DE102016003770A1 - Method for producing a memory, memory, and use of the memory - Google Patents

Abstract

The invention relates to a process for producing an electrochemical storage, characterized by the steps of: a) providing a non-conductive substrate; b) a first conductor of conductive material is disposed on the substrate; c) a porous dielectric with redox-active molecules is punctiformly arranged on the first conductor track; d) a second interconnect is arranged orthogonally to the first interconnect, wherein at the intersection point the interconnects have an electrode function between which the dielectric is arranged; e) on the substrate, the first interconnect, the dielectric and the second interconnect, a passivation layer is arranged, wherein the first and the second interconnect at its intersection with the interposed dielectric form a memory in which by applying voltage across the interconnects the Redox reaction of the redox-active molecules at the electrodes is driven to produce a bit. Various memories and their uses are disclosed.

Description

The invention relates to a method for producing a memory, a memory and the use of the memory.

State of the art

Non-volatile memory techniques are widely known in the art, which are widely used in information processing, electronic equipment, or for product labeling.

There is a commercially very widespread 1-bit ROM / WORM memory, which is established in the form of an LC resonant circuit. The resonant circuit has a natural frequency determined by the component parameters. A test wave can determine a yes / no information by the energy absorption of the resonant circuit. Such tags are very inexpensive and can be easily eliminated or destroyed by applying a strong magnetic field (1 bit ROM / WORM memory).

However, the storage techniques for more complex information processing electronic equipment rely heavily on the use of silicon (CMOS) technology to produce these memories. Non-volatile memories currently in use use solid-state physical properties in order to be able to store information and to be able to store it for long periods (> 10 years) without external power supply.

EPROM / EEPROM / Flash EPROM:

A class of memory devices based on silicon semiconductor technology. The basic element of EPROMs / EEPROMs and Flash EPROMs is a MOSFET (field effect transistor) with a gate electrode insulated on all sides. The insulation is realized by a layer of SiO ₂ . A charge applied to the gate electrode is permanently held via the insulating SiO ₂ acting as a dielectric and optionally blocks the transistor.

EPROM: This memory type is electrically writable and readable. Before an electrical re-description can take place, an irradiation of the memory chip with UV must take place. This basic type is realized by the FAMOS (floating gate avalanche injection MOS) cell using a p-channel MOSFET or an n-channel MOSFET having a control gate next to a floating gate. The description is made by hot carrier injection (HCl), which exploits that electrons with sufficient kinetic energy can break a potential barrier and are stored on the gate.

The memory cell must be UV-irradiated for a re-description. Photons with sufficient kinetic energy, depending on the gate material, excite the electrons of the floating gate so that they can leave the gate. The erasure time is of the order of 10 minutes and requires the mechanical shifting of a glare protection of the memory cell with quartz glass window, which must be covered in the normal operation / read mode.

The write power is ~ 1 μJ / bit. The writing time is ~ 3-50 ms, + 10 minutes for re-description. The operating voltage is 3-6 V and the writing voltage is ~ 12 V.

EEPROM (also E ² PROM):

The Electrically Erasable PROM (EEPROM) is a memory cell that can be both electrically read and electrically written. The charge applied to the gate is achieved by the quantum physical tunnel effect. The write / erase mode therefore runs through field emission (Fowler-Nordheim Tunneling) under application of a high electric field and is done bit by bit.

Typical storage capacities are 1-4 Mbit. The write power is ~ 1 μJ / bit. The write time is ~ 1-10 ms and the operating voltage: 3-6 V. Market leader for RFID EEPROM is Atmel, USA. The distribution runs z. B. on the commercial product ATA5575M1 16 byte EEPROM microcontroller.

Flash EPROM (also called "flash memory"):

The structure of this memory cell is based on the EPROM technology. In this type of memory, the charge is applied to the gate via the hot carrier injection process and the write / erase process is accomplished via Fowler-Nordheim tunneling (as in EEPROM). Large amounts of memory cells are arranged as a matrix and are widely commercialized in two forms of NAND flash and NOR flash architecture. Due to the matrix architecture and the use of common branches for source / drain, complete switching of individual bits is not possible. In particular, the deletion of a state can only be done in blocks.

The typical storage capacities are 1 MBit-128 GBit. The write power is ~ 1 μJ / bit, the write time is ~ 0.1-1 ms. The operating voltage is 3-6 V, 12 V.

MLC NAND (Multi Level Cell Flash) Memory:

This type of memory is based on the flash memory, which can store only 1 bit per cell, and can store several bits per cell. The higher storage density is achieved by differentiating several charge states per transistor. The MLC memories today reach up to 4-bit, with 16 charge states being distinguished. Typical feature sizes for a transistor with this type of memory are 10-20 nm. Market leaders in this segment are SanDisk and Samsung.

Ferroelectric RAM (FRAM or FeRAM):

FRAM is also based on silicon semiconductor technology. On the gate of a field-effect transistor, a ferroelectric crystal is applied, which can be polarized by applying a field and switches the transistor. The writing power is ~ 0.1 μJ / bit. The write time is ~ 1-10 μs. The operating voltage is 1-3 V. Market leader for RFID FRAM is Fujitsu, Japan.

For wireless readable memories, which are established as RW RFID / NFC tag and have more than> 1 bit memory, especially EEPROM and FRAM memory are used (market leader NXP).

There are few commercial products for printed RFID / NFC tags. The best known approach is that of Thin Film Electronics ASA. There, a printed, non-volatile memory based on ferroelectric polymers is used. In these memory cells, the ferroelectric polymer is inserted between two electrodes in a passive matrix. Thus, a ferroelectric capacitor defining a memory cell is formed at each intersection of the conductive metallic contact lines.

There are known electrochemical storage. So z. As an electrochemical metallization memory (ECM) or generally Memristor devices. These are usually based on an electrochemical process as described by Valov et al. is known ( Ilia Valov, Rainer Waser, John R. Jameson, and Michael N Kozicki, 2011. Electrochemical metallization memories: fundamentals, applications, prospects. Nanotechnology 22, 254003 (22pp) doi: 10.1088 / 0957-4484 / 22/25/254003 ).

The principle of operation of the ECM is that of an electrochemical memristor. A memristor is a passive electronic device whose electrical resistance is not constant but depends on its past. The current resistance of this device depends on how many charges have flowed in which direction in the past. Thus, the resistance value over the time course of the current is adjustable. This resistance is permanently maintained even without energy. Between the electrodes of a memory cell is an electrolyte of metal salts or metal oxides, sometimes liquid but mostly solid. When a reducing potential is applied between the electrodes, metal cations reduce to elemental metal and form fibers, called filaments, from the working electrode to the counter electrode, which establish a conductive connection between the electrodes. Upon application of the reverse oxidizing potential, the metals oxidize and redissolve these fibers. If both electrodes are bridged by a conductive fiber, the current can flow with little resistance and the memory cell has the state "1". If the electrode in the cell is not connected by fibers, the resistance is large and the memory cell is at "0".

The prior art has the following disadvantages:
The disadvantage of LC resonant circuits is the limitation of the storage capacity to 1 bit.

EEPROM and FRAM are both expensive to manufacture due to the silicon semiconductor technology required (eg, ~ 0.5-1.0 € per RFID tag). In addition, the energy required for switching is very high, especially with the EEPROM. Since when used as RFID only about 1/1000 of the energy used for reading can be recorded by the tag antenna, the range and conductivity is limited by the energy requirements of the memory cell.

By contrast, FRAM is more energy-efficient, but due to the more complex production and the required ferroelectric crystal structure, it is still considerably more expensive than EEPROM. These methods are economically viable, if at all, only if a very high integration density and high absolute storage quantities can be produced. Storage volumes in the range of a few bits up to the megabit range are not economically feasible with this technology, since the basic costs are already very high with this technology.

Electrochemical metallization memories are produced by photolithographic micro- and nanostructuring methods. These methods are also per se consuming and therefore expensive.

Disadvantages are the previously published methods for producing non-volatile memory consuming and thus time and / or costly.

Object of the invention

The object of the invention is to provide a novel memory which does not have the disadvantages of the prior art. Furthermore, it is an object of the invention to provide a method for producing the memory and to indicate uses therefor.

Solution of the task

The object is achieved with the method according to claim 1 and the memory and the use of the memory according to the independent claim. Advantageous embodiments of this result in each case from the back-related claims.

Description of the invention

• a) ein nicht leitendes Substrat wird bereit gestellt;
• b) eine erste Leiterbahn aus leitendem Material wird auf dem Substrat angeordnet;
• c) auf die erste Leiterbahn wird punktförmig ein poröses Dielektrikum mit einer bestimmten Konzentration der redoxaktiven Molekülen angeordnet, wobei in manchen Fällen die Konzentration der redoxaktiven Molekülen Null aufweisen soll;
• d) orthogonal oder mit einem Winkel von 5–90 Grad zur ersten Leiterbahn wird eine zweite Leiterbahn angeordnet, wobei am Kreuzungspunkt die Leiterbahnen eine Elektrodenfunktion aufweisen, zwischen denen das Dielektrikum angeordnet wird; dies erlaubt die spannungsgetriebene Redoxreaktion der redoxaktiven Moleküle an den Elektroden.
• e) auf das Substrat, die erste Leiterbahn, das Dielektrikum und die zweite Leiterbahn wird eine Passivierungsschicht angeordnet.

The inventive method for producing an electrochemical storage is characterized by the steps:

• a) a non-conductive substrate is provided;
• b) a first conductor of conductive material is disposed on the substrate;
• c) a porous dielectric with a specific concentration of the redox-active molecules is punctiformly arranged on the first conductor track, in which case the concentration of the redox-active molecules should in some cases be zero;
• d) orthogonal or at an angle of 5-90 degrees to the first trace a second trace is arranged, wherein at the crossing point the traces have an electrode function between which the dielectric is arranged; this allows the voltage-driven redox reaction of the redox-active molecules at the electrodes.
• e) a passivation layer is arranged on the substrate, the first interconnect, the dielectric and the second interconnect.

Additional passivation layers for passivation of the printed conductors can be arranged. With the exception of the dielectric z. B. the structure in the connection of step c) are optionally passivated by arranging a passivation layer.

Unlike redox-active sensors, there are no accesses to the electrodes with the exception of the contact points at the end of the tracks, but it is a completely closed system without leakage currents and / or evaporation of the redox-active molecules.

According to the invention, the first and the second conductor track form a non-volatile 1-bit memory at their point of intersection with the dielectric arranged therebetween, in which the redox reaction of the redox-active molecules at the electrodes is driven by applying voltage across the conductor tracks.

As the electrically non-conductive substrate, glass, silica, or polymers, etc. can be used.

As a material for the first conductor or electrode z. Gold, platinum, carbon, conductive polymers and so on.

It can be arranged a plurality of interconnects preferably orthogonal to each other. Then advantageously a "crossbar" structure is provided.

The first conductive trace (s) are preferably applied by ink jet printing or other printing in a finished pattern. This preferably reduces the cost of manufacturing the memory.

A printing process, especially digital printing methods such as ink jet printing or aerosol jet printing, is advantageous fast, inexpensive, and very easy to reproduce. In addition, the digital method allows an arrangement of different inks, or inks with different concentrations of the redox molecules, punctiform on the cross points, which is particularly advantageous for the memory.

The method for producing the redox cycling memory thus comprises in particular, but not exclusively, a choice of conductive and / or insulating, in particular printable, particles with which the printed conductors or electrodes and / or the dielectric can be arranged in a structured manner one above the other. In particular, but not exclusively, ink jet, aerosol jet, screen, gravure, offset, microcontact, nanoimprint or hot stamp are used. Combinations of coating and ablation steps which apply the same layers in combination with various coating methods such as e.g. As slot-die, laser ablation and so on. In particular, only inkjet printing is used, as it is a particularly inexpensive method.

The first trace has either no or very small pores compared to the dielectric to be placed above it. The conductor must be correspondingly conductive, chemically inert and have low charge transfer resistance and thus good electrochemical properties. This advantageously has the effect that trouble-free cyclic voltammograms can be measured with standard redox molecules.

In the case of inkjet printing, the ink should preferably be sintered after printing on the substrate. For this purpose, thermal, photonic, UV or other sintering methods can be carried out so that a homogeneous conductive layer is formed for the first printed conductor.

Subsequently, a dielectric ink is printed on the future crossing point. The ink used is preferably a sol-gel or hydrogel, and optionally the ink already contains redox molecules as the active material and later forms the nanoporous dielectric.

The ink may additionally include other additives such as soft polymers for increased elasticity, as well as fabrics which improve the printing properties of the inks.

The ink should cure before the next conductive top trace or electrode is printed but not dry out. In the case of sol-gel based ink, this means that the condensation reaction of the sol gel should take place. The pores formed by the sol-gel reaction are smaller in one embodiment of the invention than the particles for the conductive ink of the upper trace or electrode.

Following this, the second trace is preferably printed from a conductive ink orthogonal to the lower trace, so that at the intersection of the upper trace and the lower trace the traces can perform electrode function. In crossbar structures, a multiplicity of second conductor tracks are preferably arranged orthogonally to the first conductor tracks.

At the respective crossing points of the lower, first interconnects to the upper, second interconnects, the nanoporous dielectric layer is embedded in a "sandwich" structure, which thus forms the memory matrix.

In the case of ink-jet printing, the ink for the second conductive trace (s) is preferably comprised of nanoparticles of gold, platinum, carbon, conductive polymers, and so forth. The nanoparticle size for The ink of the second trace should then be no smaller than the pores in the underlying dielectric layer, so that the nanoparticles for building the second trace can not penetrate into the intermediate layer, the dielectric and cause a short circuit between the electrodes. Also, this ink should preferably be sintered, for. Thermally, photonically, by UV or other method to form a conductive web.

Optionally, the invention uses the high resolution of the Z-axis printing process to produce electrochemical nanostructures.

Preferably, the inventive method for producing a printed or redox cycling cell is transferred to a cross-bar architecture and extended to this, so that not only 1 × 1 conductor and cell, but z. B. 2 × 2 or z. B. 4 × 4, tracks and a corresponding number of redox cycling cells are arranged side by side.

In these configurations, at the crossing points, the surface areas of the first conductive lines serve as lower electrodes for a plurality of redox cycling cells simultaneously. Furthermore, at the crossing points, the surface regions of the second interconnects serve as upper electrodes for the necessary redox cycling process.

• a) Auf dem Substrat wird eine erste, elektrisch leitfähige Elektrode angeordnet,
• b) Auf die erste Elektrode wird eine poröse dielektrische Schicht angeordnet, bei der die Poren bis auf die Oberfläche der ersten Elektrode führen,
• c) Auf die dielektrische Schicht wird eine zweite, elektrisch leitfähige Elektrode angeordnet, wobei mindestens einer der Schritte a) bis c) mit einem Druckverfahren von elektrisch leitfähigen und/oder elektrisch isolierenden Partikeln durchgeführt wird,
• d) Passivierung der Schicht-Struktur.

The structure of individual redox cycling cells or memory cells by means of printing z. Example by means of an ink jet printing process can thus have at least the following steps:

• a) a first, electrically conductive electrode is arranged on the substrate,
• b) a porous dielectric layer is arranged on the first electrode, in which the pores lead to the surface of the first electrode,
• c) a second, electrically conductive electrode is arranged on the dielectric layer, wherein at least one of the steps a) to c) is carried out with a printing method of electrically conductive and / or electrically insulating particles,
• d) Passivation of the layer structure.

It is understood that the electrodes are made contactable by being contacted over the free ends of the conductor paths which can be contacted in this way.

Thus, at a point of intersection, at least at the first electrode, a porous dielectric layer is arranged at a point of intersection, in which the pores of the dielectric extend to the surface of the first electrode. The pores in the dielectric layer are filled with redox-active molecules. Preferably, the redox-active molecules are already contained in the ink.

The nanoscale redox cycling memory is preferably made only by printing technologies and without additional etching steps or sacrificial layers. In this design, the Z-axis electrodes, which are separated by a nano-scaled dielectric, are preferably completely printed.

There are advantageously no etching steps in this method. This is achieved by virtue of the fact that the three layers, the first first lower interconnect, the second nanoporous dielectric layer and the third, second interconnect preferably have different porosity.

The ink of each further layer should have a larger particle size than the previous lower layer, so that the layers during printing in the liquid phase, for. B. by ink jet printing, can not flow into the underlying layer. This makes it possible to reliably electrically disconnect the various layers and to prevent short circuits through bridges of conductive material flowing through the dielectric separation layer.

Advantageously, the ink for the porous dielectric layer is prepared so that after its deposition on the first conductor, z. B. by means of an ink jet, cures and has the desired porosity, but at the same time still contains enough liquid to ensure transport of the redox molecules to the electrodes.

This effect can be achieved according to the invention by the use of sol-gel materials and a sol-gel ink with a solvent having a low vapor pressure or low vapor pressure or high boiling point.

Preferably, the ink for this z. As tetramethyl orthosilicate (TMOS), tetraethyl orthosilicate (TEOS), tetraisopropyl orthosilicate (TPOS) as a silicate for the sol-gel. The ink can also be mixed with other materials such. For example, with aluminum (2-propylate), aluminum (2-butylate), zirconium propylate, titanium ethylate, titanium (2-propylate) or the like can be prepared. In general, for a sol-gel ink for the nanoporous dielectric, an alkoxide of an element should be used selected from the group of silicon, titanium, aluminum, zirconium, germanium, tin, lead and antimony.

Another aspect of making this ink involves mixing the alkoxide solution with an alcohol, deionized water and an acid catalyst and / or base catalyst with the redox-active molecule to make the ink for the nanoporous dielectric.

The ink for the nanoporous dielectric but z. B. include hydrogel as a dielectric nanoporous material, or the components for the formation of hydrogels.

As a solvent for the remaining after drying of the nanoporous dielectric liquid in particular but not exclusively glycerol, different glycols such. As ethylene glycol, diethylene glycol, triethylene glycol, propylene glycol, dipropylene glycol, tripropylene glycol, and so on, as well as polyglycols with different chain lengths z. With up to 200, 300, 400, or more monomer units. It is therefore conceivable to use polyethylene glycol, polypropylene glycol and so on.

In addition, as noted, the ink may optionally include a redox active molecule depending on the type and location of the storage.

As redox-active molecules come z. B. hexacyanoferrate, iridium hexachloride, ferrocene and their derivatives such. As ferrocene methanol, ferrocene dimethanol, ferrocenecarboxylic acid, ferrocenedicarboxylic acid, quinones and their derivatives such. Orthoquinones, dopamine, benzoquinone and so on, as well as other redox-active molecules.

After the deposition of the second trace as the upper electrode over the dielectric, the redox molecules in the remaining liquid of the dielectric nanoporous layer should diffuse freely and reach the electrodes for the redox reaction.

• 1.) Sie entkoppelt die beiden Elektroden, das heißt die untere und die obere, mit einer sehr dünnen, in der Regel kleiner als 1000 nm, am besten 100–300 nm dicken, dielektrischen Schicht elektrisch voneinander.
• 2.) Das Dielektrikum erlaubt im fertig ausgebildeten Zustand die Diffusion der elektrochemisch aktiven Moleküle (Redox-Moleküle) an die Elektroden, damit ein Redox-Cycling-Prozess zwischen der unteren und der oberen Elektrode stattfinden kann.

The dielectric nanoporous layer performs at least two functions:

• 1.) It decouples the two electrodes, that is, the lower and the upper, with a very thin, usually smaller than 1000 nm, preferably 100-300 nm thick dielectric layer electrically from each other.
• 2.) The dielectric in the finished state allows the diffusion of the electrochemically active molecules (redox molecules) to the electrodes, so that a redox cycling process can take place between the lower and the upper electrode.

All materials meeting these two requirements can be used for the nanoporous dielectric layer, in particular so-called sol-gel materials or hydrogels, but also other materials. The sol-gel formation of a silica gel on the point-like applied dielectric follows the steps described in the publication "The Sol-Gel Preparation of Silica Gels" ( Buckley, AM, Greenblatt, M. 1994. Journal of Chemical Education. Volume 71, no. 7, 599-602 ) and the contents thereof, in particular for the preparation of the sol gel, are hereby incorporated by reference into this patent application.

In a further step, the memory structure consisting of a large number of memory cells is coated with a non-porous dielectric, ie completely passivated. For this purpose, z. B. a lamination process can be performed. Only the contact points of the conductor tracks for the electrical contacting and the application of voltage remain free.

A passivation layer having a non-porous, and / or non-vapor and / or gas-permeable layer advantageously has the effect that the liquid, if present, does not evaporate in the dielectric nanoporous layer. In addition, the passivation layer as a protective layer ensures greater mechanical stability of the structures.

It is advantageously possible to extend the cell array also in height to increase the number of cells per area and thus the storage density. In such a design, another dielectric is deposited on top of the top electrode, followed by a second top electrode in the next step.

This configuration can be further extended in the Z direction, and is limited only by the problem of contacting circuit traces for memory control. The interconnects can also be linked in the Z direction.

In a two-level redox-cycling architecture, the "first upper electrode" is then used as the lower electrode for the upper cell, while the "second upper electrode" acts as the upper electrode for this cell.

During redox cycling, the electrochemically active molecules, the redox molecules, can be alternately oxidized and reduced in the liquid volume of the dielectric nanoporous intermediate layer of the dielectric at the electrodes. These reactions therefore take place between the electrodes. For this purpose, corresponding oxidizing and reducing potentials are applied to the electrodes so that the redox molecules are directly oxidized or reduced when the electrodes are touched. By allowing each molecule to be alternately oxidized at one electrode and then reduced again at the other electrode, each molecule contributes multiply to the charge transport between the electrodes, which in turn leads to an increase in the total current.

For certain applications (eg for WORM - write once read many memory) irreversible oxidation states can also be sought. In this case the reverse reaction is not possible. For example, the molecules of the Viologen family advantageously have such properties. They usually occur in three oxidation states in which the second oxidation reaction is usually irreversible: V ²⁺ ↔ V ⁺ → V ⁰

All molecules that follow this reaction can basically be used for WORM memory cells.

The movement of the redox molecules in the liquid medium of the dielectric between the electrodes is driven by the diffusion in the nanoporous dielectric. In a diffusion-driven movement, the transport time scales with the square of the distance, therefore, for the realization of such a memory, a special arrangement of the electrodes is necessary, in which the electrodes are very close to each other. The efficiency of the redox cycling or the amplification of the signal depends on the square of the distance between the electrodes.

For the electrodes, a distance in the nanometer to micrometer range is ideally used, which allows a large current gain. The absolute redox cycling current scales linearly with the concentration of the redox molecules.

• – nicht-volatiler Speicher;
• – niedrige Produktionskosten (< 1 Cent) durch Druckverfahren (Rolle zu Rolle);
• – energiesparende (fJ – nJ je Vorgang) Auslese- und Schreibvorgänge;
• – Verschiedene Konfigurationen in Abhängigkeit vom verwendeten redoxaktiven Molekül: ”read-only memory” (ROM), ”write once read many memory” (WORM) und ”re-writable memory” (RW)
• – Typische absolute Speicherkapazitäten von wenigen Bit bis zu mehreren KBytes;
• – Sehr große Varianz in der Realisierung der Speicherdichte möglich in Abhängigkeit von der Herstellungstechnik und des Ausleseverfahrens;
• – Möglichkeit von Multi-Level-Kodierung, das heißt Speicherzellen, bei denen nicht nur zwei Zustände ”0” und ”1” aufgeschrieben werden können, sondern mehr, wie z. B. ”0”, ”1”, ”2” und ”3” und so weiter. Diese können durch das Herstellungs- und/oder Ausleseverfahren erreicht werden. Unter Schreiben versteht man erfindungsgemäß die Aktivierung/Deaktivierung des Moleküls um es entsprechend redoxaktiv/redoxinaktiv zu machen. Beispielweise kann man dies durch die Änderung des Oxidationszustands des Moleküls wie beim Viologen (V²⁺ ↔ V⁺ → V⁰) erreichen.

In particular, a printed redox cycling based electrochemical storage device acting as a non-volatile memory is provided in this way. Particular features that this device has are:

• - non-volatile memory;
• - low production costs (<1 cent) through printing (roll to roll);
• - energy saving (fJ - nJ per process) read and write operations;
• - Different configurations depending on the used redox-active molecule: "read-only memory" (ROM), "write once read many memory" (WORM) and "re-writable memory" (RW)
• - Typical absolute storage capacities from a few bits to several KBytes;
• - Very large variance in the realization of the storage density possible depending on the manufacturing technique and the readout method;
• - Possibility of multi-level coding, that is memory cells in which not only two states "0" and "1" can be written down, but more, such as. "0", "1", "2" and "3" and so on. These can be achieved by the manufacturing and / or Ausleseverfahren. Writing is understood according to the invention to be the activation / deactivation of the molecule in order to make it correspondingly redox-active / redoxin-active. For example, this can be achieved by changing the oxidation state of the molecule as in the viologen (V ²⁺ ↔ V ⁺ → V ⁰ ).

• – Typische erreichbare Speicherdichten des Sensors (ohne Zuleitung): 10–10⁵ pro Bit/mm² in Abhängigkeit vom Herstellungsverfahren. Bei multi-level Speicherzellen ist die Speicherdichte deutlich höher.
• – Kompatibilität mit Vorrichtungen und Protokollen für RFID/NFC und anderen drahtlosen Kommunikationssystemen, die im Internet-of-Things (IoT) benutzt werden;
• – Kompatibilität mit Vorrichtungen und Prozessen der gedruckter Elektronik für Anwendungen in IoT;
• – mögliche Nutzung nicht-toxischer/umweltbelastender Materialien für die Speicherherstellung (insbesondere für den Einsatz in Lebensmittelverpackungen);
• – Redox-cycling Verfahren ohne Referenzelektrode, da abgeschlossenes Speichersystem.

But it can also structural changes of the molecule, which are caused by the oxidation / reduction, such as. B. in different ferrocene derivatives (A ↔ B ↔ C) can be achieved. To summarize all these possible changes and to avoid charge balance problems in the formulas, the molecules are referred to schematically as A ↔ B, B ↔ C, A ↔ D, and so on.

• - Typical achievable storage densities of the sensor (without supply line): 10-10 ⁵ per bit / mm ² depending on the manufacturing process. With multi-level memory cells, the storage density is significantly higher.
• - Compatibility with devices and protocols for RFID / NFC and other wireless communication systems used in the Internet of Things (IoT);
• Compatibility with printed electronics devices and processes for applications in IoT;
• - possible use of non-toxic / polluting materials for storage production (especially for use in food packaging);
• - Redox-cycling process without reference electrode, since closed storage system.

The method for producing the memory cell (s) is characterized in particular by the fact that the first printed conductor and / or the dielectric and / or the second printed conductor can be provided by a printing method, in particular by inkjet printing. This is advantageously cheaper than silicon fabrication technologies.

The method may be carried out by selecting an ink for the dielectric which, after being applied to the first trace, is dried and forms a sol-gel or hydrogel with pores. This represents, in particular after a printing process, a comparatively simple possibility to define the nanoporous dielectric and to provide it with a certain size of pores. For this purpose, a method can be used as z. As described in The Sol-Gel Preparation of Silica Gels ( Buckley, AM, Greenblatt, M. 1994. Jouranl of chemical education. Volume 71, no. 7, 599-602 ). This publication and the process for preparing the sol gels is incorporated by reference in full in the present patent application.

According to the invention, an ink can be selected which already comprises, in addition to the components for forming the gels, also the redox-active molecules which diffuse to the electrodes in the pores of the dielectric after the pressure or the formation of the nanoporous dielectric and react electrochemically thereon. This advantageously represents a significant simplification of the process.

It is understood that the method steps according to claim 1 can be repeated. This advantageously has the effect that the steps for forming a multiplicity of interconnects arranged orthogonally relative to one another in a memory array can be used. This advantageously causes a "crossbar" structure to be provided.

By means of the "crossbar" architecture of several redox cycling memory cells at the crossing points, it is advantageously effected that for every 2n contact points, n.sub.2 redox cycling cells and correspondingly n.sup.2 bits are generated. This leads to the highest possible two-dimensional packing density.

It is important to understand that even without redox cycling reaction in this type of storage electrochemical currents can come to pass. If appropriate read-out potentials (cathodic read potential _red and anodic read potential _Ox ) are applied to two orthogonally lying printed conductors in order to read out the state of the bit, an electrochemical current is generated at all points of intersection with redox-active molecules on the printed conductors. Since, however, only one electrode is connected to one of the two readout potentials (Ox or Red) at all bits except for the intersection point of the interconnects, only a partial reaction, ie A → B, or B → A, is possible in the other bits. The current is thus not amplified and thus has a significantly lower value than for the bit at the intersection point of both addressed interconnects, where the back and the backward reaction is possible.

The threshold value to be exceeded, at which a measured current is evaluated by redox cycling amplification of the reaction A ↔ B and thus as state 1, according to the invention is preferably at least three times as high as the unamplified noise (current noise). This definition applies to all memory cells ROM / WORM / RW mentioned.

These larger differences in current at the crossing point to all other memory cells of the array results in a high signal-to-noise ratio. This signal-to-noise ratio enables redox cycling based electrochemical storage. The state of a bit as "1" is then linked to a certain minimum current which must be reached when the read-out potentials are applied. Of course, this minimum current must be defined to be larger than the background current due to the half-reaction to the unaddressed bits on the respective printed conductors.

The method can further be used in the method by the choice of ink with different concentrations and / or substances on redox-active molecules for the storage array following advantages:
Thus, it is possible to realize a memory cell with multi-level coding, in which besides two binary states "0" (no redox molecules) and "1" (present redox molecules) also other states are present. This is z. B. realized by the deposition of inks with different concentrations of redox molecules in the memory cells.

Example: In the readout method, 16 different concentrations, eg 0 μM, 1 μM, 5 μM, 10 μM, 25 μM, 50 μM, 100 μM, 200 μM, 300 μM, 400 μM, 500 μM, 600 μM, 700 μM, 800 μM, 900 μM, 1000 μM distinguished from the redox-active molecule ferrocendimethanol. As a result, advantageously 16 different states can be written into each individual memory cell. It can then for the memory instead of a binary base z. For example, a hexadecimal system may be used and units of the hexadecimal 16-base (HEX) system written: 0, 1, 2, 3, 4, 5, 6, 7, 8, 9, A, B, C, D, E , F, short 0 to F.

This advantageously increases the storage density, that is to say the possible type of memory state per physical area. Thus, these memory cells with multi-level coding offer much higher capacities, although the number of storage nodes remains the same. For example, you usually put eight bits together in one byte. For binary ("0" or "1") bits, one byte corresponds at most to the number Sum (n = from 0 to 7) 2 ^ n = 2 ^ 0 + 2 ^ 1 + 2 ^ 2 + 2 ^ 3 + 2 ^ 4 + 2 ^ 5 + 2 ^ 6 + 2 ^ 7 = 255. That means you can store numbers from 0 to 255 with a binary byte.

und so weiter. Tabelle 2: Hexadezimales System

und so weiter.According to the invention, by comparison with the use of hexadecimal bytes ("A" to "9"), the maximum representable number in a byte number Sum (n = from 0 to 7) 16 ^ n = 16 ^ 0 + 16 ^ 1 + 16 ^ 2 + 16 ^ 3 + 16 ^ 4 + 16 ^ 5 + 16 ^ 6 + 16 ^ 7 = 286,331,153, which means you can store the numbers from 0 to 286,331,153. This leads according to the invention to a significantly higher storage density than in conventional storage systems. The differences between binary and hexadecimal storage cell systems can be seen in Table 1 (binary) and 2 (HEX): Table 1: Binary system

and so on. Table 2: Hexadecimal system

and so on.

According to the invention, both storage systems and other systems are realized.

An inventive electrochemical memory or such a memory cell initially comprises at least one arrangement of a first electrically contactable conductor on a non-conductive substrate, a orthogonal thereto arranged second electrically contactable conductor, wherein the conductors have at the intersection point electrode function and wherein at the intersection between the two Conductors a porous dielectric is arranged with freely diffusible in the pores redox-active molecules. These redox-active molecules are oxidized and reduced at the electrodes of the interconnects by applying electrical voltage (read-out potentials Ox and Red) to generate a defined state or to read out the state, whereby the memory of substrate, interconnects and dielectric is completely passivated by a passivation layer. There are no separate leads or derivatives for redox-active molecules, so it is a non-volatile memory, more specifically a single memory cell thereof. Writes for WORM and RW memory cells are also possible, as will be shown.

A memory array according to the invention has a multiplicity of such passivated electrochemical memory cells) in "crossbar" configuration, produced by the method described above.

Preferably, in one embodiment of the invention, a memory array is characterized by a plurality of memory cells. A memory array according to the invention is characterized in particular by the fact that the memories or the individual memory cells have dielectrics each having different concentrations of redox-active molecules and / or different and / or no redox-active molecules. This advantageously develops memory cells with a multi-level coding from the proposed device.

The configuration with multiple types of redox molecules can also be advantageously combined with the configuration of multiple concentrations of the redox-active molecules to produce even more complex memory cells with multi-level coding.

This advantageously provides for a non-volatile memory to be provided by a printed device based on redox cycling and crossbar architecture. The memory is for this purpose made with dielectric nanoporous ink comprising redox molecules.

The storage array can comprise at least one storage state without a redox-active molecule for forming state 0 and at least one storage state with a redox-active molecule, wherein the redox-active molecule is reversibly oxidized by applying voltage (read potential Ox and Red) via the strip conductors to the electrodes forms a reduced form for the formation of state 1.

In the following Tables 3 to 5, the symbol question mark ("?") Represents an unknown memory cell value. GND = ground; Read potential = potential at the cathode and at the anode, which is sufficient only for driving the reversible redox reaction A ↔ B for reading out the memory cell value or state "1", but not for generating write operations; Write potential = potential at the cathode and at the anode, which is used to drive a reversible or irreversible redox reaction to rewrite the redox-active molecules A and / or B by oxidation and / or reduction and generation of the memory cell value or state "0" (binary) or a Intermediate value (hex or other system) is sufficient. In the case of WORM, a write potential directionally produces a state of progressively lower concentrations of [A and B] in the memory cell, in the case of RW the concentration of [A and B] is renewable.

ROME

• 1.a. ROM: A ROM memory cell is advantageously characterized in that it comprises a nanoporous dielectric either without or with a redox-active molecule. In the ROM with a redox-active molecule, a redox reaction takes place, reversibly switching the redox-active molecule between its two oxidation states, the oxidized state and the reduced state.

The following modes for writing and reading are possible, see Tables 3a, 3b and 1a with corresponding cyclic voltammogram. Table 3a: ROM - Binary system Value (binary) 0 1 Current value (discrete) 1 2 Concentration [A + B] (discrete) 0 1 mode Redox molecule concentration (discrete) Potential cathode Potential anode Current value (discrete) Memory cell data standstill [A + B] = 0 none none none ? select [A + B] = 0 V _{readout (red)} V _{readout (ox)} 1 "0" standstill [A + B] ≠ 0 none none none ? select [A + B] ≠ 0 V _{readout (red)} V _{readout (ox)} 2 "1" Write not possible

• 1.b. Multi-level ROM: wie 1.a. aber mit mehreren Konzentrationen von Redox-Molekülen, z. B. mit 16 verschiedenen Konzentrationen von z. B. 0 μM, 1 μM, 5 μM, 10 μM, 25 μM, 50 μM, 100 μM, 200 μM, 300 μM, 400 μM, 500 μM, 600 μM, 700 μM, 800 μM, 900 μM, 1000 μM an z. B. Ferrocendimethanol als redoxaktives Molekül für die Erzeugung von hexadezimalen Speicherzellen im Speicherarray.

Thus, for a binary ROM configuration, two different types of dielectric nanoporous layer are deposited at the crossing points: one with redox-active molecules (so-called state "1") and one without redox-active molecules (state "0"). If oxidation and reduction potentials exceeding the normal potential of the redox-active molecule are applied to the lower and the upper electrode via read-out potentials, redox-cycling will take place in the cells with redox-active molecules, whereas in the cell without redox-active molecules only Background noise occurs. By reading out the current signal of the electrodes, one can correspondingly detect "1" and "0" states of the memory cell. The redox-active molecule should accordingly have the lowest possible reversible oxidation and reduction potential in order to reduce the readout energy. For ROM, only the following reaction takes place: A ↔ B

• 1.b. Multi-level ROM: as 1.a. but with multiple concentrations of redox molecules, e.g. B. with 16 different concentrations of z. 0 μM, 1 μM, 5 μM, 10 μM, 25 μM, 50 μM, 100 μM, 200 μM, 300 μM, 400 μM, 500 μM, 600 μM, 700 μM, 800 μM, 900 μM, 1000 μM z. B. ferrocendimethanol as a redox-active molecule for the generation of hexadecimal memory cells in the memory array.

Then, the following modes for reading out the ROM memory array according to Table 3b result. Table 3b: ROM - Hexadecimal System (HEX). This part is also valid after writes to the WORM memory array and RW memory array. Value (HEX) 0 1 2 3 4 5 6 7 8th 9 A B C D e F Current value (discrete) 1 2 3 4 5 6 7 8th 9 10 11 12 13 14 15 16 Concentration [A + B] (discrete) 0 1 2 3 4 5 6 7 8th 9 10 11 12 13 14 15

Multi-level ROM (HEX) - exemplary, that is not all 16 states are described: mode Redox molecule concentration (discrete) Potential cathode Potential anode Current value (discrete) Memory cell data standstill [A + B] = 0 none none none ? select [A + B] = 0 V _{readout (red)} V _{readout (ox)} 1 "0" standstill [A + B] = 3 none none none ? select [A + B] = 3 V _{readout (red)} V _{readout (ox)} 4 "3" standstill [A + B] = 10 none none none ? select [A + B] = 10 V _{readout (red)} V _{readout (ox)} 11 "A" standstill [A + B] = 15 none none none ? select [A + B] = 15 V _{readout (red)} V _{readout (ox)} 16 "F" Write not possible

According to readout potentials at the cathode V _{read (red)} and / or at the anode V, _{readouts (ox) are} applied which are above (oxidation) or below (reduction) the normal potential of the redox-active molecule and thus the cyclic redox reaction A ↔ B drift in the dielectric.

The absolute value of the measured current is linearly dependent on the concentration of the redox-active molecule and can be assigned to different memory cell values (see also US Pat 4 ).

WORM

• 2.a. WORM: The WORM memory cell is advantageously characterized by comprising a nanoporous dielectric having a redox-active molecule.

For the production of a WORM memory cell, the dielectric is produced with a redox-active molecule of high concentration. The concentration of the redox-active molecule is chosen to be particularly high in order to enable as many irreversible write operations as possible.

In each WORM memory cell there is a reversible redox reaction, and at least one irreversible oxidation and / or reduction.

The following modes for writing and reading are possible, see tables 4a, 4b and the 1b with corresponding cyclic voltammogram. Table 4a: WORM (binary): Read 1 - before writing; Read 2 - after writing; Writing - creating a write potential; Default - all cells are "1" mode Redox molecule concentration (discrete) Potential cathode Potential anode Current value (discrete) Memory cell data standstill [A + B] ≠ 0 none none none ? Read out 1 [A + B] ≠ 0 V _{readout (red)} V _{readout (ox)} 2 "1" Irreversible writing by reduction Write [A + B] ≠ 0 V _{letter (red)} GND - - Read out 2 [A + B] = 0 V _{readout (red)} V _{readout (ox)} 1 "0" or irreversible writing by oxidation Write [A + B] ≠ 0 GND V _{writing (ox)} - - Read out 2 [A + B] = 0 V _{readout (red)} V _{readout (ox)} 1 "0"

It is possible by the choice of a suitable redox-active molecule in the ink to provide WORM memory with the inventive method. For this purpose, the memory cell has at least two memory states, each having the same molecule, the molecule forming a first state 0 (not redox active) or 1 (redox active) without voltage being applied and irreversibly corresponding to the second state 1 (redox active) by applying voltage. or 0 (not redox active). The redox-active state "1" indicates the reversible redox reaction that results from a device according to the invention in redox cycling current.

For the WORM configuration, a dielectric nanoporous layer with a soluble redox-active molecule (state "1") with a correspondingly large diffusion constant is formed at all crossing points the first trace and closed with the second upper trace to form the redox cycling memory cell. This means that immediately after production, all these cells have a default value of "1".

In order to generate a value "0" from the value "1", a potential should be applied between the two electrodes of the crossing point, which irreversibly alters the redox molecules, so that when applying a "normal" low readout potential, no redox -Cycling reaction takes place at this crossing point more. The redox cycling reaction no longer takes place because the irreversibly altered molecules are no longer redox active. See also the 1b with corresponding cyclic voltammograms for WORM memory cells.

Accordingly, these redox molecules must have a low reversible oxidation and reduction state and, in addition, a higher irreversible oxidation or reduction state.

For example, the molecules of the Viologen family have such properties. These molecules have up to three oxidation states, in which the second oxidation reaction usually proceeds irreversibly.

• 1.) A ↔ B und
• 2.) B → C und/oder
• 3.) A → D

In general, the following voltage-driven reactions must take place in WORM memory cells:

• 1.) A ↔ B and
• 2.) B → C and / or
• 3.) A → D

Wherein step 1) is reversible and steps 2) and / or 3) are irreversible.

Alternatively, for a WORM configuration at all crossing points, a dielectric nanoporous layer with an insoluble redox molecule or a redox-active molecule with a correspondingly small diffusion constant in the so-called default state "0" can be arranged.

• 1.) C → B und/oder
• 2.) D → A und
• 3.) A ↔ B

That is, these memory cells have a value of "0" immediately after their manufacture. In order to generate a value of "1" from the value "0", a potential should be applied between the two electrodes of the crossing point, which irreversibly alters the redox molecules, so that they become soluble and get a large diffusion constant. Then, upon application of a "normal" low readout potential, a redox cycling reaction according to step 3.) will occur at this crossover point.

• 1.) C → B and / or
• 2.) D → A and
• 3.) A ↔ B

The redox-active molecule must accordingly be irreversibly oxidizable or reducible. In the first oxidation state, the redox-active molecules are not soluble and can not diffuse between the electrodes. After irreversible oxidation, the molecules become soluble or mobile and can freely diffuse between the electrodes. They can then be reversibly oxidized and reduced by applying readout potentials, see step 3).

• 2.b. Multi-level WORM: Der Multi-Level-WORM wird wie für 2.a. bereit gestellt. Beginnend mit einer hohen Konzentration an dem redoxaktiven Molekül im Dielektrikum wird dieses nach und nach durch unterschiedlich hohe Schreibpotentiale zu verschiedenen fortlaufend geringeren Konzentrationen an dem Redox-Molekül umgeschrieben.

The redox molecules should be suitably selected to match the nanoporous dielectric layer because the solubility depends on the medium in which the substance is dissolved. There may also be residual oxidation or reduction currents in the "0" state resulting from the molecules that are in the immediate vicinity of the electrodes. They should then also have a reversible oxidation and reduction state in order to be read out (see 1b ).

• 2 B. Multi-level WORM: The multi-level WORM will work as for 2.a. provided. Starting with a high concentration of the redox-active molecule in the dielectric, this is gradually rewritten by different writing potentials to various continuously lower concentrations of the redox molecule.

Writing potentials are generally applied over the applied time, the level of the writing potential or the current in order to give rise to different concentrations of redox-active molecules. This statement also applies to RW memory cells.

The extent of the description depends on these parameters. Unlike binary WORM memory cells in which once can be a paraphrase, is in the multi-level encoding by Write a variety of concentrations of redox-active molecules possible. In this way, for example, a total of 16 different concentrations for the generation of hexadecimal memory cells are irreversibly rewritten and read out.

The following partial reading and storage modes according to the examples given in Table 4b result (see also 1b and the upper part of the table 3b, which is also applicable to WORM memory cells). Table 4b: Multi-level WORM (HEX), exemplified for only two writes, that is not all 16 states are given in the table: Read 1 - before writing; Read 2 - after writing; Write 1 - Voltage pulse 1 with certain parameter; Write 2 - Voltage pulse 2 with another parameter; Default - all cells are "1". mode Redox molecule concentration (discrete) Potential cathode Potential anode Current value (discrete) Memory cell data standstill [A + B] = 15 None none none ? Read out 1 [A + B] = 15 V _{readout (red)} V _{readout (ox)} 16 "F" Write 1 by reduction Write 1 [A + B] = 4 V _{letter (red)} GND - - Read out 2 [A + B] = 4 V _{readout (red)} V _{readout (ox)} 5 "4" or write 1 by oxidation Write 1 [A + B] = 4 GND V _{writing (ox)} - - Read out 2 [A + B] = 4 V _{readout (red)} V _{readout (ox)} 5 "4" Write 2 by reduction Write 2 [A + B] = 0 V _{letter (red)} GND - - Read out 2 [A + B] = 0 V _{readout (red)} V _{readout (ox)} 1 "0" or write 2 by oxidation Write 2 [A + B] = 0 GND V _{writing (ox)} - - Read out 2 [A + B] = 0 V _{readout (red)} V _{readout (ox)} 1 "0" rewrite not possible

Thus, as described, the WORM memory cell is made with a large concentration of redox-active molecule. A first write is initiated by applying the write potential and rewrites the redox-active molecule irreversibly until it has a lower concentration of [A + B], here the discrete concentration 4. The writing is done either by oxidation or reduction.

Advantageously, all intermediate levels can be adjusted by the shape of the write pulse. For example, by varying the shape and strength of the voltage pulse from the value "1" only a part of the redox molecules can be irreversibly changed and thereby achieve a certain concentration. A peculiarity in this case is that one can rewrite the memory cell again and again with the right pulse in the direction of a lower concentration of the redox molecules (less "1", more "0"), but not in the other direction.

Alternatively, the initial state is default "0" in the case of initially using insoluble molecules. This means that further voltage pulses can be used to oxidize more and more insoluble molecules and to increase the concentration of soluble redox-active molecules which produce more "1" and less "0" states according to the reaction A ↔ B.

RW

It is possible with the method according to the invention to provide an RW memory array, that is to say rewritable memory cells.

An RW memory cell is advantageously characterized in that it comprises a nanoporous dielectric with a redox-active molecule. By applying voltage, at least two reversible redox reactions take place in the RW memory cell, with which the redox-active molecule changes between at least three oxidation states. Only the redox reaction A ↔ B produces state 1.

• 3.a. RW-Speicher (binär): Für eine RW-Konfiguration wird an allen Kreuzungspunkten eine dielektrische nanoporöse Schicht mit redoxaktiven Molekülen (Default ”1”) aufgebracht. Das heißt, direkt nach der Herstellung haben alle Zellen einen Wert ”1”.

This memory array has at least two memory states each having the same molecule, the molecule forming a first state "0" (not redox active) or "1" (redox active) without application of voltage and reversibly reversing the second state "1" by applying voltage. (redoxaktiv) or "0" (not redoxaktiv) change.

• 3.a. RW memory (binary): For a RW configuration, a dielectric nanoporous layer with redox-active molecules (default "1") is applied at all crossing points. That is, immediately after production, all cells have a value of "1".

To generate a value "0" from the value "1", a potential should be applied between the two electrodes of the crossing point, which reversibly changes the redox molecules, eg. B. by oxidation or reduction in a second higher or lower oxidation state.

With a low readout potential no redox cycling reaction takes place at this crossing point, ie the redox cycling reaction does not take place because the altered molecules are no longer redox active.

In order to generate a value "1" again from the value "0", an inverse potential between the two electrodes of the crossing point is to be applied, which reversibly returns the redox molecules to the lower (or higher) oxidation state the application of a read-out potential of the storage state, again a redox cycling reaction according to the redox reaction A ↔ B takes place at this crossing point, see 1c with corresponding cyclic voltammograms, and Table 5a. Table 5a: RW (binary). Reading 1 - before writing; Read 2 - after writing; Reading 3 - after rewriting; Writing - creating a write potential; Rewriting - applying a reverse writing potential; all cells with default "1" mode Redox molecule concentration (discrete) Potential cathode Potential anode Current value (discrete) Memory cell data standstill [A + B] ≠ 0 None none none ? Read out 1 [A + B] ≠ 0 V _{readout (red)} V _{readout (ox)} 2 "1" Reversible writing by reduction Write [A + B] ≠ 0 V _{letter (red)} GND - - Read out 2 [A + B] = 0 V _{readout (red)} V _{readout (ox)} 1 "0" or reversible writing by oxidation Write [A + B] ≠ 0 GND V _{writing (ox)} - - Read out 2 [A + B] = 0 V _{readout (red)} V _{readout (ox)} 1 "0" Rewriting by reoxidation (after reduction) rewrite [A + B] ≠ 0 GND V _{writing (ox)} - - Reading 3 [A + B] ≠ 0 V _{readout (red)} V _{readout (ox)} 2 "1" Rewriting by re-reduction (after oxidation) rewrite [A + B] ≠ 0 V _{letter (red)} GND - Reading 3 [A + B] ≠ 0 V _{readout (red)} V _{readout (ox)} 2 "1"

The redox-active molecule must accordingly have a low reversible oxidation and reduction state, and in addition also a higher or lower second reversible oxidation or reduction state.

Such molecules are z. B. aromatic azo compounds such as azobenzene, azotoluene, and so on, which are dissolved in a solution of dimethylformamide with tetra-n-butylammonium perchlorate (TBAP) and can form RW memory cells according to the invention.

• 1.) A ↔ B und
• 2.) B ↔ C und/oder
• 3.) A ↔ D

In general, the following reactions must take place for RW memory cells:

• 1.) A ↔ B and
• 2.) B ↔ C and / or
• 3.) A ↔ D

• 1.) C ↔ B und/oder
• 2.) D ↔ A und
• 3.) A ↔ B

Alternatively, for a RW configuration, a dielectric nanoporous layer with insoluble redox molecules is initially introduced at all crossing points during production in the dielectric, which correspondingly have a small diffusion constant. These memory cells therefore have a default "0" directly after their manufacture. In order to generate a value "1" from the value "0", a potential should be applied between the two electrodes of the crossing point, which reversibly changes the redox molecules, so that they become soluble and get a large diffusion constant.

• 1.) C ↔ B and / or
• 2.) D ↔ A and
• 3.) A ↔ B

Then, upon application of a "regular" low readout potential, a redox cycling reaction will occur at that crosspoint according to step 3) and generate state "1". In the state "0", the diffusion of the molecules is so limited that only a half-reaction of the molecules in the immediate vicinity of the electrode surface, but no redox cycling amplified current can take place. So that these currents are not recognized as "1", a signal is only read as "1" if it exceeds a threshold.

• 3.b. Multi-level RW: Dieses Speicherarray wird wie in 3.a. gezeigt bereit gestellt, aber mit mehreren Konzentrationen von Redox-Molekülen, z. B. mit bis zu 16 verschiedenen Konzentrationen für die Erzeugung von hexadezimalen Speicherzellen. Die verschiedenen Konzentrationen an redoxaktiven Molekülen werden wie bei der WORM-Speicherzelle durch Schreiben erzeugt. Bei der Herstellung wird das Dielektrikum mit höchster Konzentration an redoxaktiven Molekülen bereitgestellt.

In order to generate a value "0" again from the value "1", a reversed potential should be applied between the two electrodes of the crossing point, which reversibly returns the redox molecules to the lower (or higher) oxidation state, so that the molecules return insoluble and thus not become mobile in the dielectric and when applying a "normal" low read-out potential of the memory state, again no redox cycling reaction can take place at this crossing point more. The redox molecules for RW memory cells must be correspondingly reversibly oxidizable. In the first oxidation state, the redox molecules are insoluble and can not diffuse between the electrodes. They should then also have a higher second reversible oxidation or reduction state to be read. For example, this can ferrocene (not soluble in water) and their derivatives such as ferrocene methanol (soluble in water) can be used, which can be reversibly converted from a state in the.

• 3.b. Multi-level RW: This storage array will work as described in 3.a. shown but with multiple concentrations of redox molecules, e.g. With up to 16 different concentrations for the generation of hexadecimal storage cells. The various concentrations of redox-active molecules are generated by writing as in the WORM memory cell. During production, the dielectric is provided with the highest concentration of redox-active molecules.

The following readout and write modes are possible, see the examples shown in excerpts in Table 5b and the 1c : Table 5b: Multi-level RW (HEX), for example (not all 16 states are displayed): Read 1 - before writing; Read 2 - after writing; Reading 3 - after rewriting; Write - voltage pulse with certain parameters; Rewriting - reverse voltage pulse with certain parameters; Default - all cells "1". mode Redox molecule concentration (discrete) Potential cathode Potential anode Current value (discrete) Memory cell data standstill [A + B] = 15 none none none ? Read out 1 [A + B] = 15 V _{readout (red)} V _{readout (ox)} 16 "F" Reversible writing 1 by reduction Write [A + B] = 3 V _{letter (red)} GND - - Read out 2 [A + B] = 3 V _{readout (red)} V _{readout (ox)} 4 "3" or reversible writing 1 by oxidation Write [A + B] = 3 GND V _{writing (ox)} - - Read out 2 [A + B] = 3 V _{readout (red)} V _{readout (ox)} 4 "3" Rewriting by reoxidation (after reduction) rewrite [A + B] = 11 GND V _{writing (ox)} - - Reading 3 [A + B] = 11 V _{readout (red)} V _{readout (ox)} 12 "B" Rewriting by re-reduction (after oxidation) rewrite [A + B] = 11 V _{letter (red)} GND - - Reading 3 [A + B] = 11 V _{readout (red)} V _{readout (ox)} 12 "B"

Thus, for example, the RW memory cell is produced with a large concentration of redox-active molecule in the dielectric. A first write is accomplished by applying the write potential and reversibly rewrites the redox-active molecule until it has a lower concentration of [A + B], here the discrete concentration 3. The first write is either by oxidation or reduction.

You can set the intermediate levels by the shape of the write pulse. For example, by varying the shape and the magnitude of the voltage pulse from the value "1" only a part of the redox molecules can be reversibly changed and thereby achieve a certain concentration, that is, the concentration of the readable redox molecules is reduced.

Then, with a suitable pulse, the altered, non-readable redox molecules can be reversibly changed back to the original state, or into another intermediate state here of the discrete concentration with a value of 11, and the concentration of the redox molecules readable can be increased again.

It is understood that the molecules A and B, which are only schematically indicated for the ROM memory cells, WORM memory cells and RW memory cells, as well as, if appropriate, also C and D, must actually be different in order to fulfill the prerequisites described above. So, there is actually A _ROM , B _ROM , A _WORM , B _WORM , C _WORM , D _WORM , A _RW , B _RW , C _RW , D _RW .

The memory cells ROM, WORM and RW thus differ by chemically different redox-active molecules or oxidation states A _ROM , B _ROM , A _WORM , B _WORM , C _WORM , D _WORM , A _RW , B _RW , C _RW , D _RW , wherein the cyclic reaction for generating current above the threshold value and thus memory state 1 between A ↔ B takes place, as indicated inter alia in the embodiments.

The memory cells WORM and RW thus likewise differ by chemically different molecules C and D which can be generated by writing potential.

The memory arrays of the present invention are used as non-volatile memory for generating bits and bytes on a binary or hexadecimal basis or any other basis.

The ink for producing such a memory already achieves the object of the invention. For this purpose, the ink comprises the precursors of a sol-gel reaction or hydrogels and the redox-active molecules, which can diffuse between the electrodes by drying or hardening of the ink, either immediately as default "1" or after application of a voltage pulse , which makes the molecules soluble in the dielectric (default "0").

As inks, any liquid composition which can be arranged in particular by a printing process can be used, which after deposition allows a dielectric layer, preferably with redox-active molecules, and which allows the diffusion of the redox-active molecules. The diffusion passes through the nanopores of the nanoporous dielectric. For this purpose, the pore size should be greater than the molecular size of redox-active molecules. In the usual sizes of the redox-active molecules of about 1-5 nm, the pore size z. B. at least 2-10 nm or more (factor 2). If the dielectric layer has a solid but quasi-liquid phase as in the case of the hydrogel, the diffusion proceeds directly through this phase.

Regardless of the physical mode of action, the diffusion constant of the redox-active molecules in diffusion through the pores should be as similar as possible to the diffusion constant in the liquid phase without pores and at least 1% thereof or better, for. At least 5%, 10%, 15%, 20%, 30%, 40%, 50%, 60%, 70%, 80%, 90% or even 100%.

According to the definition according to the invention for the redox-active molecules, these are not enzymes.

The ink should have a solids content so that the layer thickness of the dielectric layer after printing and drying is about 50-1000 nm, most preferably 100-300 nm.

Preferably, the nanoporous dielectrics are produced between the electrodes with the specified components after printing on the first conductor track. For this purpose, the ink advantageously comprises the starting components for a sol-gel reaction, for. Tetramethylorthosilicate, tetraethoxysilane and so on). In the first step, the hydrolysis in water and alcohol and then in a further step, the condensation to the sol-gel matrix with residual liquid.

For this purpose, the ink preferably has a solvent with a low vapor pressure and a correspondingly high boiling point.

embodiments

In addition, the invention will be explained in more detail with reference to embodiments and the accompanying figures, without this being intended to limit the invention.

Show it:

1 : Dependence of the measured redox cycling current on the applied voltage, represented by cyclic voltammograms of the redox-active molecules used.
a) ROM; b) WORM and c) RW.

2 : Schematic representation of inventive memory array
a) ROM binary. b) WORM / RW binary. c) Multi-level ROM / WORM / RW

3 : Method for producing a memory array (ROM / WORM / RW)

4 : Dependence of the measured CV-voltammogram on the concentration of the redox-active molecules, here using the example of ferrocendimethanol.

1 a) shows the cyclic voltammogram of a ROM cell. 1 b) shows the cyclic voltammogram of a WORM cell. 1 c) shows the cyclic voltammogram of an RW cell. The modes for writing and reading are given in the descriptions of Tables 3 to 5.

2a schematically shows the arrangement of the nanoporous dielectrics at 16 crossing points between the electrodes in a ROM memory array, binary. The memory states - here only "0" and "1" are given to the right. Only the states "0" and "1" are possible according to the ROM, see Table 3a. Bright circles affect the memory cells without redox-active molecules, filled circles, however, memory cells with redox-active molecules, ie those in which electricity can be generated.

2 B schematically shows the arrangement of the nanoporous dielectrics at 16 crossing points between the electrodes in a WORM or RW memory array. The memory states - here only "0" and "1" (binary) are indicated to the right, see Tables 4a and 5a. The black circles symbolize the memory cells with redox-active molecules in which electricity can be generated, the gray shaded circles the memory cells without sufficient power.

2c schematically shows the arrangement of nanoporous dielectrics at 16 crossing points between the electrodes in a multilevel ROM / WORM or RW memory array. The different memory states - from "0" to "F" - are indicated to the right. These memory states are shown in the left part of the figure by the different hatching at the crossing points and thus by the different concentrations.

Embodiment 1

25-bit (binary) printed redox cycling based electrochemical ROM memory module ( 1a . 2a and 3 to 4 ).

• 1. Goldtinte wird als Material für die erste Leiterbahn 2 bzw. für die untere Elektrode benutzt. Die leitenden Strukturen 2, dargestellt durch insgesamt fünf Linien (100 μm breit, 100 μm Abstand zwischen den Leiterbahnen) aus der Goldtinte werden auf einem PEN(Polyethylennaphthalat)-Substrat 1 mit einem Tintenstrahldrucker gedruckt und danach bei 125°C für 1 Stunde gesintert.
• 2. Zwei Sol-Gel-Tinten werden vorbereitet: a. Ohne Redox-Moleküle: TMOS (Tetramethylorthosilicat) wird 1:1:1 (Gewichtsanteile) mit deionisiertem Wasser und Glyzerol in einer 100 mL Flasche gemischt und eine Stunde lang mit einem magnetischen Rührer auf einer Magnetplatte bei Raumtemperatur gerührt. Danach wird 100 mM-Lösung der Salzsäure im 500:1 (Sol-Gel:Säure, Gewichtsanteile) für den Start der Kondensationsreaktion dazu gegeben. b. Mit Redox-Moleküle: Eine 1 mM-Lösung einer Mischung Redox-Moleküle Hexacyanoferrat (II)/Hexacyanoferrat (III) (Normalpotential +200 mV gegen Kalomelelektrode SCE) wird im deionisiertem Wasser vorbereitet. TMOS wird 1:1:1 (Gewichtsanteile) mit der Ferrocendimethanollösung und Glyzerol in einer 100 mL Flasche gemischt und eine Stunde lang mit magnetischem Rührer auf einer Magnetplatte bei Raumtemperatur gerührt. Danach wird 100 mM-Lösung der Salzsäure im 500:1 (Sol-Gel:Säure, Gewichtsanteile) für den Start der Kondensationsreaktion dazu gegeben. Bei Bedarf wird eine 10–100 mM Salzlösung (z. B. NaCl) dazu gegeben, um die elektrische Leitfähigkeit der Schicht und Elektronentransfer zu verbessern.
• 3. Die Sol-Gel-Tinten werden auf den zukünftigen Kreuzungspunkten gedruckt. An Stellen, an denen ein Zustand ”1” (3b) ausgelesen werden soll, wird die Tinte mit den Redox-Molekülen für das nanoporöse Dielektrikum 3a gedruckt, dargestellt durch die dunklen Kreise im linken Teil der 3b. An den Stellen, an denen der Zustand ”0” ausgelesen werden soll, wird die Tinte ohne Redox-Moleküle gedruckt für das nanoporöse Dielektrikum 3b, dargestellt durch die nicht ausgefüllten Kreise.
• 4. Nach dem Drucken der nanoporösen Dielektrika 3a, 3b wird die gedruckte Tinte bei Raumtemperatur für eine Stunde ausgehärtet, so dass die gedruckten Dielektrika noch Rest-Flüssigkeit enthalten. Entsprechend bildet sich das punktförmige nanoporöse Dielektrikum mit (Bezugszeichen 3a) und ohne (Bezugszeichen 3b) ein redoxaktives Molekül.
• 5. Karbontinte aus 300–400 nm großen Kohlenstoffnanopartikeln wird als Material für die zweite Leiterbahn 4 bzw. die obere Elektrode benutzt. Diese Tinte wird orthogonal zu den ersten Leiterbahnen 2 über die nanoporösen Dielektrika gedruckt um die ”cross-bar”-Struktur (Matrix) und die Kontaktstellen zu formen. Insgesamt sind wie für die Leiterbahn 2 wiederum fünf Leiterbahnen bzw. Elektroden 4 gezeigt. Die Tinte wird nach dem Druck bei 125°C für 1 Stunde gesintert.
• 6. Als Passivierungsschicht 5 wird eine Schicht aus Polyimid großflächig aufgebracht, z. B. gedruckt, so dass nur die Kontaktstellen am Ende der Elektroden offen bzw. kontaktierbar bleiben, wie in der 3 im linken Teil gezeigt.
• 7. Um den Speicher auszulesen, werden alle Elektroden gleichzeitig oder eine nach der anderen (durch einen Schalter) kontaktiert. Bei jedem Kreuzungspunkt wird auf der unteren Elektrode ein Potential von 0 mV gegen Masse angelegt, während auf der oberen Elektrode ein Potential von +300 mV gegen Masse angelegt wird. Die Referenzelektrode entfällt in diesem Design und der Strom wird ohne Referenzelektrode gegen Masse gemessen. Wenn der ausgelesene Strom einen bestimmten Stromwert (abhängig von dem Zellendesign und der Konzentration der Redox-Moleküle, für diese Beispiel ca. 40 pA) nach einer bestimmten Zeit (abhängig vom Zellendesign, für dieses Beispiel ca. 3 ms) überschreitet, wird diese Zelle als Zustand ”1” identifiziert. Wenn nicht, wird sie als Zustand ”0” identifiziert.

2a schematically shows the arrangement of the ROM memory array in the finished state. 3 shows the associated manufacturing process in plan left and in section along the line AA right in the figure. For reasons of space is in the 3 only one structure has been provided with a reference numeral.

• 1. Gold ink is used as material for the first trace 2 or used for the lower electrode. The conductive structures 2 , represented by a total of five lines (100 microns wide, 100 microns distance between the tracks) from the gold ink on a PEN (polyethylene naphthalate) substrate 1 printed with an ink jet printer and then sintered at 125 ° C for 1 hour.
• 2. Prepare two sol-gel inks: a. Without Redox Molecules: TMOS (tetramethyl orthosilicate) is mixed 1: 1: 1 (parts by weight) with deionized water and glycerol in a 100 mL bottle and stirred for one hour on a magnetic disk at room temperature with a magnetic stirrer. Thereafter, 100 mM solution of hydrochloric acid in 500: 1 (sol-gel: acid, parts by weight) is added for the start of the condensation reaction. b. With redox molecules: A 1 mM solution of a mixture of redox molecules hexacyanoferrate (II) / hexacyanoferrate (III) (normal potential +200 mV against calomel electrode SCE) is prepared in deionized water. TMOS is mixed 1: 1: 1 (parts by weight) with the ferrocene dimethanol solution and glycerol in a 100-mL bottle and stirred for one hour with a magnetic stirrer on a magnetic disk at room temperature. Thereafter, 100 mM solution of hydrochloric acid in 500: 1 (sol-gel: acid, parts by weight) is added for the start of the condensation reaction. If necessary, a 10-100 mM saline solution (eg NaCl) is added to improve the electrical conductivity of the layer and electron transfer.
• 3. The sol-gel inks will be printed on the future crossing points. In places where a state "1" ( 3b ) is to be read, the ink with the redox molecules for the nanoporous dielectric 3a printed, represented by the dark circles in the left part of the 3b , At the points where the state "0" is to be read, the ink is printed without redox molecules for the nanoporous dielectric 3b represented by the unfilled circles.
• 4. After printing the nanoporous dielectrics 3a . 3b For example, the printed ink is cured at room temperature for one hour so that the printed dielectrics still contain residual liquid. Accordingly, the point-shaped nanoporous dielectric forms with (reference numerals 3a ) and without (reference number 3b ) a redox-active molecule.
• 5. Carbon ink from 300-400 nm sized carbon nanoparticles is used as material for the second trace 4 or the upper electrode used. This ink becomes orthogonal to the first traces 2 printed over the nanoporous dielectrics to form the cross-bar structure (matrix) and the contact points. Overall, as for the track 2 again five tracks or electrodes 4 shown. The ink is sintered after printing at 125 ° C for 1 hour.
• 6. As a passivation layer 5 a layer of polyimide is applied over a large area, for. B., so that only the contact points at the end of the electrodes remain open or contactable, as in the 3 shown in the left part.
• 7. To read the memory, all electrodes are contacted simultaneously or one at a time (through a switch). At each crossing point, a potential of 0 mV to ground is applied to the lower electrode while a potential of +300 mV to ground is applied to the upper electrode. The reference electrode is omitted in this design and the current is measured without reference electrode to ground. If the read-out current exceeds a certain current value (depending on the cell design and the concentration of the redox molecules, for this example about 40 pA) after a certain time (depending on the cell design, for this example about 3 ms), this cell becomes identified as state "1". If not, it is identified as state "0".

Embodiment 2:

A 25-bit (HEX) printed redox cycling based electrochemical ROM memory module ( 1a . 2c . 3 )) is produced.

Steps 1 to 6 are the same as in Embodiment 1. However, up to 16 different inks containing up to 16 different concentrations of redox molecules are provided. There are 16 inks with 16 concentrations of z. 0 μM, 1 μM, 5 μM, 10 μM, 25 μM, 50 μM, 100 μM, 200 μM, 300 μM, 400 μM, 500 μM, 600 μM, 700 μM, 800 μM, 900 μM, 1000 μM). the redox molecules Hexacyanoferrate (II) / hexacyanoferrate (III) prepared. The concentrations are in the 2c shown with circles with different fillings.

• 7. Um den Speicher auszulesen, werden alle Elektroden gleichzeitig oder eine nach der anderen (durch einen Schalter) kontaktiert. Bei jedem Kreuzungspunkt wird auf der unteren Elektrode 2 ein Potential von –300 mV gegen Masse angelegt, während auf der oberen Elektrode 4 ein Potential von +300 mV gegen Masse angelegt wird. Die Referenzelektrode entfällt in diesem Design und der Strom wird ohne Referenzelektrode gegen Masse gemessen. Der ausgelesene Strom nach einer bestimmten Zeit (abhängig von dem Zellendesign, für dieses Beispiel ca. 3 ms) wird zu einer bestimmten Konzentration (aus 16 vorkalibrierten Stromwerten insgesamt, siehe auch die 4) zugeordnet um den Speicherzustand der Zelle auszulesen. Es sind dann Zustände zwischen ”0” und ”F” (0, 1, 2, 3, 4, 5, 6, 7, 8, 9, A, B, C, D, E, F) möglich.

These inks are passed through multiple printheads or nozzles into the crossing points of the memory array as a dielectric nanoporous layer 3 printed consecutively to generate 25-bit (HEX) memory modules.

• 7. To read the memory, all electrodes are contacted simultaneously or one at a time (through a switch). At each crossing point is on the lower electrode 2 a potential of -300 mV is applied to ground while on the top electrode 4 a potential of +300 mV is applied to ground. The reference electrode is omitted in this design and the current is measured without reference electrode to ground. The read current after a certain time (depending on the cell design, for this example about 3 ms) becomes a certain concentration (from 16 pre-calibrated current values in total, see also the 4 ) to read the memory state of the cell. Then states between "0" and "F" (0, 1, 2, 3, 4, 5, 6, 7, 8, 9, A, B, C, D, E, F) are possible.

Is exemplary in the 4 shown that the measured currents can be assigned to the concentrations for reading the 16 states.

Embodiment 3

A 25-bit (binary) redox cycling-based WORM electrochemical storage module ( 1b . 2 B . 3 ) is produced.

• 1. Wie Ausführungsbeispiel 1, aber Schritt 2a entfällt. Jedes nanoporöse Dielektrikum enthält also Tinte mit redoxaktiven Molekülen. In Schritt 2b wird Heptylviologenbromid mit einem Normalpotential der ersten reversiblen Reduktionsreaktion von –300 mV und einem Normalpotential der zweiten irreversiblen Reduktionsreaktion von –700 mV anstatt Hexacyanoferrat (II)/Hexacyanoferrat (III) Mischung verwendet. 2. Diese Sol-Gel-Tinte wird auf allen zukünftigen Kreuzungspunkten gedruckt um den Default „1” zu bilden (Schritt 3.).

Steps 1-7 are the same as those of Embodiment 1. The differences are in Steps 2 and 3, and Step 6 * is repeated:

• 1. As Embodiment 1, but step 2a is omitted. Each nanoporous dielectric thus contains ink with redox-active molecules. In step 2b, heptylviologen bromide having a normal potential of the first reversible reduction reaction of -300 mV and a normal potential of the second irreversible reduction reaction of -700 mV instead of hexacyanoferrate (II) / hexacyanoferrate (III) mixture is used. 2. This sol-gel ink will be printed on all future crossing points to form the default "1" (step 3.).

• 6*. Um aus dem Zustand „1” Zustand „0” an einigen bestimmten Kreuzungspunkten (Speicherzellen aus dem Array) zu erzeugen, wird zwischen der oberen und unteren Elektrode an diesen Kreuzungspunkten ein Reduktionspotential von –1 V für 1 Sekunde angelegt. Das angelegte Potential reicht aus, um die irreversible Reaktion A → D zu treiben. Danach sind die Moleküle in dem Bereich von Auslesepotentialen im Sinne der Redoxreaktion A ↔ B nicht mehr redoxaktiv. Stattdessen liegen die redoxaktiven Moleküle in der inaktiven reduzierten Form D vor.
• 7. Wie in Ausführungsbeispiel 1.

If necessary, an irreversible writing step may be carried out before step 7:

• 6 *. In order to generate state "0" from the state "1" at some specific crossing points (memory cells from the array), a reduction potential of -1 V is applied for 1 second between the upper and lower electrodes at these crossing points. The applied potential is sufficient to drive the irreversible reaction A → D. Thereafter, the molecules in the range of readout potentials in the sense of the redox reaction A ↔ B no longer redoxaktiv. Instead, the redox-active molecules are present in the inactive reduced form D.
• 7. As in embodiment 1.

Embodiment 4

A 25-bit (binary) redox cycling-based electrochemical RW memory module ( 1c . 2 B . 3 ) is produced.

• 2. An Stelle von Heptylviologenbromid wird Azotoluen der ersten reversiblen Oxidationsreaktion von +300 mV und einem Normalpotential der zweiten reversiblen Reduktionsreaktion von +700 mV verwendet. Nach dem Aufbringen der Tinte mit dem redoxaktiven Molekül weisen alle Speicherzellen den Default „1” auf.
• 6*. Um aus dem Zustand „1” einen Zustand „0” an bestimmten Kreuzungspunkten bzw. Speicherzellen aus dem Array zu erzeugen, wird zwischen den oberen und den unteren Elektroden an diesen Kreuzungspunkten ein Oxidationspotential von +1,5 V für 1 Sekunde angelegt. Danach sind die Moleküle in dem Bereich vom Auslesepotential der Redoxreaktion A ↔ B nicht mehr redoxaktiv, da Molekül C erzeugt wurde. Es kann alternativ auch das Molekül D durch ein entsprechendes Reduktionspotential von –1,5 V erzeugt werden.
• 8. Um aus dem geschriebenen Zustand „0”, wieder den Zustand „1” an bestimmten Kreuzungspunkten zu erhalten, wird zwischen den oberen und den unteren Elektroden an diesen Kreuzungspunkt ein umgekehrtes Reduktionspotential (bzw. Oxidationspotential) von –1,5 V für 1 Sekunde angelegt. Danach werden die Moleküle in dem Bereich von Auslesepotentialen der Redoxreaktion A ↔ B wieder redoxaktiv.

Steps 1 and 3-5 as well as 7 correspond to those from embodiment 3. The difference lies in steps 2 and 6 *, in addition there is a step 8:

• 2. In place of heptylviologen bromide azotoluene of the first reversible oxidation reaction of +300 mV and a normal potential of the second reversible reduction reaction of +700 mV is used. After the application of the ink with the redox-active molecule, all memory cells have the default "1".
• 6 *. In order to generate a state "0" from the state "1" at certain crossing points or memory cells from the array, an oxidation potential of +1.5 V is applied for 1 second between the upper and the lower electrodes at these crossing points. Thereafter, the molecules are no longer redox active in the region of the readout potential of the redox reaction A ↔ B, since molecule C was generated. Alternatively, the molecule D can also be generated by a corresponding reduction potential of -1.5 V.
• 8. In order to obtain the state "1" again from the written state "0" at certain points of intersection, a reversed reduction potential (or oxidation potential) of -1.5 V for 1. Is obtained between the upper and the lower electrodes at this intersection point Second created. Thereafter, the molecules in the range of readout potentials of the redox reaction A ↔ B again redoxaktiv.

Embodiment 5:

A 25-bit (hexadecimal) printed redox cycling based electrochemical RW memory module is manufactured.

The procedure corresponds to embodiment 2, steps 1-6. There is a difference only in the redox-active molecules in step 2. The molecules used here correspond to those from step 2. in the embodiment 4 (RW binary memory). To write the memory (step 7) a writing potential is created for a defined short period of time, e.g. For 100 mS. To reduce the stored state, an oxidation potential is applied, so that a part of the particles is converted into a reversible, higher oxidation state (B → C). To increase the stored state, a reducing potential is applied, so that a part of the particles is transferred from the higher to the lower state (C → B) and is available again for the redox cycling. This process can also be done in the direction of a second reduction. The potentials are created exactly the opposite. The readout then takes place analogously to step 7, exemplary embodiment 2.

Further embodiments:

Further embodiments relate to the production and use of such storage cells for the food industry. The ROM memory cells disclosed in the exemplary embodiments, and in particular in the embodiments 1 and 2, are particularly suitable for the production of printed memory cells on outer packaging in the food industry. The reason is that the redox molecules, such as the hexacyanoferrate (II) / hexacyanoferrate (III) are largely food safe, especially when they are placed on the surface of the food-containing cardboard or other outer packaging.

Such ROM memory cells can then be used to provide information about the food, such as the date of manufacture, the expiration date, the date of filling of the food and other relevant parameters such. B. origin of the food or transport routes, etc. store during manufacture and read in a simple way again.

It is conceivable to apply the necessary voltage to the memory cell for reading even without electronic leads and read out the corresponding parameters again, for. B. over radio waves, as in an RFID tag.

The method according to the invention can be carried out for this purpose by providing a component of such a cardboard as non-conductive substrate.

The inventive use of the memory cells then relates to the additional feature, according to which this memory cell is arranged on an outer packaging from the food industry. The memory cell could also be glued to the packaging.

QUOTES INCLUDE IN THE DESCRIPTION

This list of the documents listed by the applicant has been generated automatically and is included solely for the better information of the reader. The list is not part of the German patent or utility model application. The DPMA assumes no liability for any errors or omissions.

Cited non-patent literature

• Ilia Valov, Rainer Waser, John R. Jameson, and Michael N Kozicki, 2011. Electrochemical metallization memories: fundamentals, applications, prospects. Nanotechnology 22, 254003 (22pp) doi: 10.1088 / 0957-4484 / 22/25/254003 [0017]
• Buckley, AM, Greenblatt, M. 1994. Journal of Chemical Education. Volume 71, no. 7, 599-602 [0063]
• Buckley, AM, Greenblatt, M. 1994. Jouranl of chemical education. Volume 71, no. 7, 599-602 [0077]

Claims (19)

Method for producing a memory cell, characterized by the steps: a) a nonconductive substrate ( 1 ) will be provided; b) a first conductor track ( 2 ) of a conductive material is disposed on the substrate; c) a porous dielectric is punctiform on the first conductor track ( 3a . 3b ) with or without redox-active molecules; d) orthogonal to the first interconnect, a second interconnect ( 4 ), wherein the first and the second conductor track ( 2 . 4 ) have an electrode function at their crossing point, and between the electrodes the dielectric ( 3a . 3b ) is arranged; e) on the substrate, the first interconnect, the dielectric and the second interconnect is a passivation layer ( 5 ), so that the conductor tracks remain contactable; wherein the first and the second conductor track form a memory at their point of intersection with the dielectric disposed therebetween, in which the redox reaction of the redox-active molecules at the electrodes is driven by the application of voltage via the conductor tracks in order to generate a memory state. A method according to claim 1, characterized in that the first conductor track and / or the dielectric and / or the second conductor track and / or the passivation layer are arranged with a printing process. Method according to one of the preceding claims, characterized by the choice of a sol-gel ink or a hydrogel ink for the dielectric, which is dried after application to the first conductor and forms the nanoporous layer with pores. Method according to the preceding claim, characterized by the choice of an ink which comprises redox-active molecules which can diffuse to the electrodes after the formation of the nanoporous dielectric in the pores of the dielectric and can be converted to these. Method according to one of the preceding claims, characterized by repeating the steps for forming a plurality of orthogonally arranged interconnects in a memory array. Method according to the preceding claim, characterized by selection of ink with different concentrations and / or substances on redox-active molecules for the storage array. Memory cell, comprising at least one arrangement of a first electrically contactable conductor track ( 2 ) on a non-conductive substrate ( 1 ), an orthogonal thereto arranged second electrically contactable conductor track ( 4 ), wherein the interconnects have electrode function at the intersection point and wherein at the intersection point between the two interconnects a porous dielectric ( 3a . 3b ) is arranged with redox-active molecules which are freely diffusible in the pores and which can be oxidized and / or reduced at the electrodes of the conductor tracks by the application of electrical voltage to generate memory states, and wherein the memory of substrate, interconnects and dielectric is completely filled by a passivation layer ( 5 ) is passivated. Memory array with a plurality of passivated electrochemical storage cells according to the preceding claim in a "crossbar" configuration. Memory array according to the preceding claim, characterized in that different memory cells of the memory array have a plurality of different redox-active molecules and / or different concentrations of a redox-active molecule. ROM memory array according to one of the preceding claims 8 or 9, characterized by a plurality of nanoporous dielectrics ( 3b ) without redox-active molecules and a large number of nanoporous dielectrics ( 3a ) with a redox-active molecule, wherein the redox-active molecule can be reversibly oxidized and reduced exclusively between its two oxidation states according to a reaction A ↔ B. WORM memory array according to one of the preceding claims 8 or 9 characterized by a plurality of nanoporous dielectrics ( 3a ) with a redox-active molecule, wherein the redox-active molecule can be reversibly oxidized and reduced between its two oxidation states A and B according to a first reaction A ↔ B and in which the redox-active molecule additionally at least one further irreversible oxidation state according to a second irreversible reaction B → C and / or a third irreversible reaction A → D. RW memory array according to one of the preceding claims 8 or 9, characterized by a plurality of nanoporous dielectrics ( 3a ) with a redox-active molecule, wherein the redox-active molecule can be reversibly oxidized and reduced between its two oxidation states A and B according to a first reaction A ↔ B and in which the redox-active molecule additionally at least one further reversible oxidation state according to a second reversible reaction B ↔ C and / or a third reversible reaction A ↔ D has. Use of a ROM memory array according to claim 10, wherein the redox-active molecule is reversibly oxidized and reduced by applying voltage to the electrodes having a read-out potential above and below the normal potential of the redox-active molecule according to the reaction A ↔ B and the measured current when a threshold value is exceeded is read out as state "1" and in which the state 0 is read out for memory cells without redox-active molecules. Use of a WORM memory array according to claim 11, wherein the redox-active molecule is reversibly oxidized and reduced by applying voltage to the electrodes having a read-out potential above and below the normal potential of the redox-active molecule according to the reaction A ↔ B and the measured current when a threshold value is exceeded is read as state 1. Use of a WORM memory array according to the preceding claim, wherein the redox-active molecule is rewritten by applying voltage to the electrodes with a writing potential in the irreversible oxidation state according to the reaction B → C or A → D, so that the redox-active molecule by applying voltage is no longer oxidized or reduced to the electrodes with a read-out potential above and below the normal potential of the redox-active molecule according to the reaction A ↔ B and the measured current is read out as state 0. Use of an RW storage array according to claim 12, wherein the redox-active molecule is reversibly oxidized and reduced by applying voltage to the electrodes having a read-out potential above and below the normal potential of the redox-active molecule according to the reaction A ↔ B and the measured current when the threshold value is exceeded is read as state 1. Use of an RW memory array according to the preceding claim, wherein the redox-active molecule is rewritten by applying voltage to the electrodes with a write potential in the reversible oxidation state according to the reaction B ↔ C or A ↔ D, so that the redox-active molecule by applying voltage is no longer oxidized or reduced to the electrodes with a read-out potential above and below the normal potential of the redox-active molecule according to the reaction A ↔ B and the measured current is read out as state 0. Use of an RW memory array according to the two preceding claims, characterized in that the redox-active molecule is rewritten by applying voltage to the electrodes in the reversible oxidation state according to the reaction A ↔ B, so that the redox-active molecule by applying voltage to the electrodes with a read-out potential above and below the normal potential of the redox-active molecule in accordance with the reaction A ↔ B is again reversibly oxidied and reduced and the measured current is read out as state 1 when the threshold value is exceeded. Use of a memory array according to one of the preceding claims 13 to 18, characterized in that provided for ROM memory array a plurality of memory cells with different concentrations of redox-active molecules in the dielectrics and addressed by the electrodes by readout potentials and the measured current when the threshold value is exceeded assigned to different memory states according to the concentrations and for RW memory array and WORM memory array a plurality of memory cells with different concentrations of redox-active molecules in the dielectrics by write potentials reversible (RW) or irreversible (WORM) are generated, and the electrodes addressed by readout potentials and the measured current is assigned to different memory states when the threshold value is exceeded.

Images