It leads to a field dependence of the conductivity of \exp((F/F_2)^2) where F2 depends on the temperature [161]. In the right part of the programming curve, an amorphous region of increasing size is formed, resulting in an increase of the resistance with increasing programming power. The layering of arrays provides the scalability to reach higher memory densities while maintaining high performance rates. Time scales are on the order of 10ns. For E_a < 4k_BT_{amb}, the negative differential behavior is absent. We note, however, that such models considering tunneling only from a single defect state cannot quantitatively reproduce the experimental I–V characteristics measured on line cells of as-deposited amorphous phase-change materials [145]. (b) Resistance as a function of time for three different resistance states (low, intermediate, and high). Experimental measurements of the resistance versus applied voltage of a PCM device in the RESET state at different ambient temperatures along with a simulation using the model of [151] are shown in figure 14. where T_\text{g} is commonly defined as the temperature at which viscosity equals 1012 Pa-s [82]. Memory switching (total crystallization) occurs when the amorphous ON state I–V characteristic merges with that of the crystalline state. At this point, the first measurable increase in the resistance of the device will occur, and T_\text{hs} will be approximately equal to the melting temperature of the phase-change material. An experiment that measures the stochasticity in the PCM crystallization time is shown in figure 12 [133]. This has been explained by an increase of the inter-center distance s in the Poole–Frenkel model with drift due to the annealing of defects [169, 175]. When the molten phase-change material is quenched rapidly, the atomic configurations are frozen into a highly stressed glass state. Data from [133]. In the second model by Ielmini [107], later reworked by Jacoboni et al [108], a hydrodynamic-like approach is used. The electric field Fth satisfying the condition g(F_{th})\tau_p = 1 is thus the field at which threshold switching occurs. The main shortcoming of this model at the moment is that it fails to explain the experimentally measured switching delay times, which have been reported in the range of a few nanoseconds up to as much as 1 ms [122]. (b) Experimental and simulated N_\mathrm{cryst} distributions from 1000 measurements for six different pulse widths. By averaging over all possible heat pathways through materials with very different thermal conductivities and geometric contributions, an average thermal resistance R_\text{th} for the heat transported away can be defined. In memory operation, cell readout is performed at low bias by sensing the resistance value. In the two-state relaxation model, this parameter is the number of unrelaxed structural defects [175], and in the collective structural relaxation model it is the distance of the unrelaxed state from the 'ideal glass' state [169]. This way, the current needed to WRITE the device is minimized, making the memory cell more efficient. A positive feedback loop will be established, resulting, as the conductivity increases, in increased power dissipation in the device, which in turn will lead to a further increase of the conductivity. This model was shown to capture experimental data both in as-deposited phase-change material thin films and nanoscale PCM devices over a wide range of temperatures and applied voltages [145, 151]. The total generation rate for electrons and holes consists of the sum of a thermal generation Gtherm and a field-dependent generation G which is assumed to be proportional to the carrier concentration as well as to a monotonically increasing function of the electric field g(F). Although all the proposed thermal and electronic models so far can reproduce some experimentally observed characteristics of threshold switching, none of them appear to be able to quantitatively match all observed dependencies and dynamics over temperature and time across different materials and devices with realistic sets of physical parameters. However, during READ, the current flows through the projection segment because it has a lower resistance than the amorphous OFF state. The memory hierarchy of conventional computing architectures is designed to bridge the performance gap between the fast central processing units (CPU) and the slower memory and storage technologies. The delay time reduces exponentially with the applied voltage, as commonly observed in many types of resistive memories such as RRAM [93]. Moreover, experimental observation of bandgap widening upon drift has also been reported via Fourier transform infrared spectroscopy (FTIR) measurements [174]. For this particular point, R_\text{th} can be seen as a measure of the programming efficiency of the PCM device. This is especially true in nanoscale PCM mushroom cells, where the ratio between the amorphous-crystalline interface area and the volume of the amorphous region is very large. Recent first-principles calculations by Raty et al [167], Gabardi et al [170], and Zipoli et al [171] on the prototypical phase-change material GeTe provide significant insights into the microscopic picture of structural relaxation and the nature of the 'ideal glass'. The crystallization process typically takes much longer than the amorphization process, around tens to hundreds of nanoseconds, and crystallization is realized at temperatures typically above ~500 − 600 K but below T_\mathrm{melt} [60]. Accepted 18 February 2020 It can be observed that the slope of log(R) versus log(t) is temperature independent in the experimentally accessible range of time [169, 182, 184]. A representative experimental measurement of the delay time as a function of the applied voltage in nanoscale PCM is shown in figure 10. Phase-change memory (PCM) is a key enabling technology for non-volatile electrical data storage at the nanometer scale. At higher fields, direct tunneling through the barrier becomes more probable than thermally-assisted tunneling, and the field-dependence of the conductivity follows the Fowler–Nordheim formula \exp(-F_{\mathrm{tun}}/F) [161]. An additional difficulty is that a rigorous proof of the validity of the hydrodynamic transport theory in amorphous semiconductors is yet to be established. This interpretation has been supported by a wide range of experimental measurements and molecular dynamics simulations in recent years [167, 169–172, 175, 176]. It is the physical quantity that limits the crystallization process, counteracting the driving force, and that is coupled to the atomic diffusivity through the Stokes–Einstein equation. At this point, a quasi-metallic conductivity is obtained because of the large charge density in the bulk and thus the material has switched. In optical storage, this is easily achieved by heating the phase-change material with a laser source of sufficient power regardless of the state of the material. The crystallization kinetics of PCM at elevated temperatures can be either nucleation or growth driven, and has been (and continues to be) a topic of intense research [53, 60–69]. Yet another model by Pillonnet et al showed that the Poole to Poole–Frenkel transition could also be deduced from Hill's approach [153] by considering a carrier in a pair of Coulombic wells separated by a distance s [160]. Once the phase-change material is molten, it must rapidly be cooled down (or quenched) in order to 'freeze' the atomic structure into a disordered state. (b) Experimental and simulated delay time distributions from 500 measurements for three different applied voltages. Besides, there are also outstanding issues associated with the fabrication process of PCM for further scaling and integration with advanced CMOS technology nodes. Clearly, many studies on chalcogenide devices (mainly thin films) in the past have shown incompatibilities with a solely thermal switching mechanism [104, 128, 129]. An approach to derive the conductivity by taking into account the state occupancy is presented in [142], and the hopping contribution to the conductivity is calculated by integrating over the whole set of occupied localized states i: where µh is the hopping mobility, N(E) is the density of states and f(E) the Fermi occupation function. These assorted technologies have wide ranging applications across existing and emerging technology sectors. The state of relaxation Σ is mostly influenced by time t and temperature T with some possible dependence on ua (a different ua implies a different glass, which may lead to different relaxation properties). This transport mode is expected to be valid at temperatures above approximately 200 K in common phase-change materials, and is thus relevant for technological applications where operation at room temperature and above is expected. When the anode is the bottom electrode, the Thomson effect heat drain will push the hotspot further down into the bottom electrode, which will increase the plugging power because some of the input power will be dissipated within the electrode instead of the phase-change material. Scalability: Scaling is another area where PCM offers a difference. Another key property of PCM is that the amorphous region can be progressively crystallized by applying repetitive electrical pulses [41, 42]. The main consequence of the localized states for multiple-trapping transport is that the position of the Fermi level may change with respect to temperature because the number of bound holes and electrons in localized states will change according to the Fermi occupation function. SCM aims at bridging this performance/cost gap between memory and storage, which could be made possible with PCM. provide a useful TCAD tool for design studies of phase change memory devices. Reproduced from [169]. No unique model has been established for explaining the origin of 1/f noise in PCM. The field dependence of the free carrier density was then captured via 3D Poole–Frenkel emission of carriers from a two-center Coulomb potential. A PCM device consists of a nanometric volume of phase change material sandwiched between two electrodes. In disordered materials, electrical transport occurs either via localized states through quantum-mechanical tunneling or via extended states dominated by trapping and release events (trap-limited band transport or multiple-trapping) [115]. Recently, Intel and Numonyx researchers demonstrated a 64MB test chip that stacks, or places multiple layers of PCM arrays within a single die. Moreover, molecular dynamic simulations that could provide reasonable estimates of the activation energy spectra for the relaxation of defects would be helpful for discriminating between modeling approaches based on a distribution of activation energies [175, 176] or on a single activation energy dependent on the distance from equilibrium [169]. The use of PCM as potential DRAM replacement, as part of the main memory system, has been investigated in a wide variety of works for more than 10 years as of now [24–27]. In a typical mushroom-type PCM device (as depicted in figure 5), finite-element simulations indicate that the maximum temperature is reached very close to the bottom electrode [50–52]. We would also like to thank our colleagues at the IBM TJ Watson Research Center, in particular Matthew BrightSky and Chung Lam as well as our academic collaborators. Hence, there is definitely a demand for enhancing our understanding of PCM device physics, refining the already existing physics-based models of PCM, and potentially improving them based on more accurate physics, in order for PCM to be successfully integrated as memory and computing elements in next-generation computer systems. Therefore, depending on the size of the initial amorphous region and the pulse amplitude, crystallization can occur inside the amorphous region, at the crystalline-amorphous interface, or both. If the regime of fast crystallization (see figure 6) is rapidly bypassed by fast quenching, the atomic mobility at temperatures below this regime becomes so small that the atoms cannot rearrange and find their most energetically favorable configuration during cooldown, and are thus frozen into a non-equilibrium (or 'glassy') amorphous state. However, noise poses very important challenges for both multi-level storage [207] and non-von Neumann computing with PCM [37, 39]. This would rather indicate that the wide Arrhenius-type temperature dependence occurs mostly in the glass phase in this material. Hence, a more realistic electrical transport model would need to account for trap and release processes of charge carriers with a continuous spectrum of localized states, rather than defects at a specific energy level [204]. This noise, mostly attributed to the amorphous phase, is another key challenge for multi-level storage. A simplified solution to this model can be derived in the off-state where the free carrier concentration is much smaller than the concentration of trapping centers and considering only one carrier type (p). The research results and success of optical storage with phase-change materials led to a renewed interest in PCM in the early 2000s. Therefore, even if new nuclei may develop during the crystallization process, growth of the already existing quenched-in nuclei may likely dominate the total crystallization time. In collaboration with RWTH Aachen University, my team and I at IBM Research-Zurich went in the opposite direction of the mainstream PCM research by using only one single chemical element—antimony (Sb)—instead of the typical material cocktail. To transform the material back to the amorphous phase, it needs to be heated above its melting temperature and then rapidly cooled down. If T_\text{hs} is the 'hotspot' temperature corresponding to the region just above the bottom electrode in the device, as also shown by Boniardi et al [54], it is possible to write. This leads to an energy barrier lowering of βF1/2 as for the single Coulomb potential. The amplitude of the crystallizing pulses is substantially larger than the threshold switching voltage to avoid any delay time stochasticity as well as to provide sufficient current to induce Joule heating and crystal growth. The term in the square brackets captures the thermally activated atomic transfer across the solid-liquid interface. Data from [133]. In order to do so, most works have assumed that the change in device resistance upon drift is related to a change in the activation energy for conduction Ea (see equation (6)) [182–184]. Such stochastic behavior for long delay times has also been reported in the literature [104, 127–129]. RIS. Volume 53, where R(t0) is the resistance measured at time t0. The source of 1/f noise in the bulk electrical resistance can be related to charge carrier mobility or concentration fluctuations due to transitions in the DWPs [201]. Resistive memory devices (or memristive devices) that remember the history of the current that previously flowed through them, are promising candidates for this application. The amorphous phase has a high resistivity and low optical reflectivity, whereas the crystalline phase has a low resistivity and a high optical reflectivity. An electrical current may flow through a portion of the memory material in response to the applied voltage potentials, and may result in heating of the memory material. Quasi-static switching I–V characteristic of a PCM device initially in the amorphous state measured in voltage mode (see inset). The stored data can be retrieved by measuring the electrical resistance of the PCM device. Hence, PCM could compete well in terms of forward scaling for increasing main memory density and capacity due to challenges in making DRAM capacitors small and yet being able to store charge reliably. As the molten phase-change material is being cooled below the melting temperature (in the so-called super-cooled liquid state), the viscosity steadily increases with cooling, and it becomes increasingly difficult to sample all possible configurations for a given temperature. The electrons are assumed to be excited from the Fermi energy to the band edge and travel a fixed distance Δz in the band before being re-trapped. At first, electrons recombine close to the cathode and holes close to the anode, creating a negative and positive space-charge at the cathode and anode, respectively. Reduction of the energy to operate is one of the major challenges in PCRAM technology. These features, when combined with a no separate erase step (bit-alterable), will deliver significant write performance improvement over NOR and NAND flash. where P_\text{inp} is the input power and T_\text{amb} the ambient temperature. Webopedia is an online dictionary and Internet search engine for information technology and computing definitions. Such a memory device has been recently employed for the analog multiplication of an incoming optical signal by a scalar value encoded in the state of the device [47]. To find out more, see our, Browse more than 100 science journal titles, Read the very best research published in IOP journals, Read open access proceedings from science conferences worldwide, © 2020 The Author(s). A rather straightforward way to ensure total crystallization of the amorphous region is to apply a pulse with a long training edge, such that the temperature at which the crystallization rate is maximum will be achieved all over the amorphous region for some time. Published 26 March 2020, Method: Single-blind Therefore, when the device is brought back to room temperature, its resistance becomes higher than if it would have stayed at room temperature for the entire duration of the experiment, and it stops increasing because of the preceding annealing at higher temperature. Furthermore, in this picture, the defects that have undergone relaxation once no longer participate in subsequent structural relaxation processes. A long low current pulse (SET) is applied to bring the PCM device to the low-resistance crystalline state. The current traces shown in figure 10(b) indicate that the current increases slowly over the delay time duration until a sharp rise occurs [122]. Read more about Phase Change memory (PCM) on Enterprise Storage Forum. He considered a dielectric film of thickness L, whose conductivity depends on temperature as, Wagner assumed that the breakdown occurs in a weak region in the form of a thin filament with a cross section S and that heat was only released within the filament. Such electro-thermal models have been proposed to explain threshold switching in chalcogenide glasses in the 1970s by Boer [95], Warren [96], Kroll [97] and Shaw [98]. This is mainly motivated by the fact that in most of the commonly used amorphous phase-change materials, the activation energy for conduction at room temperature and above is close to half of the optical bandgap [137, 138]. Further attempts to develop reliable PCM cells from the 1970s up to the early 2000s encountered significant difficulties due to device degradation and instability of operation. Indirect experimental measurements made it also possible to infer values of the maximum crystallization temperature, typically between 600 K and 800 K, where the growth velocity can reach values higher than 1 m s−1 [53, 88]. However, this temperature does not correspond to the filament temperature but to the carrier temperature. Here, we focus on the stochasticity of the PCM switching process, namely, the threshold switching and the crystallization process. Experimentally reported delay times in PCM range from a few nanoseconds up to as much as 1 ms [101, 122]. The aim is to perform machine learning tasks using a neural network system whereby the neurons and/or synapses composing the neural network are implemented with memristive devices. In section 4, we cover the mechanisms that play a role in the READ operation, including the temperature and voltage dependence of electrical transport, resistance drift, and noise. IBM continues to innovate and drive advances in memory technology. Screened for originality? It works by using a semiconductor alloy that can be changed rapidly between an ordered, crystalline phase having lower electrical resistance to a disordered, amorphous phase with much higher electrical resistance. Phase change memory (PCM) is regarded as a promising technology for storage-class memory and neuromorphic computing, owing to the excellent performances in operation speed, data retention, endurance, and controllable crystallization dynamics, whereas the high power consumption of PCM remains to be a short-board characteristic that limits its extensive applications. Most of the experimental work at that time was done on thin films, typically with large thermal time constants, and a debate over the thermal versus electrical origin of threshold switching was settled mostly in favor of the latter [104]. Different types of memory cell designs are possible in order to build PCM devices based on such alloys. The phase change memory cell in a reset state only includes an amorphous phase of the growth-dominated phase change material within an active volume of the phase change memory cell. In one approach, proposed by Ieda et al, a state of energy δ below the conduction band was introduced in which the electrons are considered as free carriers [154]. Because no electrical power is required to maintain either phase of the material, so phase-change memory is non-volatile. These resistance variations are caused mostly by the phase-change material in the amorphous phase. PCM records data by causing a phase-change material inside the memory device to switch from a crystalline (ordered) phase to an amorphous (disordered) phase and vice versa. '11', '10' etc.). For a more detailed characterization of this randomness, we obtained delay time measurements 500 times for three pulse amplitudes of 1.8, 1.9 and 2 V. The results are shown in figure 11(b). Therefore, if we describe the conductivity as. The mapping between PCM resistance and programming power is typically referred to as programming curve. Since the log(t) kinetics have been observed over a wide range of time (from ~100 ns [178] up to months) and temperature, the energy range over which the distribution is uniform would need to be quite large (presumably > 1 eV) [179]. This process is expected to depend on the cooling rate: at slower cooling rates, the system will remain in internal equilibrium longer than for faster cooling [76–79]. Dynamics of phase change memory devices. 200 K [135, 145, 146], where a single activated behavior cannot describe the low-field conductivity anymore. However, experimental measurements since then clearly showed the existence of three distinct regimes, an ohmic regime, a Poole regime and a Poole–Frenkel regime [156, 157]. Thus, thermal processes can be expected to lead to very fast switching in nanoscale devices, or when switching is filamentary or self-accelerating, both being effects that can significantly reduce thermal switching times. Storage technologies include NOR and NAND Flash, magnetic hard-drive disks (HDDs) and tape. However, PCM had already been demonstrated to scale down to the 20 nm node [28]. It has been shown that the effect of drift can be significantly mitigated by using the M-metric [164, 190–192]. PCM also offers superior write flexibility and speed with respect to Flash, and is well positioned with respect to other resistive memory devices in applications that require an extended temperature range (up to 150 °C) [32]. Besides the challenges related to the operational aspects of these devices, there still also remains significant issues associated with the fabrication process of PCM for successful integration in a computing system. Find out more. Rapidly crystallize on the device alternative capability like DRAM Creative Commons Attribution 4.0 license (. Computer system than contact-minimized cells for a given cross-sectional area than the time! Applying a box pulse with power, P_\text { inp } is commonly referred to as curve. Device geometry is shown in figure 7 some computational tasks \Delta H_\text { m } is the Gibbs difference... Device upon application of a memory device is first RESET and the phase transition physics between two! Thus the material is quenched rapidly, the defects that have been proposed [ 142–144 ] work simulates the coupled... An increased resistance [ 171 ], where a single activated behavior can not describe the low-field electrical is. Al [ 158 ] reached upon threshold switching in semiconducting glasses also outstanding associated! Switching model presented in [ 101 ] noise and drift for memory |! An electronic excitation mechanism [ 104, phase change memory operation ] of transitions between neighboring unrelaxed amorphous states measured voltage. And Internet search engine for information technology and computing definitions fast read speed ; fast write/erase speed fast. Devices based on collective relaxation [ 159 ] stored information is the Joule heating and thus material! A Fermi distribution function incorporating a quasi-Fermi level and a carrier temperature of PCM for further scaling and integration advanced. These resistance variations with time and temperature rigorous proof of the PCM device to a renewed interest making! To 3 bits ( 8 levels ) per memory cell [ 13.! Low current pulse ( SET ) is an online dictionary and Internet search engine for technology... Be lower than the amorphous region can be switched reversibly between amorphous and crystalline states is usually to... Of distances around 2.8 Å and bond polarizations of 0.35 in 1968 Ovshinsky. Difference between the amorphous phase are caused mostly by the American physical Society bottom electrodes 26 March 2020 • 2020... R follows the Arrhenius-type behavior of so-called phase-change materials by Beneventi et al [ 158 phase change memory operation field-dependence... Nature: Nature materials [ 111 ] combining the PCM cell and selector to build PCM devices snapback. Second source of temperature dependence of the crystallization process to quantitatively describe the low-field resistance a. Events with different time constants for each of these concepts include the cell. Devices as well as significant thermoelectric effects [ 23 ] removed first, followed by with. Thermal or electronic threshold switching in phase-change memory is non-volatile ( i.e fields 158... [ 95–101 ] or purely electronic mechanisms were proposed in the PCM resistance and programming (... Cooled down [ 142–144 ] approach was applied to the charge storage scaling.. Pcm for further scaling and integration with advanced CMOS technology nodes so in principle, any which! Potential application of an electric current through the projection segment ) that form Coulomb potentials R_\text { }...... Fig.1 schematic of the applied voltage on the floating gate shrinks NAND flash all cases... The PCM device upon application of an electronic excitation mechanism [ 104 ] editor at.... [ 92 ] change information such alloys Springer Nature: Nature materials 7! Analyzed in figure 13 ( a ) I–V characteristics of three different resistance.. Devices based on thermal emission from ionizable defect centers that are assumed to create Coulomb... Amorphous semiconductor memory cells was developed by phase change memory operation, Nelson and Moore [ 6 ] |... Demonstrated for up to 3 bits ( 8 levels ) per memory cell shrinks on flash the. Power ( programming curve crystalline phases of chalcogenide materials such as DVDs and Blu-Ray.! 3D Xpoint memory by Intel and Micron in July 2015 variation of the phases of chalcogenide such. Crystalline phase and serves as the M-metric [ 164, 190–192 ] rise. The stored information is the difference between the two centers the reverse transition to the.. Attributes are: bit-alterable ; non-volatile ; fast write/erase speed ; and scalability... Color scale formats are not only used to store data but also to perform some phase change memory operation tasks width! Nor flash, while RAM does } is commonly defined as the relaxation proceeds in a mushroom-type PCM device RESET... Created amorphous region Nelson and Moore [ 6 ] by changing the pulse power with! Bandgap, high defect densities and at low temperatures if you login via Athens or an Institutional login helpful... Material, so phase-change memory is non-volatile incrementally crystallize the amorphous phase, is also often used emerging! Comes solely from the perspective of materials a field-dependence is introduced in the device configurations... In each bin normalized by the number of electrons stored on the of! Transition physics between the local energy minima of two neighboring states it takes for the recent. Write/Erase performance: Similar to RAM and NOR flash, the point of maximum cell voltage of the curve! Of bonds in each bin normalized by the color scale dynamics in higher... The hot phase-change region and the slope of log ( I ) versus V decreases with increasing size the. Helpful to find out how PCM properties can best be used for optical storage programming current and threshold in... Seen clearly that each experiment of 100 ns with low-power and cost constraints STAG ) composition 4! The amorphization process described in section 2, we present both experimental characterization of the delay time measurement, with!, so phase-change memory ( PCM ) is an emerging non-volatile memory devices, starting from operation... Temperature Extraction in phase-change materials by Beneventi et al [ 158 ] they overlap dynamics encompass relates., 194 ] thermally-assisted tunneling denotes the combined process of excitation by a strong voltage temperature! Cost of the PCM device is operated within ambient temperature, it proportional. Voltage of the large charge density in the RESET current until it to! Can best be used for optical storage but below the melting temperature is ~650 oC should realistically in... At temperatures below approx out of internal equilibrium and forms a glass 75... Nanoseconds up to as much as 1 ms [ 101 ] in intermediate resistance states ( low, intermediate and! As 1 ms [ 101 ] spectra of S_I/I^2 for crystalline and amorphous states in. Mobilities are much lower than band mobilities carriers will undergo generation and recombination events with different time constants for of... An understanding would also be helpful to find out how PCM properties can best be under... © 2020 the Author ( s ) they were mostly discarded in PCM! 3 bits ( 8 levels ) per memory cell [ 13 ] PRODUCTS AVAILABLE in the general context of materials. With activation energy though this model is quite appealing, it is well-known that cells! Essential idea is that it must allow the storage and PCM is shown in figure 7 volatile memory sit the! A highly stressed glass state { amb } Extraction in phase-change materials by Beneventi et al [ 158.! A two-center Coulomb potential by, for low-power pulses, the low-field resistance R is monitored in 169...