 So, we have come to the second lecture of module 5. In the first lecture, I discussed with you along the same lines of how a material can drop in resistance, especially in a rareth oxide such as lanthanum manganite. How this particular unit cell displays a unique property of both becoming ferromagnetic as well as becoming metallic from a non-magnetic and a insulating state. I call this as a genie inside the lattice because two things are operative in a same single lattice. Now, today I want to talk along the same lines of losing resistance in a different class of compounds called metallic multilayers metallic multilayers. And the way the material loses the electrical resistivity brings a unique nature of giant magneto resistance which can be used for a variety of applications. To draw your attention more this metallic multilayers are presently used in our computer read heads and it has brought about a tremendous revolution in the magnetic storage system. So, quickly let me go through and tell you about the last lectures brief. Here we told that the CMR oxide or colossal magneto resistive manganese shows loss in resistance at the curie temperature if you apply a very high field and this can be used for magneto resistivity or giant magneto resistivity or colossal magneto resistivity. It is mentioned in different ways and by and large colossal M R is referred to manganese whereas giant magneto resistivity is actually referred to multilayers. Now, this is the application that I drew your attention to that it can be used as a read head and you can also write information using this sort of devices. So, the compounds which show similar property colossal M R property is compounds based on L A M N O 3 when substituted with the strontium barium or copper. Now, what is important as far as the metallic multilayers are concerned this is a typical multilayer that is made and it is presently used in our computer hard disk and if you want to see the animation of this you should visit this website this is displayed in IBM.com website where they will show you how the resistance varies if varies with information and how that can be used for reading magnetic information. In this device you would see there is this is a bit that has to be read we have a free layer which is this one this is a free magnetic layer and then you have a spacer layer in between and you have a pinned layer this is also a ferromagnetic layer but this is actually pinned to the exchange layer or anti ferromagnetic layer this is also shown in this view graph where you have the copper spacer and you have the GMR free film a nickel iron which is called perma-loy and this is there on the top and you also have cobalt which is actually pinned to a anti ferromagnetic exchange film. So, when you have a ferromagnet and a ferromagnet divided by a spacer and this ferromagnetic layer is actually pinned by a anti ferromagnet then this top layer alone is free to rotate I will come to the physics of it later but what happens is in such a device system the reading capacity of this head becomes much much faster than the magneto resistive head that is used now. So, the implications are phenomenal same thing you can do you can pin with an anti ferromagnetic layer a magnetic layer therefore this moment will be fixed this magnetic moment is fixed and then you have a instead of copper in the previous case you can put an insulating layer if you can pin this magnetization in this direction then when you apply magnetic field actually the magnetic field can either rotate this in this direction or it can rotate in this direction. So, this is free to rotate depending on the field of magnetization and in such case when the moments of this ferromagnetic layer and this is aligned then the electron that is flowing across this layer will easily tunnel whereas in this case because these two ferromagnetic layers are in opposite direction then the tunneling of this electron becomes difficult as a result you have a high resistance case and a low resistance case we will come to this later this is called TMR devices tunneling magneto resistance in the previous case we talked about giant magneto resistance both have tremendous application and this is this TMR device is actually used for magnetic random access memory devices which is a major breakthrough in today's technology because an enlargement of the active layers in M-RAM device is actually shown here and this is exactly the way a M-RAM device will look like where you have a anti ferromagnet and this is actually pinning this is pinning a ferromagnetic layer whose electrons have a spin in this direction and then you have a coupling layer like this and then you have the other ferromagnetic layer here so you have the ferromagnetic layer and the anti ferromagnetic layer which is actually pinning then you have a tunneling insulator and then another ferromagnetic layer on the top. So this is the way a M-RAM device is actually used in IBM the red and green spears represents electrons spinning in opposite directions in the magnetic layers the very thin insulator allows electrons to quantum mechanically tunnel across this interface and information is actually stored in the top layer by forcing its electron to spin in one direction or other so this is the free layer where the flip can either be clockwise or anti clockwise depending upon the configuration and depending upon the magnetic field direction. So this is one of the major development in the magnetic storage where thin layers of ferromagnetic metals are actually stacked across a non-magnetic layer or a insulating layer I will come to the details of it later. Now according to Moore's law number of transistors increases per integrated circuit as a function of the number of years. So if you see the transistors that can be accommodated in the integrated circuit has almost linearly increased and this is called Moore's law so you can actually have hundreds and thousands of transistors stacked in a integrated circuit now and this is bound to keep going in the future years. The same analogy can be extrapolated for magnetic recording which is also equivalent to Moore's law the aerial density actually is keep increasing with the years it is almost you see a linear dependency therefore magnetic recording more and more information can be stored in the hard disk because you have a powerful read head memory device now and the current memory device that is used by IBM is called Spinvalve MR head. If you have opportunity you can visit the IBM website to understand how this multilayers are used to improve the aerial density so this is bound to bring lot of revolution into the magnetic recording market. Now what is fundamental to this application is in the in the case of electronics so far people have worried more about the charge and number of charge are very important for a semi conducting industry whereas if you look at another missing link in this whole application is the spin part of the electron. Electron has spin either plus half or minus half and this spin actually can control the charge of the electron so far the spin part is actually forgotten. Magnetic industry has exploited the issue of charge number of charge carriers but never it has bothered about controlling the charges or the mobility of electrons with respect to spin part therefore if you apply a magnetic field then either it will be up spin or down spin and depending upon the population of electrons that you are going to force then you can modify the electronic part so that is what is called as spin electronics or spin electronics it is also called as magneto electronics because you are trying to control the electronic property using magnetic field and this was actually proposed the issue of spin part was proposed as early as 1926. Now if you take any metal in the periodic table you can easily classify whether it is a non magnetic metal or whether it is a magnetic metal or it is a ferromagnetic metal. Now in the case of copper chromium ruthenium we know that it is a metal but not a ferromagnetic metal why because you have the number of up spin electrons and the down spin electrons they are exactly having the same sub bands spin bands across the Fermi level so because they are equal then this can be actually called a non magnetic metal because the number of up spin electron cancels the number of down spin electrons therefore there is no net ferromagnetic moment there is no net magnetic moment whereas in the case of iron cobalt and nickel as we saw in the first module the whole thing can be explained based on molecular orbital theory where you can see the spin up band has the spin band like this across the Fermi level and the spin down band has across the Fermi level band which is lower than the spin up band as a result there is a net moment which makes this material ferromagnetic. Now if you look at these compounds chromium dioxide which is actually used in videotape and this is the colossal magneto resistive CMR oxides one of the interesting feature is one of the spin up band here is actually well above the Fermi level and what matters finally is the spin down band which is cutting across the Fermi level therefore the conduction can actually come from one of the spin bands which is completely devoid of the Fermi level in such case this is called as half metallic ferromagnets this is called as half metallic ferromagnets this sort of magnets have 100 percent polarization compared to the traditional ferromagnets these ferromagnets have 100 percent polarization because in one of the configuration the spin up band is actually 100 percent spin polarization. So this can be used for the GMR or CMR applications so this is very important as we think about what sort of material you want to use for tunneling magneto resistance because if you want to tunnel the electron then if it is actually 100 percent spin polarized then it can easily tunnel because there would not be any scattering process across the interface. Now I will have to register this issue that the whole idea of CMR or GMR or TMR all this came to prominence because of the discovery of magneto resistance by these two gentlemen this is Albert Fert from France and this is Peter Groenberg from Foscheng centre Mulesh both of them found out that there is a strange coupling mechanism happening if you can maintain this metallic multilayers in a very very thin dimension and they found there is a huge response when you measure such such stackings when you measure the electrical conductivity and they found there is a huge loss in resistance in the presence and absence of field and that is what brings to effect the GMR spin walls. Now what really they did this is Peter Groenberg who is receiving his noble prize in 2007 which marks the birth of spintronics also I should say because they emphasized on thin magnetic or nanomagnetic layers as small as 1 nanometer the issue of nanotechnology became more prominent after the discovery. Now what did Peter Groenberg report in his report he said if you take only iron of this dimension say 250 nanometer and if you measure the resistance as a function of magnetic field in both directions you see a very small change in the resistance which is called anisotropic magneto resistance which is of a very low order whereas if you now separate the same amount thickness of iron but you divide it as 120 nanometer 120 nanometer of iron and you put one single small layer of chromium which is only 1 nanometer in this form then you can clearly see this same feature which is supposed to be there is now a more pronounced feature like this. So what happens when there is zero field here when there is zero field these two ferromagnets are actually anti-parallely they are coupled they are anti-ferromagnetically coupled but they are not actually anti-ferromagnets the way they are coupled is of the opposite form therefore we can call this as anti-ferromagnetic coupling. Now as you sweep the magnetic field in both directions you see that these two moments get ferromagnetically ordered in this fashion and at that point the resistance is really low and the same way it happens if you flip it to the other direction therefore there is a tremendous fall in resistance as you sweep the magnetic field. Suppose you can keep reverting this at very low field then it becomes a real magnetic switch. So Albert Fert actually brought out this notion he said I can try to make several of these bilayers several of this bilayers instead of just a trial layer and everywhere I will try to change the spacer layer thickness. So that is what he did if you have chromium as 1.8 nanometer or 1.2 nanometer or 0.9 nanometer as you bring down the spacer layer smaller and smaller you can see there is a tremendous fall in resistance and also the field sensitivity is quite bright compared to even this example. So if you make such bilayer repeats of 30 angstrom iron and 9 angstrom chromium and if you make repeats like 60 times or 30 times or 30 times like this then you see a tremendous fall in resistance and that makes the application much more prospective where you can now look for a 0 1 switch so that you can write and read the magnetic information as a 0 1 bit. So this is the birth for spin electronic applications. Now what really happens is this cartoon tells you that if you have this ferromagnetic layer aligned in this direction and this ferromagnetic layer aligned in opposite direction they are anti ferromagnetically coupled whereas in this case they are ferromagnetically coupled. Now in both case you will see when they are anti ferromagnetically coupled then they have a different resistance that is what we show here but once you force with the magnetic field in this direction all this anti ferromagnetically coupled layers they also go ferromagnetic. So as a result you have a low resistance state and a high resistance state at h is equal to 0 which is important for magneto resistive property but what is needed is this saturation cannot take so much time it has to saturate very sharply for applications then only you can use such materials for 0 1 bit reading or writing otherwise if it takes too much of a field saturation then that cannot really act as a very good device. So to transform this people have made several structures and they have made something called spin valve where the top ferromagnetic layer is more like a valve and the bottom ferromagnetic layer is actually pinned to a anti ferromagnetically. So even with small magnetic field like a valve you can rotate the moment of the top magnetic layer so that is the technological challenge. Let me run through some of the issues just to register that in your memory this new field of electronics which is not based on conduction of electrons or holes but relies on the different transport properties of majority and minority spin electrons actually forms the basis for spin electronics. Add to electronics an additional degree of freedom that is the spin character so you are actually using the spin of the electron for governing the electronic properties. So in magneto electronics you actually have passive elements which are resistors change in resistance happens upon application of magnetic field whereas in spin electronics you actually have active elements which are spin transistors this amplify a current rather than merely switching it on or off. So what are such characters you have a metal ferromagnet semiconductor non ferromagnet and then you also have a ferromagnet. So this sort of structures can actually bring about the spin tronic applications. Another key factors that I want to emphasis is spin electronics in semiconductors are also possible an obstacle for spin tronics is that electronic companies are geared up for semiconductors they are traditionally they know how to handle a semiconductor and to run the industry without any interruption. An important goal is to make devices using semiconductors that are compatible with existing spin technology chip technology the problem is that conventional semiconductors use in integrated circuits are not magnetic this is why several research groups are exploring ways to tune the semiconductors into ferromagnetic metals which we call it as dilute magnetic semiconductors. So another field apart from tunneling magneto resistance comes into picture which is called DMS field where they want to retain the semiconductor technology but just do a careful manipulation so that you can make the semiconductor magnetic so that you can tune the electronic properties now controlling the spin part of the semiconductor. The big problem here is spin polarized transport across the interfaces between different materials because interfaces are very sensitive between semiconductors and ferromagnetic metals presently they induce a Scott key barrier that is leading to loss of spin polarization. So one has to overcome this Scott key barrier which is the challenge if you want to realize a DMS situation here ferromagnetic semiconductors injecting spin across the interface between two semiconductors one of them ferromagnetic should be easier because there is no Scott key barrier ZNSE doped with beryllium manganese or cobalt doped with TAO2 manganese doped with SNO they are all candidates for ferromagnetic semiconductors. What is the aim fully switchable all semiconductor spin walls are possible semiconductor spin transistors are possible so there are tremendous scope that is lying if we can generate newer materials. A bit of history in the anisotropic magneto resistance was reported as early as 1857 and in 1947 there was discovery of transit action in germanium 1952 germanium transistors were discovered then 1950 and 60s this random axis memories in computers were brought in in USA then 1975 first time a tunneling magneto resistance response was reported by juliary and 1979 IBM introduced thin film heads which is called MR read heads which started coming and affecting the memory storage and that is where we saw the computers coming into every home and a bit of very recent history 1988 1990 GMR was reported by both this Nobel laureates and 1991 IBM introduced the the MR effect to for read out in hard disk drive 1991 the spin wall wave effect was also recognized 1994 first commercial product using GMR a magnetic field sensor was brought in and a bit of history till our days 1995 tunnel magneto resistance was rediscovered and in 1997 this is the most important stuff GMR using GMR property IBM brought its first hard disk drive and in 2004 we have free scale semiconductor they currently sampling a 4 megabit m ram chip for backup memory in industrial and military environments in near future you have m ram production expected in 2005 which is already set into action now by 2010 we have a generation of m ram devices which are coming into market this was actually a forecast several years back but this is actually turning out to be a reality in 2010 now GMR in spin walls is typically of this nature I have already shown you a cartoon impressing upon the importance of it so you actually have anti ferromagnetic pinning layer which is spinning a ferromagnetic layer therefore that moment is fixed and then you have a spacer layer and a free layer this free layer can rotate either this way or it can rotate this way freely changing the resist overall resistance of the device so typically if you want to look at the device the device will have a hysteresis m versus h hysteresis m versus h hysteresis like this where in one direction you see the loop and the other direction is actually pinned if you measure the resistance as a function of field you can see when this is this ferromagnet is actually pinned it is like this sorry it is coming this way and then this free layer it can rotate very sharply in other words you can achieve saturation in even with the 10 or 20 orsted because this is a free layer and this can easily rotate to align in the direction in which it wants so you can get this field sensitivity in spin wall device which is of fundamental importance for technological applications and in this cartoon what we see here is gmr up to 5 percent at just 10 orsted which is less than the field that is generated by the magnets which we try to put on the refrigerator so even with such a low magnetic field you can actually make this switch operate therefore it can be used for magnetic sensor application gmr can also be found in granular alloys granular alloys means those which are like cobalt copper if you co sputter or co deposit cobalt and copper you would see that cobalt and copper are not miscible so in a copper matrix cobalt will actually form a cluster like stuff instead of forming a continuous layer cobalt will form clusters and depending on the cluster size of this cobalt in copper matrix you can see resistance varying and that is what you see here this is cobalt in copper and you can see as a function of temperature the magneto resistance changes and also one can change the gmr property based on the cluster size cluster size of cobalt metal so this is another way gmr property can be exemplified now let me tell you briefly what really makes this useful for magnetic readed applications this is the origin of gmr and the principle that acts in this gmr is called spin dependence scattering spin is getting scattered at the interface between a ferromagnet and a metal and a metal and a ferromagnet so we can look at this situation in the following way there are two models given in this two models you see in this model this is a ferromagnet and this is a ferromagnet where the moments are aligned in same direction and in this it is a ferromagnet this is a ferromagnet and it is aligned in opposite fashion and this is the non-magnetic layer which is a metal now you have both the situations of a spin up electron spin down electron spin up electron when it is going from this layer to this layer as you are measuring the current you see because these two are up spin in this direction then there is no scattering it will just go across this interface whereas the spin down will get scattered little bit here and then it will go through again it will get scattered here so if you try to translate this to a resistivity model for the spin up there is low resistance pathway which is like this but for a spin down the resistance pathway actually comes out like this which is bigger because in both cases it is getting scattered therefore resistance in this form is a high resistivity issue so in one case you have a low resistance in other case it is a higher resistance whereas when you come here for spin up it gets easily to this stage and then it gets scattered here therefore you have a lower resistance and a higher resistance same thing happens for the spin down also here it gets more scattered and then it gets easily transmitted therefore in if you draw this resistivity model then you see in both cases there is higher resistance in either way therefore in overall if you see this resistance is going to be very high compared to this this is this resistance when the ferromagnets are aligned antiparallel in this case ferromagnets are aligned parallel so in what happens is a shock circuit therefore resistance is lower when these two are ferromagnetically aligned and when they are ferromagnetically non aligned they are antiferromagnetically aligned then you have a higher resistance and that is what is called GMR or giant magneto resistance in the presence of field and in the absence of field you see difference so when magnetic informations are actually to be read you have a low resistance case or a high resistance case which can actually flip as a 0 1 0 1 bit and this property is what is important for reading magnetic storage one thing that we need to understand if you want such GMR property to be there and if this is purely a spin dependence scattering then I will come to the previous one once more so if this has to go through without scattering this interface has to be very very sharp there should be no roughness if there are physical defects also scattering will occur therefore you need to know how to make thin layers like this so spin dependence scattering is one of the important issue and the next issue what we see here is as you increase the layer thickness of the spacer in this example it is gold which is used as a spacer layer and permaloid which is nothing but nickel iron I will write it here nickel iron is the permaloid and in this case you can see as you vary the thickness of your gold spacer layer the magnet of resistance varies like this like oscillation what does it mean at some thickness it is showing GMR property at some thickness it does not show GMR property so you can see several maximas coming anti ferromagnetic situation one anti ferromagnetic situation one anti ferromagnetic three anti ferromagnetic four so if you keep on increasing the layer thickness of gold you should actually see more and more of GMR or lesser and lesser of GMR happening but what happens suddenly you see maximum GMR and then there is no GMR property and then there is GMR property and then it comes down so it keeps on varying as a oscillatory fraction and that is because of the physics involved in it which can be interpreted based on RKKY type of coupling RKKY type of coupling therefore this is very important so you need to know what is the thickness of the non-magnetic layer that you are depositing so this thickness of the non-magnetic layer is oscillatory it can it can show maximum GMR at 2 nanometer but at 4 it may not show at all and at 8 nanometer it might show again so that is purely because of the exchange coupling which can be explained based on RKKY type of stuff I will come to this issue of origin of tunneling magnet of resistance in a few minutes from now so far I told you about the origin of giant magnet of resistance in metallic multi layers and the issue that I have emphasis there is that of spin dependence scattering therefore when you make such very thin films the interface has to be extremely flat otherwise the electrons can get scattered not just by the moment but by the interfacial roughness therefore maintaining such flat layers is very important for which people use molecular beam epitaxy as a very convenient tool to make such flat terraces and if you are if you are wanting to know whether your material is flat enough then you use scanning tunneling microscopy to study whether it is atomically flat. The other origin of TMR device which we call it as tunneling magnet of resistance this is not only based on interfacial scattering but it is based on spin dependent tunneling where the spin subbands of the ferromagnetic layers and this ferromagnetic layer is very important the spin subband of this ferromagnet and the spin subband of this ferromagnet is important so let me take it out yeah so this spin subband represents that of the ferromagnetic layer here and this spin subband determines that of this one so when two ferromagnetic layers are aligned then the spin subbands are also same so what would happen the up spin electrons can easily hop to this one and same is true for the down spin electrons they will happily go across the insulating layer in this case this is a small barrier it will tunnel through this small barrier to the other ferromagnetic layer where and this is the situation when you apply a magnetic field suppose you do not apply a magnetic wheel h is equal to 0 then you would see that this up spin electron is going reluctantly to this spin subband because they are anti ferromagnetically coupled as a result the position of this up spin band here is different from the position of the up spin band here so energetically they are not favorable therefore it is going reluctantly same is true for the down spin band in this case it is positioned here whereas in this case it is positioned here therefore energetically it is not favorable in both cases you see a reluctance in the transfer of the spins up or down spin as a result in the anti parallel configuration resistance is greater than the resistance in the parallel configuration and this is not based on the spin inter facial scattering this is based on spin dependent tunneling and one more thing that is important to note in TMR devices this should not be a metal this should be a insulator and this insulator should be thin enough so that this quantum mechanical tunneling can be effective okay so this is called as tunneling magnetor resistance TMR and this property what is happening here is a spin dependent tunneling if we have different spin bands then the first question that I would like to know is what is the spin polarization of this ferromagnets what I am using and the definition for spin polarization is spin density of the up spin states minus density of the down spin electrons divided by density of the up spin plus density of the down spin where D plus and D minus represents the density of states near the Fermi level okay now to measure this spin polarization there are two methods one is tunneling technique and the other one is and andri reflection method in both cases we can try to measure the spin polarization of the ferromagnets what you are using and for example this is perma-loy nickel iron perma-loy in case of tunneling experiment you see spin polarization up to say 40 percent whereas in the andri mode you see it is around 35 same is true for cobalt and you can measure for nickel iron and you can measure for nickel manganese antimony this is called Huesler alloy Huesler alloy or lanthanine strontium manganates that also shows up to 75 percent of spin polarization but the best one to show is CRO2 which is a ferromagnetic metal which seems to show spin polarization up to 90 percent or so because it is a half metallic ferromagnet okay. So these are good candidates for using this as a electrode for tunneling magneto resistance devices so one of the important criteria is that one of the spins should be a majority spin and if it is 100 percent spin polarized then tunneling magneto resistance can be more pronounced for such applications. There are several other models or several other trial layers which have been tried there are experiments on tunneling for a very long time as early as 1975 jewelry he found that in iron germanium cobalt you can get up to 14 percent at 4.2 K but this is interesting because he was the first one to report the tunneling magneto resistance nevertheless the numbers are not very attractive because you have to get this sort of huge values at room temperature. Now there was another report where they have used nickel, nickel oxide is a antiferromagnetic compound which is a insulator and nickel cobalt layers show magneto resistance like this and TMR is actually tunneling magneto resistance which is explained in terms of spin polarization as 2 P1 P2 by 1 minus P1 P2 that is the order of TMR and this P1 P2 is the polarization of the first electrode and the second electrode. So the TMR values entirely depends on the spin polarization more than the thickness of the metallic layers because that thickness becomes more prominent for giant magneto resistive devices for TMR devices it is the polarization which is more important. You can also get a very nice TMR response if you actually have a perma-loy alumina cobalt junction where alumina is used as a tunneling barrier because alumina is a very good insulator. So it is possible to get TMR devices like that and this is again another example where you can clearly see that there is a very nice saturation for a spin valve junction operating on a TMR property. So you can actually get this sort of response if you can try a variety of trial layer or bilayer devices. In a typical TMR device you can see how many magnetic signatures happen if you record the magnetic hysteresis. If it is a typical TMR device then you are supposed to see this sort of a staircase like magnetic hysteresis and that staircase like magnetic hysteresis has something to say. In the device that is shown here is that of gallium arsenide based device where this is the substrate and this is a MGO layer that is deposited also to provide the substrate template and then you have the ferromagnetic layer ion in 200 angstrom that is 20 nanometers and then separated by a 2 nanometer MGO and then you have 200 angstrom fxco. In such a situation you can see if this layer and this layer are ferromagnetically coupled you see the saturation reaching very fast and as you come down at this dip staircase you can see that the lower one is anti ferromagnetically coupled and then you can come down further this the top layer gets rotated therefore you see again a fully ferromagnetic layer at say 80 angstrom. So, at 80 angstrom here and 80 angstrom here in both case it is ferromagnetically coupled but in the staircase area you can see that they are anti ferromagnetically coupled and this is a true property of a tri layer TMR device. So, in any TMR device you should see a staircase like property which is a signature that you have made the device and the same is true here actually if you measure it along this axis and if you measure it along this axis you see here a double staircase phenomena will come that is because this moment is actually rotating. This spin is actually rotating therefore you see a double staircase situation if you are going to measure it along 0 1 0 plane. Now, this TMR device can also be extended to other cases one of the problem that we faces when electron goes or tunnels from one ferromagnetic layer to another ferromagnetic layer. If the electron is scattered because of short or long time scales then the spin that the electron carries from here to here will be lost or will be minimized the effect will be minimized if because of the non-magnetic layer why because if you are using a metal ferromagnetic non-magnetic metal or if you are using a heavier metal ion then there will be spin orbit coupling that is contributing as a result there will be a spin orbit coupling happening as it goes from here to here. So, it is better for us to replace this metal with the organic layer because you have the spin relaxation times are of a very larger scale because there is no metal in organic compounds. So, the electron can take its own time to carry its spin memory from here to here without any scattering. Therefore, the recent TMR devices they are trying to replace the spacer layer from a insulating inorganic layer to a organic insulating layer. In such case you can actually extend the spin memory of the electron going from this ferromagnet to this ferromagnetic layer by extending the length scale. So, lot of work is going on to understand that I will come to this issue later in one of the slides. So, you can make several such organic molecules to measure the GMR I will come to this issue shortly from now. So, you also can make other structures like European sulphide which is a non-magnetic chalcogenide people have explored and it shows TMR device like this. Now, what is the importance of TMR as I told you earlier the technological relevance of tunneling magneto resistance is to produce magnetic random as its memory. They are fast dense non-volatile cheaper and this is projected to become a 50 billion dollar industry by 2010. We are almost realizing such a market trend as far as TMR property is concerned. Nearly every major technological company now has a hand in EMRAM not only that there are good ferromac ferroelectric devices which are coming which are also useful for integration in EMRAM. So, along with the ferromagnetic research there is also ferroelectric compounds which are essential for EMRAM which makes this very very challenging venture for most of the industries. Each magnetic tunnel junction is a memory cell that stores a single bit of data. So, to write in such a cell one need only a apply a magnetic field to flip the spin orientation of one of this layers. For the tunneling magneto resistance involving organic compounds we call this as organic spin tronics because several of these compounds can be used. This is a sexy typhoon which is used here and you also have typhoon molecules substituted with this sort of substituents which makes it more interesting or you can also use the well known AL Q 3 which can also act as a very good organic insulator in this TMR devices where organic spin tronics can be demonstrated. As I told you I am showing a enlarge view of the table which I referred earlier. I can actually take two ferromagnets and now put T 6 which is sexy thionyl or you can take two ferromagnetic electrodes and you can put organic molecule like AL Q 3 or P 3 H T you can put P 3 H T is here that is poly 3 hexyl typhoon P 3 H T or you can use P 3 O T or you can also use phenyl forphyrin. Phenyl porphyrin can be used here where ferromagnet and another ferromagnet is there MR ratio of the order of 18 percent can be achieved using porphyrin and MR ratio up to 90 percent can be achieved using P 3 O T but at lower temperature. What is important is these three device configurations where you can achieve TMR property to a greater extent at room temperature that is important because for fundamental applications you need TMR properties at room temperature where you can clearly see organic molecules are coming into picture P 3 H is used, typhoon is used. So, it is a very challenging issue as of now to make lot of device combinations for TMR junctions. I will try to discuss how other organic layers can be used in this sort of device applications may be in the next lecture and see how the magnetor resistive devices can be improved with various combination of these inter layers. I will stop here what we have seen in this lecture is that making magnetic multilayers of different sort with spacer layers brings about a enormous change in the resistance in the presence and absence of magnetic field and this can be extrapolated to many devices having spaces as oxides, spaces as non magnetic metals, spaces involving organic molecules and different magnetor resistive property can be observed and this is of importance for application. So, I stop here.