Catching the quantum sound wave Rusko Ruskov and Charles Tahan Science 346, 165 (2014); DOI: 10.1126/science.1260180
This copy is for your personal, non-commercial use only.
If you wish to distribute this article to others, you can order high-quality copies for your colleagues, clients, or customers by clicking here.
The following resources related to this article are available online at www.sciencemag.org (this information is current as of October 9, 2014 ): Updated information and services, including high-resolution figures, can be found in the online version of this article at: http://www.sciencemag.org/content/346/6206/165.full.html A list of selected additional articles on the Science Web sites related to this article can be found at: http://www.sciencemag.org/content/346/6206/165.full.html#related This article cites 11 articles, 1 of which can be accessed free: http://www.sciencemag.org/content/346/6206/165.full.html#ref-list-1
Science (print ISSN 0036-8075; online ISSN 1095-9203) is published weekly, except the last week in December, by the American Association for the Advancement of Science, 1200 New York Avenue NW, Washington, DC 20005. Copyright 2014 by the American Association for the Advancement of Science; all rights reserved. The title Science is a registered trademark of AAAS.
Downloaded from www.sciencemag.org on October 9, 2014
Permission to republish or repurpose articles or portions of articles can be obtained by following the guidelines here.
QUANTUM PROCESSING
Catching the quantum sound wave A superconducting qubit built to listen as well as see By Rusko Ruskov and Charles Tahan
ILLUSTRATION: P. HUEY/SCIENCE
A
properties of sound that is reminiscent of the earliest quantum optics experiments. In recent years, the essential physics of quantum optics and cavity QED have been replicated with superconducting circuits (circuit QED) (8). In circuit QED, the photons are in the microwave regime, and anharmonic (“atom-like”) superconducting qubits that use the loss-less nonlinear element of the Josephson junction take the place of the atoms. Whereas atoms are microscopic and provided by nature, superconducting circuits are macroscopic, can be made artificially on a chip, and offer exceptional design choice. These artificial atoms can be tuned to be very sensitive to electromagnetic fields, and this capability has been used to create strong light-qubit coupling, both in a cavity and with itinerant photons. Gustafsson et al. took advantage of the flexibility of superconducting circuits to design a transmon-type qubit (designed to minimize sensitivity to charge noise) that
produce sound waves on the surface of the crystal. In this manner, Gustafsson et al. constructed a beautifully straightforward experiment. A standalone interdigitated transducer (IDT) (analogous to the wand shown in panel A of the figure) generates SAWs directed toward the sound-sensitive transmon qubit (see the figure, panel B). This IDT also detects acoustic waves that reflect back from or are generated by the sound-sensitive qubit, which one might call a “soundmon”. The IDT attached to the qubit translates the incoming sound wave to an electrical signal that can act as an input signal for the qubit. This qubit is designed to interact with phonons of precisely the same frequency as those produced by the IDT. The critical question is whether the artificial atom really couples to sound in a quantum way. The qubit can be used to probe the quantum nature of sound because its energy levels are not equally spaced. In particular,
n ultrasound transducer (“the wand”) both creates and detects sound waves that travel through the body to create images of internal organs or precious cargo (see the figure, panel A). This compact device is made possible with piezoelectric crystals that expand or contract in response to an applied voltage and thus interconvert sound waves and electrical signals. Because sound travels relatively slowly, there is time to process the reflected pulses and display an image in real time. These measurements are in the realm of classical physics, but sound could also play a useful role in quantum-based devices. On page 207 of this issue, Gustafsson et al. (1) take a major step toward that goal, demonstrating a system in which a specially engineered artificial atom in the form of a superconducting quantum bit (qubit) couples to propagating surface acoustic waves on a chip. This soundmatter system shows evidence of quantum behavior. B A A fundamental concept of quantum physics is the wave-particle duality of elementary particles; for example, a photon behaves like a particle or a wave depending on the measurement being made. The quantum nature of light is especially evident in experiments that couple moving photons with atoms (quantum optics) or quanGaAs crystal tize a confined light field, placing an atom between two mirrors (cavity quantum electrodynamics, or cavity QED). That light is composed of Classical and quantum sound. (A) Ultrasound imaging works by creating and detecting sound waves with piezoelectric photons is made clear when a single crystals in the handheld “wand.” (B) A schematic of the superconducting device of Gustafsson et al. that operates in the atom is shown to radiate one photon quantum regime at cryogenic temperatures. An acoustic transducer coupled to a superconducting qubit is fabricated on a at a time (2). piezoelectric substrate. Surface acoustic waves are shown to interact in a quantum way with the artificial atom. Although sound waves in solidstate systems such as a crystal are the colcouples strongly with surface acoustic waves the energy splitting between the ground and lective excitation of many atoms, quantum (SAWs) in the crystal on which the qubit sits. first excited states is greater than the splitmechanics tells us that these too behave First, they deliberately fabricated a qubit ting between the first and second excited at the smallest level as quantized particles on a gallium-arsenide piezoelectric crystal. states. The phonon quanta that make up the called phonons. Mechanical systems have In a piezoelectric crystal, strain and electric sound wave can be registered by the qubit, received renewed attention, especially in the fields are directly coupled, the latter being a which starts out in its ground state given the study of nanomechanical resonators as nanonatural interaction mechanism for superconvery low temperatures of the experiment. fabrication and quantum control techniques ducting qubits. They then integrated a SAW When the sound wave’s energy is in resohave improved (3–7). transducer into the qubit, which can act as nance with the qubit (which can be tuned The work of Gustafsson et al. introduces both a speaker and a microphone for SAWs. with an external magnetic field), a phonon a new test bed to investigate the quantum By fabricating metal wires in an interdigican be absorbed as a whole, and the qubit tated finger arrangement and applying an will be excited from its ground state to the oscillating voltage, the electric field drives first excited state. A subsequently impinging Laboratory for Physical Sciences, College Park, MD 20740 USA. E-mail: [email protected] the strain field via the piezoelectric effect to second phonon of the same frequency canSCIENCE sciencemag.org
10 OCTOBER 2014 • VOL 346 ISSUE 6206
Published by AAAS
165
INSIGHTS | P E R S P E C T I V E S
REFERENCES
1. M. V. Gustafsson et al., Science 346, 207 (2014). 2. H. J. Kimble, M. Dagenais, L. Mandel, Phys. Rev. Lett. 39, 691 (1977). 3. M. D. LaHaye, J. Suh, P. M. Echternach, K. C. Schwab, M. L. Roukes, Nature 459, 960 (2009). 4. M. Eichenfield, J. Chan, R. M. Camacho, K. J. Vahala, O. Painter, Nature 462, 78 (2009). 5. A. D. O’Connell et al., Nature 464, 697 (2010). 6. P. Rabl et al., Nat. Phys. 6, 602 (2010). 7. R. W. Andrews et al., Nat. Phys. 10, 321 (2014). 8. A. Blais, R.-S. Huang, A. Wallraff, S. Girvin, R. Schoelkopf, Phys. Rev. A 69, 062320 (2004). 9. O. Astafiev et al., Science 327, 840 (2010). 10. O. O. Soykal, R. Ruskov, C. Tahan, Phys. Rev. Lett. 107, 235502 (2011). 11. R. Ruskov, C. Tahan, Phys. Rev. B 88, 064308 (2013). 10.1126/science.1260180
166
CONSERVATION
Taking the measure of change Predictive models of biodiversity change are required to inform conservation policy decisions Ideally, the chain between metric, indicator, and policy should start with specific ver the past decade, numerous metrics targets. In fisheries management, targets for biodiversity—including species are often explicit, typically relating to the abundance, extinction risk, distribusustainability of fish stocks; this helps to tion, genetic variability, species turnguide fisheries policy and management inover, and trait diversity—have been terventions and to detect fishing impacts used to create indicators to track how on marine biodiversity (7). Targets can vary biodiversity has changed (1–3). widely in scale; metrics such total These indicators have made it biomass, catch, and mean trophic clear that biodiversity loss, howlevel are used to evaluate sustainever it is measured, is showing ability targets for ecosystem manlittle sign of abatement (1, 4) and agement (8), whereas changes in that humans must respond to recruitment and abundance are safeguard the provision of natuused to construct indicators for ral services on which we all rely managing specific fish stocks (9). CONSERVATION SERIES (5, 6). But which metrics provide In contrast, global biodiversity the most informative indicators targets such as those agreed to in under which circumstances? And how can the Convention on Biological Diversity (CBD) the growing list of indicators best serve con(6) tend to be less specific, and the alignment servation policy decisions? of metric, indicator, and target can be poor. For example, one CBD target calls for polluALIGNMENT OF TARGET AND INDICAtion to be reduced to levels that are not detriTOR. If we are interested in the status of an mental to ecosystem function or biodiversity. economy, we measure its performance over This laudable target does not detail which time using metrics such as cost of goods, pollutants, ecosystem functions, or particular income, and employment numbers. Those aspects of biodiversity should be addressed. metrics are then used to produce indicators The distinction is important because many such as GDP and RPI. Similarly, metrics functions will trade off with one another, and such as species abundance are used to creprioritizing some aspects of biodiversity will ate indicators of the health of biodiversity. be at the cost of others. Efforts to measure For an indicator to help achieve a particprogress toward this target, hold polluting ular conservation target, the indicator and countries and industries to account, and dithe target must be closely aligned (4). There agnose which interventions work best are is little point in measuring progress toward made more difficult. a target with an indicator based on a metThe outcomes of global biodiversity tarric that is only loosely related to the degets are perhaps inevitably less focused sired outcome. For example, ensuring that than those in specified circumstances such protected areas maintain their biodiversity as fisheries management. However, with is a fundamental goal of conservation. Targreater demand and scrutiny placed on biogets are frequently centered on the extent diversity indicators (4) through targets such of area under protection [e.g., 17% of land as CBD, how can they better support consershould be under protection by 2020 (6)]. vation efforts? One way forward is improved The implicit assumption is that the greater prediction. the area protected, the more biodiversity will prosper. However, this ignores factors PROJECTING FORWARD. Effective consersuch as governance, funding, and the type vation policy decisions require an explicit of species within the protected area that understanding of the links between desired influence its effectiveness. Protected areas outcomes of conservation, how those outdiffer greatly in the protection they afford comes can be measured, and the proposed species, but management effectiveness indiactions needed to achieve them (10). One cators are only available for less than 5% of way to accomplish this is to project forward protected areas (4). Using just one metric the impacts of a prospective policy. In doas an indicator may not achieve the desired ing so, both the impact of the policy and outcome. the ability of indicators to detect change in By Ben Collen1 and Emily Nicholson2
O
sciencemag.org SCIENCE
10 OCTOBER 2014 • VOL 346 ISSUE 6206
Published by AAAS
ILLUSTRATION: C. SMITH/SCIENCE
not scatter until the first phonon is “released” from the qubit, because it does not match the splitting between the first and second excited states. Gustafsson et al. showed this by continuously changing the rate of impinging phonons from a much lower to a much higher rate with respect to the qubit relaxation time. Their artificial atom reflected sound on resonance and did not for nonresonant conditions, in a manner similar to that seen in circuit QED (9). Gustafsson et al. also used microwave radiation to independently excite and drive the qubit without sound. The excited artificial atom is expected to relax by emitting phonons that will travel to and induce a characteristic signal at the IDT. The time the phonons take to get to the IDT definitively differentiates sound from purely electrical signals. A cross-talk signal on the IDT was seen when the electrical pulse to the qubit was turned on (and even when the qubit was off resonance). Then, ~40 ns later, the sound wave was detected, and this delay corresponded to the sound-wave propagation time. Further experiments that used both acoustic and microwave drives on the qubit showed additional analogs of atomic physics, specifically the Autler-Townes splitting related to the Stark effect. A next step would be to confirm phonon antibunching of the driven qubit in this system, which would definitively “prove” the existence of phonons. But more questions can be asked. How long will an acoustic phonon stay coherent in a specific crystal? Can phonons be entangled with each other? Can sound-sensitive qubits and cavity quantum phonodynamics be realized in other systems, for example, via the deformation potential (10, 11)? With the ability to make artificial atoms much larger than the wavelength of the phonon or to perform several qubit operations (including classical processing and feedback) while the slow phonon is still traveling, the space of available quantum experiments grows. Look out, quantum optics; the future seems loud for quantum sound. ■
This copy is for your personal, non-commercial use only.
If you wish to distribute this article to others, you can order high-quality copies for your colleagues, clients, or customers by clicking here.
The following resources related to this article are available online at www.sciencemag.org (this information is current as of October 9, 2014 ): Updated information and services, including high-resolution figures, can be found in the online version of this article at: http://www.sciencemag.org/content/346/6206/165.full.html A list of selected additional articles on the Science Web sites related to this article can be found at: http://www.sciencemag.org/content/346/6206/165.full.html#related This article cites 11 articles, 1 of which can be accessed free: http://www.sciencemag.org/content/346/6206/165.full.html#ref-list-1
Science (print ISSN 0036-8075; online ISSN 1095-9203) is published weekly, except the last week in December, by the American Association for the Advancement of Science, 1200 New York Avenue NW, Washington, DC 20005. Copyright 2014 by the American Association for the Advancement of Science; all rights reserved. The title Science is a registered trademark of AAAS.
Downloaded from www.sciencemag.org on October 9, 2014
Permission to republish or repurpose articles or portions of articles can be obtained by following the guidelines here.
QUANTUM PROCESSING
Catching the quantum sound wave A superconducting qubit built to listen as well as see By Rusko Ruskov and Charles Tahan
ILLUSTRATION: P. HUEY/SCIENCE
A
properties of sound that is reminiscent of the earliest quantum optics experiments. In recent years, the essential physics of quantum optics and cavity QED have been replicated with superconducting circuits (circuit QED) (8). In circuit QED, the photons are in the microwave regime, and anharmonic (“atom-like”) superconducting qubits that use the loss-less nonlinear element of the Josephson junction take the place of the atoms. Whereas atoms are microscopic and provided by nature, superconducting circuits are macroscopic, can be made artificially on a chip, and offer exceptional design choice. These artificial atoms can be tuned to be very sensitive to electromagnetic fields, and this capability has been used to create strong light-qubit coupling, both in a cavity and with itinerant photons. Gustafsson et al. took advantage of the flexibility of superconducting circuits to design a transmon-type qubit (designed to minimize sensitivity to charge noise) that
produce sound waves on the surface of the crystal. In this manner, Gustafsson et al. constructed a beautifully straightforward experiment. A standalone interdigitated transducer (IDT) (analogous to the wand shown in panel A of the figure) generates SAWs directed toward the sound-sensitive transmon qubit (see the figure, panel B). This IDT also detects acoustic waves that reflect back from or are generated by the sound-sensitive qubit, which one might call a “soundmon”. The IDT attached to the qubit translates the incoming sound wave to an electrical signal that can act as an input signal for the qubit. This qubit is designed to interact with phonons of precisely the same frequency as those produced by the IDT. The critical question is whether the artificial atom really couples to sound in a quantum way. The qubit can be used to probe the quantum nature of sound because its energy levels are not equally spaced. In particular,
n ultrasound transducer (“the wand”) both creates and detects sound waves that travel through the body to create images of internal organs or precious cargo (see the figure, panel A). This compact device is made possible with piezoelectric crystals that expand or contract in response to an applied voltage and thus interconvert sound waves and electrical signals. Because sound travels relatively slowly, there is time to process the reflected pulses and display an image in real time. These measurements are in the realm of classical physics, but sound could also play a useful role in quantum-based devices. On page 207 of this issue, Gustafsson et al. (1) take a major step toward that goal, demonstrating a system in which a specially engineered artificial atom in the form of a superconducting quantum bit (qubit) couples to propagating surface acoustic waves on a chip. This soundmatter system shows evidence of quantum behavior. B A A fundamental concept of quantum physics is the wave-particle duality of elementary particles; for example, a photon behaves like a particle or a wave depending on the measurement being made. The quantum nature of light is especially evident in experiments that couple moving photons with atoms (quantum optics) or quanGaAs crystal tize a confined light field, placing an atom between two mirrors (cavity quantum electrodynamics, or cavity QED). That light is composed of Classical and quantum sound. (A) Ultrasound imaging works by creating and detecting sound waves with piezoelectric photons is made clear when a single crystals in the handheld “wand.” (B) A schematic of the superconducting device of Gustafsson et al. that operates in the atom is shown to radiate one photon quantum regime at cryogenic temperatures. An acoustic transducer coupled to a superconducting qubit is fabricated on a at a time (2). piezoelectric substrate. Surface acoustic waves are shown to interact in a quantum way with the artificial atom. Although sound waves in solidstate systems such as a crystal are the colcouples strongly with surface acoustic waves the energy splitting between the ground and lective excitation of many atoms, quantum (SAWs) in the crystal on which the qubit sits. first excited states is greater than the splitmechanics tells us that these too behave First, they deliberately fabricated a qubit ting between the first and second excited at the smallest level as quantized particles on a gallium-arsenide piezoelectric crystal. states. The phonon quanta that make up the called phonons. Mechanical systems have In a piezoelectric crystal, strain and electric sound wave can be registered by the qubit, received renewed attention, especially in the fields are directly coupled, the latter being a which starts out in its ground state given the study of nanomechanical resonators as nanonatural interaction mechanism for superconvery low temperatures of the experiment. fabrication and quantum control techniques ducting qubits. They then integrated a SAW When the sound wave’s energy is in resohave improved (3–7). transducer into the qubit, which can act as nance with the qubit (which can be tuned The work of Gustafsson et al. introduces both a speaker and a microphone for SAWs. with an external magnetic field), a phonon a new test bed to investigate the quantum By fabricating metal wires in an interdigican be absorbed as a whole, and the qubit tated finger arrangement and applying an will be excited from its ground state to the oscillating voltage, the electric field drives first excited state. A subsequently impinging Laboratory for Physical Sciences, College Park, MD 20740 USA. E-mail: [email protected] the strain field via the piezoelectric effect to second phonon of the same frequency canSCIENCE sciencemag.org
10 OCTOBER 2014 • VOL 346 ISSUE 6206
Published by AAAS
165
INSIGHTS | P E R S P E C T I V E S
REFERENCES
1. M. V. Gustafsson et al., Science 346, 207 (2014). 2. H. J. Kimble, M. Dagenais, L. Mandel, Phys. Rev. Lett. 39, 691 (1977). 3. M. D. LaHaye, J. Suh, P. M. Echternach, K. C. Schwab, M. L. Roukes, Nature 459, 960 (2009). 4. M. Eichenfield, J. Chan, R. M. Camacho, K. J. Vahala, O. Painter, Nature 462, 78 (2009). 5. A. D. O’Connell et al., Nature 464, 697 (2010). 6. P. Rabl et al., Nat. Phys. 6, 602 (2010). 7. R. W. Andrews et al., Nat. Phys. 10, 321 (2014). 8. A. Blais, R.-S. Huang, A. Wallraff, S. Girvin, R. Schoelkopf, Phys. Rev. A 69, 062320 (2004). 9. O. Astafiev et al., Science 327, 840 (2010). 10. O. O. Soykal, R. Ruskov, C. Tahan, Phys. Rev. Lett. 107, 235502 (2011). 11. R. Ruskov, C. Tahan, Phys. Rev. B 88, 064308 (2013). 10.1126/science.1260180
166
CONSERVATION
Taking the measure of change Predictive models of biodiversity change are required to inform conservation policy decisions Ideally, the chain between metric, indicator, and policy should start with specific ver the past decade, numerous metrics targets. In fisheries management, targets for biodiversity—including species are often explicit, typically relating to the abundance, extinction risk, distribusustainability of fish stocks; this helps to tion, genetic variability, species turnguide fisheries policy and management inover, and trait diversity—have been terventions and to detect fishing impacts used to create indicators to track how on marine biodiversity (7). Targets can vary biodiversity has changed (1–3). widely in scale; metrics such total These indicators have made it biomass, catch, and mean trophic clear that biodiversity loss, howlevel are used to evaluate sustainever it is measured, is showing ability targets for ecosystem manlittle sign of abatement (1, 4) and agement (8), whereas changes in that humans must respond to recruitment and abundance are safeguard the provision of natuused to construct indicators for ral services on which we all rely managing specific fish stocks (9). CONSERVATION SERIES (5, 6). But which metrics provide In contrast, global biodiversity the most informative indicators targets such as those agreed to in under which circumstances? And how can the Convention on Biological Diversity (CBD) the growing list of indicators best serve con(6) tend to be less specific, and the alignment servation policy decisions? of metric, indicator, and target can be poor. For example, one CBD target calls for polluALIGNMENT OF TARGET AND INDICAtion to be reduced to levels that are not detriTOR. If we are interested in the status of an mental to ecosystem function or biodiversity. economy, we measure its performance over This laudable target does not detail which time using metrics such as cost of goods, pollutants, ecosystem functions, or particular income, and employment numbers. Those aspects of biodiversity should be addressed. metrics are then used to produce indicators The distinction is important because many such as GDP and RPI. Similarly, metrics functions will trade off with one another, and such as species abundance are used to creprioritizing some aspects of biodiversity will ate indicators of the health of biodiversity. be at the cost of others. Efforts to measure For an indicator to help achieve a particprogress toward this target, hold polluting ular conservation target, the indicator and countries and industries to account, and dithe target must be closely aligned (4). There agnose which interventions work best are is little point in measuring progress toward made more difficult. a target with an indicator based on a metThe outcomes of global biodiversity tarric that is only loosely related to the degets are perhaps inevitably less focused sired outcome. For example, ensuring that than those in specified circumstances such protected areas maintain their biodiversity as fisheries management. However, with is a fundamental goal of conservation. Targreater demand and scrutiny placed on biogets are frequently centered on the extent diversity indicators (4) through targets such of area under protection [e.g., 17% of land as CBD, how can they better support consershould be under protection by 2020 (6)]. vation efforts? One way forward is improved The implicit assumption is that the greater prediction. the area protected, the more biodiversity will prosper. However, this ignores factors PROJECTING FORWARD. Effective consersuch as governance, funding, and the type vation policy decisions require an explicit of species within the protected area that understanding of the links between desired influence its effectiveness. Protected areas outcomes of conservation, how those outdiffer greatly in the protection they afford comes can be measured, and the proposed species, but management effectiveness indiactions needed to achieve them (10). One cators are only available for less than 5% of way to accomplish this is to project forward protected areas (4). Using just one metric the impacts of a prospective policy. In doas an indicator may not achieve the desired ing so, both the impact of the policy and outcome. the ability of indicators to detect change in By Ben Collen1 and Emily Nicholson2
O
sciencemag.org SCIENCE
10 OCTOBER 2014 • VOL 346 ISSUE 6206
Published by AAAS
ILLUSTRATION: C. SMITH/SCIENCE
not scatter until the first phonon is “released” from the qubit, because it does not match the splitting between the first and second excited states. Gustafsson et al. showed this by continuously changing the rate of impinging phonons from a much lower to a much higher rate with respect to the qubit relaxation time. Their artificial atom reflected sound on resonance and did not for nonresonant conditions, in a manner similar to that seen in circuit QED (9). Gustafsson et al. also used microwave radiation to independently excite and drive the qubit without sound. The excited artificial atom is expected to relax by emitting phonons that will travel to and induce a characteristic signal at the IDT. The time the phonons take to get to the IDT definitively differentiates sound from purely electrical signals. A cross-talk signal on the IDT was seen when the electrical pulse to the qubit was turned on (and even when the qubit was off resonance). Then, ~40 ns later, the sound wave was detected, and this delay corresponded to the sound-wave propagation time. Further experiments that used both acoustic and microwave drives on the qubit showed additional analogs of atomic physics, specifically the Autler-Townes splitting related to the Stark effect. A next step would be to confirm phonon antibunching of the driven qubit in this system, which would definitively “prove” the existence of phonons. But more questions can be asked. How long will an acoustic phonon stay coherent in a specific crystal? Can phonons be entangled with each other? Can sound-sensitive qubits and cavity quantum phonodynamics be realized in other systems, for example, via the deformation potential (10, 11)? With the ability to make artificial atoms much larger than the wavelength of the phonon or to perform several qubit operations (including classical processing and feedback) while the slow phonon is still traveling, the space of available quantum experiments grows. Look out, quantum optics; the future seems loud for quantum sound. ■
