Sanchit Deshmukh

Towards ultimate scaling limits of phase-change memory

Data storage based on a reversible material phase transition (e.g. amorphous to crystalline) has been studied for nearly five decades. Yet, it was only during the past five years that some phase-change memory technologies (e.g. GeSbTe) have been approaching the physical scaling limits of the smallest possible memory cell. Here we review recent results from our group and others, which have achieved sub-10 nm scale PCM with switching energy approaching single femtojoules per bit. Fundamental limits could be as low as single attojoules per cubic nanometer of the memory material, although approaching such limits in practice appears strongly limited by electrical and thermal parasitics, i.e. contacts and interfaces.

Spatially Resolved Thermometry of Resistive Memory Devices

The operation of resistive and phase-change memory (RRAM and PCM) is controlled by highly localized self-heating effects, yet detailed studies of their temperature are rare due to challenges of nanoscale thermometry. Here we show that the combination of Raman thermometry and scanning thermal microscopy (SThM) can enable such measurements with high spatial resolution. We report temperature-dependent Raman spectra of HfO2, TiO2 and Ge2Sb2Te5 (GST) films, and demonstrate direct measurements of temperature profiles in lateral PCM devices. Our measurements reveal that electrical and thermal interfaces dominate the operation of such devices, uncovering a thermal boundary resistance of 28 ± 8 m2K/GW at GST-SiO2 interfaces and an effective thermopower 350 ± 50 µV/K at GST-Pt interfaces. We also discuss possible pathways to apply Raman thermometry and SThM techniques to nanoscale and vertical resistive memory devices.

Effect of oxygen vacancies and strain on the phonon spectrum of HfO2 thin films

The effect of strain and oxygen deficiency on the Raman spectrum of monoclinic HfO2 is investigated theoretically using first-principles calculations. 1% in-plane compressive strain applied to a and c axes is found to blue shift the phonon frequencies, while 1% tensile strain does the opposite. The simulations are compared, and good agreement is found with the experimental results of Raman frequencies greater than 110 cm−1 for 50 nm HfO2 thin films. Several Raman modes measured below 110 cm−1 and previously assigned to HfO2 are found to be rotational modes of gases present in air ambient (nitrogen and oxygen). However, localized vibrational modes introduced by threefold-coordinated oxygen (O3) vacancies are identified at 96.4 cm−1 computationally. These results are important for a deeper understanding of vibrational modes in HfO2, which has technological applications in transistors and particularly in resistive random-access memory whose operation relies on oxygen-deficient HfOx.

SANTA: Self-aligned nanotrench ablation via Joule heating for probing sub-20 nm devices

Manipulating materials at the nanometer scale is challenging, particularly if alignment with nanoscale electrodes is desired. Here, we describe a lithography-free, self-aligned nanotrench ablation (SANTA) technique to create nanoscale “trenches” in a polymer like poly(methyl methacrylate) (PMMA). The nanotrenches are self-aligned with carbon nanotube (CNT) or graphene ribbon electrodes through a simple Joule heating process. Using simulations and experiments we investigated how the Joule power, ambient temperature, PMMA thickness, and substrate properties affect the spatial resolution of this technique. We achieved sub-20 nm nanotrenches, for the first time, by lowering the ambient temperature and reducing the PMMA thickness. We also demonstrated a functioning nanoscale resistive memory (RRAM) bit selfaligned with a CNT control device, achieved through the SANTA approach. This technique provides an elegant and inexpensive method to probe nanoscale devices using self-aligned electrodes, without the use of conventional alignment or lithography steps.

Thermal transport across graphene step junctions

Step junctions are often present in layered materials, i.e. where single-layer regions meet multi-layer regions, yet their effect on thermal transport is not understood to date. Here, we measure heat flow across graphene junctions (GJs) from monolayer to bilayer graphene, as well as bilayer to four-layer graphene for the first time, in both heat flow directions. The thermal conductance of the monolayer-bilayer GJ device ranges from ~0.5 to 9.1x10^8 Wm-2K-1 between 50 K to 300 K. Atomistic simulations of such GJ device reveal that graphene layers are relatively decoupled, and the low thermal conductance of the device is determined by the resistance between the two dis-tinct graphene layers. In these conditions the junction plays a negligible effect. To prove that the decoupling between layers controls thermal transport in the junction, the heat flow in both directions was measured, showing no evidence of thermal asymmetry or rectification (within experimental error bars). For large-area graphene applications, this signifies that small bilayer (or multilayer) islands have little or no contribution to overall thermal transport.

Active metasurfaces based on phase-change memory material digital metamolecules

Tunable metasurfaces are a promising candidate for the next generation of spatial light modulators which will require higher refresh rates, smaller pixel sizes, and compact form factors. Phase-change memory materials present a unique platform for nonvolatile reconfigurable metasurfaces which could undergo phase transitions at MHz frequencies if actuated electrically, more than three orders of magnitude higher than refresh rates of existing commercial SLMs. While stable intermediate phases of GeSbTe (GST) exist which can be used for imparting differential phase shifts, the stochasticity of the material properties would limit the robustness of such a phase shifter, whereas the fully crystalline and amorphous states exhibit more consistent behavior. To overcome this, we design GST digital metamolecules comprising constituent meta-atoms which individually are in either the SET or RESET state, but which together form a tunable metamolecule with a set of robust phase shifts. We simulate active metasurface lenses based on these metamolecules, showing successful focusing, and demonstrate nano-patterning of a GST film with isolated nanoposts of material which could be electrically actuated, unlike counterparts which must be optically reconfigured.

Thermal modeling of metal oxides for highly scaled nanoscale RRAM

Resistive random access memory (RRAM) is a promising candidate for future non-volatile memory applications due to its potential for performance, scalability and compatibility with CMOS processing. The switching in the RRAM cell occurs via formation of conductive filaments composed of sub-stoichiometric oxide (SSO). In this work, we model thermal conduction in a pair of neighboring memory cells, taking into account more detailed phonon scattering effects in the SSO than previously considered. We find that for devices scaled below 10 nm in bit spacing, the neighboring filament temperature can increase significantly even when only the phononic heat conduction is considered. This increase is underestimated if using the previous state-of-the-art model of thermal conductivity of SSO, i.e. linear interpolation between metal and stoichiometric oxide thermal conductivity.