Sebastian Hassinger
Welcome to the new quantum era. I'm your host, Sebastian Hassinger, bringing you another conversation from the cutting edge of quantum science and technology. On this episode, I'll be talking with Dr. Mohammad Mirhosseini, an assistant professor of electrical engineering and applied physics at Caltech. Mohamed leads a research group developing really innovative hybrid quantum devices. These are systems that bridge the worlds of sound and superconducting circuits all at the quantum level. We'll be talking about his team's recent breakthrough: building a mechanical quantum memory for microwave photons. If that sounds wild, you're not alone. The idea that sound waves, or more specifically, mechanical vibrations at the nanoscale, could store quantum memory. Information carried by superconducting qubits really fascinated me and drew me to inviting Mohamed to come on the show. There's some really incredible implications to quantum hardware design from this work. But we'll start by unpacking what quantum memory really means in this setting. and why microwave photons are essential to superconducting quantum computers, and how mechanical systems like tiny high-frequency tuning forks can serve as robust, long-lived quantum storage. We'll also get into what makes this approach so promising for extending coherence times and potentially enabling scalable quantum networks through quantum transduction. It's a fascinating conversation that will change how you think about the intersection of quantum computing, physics, and engineering. I hope you enjoy it as much as I did. Let's get started. Hi Mohamed, thank you so much for joining me. Hi Sebastian, thank you for having me. I'm really happy that you Got on the, you know, you accepted my invitation to the podcast, Mahal, because I'm really interested in some work that you've been doing recently. I think there was a paper out at the end of last year that caught my eye. But can you start off just by sort of quickly introducing yourself and talking a little bit about how you got into the field of quantum computing?

Mohammad Mirhosseini
Hi, Sebastian. Thank you for having me.

Sebastian Hassinger
I'm really happy that you got on the you accepted my invitation to the podcast, Mahal, because I'm really interested in some work that you've been doing recently. I think there was a paper out at the end of last year that caught my eye. But can you start off just by sort of quickly introducing yourself and talking a little bit about how you got into the field of quantum computing?

Mohammad Mirhosseini
Yes, of course. Well, my name is Mohammad Mirhosseini. I am an assistant professor of electrical engineering and applied physics at Caltech. I run a lab of researchers of about ten people and I'm glad that our recent work has been of interest to you. I got into this field basically by getting very interested in electromagnetism. The theory of electrical charges and waves in the more conventional form of it when I was a high school student. it it was very fascinating to me that a very understandable range of phenomena can be abstracted into Maxwell's equations. The mathematical simplicity and the conciseness of that was very appealing to me. And I tried to pursue understanding more and learning more about DNM in my undergraduate studies. There, of course, are many technological areas or different types of systems that work based on electromagnetic fields.

Sebastian Hassinger
Uh how I understand what I'm doing.

Mohammad Mirhosseini
I found most the fact that most of them are more or less understandable, frankly, a little bit boring. But then I heard about this field of quantum optics, which seemed to offer a little bit of Maybe, yeah, yeah, yeah.

Sebastian Hassinger
Where nothing's understood, really.

Mohammad Mirhosseini
A little bit of more confusion.

Sebastian Hassinger
Yes.

Mohammad Mirhosseini
And that, plus the determinism that usually comes with, okay, Maxwell's equation, explain everything, intrigued me. And, you know. I that's how far I remember in terms of how I got into this.

Sebastian Hassinger
That's great. And so you ended up, I think you did your PhD at Caltech, is that right?

Mohammad Mirhosseini
Not really. I did my PhD at the Institute of Optics at the University of Rochester in upstate New York and joined Caltech as a postdoc previously before I started my own lab as a PI.

Sebastian Hassinger
Postdoc Rochester, that's right. Right, right, right, right. You were in Oscar Painter's group, is that right?

Mohammad Mirhosseini
During my postdoc, I was a member of Oscar Painters Group.

Sebastian Hassinger
In your postdoc, yeah, yeah, yeah, yeah.

Mohammad Mirhosseini
That's right.

Sebastian Hassinger
And of course, I I'm in AWS Quantum and Oscar leads the the Center for Quantum Computing R and D team. So I'm I'm definitely familiar with Oscar's work and the a huge number of really talented people who've come through his group too.

Mohammad Mirhosseini
And I think talking about this particular topic of our recent work, we would have to refer to a few places where his past work has been instrumental in this.

Sebastian Hassinger
Right, right, right, right. That's great. So okay, cool. And so getting to this work, what caught my eye was the paper is called A Mechanical Quantum Memory for Microwave Photons. And there's a couple of things that I thought somebody from outside of the field would find really interesting, just right in the title. Because, for one thing, I think that on a fundamental level, it's often surprising to people that we talk about photons in a microwave kind of setting. Where electromagnetic, as you said, the electromagnetic spectrum, we think about it, I think lay people think about it more in terms of Electrical currents, maybe TV, radio, right? But the spectrum goes all the way up into, I mean, sorry, all the way down into the visible spectrum, right? I mean, there's sort of this. This continuous kind of path from microwave to photon to optical photons.

Mohammad Mirhosseini
Uh yeah, but it's a concept Yeah, I mean, you're right to ask that question.

Sebastian Hassinger
And so The in the microwave regime, are there actually photons within the qubit setting that you are interacting with that you're able to access? And so Uh four, we've got a market out of the way.

Mohammad Mirhosseini
Microwave photons are completely nontrivial and still fascinating concepts. Well, as you said, in electrical systems, we usually tend to think about electrons or Some charge carriers that are perhaps flowing in wires, and we conventionally understand an electrical system based on them. But when we think about what happens, is that these electrical charge carriers or electrons. They also interact with each other via the electrical forces. And if you think about the description of this, Entire system, you can keep track of these material particles, the charge carriers, or you can keep track of the electromagnetic fields that mediate the interactions among them. And it turns out that for radio frequency waves, and in particular situations where there's not particular sources around, you can essentially think of an electromagnetic circuit as a vessel.

Sebastian Hassinger
And I'm not going to use buttons and right online.

Mohammad Mirhosseini
where it's it is holding some electromagnetic energy within it, and then the once you have electromagnetic energy stored in this what we call fields, then the photons are the quanta of energy of these fields when you try to describe it electromechan electro sorry, quantum mechanically.

Sebastian Hassinger
I think that's a lot of magic on So if you can always do it.

Mohammad Mirhosseini
So that's where the microwave photon comes into the picture. It is no different than an optical photon in its abstract form. and its properties, but the terms of the techniques and the systems where it manifests itself, they tend to be very different. They are they have much lower frequency, they they they live inside different structures, but they are photons. For the most part. And I guess, in terms of thinking about qubits, we usually do not refer to these because we think about perhaps quantum computing. where we have many qubits interacting among themselves. But what goes on is that these qubits essentially can create or absorb microwave photons So in a rough analogy to optical systems, where we have atoms emitting or absorbing optical photons, here we have qubits emitting or absorbing single microwave photons.

Sebastian Hassinger
Right, right. And so, in this, again, going back to the title, Mechanical Quantum Memory. So, the next thing that I think is really fascinating is the idea of quantum memory, right? If people are paying attention to the devices that are being built today, these are rudimentary computers because they don't really have storage. They have The ability to run a single circuit. That circuit just describes a series of gates of operations on the qubits that carry out some kind of task. But there's no longer-term storage. There's none of the, you know, we sort of are very familiar with classical computing. There's, There's both random access memory and then there's long-term storage on hard drives. There's nothing like that in the quantum computing sort of context. So, is that you're attempting to add sort of longer term storage of quantum states into the sort of the toolkit of quantum computing? Is that an accurate way of describe it?

Mohammad Mirhosseini
I would say so, but I would have to add some details.

Sebastian Hassinger
Yeah, of course.

Mohammad Mirhosseini
architecturally, where do storage come into the picture when you try to do quantum computing is a non trivial question. Depending on how you do things, you may or may not need a dedicated form of hardware just for storing quantum information. Normally, we think of qubits as Entities that not only hold and create, but also manipulate the quantum information that are in these quantum states. But when it comes down to the physical properties that we need at the basic hardware, For realizing these basic operations, you know, gate operations or creation of states or storage of them, they translate to different physical properties.

Sebastian Hassinger
I don't know if you're not going to have to find that.

Mohammad Mirhosseini
what we tried to do in our work was essentially study this type of system, in this case, mechanical oscillators, where they have really exemplary properties when it comes to the storage. of the quantum states, but not perhaps they're not as good when it comes to the other parts, which are realizing gate operations, for example. So how would it come into quantum computing if you have something that can only hold a quantum state for a long time? Is that an essential component? The answer to that question is not easy. There are cases where we think they would be needed. But our emphasis again was on the basic properties of the physical system and how it can be utilized for these quantum operations.

Sebastian Hassinger
Interesting. It reminded me a little bit about I really like the work that Von Neumann did at the Institute for Advanced Studies. That whole story of building that system there is very interesting. And of course, the von Neumann architecture is still intrinsically part of the classical computing. Context. So I believe there was a very long tube. There was a switch, a mercury switch, that it would tilt, and the wave of the mercury would Propagate from one end of the switch to the other. And that was their first sort of persisted state that would last beyond the length of just the individual classical circuit that was being carried out on the transistors or the sorry, the vacuum tubes. So is it sort of analogous to that, where it's some mechanism for being able to hold the state longer than the qubit itself would be able to hold it?

Mohammad Mirhosseini
Well, I would say that you know, in quantum computing, the People didn't really, to my knowledge, pursue building a memory just to satisfy an architectural need. There are examples In things like quantum communication, where the long propagation time of photons required for something like a memory. Essentially, right on where you begin to think about that problem. In quantum computing, it does not come up naturally in that form. All you need in the most abstract form is just the qubits. But practically, once people have tried to make better and better qubits, at least with the superconducting systems, within this field, there has been these observations that it is easier to make physical hardware that can hold the quantum state in complicated form where you can either hold it passively for a long time or you can correct for errors.

Sebastian Hassinger
And what I'm going to do is I have to See what we're doing, but it's just another one.

Mohammad Mirhosseini
But these hardware components are not perhaps best at realizing quantum gates, but On the contrary, you can make things that are really fast and they can do these gate operations, but they have a lot of dissipation. So this competition between these two different properties has essentially guided the work to areas where you have seen these primitive components, which you can call as processor and memory, in forms of, for example, a microwave cavity and a superconductor. Qubit, that's what you see more conventionally.

Sebastian Hassinger
Um so that I'd be served.

Mohammad Mirhosseini
So that has emerged. Now, architecturally, again, whether or not you need some dedicated memory to Synchronize perhaps different parts of your computer that are going at different speeds, or in situations which are more similar to classical systems.

Sebastian Hassinger
Now I'll get to the Hmm.

Mohammad Mirhosseini
That is highly likely, but it goes a little bit beyond my expertise to comment on that.

Sebastian Hassinger
Got it. So at this point then, when you're with the quantum memories in the in your paper This is a way of extending the lifespan of the coherence and the quantum state of the qubit while it's carrying out its operations. Is that more accurate?

Mohammad Mirhosseini
That's correct. That's correct.

Sebastian Hassinger
Yeah, okay, okay.

Mohammad Mirhosseini
Yes.

Sebastian Hassinger
And you're doing this by getting to the third part of the title that I found fascinating. So it's a mechanical means of storing a quantum state. You're actually you're linking it an opto through optomechanics, you're linking a um an oscillator a physical uh device to something that we think of as being this very abstract quantum level device. Talk about how like how do you manage to physically link a mechanical device to something as ephemeral as a qubit?

Mohammad Mirhosseini
Well, I ha it is perhaps a little bit surprising to think of that. But if you think about it, well, first of all, we do have mechanical oscillators. These are vibrating dielectric components, which sometimes we think of them as tuning forks.

Sebastian Hassinger
Um so I'll have this microphone and um What we've done is great.

Mohammad Mirhosseini
They oscillate at frequencies that are pretty high, you know, gigahertz frequencies. And we what we do is actually we convert electrical energy or electrical quantum states to be more specific that are created by qubits, superconducting qubits.

Sebastian Hassinger
I don't know if I'm going to find the brain.

Mohammad Mirhosseini
Into mechanical vibrations or quantum states that are encoded in the vibrations inside these mechanical oscillators It may sound a little bit surprising, but this interplay of electrical signals and mechanical vibrations is something that happens.

Sebastian Hassinger
I'm a bad person. I think we can add a good video.

Mohammad Mirhosseini
In microphones and speakers. So, in some forms, it's very natural, you know, it comes up naturally almost.

Sebastian Hassinger
It's true. I guess what's surprising is the idea that down at the quantum level, I mean, does that mean that mechanical vibration is quantized at that scale?

Mohammad Mirhosseini
Well, to our knowledge, quantum mechanics is the underlying theory for understanding basically everything around us. What makes this perhaps a little bit less usual or less familiar is that normally we do not have to invoke quantum mechanics to understand vibrations. There is this rules of acoustics, for example. We've conventionally made systems that we can fully understand with uh with uh like with classical theories.

Sebastian Hassinger
I mean, it's going to be like you actually uh log a little bit and Yeah, yeah.

Mohammad Mirhosseini
Uh what makes the quantum mechanics necessary in in in this experiment and other experiments of this type, I have to say we were not the first people who thought about doing This is that once you have a what we call a non-classical state, you essentially use this qubits in these types of systems to generate.

Sebastian Hassinger
It's like a statement.

Mohammad Mirhosseini
states of motion which have genuinely non-classical properties.

Sebastian Hassinger
Now if I get to the back of the box, I think it's a lot I think I should have And the um you just mentioned phonon, so that's the sort of Terminology for the quantized unit of mechanical oscillation, is that right?

Mohammad Mirhosseini
They manifest signatures that are not understandable without using quantum mechanics. It is this creation of these relatively exotic states that makes us talk about single excitations of energy, you know, phonons, or quantum mechanical states of mechanical oscillators. But again, in this context, if there is an oscillator of any type, whether it's electric, optical, mechanical, or any other type. If you think about it in terms of its quanta of energy, there are single excitations, they might have different names like photons or phonons for electromagnetic and mechanical variance, for example. And then if you have a quantum state, it is in principle possible to map it from one degree of freedom to another. by essentially changing uh converting the quantum state one photon or one excitation at a time. Exactly.

Sebastian Hassinger
Yeah.

Mohammad Mirhosseini
This is a phone on to a mechanical oscillator. is is a photon, is what a photon is to an optical cavity. It's a it's a single excitation of energy.

Sebastian Hassinger
Incredible. And so You mentioned tuning forks. I mean, that's a very familiar sort of, you know, it's visceral, right? You can hit the tuning fork and then put it on something that's going to resonate like a table, and you can hear the tone. Is that, I mean, at the very, very smallest scale, is that the same kind of principle that you're using to link the two-level system of the qubit to the oscillator?

Mohammad Mirhosseini
Yes, very very much so. I mean, these are the the structures we make are smaller, a little bit.

Sebastian Hassinger
Yeah, a little bit smaller.

Mohammad Mirhosseini
They are a few microns, but they are still they are still made of many, many atoms. So in that sense, they are there are still components that in some sense resemble a tuning fork.

Sebastian Hassinger
Yeah.

Mohammad Mirhosseini
The frequency they vibrate at, as I mentioned, is much, much higher than what we hear. We can perhaps hear about a few kilohertz But these are at gigahertz frequency, and they sit in vacuum for technical reasons, so we cannot hear them echoing. But but otherwise they are the similar things.

Sebastian Hassinger
You couldn't use it to tune a piano, but other than that, yeah.

Mohammad Mirhosseini
Perhaps, you know, it did.

Sebastian Hassinger
And so, of course, they would have to be in the gigahertz range because that's the. That's the microwave frequency range that is used to control and read out a qubit, typically a superconducting qubit. Is it difficult to sort of Match the microwave regime with this mechanical oscillation frequency to link the two systems?

Mohammad Mirhosseini
There is a couple of different things. I mean, you mentioned I would perhaps talk about them one at a time. In terms of frequency Usually mechanical oscillators operate at lower frequency in more standard settings like room temperature and ambient pressure, it is very difficult to operate with gigahertz frequency resonances. They tend to dissipate their energy really fast. they involve really small dimensions. So unless there is a very good reason, usually they they do not show up in the technologically relevant devices. But for cryogenic experiments at very low temperatures and conditions like vacuum, which we already need for superconducting qubits, the experiments with these gigahertz frequency oscillators are much easier. They are a little bit smaller, but otherwise they are relatively easy to work with. In terms of the mechanism for coupling them to the electrical domain, essentially, how to convert these Quantum states from a qubit into mechanical oscillator. Well, that is basically where most of our work has been. There are different processes that you can use to do this conversion. And in the quantum world, we care about ultimately how well we can get interactions between two elements while at the same time avoiding what we call decoherence or dissipation.

Sebastian Hassinger
I'm going to get that in the file.

Mohammad Mirhosseini
So there's always a competition between these interaction rates that are things we want to the dissipation rates that are things we want to avoid. And the research basically in linking these mechanical oscillators to microwave qubits is is in maximizing this ratio between coupling and dissipation.

Sebastian Hassinger
I um Hmm, interesting. And in terms of, I mean, you know, coherence times are one of the critical factors in quantum computing. It's one of the things that You know, different modalities or different designs within a particular modality are constantly seeking when you your work sort of coupling these opto these uh mechanical resonators to or oscillator sorry Yeah, mechanical oscillators to the two level system. What's the sort of potential gain in coherence times that you think that that might provide?

Mohammad Mirhosseini
Yeah, so I I guess I have to uh give some extra context here. Well, as you said, coherence time is what you care about. Ultimately, coherence time is the as you know, is the metric of how long we can preserve quantum coherence. which is essentially the resource in trying to use a quantum state for any application.

Sebastian Hassinger
Um Like the different cards out of the live line.

Mohammad Mirhosseini
There is a slightly different concept. Of a lifetime, which is related to a coherence time, but is not exactly the same, which is the amount of time a system can retain the energy you put in it.

Sebastian Hassinger
If you're making the interview, I'm just going to have a little bit of So now they're going to have the idea because I'm going to use the same thing.

Mohammad Mirhosseini
Now these are sound very similar, but they're a little bit different because for quantum states you care about preserving the energy, but you also care about preserving the phase. Phase is a very Significant component of what you care about, and there are systems that can preserve energy for a long time, but necessarily the phase. So the system we worked with is what is uh very exciting about it is that they have extremely long lifetimes. They can preserve energies in in terms of phonons for for for times that exceed What superconducting qubits can do by more than ten times, perhaps about thirty times or so. Now, when it comes to coherence, the picture is a little bit more complicated. We see signatures of Processes that can cause dephasing. So in terms of coherence time, the improvements are a little bit more modest, a factor of few with current experiments. But without going to too much detail, I would say that there are techniques where these dephasing processes can be mitigated and ultimately with proper control. Coherence time can be extended to the limits where the lifetime is the relevant time scale.

Sebastian Hassinger
Right.

Mohammad Mirhosseini
And there we would expect that this factor of 10 to 30 would be the improvement over qubit systems.

Sebastian Hassinger
Amazing. That's amazing. And you're you're talking about the difference between T one and T two essentially there with the yeah, okay.

Mohammad Mirhosseini
Exactly, exactly.

Sebastian Hassinger
So T two is the dissipation time essentially, right?

Mohammad Mirhosseini
T T two is energy dissipation plus dephasing.

Sebastian Hassinger
Oh, plus C phase.

Mohammad Mirhosseini
T one is just energy dephasing.

Sebastian Hassinger
Okay. Right.

Mohammad Mirhosseini
Energy dissipation, yeah.

Sebastian Hassinger
Okay. Okay. Okay. Okay. Yeah.

Mohammad Mirhosseini
So we have about 30 times larger T1, but our T2s are only a factor of few better.

Sebastian Hassinger
So, okay. Oh, okay. I got it. I got it. Okay. That's great. So that must be actually, I mean. you know one of the the sort of rivalries between uh like the the neutral or trapped ion uh based systems and superconducting based systems is That the atom-based systems like to say that they've got orders of magnitude more sort of lifetimes of their qubits. This is a way of potentially making superconducting more competitive with the atom-based modalities, I guess.

Mohammad Mirhosseini
Yeah, perhaps. And I have to mention that there is a a number of studies with mechanical oscillators of very different shapes oscillating at very different frequencies. And made out of very different materials where they can get to massive lifetimes, you know, several seconds, even minutes in some systems. The challenge has been in trying to observe these exciting properties at the single phonon level Where you can precisely put economist state there and look at the phase, what happens to the phase, what happens to the energy, and observe that decoherence process And that is perhaps where our work may stand out because we are a we have been able to do that observation on one variant of mechanical oscillator. And now that we, in some sense, see verification of these really exciting properties, there is hope that with different types of oscillators, we can potentially achieve a much, much larger benefit down the road.

Sebastian Hassinger
And I mean, one of it seems like there's a trade-off in general, both the atom-based and also things like NV centers, for example, have very, very long coherence times, but they're very, very slow, right? The claim to fame of superconducting qubits is the speed of gate operations and readout. Is there any kind of Speed penalty that is associated with adding this mechanical oscillator to a qubit.

Mohammad Mirhosseini
It's a very, very good question. Yeah, I mean, like you said, to get conceptually to get to very good coherence times, you want the most isolated system in the world, but then how would you use it if it's really isolated, right?

Sebastian Hassinger
Right. Yeah.

Mohammad Mirhosseini
So the challenge is into selectively Exactly, right.

Sebastian Hassinger
There might be an immortal cubit somewhere, but we would never know because he couldn't access it.

Mohammad Mirhosseini
Exactly. And there are systems out there, right, that that have hours lo uh coherence times, but they're not so easy to operate. Yes, so the key is to selectively access, I guess, interaction to to an entity you care about, and then essentially turn that interaction off when you don't want it. And the same types of trade offs do show up in fact in our work. The fact that they What enabled our experiment was has been the a specific process that we have used to couple Mechanical motion to electrical signals. And in this type of process, the optomechanical process that we are using, we are able to retain these large coherence properties of these mechanical oscillators, but the rate of Quantum state conversion from electrical domain to mechanical domain remains slow.

Sebastian Hassinger
Okay.

Mohammad Mirhosseini
So it's about a hundred times slower than what you can do with just a superconductor. Qubit.

Sebastian Hassinger
Oh, interesting.

Mohammad Mirhosseini
So, there is a speed, there is a speed bottleneck right there in terms of converting the equum states from one domain to the other. However, there is a lot of room for improvement because this is the first time we have done something like this.

Sebastian Hassinger
However, if you remove logical things, I don't know what it is.

Mohammad Mirhosseini
We think. we can improve it quite a bit. How far it would go ultimately, it's a very difficult question to answer beforehand.

Sebastian Hassinger
Yeah, of course. Of course. Sounds like a combination of sort of iteratively like the engineering and the fabrication and the design, sort of iteratively trying to. Improve it. And then, of course, not knowing scientifically the fundamental limitations of how far you can go with it. And is there any impact In this kind of coupled system on the error rate. Is there either an increase or decrease in bit flips or phase flips?

Mohammad Mirhosseini
Well, this improved lifetime is or this improved in T one is can be interpreted as a reduction of the bit flips.

Sebastian Hassinger
Uh, I'm gonna go ahead and do a little bit of a test.

Mohammad Mirhosseini
In this context, and the T2 the T2 coherence time can be the dephasing component of the T2 can be translated as. as phase flip. So in reiterating what I said previously, I would say that we are seeing an order of magnitude reduction of bit flip errors and a very modest improvement in phase flip.

Sebastian Hassinger
So potentially, these types of qubits coupled to quantum memories, optomechanical quantum memories, might need might have lower overhead to get to fault tolerance, right, in terms of an error correction?

Mohammad Mirhosseini
Um Well, maybe, but I mean, to answer that question properly, one has to remember that the error rates only become significant or relevant when we consider them in the process of realizing the quantum gate.

Sebastian Hassinger
Uh code.

Mohammad Mirhosseini
So, we need to be able to do a quantum gate and then look at the error created during that process. It's very different than our experiment where we have had this Quantum state sitting is stationary in an object and looking at it just at a later time. So, in terms of we have not yet done any sophisticated type of gate operation on these oscillators It is in principle possible to do it.

Sebastian Hassinger
I see. I see.

Mohammad Mirhosseini
There is work with microwave frequency oscillators, essentially RF cavities that are also combined with qubits. to make an architecture for long time storage plus gate operations. But for our system to achieve that realm we are, we need to improve essentially these coupling gates that we talked about before.

Sebastian Hassinger
Okay. I see. So in that context, then the gate operations would have to be performed On the optomechanical coupling itself, or on the device itself, or that aspect of the device.

Mohammad Mirhosseini
Yes, yes.

Sebastian Hassinger
Yeah, I see.

Mohammad Mirhosseini
So right now in in these more primitive types of experiment like the one we did, we shuttle the states back and forth between the qubit and the oscillator. We make something, put it there, revive it later.

Sebastian Hassinger
I see.

Mohammad Mirhosseini
But with more coupling, what you can do is you can have the qubit to be always coupled to the mechanical oscillator or the oscillator or the long-lived oscillator.

Sebastian Hassinger
Right. Um I'm not sure if you can find the children, but it's a good idea to have a good idea.

Mohammad Mirhosseini
You would. design this system such that they do not interfere with the operation of each subparty too much. So the qubit lives there, the oscillator is there, but the presence of the qubit and its interaction with the oscillator will create a some sort of nonlinearity in the spectrum of the of the oscillator, and that allows you to essentially take benefit of the qubit presence to do gate operation in the oscillator on the quantum states that are put into the oscillator. while at the same time preserve the long coherence time of the oscillator. So that it's a it's a hybrid architecture where the qubits are there essentially for these gate operations, even though you don't actively use Them all the time. And that's what we are hoping to move towards.

Sebastian Hassinger
Interesting. Interesting. And I guess the other, I mean, the background of this work. Was, I think, explorations in transduction, right? Which is sort of the idea of taking a microwave qubit or a qubit in a microwave frequency. And trying to transduce it into a telecom frequency or a flying qubit, a photonic qubit. And this has obvious Huge advantages in scalability, right? If you can, superconducting devices are limited to some degree by how much you can fit in a dilution refrigerator. If you could. Span from one dilution refrigerator to another through these photonic links, photonic to microwave links. That would be really exciting. And of course, obviously. You mentioned quantum networking and quantum communications before. It seems like that transduction would be really critical there, too. So is this quantum memory's application of optomechanical coupling, does that have any implications for the work you've done prior in transduction?

Mohammad Mirhosseini
Yes, yes, you're absolutely right there. The mechanical oscillators surprisingly keep showing up in several areas in the close vicinity of superconducting qubits. And like you said, the optomechanical nonlinearities are one of the best ways we know to map equanimous state from the mechanical domain to the optical domain. So if you have a way of connecting electrical domain to mechanics and then a way of going from mechanics to optics, that naturally gives you a way to transduce microwave frequency quantum state to optics. The in terms of practical impact, the type of mechanical oscillators that we have used in our current experiment is exactly the same type that is commonly used in optomechanical experiments. So the techniques that are developed in the course of our experiment we think can be readily applied to also microwave to optical frequency conversion essentially by using the same oscillator as an intermediary. And we look forward to exploring it.

Sebastian Hassinger
Is it safe to say sort of like, I mean, very crudely, that you're you're taking electrical, electromechanical, I mean, microwave frequencies And having that vibrate the optomechanical the device, and then that essentially is flipping on and off the or affecting the optical regime.

Mohammad Mirhosseini
So like another one.

Sebastian Hassinger
It's essentially just bridging those two worlds in a sense.

Mohammad Mirhosseini
No, exactly. It's exactly what you said.

Sebastian Hassinger
Yeah.

Mohammad Mirhosseini
It's not crude at all. You can think of an electrical signal shaking a the di the membrane on a speaker and then laser light reflecting off of it dance on the wall.

Sebastian Hassinger
Yeah, right, right, right, right.

Mohammad Mirhosseini
That's exactly what's happening.

Sebastian Hassinger
It's so mind-blowing to think about that happening at that tiny, tiny scale. It's funny because it feels like, you know. What you just described, we have direct physical experience.

Mohammad Mirhosseini
Something like that.

Sebastian Hassinger
So we have good intuition for that kind of speakers, microphones, light bouncing off of surfaces that are vibrating. And yet it still seems so impossible to try to translate that intuitively into that microscopic scale that you're working on. It's really fascinating. What do you like? That last paper was December. What's sort of the next phase of work for you and your group?

Mohammad Mirhosseini
So in terms of the applications, you already hinted a few areas where we are thinking about we think that the coupling rate between electrical and mechanical domains can be improved quite a bit, and that would enable us to do more sophisticated operations that are relevant to quantum computing. That's one area we are pursuing. Taking these mechanisms and using them for microwave optical transduction is another area which we are this is something we are also exploring actively.

Sebastian Hassinger
I'm using the data and copying out I would like to add it there.

Mohammad Mirhosseini
I would add a third component which we are very excited about, which is if these mechanical oscillators also have a lot in common with qubits when it comes to their sources of decoherence. Usually we treat decoherence as some existing environmental condition or precondition that these devices have and they they suffer from. But in terms of for superconducting qubits, we know that two level system defects are Essentially, defects within atomic defects within material systems that are the bottleneck.

Sebastian Hassinger
I told the people that do it. But um it turns out that the day of the music is right so Um allow a little bit of stuff to get money and Right.

Mohammad Mirhosseini
It turns out that the same two-level system defects also Create the coherence in mechanical oscillators. They are also the major source of the coherence in these oscillators. But despite that, the mechanical oscillators tend to achieve much better lifetimes and coherence time than the qubits. That's a little bit puzzling. It's partially understood by the community, and it's been often attributed to the size of scales. These oscillators are thousands of times smaller than qubits, and that just allows them to have less atoms and less defects within those atoms. But there are effects that are also a little bit less Less trivial than that. So, one of the things we are trying to understand is to look at the interplay between mechanical motion and these atomic defects. in these systems. And we think that would be very interesting because, well, it can potentially allow us to make better mechanical oscillators, but a better understanding of what these defects are and how they interact. With other systems can also be beneficial to superconducting qubits ultimate.

Sebastian Hassinger
Because you're seeing similar behaviors from similar roots in two very different settings. So the comparison between the two may actually be very illuminated.

Mohammad Mirhosseini
Exactly, exactly.

Sebastian Hassinger
Yeah, interesting. That's fantastic. That's terrific. Well, thank you so much, Mohammed. I've really enjoyed the conversation. I appreciate your time.

Mohammad Mirhosseini
Thanks for having me. I also enjoyed it and it was great.