�ݺ�ߣ

Editor's Notes

• #2: Hello everybody. It’s a pleasure for me to be here today and be able to speak about my work. First of all, let me thank Force For Future for the grant to attend to this congress. And now, let me introduce my topic. Magnetic pinning geometry for sensing with magnetic domain walls. This is an experiment inside the project metmags and has been partially done at the National Physical laboratory in United kingdom. It’s a project for developing metrology magnetic sensors into industry.
• #3: I’m going to speak about the motivations of this project, what is the state of the art and what are the objectives we want to achieve. I’m going to describe the fabrication process of the nanostructure we have been working with, and the identification of magnetic states inside of it. Then I’m going to switch into electrical transport measurements which include state space measurements for characterizing the device and it’s application to nano bead detection.
• #4: Conclusion
• #5: In this project ,we are focusing in magnetic domain wall technology. In the last few years, it has been a huge development in this area, and new devices have arise quite fast. In this slide we can see some of them dedicated to electronics. The main idea is substitute standard electronics with magnetic ones that operate on domain walls. In the table of the right, we can see an immediate improvement, dw electronics are simpler and in some cases easier to fabricate than the actual technology. That represents an improvement in space. But there is also an improvement in power. The picture below is an scheme of a dw based memory. It’s principal advantage is that power is needed only for moving the dw, not for maintain the information, so the heat generation is very low compared to electrical memories.
• #6: But not only electronics. Here we can see two applications of dw to biology. The big picture on the right represent a nanowire that has been curled to trap dw on it and allow them to move only one curl at a time. In the picture we can see how a nanobead is attracted by one of the dw and how it can be moved along the wire just pushing the dw. The other picture represents an early stage of a nanoparticle detector. The main idea here is that the interaction of the nanoparticle with a domain wall modifies the behavior of the dw, and is possible to detect that change in the behavior and use it to detect the presence of nanoparticles. Ok. So here is where our work starts. We want to focus on nanoparticle detection using dw. What is the best technique? how to make it standard? and what kind of parameters we need to use to characterize it? Imagine that two companies start selling nanoparticle detectors based on dw. Which one is better? What kind of nanoparticles can be detected? How accurately is possible to identify different kinds of nanoparticles?
• #7: Because NPL works on standards and measurements, all these questions are very close related to our task. So having that objetives in mind, we selected a Py nanostructre which have been reported as a good device to detect nanoparticles. The device is basically a L shape with electrical contacts. Applying a magnetic field we can place or remove a Dw at the corner. --Click-- If we apply a current through the device and measure the voltage drop, we see 2 different voltages, one when we have a Dw at the corner and a different one when there is no dw. The presence of a nanoparticle modifies the field needed to remove the dw and can be seen in the voltage. Our task is to determine the best conditions for that detection and characterize all the parameters involved. So in the future this will become a feedback with industry and new standardization and measurement techniques will arise.
• #9: Here we can see a scaning electron image of the device selected for this stage of the project. Is a L shape made of Py with disks at the end. The effect of the disks is to reduce the stray field at the end of the nanowires and to reduce also the coercive field of the arms. It has 4 gold contacts on top to perform electrical transport measurements. To pattern the nanostructure form a thin Py film it was used Standard e-beam lithography in combination with Ar-ion etching. The gold contacts were prepared by thermal evaporation, and the surface was cleared by low energy Ar-ions before being deposited.
• #11: Before performing any electrical transport measurement, it is needed to understand what are the stable states of the device and how many of them have a DW. To do that, we use a magnetic Force Microscope in combination with an electromagnetic coil. I will explain a little bit how to do it. Suppose we have a device like the big one. We don’t know how the magnetization of the disks is, but we expect the magnetization of the arms to be more or less uniform and along the arms. So, imagine we have a situation like this. To create a DW at the corner, it is enough to increase the magnetic field. --click— And then remove it. --click— Ideally, the state of the device will be now with a DW at the corner. --click— --click— --click— Choosing other orientation we can have the other 4 configurations of the arms. And in that way extract information about 4 possible stable states.
• #12: We did it and these are the results we get. The arrows indicate in which direction the field was applied. So these are four possible stable states at zero field. And the only ones we manage to found. A, and c with a DW and b and d without. From the image we can see that the biggest field gradient occurs at the corner when we have a DW. The disks behave similar with and without DW. Apart from understanding of the device, this has also proven to be a good technique to identify defects in the devices.
• #14: Now we move into the electrical transport measurements. Basically, we want to detect if there is a DW at the corner or not using a electrical current. To do that, we use the anisotropic magnetoresistance effect. The anisotropic magneto resistance effect is an interaction between magnetization and current. A simple manifestation of this interaction is a change in the resistance of a material in function of the angle between current and magnetization. This effect is particular important when we have a Domain Wall perpendicular to the current, because the creation of that Domain Wall can be seen as a jump in the resistance.
• #15: Using this effect we want to detect when there is a DW on the nanostructures, and this is the setup we used. A Lock-in Amplifier is used to generate an Ac signal. --click— Then we use a 300 kohm resistor to fix the current trough the device into a low value. --click— When the signal passes trough the device, the resistance will depend on the applied magnetic field and the magnetization state of the corner, so we measure the voltage across the corner in order to get that information. --click— On the next step, the signal from the device is amplified. --click— Finally, we compare the measured signal against an scaled version of the original signal. --click— If the signals are adjusted to match each other at zero magnetic field, --Click-- then any change in the resistance will be measured as a difference between these two signals. --click--
• #16: Ok. So I’m going to show you one kind of experiment we can perform. Suppose we have the device and we apply a high magnetic field to the left. This is going to be a state with a DW at the corner and we are going to measure the resistance. So we start the measure lowering the field to zero. --click1— We can see a change in the resistance because lowering the field we allow more spins to be allying with their natural direction, which is along the arm and parallel to the current. --click2— Now we invert the field And start rising it into the positive direction. --click3— When we reach the propagation field, the arm that is more aligned with the field reverses it’s magnetization and the DW propagates from the corner. We can detect it by a rising step in the resistance of the corner. --click4— If we keep increasing the field, eventually we reach the coerceive field for the other arm and it also switches, creating a Dw at the corner and lowering the resistance. --click5— If we keep increasing the field nothing more happens to the DW and the resistance just keeps decreasing. --click6— We can now repeat the cycle in the other direction and face the same events. --click—click—click— This is what we call a hysteresis loop.
• #17: We have seen in the previous example a hysteresis loop. We have measure how the propagation and nucleation fields depends on the orientation, and how the loop changes after crossing 90 degrees. Also we have study subhysteresis loops. When you only switch one of the arms all of the time. All of these experiments gave useful information, and allow to predict the behavior of the device in certain situations, but not in a general scenario where different magnetic fields are applied in different directions. That is why we focus on having a state space map of the device. That will allow us to predict it’s behaviour in any situation. To create the map, first we need to identify a few stable states to start the map with. We already did it when using the Magnetic force microscopy, we found at least 4 of them. Ok, so now what we are going to do is place the device in one of these stable states and apply a field in a certain direction tracking the evolution and see into what state it evolves.
• #18: So. Suppose that we start creating a state like the one on the corner. We call it state [1,1] We apply a field and place a DW at the corner. Now we are going to lower the field, chose an orientation for the magnetic field and increase it. Magnetic field can be increased into the positive direction or into the negative direction. --Click— This is what we found. Arrows in blue indicate in what direction the boundaries in this map can be crosses. That means that this map can be used only to predict the evolution when the parameters field and angle cross these lines in that direction. For more information we need more plots. And that is what we do with the next state [-1,-1] --click— Now we have even more information. Note the symmetry in the plots. And the same can be made with the other two states. --click— --click— If now we write down what states correspond to each area in the color plots. --click— We can see that each area correspond to one of the 4 stable states, and the complete evolution of the device can be predicted using this map. I have to notice, that because measurements take quite a long time (typically 2 days to do the whole map), this can be used only to predict the evolution under slow dynamics. If the field variations are considerably faster, then is possible that new unstable states appear in the evolution. Another interesting thing in this map is the noise. Some areas present a very smooth boundary, while in others the boundary is full of spikes. We are centered on this phenomenon right now, trying to quantify it and determine what is triggering it and why it appears in some areas or not. In some cases it looks like there is a second field for nucleation or propagation, and sometimes is just that the two arms switch together. Understanding this effects it’s crucial if we want to use this device as a sensor.
• #19: In order to test the state space map we have randomly chosen a path on it. --click— The path starts on point 1 with 80mT and an orientation of 22 degree. --click— Then we rotate the orientation of the field without lowering it’s intensity up to point 2. --click— Now we need to change graph because we cannot cross the boundaries for next change. We lower the field and make it negative up to point 3. --click— Point 4 is another rotation. --click— And point 5 gives us back again to the same state as point 1. Now. To test this prediction, we perform a real measurement on the device following this same evolution of fields and orientation. --click— We can see that there is some bias. This we think is because thermal drift. These experiments were perform at room temperature in a not isolated lab. So some thermal fluctuation is expected. -click— We repeated the experiment several times. --click— And finally all together. --click— Despite thermal drift, we can see that the agreement is quite well in all points, with some fluctuations between 2 and 3 and 4 and 5. Between 2 and 3 we can see that it is a special noisy area, maybe a double boundary or a not well defined one definitively not a good place for sensing. The problem between 4 and 5 is because boundary near point 5 is extremely angle dependent, and even a small change in the angle leads to a big change in nucleation and propagation fields. This is the kind of information we need to know in order to work with these devices. Specially, random switching, because what is going to be extremely important to detect nanoparticles is being able to measure very accurately any shift in these boundaries because of the presence of a magnetic nanobead.
• #21: And we came to the discussion. Detecting nanoparticles using Domain Walls. Our objective was to test one of the proposed devices in order to gain expertise in the new field and to evaluate it’s potential as a real commercial device. To do that, we need to quantify it’s properties and specially the ones related to measurements. But first, we need to characterize and understand the behavior of the device. So we have performed anisotropic magnetoresistance measurements and we found that they can be used to track the state of the device. We have identify 4 stable states and created a state space map that agrees to this 4 states interpretation. Most important, we have used this map to predict the evolution of the device under different conditions and we have spotted working conditions to improve the bead detection experiments. Our future task will be to measure the effect of the nanobeads in different working conditions and find out what are the best ones to use the device as a sensor. And maybe, in the future, explore different geometries and their effect.
• #22: About bead detection. This is our next step in the project. Until now we have been characterizing the devices and looking for good sensing parameters. Next point will be to actually place a nanobead on the corner and measure it’s influence onto the device. Right now, the nanobeads have been placed, but we didn’t performed any measurement yet. So I can only say a few words about the placement. It has been performed using a Focused Ion Beam with micromanipulators. Beads were commercial Dynal beads, this one in the picture 1micron size. To pick them we just spread the beads into a substrate and allow dissolution to evaporate. Then just look for an isolated bead and pick it with the micromanipulator. To place it into position we use Platinum deposition. We measure the devices before placing the bead, we plan to measure them with the bead and see the difference, and we plan to remove the beads using a scanning probe microscopy. And measure again to see if the behavior is the same as in the beginning.
• #23: Thank you for your attention and thanks to all the people involved in the project.