Lecture Transcript
Ashok Ajoy:
It’s a pleasure to be here and also to kick off this very exciting seminar series. I was going to tell you about some work that we’ve been doing in Berkeley on essentially quantum assisted NMR and MRI. Since this is the first talk in the series, I also wanted to keep it more introductory and give a flavor of the field to maybe students and post docs who may be joining. Let me just start by saying that our group in Berkeley is very interested in nuclear spins and, in general, in NMR and MRI technology.
NMR and MRI is actually a very versatile and mature technologies. Perhaps MRI is most familiar to most people because when you go into the hospital, the MRI machines are everywhere. Actually NMR and MRI are nuclear spins. The nuclear spins are a ubiquitous object. They can be found in everything, you take a crystal out of the ground, for instance, and you can see these nuclear spins. Here’s a crystal of fluorapatite and it’s got these chains of nuclear spins. And often because of the properties of these nuclear spins, they’re also very, very coherent. Coherence times of a few seconds are easily possible even in the solid. And often have very long lifetimes. T1 lifetimes, for instance, of these nuclear spins can be, in some samples, even up to a day.
NMR technology has been around for more than 60 years. And is the means by which you can probe these nuclear spins. Why NMR is so popular is because these nuclear spins don’t participate in chemical reactions. And each nuclear spins reports on its local chemical environment. If you’ve got some small molecule of this sort, that you can read out a chemical fingerprint of this molecule and all the nuclear spins within it. NMR became an indispensable tool for chemical structure.
But it’s very important to realize that nuclear spins are really the engines of NMR. And nuclear spins are in everything. You can take any object and it has nuclear spins. And you can control these nuclear spins by applying RF controls on the outside. While nuclear spins are in everything, the main problem of the technical challenge in both NMR and MRI technology is it’s very hard to see them. And to see these nuclear spins, you have to prepare them. And this often requires large, big magnets. Which are very expensive. And you need to detect them. And you interrogate them from the outside using RF techniques, using some sort of coils.
In some sense, therefore, these two levels of preparation detection are more technical aspects of how NMR and MRI are actually done. But the engines of NMR are these nuclear spins which are in the center. And so when you put these spins in the large magnet, essentially these spins align in the direction of the magnetic field. This is what’s called polarization, the amount of alignment, and this polarization value is very, very small. At room temperature, even in a large magnetic field that may cost you about a million dollars, the polarization level is only about few part of a million. For carbon-13 nuclear spins, for instance, at seven tesla, which you can only produce with a superconducting magnet, the polarization is 10 part of a million, which means if you take 100,000 nuclear spins, basically one spin is aligned. Everything else does not contribute to signal.
There is a polarization challenge. There’s also a detection challenge, because these coils are usually macroscopic in size. While NMR and nuclear spins have so much information, it’s very hard to get information about these nuclear spins at really length scales that are of interest to several people, including interfaces. Large frontier area in NMR is actually to shrink down the size of the NMR detector so you can see NMR at places where it’s harder to see. And finally, it’s very important to realize, when you put these nuclear spins in a magnet, all nuclear spins are polarized equally. All nuclear spins of one species are polarized equally. Which is fine and you can do very interesting things with it. But nuclear spins are more powerful, especially because these nuclear spins form a fully coupled quantum network.
So these lines here indicate nuclear spins that are coupled to each other with some sort of coupling, which naturally occurs, the dipole-dipole coupling. And if one could polarize spins in any manner that one wishes, then you can really exploit this fully connected quantum mechanical system for many other things, including studies of quantum statistical mechanics, quantum simulation and so forth.
Broadly, one could characterize this paradigm of NMR as saying that NMR is a study of nuclear spins and the paradigm of NMR that has been going on for the past 50 years is that to study the properties of the nuclear spins, you’ve got to take the sample to the NMR machine. You either take a liquid or a solid and you crush the solid and you take that, put it in a test tube, you walk down to the NMR facility, which houses this large magnets and you put the sample into the magnetic field and then you’re able to get an NMR spectrum.
But that said, this has been a very successful technology and it has touched several branches of science. Many things started at NMR first and then they translated to other fields. And this includes 2D spectroscopy, a technique for studying the structure of molecules. And MRI is now a large area in itself. And also several ideas in quantum computing and quantum control and so forth, all came in NMR first. There’s good reasons for this, related to the fact that these nuclear spins are very coherent and they’re very easy to control.
So with that small introduction to NMR technology, I want to give you some flavor of, in general, kind of ideas that we and others in the community have been thinking about. And this involves inverting the paradigm of NMR that I was just telling you about. What we want to do is to use these particles, and it’ll turn out to be particles of diamond for good reason, but these particulate materials. And the goal is to make these particulate materials to act as integrated polarization injectors, as well as detectors.
So the same particle, upon shining light on it, would play the role of a big magnet. A polarization source as well as a polarization detector. In some sense, you can compress the large NMR magnet as well as the detection coil within this particulate volume and you can address it with light. This is the broad, thematic goal that we’ve been pursuing for the past several years. And how we envision doing this is to use diamond material and to use the nuclear spins and electronic spins within this diamond material, and I’ll tell you more in a moment. The idea is to, as I’ll show you, by shining light on these diamond materials, you can actually polarize the carbon-13 nuclear spins within these diamond particles to a very high number.
It turns out that these nuclear spins can be polarized effectively to greater than thousands of tesla level. This basically means that to get an equal valent polarization level, you would have to put the sample in, or put these diamonds in a very, very large magnetic field that’s basically impossible to achieve through current other technology. And this can be done with modest laser power, as I’ll show you in a moment.
These nuclear spins are very good as polarization sources because they exist everywhere, including on the surface of the diamond material and nuclear spins are very long lived. I’ll show you, they can have T1 approaching one hour or so. Once you prepare these nuclear spins, they can act as very good polarization sources to the outside world. Then we envision using the electronic spins within the same diamond lattice. These are NV centers, as it will turn out. As the very good NMR detectors, the outside walls of the same lattice houses, in some sense, the nuclear spins detectors, the polarization source, as well as there’s these polarization detectors within the same volume.
And most importantly, one can use the fact that these are particulates. Now, I use nano diamond in a more interchangeable sense. The particle sizes can range from hundreds of nanometers to maybe even micron sizes or large macroscopic sizes. But it’s important to realize that these particles can be actually transported from one place to another. Either using several means of mechanically transporting them, or by using the surface fictionalization of the particles themselves.
And this might really allow us to access new forms of NMR that were not available before and importantly play on this paradigm, in the sense that now, potentially, one could bring the NMR spectrometer to the sample, rather than having to take the sample and bring it to the NMR spectrometer. These particles will be the NMR spectrometers, but you can actually carry them into the sample. And so now the sample could be a vast, large array, in the sense you can study things inside a cell. And other things which are usually things in their local environment, which are usually very hard to bring to NMR spectrometer. But you can still probe these nuclear spins that exist in all these materials by taking the NMR spectrometer to the sample. And that’s a broad goal that we’ve been pursuing.
As I was telling you, how we plan to accomplish this is to use a combination of the NV center and the carbon-13 nuclear spins in diamond. And this really relies on the very beautiful properties of diamond material in itself and also diamond processing techniques by which you can, for instance, deplete or enrich the carbon-13 concentration, as well as you can produce NV centers at more or less any concentration that one desires to a point. The material itself is rather versatile. And this forms the basis of this technology. Here is a picture of our 100 micron size particles, and they have this purple tinge, as many of you know, because they host these NV centers within their lattice.
Let me just say that all of these properties stem from the remarkable properties of the carbon-13 and diamond. And I think this has been less studied than the NV center, partly because the NV center is easy to interrogate through light, as I’ll show you. And there are many groups who have studied them. But the carbon-13s have very interesting properties in themselves. And together, now this forms a very exciting system. I’m just going to flash some interesting results that we got about these carbon-13 nuclear spins.
First of all, as I’ll show you in this talk, and this is going to be the subject of this talk, these nuclear spins can be very highly hyperpolarized. This is a single shot NMR spectrum from a small chunk of diamond. And that’s a remarkably strong NMR signal. Perhaps the only normalization I can state is that you get your equivalent signal, or the signal has been accelerated by a factor of about a billion to 10 per 10 times. And to get an equivalent signal one would have to wait for a few years signal averaging. You can really probe these carbon-13 nuclear spins in NMR very easily by hyperpolarizing them, as I’ll show you.
They’re also very highly coherent. And this is also somewhat surprising, if you like, so this is a bulk solid with carbon-13s that are enriched inside this diamond material. And the T2 diamonds are in excess of a few seconds. You can set them processing in a magnetic field and they process of about a billion precessions before they decay. They are very highly coherent. And they are also rather long lived. It’s very easy to produce T1 times that exceed a few minutes, tens of minutes in diamond. And I should say that several of these properties are material dependent. And they really depend on both the carbon-13 concentration and, importantly, on other paramagnetic defects that exist within the lattice.
But these are not very well specially produced diamond materials. These are diamond CBD crystals, in this particular slide. But the properties here are quite remarkable because they are all bulk properties with a large concentration of spins, and these spins are all interacting with each other but still they are very highly coherent.
And so this is going to form the basis of all of the things that we’re doing. And I just wanted to flash what things one could envision doing with this property of combined NV and carbon-13 in the diamond lattice. One could really envision making laser hyperpolarized NMR devices, and I’ll tell you more about it. But essentially using the high surface area of diamond particles so that you can shine light on these diamond particles, polarize the carbon-13 spins and through them polarize other materials. Thereby making a polarization source through light and having a means by which you can deliver hyperpolarization and conventional NMR. This is a very exciting topic, with several groups working on that.
You can also make very highly coherent quantum sensors, partly because these nuclear spins are very highly coherent, one could make nano scale NMR detectors and perhaps this is the most widely studied topic in the NV science, using the NV center as a very good nano scale NMR detector. You can also do more interesting things using the fact that these nano diamonds are actually transportable, so you can make an NMR detector that can be targeted to a particular place and moved either using optical techniques or mechanical techniques.
There are a lot of implications for a quantum information science, especially because if you can inject polarization into a material, then you can produce spins states that are far from equilibrium. And I won’t have much time to talk about it, but here is actually a spin shell that we first observed for the first time. You’re basically using the fact that you can produce these states that are far from thermal equilibrium. And finally, you can make new forms of contrast agents in MRI. And I’ll tell you more about it but you can make contrast agents in MRI that are bright in more than one dimension. This is bright in optics as well as MRI, simultaneously. And here there are some advantages of doing so.
So what I wanted to lay out, since also this was the first talk of the series, is all the beautiful properties that one could envision getting from these material. And in this talk today I’m only going to focus on this first topic. But I’ll be very happy to answer questions. With that introduction, let me just quickly flash the properties of NV center that actually make all of this happen. As many of you know, the NV center is a defect in diamond which happens when you take a diamond lattice and you remove a carbon atom, you replace with a nitrogen atom and you have a vacancy in its neighboring site. And this NV center has two very important properties, it can be polarized with light. It’s a spin one, and you shine light on it and all the spin population can be brought to this ms = 0 state. It also has an optical read out, so you can read out the spin state because of a differential contrast of the spin state with light.
So one electron spin state is slightly brighter than the other in florescence and then you can read it out. And I assume the audience knows this, I’m not spending much time about it. But let me just give you a flavor of why this happens. Of course, this is the energy level structure of the NV center in the ground state. And if you start from the ms = 0, and you shine green light, you have a cycling transition and you get fluorescence. But if you start with the ms = ±1 state, then you go through this non radiative singlet state here and this gives rise to the lower optical contrast but also spin polarizes selectively to the ms = 0 state. Both these properties are connected to the energy level structure of the NV center.
So what we really set out to try to do is exploit this first property of the NV. The fact that it can be optically polarized with light. And in some level, this is quite a remarkable took. Because these are spins which add [inaudible 00:17:12] magnetic field and with a rather low amount of optical power, can be brought to a state that is close to zero spin temperature. In fact, it would be meaning that you maybe have spins equivalent to being at zero temperature or at infinite magnetic field. And while other systems have a similar property, the power of the NV is that it can be done really at room temperature and at zero magnetic field, so with almost no infrastructure.
So the idea is that we could try to use the NV center to polarize other nuclear spins in their environment. We shine light on the NV centers to align them. And these blue arrows are supposed to note nuclear spins. And you can transfer the spin polarization from the NV to the nuclear spins. Thereby hyperpolarizing them to a level that’s far beyond what’s possible with a sub conducting magnet. Classically, what people have been trying to do is to transfer polarization from an NV to an analyte directly. This has its own challenges. We decided to go a slightly different path, to actually use the carbon-13 nuclear spins in the diamond as a relay channel so you would transfer polarization from NV to carbon and then from carbon to the analyte.
So essentially, a program of this sort where you have diamond particle, you have carbon-13s both in the lattice and maybe on the surface enriched. And then you shine light on the NV center as you transfer polarization from NV to carbon, carbon to the layer, perhaps, and then from the layer to the outside world. Exploiting the fact that these diamond particles can have a very high surface area and can be functionalized. And the carbon-13s are very long lived. And they can be selectively enriched on the surface. And so this is the general goal.
Today I’m going to show you about how we can actually do this, in a few slides. In a cartoon sense, this is what we want to do, exploit a high surface area of diamond, shine light on the particles, transfer polarization from NV to carbon and from carbon to the outside. And thereby, you can even imagine flowing a liquid through this diamond dust and then having a hyperpolarized liquid coming out.
Let me say that while this is a very interesting direction, there were some technical challenges that we had to overcome. And the most important challenge is the fact that while the NV center can be optically polarized, it polarizes in a very special axis, as you well know. It polarizes along the N to V axis in the lattice. If you consider a collection of diamond particles, and you’ve got all these particles that are pointing in random directions and the NV centers within them are polarized, but they all polarize in random directions. The net polarization is zero, effectively.
So in some sense, the technical challenge is if you have this resource of polarization, how can you transfer polarization from randomly oriented NV centers to the carbon-13s in the lattice, so the carbon-13 are always in the lattice so the carbon-13s all polarize in one direction. And technically there’s another way of looking at it, at even a modest magnetic field, the NV electronic spectrum is very broad. It’s about six gigahertz broad, let’s say. It’s very hard to control these in the electronic spins. But what we want to do is to transfer polarization to carbons, and the carbons being spin half have no orientation dependence. And we want to polarize them all in one direction.
It turned out to be actually, while it was a challenge, it turned out to be overcome by smartly choosing an operating regime that is closer to zero field, and basically the idea is, if you work closer to zero field, then all the NVs have roughly the same frequency around the zero field, I think they are 2.8 gigahertz.
I don’t want to go too much into details, but let me just show you how this experiment is performed. What we do is we shine laser light continuously on the diamond particles and we apply chopped microwave radiations. We sweep microwave frequency across the NV electronic spectrum. And this is done at low magnetic field. And then you detect the polarization by carrying it into a high magnetic field. The only reason for doing this is that you can get a gold standard measure of the exact polarization levels that’s in the diamond particles.
Here is a result, here on the right. The red line is a single shot NMR signal that’s got after hyperpolarization in the tenths of seconds of optical pumping. And the blue line is an NMR spectrum that’s obtained conventional seven tesla NMR. It takes about seven hours to get that kind of SNR. Effectively, the signals have been accelerated by a factor about 10 million or so. And at the field at which you are hyperpolarizing, which is, in this case, 38 militesla, the polarization is enhanced by a factor of about four or five orders of magnitude. It’s also important that the resources that you need for polarization is actually quite less. To polarize one micron particle, you only need about 30 nanowatts of optical power and two nanowatts of microwave power. Very benign resources.
There are also some very interesting properties of the polarization itself. I told you that to polarize we actually sweep the microwaves. It turns out, if you sweep the microwaves in one direction, the carbon-13 polarize in one way. And if you sweep the microwaves in the other direction, the carbon-13 spins polarize in the other direction. Actually sweeping a sawtooth pattern, you get zero net polarization. And so you can really control the sign of this carbon-13 polarization, so you can imagine that you’ve got diamond dust to air particle randomly oriented, but the carbon-13 spins can be made aligned or anti aligned to a magnetic field, just by the direction of the microwave sweep that one applies.
I don’t want to go too much into detail, given the short time here, but the mechanism of why this happens is actually quite interesting. It turned out to be completely orientation dependent. It doesn’t matter about the orientation of the particles themselves. And it relies on the fact that NVs have spin equals to one. I think this is the most important property that we exploit. The NV is a spin one, and there’s a hyperfine coupling only when the NV is promoted to the ms ± 1 state. In the ms = 0 state, the carbon spins are polarized more or less in the native magnetic field. And in one line, the way the mechanism works is by setting up a pair of Landau–Zener transitions in the rotating frame. And one of these Landau–Zener transitions has a large energy gap and the other one has a small energy gap.
I think hyperpolarization is a means to bias the system, you start with nuclear spins that are equally polarized, so this is what this means here, the alpha up and alpha down are nuclear spins states. And the ms = 0 NV electronic state. And these nuclear spins, you’re starting with a state which is equal probable nuclear spin up or down. And what we want to achieve through hyperpolarization is a means by which, if you start with nuclear spin up, you remain nuclear spin up. But if you start with nuclear spin down, you flip nuclear spin down to spin up, introducing bias in the system.
And this is accomplished. Because when you sweep across the Landau–Zener transitions, the gap corresponding to nuclear spin up is larger, so this transition is adiabatic. But this transition is diabatic. And so these is a bias and one transition is adiabatic, so nothing happens to the spin state. The other transition is diabatic and that spin state then will flip. And then if you do this enough number of times, it’s like building up a ratchet. And then you build up polarization in that way. But importantly, it’s completely orientation independent, off the NV.
So let me just conclude now by showing you some things that you can do with it. We’ve made a hyperpolarizer device. It’s a small, little shoebox sized device. It really exploits the properties of these NV centers, because polarization is quite easily accomplishable with off the shell technology. It’s really … now, low magnetic fields of the size of a refrigerator magnet, very low laser power, very low microwave power and it works at room temperature. The idea is that you put diamond dust into it and you press a button and then you get these particles that are hyperpolarized.
If one envisions now, transferring this polarization from the carbon to the outside world, then you can really a have a retrofit hyperpolarizer that you can stick on any NMR machine that can deliver polarization to a material that flows through this diamond particles. Here is a blow up of the device. It has a small diode laser. And the microwave sources, which are just off the shelf VCOs, that we exploit. And the coil and the magnetic field is just produced with small [inaudible 00:26:02] magnet or a coil.
I have a small movie, since I think I have two, three minutes more. Let me just play this movie. I hope it still plays when I do it remotely. But … It’s also got some music. But I’ll narrate over it. Here’s this device. It’s the size of a small shoebox. It’s got a laser source. And all the microwave amplifier and source is all within the same device. Here’s a pencil for scale. It’s the size of a small pencil and you put the sample into this device, this is diamond dust, essentially. And you shine light. We actually carrying it into an NMR machine so that we can measure the signal enhancement. That’s a one shot NMR signal from these carbon-13 spins. And you can polarize them up or down.
The important thing is to get an equivalent signal to noise, with conventional NMR, you’d have to wait for a few a years of averaging. Let me just quickly show you what else you could do with it. Maybe in one slide, already with the things that we have. And this is actually quite an interesting idea, the fact that you can now use these diamond particles, along with the fact that they can be targeted and functionalized as a contrast agents in MRI. And this is obvious, because the array is so highly hyperpolarized. Obviously, when you put them in to an NMR or MRI machine, they light up.
But it’s more interesting than just that idea, because of the fact that these diamond particles also fluoresce. In this particular, very proof of concept experiment, we’ve taken diamonds and arranged them in a ring. And you can see them fluoresce, as well as you can see them glow in MRI. You can see them bright in two dimensions. That’s why we call them dual mode contrast. But this is very interesting for a very important point. And that is that fluorescent imaging is real space imaging. You actually have a camera and you’re looking at these spins in real space. But, importantly, MRI imaging is a case based imaging modality. You’re actually collecting the image … How you do MRI imaging is you’re collecting the image and you’re collecting cake slices and then doing an inverse wire transfer to produce the image.
Importantly, you’ve got a means by which you can see these diamonds, but you can see them bright in two Fourier conjugate dimensions, simultaneously. And this really opens some new perspectives on exploiting some ideas similar to compressed sensing, but hardware compressed sensing instead of software compressed sensing. Because if you’ve got a sparse collection of these nanodiamonds, you can exploit the fact that in Fourier it’s very efficient to be able to detect them so you can do a few MRI slices and then just localize their position through fluorescence better. And both optics and MRI have complimentary advantages in the sense of depth penetration, imaging through scattering environments and so forth.
So I think this is a very interesting possibility that comes out of the fact that these nuclear spins are what’s imaged in MRI but you’re actually imaging the NV center fluorescence and optics. Let me just conclude by, I wanted to just mostly give a favor of the kind of work that we and others in the community have been thinking about. But just really exploit the wonderful properties of nanodiamond material and the fact that they host these electronic spins as well as nuclear spins. And they have very good surfaces. And they can be enriched, so they now properties that you can almost list together. But, importantly, one could envision this very exciting, new paradigm of taking an NMR spectrometer to the sample. This might really open very new perspectives of using the power of NMR and doing NMR on samples that are completely inaccessible otherwise.
There are many directions it can open in a variety of fields. Let me acknowledge all the people that have been involved in this work, several students at Berkeley. And here are the references, if that’s of interest. Thank you very much.
Male 3:
Thank you very much for this very interesting presentation. Very didactic presentation. You mentioned that the diamond, nano diamond surface can me enriched by carbon-13. Did you perform some surface treatment for that? Or-
Ashok Ajoy:
Yeah. Thanks for the question. I saw your recent paper on something similar. We have been thinking about this and we have some preliminary data but it’s not too great. Let’s just say that the idea, from a purely physical point of view is exciting for an important reason, the fact that I told you the carbon-13 nuclear spins have a long T1. But the T1 is still restricted by the fact that they interact with paramagnetic spins in the lattice. And that’s because to make the NV centers, you’re bombarding the diamond lattice with a lot of radiation and you’re producing paramagnetic spins. And these paramagnetic spins in the second order process, interact with the carbons. And that’s what causes this one minute T1, let’s say.
Ashok Ajoy:
But if take pure diamond and you look at the nuclear spins within them, the T1 in pure diamond can be very long. I think some people measured it and several hours at room temperature. It’s interesting to consider, if you can make this layer of diamond. But that layer would be pure diamond and then the carbons inside the diamond will relay polarization to this exterior layer. And potentially if one could do epitaxial growth, then you could have an interface that’s seamless enough and then you can polarize this layer.
Ashok Ajoy:
There are challenges, about epitaxial growth on diamond particles just given on the fact that the facets are different. And I think on a crystal, I think the odds are much higher to be able to accomplish this. And we are trying to do this basically on CBD crystals.
Female 1:
Ashok, I have also a question. Obviously everybody is excited about possibilities to external polarization transfer to external molecules. Maybe you can tell a little bit more about perspectives in this topic.
Ashok Ajoy:
Yeah. This is a very good questions. It’s important to realize that hyperpolarization is a really intangible resource. And it could have many applications. But you don’t need to polarize it to very large numbers to already make an impact in several fields. Just for a normalization’s sake, let’s say the carbon-13s are roughly at a polarization level of, let’s say, 5,000 tesla. You need to be able to polarize with an efficiency of less than one percent, to already be able to polarize more than what you can get with a superconducting magnet.
Ashok Ajoy:
And if your reference point is MRI, if you want to polarize water to larger than 1.5 tesla, just looks like the resources required are not very difficult. I’m actually, quite honestly, quite optimistic, myself. I think the challenges arise from two important reasons. One is a paramagnetic defects, which are likely on the surface of the diamond that can act as polarization sinks. Although you use the carbon-13s, the T1 of the surface nuclei maybe very short. And this really requires the diamond surfaces to be passivated. I think this idea that I was mentioning on maybe making an enriched layer of pure diamond could be a very interesting pathway to solve both these problems at once, allowing you to produce a very high effective surface area because you can make 100% enriched carbon-13 on the surface.
Ashok Ajoy:
And this carbon-13 could have potentially a very long lifetime and could act as a relay channel to the outside. I’m very optimistic. I think to get polarization level of few tesla through light should be accomplishable with good materials processing.
Male 4:
Sorry, a question from myself. [inaudible 00:35:27] from South Africa. How thin a layer do you need for this polarization to be effective because of the carbon-13 there?
Ashok Ajoy:
That’s a very good question, so the idea, in that modality, would be to use spin diffusion to transfer polarization from carbon-13, the diamond to the layer itself. And then the layer then, has advantages that it has a long T1, so you can build up a large amount of polarization within it as well as it’s highly enriched. It has high surface area to be outside.
Ashok Ajoy:
And so the thickness of the layer would depend on the amount of time that you have for spin diffusion. And this is a very temperature dependent thing. If you’re working at room temperature, at a low magnetic field, the carbon-13s in the diamond have a T1 of … it’s a field dependent thing, but somewhere between, let’s say a few minutes. And a few minutes I think you can travel something between 10 and 100 nanometers in the lattice. That would be the size of the layer that you would make. If you were to lower temperature, I think then you could have a thicker layer.
Ashok Ajoy:
But the important thing is that even if the carbon-13s themselves have a T1 of only a few minutes, the layer, specifically the layer was made out of pure diamond, and the interface is accomplished as an epitaxial interface, then the layer itself could likely have a T1 very long T1. You could keep building up polarization. Essentially, you would just keep the laser on. And then the polarization on that layer would grow to a large number. And then transferring polarization from that layer to the outside, there’s been a lot of work in NMR over the past 50 years. And many approaches that one could accomplish use in cross polarization or Overhauser techniques, etc. I think the main challenge so far has been the fact that the surface of the diamond may not be the most controlled at the moment.
Male 4:
Thank you.
Female 1:
We have a question from audience, it is from [Annalise Fav 00:37:37]. Question is, does functionalization of diamond affect its ability to hyperpolarize a sample?
Ashok Ajoy:
Not that we can see. But it’s important to qualify this by saying that the polarization that we measure, even though we use nano diamonds, we are still measuring bulk. We don’t have a means by which we can separate the polarization exactly on the surface and within the diamond particle. We do not see any change in the polarization levels based on whether the surface is functionalized or not. To the extent that the laser can penetrate through the small layer, the polarization is unaffected. It’s interesting to ask what the polarization would be on exactly on the carbon that’s at the surface. I think that’s still an open question. If there is a means by which one can transfer carbon from inside diamond to a layer, then that would automatically answer that question.
Female 1:
There is another question from Professor Agafonov, France. What is relation between the size of nano diamond and hyperpolarization?
Ashok Ajoy:
Yeah, that’s a very good question actually. Indeed, we observe that nano diamonds, which are smaller in size, have a lower polarizability. And we went to really deeply study, and actually in collaborative work with Olga, there are origins for this loss. And I think this is largely connected to all kinds of defects and lattice dislocations that come upon materials processing that one uses to produce these small nano diamonds. Especially crushing and milling. The lower polarization is due to the lower T1 of the carbon-13s in these nano diamonds. For reference, the carbon-13s in a micro diamond are polarized to about 1%. That’s equivalent to his 5,000 tesla level. In 100 nanometer particles, it’s about 50 times lower polarization.
Female 1:
We have one more question from Alexander [Tellair 00:39:59]. Would you be able to use silicon-vacancy defect in nano diamonds to perform hyperpolarization of C13?
Ashok Ajoy:
Yeah. That is a very good question too. So, importantly, as I told you, the NV center has two properties. The fact that it can be spin polarized and the property that it has potentially a spin contrast so you can read out the spin with light. And in this particular modality, when you use it for hyperpolarization, you’re only using the fact that it can be spin polarized with light. Any other defect that can be optically spin polarized, can be used for the same task. And I know that in silicone carbide, the defect there, that also can be optically polarized at room temperature. It does not have a very good spin contrast at room temperature, but it can be optically polarized at room temperature. And that could be used as well.
Ashok Ajoy:
But I think in a broader sense, though, the fact that you can make both the polarization injector, as well as the detector in the same nano scale platform, I think that can be very powerful, because the carbon-13s would act as the injector, and then the NVs within it could maybe sense the polarization. And you can light up NMR signals selectively from particular regions in the sample that you’re interested in. That’s a very interesting direction to go.
Female 1:
And we have one more question. What is nitrogen vacancy concentration that would improve polarization efficiency?
Ashok Ajoy:
Yeah, that’s a very good question too. And we’ve studied this in detail. The concentration that these NVs are anywhere between … I would say roughly a few part per million, so it’s like between three and seven or so. Increasing the NV concentration, I think that’s a very multifaceted question. It’s not very easy to answer. Of course, they have large NV concentration. The NVs interact with themselves and there’s dipole coupling between them. And this can make the polarization transfer somewhat inefficient. But I think even before you hit that regime, you hit the regime that the NV centers interact with a lot of other defects in the lattice. And that’s because when you try to go to these high concentrations, in the process of making these high concentrations, you’re also polluting the lattice with a large amount of lattice defects and damage.
Ashok Ajoy:
And this basically messes up the T1 of the NV. There is a sweet spot currently with materials processing, and we’ve discovered with Olga and Alex Shames that perhaps using new forms of rapid thermal annealing and solve these problems and they allow you to heal the lattice, in some sense, and allow you to go to high concentration of NV but yet maintain the T1 and T2 and therefore obtain polarization transfer boost. I think the field itself would be enriched by more materials. I’m just trying to tackle this challenging problem of asking how we can maintain spin properties but yet elevate concentration levels of the spins.
Male 3:
So it’s regarding the T1 process of the carbon-13 spins, you mention it’s keeping it in the field. And I think it’s well described in your paper, so with this [inaudible 00:46:08] measurement, where you measure, I think, up to seven tesla the realization time, do you have actually an idea what is the mechanism for T1 realization at high field? I would think about one tesla or even about 10 tesla, which I think [crosstalk 00:46:26] measure so much. But maybe one could imagine what would be the realization mechanism?
Ashok Ajoy:
Yeah. That’s a very good question. And I don’t want to hold up the next speaker, but I have also a slide on this that this work, that you mention. But indeed, this is data from two different fields and the carbon-13 T1 is field dependent. And we get profiles of this sort where you’ve measured T1 as a function of field. This is just some random sample, let’s say. And the T1 has this profile, which looks like a knee, like a step function. And the T1 becomes flat after something like 0.5 tesla or 0.1 tesla.
And it’s very interesting to ask what the mechanism of high field is. I think it’s a combination of factors. And it’s also very sample dependent. People talk about second order processes that are mediated by lattice train of phonons. I think that’s the more conventional spin lattice relaxation, as in the conventional T1, when you look in a NMR book, that’s basically related to the phonons and the lattice. But at lower field, it’s really interactions with the paramagnetics that are dictating it.
Male 3:
Do you know if the T1s are involve and I assume that … Well, it doesn’t seem to be the case here, but if you get at higher field, or on seven tesla, you are able to have this cross effect between T1s of two different [inaudible 00:47:58] spins state. The hyper [inaudible 00:48:02] of T1 is about, I think, the [inaudible 00:48:05] about 100 of megahertz. And do you think there could be a resonance here and maybe a drop in T1 or … I don’t know if it’s seven tesla or nine tesla, even maybe that addition.
Ashok Ajoy:
Yeah. I guess that mechanism also exists. It’s also a function of orientation, whether you’re looking at a crystal or a powder, I think that make a difference too there. Yeah, it’s very interesting. Yeah, but I think what’s more interesting is if you have pure diamond the T1 is likely very long. It’s just in conventional NMR, if you have pure diamond, it’s very hard to see NMR spectrum because T1 is not only the time to depolarize, it’s also the time to polarize. In conventional NMR spectrometer and it has a T1 of one day, you have to wait for five days before you can press a button and get a spectrum. And then the spectrum is so weak because the spins are not very polarized.
So this is the reason why many people dope samples with the defects to basically reduce the T1 so you can actually measure stuff. But now if you use the fact that the spins can be polarized through other means and then that it aids to have a very long T1, by just depleting the sample from any defects, have a pure carbon lattice. This is an exciting direction, we have still not pursued it fully.