Lecture Transcript
(Olga Shenderova) I’m happy to announce today a talk by Carlos Meriles. Carlos was born and raised in Cordoba Argentina. He obtained his PhD in physics at the mathematics, physics and astronomy division of the Universidad Nacional de Cordoba under the supervision of professor Brunetti. His doctoral studies focused on the use of nuclear quadrupole resonance to probe structural and dynamical disorder in organic crystals. After graduating in 2000, Carlos joined the group of professor Alexander Pines, at the university of California Berkeley to work on nuclear magnetic resonance, his present field of research. In 2004, Carlos joined the department of physics at City College New Yourk where he now teaches as a professor. (Carlos Meriles) Thank you Olga and thanks to all the Adamas Nanotechnology team for this opportunity and before I get started, let me apologize to the audience. The conditions today are not going to be optimal. There’s been a number of recent developments that forced me to deliver this talk from far away from where I was supposed to be, and that’s actually the reason why I have to wear this mask. So if you happen to do not hear me well or things become unclear, please feel free to interrupt at any time and I’ll be happy to repeat or clarify as needed. Well, as you can tell from the title in the presentation, this talk is going to be centered on recent results that we have had in the laboratory on the optical activation and observation of transport between individual defects in diamond and to give you a little bit of motivation context. I think I will first start by saying that it all began a couple years ago in my group. We knew, of course, that optical excitation could spin polarized nitrogen-vacancy (NV) centers, and we also knew that optical manipulation could change the charge state of the NV center, and so the idea that we started discussing was: is it possible to spin polarize and inject a spin polarized electron into the conduction band of diamonds. Diamond is a nearly ideal platform for spintronics applications. It has central-symmetric structure, very low spin coupling, very low content of nuclear spins; and yet there’s essentially no studies of spin injection in diamond, precisely because the injection through standard means is rather impossible. It so happened too that at that point I started having some discussions with my friend, Marcus Daugherty in Canberra, Australia, and we started to think of the carrier in perhaps more general, more ambitious ways as a bus to transport coherently information between let’s say cubits in diamond or some other white bandgap semiconductor separated by a few microns. And so over the last few years, what we’ve been doing in the laboratory is to sort of gain the know-how and put together the physical infrastructure that is needed for these type of experiments. Now before I move forward, I want to emphasize that in no way my group is the only one working along these lines and, as a matter of fact, there’s a number of other groups working on related topics and, more broadly, applications of charge manipulation to different forms of sensing. For example, and as just as an illustration, I decided to highlight this particular paper by Jelezko and Nesladek groups in this case. What they do is they use a very strong green beam to illuminate an individual NV that happens to be inside a diamond where they patterned some interdigitated electrodes here, and the idea is that, with this strong beam, you can actually change the charge state of the NV and make it cycle between the positive, sorry, between the negatively charged and the neutral states, and every time a charged state change takes place. You are inputting an electron or a hole into the conduction or valence bands, and so in this experiment, what they do is they detect the current that is produced by these photo generated carriers, and so here on the upper right. What you see is an individual NV, the confocal image of that individual NV obtained by standard themes. What you see down here is the electrically detected image of that same NV as they displace these excitation beam from one point to the other. Now this is, as I said, just one illustration there’s. This is an active area of research and down here I’ve listed some of other groups that are also working on these or related directions, and this, of course, is just a reduced list. I’m sure, there’s many others. Now I know the audience in this presentation is rather broad, although I’m pretty sure you all know about the NV center, so I’m not going to introduce so much the energy center itself, rather I’m going to refresh some of the aspects that have to do with the charge dynamics of the NV and before getting into that, let me simply start by saying that NV centers and in general defects in semiconductors take different charge states very commonly; they take different charge states and similar as when you move in the periodic table from one atom to the other one and you add or subtract an electron, the same applies to color centers when they change the contents of electrons, their properties, optical and spin are changed. In particular, they change quite dramatically, and the relevant example for this presentation is, of course, the NV center in the negatively charged state; the NV minus has a ground and an excited state are separated by 1.9 eV, the equivalent of 637 nanometers. However, if you go to the neutral charge state, what you find is two states that are separated by 2.2 eV or the equivalent of 575 nanometers, and that essentially allows you to distinguish, for example, just using fluorescence one charge state from the other, in particular. In all the experiments we’re going to set our microscope so that neutral state appears as dark and the bright state is the negatively charged. Now there’s different ways that you can imagine using for controlling that charge state and here I’m just going to give you an example. So, let’s imagine that we start from the negatively charged state and we use red illumination and in particular, let’s imagine that the beam is pretty strong and so it’s sufficiently strong so that when we first excite the NV center from its ground to the first excited state, photon arrives before the system had time to relax, and so when that happens, then there’s an injection of an electron into the conduction band and the system as a whole transitions to the neutral state. Now you realize that once in the neutral state red photons don’t have the energy to excite can be zero any longer, and so the transformation is in this sense one-directional in the sense that it goes from NV minus to NV zero, but it cannot go back now. Of course we can recover the negatively charged state and, for example, we could use a green illumination for doing that, and so in that case, one first photon would excite the neutral NV into the first excited state and then a second photon would excite an electron from the valence band, and so we transition to a negatively charged state, and we also are left with a diffusing hall; so now notice that green excitation also has the energy for exciting and eventually ionizing and be minus. And so then, green illumination creates a back and forth between NV minus and NV zero. Now it so happens that the probability for producing NV minus is under green excitation, a little bit higher than the opposite, and so in throughout the presentation we’re going to see experiments where, essentially, we initialize, we say we initialize into NV minus that should be understood as a probabilistic initialization of the charge state of the NV. So before these experiments that I’m going to be presenting today, we’ve done a lot of work along these lines, and here I simply wanted to give you some taste. These are images confocal images where we use, for example, a strong green beam to initialize the left side of this field of view into the negatively charged state, a red beam to initialize into the dark, neutral charge state. And then we park a strong beam of different colors here right at the interface and then we use another weak red beam to read out what the distribution of charge looks like and what we find are different patterns that contain a lot of information in terms of how carriers, electrons and holes diffuse and get generated and ultimately get trapped, and so these papers over here somehow reflect the gradually improved understanding that we’ve gained over time about these dynamics. Now one thing that perhaps you already noticed is that the ability to show these images already implies that these charge patterns that you’ve created stay for long and so of course, then, with that in mind, you could imagine patterning these discharge states in a way so that you can store classical information, and so this is an example where we first prepare the ensemble of NVs into the negatively charged state. And then we use a red beam to go point by point and store some arbitrary pattern. That later on, we can erase and we write – and we can do this at will – and of course this can also be done in three dimensions and it has a resolution of it’s very much limited by diffraction. Now, most recently, we’ve repeated these same type of experiments, but in the presence of potential of electric fields created by electrodes in the vicinity, and we see these patterns evolving depending on the voltage that we apply and in this process we can extract information about the space charge potentials that are created through the ionization and diffusion of these carriers. Now, today, we’re going to move to something different, and that is different because, rather than looking at ensembles we’re going to be looking at very small ensembles, ideally individual NV centers, and to make this happen, we teamed up with people at Sandia National laboratory and used a very strong ion implanter to create these islands, some of which I show over here controlling the fluence. In that implementation process you can, of course, control how many NVs you create, and on occasions we find islands that contain just a few and these, as shown over here. Sometimes we manage to identify individual NVs that happen to be separated by a few microns, and we can use them now to start trying to see how the transport of carriers, the photo generation and transport of carriers between these individual NVs takes place. Now that these are individual NVs and not some a bunch of NV’s all clustered together, you can see here when we take the autocorrelation functions. You see, dips that go below 0.5 and indicate presence of an individual defect. If you take, for example, a time trace like this one, you see blinking that is characteristic of a defect, changing its charge, state from a dark or neutral state to a bright or negatively charged state. So with this sort of system now you can imagine the following type of experiment. So, let’s imagine we use two beams and we first use a green beam to initialize, both NV centers into the negatively charged states. So this will correspond to this initialization stage over here by the way, I’m going to be referring to these NVa and NVb as the source and the target, meaning that these is the one where carriers are going to be generated. This is the one that ultimately is going to reflect the capture of one such carrier. So the idea is that after we’ve initialized these two into NV minus, then we park the green beam into say, NVa, and we leave it there for time that it typically is on the order of a couple hundred milliseconds. And then we use a orange beam to go and see what the charge state is now. This is the type of results that we observe, so this is what we first get if the there’s no parking at zero, milliseconds, the evolution time or the park time. This is what happens as these park-time increases and what you see is essentially that the target NV becomes less and less bright, and this we interpret as the result of the capture of a hole that was generated from NVa and that rendered the target NV. Initially, in binary state now a neutral NV, this plot over here shows perhaps a more quantitative evolution for different laser powers, and perhaps the takeaway message when looking at this image is not so much the main, the main plot itself. Rather, this insert that I’m showing here in the in the upper corner. These bleaching rates are a function of the laser power during the park. What we see is a quadratic behavior, and this quadratic behavior is already consistent with the fact that there’s a two photon process associated with the ionization and recombination at the NV at the source energy, and so this is already indicating that the source NV is actually the only source and there’s nothing else over there. That is also creating some background defects. As experimentalists, I think we all know that it’s very difficult to guarantee this, that I was just saying that the NV is the only one creating defects, and so we went to great lengths to make sure that was actually the case and one strategy that we used for. That was to make use of the spin states of the NV and a protocol that is called spin to charge conversion. Now I won’t get into the details as to how spin to charge conversion works, but I will just give you the main idea, which is the fact that the NV has in the ground state these essentially doublet between the zero and plus minus one. If the NV is in zero chances are that the that once excited the state of the NV is going to be this one and there is going to be no cross-relaxation with this manifold of singular states. And so, if that is the case, then there’s a pretty high chance that the that this second photon arrives, while in the excited state and then ejects the electron into the conduction band, whereas in the opposite case, if it was say, plus minus one, then there’s there’s a larger chance of this into system crossing and because the shelving time in this singlet manifold is higher, the rate of generating electrons in this case reduces okay. So essentially, what we have is a way to ionize the NV minus conditional on the spin state. It is all right, so this is actually the protocol t that we’re using it’s a bit more complicated that one we had before, but conceptually it’s relatively simple as before. What we do is we initialize the both the source and the target and these into the negatively charged state, and we finally will probe in the same way, but instead of just sparking a beam for a long time. What we’re going to do is we first spin and initialize the source NV into say ms equals zero. Then we will apply or not microwave at some frequency, and then we will apply a pulse. You can think of it as a red pulse that is going to ionize the NV and hopefully that ionization is going to be conditional on the spin state, and we then repeat that multiple times and we look at the result with an orange scan. So these that I’m, showing here on the upper right, are two images of the source, sorry of the target NV. This is before we apply these sequence, and this is essentially reference fluorescence that we start with, and this is what happens after a couple repetitions. There’s no way no microwave applied in this case and what we see is just some bleaching coming from the holes that are being generated in this process. Now we turn on the microwave, and so let’s imagine first, we set the microwave away from the resonance frequency of the ground state. It’s this case here on the right, and what we see is that the fluorescence that we obtain is pretty much the same as what we had in the absence of microwave, and that is simply because off-resonance frequency or microwave doesn’t have any impact on the NVb charge state, on the other hand, if we set the frequency to be on resonance with that transition in the ground state. What happens now is that the population of ms plus minus one grows, and so, as a result, we eject less carriers into the conduction band, and so we have we have less chance of a capture. So then, the target NV remains brighter than otherwise, and we can repeat that same that same protocol as a function of frequency, and so what we get down here is the equivalent of an ancilla aided magnetic resonance spectrum of the source NV , which agrees reasonably well with the traditional ODMR or optically detected magnetic resonance spectrum of that segment. Now, a comment that I want to make before moving forward is that the contrast that we get over here tells us that, essentially, is just this one NV , the one that is serving as the source for the ionization or for the charge state change of the target. Then so the picture that emerges is one where, essentially, we have a sourced NV that is sort of generating carriers radially and because we know the distance between the two, the source and the target, and because we also know the rate at which these carriers are being generated then we can calculate what is the probability of the capture and from it calculate what is the capture cross section. The whole capture cross section for an NV minus now. The number that we get is actually quite large and I’m going to get back to that in a couple minutes. It’s about 50 by 50, nanometers squared, which is a number that is surprisingly large. If you go to the literature and compare with what people have measured in samples, but let me let me get back to that in a second and before let me address the fact that we have this image of carriers emitting radially, and this is also something that we’ve confirmed by taking different pairs of NV samples. Different pairs of NV centers are separated by different distances, say, for example, this one has five microns. In this case, the two are like almost nine microns away and in each case we repeat the same type of experiments we had before and we measure what the bleaching rate of this the target energy is, and we plot that as a function of the distance and what we get is a dependence that goes with the inverse of the distance square, which is what we’d anticipate for a sort of isotropic emission of the of the carrier from the source end. Now the physical picture behind the observation of these large capture cross sections can be somewhat rationalized. If we start, if we realize that as a whole approaches the negatively charged NV, there’s a strong coulombic interaction that takes place, and so essentially what happens is now we’re going to be dealing with hydrogenic states and these hydrogenic states because of the long range of the Coulomb interaction is going to essentially create, is going to lead to these large capture cross sections that we see here. I don’t have the time to go into that into these in detail, but we’ve done calculations that partly – and I say partly because DFT doesn’t have the capability to really go to long distances. But at least the results that we get are consistent with this understanding of the formation of what we call a bound exiting an idea that, by the way, was introduced first by Mellax in the 1960s, and that I have to say most recently has been invoked in some other related articles here I’m highlighting two in this particular case. They sort of invoked this idea to explain observations of calorie recombination at single NVs in a pin junction. Most recently, the group of Natalie de Leon reported the observation of bound exiting states in silicon vacancies. So these are ideas that have been around and our observations are consistent with these. With these ideas now I’m going to use the last minutes that I have left to tell you about ongoing research activities that are related to these observations. Two in particular, one is more involved than the other one I’m going to start with a less complete picture, because I think it’s just going to be clearer in this particular case. We return to the geometry that we’ve been discussing so far and we keep everything unchanged, except that we’re going to be using electrodes to control the carriers that we generate induce a bias in the propagation. So in this plot over here, what we see is the is the bleaching rate for a given fixed laser power during the part, but different voltages that are applied now. In this case, we apply positive voltages, meaning that carriers that are generated here by the left, NV the source and e are tend to propagate actually to the left. That is against not towards the so the target ended and so, as expected. What we find is that the bleaching rate decreases to finally disappear, so we preserve the charge state by simply of the target of the target simply by redirecting the carrier’s holes in this case, in the opposite direction. Now something less intuitive is what happens when we change the sign of that potential. In this case, we apply positive, sorry and negative potentials, and in that case we buy as the carriers towards not against towards the target energy. And what we see is that the as we increase the voltage, the bleaching rate also decreases as well. And all these has been captured in this plot over here, where we show the a function of the applied voltage, and we find that there’s a maximum precisely when we don’t have any voltage applied now; I don’t think this is fully ready yet for prime time I won’t get into a detailed explanation. All I say is that, for now we have monte carlo calculations that sort of confirm almost quantitatively this picture and briefly there’s. There seems to be an interplay between the enhanced carrier density that we generate with negative potentials and the acceleration that we induce in these carriers that somehow combine to yield a lower capture probability. But this is something that I’m going to postpone. For later now, another aspect that I think is interesting and where we’ve done some more work, is this one. I think this one is interesting because it shows I’d say that the first application of something that may seem too many too fundamental. So in this particular case, we are repeating the same type of experiments as before, except that now we’re going to move the position of the part beam around the source and we’re going to be monitoring the target NV notice. We are changing slightly the language here because we’re using this to see what’s happening here at the source, we’re going to be referring the to the target as the probe NV , and so this one is just a standard, confocal image of that probe NV; this one is what we call the carrier to photon image or cpc image of that same source NV, as measured by looking at the probe NV charge. So we see essentially and notice that there’s an inversion in terms of the colors here when we have a source of carriers. What we find is a dip rather than a peak like over here right and that’s important, because now things are going to become a little bit more complicated. So you want to actually keep that in mind a comment that I want to make. Also is that if we were to change here, I’m indicating down here, we’re indicating what color we’re using for the part or in this case, what color we’re using for the readout during the confocal? And so, if, instead of using a green part, we transition to a red product. What we see is that essentially this image disappears, and that is consistent with the fact that the NV gets ionized once and there’s nothing else happening. So the charge state of the of the probe NV remains unchanged, and this is actually confirmed directly here in the confocal image. We don’t see the source NV simply because we ionize it almost immediately and we don’t have time to collect photons now. With this in mind, what we can do is, of course, image not just individual entities, we can image, say collections of entities, and this is an example. So, in this case we have the program be located over here and we have these islands of neighboring NVs. The cpc image that emerges is this one that I’m encircling here you can see. If I know the resolution is less good as or worse as compared to this one, but still you can you can see that there’s a one-to-one correspondence between the source and these that are acting here and what we can see in the cpc image now so far. We’ve been looking at say, other NVs, but let’s now make things a little bit more interesting. Let’s, let’s try to go and see, defects that are not necessarily the NVs, and one idea that we had was to look at silicon vacancies and see if silicon vacancies could also charge cycle. So what I’m showing up here is a region that we implanted with silicon to produce silicon vacancies and that these are silicon negatively charged silicon vacancies is shown in this spectra on this spectrum. There’s this peak that is characteristic of negatively charged silicon vacancies. Now, if yes and this image can be selectively, made selective to silicon vacancies using red excitation now, if we were to use green excitation and here we’re looking at confocal images, we already find something interesting and that’s the fact that there’s NVs over a region that is much larger than the region that we implanted with silica the way we interpret this observation is that the implantation of silicon produces lots of vacancies which subsequently diffuse during the annealing or produce. That is typical for producing the silicon vacancies or NV centers, and so the result is that then, these vacancies diffuse and pair up with some existing nitrogen that leads ultimately to the formation of a negatively charged NV . So this is already something that caught our attention. Now, let’s go back to these indirect imaging, and so let’s first focus on this particular region. Over here remember we were trying to see if the probe NV that is located here is not visible, because we’re illuminating with red. Here you can see a faint image of the program b when we illuminate the green okay, so this is our probe. This is what we scan, and this is the cpc image that comes out, and what we find is that there are some regions over here that seem to be emitting charges, and so that essentially translates into a bleaching of the probe NV . Now, if we were to compare this image with now the image of silicon vacancies in this region, what we see is that there’s essentially no correspondence. As a matter of fact, you see that there’s emission from places where there’s no silicon vacancies remember that red excitation precludes NVs from even producing a meaningful number of carriers. But if we were to also look at the at the confocal image via green excitation of that same reason, region, we also find that there’s no correspondence between say what we see here in the cpc image and what we have here for the NVs. Now perhaps we can make the case clearer if we actually move our spot from this point, where there’s lots of silicon vacancies to this other spot, where there’s no silicon vacancy there’s, essentially no fluorescence, and what we see here is again the presence of charge emitters that lead to a bleaching of the of the program, the confocal image of that same region yields no fluorescence virtually because, as I said, there’s no silicon vacancies. If we were to look at the NV centers, once again, we see that there’s no correspondence. We have emission from regions that do associate with NV centers and once again, I insist, NV centers via red excitation, cannot yield any charges, as we’ve already seen. So, of course, we are left with the question what those what those defects are. I mean on the bright side, there’s the fact that we can now see charging meters that are dark that essentially they do not fluoresce, at least in the range that we’ve looked at, but we are at a point where we don’t know exactly what those are. We have some ideas and some possibilities that I think are worth exploring, but there’s nothing concrete at this point now, on the other hand, there’s you can think of this sort of imaging as a sort of alternative to photo luminescence excitation and in the same way, you can imagine doing a sort of spectroscopy and through that spectroscopy gain some clues as to what that defect, that source charge emitter happens to be, and well we haven’t done this yet, but we’ve done some steps in that direction, and these are results here. What we do is we plot the bleaching rate for different red excitation powers, and what we see is this change if we were to plot the rate now as a function of the power that we use, what we find is this linear regime, so remember In the case of the np, we saw a quadratic dependence in this case. This is telling us that the process is one for not too far and then somehow mirroring what happened with the NV , which was active and the green excitation, and not on the red. In this case, if we were to use green excitation, we see that there’s no bleaching, so essentially the bleaching rate of the probe and the remains at virtually zero. So, whatever model, essentially we come up for what that charge defect or that charge can meter is has to now fit these spectroscopic observations. That, of course, can be completed with additional measurements at different ways. So, well, I think I’m going to conclude here. I can summarize these presentation by saying that we’ve been able to observe, to generate and observe the transport of carriers between individual defects, small groups in general and down to the level of individual defects. We use spin to charge conversion to filter out any potential contribution from background defects that are not taken care of, and we found that well in the examples that we looked at the NV was the sole source of these carriers. We also looked at dependence with distances to find out to conclude, essentially that the emission of carriers from the source defect is isotropic, at least in the unbiased case. We’ve also used all these that we learned to highlight or to expose charge emitters that well are not fluorescent and therefore are invisible for practical purposes and as a concluding point, I would say that we are now left with pretty, at least in my view, a pretty interesting prospect, because we I think you know pointing in different directions. There’s the fact that the very formation of these large orbits that translate into these large capture cross sections points to the existence of redberg states in the NV and whether those redwork states can be generated and controlled is to mean something that seems quite intriguing and that direction that we will certainly explore in the future and there’s, of course, all the all that’s related to the spin degrees of freedom either for the injected carriers or, alternatively, this the spin state of the source, sorry of the target NV that can be also manipulated and where we think ultimately, we’re also going to see an impact in the capture probability allowing us perhaps to identify the very injection of spin fertilization or spin polarized carriers that motivated somehow this presentation so with that I’m just going to acknowledge my group – and this is a rather old photo. But I have people who have already left and I wanted to make sure that they are properly acknowledged, harry among them and aisha as well as damon, but, most importantly, arthur who’s been leading all this effort, and I of course have to also think my collaborators at Sandia, marcus, doherty and I’m forgetting and I’m embarrassed by that, but I’m forgetting about johannes flick, who has been helping us being central to all the DFT calculations that we did on this topic. And with that I want to just thank the funding agencies and you all for your patience again. (Olga) Thank you on behalf of everybody, our applauses. It was brilliant presentation and amazing work. Thank you very much, so I will start with questions so jim Butler asked: very elegant work. Can you comment on the role of other impurities on your observations, namely Ns and boron? (Carlos) Yes, so that’s actually a good point. I have to confess that we haven’t looked at the boron content. All I can say is that the sample that we used is of the highest quality. It’s electronic grade, and we’ve been told by the manufacturer that background concentration of nitrogen is something in the order of 10 to the 14 per centimeter cube or less. As a matter of fact, the notion that I was trying to lay out when we saw that the formation of NV centers around the spot of silicon implantation, if we, where now to go and directly count how many NV ’s formed and assume that all these NV’s formed because there was already a nitrogen there. We obtain a concentration that is very much in agreement with what the manufacturer told us. So that seems to be consistent and typically – and here I may be wrong because I’m not you know that expert in diamond growth, but my understanding is that nitrogen is by far the more common impurity. So if nitrogen is at that level, then I would anticipate that boron is actually way below that, unless otherwise incorporated, and even though that’s typical, especially if you want to create SiV zeros, for example. Well, this is certainly not the case. We’ve kept the sample pristine, except for the implantation of nitrogen. (Olga) Thank you and the next question is from stefan diggs: very nice talk. I have a question concerning the scc protocol, which kind of laser. Do you use to ionize the negative NV center I’ve seen people showing scc protocols with a resonant shelving pulse at 637 nanometer, but others use protocols with 589 594 nanometer laser? Is there an advantage to use one over the other yeah? (Carlos) So here I have to say that, yes, all these, that is being said is correct, and probably you know this person knows that the conditions for a spin to charge for an optimal screen to charge protocol are still very much. Optimization process there’s the process is known, but not perfectly understood, and so, for example, this combination of green and red is something that was introduced rather empirically with the idea that, for example, this weak green beam would help the excitation of the NV to the first excited State and then the red would actually create the ionization. That was the rationale for it, and it turns out that it works reasonably okay and we get better contrast doing it this way, rather than going. You know in some other possible way, but that is the only reason. It’s just that experimentally. We find that these conditions seem to be optimal, and this is what allowed us to see these contrasts that otherwise is pretty difficult to pick up. (Olga) Thank you. I also have a question: how robust is this charge, cycling process? So can you repeat it over time and it will perform similar or it can degrade over time due to some reason? (Carlos) we haven’t seen, we haven’t seen any change, but here I have to emphasize something that perhaps I didn’t highlight enough, which is that when we, let me go back to this slide. When we created these islands, we made sure that these islands were sufficiently deep inside the dime. So you see here in the cartoon it’s about 10 microns, 10 to 12. Roughly is what we find when we go and look with the confocal microscope and the reason we wanted to do it. This way was that we didn’t want to deal with processes that alter the charge dynamics close to the surface and in prior work. We’ve seen that changes in the photodynamics of NV centers can be quite dramatic. Then you can even see one photon processes and linear growth with power. You can see all sort of things, and so this would only complicate the overall picture, so so by going sufficiently deep – and this was possible at sandia because they have these high energy implanters that we took advantage of. Then we put those concerns to rest, at least for now. Of course, if now we were to go say closer to the surface and there’s a second round of samples that are being engineered with these conditions, because we want to couple to structures that we create on the on the surface and move these little bits higher. In terms of sophistication, but it could well be that this all fails, because the surface termination of diamond is not the right one and we’ve already seen not in these samples that I was talking about. But in some other samples that were given to us for tests. We’ve seen that you can actually create non-reversible changes in the charge state of the end when you are very close to the surface, so that’s the only that’s the only caveat what causes that and whether that can be avoided. Personally, I think that is the case, but exactly what causes it and how to circumvent the problem. Yes, it’s unclear yet. (Jim Butler) I think you’ve developed some really powerful tools here, but I would be interested in seeing them eventually get applied to diamonds that are not as pure as you’re working on now, such as natural diamonds or lab-grown diamonds. Where we know we have phosphorescent processes that are involving charge transfer may be very interesting to apply your techniques to those kinds of stones. (Carlos) Yes, I fully agree, and that is somehow a line that branched out from this work and so now, as we speak, actually we’re completing measurements once again in we’re dealing with type 1b samples. For now, because there’s the caveat that if the impurity concentration is too high, then these metastable charge states that last pretty much forever in the dark in this type, 1b samples may not be so long-lived in samples that have a very high impurity content. So that’s something to keep in mind. We’ve seen that happening with samples where the NV concentration is really high on the order of say tens of ppms and nitrogen is on the order of hundreds of PPMs. So we know that we can ionize, but we can actually see on the say, I’d say a few milliseconds tenths of milliseconds I’d say time scale that the chart states that we create sort of bleach out. So there’s that point to keep in mind. But if the concentration of impurities is sufficiently low and these type of images become possible, then there’s the number of opportunities and one that’s actually we were discussing yesterday in the laboratory was the fact that we can now tell with via the help of electrodes, to buy Us the the carriers in in one direction. We could actually tell what type of charge state the silicon vacancy is taking when exposed to two carriers. So initially we were thinking. It would be that we know for sure these are non-fluorescent states. We were very much hopeful. They would be neutral silicon vacancies, we’ve now seen via these electrodes that that’s actually not the case. We have to believe it. We believe it’s the doubly negatively charged silicon, but that’s just you know an application that comes out in ensemble, well I’ll jump in with. I guess one quick question kind of a higher level, because I think these characterizations that you’re doing are, I mean they’re, I think, as jim said very elegant. But you open saying that you know this was in the vein of a almost a communication bus, and that implies that you’re, actually thinking about an application or some sort of information transfer and I’m wondering if you foresee something like that in the future. And what, if you have any applications in mind for actually communicating between the different centers yeah, so that I would say, is the silver line of these of this work. And I don’t want to make statements that later on I’m going to have to withdraw. When we see that is not possible, at least on paper – and this was mostly the work of marcus – that I highlighted early on, things are possible. There’s one important parameter that here I haven’t actually exploited and that we are now trying to make part of these observations, which is, of course the temperature. I mean all these that I described took place at room temperature. We believe we have reasons to believe based in part on these monte, carlo simulations that we’ve been running, that things are going to improve for the better as we lower the temperatures, without necessarily going to ultra low temperatures. We think that something in the order of liquid nitrogen would already make quite a positive impact. This is in terms of propagation distances and scattering processes, so we think that there’s a bunch of things that are going to happen as we lower the temperature first off. I have to say that the speed state of the NV , even under near ambient or near ambient conditions, are predicted to be ten times better than silicon already. So not nobody, as far as I know, has been able to prove that and so there’s experiments. I think that we can imagine where we first try to see if we can deterministically inject spin polarized carriers into diet. That was, I think, the initial idea, and this is the simpler one that in itself, I think – and this is of course not quantum in the sense that it’s still classical spintronics, but I think it would be interesting to see that happening. Given these near ideal features that diamond has connected to that, there’s of course, the fundamental problem, the fundamental knowledge that comes from say, looking at capture cross sections that can be made spin dependent, for example, by controlling the spin state of the target NV . If all that is put in place and that’s a big, if then, you can start imagining processes where now you can encode information in these flying qubits that take the form of either holes or electrons and imagine transporting these information coherent. That would be a dream for now, so I have to be clear on that. So that’s pretty great forward-thinking perspective. (Olga) Thank you very much for your excellent presentation and we wish you good luck in your research and all your dreams that you outline. Thank you, okay, bye everybody. Thank you for your attendance.
