Periodic Reporting for period 2 - LEIT (Lossless information for Emerging Information Technologies)
Reporting period: 2022-07-01 to 2023-09-30
Summary for publication
The problem being addressed is the use of phonons (quanta of lattice vibrations) as information carriers, focusing on information transport in interconnects. The information is represented by coordinated atomic displacements in a semiconductor crystal of sub millielectron volts travelling along an interconnect, here represented by a waveguide. The challenge is the highly dissipative nature of such atomic vibrations thus, in this action we focus the research to test to what degree, the new physics based on the phases of matter, can deliver sufficient robustness so that phonons could carry information as electron do in electronics and photons do in photonics.
The relevance to society is the increasing share of the world energy currently increasing with the use of data centres, the huge and increasing number of bits flying to and fro, and the more demanding computation performance, such as that required to train networks for artificial intelligence. At some point, in one of the smaller EU countries, data centres were consuming nearly one third of the country’s electricity. The energy demands of the digital age place a huge burden on the world’s electricity supply and phonons, with a fraction of the energy compared to electrons and photons, could alleviate this demand at least in the less demanding information processing tasks.
The overall objective is to demonstrate information flow transported by lattice vibrations or phonons in phononic waveguides, akin to optical fibres for photons or to metal interconnects in printed circuit board, taken to the limit of integration, targeting in-chip interconnects. The challenge is to maintain the information: by their very nature lattice vibrations couple to neighbouring atoms and defects thus losing energy.
Newly identified concepts in solid state physics involving novel phases of matter determined by the symmetry of the atoms in a crystal and the way they bond with their neighbours, gave rise to mathematical concepts that won the Nobel Prize of Physics to three scientists in 2016 (Profs. Thouless, Haldane and Kosterlitz) and heralded the possibility of signal robustness. While the concepts of “topology” were applied to electron transport, in this action we proposed to extend them to lattice vibrations and thus targeted topological phonon transport at frequencies relevant for information technologies, here in the GHz range.
The research carried out so far has focused on source of phonons and phonon waveguides. Several approaches were selected to begin with and in the course of the work they have been narrowed down. Some have been proof-of-concepts helping us to understand the scope of the physical realisations that may form a future phononic circuit. In a circuit a signal is needed and here we investigate “phonon sources”, as well as a waveguide to carry it. This means that the source needs to be coupled to a waveguide, which has been the focus of our research in this period.
We have succeeded in selecting surface acoustic waves as a source of phonons, which can be integrated with a topological interface phononic waveguide. Tests are running at 2.1 GHZ and others at 13 GHz, depending on design and means of observation. The preliminary results are promising and we will continue to improve our understanding on the physics of the incoherent phonons and that of the coherent, topological, phonons.
Despite the progress made so far, we have much to understand conceptually to develop models and advance simulations. We also face technological challenges since, despite working with silicon technology and at room temperature, we come up against the limit of fabrication tolerances were a variation of a few nanometers in critical dimensions matter. Novel concepts and avenues are under study to enable reliable designs, which in turn will allow us to realise circuits in the low GHz frequency range and in the 10s of GHz, the latter being a frequency close to applications in filters and interconnects for 5G and 6G.
The LEIT team and associates has being meeting regularly online and later in person. We held our first presencial workshops in October last year with a guest from another European project (see image 1). We anticipate the spirit of cooperation and curiosity will continue in the remaining of the action.
The problem being addressed is the use of phonons (quanta of lattice vibrations) as information carriers, focusing on information transport in interconnects. The information is represented by coordinated atomic displacements in a semiconductor crystal of sub millielectron volts travelling along an interconnect, here represented by a waveguide. The challenge is the highly dissipative nature of such atomic vibrations thus, in this action we focus the research to test to what degree, the new physics based on the phases of matter, can deliver sufficient robustness so that phonons could carry information as electron do in electronics and photons do in photonics.
The relevance to society is the increasing share of the world energy currently increasing with the use of data centres, the huge and increasing number of bits flying to and fro, and the more demanding computation performance, such as that required to train networks for artificial intelligence. At some point, in one of the smaller EU countries, data centres were consuming nearly one third of the country’s electricity. The energy demands of the digital age place a huge burden on the world’s electricity supply and phonons, with a fraction of the energy compared to electrons and photons, could alleviate this demand at least in the less demanding information processing tasks.
The overall objective is to demonstrate information flow transported by lattice vibrations or phonons in phononic waveguides, akin to optical fibres for photons or to metal interconnects in printed circuit board, taken to the limit of integration, targeting in-chip interconnects. The challenge is to maintain the information: by their very nature lattice vibrations couple to neighbouring atoms and defects thus losing energy.
Newly identified concepts in solid state physics involving novel phases of matter determined by the symmetry of the atoms in a crystal and the way they bond with their neighbours, gave rise to mathematical concepts that won the Nobel Prize of Physics to three scientists in 2016 (Profs. Thouless, Haldane and Kosterlitz) and heralded the possibility of signal robustness. While the concepts of “topology” were applied to electron transport, in this action we proposed to extend them to lattice vibrations and thus targeted topological phonon transport at frequencies relevant for information technologies, here in the GHz range.
The research carried out so far has focused on source of phonons and phonon waveguides. Several approaches were selected to begin with and in the course of the work they have been narrowed down. Some have been proof-of-concepts helping us to understand the scope of the physical realisations that may form a future phononic circuit. In a circuit a signal is needed and here we investigate “phonon sources”, as well as a waveguide to carry it. This means that the source needs to be coupled to a waveguide, which has been the focus of our research in this period.
We have succeeded in selecting surface acoustic waves as a source of phonons, which can be integrated with a topological interface phononic waveguide. Tests are running at 2.1 GHZ and others at 13 GHz, depending on design and means of observation. The preliminary results are promising and we will continue to improve our understanding on the physics of the incoherent phonons and that of the coherent, topological, phonons.
Despite the progress made so far, we have much to understand conceptually to develop models and advance simulations. We also face technological challenges since, despite working with silicon technology and at room temperature, we come up against the limit of fabrication tolerances were a variation of a few nanometers in critical dimensions matter. Novel concepts and avenues are under study to enable reliable designs, which in turn will allow us to realise circuits in the low GHz frequency range and in the 10s of GHz, the latter being a frequency close to applications in filters and interconnects for 5G and 6G.
The LEIT team and associates has being meeting regularly online and later in person. We held our first presencial workshops in October last year with a guest from another European project (see image 1). We anticipate the spirit of cooperation and curiosity will continue in the remaining of the action.
So far we have:
(a) demonstrated two phonon sources approaches, one of which is amenable to integration,
(b) demonstrated one signal modulation scheme although more integrable designs are needed,
(c) identified and parametrically characterised the impact of fabrication in terms of critical dimensions upon the properties of the topological waveguides,
(d) realised state-of-the-art electrical signal to mechanical signal convertors in source-waveguide configurations to quantify coupling efficiency,
(e) advanced simulations to model topological non-trivial phononic waveguides and
(f) developed experimental methods to make them suitable to quantify the observation with our experimental techniques.
(a) demonstrated two phonon sources approaches, one of which is amenable to integration,
(b) demonstrated one signal modulation scheme although more integrable designs are needed,
(c) identified and parametrically characterised the impact of fabrication in terms of critical dimensions upon the properties of the topological waveguides,
(d) realised state-of-the-art electrical signal to mechanical signal convertors in source-waveguide configurations to quantify coupling efficiency,
(e) advanced simulations to model topological non-trivial phononic waveguides and
(f) developed experimental methods to make them suitable to quantify the observation with our experimental techniques.
Most progress has been made in the physics of optomechanics to generate phonons and to modulate phonon signals. While this progress is pushing the state of the art since we obtained phonon lasing at 7 GHz which is beyond the state of the art in ambient conditions, the main challenge is to capitalise on this achievement to ensure the phonon laser can be integrated in the platform for phononic circuits of this action. This is one task for the remaining of the project. With respect to modulation, a similar situation arises, namely, while a multi-mode approach based on mechanical-optical-mechanical coupling, even if using a single optomechanical cavity, which our work has shown probably for the first time, the challenge is also integration.
With the contributions from the theoretical work to ensure the waveguide designs lead to the observation of non-trivial topological modes and from the progress in fabrication. Based on a learning curve walked during the first half of the project, we plan to realise bespoke topological waveguides to enable the experimental demonstration of guiding lossless phononic signals.
Thus, we expect to advance the research working with surface acoustic waves as phonon sources and with strain as the signal modulating mechanism. The character of the topological phononic waveguides is planned to be demonstrated unambiguously at 2.1 GHz by means of vibrometry and above 10 GHz by means of Brillouin light scattering. We shall advance the methodology since signal amplitude alone may not be sufficient to demonstrate phonon signal robustness and thus reach or goal of demonstrating the viability of phonons as tokens of information.
With the contributions from the theoretical work to ensure the waveguide designs lead to the observation of non-trivial topological modes and from the progress in fabrication. Based on a learning curve walked during the first half of the project, we plan to realise bespoke topological waveguides to enable the experimental demonstration of guiding lossless phononic signals.
Thus, we expect to advance the research working with surface acoustic waves as phonon sources and with strain as the signal modulating mechanism. The character of the topological phononic waveguides is planned to be demonstrated unambiguously at 2.1 GHz by means of vibrometry and above 10 GHz by means of Brillouin light scattering. We shall advance the methodology since signal amplitude alone may not be sufficient to demonstrate phonon signal robustness and thus reach or goal of demonstrating the viability of phonons as tokens of information.