REsistive Memories on FLEXible substrates – REFLEX
REsistive memories on FLEXible substrate
The project aims at the development of resistive memories on flexible substrate. With respect to standard memory technology used in RFID applications (Flash memory), resistive memories offer lower programming voltages (sometimes loxer than 1V) with a cheaper fabrication cost. Moreover the project focuses on the development of memory cells on plastic substrate allowing to bypass the use of a dedicated silicon chip embedding the memory functionnality.
Scientific goals and objectives
The main objective of this project is to develop a proof-of-concept of various small non-volatile memory arrays on plastic substrates. Chalcogenide materials will be used as functional material. Three different memory “flavors” will be targeted: conductive-bridge random access memories (CBRAM); phase-change memories (PCM); hybrid CBRAM/PCM cells potentially offering 4-electrical states.<br />Considering our choice to use chalcogenides as active layer, GeS, GeTe and Ge2Se2Te5 (GST) will be deposited by RF magnetron sputtering on metallic electrodes (soluble or inert). Then a metallic (or organic) conductive top electrode will be deposited (or printed) on the upper surface of the device. The nature of the combination electrodes/active determines the type of memory:<br />1. CBRAM devices: the memory element will rely on Metal/Insulator/Metal (MIM) capacitors consisting in a passive electrode (e.g. W, Pt, PEDOT...), a GeS amorphous layer used as solid electrolyte and a soluble electrode (Ag). In that case, GeS is chosen as the active material as a good glass former which present a better stability than GeSe glass. <br />2. PCM devices: the memory element will be based on MIM structures in which both electrodes are inert and GeTe alloys will be employed as the phase change materials. The binary GeTe is chosen since it is one of the limiting case of the pseudo binary line GeTe-Sb2Te3 on which most of the prominent PC materials are located. Moreover this system presents the advantage to be exhibit stable crystalline phase. <br />3. Hybrid CBRAM/PCM devices: the innovative memory element will use GST alloy as a phase change materials and as a solid electrolyte with one active silver electrode to mix conductive bridges. The choice of GST is motivated by the fact that the metastable crystal structures are characterized by a large amount of vacancies. This should ensure the diffusion of Ag+ species leading to the formation of the bridge, in both the amorphous and crystalline phases.<br />
The project is divided into 4 scientific tasks aiming at i) the understanding of the physical and chemical properties of the functional materials, ii) optimizing the fabrication steps of the memory devices and iii) develop 3 types of memory cells as mentioned previously. More specifically :
1)Task 1 : Sample Elaboration
The first goal of this task concerns the elaboration of samples dedicated to the physical and chemical characterization of functional materials (AFM, XRD, in-situ synchrotron measurements…). Secondly, the development of an heterogeneous process will be undertaken, combining inkjet printing and sputtering, in order to fabricate individual memory cells or arrays onto plastic substrate.
2)Task 2 : CBRAM memories (GeS)
The objective is here to develop CBRAM-type memory cells in which a metallic silver-rich filament can be created/dissolved within a solid electrolyte (GeS) in order to store two binary states. In this task, the fabricated devices will be characterized and their electrical behavior modeled.
3)Task 3 : PCM memories (GeTe)
The use of GeTe as active material will lead to the memory cells in which the reversible amorphous to crystalline phase-transition is used to store two binary states. This task focuses on the electrical characterization and modeling of this kind of memory devices.
4)Task 4 : Hybrid PCM/CBRAM memories (GST)
By combining the use of GT (acting either as a phase-change material and as solid electrolyte) and a soluble Ag electrode, this task aims at demonstrating that it is possible to exploit the interplay between phase-change and conductive-bridge mechanisms to develop advantageously a 4 states memory devices.
During the 6-month period after the beginning of the project, the following results were obtained:
1. Task 1: Sample Elaboration
The deposition process of silver layers was set-up. Thin films consisting of 50 to 200nm thick Ag layers can be deposited in view of fabricating individual hybrid or CBRAM memory devices. Mechanical adhesion tests were conducted and revealed a good adhesion of Ag layers on plastic substrate. In parallel, GST deposition process was also set-up and a first batch of test-samples was fabricated to validate the memory functionality of theses films. Finally, laser ablation test were performed on full sheet platinum layers in order to define sub-10µm thick Pt electrode that could serve as electrode in memory devices or arrays.
2. Task 2: CBRAM memories (GeS)
In the absence of sufficient guaranties concerning its mechanical properties and the stoichiometry, the GeS target was not bought yet. This buying necessitates a deeper investigation and i twill be performed between T0+6 and T0+12. In order to keep the project unaffected, task 4 was anticipated with respect to the initial schedule (T0+20); this task is dedicated to hybrid memory devices based on GST, which is available for deposition in our lab. The project should not be affected, Task 2 and 4 being relatively independent at this stage of the project
3. Task 4: Hybrid PCM/CBRAM memories (GST)
Samples consisting of Ag\GST(35nm) were studied by means of Atomic Force Microscope measurements. Beside the measurement of surface roughness of the memory stack (in the order of few nm), electrical measurements consisting of conductive-AFM measurements revealed local reversible resistive switching effects attributed to CBRAM-type memory effects.
During the following 6 months, the perspectives are described below:
1. Task 1: Sample Elaboration
Elaboration of stacks consisting of Ag(5ànm)\GST with various GST thickness in order to evaluate the impact of the functional material on the memory effect (CBRAM or PCM).
A comparison between silver layers obtained either by means of inkjet printing or sputtering will be conducted in terms of surface roughness and electrical resistivity.
Since we just received a nickel mask allowing the definition of top electrodes during deposition, individual memory devices will be fabricated in order to assess their electrical performances.
The etching of memory stacks by means of laser ablation will be studied in order to define individual lines or memory cells. The studied stack will consist of Ag\GST or Pt\GST.
After buying and receiving the GeS target, a deposition procedure will be set-up in order to fabricated CBRAM memory stacks dedicated to AFM measurements or individual memory cells for electrical characterization.
2. Task 2: CBRAM memories (GeS)
Ag\GeS stacks will be studied by means of C-AFM measurements and under probe station for layers embedding a top electrode.
3. Task 4: Hybrid PCM/CBRAM memories (GST)
Ag\GST stacks with different GST thickness will be tested by means of C-AFM. C-AFM measurements at various temperature will be conducted to evaluate the “phase-change” behavior of the fabricated devices. Individual devices will be tested under probe station to assess their electrical performances.
1) M. Putero, T. Ouled-Khachroum, M.-V. Coulet, D. Deleruyelle, E. Ziegler, C Muller: ”Evidence for correlated structural and electrical changes in chalcogenide thin films from combined X-ray techniques and sheet resistance measurements during in situ thermal annealing”, European Material Research Society (EMRS) Spring Meeting, Symposium L, May 2012, Strasbourg (France) – oral presentation
2) M. Putero, T. Ouled-Khachroum, M.-V. Coulet, D. Deleruyelle, E. Ziegler, C Muller: ”Evidence for correlated structural and electrical changes in chalcogenide thin films from combined X-ray techniques and sheet resistance measurements during in situ thermal annealing”, Material Research Society (MRS) Spring Meeting, , April 2012, San-Francisco (US) – oral presentation
3) D. Deleruyelle, M. Putero, T. Ouled-Khachroum, M. Bocquet, E. Bergeret, X. Boddaert, C. Calmes, C. Muller: ”Ge2Sb2Te5 layers used as solid electrolytes for the development of programmable metallization cells on flexible substrates”, European Material Research Society (EMRS) Spring Meeting, Symposium L, May 2012, Strasbourg (France) – oral presentation
4) D. Deleruyelle, M. Putero, T. Ouled-Khachroum, M. Bocquet, E. Bergeret, X. Boddaert, C. Calmes, C. Muller: ”Ge2Sb2Te5 layers used as solid electrolytes for the development of programmable metallization cells on flexible substrates”, Material Research Society (MRS) Spring Meeting, , April 2012, San-Francisco (US) – poster
5) T. Ouled-Khachroum, M. Putero, M.-V. Coulet, D. Deleruyelle, E. Ziegler, C Muller: ”Fabrication and Characterization of Chalcogenide-based Non Volatile Memory Devices on Flexible Substrate ”, European\Phase Change and Ovonics Symposium, July 2012, Tampere (Finland) – poster
6) D. Deleruyelle, M. Putero, T. Ouled-Khachroum, M. Bocquet, , X. Boddaert, C. Calmes, C. Muller: ” Ge2Sb2Te5 layer used as solid electrolyte in conductive-bridge memory devices fabricated on flexible substrate”, Solid-State Electronics, 2012 – accepted
The present project aims at developing a proof-of-concept of small density arrays of non-volatile memory cells on flexible substrate dedicated to future Radio Frequency Identification (RFID) applications such as “smart-tags”. The work relies on the joint collaboration of 5 researchers belonging to IM2NP laboratory (Institute of Microelectronic and Nanosciences of Provence) and covering complementary scientific fields: thin film deposition using inkjet printing or sputtering methods; fundamental material physics (e.g. thermodynamics); in-situ characterization techniques using synchrotron radiation; device physics; electrical characterization and physical modeling.
By using chalcogenide alloys as active layers, three kinds of resistive memory technologies governed by distinct physical phenomena will be fabricated on plastic substrates: conductive-bridge random access memories (so-called CBRAM); phase-change memories (so-called PCM); by combining appropriately the two latter mechanisms, mixed CBRAM/PCM memories which are expected to exhibit 4 distinct electrical states.
The nature of the electrodes and the active layer and the programming operations will directly determine the type of memory:
1. CBRAM devices: the memory element will rely on MIM (Metal/Insulator/Metal) capacitors consisting in a passive electrode (e.g. W, Pt, PEDOT...), GeS chalcogenide alloy used as solid electrolyte and a soluble electrode (Ag). In that case, GeS active material is preferred since it is a good glass former. Bipolar operations are required.
2. PCM devices: the memory element will be based on MIM structures in which both electrodes are inert and GeTe alloys will be employed as phase change layers. Mimicking silicon-based solid state devices, unipolar operations are required.
3. Hybrid CBRAM/PCM devices: this innovative memory element will use GST alloy as solid electrolyte together with one active silver electrode to mix conductive bridges and phase change mechanisms. Both unipolar and bipolar operations are necessary.
Physical and microstructural analyses as well as extensive electrical characterization of the fabricated devices will be carried out to get insights on the physical mechanisms occurring at nanoscale as well as on memory devices. From obtained results, comprehensive physical models will be proposed and subsequently used to optimize the fabrication of memory cells.
Monsieur Damien DELERUYELLE (CENTRE NATIONAL DE LA RECHERCHE SCIENTIFIQUE - DELEGATION REGIONALE PROVENCE ET CORSE) – firstname.lastname@example.org
The author of this summary is the project coordinator, who is responsible for the content of this summary. The ANR declines any responsibility as for its contents.
CNRS DR 12 - IM2NP CENTRE NATIONAL DE LA RECHERCHE SCIENTIFIQUE - DELEGATION REGIONALE PROVENCE ET CORSE
Help of the ANR 179,679 euros
Beginning and duration of the scientific project: December 2011 - 36 Months