METHOD FOR PROGRAMMING A MEMORY WITH PHASE CHANGE

Description

Domaine technique

[0001] The present invention relates generally to non-volatile memories, more specifically to phase change memories (or PCM from the English Phase Change Memory) and more particularly to the memory programming method. Technique antérieure

[0002] Phase-change memory (PCM) is a type of non-volatile memory that relies on the properties of phase-change materials. These materials have the ability to switch from a low-resistance state to a high-resistance state through heating. PCMs take advantage of the fact that the electrical resistances of the different states of these materials vary to store data.

[0003] Programming phase-change memories is generally lengthy, so it would be desirable to reduce this programming time in order to increase programming speed.

[0004] US patents 8634235 B2 and 2010 / 097851 A1 describe methods for programming a multilevel phase-change memory device. Summary of the invention

[0005] There is a need to improve the processes for writing data to phase-change memories.

[0006] One embodiment overcomes all or part of the disadvantages of known processes.

[0007] The invention provides a method for programming a phase-change memory comprising a first layer of a phase-change material capable of switching between a crystalline state and an amorphous state and vice versa, the method comprising applying a programming current through the first layer such that the surface density of this current decreases with time t from a first level, between a first instant and a second instant, following a first time evolution respecting, or being close to, J 0 t = K t where K is a constant.

[0008] The first development is close to J 0 t = K t if, when superimposed on it, it creates one or more surfaces between the first evolution and J 0 t = K t , and the ratio of the area of ​​each surface to the integral of J 0 t = K t The difference between the first and second times is less than 0.2.

[0009] According to one embodiment, the first evolution cuts J 0 t = K t at least at two points between the first and second moments.

[0010] According to one embodiment, before the first instant, the evolution of the surface density of the current as a function of time t follows a second linear increasing evolution between a third instant and a fourth instant, then a third constant evolution equal to the first level between the fourth instant and the first instant.

[0011] According to one embodiment, the first evolution comprises several strictly decreasing linear portions and / or constant proportions, between the first instant and the second instant.

[0012] According to one embodiment, the first layer rests on a heating element or an electrode.

[0013] According to one embodiment, the first layer rests on a second layer based on germanium nitride.

[0014] According to one embodiment, the programming current is applied through the first layer and a channel in the second layer, a contact area between the channel and the lower face of the first layer having an area smaller than that of the lower face of the first layer.

[0015] According to one embodiment, the channel is a confined area of ​​a phase-change material and is located in a filament of the second layer, the process comprising, after the second instant, a crystallization step of the confined area.

[0016] According to one embodiment, crystallization of the confined zone is carried out by a fourth constant linear evolution of the current density as a function of time, over a first duration and in which the first duration is equal to t filament = h filament v g with h filament the average height of the confined area and with v g the crystallization rate of the phase-change material.

[0017] According to one embodiment, the crystallization of the confined zone is carried out by a strictly decreasing linear evolution of the current density as a function of time.

[0018] According to one embodiment, K = 1 H Or H = v g . R th . A 2 . R h h . T melt − T h with R th the thermal resistance of the cell, R h the electrical resistance of the phase-change material in the filament, A the surface area of ​​the contact zone between the filament and the underside of the first layer, h the maximum height of the phase-change material, T meit the melting temperature of the phase-change material, and with T h the temperature of the first layer on its upper face, that is to say on the face opposite to the lower face.

[0019] According to one embodiment, K = 1 H ′ Or H ′ = v g . R th . A 2 . R h ′ h . T melt − T h with R th the thermal resistance of the cell, R h ' the electrical resistance of the heating element, A the surface area of ​​the contact zone between the electrode or heating element and the underside of the first layer, h the maximum height of the phase-change material, T meit the melting temperature of the phase-change material, and with T h the temperature of the first layer on its upper face, that is to say on the face opposite to the lower face.

[0020] According to one embodiment, the duration between the first and second instants is equal to a second duration t d ô me = h . T melt − T h v g . T melt .

[0021] According to one embodiment, the duration between the first and second instants is between 1 ns and 1 ps.

[0022] According to one embodiment, the first level is a maximum level that the current density reaches through the cell during the programming cycles.

[0023] According to one embodiment, the application of the programming current corresponds to a crystallization of a part of the first layer. Brève description des dessins

[0024] These features and advantages, as well as others, will be described in detail in the following description of particular embodiments, given by way of non-limiting example, in relation to the attached figures, among which: there figure 1 represents, through a partial and schematic perspective view, a phase-change memory device; the figure 2 represents, through a partial and schematic cross-sectional view, the phase-change device of the figure 1 ; there figure 3 represents, through a partial and schematic cross-sectional view, another embodiment of a phase-change device; the figure 4 represents, through a partial and schematic cross-sectional view, yet another embodiment of a phase-change device; the figure 5 is a block diagram illustrating a phase-change memory system; the figure 6 is a graph representing an example of the evolution of the current density propagating in the device illustrated in figure 3 depending on time in order to switch some of the materials in this device from a resistive state to a less resistive state; the figure 7 is a graph representing an example of the evolution of the current density propagating in the device illustrated in figure 3 depending on time in order to switch some of the materials in this device from a resistive state to a less resistive state; the figure 8 is a graph representing an example of the evolution of the current density propagating in the device illustrated in figure 3 depending on time in order to switch some of the materials in this device from a resistive state to a less resistive state; the figure 9 is a graph representing an example of the evolution of the current density propagating in the device illustrated in figure 3 depending on time in order to switch some of the materials in this device from a resistive state to a less resistive state; the figure 10 is a graph representing an example of the evolution of the current density propagating in the device illustrated in figure 3 depending on time in order to switch some of the materials in this device from a resistive state to a less resistive state; the figure 11 is a graph representing an example of the evolution of the current density propagating in the device illustrated in figure 2 depending on time in order to switch some of the materials in this device from a resistive state to a less resistive state; and the figure 12 is a graph representing an example of the evolution of the current density propagated in the device illustrated in figure 2 depending on the time in order to switch a part of the materials of this device from a resistive state to a less resistive state. Description des modes de réalisation

[0025] The same elements have been designated by the same reference numerals in the different figures. In particular, structural and / or functional elements common to the different embodiments may have the same reference numerals and may have identical structural, dimensional and material properties.

[0026] For the sake of clarity, only the steps and elements useful for understanding the implementation methods described have been represented and are detailed.

[0027] Unless otherwise specified, when referring to two connected elements, this means directly connected without any intermediate elements other than conductors, and when referring to two coupled elements, this means that these two elements can be connected or linked through one or more other elements.

[0028] In the description that follows, when referring to absolute positional qualifiers, such as the terms "front", "back", "top", "bottom", "left", "right", etc., or relative positional qualifiers, such as the terms "above", "below", "superior", "inferior", etc., or to orientational qualifiers, such as the terms "horizontal", "vertical", etc., unless otherwise specified, it refers to the orientation of the figures.

[0029] Unless otherwise specified, the expressions "approximately", "roughly", "about", and "on the order of" mean within 10%, preferably within 5%.

[0030] There figure 1 represents by a partial and schematic perspective view, a 2-memory phase-change device.

[0031] More specifically, the figure 1 represents a phase change memory device comprising a first electrode 23, which corresponds for example to a high electrode, a layer 25 of a phase change material and a second electrode 21, which corresponds for example to a low electrode.

[0032] The two electrodes 23 and 21 are, for example, metallic.

[0033] The phase change material is, for example, a GST material, that is, a material composed of germanium (Ge), antimony (Sb) and tellurium (Te).

[0034] The device 2 further includes a heating element (heater) 24 between the second electrode 21 and the layer 25 of the phase-change material.

[0035] The heating element 24 is, for example, made of tungsten or a titanium or tantalum-based alloy.

[0036] The phase change material has the ability to switch from a resistive state (HRS, High Resistive State) to a weakly resistive state (LRS, Low Resistive State) under the effect of heat and the Joule effect generated by the application of a current between electrodes 21 and 23. The switch from a resistive state to a weakly resistive state corresponds to an activation operation, called SET, i.e. an operation of writing a binary data value, for example a logic value "1", and the switch from a weakly resistive state to a resistive state corresponds to a deactivation operation, called RESET, i.e. an operation of writing an opposite binary data value, for example a logic value "0".

[0037] In this description, the so-called "resistive" state can, for example, be an amorphous state, that is, a disordered state with high electrical resistance, and the so-called "less resistive" or "weakly resistive" state can, for example, be a crystalline or partially crystalline state, that is, an ordered state with lower electrical resistance.

[0038] According to the embodiment illustrated in figure 2 , the volume of layer 25 affected by the phase changes corresponds, for example, to a dome 31. In the example of the figure 1 The heating element 24 is cylindrical in shape. In a variant, the heating element 24 is rectangular in shape. The heating element is, for example, surrounded by an insulator 26, usually SiO2, so as to increase the thermal resistance of the device and prevent heat dissipation through the edges of the heating element 24.

[0039] There figure 2 is a partial and schematic cross-sectional view of device 2, the section passing through heating element 24.

[0040] The upper face of the heating element 24 has a smaller area than the lower face of the layer 25.

[0041] Layer 25, for example, has a thickness e1 between 10 nm and 200 nm and is equal, for example, to approximately 50 nm. Layer 25 has, for example, a width between 3 nm and 500 nm and is equal, for example, to approximately 50 nm.

[0042] In figure 2 The heating element 24 has been represented as a layer. However, in other embodiments, it can have any shape, for example, the shape of an "L".

[0043] According to one variant of the embodiment of the figure 2 , device 2 does not include a lower electrode 21 and the heating element 24 acts as an electrode or vice versa.

[0044] There figure 3 is a partial, schematic cross-sectional view of a variant embodiment of device 2 illustrated in figure 2 .

[0045] More specifically, the figure 3 illustrates a device 3 similar to device 2 illustrated in figure 2 with the difference that device 3 includes a layer 27, between layer 25 and the lower electrode 21.

[0046] Layer 27 is located, for example, on and in contact with the heating element 24 and layer 25 is located between layer 27 and electrode 23.

[0047] In one embodiment, layer 27 is composed of a germanium nitride (GeN)-based material or a dielectric material. A filament is created in layer 27 by a process known as breakdown. The filament extends, for example, from the bottom face of layer 27 to the top face of layer 27. Layer 27 is, for example, composed entirely of germanium nitride. The filament is created during a step at the beginning of the life of device 3, called the forming step, during which a voltage is applied to device 3. As soon as this voltage reaches a breaking voltage, the filament is created in layer 27. During forming, the current is below a threshold value at which the materials of layers 25 and 27 liquefy and fuse.

[0048] The nitrogen content in layer 27 is chosen based on the desired tensile strength and electrical resistance of the layer material. In one embodiment, the nitrogen content is between 25% and 35%. Layer 27 has a thickness e3 chosen, for example, such that the tensile strength is sufficiently low without significant variability. The thickness e3 of layer 27 is then, for example, between 2 nm and 30 nm, and in some examples, between 3 nm and 25 nm.

[0049] During the initialization step, and following the creation of the filament, part of the phase change material of layer 25 is liquefied and fills the inside of the filament to form a volume 29 called "confined zone" or channel.

[0050] According to another embodiment, layer 27 is a layer in a dielectric which includes a confined area 29 formed, for example, by etching followed by filling with a phase-change material, for example, the same material as layer 25.

[0051] The PCM 25 layer, for example, has a thickness e1 between 10 nm and 200 nm and is equal, for example, to approximately 50 nm. According to the embodiment illustrated in figure 3 , layer 25 and layer 27 have, for example, a width substantially equal to the width of electrode 23, electrode 21 and, possibly, heating element 24. The PCM 25 and GeN 27 layers have, for example, a width between 3 nm and 500 nm and equal, for example, to about 50 nm.

[0052] According to the embodiment illustrated in figure 3 Volume 31 is located opposite volume 29. The surface area of ​​the base of dome 31 is slightly larger than the surface area of ​​the upper face of volume 29.

[0053] One advantage that appears from the presence of a germanium nitride layer between the phase memory material layer and the bottom electrode is that germanium nitride has high physico-chemical compatibility.

[0054] There figure 4 is a partial, schematic cross-sectional view of a variant embodiment of device 3 illustrated in figure 3 .

[0055] More specifically, the figure 4 illustrates a device 4 similar to device 3 illustrated in figure 3 the difference being that it includes a layer 35 between layer 27 and heating element 24.

[0056] According to the embodiment illustrated in figure 4 , layer 35 is composed of a phase change material, for example, of the same nature as layer 25.

[0057] According to the embodiment illustrated in figure 4 A volume 37 of layer 35 is also involved in the switching between the resistive and weakly resistive phases. Volume 37 has the shape of an inverted dome, meaning that its base is at the interface between layer 27 and layer 35.

[0058] Device 4 thus comprises, when the temperature increases, two domes in head-to-tail, that is to say that their flat faces, or bases, face each other.

[0059] To perform the RESET operation, the dome or volume 31 illustrated in figures 2 à 4 is switched into a resistive state and to perform the SET operation, the dome or volume 31 is switched into a weakly resistive state.

[0060] Each of the SET and RESET operations, if it follows a previous RESET operation, includes for example a first so-called electronic transition during which a relatively long (a few nanoseconds) and low intensity (but voltage greater than the threshold voltage) electrical pulse is applied through electrodes 21 and 23. This first transition aims to switch the volume 31 from a resistive state to a conductive state, also called the "threshold" transition, the conductive state allowing the passage of current and then the driving of heating.

[0061] Subsequently, a second so-called "phase" transition is implemented either within the framework of a SET operation during which the phase-change material switches from a resistive state to a less resistive state, or within the framework of a RESET operation, where the phase-change material switches from a less resistive state to a resistive state.

[0062] According to one embodiment, the evolution of the current density J 0 ( t ) propagated through devices 2, 3 and 4 over time in order to switch volume 31 into a less resistive state (SET operation) includes: a step of increasing the current density as a function of time in order to make the material of volume 31 conductive; and a step of monotonically decreasing the current density as a function of time in order to switch the material of volume 31 into a less resistive state.

[0063] If, during the second time, the speed is not slow enough, volume 31 returns to a resistive state.

[0064] The current applied between the two electrodes 21 and 23 and the surface current density are related by the formula: I prog = A . J prog in which I prog is the programming current 27 applied between electrodes 21 and 23, I prog is the programming current density, and A is the surface mentioned above and A is a surface defined as: the area of ​​the contact zone between the volume 31 and the heating element 24 or the electrode 21 in the case of device 2; or the area of ​​the smallest section of the channel 29 of the layer 27 in the case of devices 3 and 4.

[0065] One advantage of having a filament layer between the phase shape memory material layer and the base electrode is that the surface area A is reduced. The same current density can therefore be achieved with a lower programming current.

[0066] One advantage that appears with the presence of layer 35 between layer 27 and heating element 24 is that the heat losses from the interface between layer 27 and element 24 are reduced thanks to the relatively low thermal conductivity of the material of layer 35, thus increasing thermal insulation.

[0067] There figure 5 is a block diagram illustrating a phase change memory circuit 5 comprising a part dedicated to phase change memories and a part dedicated to programming them.

[0068] More specifically, the system 5 includes a programming unit 45 and a matrix 46 of one or more phase-change memory cells. Each cell of the matrix 46 includes a device as illustrated in figures 1 à 4 The devices are, for example, electrically connected in parallel with each other.

[0069] According to the embodiment illustrated in figure 5 The programming unit 45 generates a current I prog which passes through each cell of matrix 46. Within each cell, the device is traversed by a surface current density J prog , corresponding to the current I prog divided by the area A.

[0070] According to one embodiment, the evolution of the current density propagating in the device as a function of time in order to switch the volume 31 totally or partially from a resistive state to a less resistive state (SET operation) comprises: a step of increasing the current density over time to make the material in volume 31 conductive; and a step of nonlinearly decreasing the current density over time to make the material in volume 31 less resistive, this step being described in more detail below in relation to the figures 6 à 12 .

[0071] As an example, the non-linear decrease in current density over time is controlled by a digital-to-analog converter that is part of programming unit 45. In particular, the shape of the decrease is, for example, defined by a sequence of numerical values, which are subsequently converted into an analog control signal by this converter.

[0072] There figure 6 represents one embodiment of the evolution of the current density propagated in device 3 illustrated in figure 3 depending on time, in order to switch a portion of the phase-change materials in this device from a resistive state to a less resistive state. This embodiment applies to devices similar to device 3 of the figure 3 including a confined area, i.e. a channel with a width much less than the width of layer 25.

[0073] More specifically, the figure 6 represents a curve 48 illustrating the evolution of the surface current density J 0 ( t ) in channel 29, imposed by a current I(t) passing through the electrodes as a function of time.

[0074] According to the embodiment illustrated in figure 6 Curve 48 includes: a first portion of linear growth of the current density (second evolution) between time t0 and time t1; a second portion of approximately constant current density (third evolution), at the level J0_max between time t1 and time t2; a third portion of non-linear decrease of the current density (first evolution) between time t2 and time t3, the time between times t2 and t3 being called t dome ; and a fourth portion of approximately constant current density (fourth evolution), at the J0_wire level, between time t3 and time t4, the duration between times t3 and t4 being named t filament .

[0075] There figure 7 represents the same curve 48 as the figure 6 and further shows the structural evolution of the change of state of volumes 31 and 29 of the phase-change material.

[0076] At time t0, the phase-change material in volume 31 is resistive and the current density is, for example, at level 0. Between times t0 and t1, the current density increases so as to reach, at time t1, a current density value J0_max which will remain essentially constant until time t2 (point B). Between times t0 and t2, the phase-change material in volume 31 becomes conductive.

[0077] Between time t2 and time t3, during the duration t dome , The phase-change material of volume 31, in particular the dome corresponding to volume 31, switches to a less resistive state. The phase change of volume 31 occurs from the outside of volume 31 to the inside of volume 31, that is, from the domed face of the dome to its flat face.

[0078] During the duration t dome , Curve 48 decreases, meaning the current density decreases over time. To account for a complete phase change of volume 31, the crystallization growth rate for a dome-shaped volume follows the formula below: ∂ J 0 t ∂ t < v g 2 h . T melt − T h δ J 0 t 2 − T h 2 δ J 0 t in which: v g is the crystallization rate of the phase-change material, h is the height of the layer of the phase-change material, T meit is the melting temperature of the phase-change material and T h is the temperature at the interface between the second electrode and the layer of the phase-change material and δ follows the formula: δ = R th . A 2 . R h in which R th is the thermal resistance of the device and R h is the electrical resistance of the dome material 29.

[0079] Thus, if δ ( J 0 ( t )) 2< » T h , The formula [Math 2] can be simplified to give: ∂ J 0 t ∂ t < v g δ J 0 t 3 2 h . T melt − T h or − H 2 J 0 t 3 < ∂ J 0 t ∂ t < 0 in which: H = v g . R th . A 2 . R h h . T melt − T h .

[0080] As an example, the solution to formula [Math 5] is: J 0 t = K 1 t where K is a constant and is the slope of the curve. In some embodiments, the constant K is equal to: K = 1 H hence J 0 t = 1 Ht .

[0081] The switchover to a less resistive state is complete, for example, when the following formula is verified: J 0 t = T melt R th . A 2 . R h .

[0082] So, t dome check the equation: t dome = h . T melt − T h v g . T melt .

[0083] So t dome is optimized for each phase-change material. For example, the duration t dome is between 1 ns and 1 ps.

[0084] Between times t3 and t4, during the duration t filament , The phase-change material of the dome 29, that is, the material present inside the filament, switches to a less resistive state. The phase change of the dome 29 takes place, for example, from the upper face of layer 27, that is, from the interface between layer 27 and layer 25, towards the lower face of layer 27, that is, towards the interface between layer 27 and element 24.

[0085] During the duration t filament , according to the embodiment illustrated in figures 6 et 7 The curve 48 is constant, and for example at level J 0_fil. The duration t filament depends on the height h filament of layer 27 and therefore of the filament, and corresponds to: t filament = h filament v g .

[0086] At time t4, the phase-change material in volumes 29 and 31 transitions to a low-resistance state, for example, is complete, and the current density decreases. The device is thus ready for a new writing cycle.

[0087] For example, the duration t filament is between 1 ns and 10 ps for a filament with a height between 5 nm and 30 nm.

[0088] Within the framework of a device comprising a layer 35 and a double dome as illustrated in device 4 shown in figure 4 , volume 37 switches into a less resistive state between times t 2 and t 3.

[0089] One advantage of monitoring the evolution of current density over time, as described above, is that it allows for the optimization of crystallization in the active zone of the device. In other words, it enables complete crystallization of the active zones in a shorter time and using a lower current compared to existing techniques. Another advantage of monitoring the evolution of current density over time, as described above, is that it allows for precise control of the crystallization of the phase-change material volumes, ensuring complete crystallization within the considered volumes.

[0090] The graph illustrated in figures 6 et 7 has been described and explained by considering a 27-layer germanium nitride in which a filament has been formed, however this evolution also adapts to a device comprising a 27-layer dielectric in which a PCM channel has been formed by etching.

[0091] THE figures 8 à 10 represent other ways of implementing the evolution of the current density propagated in devices 3 and 4 illustrated in figures 3 et 4 depending on the time in order to switch a part of the phase change materials of these devices from a resistive state to a weakly resistive state.

[0092] More specifically, the figure 8 represents a curve 49 illustrating a method of realizing the evolution of current density as a function of time similar to that illustrated in figures 6 et 7 with the difference that curve 49 is represented in figure 10 is different, but close, to curve 48 between times t 2 and t 3.

[0093] The example of the figure 8 is based on a step-like decay. Such a form is, for example, simple to implement using a digital-to-analog converter as described in relation to the figure 5 .

[0094] According to the embodiment illustrated in figure 8 The curve 49 passes through points B and C at times t2 and t3 respectively. figure 8 Between times t2 and t3, curve 49 is segmented into decreasing segments. In other words, between times t2 and t3, curve 49 corresponds, for example, to a succession of constant portions and portions of increasingly lower current density levels. figure 8 Curve 49 intersects curve 48, between points B and C, three times at points D1, D2, and D3. According to the embodiment, curve 49 intersects curve 48 between times t2 and t3, forming surfaces between these two curves. figure 8 , between curves 48 and 49, there is a surface S1 between points B and D1, a surface S2 between points D1 and D2, a surface S3 between points D2 and D3 and a surface S4 between points D3 and C.

[0095] For example, each of the surfaces S1, S2, S3 and S4 has an area less than 20% of the integral of the curve 48 between times t2 and t3. In some cases, each of the surfaces S1, S2, S3 and S4 has an area less than 10% of the integral of the curve 48 between times t2 and t3.

[0096] In figure 8 , curve 49 intersects curve 48 three times between times t 2 and t 3 to form four surfaces however in practice, curve 49 can intersect curve 48 once, twice or more than twice so as to form two surfaces, three surfaces or more than three surfaces between curves 48 and 49.

[0097] Similarly, the figure 9 represents a curve 51 illustrating a method of realizing the evolution of current density as a function of time similar to that illustrated in figures 6 et 7 with the difference that curve 51 is represented in figure 9 is different from curve 48 between times t2 and t3.

[0098] According to the embodiment illustrated in figure 9 The curve 51 passes through points B and C at times t2 and t3 respectively. figure 11 Between times t2 and t3, curve 51 is segmented into strictly decreasing portions and intersects curve 48, between points B and C, twice at points D4 and D5. According to the embodiment, curve 51 intersects curve 48 between times t2 and t3, forming surfaces between these two curves. figure 9 , between curves 48 and 51, there is a surface S5 between points B and D4, a surface S6 between points D4 and D5 and a surface S7 between points D5 and C.

[0099] Each of the surfaces S5, S6 and S7 has an area less than 10% of the integral of the curve 48 between times t2 and t3. For example, each of the surfaces S5, S6 and S7 has an area less than 5% of the integral of the curve 48 between times t2 and t3.

[0100] In figure 9 , curve 51 intersects curve 48 twice between times t 2 and t 3 to form three surfaces however in practice, curve 51 can intersect curve 48 once, twice or more than twice so as to form two surfaces, three surfaces or more than three surfaces between curves 48 and 51.

[0101] As an example, the average slope of the first evolution between B and C is close to a slope that respects the relationship J 0 t = K 1 t .

[0102] Similarly, the figure 10 represents a curve 53 illustrating a method of realizing the evolution of current density as a function of time similar to that illustrated in figure 9 with the difference that curve 53 is represented in figure 10 is different from curve 51 after time t 3.

[0103] According to the embodiment illustrated in figure 10 , curve 53 is decreasing after time t 4 .

[0104] There figure 11represents a way of implementing the evolution of current density J 0 ( t ) propagated in device 2 illustrated in figure 2 depending on time and more specifically, in volume 31, in order to switch some of the materials of these devices from a resistive state to a weakly resistive state.

[0105] More specifically, the figure 11 represents a curve 47 illustrating the evolution of the current density J 0 ( t ), imposed by a current I ( t ) passing through the electrodes as a function of time.

[0106] According to the embodiment illustrated in figure 11 Curve 47 includes, similarly to curve 48 illustrated in figures 6 and 7 : a first portion of linear growth of the current density between time t 0 and time t 1; a second portion of current density substantially constant between time t 1 and time t 2; and a third portion of non-linear decrease of the current density between time t 2 and time t 3.

[0107] There figure 12 represents the same curve 47 as the figure 11 and further shows the structural evolution of the change of state in volume 31 of phase-change material.

[0108] Similar to curve 48, during the duration t dome Curve 47 decreases, meaning the current density decreases over time. To ensure good crystallization of volume 31, the current density satisfies, for example, the formula: J 0 t = K 1 t in which: K = 1 H ′ avec : H = v g . R th . A 2 . R h ′ h . T melt − T h with R h ' is the resistance of the heating element 24.

[0109] So, t dome check the equation: t dome = h . T melt − T h v g . T melt .

[0110] For example, the duration t dome is between 1 ns and 1 µs.

[0111] At time t3, all the phase-change material present in volume 31 switches to a low-resistance state. The device is thus ready for a new writing cycle.

[0112] One advantage of the described embodiments and implementation methods is that they are compatible with the usual production lines for microelectronic components.

[0113] Various embodiments and variants have been described. Those skilled in the art will understand that certain features of these various embodiments and variants could be combined, and other variants will become apparent to them. In particular, the variants of curve 48 illustrated in figures 8 to 10 can be applied to the evolution of curve 47 represented in Figures 11 and 12 .

[0114] Finally, the practical implementation of the described methods and variants is within the reach of the person in the trade, based on the functional indications given above.

Claims

1. A method of programming a phase change memory comprising a first layer (25) of a phase change material capable of switching between a crystalline and an amorphous state and vice versa, the method comprising the application of a programming current through the first layer such that an evolution of the areal current density (J0) as a function of time t decreases from a first level (J0_max), between a first time instant (t2) and a second time instant (t3), following a first evolution in time respecting, or being close to, J0(t) = K t where K is a constant, characterized in that, the first evolution is close to J0(t) = K t if it creates, when superimposed thereon, one or more surfaces between the first evolution and J 0 t = K t , and the ratio of the area of each surface to the integral of J 0 t = K t between the first and the second time instants (t2 , t3) is less than 0.2.

2. The method according to claim 1, wherein the first evolution intersects J 0 t = K t at least at two points between the first and second time instants (t2, t3).

3. The method according to claim 1 or 2, wherein, prior to the first time instant (t2) the evolution of the areal current density (J0) as a function of time t, follows a second linear increasing evolution between a third time instant (t0) and a fourth time instant (t1) then a third constant evolution equal to the first level between the fourth time instant (t1) and the first time instant (t2).

4. The method according to any one of claims 1 to 3, wherein the first evolution comprises several strictly decreasing linear portions and / or constant proportions, between the first time instant (t2) and the second time instant (t3).

5. The method according to any one of claims 1 to 4, wherein the first layer (25) rests on a heating element (24) or on an electrode (21).

6. The method according to any one of claims 1 to 4, wherein the first layer (25) rests on a second germanium nitride based layer (27).

7. The method according to claim 6, wherein the programming current is applied through the first layer (25) and a channel in the second layer (27), a contact zone (A) between the channel and the lower face of the first layer having a smaller area than the lower face of the first layer.

8. The method according to claim 7, wherein the channel is a confined zone of a phase change material and is located in a filament of the second layer, the method comprising, after the second time instant (t3), a step of crystallizing the confined zone.

9. The method according to claim 8, wherein a crystallization of the confined zone (29) is brought about by a fourth constant linear evolution of the current density (J0) as a function of time (t), over a first duration (tfilament) and wherein the first duration is equal to t filament = h filament v g with hfilament the average height of the confined zone (29) and with vg the crystallization velocity of the phase change material.

10. The method according to any one of claims 8 to 9, wherein the crystallization of the confined zone (29) is brought about by a strictly decreasing linear evolution of the current density (J0) with time (t) .

11. The method according to any one of claims 8 to 10, wherein K = 1 H where H = v g . R th . A 2 . R h h . T melt − T h with Rth the thermal resistance of the cell, Rh the electrical resistance of the phase change material in the filament, A the area of the contact zone (A) between the filament and the lower face of the first layer, h the maximum height of the phase change material, Tmeit the melting temperature of the phase change material, and with Th the temperature of the first layer at its upper face, i.e., at the side opposite the lower face.

12. The method of claim 5, wherein K = 1 H ′ where H ′ = v g . R th . A 2 . R h ′ h . T melt − T h with Rth the thermal resistance of the cell, Rh' the electrical resistance of the heating element, A the area of the contact zone (A) between the electrode or heating element and the lower face of the first layer, h the maximum height of the phase change material, Tmelt the melting temperature of the phase change material, and with Th the temperature of the first layer at its upper face, i.e., at the face opposite the lower face.

13. The method according to claim 5 or 12, wherein the duration between the first and second time instants (t2, t3), is equal to a second duration t d ô me = h . T melt − T h v g . T melt .

14. The method according to any one of claims 1 to 13, wherein the duration (tdôme) between the first and second time instants (t2, t3) is between 1 ns and 1 µs.

15. The method according to any one of claims 1 to 14, wherein the first level is a maximum current density level that the cell reaches during the programming cycles.

16. The method according to any one of claims 1 to 15, wherein the application of the programming current (I) corresponds to a crystallization (SET) of a portion of the first layer.