterostructures controlled by charge and electric fields for sensors, nanoionic switches and fused memory-logic elements in which ferroelectric polarization and/or redox processes are used to create hysteretic resistance changes.
(3) Multifunctional multiferroic oxide heterostructures controlled either by charge or by spin and ultimately by ultrashort light pulses and capable of supplying both electric and magnetic fields on a very short timescale. These will be integrated into concepts for novel magneto-electronics and opto-magneto-electronics devices with unprecedented functionality.
(4) Growth of high quality oxide layers on large silicon substrates, essential for the implementation of all three material classes into electronic applications.
The major technological challenges to be mastered in order to realise advanced industrial applications were defined as:
• grow the oxide films of interest on Si and SiOx with a quality comparable to state-of-the-art semiconductor growth;
• increase the substrate size for typical oxide deposition methods to 200 mm and beyond;
• establish processing conditions for the relevant oxide heterostructures to allow nanoscale control and reproducibility of material parameters whilst maintaining CMOS compatibility;
• control the electronic reconstruction at the interfaces;
• understand and control the influence of interfacial strain.
The above are underpinned by fundamental challenges in understanding and predicting:
• electronic reconstructions at the interfaces;
• spin tunnelling and transport across oxide interfaces;
• fundamental properties of nanoscale oxide elements with respect to the ionic and electronic contributions to the screening of polarization charges at the electrode -oxide interfaces;
• the interaction of redox-centers and extended defects with ferroelectric polarization charges;
• interface coupling between ferromagnetic and ferroelectric oxides in artificial multiferroics.
Each challenge relates to interfacial properties, starting from the substrate/oxide interface that determines the quality of epitaxial growth, the internal oxide/oxide interfaces that confer functionality, the oxide/electrical contact interface at which charge redistribution and intermixing can influence device performance, to the oxide/air interface at the surface. The IFOX methodology was to achieve full control and understanding of each interface through extensive and imaginative investigations. The wide range of methods available in IFOX and their application to charge and spin controlled oxides should give rise to new insights and lead to novel device concepts. Finally, unprecedentedly fast optical control of certain oxides - particularly artificial multiferroics - will be explored, leading to new concepts for magneto-optic electronics.
Four target research areas were defined for which suitable heterostructures and concepts should be developed:
Magnetoresistive tunnel junctions
Magnetoresistive tunnel junctions are still the major candidates for magnetic random access memory (MRAM). At the beginning of IFOX the state of the art MRAM using MgO barriers and spin transfer torque for programming yielded prototypes almost ready for the market. However, the parameters in terms of the necessary currents for switching were too high. IFOX was to investigate and develop novel oxide materials which, when integrated in heterostructures, should reduce the switching current due to their intrinsic high spin polarization. This included also a major effort in the optimization of novel ferromagnetic oxides which are conducting and ferromagnetic well beyond room temperature.
Ferroelectric and valence-change effects in oxides
Ferroelectrics (FEs) and related oxides have been subject to intensive studies because of possible applications for FE memories. Redox-based resistance switching effects, however, exhibit much better scalability but at the potential expense of retention time. A resistance-based memory effect in which the FE polarization charge interacts with the space-charge region of a redox-controlled interface might lead to a novel memory device offering both high scalability and retention time. This concept has the potential to open a completely new road towards ultra-dense, energy efficient non-volatile memory devices as well as the fusion of memory and logic within new architectures.
Artificial multiferroics for new multifunctional electronics
Intrinsic multiferroic (MF) materials, which display both magnetic and electronic order, are of particular technological interest since they offer the prospect of multi-state logic and memory. However, the operation of today’s IMF materials was and still is limited to temperatures much below room temperature, largely due to the mutually-exclusive nature of the ferroic orderings. In contrast, IFOX has chosen artificial MF materials which hold much more promise. IFOX aimed to advance the state-of-the-art on several levels. It should develop the still-incomplete understanding of the interface coupling and improve its control. Moreover, IFOX should focus on the preparation of new magnetoelectric oxide film materials with advanced characteristics (e.g. enhanced Curie temperature) by means of interface and atomic sequence engineering within the SL growth approach. Multiferroic tunnel junctions were another promising approach, enabling control of spin-dependent carrier flow by electric field, and thereby the construction of four-state memories. Hence IFOX should finally target the integration of artificial MFs into completely new magneto-electronic multifunctional demonstrators, controlled by electric and/or magnetic fields and, ultimately, by light.
Oxide growth on silicon
Silicon is the most desirable platform for the integration of epitaxial multifunctional oxide devices with current semiconductor electronic devices. A capability to grow high quality layers on large area silicon substrates is therefore a major prerequisite for the integration of the novel oxide materials into mainstream electronics. Difficulties in coherent oxide growth are one of the biggest technological hurdles to oxide exploitation. IFOX was to extend the already existing research to relevant FEs, FMs and MF heterostructures. The ultimate goal was to achieve controlled epitaxial growth of the most promising oxides on 150 mm Si substrates and to define the necessary boundary conditions to access wafer sizes of 300 mm, which are currently used in production semiconductor facilities.
Furthermore the IFOX consortium was not only to use PLD for oxide growth but also pulsed plasma deposition or Aerosol deposition and sputtering. In the case that PLD would be unable to achieve the necessary buffer quality, both these processes can be used for alternative approaches. Especially Aerosol deposition with subsequent annealing may present a promising alternative because of its completely different technology and growth physics.
The above mentioned goals were defined at the beginning of the project. Because IFOX was planned for four years an assessment of the results and a revision of the workplan was foreseen after year three in which the goals should be defined in more detail, now focusing on well-defined heterostructures with novel functionalities belonging to the objectives and target research areas of IFOX.
This assessment was done and resulted in the following definition seven heterostructures:
1) Manganite/TiO based heterostructures which exhibit resistive switching suitable for novel storage and possibly logic applications
2) Heterostructures based on LaAlO3/SrTiO3 with additional ferromagnetic properties to combine transistor functionality with non-volatile storage for example for logic applications
3) Heterostructures based on BaTiO3 and NiFeO3 with inherent tunneling magnetoresistive functionality which may allow for MRAM cells which are switched by the magnetoelectric coupling between the BTO and the NFO.
4) Heterostructures of (La,Sr)MnO3 (LSMO) and ZnO which combine magnetoresistive sensing and gas sensing properties for automotive applications
5) Heterostructures using a bilayer of (La,Sr)MnO3 and BiFeO3 in which the magnetization of the LSMO can be switched via magnetoelectric coupling to the BFO which again can be switched by an electrical pulse. This structure would also address the area of magnetic memory with low switching power
6) Superlattices of LaMnO3 and SrMnO3 in which the magnetic properties compared to LSMO are much improved at room temperature and very high sensitivity in sensor applications (MR-sensors and Bolometers are to be expected).
7) Heterostructures using ferromagnetic (La,Sr)MnO3, a barrier of SrTiO3 and a ferromagnetic metal counter electrode which realise a combined element for resistive switching and magnetoresistance which can serve as a stateful logic device to combine logic and memory in a single cell.
Under the guidance of one lead partner per structure, each heterostructure was to be fabricated and investigated taking care of the following aspects:
- The structures should be optimized for maximum performance
- Growth on silicon substrates, if possible on large scale, was to be demonstrated and the results compared to those on matching oxide substrates
- Theory support was to be given wherever necessary to help design the structures and understand the results of characterization
- All results should be assessed with the help of the industrial partners in terms of maturity, technology readiness level of growth and processing, target application and competitiveness.
- For all heterostructures growth and industry compatible processing should be documented and made available as catalogues of
o Growth processes with detailed parameters
o Processing tools and recipes with detailed parameters
o Chemicals used and possible environmental impact.
An additional task of upscaling was added which was to be addressed by a third SME that joined the project at this stage with an expertise in industrial thin film deposition tools. This task should concentrate on the development of suitable buffer layers on large area silicon.
By this newly defined workplan the more general optimization and identification of promising heterostructures was guided towards concrete plans on heterostructures with defined functionality for which all technology should be ready for transfer to industrial application. The final results of IFOX can be summarized as (see for details the rest of the report):
1) All seven heterostructures were further developed including growth on silicon substrates and a thorough assessment was performed.
2) For two of the seven heterostructures (ZnO/LSMO-based gas sensor and PCMO/TiO based ReRAM) all important factors namely functionality at room temperature or above, fabrication on large area silicon wafers and industry compatible processing were demonstrated.
3) three catalogues were established, one on growth processes, one on processing and one on possible environmental impact of chemistry which allow for immediate assessment of needs and risks when transferring these heterostructures to industrial applications.
Project Results:
Description of the main S & T results/foreground from the IFOX Project
The S&T results of IFOX can be divided in several classes
- Growth of heterostructures by various technologies and development of growth technologies for oxides, including growth on large area silicon substrates
- Development of processing technologies for oxide heterostructures using industry compatible processes and equipment
- Characterization of the resulting structures and further development of characterization techniques for the requirements in IFOX
- Guidance and feedback by theory, including the development of theorical methods adequate for the goals of IFOX
- Industrial assessment of the results obtained
- Collection and cataloguing of growth and processing data for industrial use
- Oxide heterostructures with novel or improved functionalities
In the following an overview of these results is given, following the order of the workpackages in IFOX. Because these descriptions often refer to the seven heterostructures (HS) that the IFOX partners were targeting during the final period they are again listed here for further reference:
HS#1: Si/YSZ/SRO/PCMO/TiO: Resistive Random Access Memory (ReRAM) based on resistive switching.
HS#2: Si/STO/LaAlO3/SrTiOs: Field effect transistor based on ferromagnetic 2D electron gas in LaAlO3 (LAO) /SrTiO3 (STO) heterostructures.
HS#3: Si/STO/SRO/BTO/LSMO: Magneto-electric coupled memory device for room-temperature operation
HS#4: Si/YSZ/LSMO/ZnO: Multifunctional automotive sensor.
HS#5: Si/STO/BFO/LSMO: Multiferroic memory element.
HS#6: Si/STO/LMO/SMO: AMR field sensors and bolometers.
HS#7: Si/STO/LSMO/STO/FM1 (FM1=Co, Fe3O4): Resistive switching device (ReRAM) enhanced with magnetic functionality.
Main results of activities within IFOX
WP1: Theory and Modelling
The theory activity within IFOX aimed at the description of heterostructures both in terms of model parameters derived from the bulk as well as by calculating properties from setups modelling of realistic junctions and interfaces. The main tool used is the Density Functional Theory (DFT) that has transformed material physics and likewise computational chemistry, surface science and nanoscience. This framework, which provides the capability to describe the electronic structure, interatomic forces and in part also the electronic excitations of oxide materials containing hundreds of atoms in a computational volume, has developed into the standard method for realistic descriptions of materials from a quantum mechanical point of view.
At first, the calculations performed by the theory groups within IFOX concentrated on calculating the properties of bulk oxide materials. In particular a strong focus was on the determination of the complex bandstructure of several oxides (BaTiO3/PbTiO3/SrTiO3/KaNbO3/ZnO/CFO/NFO/PZT/...) which all could be used as possible materials in tunnelbarriers. These results provided the key input for the correct interpretation of tunnelling experiments and can additionally be used in the design process of heterostructures. In this first stage of the project the computational results were collected in the materials data-base developed within the consortium.
After these fundamental studies on bulk materials theory concentrated on all-oxide tunnel-junctions. In such setups we demonstrated the mechanism through which the TER (tunnelling electro-resistance) effect and the TMR (tunnelling magnetoresistance) effect can coexist in a tunnel junction. We presented the ab initio demonstrations of the mechanisms with our calculations of the SRO/STO/BTO/SRO junction. Using these results we proposed and demonstrated that the Fe/BTO/MgO/Fe junction could also function in this manner.
The study of interfaces between different oxides as well as the closely related field of domain walls in ferroelectrics was another field of active research. In particular, the influence of additional defects in such interfaces was studied. We found that the presence, concentration and nature of such defects play important roles in the stability of the interfaces and their electronic configuration.
The most intriguing effect of magneto-electric coupling at interfaces of oxides was investigated theoretically in different situations:
a) For the ferrites (CFO and NFO) we showed that it is the band alignment with the electrodes which determines the dominant spin character of the current, highlighting the importance not only of the tunnelling material but also the electronic structure of the electrodes and the interface. This understanding has led to the concept that by using a ferroelectric material the band alignment to CFO or NFO can be altered, this in turn changing the sign of the dominant tunnelling electrons resulting in a massive coupling between the magnetic spin of the tunnelling state and the electrical polarization of the ferroelectric.
b) Also the two dimensional electron gas (2DEG) formed at the interface between oxides of different polar character can be manipulated. In a BiAlO3/Sr2NiWO6 interface we showed that one can realize a magneto-electric coupling across the all-oxide interface, if we consider the effect of an external (exchange) magnetic field on the Rashba-type spin-orbit split 2DEG. Unfortunately growth efforts showed that this interface was not to be realized within the scope of IFOX.
In the final phase of the project, in which the work was concentrated on several well-defined heterostructures the theory activities also focused accordingly. For example, there were multiple simulations performed to clarify the nature and the properties of the 2D electron gas formed at the LAO/STO interface. A particular focus has been the magnetic properties of this interface and the question how the magnetism can be stabilized and manipulated. Our calculations indicate that magnetism can indeed be expected at this interface due to the localization of the electrons and the formation of Ti3+ ions, but that at the same time the ordering temperatures will be rather low. Our estimates give a critical temperature of about 35K for the ideal interface. This can be improved by doping with magnetic Mn ions and further enhanced by gating the 2D electron gas appropriately, but our results do not indicate that these effects can lead to effects observable at room temperatures.
Additionally, in this last part of the project, substantial work on the physics of oxygen defects in ZnO has been performed. Such vacancies are believed to be the key players in effects like gas sensing devices and resistive memories and the understanding of details of the formation and the diffusion of such vacancies thus is of significant practical importance. Our simulations were performed by first studying the different interfaces and surfaces and by investigating diffusion barrier heights with the nudged elastic band method. To a large extend these diffusion processes studied proved to be independent of external influences and hence the optimization of heterojunctions including ZnO are believed to by achievable by independent optimization of the constituents.
These results nicely showed that theory was able to supply new concepts to the experimentalists. Some of these were shown to be possibly realized, some were later shown to be inaccessible to current growth techniques. Those where realization seemed possible made their way into the finally chosen heterostructures. Also the simulations for example of the diffusion of vacancies are a good example how theory can and did help to interpret results of characterization.
WP2 Growth
The main objectives of this workpackage were associated to the development of state of the art depositions of functional complex oxides stacks for integration with existing CMOS technology (i.e. deposition on large area Silicon substrates). In details, the WP2 objectives were:
• To deliver high class transition oxide FM, FE and MF materials for all workpackages
• To establish a new state-of-the-art for the deposition of novel materials such as double perovskites and oxide super-lattices (SL)
• To assess the efficacy and industrial scalability of a range of oxide deposition processes for the first time
• To develop an all-oxide TMR stack
• To perfect the deposition of a range of oxides onto industry-standard Si wafers
• To progress in the areas of equipment and transfer of knowledge to the market
The above objectives have been achieved, with partial amendment for what concerns double perovskite growth, and the main results are described in the following.
2.1 Growth of high quality transition oxides
A large number of high quality thin films and heterostructures have been produced, and the feedback with other WPs (WP1 for theory, WP4 for characterization and WP3 for patterning) allowed the optimization process. The overall high quality growth obtained by IFOX enabled the functionalities claimed for heterostructures (HS) 1-7. In particular the growth of functional interfaces (between them STO/LAO for application in FET; integration of Ferrolectric oxides (BTO, BFO) with conducting perovskitic oxide (SRO and LSMO) for TER; LSMO/STO for MTJ, LSMO and ZnO for sensing) has been obtained through the control of structure and composition down to the interatomic scale as confirmed by high resolution TEM images provided within the experimental activity carried out in WP4.
2.2 New state-of-the-art for the deposition of novel materials
WP2 produced high quality functional superlattices with control down to the atomic scale. Some of the growth methods used in the project were designed as particularly useful for the purpose: a powerful example is the use of metal-organic aerosol deposition (MAD) for LMO/SMO superlattices with novel ferromagnetic high temperature phase (see for more details HS6).
2.3 Efficacy and industrial scalability of a range of oxide deposition processes
WP2 activities allowed to set up key-parameters for deposition processes depending on the specific technique (sputtering, PLD, PPD, CSA, MAD, ...). In collaboration with WP6, the scalability of the lab process to industrial one has been evaluated to define the main constraints (see the process database available on the web). A more industrial friendly technique (sputtering) has been successfully used for a full oxide HS growth (HS3) and high quality interfaces have been obtained.
2.4 All-oxide TMR stack
A full oxide TMR device has been obtained by TCD. The growth parameters of device constituents were first optimized on individually grown layers, followed by the incorporation of these layers into a full stack. The SRO/NFO/BTO/LSMO structure has been demonstrated to act as a TMR element.
2.5 Deposition of a range of oxides onto industry-standard Si wafers
A consistent effort of the IFOX project has been devoted to the development of appropriate buffer layers for the growth of complex functional oxide on Si substrates. A number of buffer layers (STO, YSZ/CeO2) have been obtained by varying the deposition conditions and the selection of the optimal process sequence has been set, allowing the deposition of epitaxial layer and heterostructures on Silicon substrates. Compatibility up to 4” wafers has been attained by PLD and PPD techniques, allowing to reach the project goal of industrial scalability of HSs.
2.6 Equipment and transfer of knowledge to the market
The active involvement of private companies (OS, CRF, TSST and PST) in the development of functional HSs has boosted the transfer of developed technologies to possible market applications. The optimization of large area deposition of oxides on Si opens then the possibility of implementing standard CMOS technology with new and functional materials. OS as well as TSST have developed new prototypes for large area deposition during the course of IFOX.
In relation to the functional Heterostructures (HSs 1-7) identified in the second part of the project, we summarize in the following table the main results achieved.
The integration onto Silicon substrates have been obtained for all the HSs, with the exception of HS3, indicating the successful approach of the activity. It should be noted that the extension of growth of HSs to large area substrates is mainly related to the processing steps: this is the main reason for having two HSs growth processes tested on 4’’ substrates.
HS
Growth on single crystal substrates Integration with Si Compatibility with Large Area Comments
HS1 X X X Leakage depends on crystallinity of buffer layer on Si
HS2 X X Different parameters for Si and single crystal substrates
HS3 X - Criticalities on the deposition systems
HS4 - X X Growth parameters allow independent functionalities
HS5 X X Working temperature below RT
HS6 X X Growth on Si attained but some criticalities on functionality
HS7 X X Role of STO crystallinity in functionality
WP3 Processing
In order to functionalise the heterostructures created for the IFOX project, various processing steps are typically required. Such steps often require a variety of technologies, from the methods used to clean substrates before deposition to device fabrication using UV and electron-beam lithography to pattern devices of micron and sub-micron dimensions, combined with various etching techniques to create the final devices. Multiple steps are usually required in order to create sensors which can be measured and integrated into a testing rig. Such processing often creates difficulties due to tight tolerances of the materials and equipment used when working on such small dimensions. Each process must also be fully compatible with the oxide materials used in the heterostructures.
During the course of the project, various factors were considered for processing of device structures, namely compatibility of technology, ease of integration into existing manufacturing, as well as the environmental impact of each step. The scale of the project meant that many partners had to work on various aspects of the processing in parallel. Partners involved in processing are Halle (MLU), Gottingen (OGOE), Twente (TSST), Glasgow (UGla), Dublin (TCD), Mainz (JGUM), CRFiat (Turin), PST (Beddizzole) and PSI (Villigen).
Work carried out during the project involved the deposition of thin films of oxide materials by a number of different physical and chemical vapour deposition techniques. This involves the use of various materials as substrates. Before deposition can occur, the proper steps must be taken to ensure that the surface is clean and properly terminated for the desired purpose. To this end, standard methods have been developed or modified to prepare Si, SrTiO3 and GGG substrates.
Cleaning of Si and GGG substrates was done by immersion in acetone and sonication to remove any organic surface contaminants, followed by rinsing in isopropanol. GGG was annealed in a furnace at 950°C for 30 minutes. Si was treated with RCA solution, followed by buffered HF to remove any surface oxide. SrTiO3 was heated in a furnace for 30 minutes at 950°C followed by HF treatment to produce a TiO2 terminated surface.
Following deposition of the heterostructure, the fabrication of devices is completed in a series of etching and deposition steps, depending on the device structure. Of the seven device heterostructures investigated during the project, three used a hall bar structure, two used small diameter metallic pillars as contact pads, one used large area metallic pads for contact, and one used a current-perpendicular-to-plane tunnel junction design.
In the first instance, hall bars were constructed by two different methods – by lift-off or by etching. HS2 used the lift-off method to form an open window to the substrate surrounded by an insulating layer of amorphous LAO. This window was defined in PMMA photoresist using e-beam lithography, with critical dimensions as low as 100nm. By depositing amorphous LAO, and removing the PMMA, the defined shape is left open for subsequent steps. Crystalline LAO was then deposited into the window to form a 2DEG, followed by deposition of a metal gating layer.
This method was employed due to reduced likelihood of issues arising from etching of the STO substrate, known to cause conduction in the surface, as well as improved performance of the 2DEG interface between the STO and LAO.
The hall bars in HS5 and HS6 were constructed by defining a hall pattern in photoresist on the top surface of the finished heterostructure stack, and etching using Ar+ ion milling or inductively-couple-plasma reactive ion etching (ICP-RIE). In this case, the etching rate of each material in the stack must be well calibrated, or an endpoint detection system must be in place, in order to ensure that etching beyond the desired layer does not occur. HS5 requires one lithographic step, which incorporates contact pads into the shape of the hall bar. HS6 requires a two step process, by etching a segment of the top surface followed by deposition and lift-off of Cr/Au contact pads. Wet etching of the manganite surface of HS6 was also tested with KI buffered HCl solution. It was found that, while effective, the reproducibility and quality was much more varied as compared to RIE, and so it is considered to be less desireable.
Both HS1 and HS7 use small diameter pillars for testing. The heterostructures include a metal as the top layer to ensure good contact. In both cases, pillars are patterned in photoresist and the surrounding material is etched down to the oxide layers. Thus, the processing of HS7 is completed and testing is carried out via probe station contacting of pillars. HS1, on the other hand, goes on to deposit a hard insulating AlOx layer, followed by deposition and lift-off to define larger contact pads.
It is also possible to pattern devices without using photoresist, as in the case of HS4. Shadow masks in close contact to the substrate have been used during deposition to produce the desired structure. This process is significantly cheaper and more straightforward than the use of photoresist, however it is also suitable only for larger structures that do not require precise placement of features.
HS3 proved difficult to process due to the complexity of its structure, as well as the incompatibility of the substrate and etching process. The steps required to form a tunnel junction device involve the patterning of a bottom electrode, followed by small mesa structures and finally top contact pads, while ensuring that there is no shorting between top and bottom electrodes. Normally Ar+ ion milling would be used to etch the material, however it has been seen that this type of etching method causes oxygen vacancies in SrTiO3, the substrate material, leading to conductive behaviour. This can cause issues with shunting of current and significantly increases noise during measurements. As such, ICP-RIE was employed as an alternative method. RIE uses chemical and physical means to achieve etching of materials, and so has a selectivity towards certain materials, allowing it to be used in situations where purely physical etching methods are not possible. BCl3 was used as the etchant in this case, and proved highly suited to etching of complex oxides. It provided demonstrably improved etching of SrTiO3 substrates without the issue of conductive pathways, and is thus considered a vital technique developed during the project for processing in future works.
In the course of the project we investigated the quality and magnetic properties of SrTiO3 substrates of different cuts from various suppliers. Only (100) substrates from just two of the suppliers were free of ferromagnetic Fe-Ni inclusions in the surface. These inclusions, present at the 1 ppm level, give spurious hysteretic ferromagnetic signals. A new type of non-hysteretic surface magnetism of up to 10 Bohr magnetons per square nanometre, associated with oxygen vacancies, was discovered in vacuum annealed crystals.
One of the major aims of the project was integration of complex oxide heterostructures onto Si substrates to establish their feasibility in future large scale manufacturing alongside other CMOS features and devices. In order to do this, a method was developed to create STO seed layers on buffered Si substrates. Using a combination of CeO2 and yttrium-stablised zirconia as buffer layers, highly textured complex oxides were able to grow. Experiments were then carried out to elucidate how these seed layers would react to the various etching techniques employed throughout the project. It was found that both Ar+ ion milling and ICP-RIE were suitable techniques, however wet chemical etching with buffered HF showed significant damage to the layer edges after etching.
A number of etching techniques have been identified during the project, which are easily integrated into modern silicon and wafer manufacturing setups. The materials used in the project need special consideration when considering the etching method, and ICP-RIE using BCl3 has been identified as a flexible and powerful tool for processing of complex oxides and sensitive substrates.
WP4 – characterization
The successful fabrication and application of novel transition metal oxide heterostructures depends on atomic scale control of the structure and composition of the produced materials. The goal of WP4 was to analyze the structure, chemical composition and electronic, magnetic, ferroelectric, multiferroic, opto-electronic and opto-magnetic properties of the as-grown and processed materials. During the course of the IFOX project close to 100 different material systems have been characterized with a variety of techniques, and the knowledge for an extensive materials database was created. The extensive knowledge about the characteristics of the heterostructures served as the basis for selecting the seven most promising functional oxide heterostructures to work on in year 4. Measurements directly related with the sample growth, the interfacial characterization and measurements related with the functionality of the heterostructure have been performed.
Due to the scientific complexity and the challenging measurement conditions of oxide heterostructures several instrumentation developments were needed. As a result, major achievements were obtained in the structural characterization of interfaces. For example, high-resolution transmission microscopy combined with electron energy-loss spectroscopy enabled the study of strain and oxygen vacancies effects at interfaces. The local structure with atomic resolution was studied by means of high-angle annular dark-field (HAADF) scanning transmission electron microscopy (STEM), energy-loss-spectroscopy (EELS) and energy-dispersive X-ray spectroscopy (EDX), performed using a FEI Titan microscope, equipped with aberration correctors for both image and probe forming lenses. This was used for a variety of scientific questions. For example transition metal L2;3 electron energy-loss spectra for a wide range of V-, Mn- and Fe- based oxides were recorded and carefully analyzed for their correlation with the formal oxidation states of the transition metal ions. Also the structural phase transition and spontaneous interface reconstruction in superlattices was studied on an atomic level. The combination of these atomic level characterization techniques with advanced sample growth enabled to create enhanced local magnetization by interface engineering in perovskite-type correlated oxide heterostructures, to discover the extreme mobility enhancement of two dimensional electron gases at oxide interfaces by charge-transfer-induced modulation doping, or control the lateral anisotropy in correlated manganite heterostructures by oxygen octahedral coupling. These techniques also enabled to study the oxygen vacancy transport mechanism and to make detailed comparison with predicted interface structures.
Impressive developments have been achieved in the nanoscale imaging of the magnetization states of oxide materials. Using scanning transmission electron microscopy (STEM) and Lorentz DPC microscopy it was possible to correlate the structural and functional properties of ferromagnetic oxide thin films. The novel capability of collecting magnetic, structural, and spectroscopic information with nanometre spatial resolution in a single experiment was demonstrated, complementing recent advances in analytical electron microscopy and spectroscopy with aberration corrected Lorentz DPC imaging. Variations in Curie temperatures were mapped and the role of defects and damage in modulating local magnetic properties on the nanoscale were assessed, including the pinning of magnetic domain walls.
Another important aspect was the development of the in-situ control of parameters such as electric fields for different types of imaging experiments. This enabled for example investigating the role of the electroforming process in the establishment of resistive switching behaviour for layered heterostructures acting as nonvolatile Resistance Random Access Memories (RRAMs). Another example is the study of interfacial magnetoelectric coupling in artificial multiferroics, which are meta-materials that show the existence of a magneto-electric coupling, obtained by the interfacial coupling of a ferroelectric and a ferromagnetic constituent. This enabled to investigate the ultimate mechanisms responsible for artificial magnetoelectric coupling in different interfaces and possible routes to optimize the coupling and extend it to the nanoscale. A 90 degree electric field-induced uniform magnetization rotation in single domain 100*200 nanometer ferromagnetic islands grown on a ferroelectric single crystal was demonstrated using x-ray photoemission electron microscopy.
The experimental findings are well correlated with micromagnetic simulations, showing that the reorientation occurs by the strain-induced magnetoelectric interaction between the ferromagnetic nanostructures and the ferroelectric crystal. Specifically, the ferroelectric domain structure plays a key role in determining the response of the structure to the applied electric field, resulting in three strain-induced regimes of magnetization behavior for the single domain islands. Furthermore, it was demonstrated that such correlations can be extended to oxide heterostructures deposited on a silicon substrate, making it particularly interesting for applications. This paves the way towards magnetoelectric-MRAM devices containing an artificial multiferroic film stack with low power consumption and high switching reliability. Understanding the dynamics of the interfacial coupling in heterostructures is important to optimize the samples for best performance and the imaging of magnetization dynamics in oxide systems with sub-ns time resolution has been developed in the IFOX network using X-ray microscopy and magneto-optical imaging techniques. For example optical control of a magnetic tunnel junction has been tested with a specially designed set-up that allows optical access and pump-probe experiments on a magnetic tunnel junction, while applying bias voltages.
Another example is the imaging of the propagation dynamics of spinwaves (SWs) excited by a femtosecond pump laser studied with a new developed Time-Resolved Magneto-Optical (TRMO) imaging system.
A large fraction of the work in WP4 in the last year was focused towards characterizing and testing the functional properties of the seven selected heterostructures for a range of applications going from sensing of gases, or as bolometers to novel ways of data storage. For this, the IFOX network could make full use of the variety of standard characterizations techniques available and the new advanced techniques developed during the project, which were pushing the limits. The IFOX partners have made significant, internationally recognized contributions to the advancement of state of the art characterization methods of oxide materials, which enabled to generate systematic data sets. The large size of the WP allowed characterizing the samples with many techniques providing deep insights into the materials and in particular into their interfaces. The information about the sample quality was reported back to WP1 (theory) which enabled modelling inspired growth and improved understanding of interfacial effects and WP2 (growth) and WP3 (processing) which helped to improve the sample quality. The functionality of the materials originating from the interfacial coupling has been investigated and was provided as input for WP5 (Industrial supervision) and WP6 (Exploration of new functionalities).
WP 5 Industrial control and supervision concerned the industrial aspects involved in the production of finalized heterostuctures selected within the project and the corresponding achieved functionalities for the appropriate applications. Also collection and management of related information were carried out. Main activities can be summarized in the following categories:
• Procedure and standardization, related to: deposition or growth techniques, fabrication processes, interface functionality and potential hazard involved.
• Prototyping, related to machinery and large area issues.
• Techical Data Collection and Data Bases Management, related to: chemicals, materials, growth and fabrication processes.
Procedure and standardization
Concerning Growth processes procedures, the involved partners have described in details the adopted techniques in order to make comparison between them related to working conditions parameters, process timing and potential hazard involved. The specific goals of depositions on wide area substrates, Silicon platform and Growth technique efficiency are clearly indicated. Sputtering, which is already an industrial technique, has been used to prove heterostructures viability and reached required functionality. Pulsed technologies such as PLD and PPD have shown a high grade of versatility and more over offer the advantage of a common deposition platform. Those last two techniques seem to offer special features that enable industrial production: the most important one is the capability in transferring complex oxides stoichiometry and crystalline structure, from bulk material to thin film form.
All the adapted techniques have assured a good interface quality with a high control of space localization of chemical compounds which have guaranteed final functionality. Another important achievements obtained by IFOX partners was that depositions were performed on Silicon substrates enabling in this way integration in Semiconductor Industry.
Regarding adopted fabrication processes most of them are already developed at industrial level (Etching, Photolithograpy, E-beam Lithography-EBL). For this reason their compatibility with previous thin film formation processes was verified in order to guarantee the final functionality.
For H#1 fabrication process has been demonstrated for samples of any area dimension, demonstrating a full technological compatibility with previous processes.
For H#2 different patterning techniques were developed and tests were performed on obtained heterostructures, demonstrating that the interface quality is not affected at all. Also a new dry etching technique was applied which allows the patterning of the material on lateral sizes while maintaining the metallic behavior. This patterning technique is compatible to standard industrial etching. Also for the device containing
H#3 a lithographic process was applied, patterning it for vertical transport measurement. It has been demonstrated that the metal contacts can be achieved without damaging the interface properties.
Substrates of H#4 were specifically prepared to host gas sensing material and were patterned by chemical etching and UV photolithography.
Hs#5 were patterned by electron beam lithography followed by ion milling down to 250nm scale. This final device processing does not have particular restrictions and, as it was verified, does not alter the layers properties and interface’ ones as well.
The device containing H#6 was fabbricated trough E-beam nanopatterning followed by dry etching for different oxide films including magnetite. Also a direct patterning processes by using lift off techniques with sacrificial layers were applied. No particular boundary conditions arise from such processing techniques.
For H#7 patterning and nano-patterning processing were successfully applied.
Concerning interface functionality of H#1, the required characteristic of “ON-OFF” switching was very near to the goal over the entire device area of this specific application.
The H#2 required functionality was obtained through the control of the two dimensions electron gas that can be used to form a field effect transistor.
Concerning H#3, the coexistence of electronic polarisation and magnetisation in adjacent thin films has been shown during the IFOX project. The coexistence of TER and TMR has not been proven due to the absence of TMR in the measured devices.
H#4 has shown both: magneto-resistance effect when subjected to a variable magnetic field and change in the conductivity when exposed to proper gas mixtures. This combination generates a multifunctional device for which the magneto-resistive behavior is a function of the gas adsorption.
For the H#5 device concept two different spin configurations were coupled due to exchange bias at the interface. Therefore, by applying an electric field it is possible to alter the magnetization in one layer and vice versa, obtaining the required final functionality for low energy consumption memory devices.
Concerning H#6, interface functionality was based on the enhanced magnetism and anisotropic magneto-resistance in a Digitalsuperlattice (DSL) composed of different oxides layers. For this heterostructure, application as a bolometer and AMR at room temperature must require additional optimization steps.
Related to H#7, the final device functionality, switching between two or more distinctive resistive states, has been demonstrated for growth on a single crystal. TMR behavior is still too low and at low temperature.
Concerning potential hazard involved in materials handling and processes set-up, all partners contributed in collecting info on chemicals used in their premises and in the cleaning procedure standards for industrial environment. The research done had led to the identifications of a number of dangerous substances used within IFOX, which probably, in the future, will be banned. In particular two of them have to be replaced with lower risks ones and, because they involve rare earth elements, maybe could give rise to problems concerning their availability. Related to deposition processes, most of them are performed in high vacuum conditions preventing the dispersions of nanoparticles.
Prototyping
WP5 partners have faced the up-scaling process for large area deposition, developing industrial prototype and production procedures. Another important achievement obtained by IFOX partners was that depositions were performed on Silicon substrates enabling in this way integration in the Semiconductor Industry. Related to deposition techniques up-scaling process, magnetron sputtered (MS) buffer layers deposited on 4“ silicon with suitable homogeneity and crystal quality under standard industrial conditions has been reviewed and its procedure has been defined.
Once moved away from early stage DC to RF driven sputter deposition, process initial parameters and hardware were refined to improve homogeneity and, at the same time, feasibility tests, quality control and production of HSs on 4” full scale wafer were carried out. With such technique, in order to further improve crystal quality to a level similar to those achievable by other more sophisticated techniques (such as PLD and PPD), a large optimization work has been attempted on the pre-processing cleaning conditions and on the early stage temperature of the deposition processes.
Since magnetron sputtering is not the most suitable technique to achieve this kind of growth, the lower level of crystallinity respect to more sophisticated techniques is not a surprisingly negative result.
PLD has successfully applied to obtain very complex oxides heterostructures. Within IFOX projects PLD main limits have been partially overcome: although its maximum repetition rate of 20Hz is still too low, depositions were performed on 4” wide substrates overcoming the small area limitation. This was one of the most important goals for the IFOX project, a common platform technique which enables depositions on wide areas.
Also with PPD systems depositions can be performed on large area substrates, furthermore an industrial prototype system has been set-up which is already working in a continuous way on (infinite) strips of 360 mm width. Related to the desired thin film properties, it is possible to adjust different PPD parameters and select, in a short time and with few attempts (about 5 samples), the optimal values to perform growth processes. In order to improve deposition quality, in particular the distance between target material and the substrate was modified, reaching, a result of 6% of homogeneity at 3 σ. An “in-situ” RF Cleaner system was realized to improve oxides layers adhesion. For adopted fabrication processes most of them are already developed at industrial level (Etching, Photolithograpy, EBL) and there’s no particular need in developing prototyping models.
Techical Data Collection and Data Bases Management
WP5 work was also related to collect information from all partners on all chemicals, materials, growth processes, fabrication procedures and interfaces validation for its technological success. Those fundamental data were elaborated and organized in on-line data bases used during the project development. Technological databases are a powerful instrument to sustain the driving of the innovative aspects of reached technological results close to industrial organizations which can cover the final steps to the market.
Main databases content are related, but not limited, to the seven (or more) “concept devices”, based on seven different oxide heterostructures, which applications are clearly indicated and, in most cases, their functionality completely described and obtained. For such cases where the results do not match the expected behaviour, they however represent a resource of information for not to “waste time and money”: this is very useful from an industrial point of view. Data were also used for the transfer on Silicon platform of the oxide heterostructures, which was successful in most cases, with the help of a buffer layer, allowing the integration with the semiconductor industry. WP5 Partners also elaborated an evaluation of TRL (Technology Readiness Levels) in order to give an idea of the global results reached in a common used language.
WP6 Exploration of new functionalities
WP6 was dedicated to the exploitation of the results of IFOX, and this exploitation happened in three different ways:
1) Assessment of results obtained in WP1 through WP4 with focus on the identification of promising materials as well as processing technologies.
2) Identification of new functionalities and possible emerging applications
3) Deliver the necessary boundary conditions for industrial application of oxides.
Furthermore, within WP6 the goals and deliverables of IFOX have been checked throughout the project to guide necessary revisions of the work and tasks. WP6 created a platform for the use of new functionalities for new device concepts, with the output of other work packages as input.
Most important deliverables (next to intermediate deliverables) are:
D6.1.1 Report on promising materials for use inside IFOX
D6.2.2 Final catalogue of processing technologies necessary for the industrial application of oxides
D6.4.1 Definition of new concepts for novel functionalities as deliverables for month 48
D6.5.2 Report on assessment of SRO/PCMO/TiO based ReRAM devices
D6.6.2 Report on assessment of LAO/STO based field effect transistors
D6.7.2 Report on assessment Si/STO/SRO/NFO/BTO/LSMO based device structure
D6.8.2 Report on assessment of gas sensing reactivity of devices based on Si/YSZ/LSMO/ZnO and Si/YSZ/FO/ZnO
D6.9.2 Report on assessment Si/STO/BFO/LSMO based memory device
D6.10.2 Report on assessment of Si/STO/LMO/SMO based demonstrators for AMR and bolometers
D6.11.2 Report on the assessment of Si/STO/LSMO/STO/FM1 (FM1=Co, Fe3O4) based ReRAM with magnetic functionality
Within the first years of IFOX the project has focused on the deliverables on most promising materials and the definition of new concepts for novel functionalities as deliverables. An important result was the definition of the most promising heterostructures, which were the basis for the device concepts for novel functionalities. The report on promising materials in IFOX lists all oxides that are grown by the partners and the reasons why they are deemed promising and for what reason. It also gives a short overview of the properties of the different materials. The materials database contains many different compounds out of which those have been selected which were investigated in the heterostructures.
In the final year within IFOX 7 heterostructures have been explored, and these are:
HS#1: Si/YSZ/SRO/PCMO/TiO: Resistive Random Access Memory (ReRAM) based on resistive switching.
HS#2: Si/STO/LaAlO3/SrTiOs: Field effect transistor based on ferromagnetic 2D electron gas in LaAlO3 (LAO) /SrTiO3 (STO) heterostructures.
HS#3: Si/STO/SRO/BTO/LSMO: Magneto-electric coupled memory device for room-temperature operation
HS#4: Si/YSZ/LSMO/ZnO: Multifunctional automotive sensor.
HS#5: Si/STO/BFO/LSMO: Multiferroic memory element.
HS#6: Si/STO/LMO/SMO: AMR field sensors and bolometers.
HS#7: Si/STO/LSMO/STO/FM1 (FM1=Co, Fe3O4): Resistive switching device (ReRAM) enhanced with magnetic functionality.
Achieved results
HS#1: Functional epitaxial heterostructures (down to 200 nm) have been fabricated for the first time using heterostructures grown on 4” Si. The IFOX partners have actively worked together and this resulted in functional resistive switching devices within epitaxial heterostructures grown on 4” Si wafers. The large area epitaxial integration on Si has not been shown before. It was found that the interface between the TiOx and the PCMO is important in the switching characteristics. Most important, epitaxial integration of the devices in 4” Si wafers has shown that high crystalline quality (in terms of epitaxy and crystallographic purity) is very required. PLD has shown to be capable of achieving such quality. Furthermore, sub-micron devices down to 200 nm have been processed from 4” heterostructures, showing resistive switching.
HS#2: IFOX showed that integration of LAO/STO with Si is possible, and STO films of high quality can be epitaxially grown on Si wafer by molecular beam epitaxy. LAO films were then grown on STO/Si by Pulsed laser deposition (PLD). A two dimensional electron gas was revealed at both amorphous LAO and crystalline LAO and STO interfaces with suitable mobility at room temperature. Furthermore, nano-patterning down to 100 nm of LAO/STO interface by E-beam lithography has been achieved, specifically to test short scale lateral electrical insulation.
HS#3: Tunnel junction devices incorporating a synthetic multiferroic tunnel barrier have been fabricated and tested. This is the first known attempt to create a voltage controlled TMR/TER device. IFOX has identified an all-transition-metal-oxide structure that shows promise for future storage device applications. Growth and fabrication of highly crystalline and epitaxial device stacks has been demonstrated.
HS#4: LSMO/ZnO heterostructures exhibit both magneto-resistance, when subjected to a variable magnetic field, and a sensible change of conductivity when exposed to proper gas mixtures. The two effects were expected to be interwoven, but from modelling and physical experiments it can be stated that the interplay is limited and not affecting significantly the HS4 functional behaviour, which is dominated by the intrinsic properties of the materials employed, in such a way that the interfacial properties are reflecting only slight changes at the atomic level due to gas absorption. The present achievement is totally new and represents a step forward of the state of the art.
HS#5: Devices prepared on both STO and Si substrates have been fabricated and characterized. IFOX showed that the coupling between the ferroelectric BFO layer and the ferromagnetic LSMO layer occurs preferentially for rough interfaces. This indicates that device fabrication can be implemented even for the crystallinity that can be achieved on Si, which is necessary for realistic device applications.
HS#6: Applications as AMR sensor or bolometer have been completed with success. The interfacial nature of the high temperature (TC ≈ 350 K) ferromagnetic phase (HTP) confines the ferromagnetic (FM) state to two dimensions. Since the two-dimensional FM phase is expected to strongly enhance spin-orbit coupling via the reduced symmetry of the MnO6 octahedra, the anisotropic magnetoresistance (AMR) within the interface is highly increased. The AMR- effect directly links the sample’s resistance to changes in the magnetization direction with respect to the probe current. External magnetic fields affect the magnetization direction and the FM domain landscape and hence can be sensed by a simple resistance 4-point-measurement method. Within IFOX, two different approaches to make functional use of the material inherent AMR were tested.
HS#7: The heterostructures showed for the first time the intercorrelation between magnetoresistance and bistability. The magnetic functionality added to a resistive switching system enables the latter with an additional degree of freedom and opens a series of various advantages considering device applications. Firstly, standard resistive switching elements, working in either memristor or different regime, represent a two terminal device with all its advantages and disadvantages. The magnetic functionality, that is, the possibility to modulate with magnetic field the resistive states established by non-volatile switching mechanisms, can be considered as a third terminal (external magnetic field) promoting the utilization of such devices in more complex electronic circuits and systems.
Conclusion:
The results described above show that IFOX has achieved a lot of results not only on the academic side but also on the side of industrial application or in paving the way there to. For two of the seven heterostructures (ZnO/LSMO-based gas sensor and PCMO/TiO based ReRAM) all important factors namely functionality at room temperature or above, fabrication on large area silicon wafers and industry compatible processing were demonstrated
As a further important result, all data on growth and processing were collected in several web-based interactive databases and evaluated by the industrial partners. From these data three catalogues were established:
- A materials database listing possible candidate oxide materials including relevant physical properties
- A chemicals database where all chemicals are listed which are used in one of the growth or processing steps investigated
- A process database where for each process an infinite number of process steps can be entered, and each process step includes the equipment used, the environmental conditions, materials involved, process times and other details if necessary. The tree format allows for an easy overview, however, with fast access to the details and allows for immediate assessment of needs and risks when transferring these heterostrctures to industrial application.
In addition, a questionnaire to evaluate necessary costs and equipment for oxide electronics was created.
Furthermore, the relevant results have been published in peer reviewed international journals (91 manuscripts) and disseminated to the major inter- national conferences (almost the same number of talks). During the 2015 E-MRS Fall Meeting, held in Warsaw University of Technology, from September 15 to 18 2015, the IFOX consortium presented a session within symposium L "Materials for electronics and optoelectronic applications away from silicon”. IFOX also setup a booth at this E-MRS meeting with posters on all heterostructures in which discussions with possibly interested industry partners were held. In addition, a massive search in patent databases was done in order to evaluate possible restrictions and possible areas for protection of IP and to identify competing technologies.
January 2016
Potential Impact:
Potential impacts
Socio-economic impact
At present oxides are important for semiconductor technology mainly in the context of improving the dielectric/insulating and interface properties of gate oxides for use in MOSFET devices. However, their real promise for novel future applications is that they can be engineered with additional functionality, e.g. ferroelectric, multiferroic, and magnetic properties, by tailoring their chemical composition and making multilayer stacks of a range of different oxides.
The aim of IFOX was to explore, create and control novel electronic and magnetic functionalities that result from the rich interplay of charge, spin and orbital degrees of freedom in transition metal oxide heterostructures and their interfaces. The outcome aimed at the establishment of a material platform for novel ‘More than Moore’ and ‘beyond CMOS’ electronics, compatible with VLSI and which delivers performance and functionality far beyond the state-of-the-art.
To achieve these goals we carried out a research programme with workpackages on theory and modelling, epitaxial growth, processing, characterization, device fabrication and knowledge transfer. Crucially, technological relevance and impact was guaranteed by the active participation of a world-leading company, Fiat and three SMEs, namely TSST and Organic Spintronics, with expertise and infrastructure for epitaxial oxide growth on Si and Protec Surface Technology, with expertise on engineering and production of PVD and PECVD machinery for thin films and coatings deposition and customised deposition process developments. Focusing the research on industry needs ensured that IFOX worked on relevant problems with tangible results. Success for the project means delivering useable new technology and improved understanding to industry beneficiaries which will allow them to make significant progress on the roadmap for future generations of semiconductor devices.
While the first three periods of IFOX were dedicated to the growth, processing, and investigation of various oxide thin films and their heterostructures, the final period focused on seven different heterostructures, chosen for their largest potential for possible future applications in the target research areas:
1) Manganite/TiO based heterostructures which exhibit resistive switching suitable for novel storage and possibly logic applications
2) Heterostructures based on LaAlO3/SrTiO3 with additional ferromagnetic properties to combine transistor functionality with non-volatile storage for example for logic applications
3) Heterostructures based on BaTiO3 and NiFeO3 with inherent tunneling magnetoresistive functionality which may allow for MRAM cells which are switched by the magnetoelectric coupling between the BTO and the NFO.
4) Heterostructures of (La,Sr)MnO3 (LSMO) and ZnO which combine magnetoresistive sensing and gas sensing properties for automotive applications
5) Heterostructures using a bilayer of (La,Sr)MnO3 and BiFeO3 in which the magnetization of the LSMO can be switched via magnetoelectric coupling to the BFO which again can be switched by an electrical pulse. This structure would also address the area of magnetic memory with low switching power
6) Superlattices of LaMnO3 and SrMnO3 in which the magnetic properties compared to LSMO are much improved at room temperature and very high sensitivity in sensor applications (MR-sensors and Bolometers are to be expected.
7) Heterostructures using ferromagnetic (La,Sr)MnO3, a barrier of SrTiO3 and a ferromagnetic metal counter electrode which realise a combined element for resistive switching and magnetoresistance which can serve as a stateful logic device to combine logic and memory in a single cell.
In addition the project foresaw an effort to attempt and assess the realization of all these heterostructures on large area silicon wafers dramatically reducing possible production costs compared to heterostructures deposited on single crystalline oxide substrates.
All these seven heterostructures were further developed including growth on silicon substrates and a thorough assessment was performed. For two of the seven heterostructures (ZnO/LSMO-based gas sensor and PCMO/TiO based ReRAM) all important factors namely functionality at room temperature or above, fabrication on large area silicon wafers and industry compatible processing were demonstrated. As follow up, CRF will continue the study to increase the TRL of the HS4 up to 6. Besides, to go deep inside into the measurement repeatibility of the gas sensitivity, and the life cycle costing of the production will be part of the future investigation to evaluate the value for money in the replacement of the standard gas sensor today employed. The evaluation will take into account the improved absolute sensitivity of HS4 demonstrated in IFOX. The exploitation of the coupled magneto resistance funtion of HS4 will be requiring further developments and analysis to gauge at what extent the airgap variation sensitivity of the MR response will affect the precision of the absolute position of a mangnetized valve. It is expected to go deeper inside these issue by a second ICT project employing novel advance designed smart systems.
To facilitate the transfer of results to industry all partners contributed to a huge effort, which was guided by the industry partners, to collect the knowledge on growth and processing in a suitable format. This resulted in three major deliverables
1) A catalogue of growth techniques and growth parameters for all heterostructures
2) A catalogue of processing techniques and parameters for all heterostructures
3) A catalogue of all chemicals involved in growth and processing, assessing the environmental impact of the chemical compounds
In addition a tentative assessment of growth and processing in terms of technology readiness level (TRL) is provided, resulting in typical values ranging from 3 to 5 for all steps (see WP5 for details).
Exploration of new functionalities
For all heterostructures an overall assessment is provided which not only addresses the results of characterization. It also provides a placement of the heterostructure within the targeted research area and an evaluation of the maturity of the necessary technology and a competitive comparison to possible alternative devices on the market or in the development chain.
The data of the three catalogues allow for immediate assessment of needs and risks when transferring these heterostrctures to industrial application.
After completion of the programme, we can state that IFOX has delivered results that are highly relevant for our industrial partners in terms of novel sensors, growth and processing and integration to Si.
Societal implications
The importance of the ICT sector from an economic, societal, and (increasingly) environmental viewpoint is obvious: Europe needs a strategic and innovative semiconductor foundation underlying the ICT portfolio to bolster industry, increase innovation and take a lead in economic, societal and environmental terms. The semiconductor market is currently worth 200 billion Euro (256 billion USD) globally on an annual basis. There is a multiplier effect of approximately 25 times that amount in terms of equipment and services ranging from television sets to IT. In other words, semiconductor technology provides the knowledge and technologies that generate some 10% of global GDP. The European semiconductor market attained a value of approximately Euro 30 billion in 2007, representing just 16% of the worldwide total .
IFOX has improved the competitiveness of the EU semiconductor industry by its development of novel device technologies based on oxide materials with novel device functionalities beyond the current state-of-the-art. This will support the need for industry linked technological innovation in the EU as described in the 2008 Europe Innova Innovation watch report: "A prospective innovation challenge for the ICT sector is the insufficient collaboration between public sector and business sector researchers".
The results obtained in IFOX is multi-disciplinary and beyond what could have been accomplished by one or two individual research groups as they do not have the expertise nor infrastructure for all the various aspects of theory, growth, structuring and characterization necessary for the achieved functional structures. In addition, without the critical evaluation and feedback of the industrial partners, the focusing on the most relevant and prospective structures and their compatibility with Si technology would not have been accomplished.
Regarding impact of research, the EU still ranks as the world’s largest producer of scientific knowledge (measured by publications), but contributes less than the US to high impact publications . These data provide the backdrop against which the consortium has tackled its research goals. The best research conducted within IFOX has been published in high-profile peer reviewed journals and the researchers have presented their findings at leading international conferences.
IFOX has provided the groundwork and knowledge necessary for magnetic oxide inclusion in future generations of ICT products, critically, by engaging a key end-user company Fiat. The new knowledge generated by IFOX is targeted in such a critical area that it will be truly groundbreaking in worldwide terms and put Europe ahead of other regions, safeguarding existing ICT jobs in Europe and working to create new markets for job increases. This should help the stagnation of the EU-27's R&D investment (R&D expenditure as % of GDP) which is currently at 1.84% vs. the Lisbon target of 3%.
Dissemination and/or exploitation of project results, and management of intellectual property
The IFOX programme is centred on increasing the knowledge and understanding about transition metal oxides and their properties and also on applying this knowledge to create new oxide functionalities for potential device application. A key component of this effort is to generate and protect intellectual property (IP) with a view towards exploitation and delivering novel device functionalities in industry.
Dissemination
IFOX contributed during its whole duration to create and disseminate knowledge on complex oxides and interfaces to the overall community. In particular, the consortium published on peer reviewed international journals 91 manuscripts, and dissemination to the major inter- national conferences almost the same number of talks.
IFOX also disseminated results in the frame of WP9 by organizing a Summer school which took place in Bildungszentrum Hesselberg from July 29° to August the 3° 2014 that count- ed 35 students recruited within IFOX and 10 external students recruited by other projects attending. Nine speakers were experts of IFOX and one was invited from the ORAMA project.
Towards the end of the project the dissemination became more and more important. During the 2015 E-MRS Fall Meeting, held in Warsaw University of Technology, from September 15 to 18 2015, the IFOX consortium was presented within symposium L "Materials for electronics and optoelectronic applications away from silicon”. The IFOX session had the following talks:
11:00 Gertjan Koster, University of Twente, Enschede, the Netherlands, "Integrating epi- taxial oxides with Si"
11:15 Rik Groenen, TSST, Enschede, the Netherlands, "large area epitaxial oxide growth using PLD"
11:30 Gwenael Atcheson, TCD, Dublin, Ireland, "All-oxide multiferroic tunnel junctions"
11:45 Patrizio Graziosi CNR-ISMN, "Memristive and magnetoresistive properties of
SrTiO3 based junctions" 12:00 Nicolas Gauquelin, EMAT, University of Antwerp, "Application of advanced mi- croscopy techniques to the study of oxide-based electronic devices"
12:15 Mehran Vafaee, Johannes-Gutenberg-Universität Mainz, Mainz, Germany "The cor- relation between the interface roughness and exchange bias coupling strength in BiFeO3/La0.7Sr0.3MnO3 heterostructures"
IFOX also setup a booth at this E-MRS meeting with posters on all heterostructures in which discussions with possibly interested industry partners were held. In addition, a massive search in patent databases was done in order to evaluate possible restrictions and possible areas for protection of IP and to identify competing technologies.
Exploitation of results
With respect to exploitation, the industrial partners were mostly responsible for the analysis according to the respective man power allocated. Magnetron sputtering is acknowledged as being one of the most promising techniques for industrial use perspective, and it was used here by PST to exploit the production of complex oxide buffer layers for many industrial sectors (both for low- and high-end prod- ucts) where thin films are a mass-production major issue. On the other hand, TSST and OS are actively involved in the production of deposition equipment for these applications. Finally, CRF is one of the major end users for the novel sensors developed within IFOX.
PST, besides analysing more than 1000 valid patents to preview possible technological bottlenecks possibly preventing use of the know-how developed (buffer layers), also found through the analysis the possible market incubators to place their machinery and processes. The analysis carried out to foresee the use potential of the complex oxide buffer materials developed in the project took into account the material quality level in terms of (i) crystallinity (H/E: highly crystalline/Epitaxial, HT: highly tex- tured, LT: low texturing, P: polycrystalline), (ii) adhesion, and (iii) homogeneity over area as large as 4” to address that single applications according to the material required need. This analysis is essential to tailor the material production according to the chosen use and properly penetrate the markets addressed.
From this analysis the exploitation strategy of PST is to offer to the market buffer materials of tailored crystallinity to allow significantly decrease their cost for unit and let novel product penetrate the market.
TSST has built a prototype for large area deposition PLD and developed the necessary processing technology in the course of IFOX. Upscaling is demonstrated of the synthesis of complex oxide films with perfect control over the structure and composition of the material on large wafer size Si up to 4”. This has led to new standards for the precision synthesis of oxide films, and this removed in a radical manner the limitations on the availability of such films on industrial scale Si wafers. Important technological steps, including technologies to uniformly deposit the material on a uniformly heated substrate, have been made. These steps have enabled the epitaxial growth of crystalline complex oxide thin films on Si. The PLD system is at present ready to enter the market, enabling TSST to increase the market share. TSST hopes to expand their sales by 30% in the next coming years.
OS has in the course of IFOX built a prototype for a multisource pulsed plasma deposition system, (PPD - 5 parallel sources), able to deposit oxides with 6% of homogeneity on 3‘‘ wafers and on an ‘‘infinite‘‘ strips of 300mm width. This industrial prototype is already working in an Industrial Research Centre and almost ready to enter the market in 2017. At the same time, OS has started to develop an enlargement of the area covered by a single source. The result is that the last prototype system uses only two sources to deposit on substrate areas of 300mm width instead of the previous 5 sources.
OS hopes to expand their sales by 2018 by dealing this type of system specifically adapted for customer applications.
Another important business strategy that OS will follow is the possibility to insert a PPD source in an already used industrials system since PPD deposition rate has become compatible with similar deposition techniques.
For CRF, exhaust gas sensors is the main application of the IFOX complex oxides developed, namely: YSZ, CeO2, ZnO, and LSMO; oxides which are all sensitive to the gas exhausted by internal combustion engine powered cars, which do guarantee improved sensitivity and selectivity in the shape of sandwiches as e.g. HS4 addressed by IFOX. A plurality of sensor types and electrochemical detection systems are in use for the polluting gaseous compounds exhausted by the current production internal combustion engines. In particular, the focus is placed on the detection systems dedicated to the detection of carbon mono-oxide and di-oxide (CO, CO2), and of nitrogen oxides (NOx), used to limit noxious emissions, to control the EGR flux, and to detect the combustion misfire. Nitrogen oxide gases, commonly referred to collectively as NOX, are common pollutants formed in internal combustion engines and industrial combustion systems by thermal fixation and oxidation of atmospheric nitrogen1. NOX, which includes NO2 and NO gases, have adverse effects on the environment, and are the leading cause of green-house effect, acid rain, and photochemical smog. In humans, exposure to more than 3 ppm of NO2 gas for periods longer than 8 hours can cause respiratory and cardiovascular diseases. Hence, better detection of such gases is of utmost importance, and that calls for better understanding of detection mechanism to facilitate development and optimization of sensing devices. Semiconducting metal oxides based thin films have traditionally been used in detection of NOX gases, with the common measuring arrangement relying on changes in electrical conductivity upon interaction with the gas. In particular, ZnO based gas sensors have been employed extensively in the detection of both reducing and oxidizing gases due to the superior properties of this material such as enhanced sensitivity, stability, and low cost. A widely adopted route to reduce NOx emissions is Exhaust Gas Recirculation (EGR). The EGR system has a valve which cannot respond instantly to changes in demand, and the exhaust gas takes time to flow around the EGR circuit. This makes the calibration of transient EGR behaviour particularly complex. To have a better control of EGR it would be very important to be capable of sensing the NOx content, or relatively knowing the actual O2 concentration in the exhaust air to be recirculated, and at the same time controlling precisely the quantity of gas recirculated. This in principle can be done if a gas sensor to probe precisely the quantity of O2 or the NOx within the exhausted mixture, and at the same time, a contactless magnetic sensor to instantly know the throttle-valve position which adjust the quantity of EGR. For all these reasons to have access to a multifuctional sensor capable to probe, also non simultaneously, a recirculated-gas concentration and the position of a magnetized throttle valve on move would allow to close-loop drive the EGR precisely. The IFOX HS4 is the bases of a multifunctional sensor conceived of probing precisely the quantity of NOx within the exhausted mixture, and at the same time, measuring the position of the EGR valve which adjusts the quantity of EGR. With the presence in the consortium of PST, and TSST, the supply chain of the manufacturing of the HS4 multifunctional sensor is complete, thus paving the way for a promising exploitation strategy of the result which can lead to a start of a new product to serve the worldwide automotive market as a whole if quality and cost will be sufficient over the full life cycle, thus to include to fuel consumption and emission reduction of the future vehicle employing the HS4 assisted combustion control. Although to give a honest assessment on the potential use of HS4 solution other development and analysis including the optimized models and engine control algorithms, as well as a real testing on a running vehicle are required, as well as the evaluation of pilot production of the device to allow the calculation of cost of production. It is a serious intention to plan the further understanding of these open points in a future collaborative research project, possibly in the context of the ICT theme.
Management of intellectual property
PST, has analysed more than 1000 valid patents to preview possible technological bottlenecks possibly preventing use of the know-how developed in the area of buffer layers. To this end, patents today still valid have been matter of intensive investigation, and the solutions protected compared with the new publications in order to evaluate beforehand the possible bottlenecks of the technological choices undertaken in the project, and assess better the interim results of the project against the prior-art. Particular attention was paid to the protected use of the addressed buffer layers to see the perspective of exploitation and screen possible clients of customized materials for post-project co-designing of new potential productions.
From this patent analysis the complex oxide buffer layer developed in the project is very often a prerequisite to conceive a product rather than the core of the idea. This issue opens up the perspective for the buffer material developers (in particular PST, but also TSST more specifically for oxide electronics) in the project to be part of new product supply chains. As one can see from this report, most of products employ today oxide buffer layer of extremely controlled crystallinity and quality standards, while, aside from microelectron- ics, most of them would find suitable, at a lower cost, less quality materials (sensors, dis- crete devices, coatings and electrolyte, fuel cells, etc.). From this assumption the prior ex- ploitation strategy of PST is to offer to the market buffer materials of tailored crystallinity to allow significantly decrease their cost per unit and let novel products penetrate the market.
CRF contributed 6PM on studying the IPR, and the current technologies of sensing to gauge the applicability perspective of the IFOX achievements, and in particular for the heterostructures more related to sensing functionalities, namely HS4 and HS6. A report on sensors for automotive was delivered to the commission at month 24.
CRF expects to generate better and solid IPR on the novel design based on HS4. Right now no patent has been filed yet as more development is needed to increase confidence.
Potential impacts
Socio-economic impact
At present oxides are important for semiconductor technology mainly in the context of improving the dielectric/insulating and interface properties of gate oxides for use in MOSFET devices. However, their real promise for novel future applications is that they can be engineered with additional functionality, e.g. ferroelectric, multiferroic, and magnetic properties, by tailoring their chemical composition and making multilayer stacks of a range of different oxides.
The aim of IFOX was to explore, create and control novel electronic and magnetic functionalities that result from the rich interplay of charge, spin and orbital degrees of freedom in transition metal oxide heterostructures and their interfaces. The outcome aimed at the establishment of a material platform for novel ‘More than Moore’ and ‘beyond CMOS’ electronics, compatible with VLSI and which delivers performance and functionality far beyond the state-of-the-art.
To achieve these goals we carried out a research programme with workpackages on theory and modelling, epitaxial growth, processing, characterization, device fabrication and knowledge transfer. Crucially, technological relevance and impact was guaranteed by the active participation of a world-leading company, Fiat and three SMEs, namely TSST and Organic Spintronics, with expertise and infrastructure for epitaxial oxide growth on Si and Protec Surface Technology, with expertise on engineering and production of PVD and PECVD machinery for thin films and coatings deposition and customised deposition process developments. Focusing the research on industry needs ensured that IFOX worked on relevant problems with tangible results. Success for the project means delivering useable new technology and improved understanding to industry beneficiaries which will allow them to make significant progress on the roadmap for future generations of semiconductor devices.
While the first three periods of IFOX were dedicated to the growth, processing, and investigation of various oxide thin films and their heterostructures, the final period focused on seven different heterostructures, chosen for their largest potential for possible future applications in the target research areas:
1) Manganite/TiO based heterostructures which exhibit resistive switching suitable for novel storage and possibly logic applications
2) Heterostructures based on LaAlO3/SrTiO3 with additional ferromagnetic properties to combine transistor functionality with non-volatile storage for example for logic applications
3) Heterostructures based on BaTiO3 and NiFeO3 with inherent tunneling magnetoresistive functionality which may allow for MRAM cells which are switched by the magnetoelectric coupling between the BTO and the NFO.
4) Heterostructures of (La,Sr)MnO3 (LSMO) and ZnO which combine magnetoresistive sensing and gas sensing properties for automotive applications
5) Heterostructures using a bilayer of (La,Sr)MnO3 and BiFeO3 in which the magnetization of the LSMO can be switched via magnetoelectric coupling to the BFO which again can be switched by an electrical pulse. This structure would also address the area of magnetic memory with low switching power
6) Superlattices of LaMnO3 and SrMnO3 in which the magnetic properties compared to LSMO are much improved at room temperature and very high sensitivity in sensor applications (MR-sensors and Bolometers are to be expected.
7) Heterostructures using ferromagnetic (La,Sr)MnO3, a barrier of SrTiO3 and a ferromagnetic metal counter electrode which realise a combined element for resistive switching and magnetoresistance which can serve as a stateful logic device to combine logic and memory in a single cell.
In addition the project foresaw an effort to attempt and assess the realization of all these heterostructures on large area silicon wafers dramatically reducing possible production costs compared to heterostructures deposited on single crystalline oxide substrates.
All these seven heterostructures were further developed including growth on silicon substrates and a thorough assessment was performed. For two of the seven heterostructures (ZnO/LSMO-based gas sensor and PCMO/TiO based ReRAM) all important factors namely functionality at room temperature or above, fabrication on large area silicon wafers and industry compatible processing were demonstrated. As follow up, CRF will continue the study to increase the TRL of the HS4 up to 6. Besides, to go deep inside into the measurement repeatibility of the gas sensitivity, and the life cycle costing of the production will be part of the future investigation to evaluate the value for money in the replacement of the standard gas sensor today employed. The evaluation will take into account the improved absolute sensitivity of HS4 demonstrated in IFOX. The exploitation of the coupled magneto resistance funtion of HS4 will be requiring further developments and analysis to gauge at what extent the airgap variation sensitivity of the MR response will affect the precision of the absolute position of a mangnetized valve. It is expected to go deeper inside these issue by a second ICT project employing novel advance designed smart systems.
To facilitate the transfer of results to industry all partners contributed to a huge effort, which was guided by the industry partners, to collect the knowledge on growth and processing in a suitable format. This resulted in three major deliverables
1) A catalogue of growth techniques and growth parameters for all heterostructures
2) A catalogue of processing techniques and parameters for all heterostructures
3) A catalogue of all chemicals involved in growth and processing, assessing the environmental impact of the chemical compounds
In addition a tentative assessment of growth and processing in terms of technology readiness level (TRL) is provided, resulting in typical values ranging from 3 to 5 for all steps (see WP5 for details).
Exploration of new functionalities
For all heterostructures an overall assessment is provided which not only addresses the results of characterization. It also provides a placement of the heterostructure within the targeted research area and an evaluation of the maturity of the necessary technology and a competitive comparison to possible alternative devices on the market or in the development chain.
The data of the three catalogues allow for immediate assessment of needs and risks when transferring these heterostrctures to industrial application.
After completion of the programme, we can state that IFOX has delivered results that are highly relevant for our industrial partners in terms of novel sensors, growth and processing and integration to Si.
Societal implications
The importance of the ICT sector from an economic, societal, and (increasingly) environmental viewpoint is obvious: Europe needs a strategic and innovative semiconductor foundation underlying the ICT portfolio to bolster industry, increase innovation and take a lead in economic, societal and environmental terms. The semiconductor market is currently worth 200 billion Euro (256 billion USD) globally on an annual basis. There is a multiplier effect of approximately 25 times that amount in terms of equipment and services ranging from television sets to IT. In other words, semiconductor technology provides the knowledge and technologies that generate some 10% of global GDP. The European semiconductor market attained a value of approximately Euro 30 billion in 2007, representing just 16% of the worldwide total .
IFOX has improved the competitiveness of the EU semiconductor industry by its development of novel device technologies based on oxide materials with novel device functionalities beyond the current state-of-the-art. This will support the need for industry linked technological innovation in the EU as described in the 2008 Europe Innova Innovation watch report: "A prospective innovation challenge for the ICT sector is the insufficient collaboration between public sector and business sector researchers".
The results obtained in IFOX is multi-disciplinary and beyond what could have been accomplished by one or two individual research groups as they do not have the expertise nor infrastructure for all the various aspects of theory, growth, structuring and characterization necessary for the achieved functional structures. In addition, without the critical evaluation and feedback of the industrial partners, the focusing on the most relevant and prospective structures and their compatibility with Si technology would not have been accomplished.
Regarding impact of research, the EU still ranks as the world’s largest producer of scientific knowledge (measured by publications), but contributes less than the US to high impact publications . These data provide the backdrop against which the consortium has tackled its research goals. The best research conducted within IFOX has been published in high-profile peer reviewed journals and the researchers have presented their findings at leading international conferences.
IFOX has provided the groundwork and knowledge necessary for magnetic oxide inclusion in future generations of ICT products, critically, by engaging a key end-user company Fiat. The new knowledge generated by IFOX is targeted in such a critical area that it will be truly groundbreaking in worldwide terms and put Europe ahead of other regions, safeguarding existing ICT jobs in Europe and working to create new markets for job increases. This should help the stagnation of the EU-27's R&D investment (R&D expenditure as % of GDP) which is currently at 1.84% vs. the Lisbon target of 3%.
Dissemination and/or exploitation of project results, and management of intellectual property
The IFOX programme is centred on increasing the knowledge and understanding about transition metal oxides and their properties and also on applying this knowledge to create new oxide functionalities for potential device application. A key component of this effort is to generate and protect intellectual property (IP) with a view towards exploitation and delivering novel device functionalities in industry.
Dissemination
IFOX contributed during its whole duration to create and disseminate knowledge on complex oxides and interfaces to the overall community. In particular, the consortium published on peer reviewed international journals 91 manuscripts, and dissemination to the major inter- national conferences almost the same number of talks.
IFOX also disseminated results in the frame of WP9 by organizing a Summer school which took place in Bildungszentrum Hesselberg from July 29° to August the 3° 2014 that count- ed 35 students recruited within IFOX and 10 external students recruited by other projects attending. Nine speakers were experts of IFOX and one was invited from the ORAMA project.
Towards the end of the project the dissemination became more and more important. During the 2015 E-MRS Fall Meeting, held in Warsaw University of Technology, from September 15 to 18 2015, the IFOX consortium was presented within symposium L "Materials for electronics and optoelectronic applications away from silicon”. The IFOX session had the following talks:
11:00 Gertjan Koster, University of Twente, Enschede, the Netherlands, "Integrating epi- taxial oxides with Si"
11:15 Rik Groenen, TSST, Enschede, the Netherlands, "large area epitaxial oxide growth using PLD"
11:30 Gwenael Atcheson, TCD, Dublin, Ireland, "All-oxide multiferroic tunnel junctions"
11:45 Patrizio Graziosi CNR-ISMN, "Memristive and magnetoresistive properties of
SrTiO3 based junctions" 12:00 Nicolas Gauquelin, EMAT, University of Antwerp, "Application of advanced mi- croscopy techniques to the study of oxide-based electronic devices"
12:15 Mehran Vafaee, Johannes-Gutenberg-Universität Mainz, Mainz, Germany "The cor- relation between the interface roughness and exchange bias coupling strength in BiFeO3/La0.7Sr0.3MnO3 heterostructures"
IFOX also setup a booth at this E-MRS meeting with posters on all heterostructures in which discussions with possibly interested industry partners were held. In addition, a massive search in patent databases was done in order to evaluate possible restrictions and possible areas for protection of IP and to identify competing technologies.
Exploitation of results
With respect to exploitation, the industrial partners were mostly responsible for the analysis according to the respective man power allocated. Magnetron sputtering is acknowledged as being one of the most promising techniques for industrial use perspective, and it was used here by PST to exploit the production of complex oxide buffer layers for many industrial sectors (both for low- and high-end prod- ucts) where thin films are a mass-production major issue. On the other hand, TSST and OS are actively involved in the production of deposition equipment for these applications. Finally, CRF is one of the major end users for the novel sensors developed within IFOX.
PST, besides analysing more than 1000 valid patents to preview possible technological bottlenecks possibly preventing use of the know-how developed (buffer layers), also found through the analysis the possible market incubators to place their machinery and processes. The analysis carried out to foresee the use potential of the complex oxide buffer materials developed in the project took into account the material quality level in terms of (i) crystallinity (H/E: highly crystalline/Epitaxial, HT: highly tex- tured, LT: low texturing, P: polycrystalline), (ii) adhesion, and (iii) homogeneity over area as large as 4” to address that single applications according to the material required need. This analysis is essential to tailor the material production according to the chosen use and properly penetrate the markets addressed.
From this analysis the exploitation strategy of PST is to offer to the market buffer materials of tailored crystallinity to allow significantly decrease their cost for unit and let novel product penetrate the market.
TSST has built a prototype for large area deposition PLD and developed the necessary processing technology in the course of IFOX. Upscaling is demonstrated of the synthesis of complex oxide films with perfect control over the structure and composition of the material on large wafer size Si up to 4”. This has led to new standards for the precision synthesis of oxide films, and this removed in a radical manner the limitations on the availability of such films on industrial scale Si wafers. Important technological steps, including technologies to uniformly deposit the material on a uniformly heated substrate, have been made. These steps have enabled the epitaxial growth of crystalline complex oxide thin films on Si. The PLD system is at present ready to enter the market, enabling TSST to increase the market share. TSST hopes to expand their sales by 30% in the next coming years.
OS has in the course of IFOX built a prototype for a multisource pulsed plasma deposition system, (PPD - 5 parallel sources), able to deposit oxides with 6% of homogeneity on 3‘‘ wafers and on an ‘‘infinite‘‘ strips of 300mm width. This industrial prototype is already working in an Industrial Research Centre and almost ready to enter the market in 2017. At the same time, OS has started to develop an enlargement of the area covered by a single source. The result is that the last prototype system uses only two sources to deposit on substrate areas of 300mm width instead of the previous 5 sources.
OS hopes to expand their sales by 2018 by dealing this type of system specifically adapted for customer applications.
Another important business strategy that OS will follow is the possibility to insert a PPD source in an already used industrials system since PPD deposition rate has become compatible with similar deposition techniques.
For CRF, exhaust gas sensors is the main application of the IFOX complex oxides developed, namely: YSZ, CeO2, ZnO, and LSMO; oxides which are all sensitive to the gas exhausted by internal combustion engine powered cars, which do guarantee improved sensitivity and selectivity in the shape of sandwiches as e.g. HS4 addressed by IFOX. A plurality of sensor types and electrochemical detection systems are in use for the polluting gaseous compounds exhausted by the current production internal combustion engines. In particular, the focus is placed on the detection systems dedicated to the detection of carbon mono-oxide and di-oxide (CO, CO2), and of nitrogen oxides (NOx), used to limit noxious emissions, to control the EGR flux, and to detect the combustion misfire. Nitrogen oxide gases, commonly referred to collectively as NOX, are common pollutants formed in internal combustion engines and industrial combustion systems by thermal fixation and oxidation of atmospheric nitrogen1. NOX, which includes NO2 and NO gases, have adverse effects on the environment, and are the leading cause of green-house effect, acid rain, and photochemical smog. In humans, exposure to more than 3 ppm of NO2 gas for periods longer than 8 hours can cause respiratory and cardiovascular diseases. Hence, better detection of such gases is of utmost importance, and that calls for better understanding of detection mechanism to facilitate development and optimization of sensing devices. Semiconducting metal oxides based thin films have traditionally been used in detection of NOX gases, with the common measuring arrangement relying on changes in electrical conductivity upon interaction with the gas. In particular, ZnO based gas sensors have been employed extensively in the detection of both reducing and oxidizing gases due to the superior properties of this material such as enhanced sensitivity, stability, and low cost. A widely adopted route to reduce NOx emissions is Exhaust Gas Recirculation (EGR). The EGR system has a valve which cannot respond instantly to changes in demand, and the exhaust gas takes time to flow around the EGR circuit. This makes the calibration of transient EGR behaviour particularly complex. To have a better control of EGR it would be very important to be capable of sensing the NOx content, or relatively knowing the actual O2 concentration in the exhaust air to be recirculated, and at the same time controlling precisely the quantity of gas recirculated. This in principle can be done if a gas sensor to probe precisely the quantity of O2 or the NOx within the exhausted mixture, and at the same time, a contactless magnetic sensor to instantly know the throttle-valve position which adjust the quantity of EGR. For all these reasons to have access to a multifuctional sensor capable to probe, also non simultaneously, a recirculated-gas concentration and the position of a magnetized throttle valve on move would allow to close-loop drive the EGR precisely. The IFOX HS4 is the bases of a multifunctional sensor conceived of probing precisely the quantity of NOx within the exhausted mixture, and at the same time, measuring the position of the EGR valve which adjusts the quantity of EGR. With the presence in the consortium of PST, and TSST, the supply chain of the manufacturing of the HS4 multifunctional sensor is complete, thus paving the way for a promising exploitation strategy of the result which can lead to a start of a new product to serve the worldwide automotive market as a whole if quality and cost will be sufficient over the full life cycle, thus to include to fuel consumption and emission reduction of the future vehicle employing the HS4 assisted combustion control. Although to give a honest assessment on the potential use of HS4 solution other development and analysis including the optimized models and engine control algorithms, as well as a real testing on a running vehicle are required, as well as the evaluation of pilot production of the device to allow the calculation of cost of production. It is a serious intention to plan the further understanding of these open points in a future collaborative research project, possibly in the context of the ICT theme.
Management of intellectual property
PST, has analysed more than 1000 valid patents to preview possible technological bottlenecks possibly preventing use of the know-how developed in the area of buffer layers. To this end, patents today still valid have been matter of intensive investigation, and the solutions protected compared with the new publications in order to evaluate beforehand the possible bottlenecks of the technological choices undertaken in the project, and assess better the interim results of the project against the prior-art. Particular attention was paid to the protected use of the addressed buffer layers to see the perspective of exploitation and screen possible clients of customized materials for post-project co-designing of new potential productions.
From this patent analysis the complex oxide buffer layer developed in the project is very often a prerequisite to conceive a product rather than the core of the idea. This issue opens up the perspective for the buffer material developers (in particular PST, but also TSST more specifically for oxide electronics) in the project to be part of new product supply chains. As one can see from this report, most of products employ today oxide buffer layer of extremely controlled crystallinity and quality standards, while, aside from microelectron- ics, most of them would find suitable, at a lower cost, less quality materials (sensors, dis- crete devices, coatings and electrolyte, fuel cells, etc.). From this assumption the prior ex- ploitation strategy of PST is to offer to the market buffer materials of tailored crystallinity to allow significantly decrease their cost per unit and let novel products penetrate the market.
CRF contributed 6PM on studying the IPR, and the current technologies of sensing to gauge the applicability perspective of the IFOX achievements, and in particular for the heterostructures more related to sensing functionalities, namely HS4 and HS6. A report on sensors for automotive was delivered to the commission at month 24.
CRF expects to generate better and solid IPR on the novel design based on HS4. Right now no patent has been filed yet as more development is needed to increase confidence.
Potential impacts
Socio-economic impact
At present oxides are important for semiconductor technology mainly in the context of improving the dielectric/insulating and interface properties of gate oxides for use in MOSFET devices. However, their real promise for novel future applications is that they can be engineered with additional functionality, e.g. ferroelectric, multiferroic, and magnetic properties, by tailoring their chemical composition and making multilayer stacks of a range of different oxides.
The aim of IFOX was to explore, create and control novel electronic and magnetic functionalities that result from the rich interplay of charge, spin and orbital degrees of freedom in transition metal oxide heterostructures and their interfaces. The outcome aimed at the establishment of a material platform for novel ‘More than Moore’ and ‘beyond CMOS’ electronics, compatible with VLSI and which delivers performance and functionality far beyond the state-of-the-art.
To achieve these goals we carried out a research programme with workpackages on theory and modelling, epitaxial growth, processing, characterization, device fabrication and knowledge transfer. Crucially, technological relevance and impact was guaranteed by the active participation of a world-leading company, Fiat and three SMEs, namely TSST and Organic Spintronics, with expertise and infrastructure for epitaxial oxide growth on Si and Protec Surface Technology, with expertise on engineering and production of PVD and PECVD machinery for thin films and coatings deposition and customised deposition process developments. Focusing the research on industry needs ensured that IFOX worked on relevant problems with tangible results. Success for the project means delivering useable new technology and improved understanding to industry beneficiaries which will allow them to make significant progress on the roadmap for future generations of semiconductor devices.
While the first three periods of IFOX were dedicated to the growth, processing, and investigation of various oxide thin films and their heterostructures, the final period focused on seven different heterostructures, chosen for their largest potential for possible future applications in the target research areas:
1) Manganite/TiO based heterostructures which exhibit resistive switching suitable for novel storage and possibly logic applications
2) Heterostructures based on LaAlO3/SrTiO3 with additional ferromagnetic properties to combine transistor functionality with non-volatile storage for example for logic applications
3) Heterostructures based on BaTiO3 and NiFeO3 with inherent tunneling magnetoresistive functionality which may allow for MRAM cells which are switched by the magnetoelectric coupling between the BTO and the NFO.
4) Heterostructures of (La,Sr)MnO3 (LSMO) and ZnO which combine magnetoresistive sensing and gas sensing properties for automotive applications
5) Heterostructures using a bilayer of (La,Sr)MnO3 and BiFeO3 in which the magnetization of the LSMO can be switched via magnetoelectric coupling to the BFO which again can be switched by an electrical pulse. This structure would also address the area of magnetic memory with low switching power
6) Superlattices of LaMnO3 and SrMnO3 in which the magnetic properties compared to LSMO are much improved at room temperature and very high sensitivity in sensor applications (MR-sensors and Bolometers are to be expected.
7) Heterostructures using ferromagnetic (La,Sr)MnO3, a barrier of SrTiO3 and a ferromagnetic metal counter electrode which realise a combined element for resistive switching and magnetoresistance which can serve as a stateful logic device to combine logic and memory in a single cell.
In addition the project foresaw an effort to attempt and assess the realization of all these heterostructures on large area silicon wafers dramatically reducing possible production costs compared to heterostructures deposited on single crystalline oxide substrates.
All these seven heterostructures were further developed including growth on silicon substrates and a thorough assessment was performed. For two of the seven heterostructures (ZnO/LSMO-based gas sensor and PCMO/TiO based ReRAM) all important factors namely functionality at room temperature or above, fabrication on large area silicon wafers and industry compatible processing were demonstrated. As follow up, CRF will continue the study to increase the TRL of the HS4 up to 6. Besides, to go deep inside into the measurement repeatibility of the gas sensitivity, and the life cycle costing of the production will be part of the future investigation to evaluate the value for money in the replacement of the standard gas sensor today employed. The evaluation will take into account the improved absolute sensitivity of HS4 demonstrated in IFOX. The exploitation of the coupled magneto resistance funtion of HS4 will be requiring further developments and analysis to gauge at what extent the airgap variation sensitivity of the MR response will affect the precision of the absolute position of a mangnetized valve. It is expected to go deeper inside these issue by a second ICT project employing novel advance designed smart systems.
To facilitate the transfer of results to industry all partners contributed to a huge effort, which was guided by the industry partners, to collect the knowledge on growth and processing in a suitable format. This resulted in three major deliverables
1) A catalogue of growth techniques and growth parameters for all heterostructures
2) A catalogue of processing techniques and parameters for all heterostructures
3) A catalogue of all chemicals involved in growth and processing, assessing the environmental impact of the chemical compounds
In addition a tentative assessment of growth and processing in terms of technology readiness level (TRL) is provided, resulting in typical values ranging from 3 to 5 for all steps (see WP5 for details).
Exploration of new functionalities
For all heterostructures an overall assessment is provided which not only addresses the results of characterization. It also provides a placement of the heterostructure within the targeted research area and an evaluation of the maturity of the necessary technology and a competitive comparison to possible alternative devices on the market or in the development chain.
The data of the three catalogues allow for immediate assessment of needs and risks when transferring these heterostrctures to industrial application.
After completion of the programme, we can state that IFOX has delivered results that are highly relevant for our industrial partners in terms of novel sensors, growth and processing and integration to Si.
Societal implications
The importance of the ICT sector from an economic, societal, and (increasingly) environmental viewpoint is obvious: Europe needs a strategic and innovative semiconductor foundation underlying the ICT portfolio to bolster industry, increase innovation and take a lead in economic, societal and environmental terms. The semiconductor market is currently worth 200 billion Euro (256 billion USD) globally on an annual basis. There is a multiplier effect of approximately 25 times that amount in terms of equipment and services ranging from television sets to IT. In other words, semiconductor technology provides the knowledge and technologies that generate some 10% of global GDP. The European semiconductor market attained a value of approximately Euro 30 billion in 2007, representing just 16% of the worldwide total .
IFOX has improved the competitiveness of the EU semiconductor industry by its development of novel device technologies based on oxide materials with novel device functionalities beyond the current state-of-the-art. This will support the need for industry linked technological innovation in the EU as described in the 2008 Europe Innova Innovation watch report: "A prospective innovation challenge for the ICT sector is the insufficient collaboration between public sector and business sector researchers".
The results obtained in IFOX is multi-disciplinary and beyond what could have been accomplished by one or two individual research groups as they do not have the expertise nor infrastructure for all the various aspects of theory, growth, structuring and characterization necessary for the achieved functional structures. In addition, without the critical evaluation and feedback of the industrial partners, the focusing on the most relevant and prospective structures and their compatibility with Si technology would not have been accomplished.
Regarding impact of research, the EU still ranks as the world’s largest producer of scientific knowledge (measured by publications), but contributes less than the US to high impact publications . These data provide the backdrop against which the consortium has tackled its research goals. The best research conducted within IFOX has been published in high-profile peer reviewed journals and the researchers have presented their findings at leading international conferences.
IFOX has provided the groundwork and knowledge necessary for magnetic oxide inclusion in future generations of ICT products, critically, by engaging a key end-user company Fiat. The new knowledge generated by IFOX is targeted in such a critical area that it will be truly groundbreaking in worldwide terms and put Europe ahead of other regions, safeguarding existing ICT jobs in Europe and working to create new markets for job increases. This should help the stagnation of the EU-27's R&D investment (R&D expenditure as % of GDP) which is currently at 1.84% vs. the Lisbon target of 3%.
Dissemination and/or exploitation of project results, and management of intellectual property
The IFOX programme is centred on increasing the knowledge and understanding about transition metal oxides and their properties and also on applying this knowledge to create new oxide functionalities for potential device application. A key component of this effort is to generate and protect intellectual property (IP) with a view towards exploitation and delivering novel device functionalities in industry.
Dissemination
IFOX contributed during its whole duration to create and disseminate knowledge on complex oxides and interfaces to the overall community. In particular, the consortium published on peer reviewed international journals 91 manuscripts, and dissemination to the major inter- national conferences almost the same number of talks.
IFOX also disseminated results in the frame of WP9 by organizing a Summer school which took place in Bildungszentrum Hesselberg from July 29° to August the 3° 2014 that count- ed 35 students recruited within IFOX and 10 external students recruited by other projects attending. Nine speakers were experts of IFOX and one was invited from the ORAMA project.
Towards the end of the project the dissemination became more and more important. During the 2015 E-MRS Fall Meeting, held in Warsaw University of Technology, from September 15 to 18 2015, the IFOX consortium was presented within symposium L "Materials for electronics and optoelectronic applications away from silicon”. The IFOX session had the following talks:
11:00 Gertjan Koster, University of Twente, Enschede, the Netherlands, "Integrating epi- taxial oxides with Si"
11:15 Rik Groenen, TSST, Enschede, the Netherlands, "large area epitaxial oxide growth using PLD"
11:30 Gwenael Atcheson, TCD, Dublin, Ireland, "All-oxide multiferroic tunnel junctions"
11:45 Patrizio Graziosi CNR-ISMN, "Memristive and magnetoresistive properties of
SrTiO3 based junctions" 12:00 Nicolas Gauquelin, EMAT, University of Antwerp, "Application of advanced mi- croscopy techniques to the study of oxide-based electronic devices"
12:15 Mehran Vafaee, Johannes-Gutenberg-Universität Mainz, Mainz, Germany "The cor- relation between the interface roughness and exchange bias coupling strength in BiFeO3/La0.7Sr0.3MnO3 heterostructures"
IFOX also setup a booth at this E-MRS meeting with posters on all heterostructures in which discussions with possibly interested industry partners were held. In addition, a massive search in patent databases was done in order to evaluate possible restrictions and possible areas for protection of IP and to identify competing technologies.
Exploitation of results
With respect to exploitation, the industrial partners were mostly responsible for the analysis according to the respective man power allocated. Magnetron sputtering is acknowledged as being one of the most promising techniques for industrial use perspective, and it was used here by PST to exploit the production of complex oxide buffer layers for many industrial sectors (both for low- and high-end prod- ucts) where thin films are a mass-production major issue. On the other hand, TSST and OS are actively involved in the production of deposition equipment for these applications. Finally, CRF is one of the major end users for the novel sensors developed within IFOX.
PST, besides analysing more than 1000 valid patents to preview possible technological bottlenecks possibly preventing use of the know-how developed (buffer layers), also found through the analysis the possible market incubators to place their machinery and processes. The analysis carried out to foresee the use potential of the complex oxide buffer materials developed in the project took into account the material quality level in terms of (i) crystallinity (H/E: highly crystalline/Epitaxial, HT: highly tex- tured, LT: low texturing, P: polycrystalline), (ii) adhesion, and (iii) homogeneity over area as large as 4” to address that single applications according to the material required need. This analysis is essential to tailor the material production according to the chosen use and properly penetrate the markets addressed.
From this analysis the exploitation strategy of PST is to offer to the market buffer materials of tailored crystallinity to allow significantly decrease their cost for unit and let novel product penetrate the market.
TSST has built a prototype for large area deposition PLD and developed the necessary processing technology in the course of IFOX. Upscaling is demonstrated of the synthesis of complex oxide films with perfect control over the structure and composition of the material on large wafer size Si up to 4”. This has led to new standards for the precision synthesis of oxide films, and this removed in a radical manner the limitations on the availability of such films on industrial scale Si wafers. Important technological steps, including technologies to uniformly deposit the material on a uniformly heated substrate, have been made. These steps have enabled the epitaxial growth of crystalline complex oxide thin films on Si. The PLD system is at present ready to enter the market, enabling TSST to increase the market share. TSST hopes to expand their sales by 30% in the next coming years.
OS has in the course of IFOX built a prototype for a multisource pulsed plasma deposition system, (PPD - 5 parallel sources), able to deposit oxides with 6% of homogeneity on 3‘‘ wafers and on an ‘‘infinite‘‘ strips of 300mm width. This industrial prototype is already working in an Industrial Research Centre and almost ready to enter the market in 2017. At the same time, OS has started to develop an enlargement of the area covered by a single source. The result is that the last prototype system uses only two sources to deposit on substrate areas of 300mm width instead of the previous 5 sources.
OS hopes to expand their sales by 2018 by dealing this type of system specifically adapted for customer applications.
Another important business strategy that OS will follow is the possibility to insert a PPD source in an already used industrials system since PPD deposition rate has become compatible with similar deposition techniques.
For CRF, exhaust gas sensors is the main application of the IFOX complex oxides developed, namely: YSZ, CeO2, ZnO, and LSMO; oxides which are all sensitive to the gas exhausted by internal combustion engine powered cars, which do guarantee improved sensitivity and selectivity in the shape of sandwiches as e.g. HS4 addressed by IFOX. A plurality of sensor types and electrochemical detection systems are in use for the polluting gaseous compounds exhausted by the current production internal combustion engines. In particular, the focus is placed on the detection systems dedicated to the detection of carbon mono-oxide and di-oxide (CO, CO2), and of nitrogen oxides (NOx), used to limit noxious emissions, to control the EGR flux, and to detect the combustion misfire. Nitrogen oxide gases, commonly referred to collectively as NOX, are common pollutants formed in internal combustion engines and industrial combustion systems by thermal fixation and oxidation of atmospheric nitrogen1. NOX, which includes NO2 and NO gases, have adverse effects on the environment, and are the leading cause of green-house effect, acid rain, and photochemical smog. In humans, exposure to more than 3 ppm of NO2 gas for periods longer than 8 hours can cause respiratory and cardiovascular diseases. Hence, better detection of such gases is of utmost importance, and that calls for better understanding of detection mechanism to facilitate development and optimization of sensing devices. Semiconducting metal oxides based thin films have traditionally been used in detection of NOX gases, with the common measuring arrangement relying on changes in electrical conductivity upon interaction with the gas. In particular, ZnO based gas sensors have been employed extensively in the detection of both reducing and oxidizing gases due to the superior properties of this material such as enhanced sensitivity, stability, and low cost. A widely adopted route to reduce NOx emissions is Exhaust Gas Recirculation (EGR). The EGR system has a valve which cannot respond instantly to changes in demand, and the exhaust gas takes time to flow around the EGR circuit. This makes the calibration of transient EGR behaviour particularly complex. To have a better control of EGR it would be very important to be capable of sensing the NOx content, or relatively knowing the actual O2 concentration in the exhaust air to be recirculated, and at the same time controlling precisely the quantity of gas recirculated. This in principle can be done if a gas sensor to probe precisely the quantity of O2 or the NOx within the exhausted mixture, and at the same time, a contactless magnetic sensor to instantly know the throttle-valve position which adjust the quantity of EGR. For all these reasons to have access to a multifuctional sensor capable to probe, also non simultaneously, a recirculated-gas concentration and the position of a magnetized throttle valve on move would allow to close-loop drive the EGR precisely. The IFOX HS4 is the bases of a multifunctional sensor conceived of probing precisely the quantity of NOx within the exhausted mixture, and at the same time, measuring the position of the EGR valve which adjusts the quantity of EGR. With the presence in the consortium of PST, and TSST, the supply chain of the manufacturing of the HS4 multifunctional sensor is complete, thus paving the way for a promising exploitation strategy of the result which can lead to a start of a new product to serve the worldwide automotive market as a whole if quality and cost will be sufficient over the full life cycle, thus to include to fuel consumption and emission reduction of the future vehicle employing the HS4 assisted combustion control. Although to give a honest assessment on the potential use of HS4 solution other development and analysis including the optimized models and engine control algorithms, as well as a real testing on a running vehicle are required, as well as the evaluation of pilot production of the device to allow the calculation of cost of production. It is a serious intention to plan the further understanding of these open points in a future collaborative research project, possibly in the context of the ICT theme.
Management of intellectual property
PST, has analysed more than 1000 valid patents to preview possible technological bottlenecks possibly preventing use of the know-how developed in the area of buffer layers. To this end, patents today still valid have been matter of intensive investigation, and the solutions protected compared with the new publications in order to evaluate beforehand the possible bottlenecks of the technological choices undertaken in the project, and assess better the interim results of the project against the prior-art. Particular attention was paid to the protected use of the addressed buffer layers to see the perspective of exploitation and screen possible clients of customized materials for post-project co-designing of new potential productions.
From this patent analysis the complex oxide buffer layer developed in the project is very often a prerequisite to conceive a product rather than the core of the idea. This issue opens up the perspective for the buffer material developers (in particular PST, but also TSST more specifically for oxide electronics) in the project to be part of new product supply chains. As one can see from this report, most of products employ today oxide buffer layer of extremely controlled crystallinity and quality standards, while, aside from microelectron- ics, most of them would find suitable, at a lower cost, less quality materials (sensors, dis- crete devices, coatings and electrolyte, fuel cells, etc.). From this assumption the prior ex- ploitation strategy of PST is to offer to the market buffer materials of tailored crystallinity to allow significantly decrease their cost per unit and let novel products penetrate the market.
CRF contributed 6PM on studying the IPR, and the current technologies of sensing to gauge the applicability perspective of the IFOX achievements, and in particular for the heterostructures more related to sensing functionalities, namely HS4 and HS6. A report on sensors for automotive was delivered to the commission at month 24.
CRF expects to generate better and solid IPR on the novel design based on HS4. Right now no patent has been filed yet as more development is needed to increase confidence.
“Mastering Innovation – Shaping the Future”, ESIA 2008 Competitiveness Report
ii Technology and Competitiveness (STandC) key figures 2008 (http://ec.europa.eu/research/era/)
List of Websites:
www.ifox-project.eu