00:00:00 Our next speaker is Art Sleight of DuPont.

00:00:04 Art has been exploring metal oxide chemistry for many years

00:00:08 and has run over 100 publications in this area.

00:00:11 He has synthesized large numbers of new metal oxide compounds

00:00:15 and is especially interested in relating structure, bonding,

00:00:19 and properties in metal oxides.

00:00:21 Properties considered over the years include

00:00:24 magnetism, ionic conduction, heterogeneous catalysis,

00:00:28 and, of course, superconductivity.

00:00:30 In 1974, he discovered high-temperature superconductivity

00:00:35 in the barium-lead bismuth oxide system.

00:00:39 Many consider his work to be extremely significant

00:00:42 because it was the first report of high-temperature superconductivity

00:00:46 for the perovskite structure.

00:00:48 Recent discoveries of high-temperature superconductors

00:00:51 are based on oxides with closely related structures.

00:00:55 Tonight, Art will give us an overview of superconductivity in metal oxides.

00:01:00 Art.

00:01:01 Well, I think I am one of those people who has always waited

00:01:26 to the last minute to make up slides,

00:01:29 but I think I finally have an extremely good excuse for that habit.

00:01:35 Things in the superconducting field have been moving so fast recently

00:01:41 that any talk one makes up is out of date a couple days later.

00:01:48 But what I am going to do is to briefly review the history

00:01:55 of what has happened in superconductivity in metal oxide over the years,

00:01:59 point out some general features of these oxide superconductors,

00:02:05 and then show you how we can use the kind of thinking

00:02:12 that chemists are familiar with to understand superconductivity,

00:02:17 where it occurs in the periodic table, for example,

00:02:20 and understand some of the transitions which may be important for these materials.

00:02:28 I think you have perhaps seen enough of this kind of thing at this point.

00:02:36 I would emphasize again that after the initial discovery of superconductivity

00:02:42 first in mercury and then in lead,

00:02:45 niobium really dominated everything to the point where

00:02:50 the vast majority of people in the world working on superconductors

00:02:55 were absolutely convinced that if you did not have niobium in there,

00:02:59 you would not have a high Tc superconductor.

00:03:05 But then we went through this last period of lean years,

00:03:09 more than ten years, with nothing much happening at all in superconductivity

00:03:14 after this niobium germanide was discovered at about 23K.

00:03:22 And then, of course, last year this IBM group discovered superconductivity

00:03:30 in these copper oxides which were initially not well identified

00:03:36 in terms of composition and structure,

00:03:39 and there's basically two types now that are known.

00:03:44 The first type based on lanthanum copper oxide,

00:03:48 but doped with some alpha on earth such as strontium,

00:03:52 and then higher Tc materials which have become known as the 1,2,3 compound

00:04:00 indicated by yttrium barium copper oxide at the top.

00:04:05 Now I'm going to basically go through the history,

00:04:10 up through the lanthanum strontium copper oxide I've indicated there,

00:04:16 but not get too much into the highest Tc materials that are well established

00:04:21 at this point, the so-called 1,2,3 compounds.

00:04:26 I think the speakers after me will be talking primarily about the 1,2,3

00:04:32 because, of course, it's the most exciting of the highest Tc.

00:04:35 But it's so new that we don't know a lot about it in many ways.

00:04:42 Most of us have only known the composition of this,

00:04:45 well, all of us really have only known the composition of this

00:04:48 for a few weeks at best,

00:04:50 and the structural information on this material is still not what we would really like.

00:04:56 We have reasonable structural information,

00:04:59 but there are some structural details which are still being debated.

00:05:02 Whereas the lanthanum strontium copper oxide there is an old material

00:05:08 that we're all familiar with now.

00:05:10 There's a lot of papers, preprints out.

00:05:12 We've known about it for a couple of months now,

00:05:15 so we can begin to come up with some rational thinking

00:05:21 about what makes it a good superconductor.

00:05:26 Well, the history of oxide superconductors is indicated here.

00:05:32 We'll begin the third column for the moment.

00:05:34 As you've already heard, strontium titanate is what really started it all for oxide superconductors.

00:05:43 Strontium titanate, if it were undoped and stoichiometric, would be an insulator,

00:05:48 but if you dope it in some way or reduce it and get a few electrons

00:05:53 into the titanium deorbos, it does become conducting and superconducting.

00:05:59 It never quite hits 1K, but it's close, 0.7K.

00:06:07 But then along came the tungsten bronzes,

00:06:11 and there's a large number of tungsten bronzes

00:06:15 where that A cation can be usually an alkali cation

00:06:20 but it might also be an alkaline earth cation.

00:06:23 X in that formula is always considerably less than 1.

00:06:28 In fact, it's always less than 0.5.

00:06:32 The highest TC is about 7.

00:06:35 In fact, looking around, the highest TC I could find was 6.6K.

00:06:42 Then at DuPont, about 20 years ago,

00:06:46 we extended that to molybdenum bronzes and rhenium bronzes

00:06:51 but without very exciting critical temperatures.

00:06:55 Then we had this exciting development about 12 years ago

00:07:00 from Johnson & Matthias, this lithium titanium oxide with a TC of 13K.

00:07:10 Now that was a different structure, a spinel structure.

00:07:15 We still had the common feature of titanium in an octahedral environment,

00:07:20 but the previous materials had been perovskite or perovskite-related in some sense.

00:07:25 Then with our discovery at DuPont of the beryllium lindesmith oxide systems,

00:07:30 we were back to perovskites again with a TC again of about 13K.

00:07:36 This was not so dramatic after the recent discovery of the lithium titanium oxide at 13K.

00:07:42 What really broke things loose was the discovery of these copper oxides

00:07:47 with the very high critical temperatures.

00:07:51 Now if we go to the third column,

00:07:54 what I'm indicating there is the number of electrons per transition metal cation,

00:08:03 basically, since most of these are transition metal cation.

00:08:07 Before we came onto the copper oxides,

00:08:10 this d-shell was only slightly occupied,

00:08:14 at best one electron per transition metal,

00:08:17 and in general, much less than one electron per transition metal.

00:08:22 What we had failed to realize was that if we had gone to the other extreme,

00:08:27 where we had just about one hole in the d-band,

00:08:31 we would have even more interesting oxide superconductors.

00:08:36 There's perhaps another key we should have picked up sooner,

00:08:41 and that is that we should have been working in the first row instead of the second and third rows.

00:08:47 If we had just stayed in the first row after discovering lithium titanium

00:08:51 and gone to the other end, we would have been there years ago.

00:09:01 This is the periodic table, I'm sure this audience recognizes,

00:09:05 but what I want to point out is that it has been generally accepted

00:09:14 that if one wants high Tc superconductors,

00:09:17 one has to work around the elements which are known to be good superconductors.

00:09:23 So we have niobium here, the element with the highest transition temperature,

00:09:29 and it's been well known for years that if one wants to make compounds

00:09:34 that are high T superconductors, you've got to stay close to niobium.

00:09:39 The only departure at all has been, well, you can go over here to lead,

00:09:44 which is another area of reasonably high Tc for elements,

00:09:48 and again, you can make some compounds with high Tcs,

00:09:52 Tcs that exceed that of the element, as you can with niobium.

00:09:57 Now, I never came across anyone who had a particularly good explanation

00:10:02 for why with compounds you had to stay close to the high Tcs found for the elements,

00:10:09 but that did seem to be a perfectly reliable rule.

00:10:14 But now we have copper. Copper is not superconducting,

00:10:19 and I think no one has ever done anything to copper to make it,

00:10:22 because some of these elements will become superconducting under pressure,

00:10:25 but not copper as far as I know.

00:10:28 So we have, with the discovery of these copper oxides,

00:10:33 a violation of one of the well-accepted rules in searching for new superconductors.

00:10:40 This is a phase diagram that the IBM people were, they stated in their first paper,

00:10:49 this is the diagram that they were looking at when they decided to look at these copper oxides.

00:10:59 And I don't want to go into the details of this.

00:11:03 I think it's not important to really talk about what bipolarons are,

00:11:09 but basically what we have is a phase diagram of the metal on one side,

00:11:14 an insulator, a semiconductor on the other side.

00:11:17 The point is that at the boundary between the metal and the semiconductor,

00:11:20 we have a superconducting region.

00:11:23 Now this is a phase diagram which was postulated and certainly is not in detail well established,

00:11:32 but it is similar to phase diagrams that I have used in the past

00:11:39 to search for new superconducting materials,

00:11:41 where we have then as a function of temperature here,

00:11:44 a metallic region, semiconductor region, some slope to this line,

00:11:47 so that we do allow for metal insulator transitions, metal semiconductor transitions,

00:11:52 where we have a metallic state at high temperatures

00:11:55 and then go semiconducting at low temperatures.

00:11:58 But what is frequently found is right at this boundary line,

00:12:03 just before one loses the metallic properties,

00:12:07 one has superconductivity which then quickly disappears

00:12:11 as one reaches this boundary of semiconducting properties.

00:12:17 I have indicated a possible chemical handle here called an ionicity,

00:12:25 and I'll come back to that later.

00:12:28 It does not necessarily have to be an ionicity.

00:12:31 I think there's a number of chemical handles that would drive you back and forth

00:12:35 across a phase diagram like this.

00:12:39 One thing that does tend to happen at this boundary

00:12:42 is that you will split into two phase regions.

00:12:45 That's a common problem actually in searching for superconductors.

00:12:52 And that's what we in fact found in this barium lead bismuth oxide system.

00:12:58 When we started in on this, barium lead oxide was known to be a metal.

00:13:03 Barium bismuth oxide was not well characterized,

00:13:07 and I was wondering whether it might be a case of a bismuth 4 plus cation,

00:13:15 which being chemistry known, bismuth 4 plus doesn't exist.

00:13:19 But if it were a metal and one had a half-filled S-band, that might be acceptable.

00:13:26 So we grew crystals of this material, and they looked metallic,

00:13:32 and they're luster and so forth, but in fact they turned out to be semiconducting.

00:13:37 Once we knew we had a semiconductor for this compound,

00:13:42 and of course it was already known that this was a metal,

00:13:45 we started in with a solid solution to work on the space diagram

00:13:51 looking for the metal to semiconductor transition.

00:13:55 And the first compound we made was at 0.5.

00:13:58 It was still a semiconductor.

00:14:00 The second one we made was 25% bismuth, and that was a superconductor.

00:14:04 So it took us two experiments in the space diagram to find a superconductor at 13K.

00:14:15 The structure of this is the perovskite structure, which hasn't really been described yet.

00:14:24 In this structure, the oxygens are not shown.

00:14:27 They're halfway between every open and filled circle here.

00:14:33 There's an oxygen.

00:14:35 So that all these bismuth cations are in octahedral coordination.

00:14:40 This is the actual structure that we determined from neutron diffraction

00:14:44 for aspirin bismuth oxide.

00:14:48 As one substitutes lead into there up to getting way over on the leverage side of that phase diagram,

00:14:56 this ordering disappears.

00:14:59 You have only one type of site for the octahedral cation,

00:15:04 and at that point you probably really should consider this bismuth 4+.

00:15:09 It's a devocalized S electron.

00:15:15 Now, going on into the copper oxide systems, which are frequently referred to as perovskite-related,

00:15:23 I'd like you to focus on these planes of cations, which have, of course, oxygen ligands,

00:15:32 but this would be one plane and another plane here and another plane down here.

00:15:37 They're linked, in this case, by oxygens, which are not shown,

00:15:41 and that gives strong bonding in all three dimensions, and one ends up with a cubic structure.

00:15:46 But these planes are indicated here, and in some of these copper oxide structures,

00:15:51 these planes end up being very two-dimensional in nature.

00:15:55 One does not necessarily have strong bonding in the other dimension.

00:16:00 So we have copper here in square planar coordination to oxygen, infinite sheets here.

00:16:08 And if we put a 5th and 6th oxygen ligand on that copper,

00:16:13 one could build up the perovskite structure shown on the previous slide.

00:16:21 Well, with this sheet, if this is dimel and copper now,

00:16:27 depending on the other cations around, perhaps on the structure and other factors,

00:16:36 one can either have localized d-electron behavior for that copper, too,

00:16:41 or either delocalized or localized behavior.

00:16:45 And this is important to develop the phase diagram that I'll be showing a little later,

00:16:51 so we need to discuss the inductive effect that can occur with other cations present in these materials.

00:17:01 By this I mean if we have this M cation might be copper or some other transition metal

00:17:07 with a partially filled D shell, and then we have some other cation bound to that oxygen,

00:17:12 such as a rare earth or an alkaline earth, and as this A cation becomes more electropositive,

00:17:19 which is the same thing as saying as this bond to A to O becomes more ionic,

00:17:25 then the O to M bond becomes more covalent, and that drives the system towards delocalized D states.

00:17:34 And there are a number of systems where one can see the effect of that.

00:17:39 As the perovskite systems, for example, with the rare earth titanium oxide series,

00:17:46 for lanthanum titanium oxide with the most electropositive rare earth,

00:17:52 one sees delocalized D electron behavior for titanium going to a smaller, less electropositive rare earth cation.

00:18:01 One sees localized behavior for titanium.

00:18:03 Again, in a different structure, the pyrochlor structure, the 3D2 system,

00:18:09 one sees delocalized electron behavior with the larger rare earths,

00:18:14 the more electropositive ones, and localized behavior for the smaller, somewhat less electropositive rare earth cations.

00:18:24 Now, as the covalency of the copper-oxygen bonds increase,

00:18:34 one is driven from what on this side we see basically an animal diagram,

00:18:41 but as we increase the covalency, the copper-oxygen bonds were driven towards bands,

00:18:47 and in the case of copper 2, we have the sigma star level,

00:18:54 which is basically the dx squared minus y squared level,

00:18:58 half-filled and a half-filled band of metallic properties.

00:19:02 Since it's square planar, I've indicated the dz squared as a non-bonding state,

00:19:08 and therefore it does not broaden significantly.

00:19:14 Okay, now I think we're ready for this.

00:19:17 Now we have to first look again at the structures, okay, before we get to the phase diagram.

00:19:23 In the case of lanthanum copper oxide, which is what's really started all this superconductivity,

00:19:32 and the copper oxide systems, we have these sheets of square planar copper here,

00:19:38 another one here, another one here, infinite sheets,

00:19:42 and between these sheets, in this case, we're not strongly bound in the third direction.

00:19:49 We simply have weak ionic-type oxygen bonding and bonding to the lanthanums here.

00:19:58 Now, for other rare earths other than lanthanum, the structure is slightly different,

00:20:04 but the essential features are exactly the same.

00:20:07 For the praseodymium through gadolinium compounds of this same formula,

00:20:12 one again has exactly the same copper square planar sheets here, here, and here.

00:20:18 The only difference is that the exact placing of the oxygens in between the sheets is a little bit different.

00:20:29 Finally, the phase diagram that I've been leading up to.

00:20:33 This is a phase diagram that I just put down on paper for the first time last week.

00:20:41 This is its first public airing, and I'm just getting to be somewhat comfortable with it myself.

00:20:47 Basically, what is known is that for this formula, if we go from praseodymium through gadolinium,

00:20:54 these are well-established semiconducting materials.

00:20:58 This should be, as is one of the titles, this should be copper-oxygen.

00:21:03 The bond length is 1.95.

00:21:06 Then we go to lanthanum.

00:21:09 We have a somewhat different structure, but we still have the square planar sheets,

00:21:16 but we now have metallic properties, and the copper-oxygen distance has shrunk to 1.90.

00:21:22 That, in fact, is typical of what happens when one goes across one of these boundaries.

00:21:27 When one goes from a semiconductor to a metal,

00:21:30 one frequently does see the metal-oxygen distances decrease by about that much.

00:21:36 So, based on the inductive arguments I was using before,

00:21:42 as we go from gadolinium to lanthanum, I would be expecting this is increased ionicity,

00:21:49 bonding from the rare earth to oxygen,

00:21:52 but this causes an increased co-balancing of the bonding and the copper-oxygen bonds,

00:21:57 and that's driving us toward this metallic region.

00:22:01 But with lanthanum, we didn't quite make it.

00:22:04 We made it to an interesting area.

00:22:06 We made it right to the boundary, but we did not get to the superconducting boundary.

00:22:12 Unfortunately, we were out of steam with trivalent cations.

00:22:17 There is no trivalent cation, which is any more electropositive than lanthanum.

00:22:23 So the only thing that could be done would be to go to alkaline earths

00:22:27 or possibly alkalize to push us further in this direction.

00:22:33 And it is, in fact, only barium, strontium, and calcium

00:22:36 that have been found to push us further in that direction

00:22:39 and get into the superconducting region.

00:22:42 If one uses other divalent cations, such as cadmium or lead,

00:22:46 which are not more electropositive than the rare earths,

00:22:51 one does not push into the superconducting region.

00:22:55 Now, you notice here that I put lanthanum right on the boundary line,

00:23:06 which means then that something should occur as a function of temperature.

00:23:11 Of course, I would not have done that if I hadn't known about this.

00:23:17 In the case of lanthanum copper oxide, it is observed at low temperatures

00:23:23 that it does appear to become—it's usually described as metallic at higher temperatures,

00:23:28 but somewhere below 200 K, the resistivity starts shooting up.

00:23:33 And so I am, at the moment, speculating that this may be a metal-to-semiconductor transition.

00:23:41 Now, there's another transition at higher temperatures

00:23:44 that there's been a lot of talk about.

00:23:47 I think a number of us have concluded that it's not an important transition,

00:23:51 although I'm still not totally convinced that it's a transition we should be ignoring in these systems.

00:23:58 There is a symmetry change, and I think that this transition is, in fact,

00:24:03 one that's very easy to understand based on chemistry.

00:24:08 In the tetragonal phase, we have linear copper oxygen, copper bonds,

00:24:15 and then when we go orthorhombic, we develop an angle there.

00:24:21 Now, remember from—well, I mean, you know, this is basically a d-9 system.

00:24:27 The copper d-orbitals are essentially filled,

00:24:30 and the pi bond that would result from overlap of copper into this p orbital indicated here

00:24:37 would be strongly anti-bonded in nature if the d-orbital extent was large enough to make this a consideration.

00:24:45 So that should drive it towards a bent bond,

00:24:49 and there's, of course, plenty of analogies like this in chemistry.

00:24:53 What we're doing is getting rid of an anti-bonding situation

00:24:56 and going to a non-bonding situation by pushing those p lobes out of the bond,

00:25:01 essentially off to the side.

00:25:04 I could say, why doesn't this also happen for the other rare earths,

00:25:07 and I would say that when we go from a bond distance which is 1.90 as in lanthanum copper oxide,

00:25:15 there we've got a marginal situation.

00:25:17 We go to 1.95 copper oxygen distance,

00:25:21 and then that anti-bonding pi interaction has become small enough

00:25:26 so that there's no longer a big driving force to bend the bond.

00:25:32 But all of this is actually well below the Fermi level in the case of these materials.

00:25:40 That is, we're changing the situation with regard to pi bonding,

00:25:47 but these pi bonding levels are well below the Fermi level,

00:25:52 and I don't see how this could basically be a factor in electron-phonon interactions.

00:26:01 I think a number of other people have concluded the same.

00:26:05 In summary, what I've tried to show is that there are some chemical concepts

00:26:12 which can be used to control how close one is to the semiconductor metal boundary,

00:26:21 and that if one manipulates systems through chemistry to get close to this boundary,

00:26:27 there's a good chance of finding superconductivity just on the metal side,

00:26:32 provided one does not run into complications such as magnetism.

00:26:39 Also, it appears that, at the moment at least, that the D-shell must be nearly empty,

00:26:45 and we've known that for years, that side of it.

00:26:47 But what we've just learned is that we can work on the other side where the D-shell is nearly filled.

00:26:54 There are a lot of questions remaining, and as the previous speaker indicated,

00:27:01 there may be a new mechanism involved here.

00:27:04 I certainly don't feel comfortable that we've explained the high-T superconductivity in these systems

00:27:10 using the old ways of explaining superconductivity.

00:27:16 Thank you.

00:27:17 Thank you.

00:27:26 Tonight, Don will speak to us on the chemistry of cuprate superconductors.

00:27:32 Thank you, Belle.

00:27:37 I'm going to try to talk about mainly the chemistry and materials of these cuprate superconductors.

00:27:44 On the first slide here, I've shown a number of the people that have been involved in this work,

00:27:49 mainly the people who have prepared materials on the left-hand side here,

00:27:53 and a number of the physicists.

00:27:55 There are many more physicists, but these are the ones that I'll be drawing on their results tonight.

00:28:03 I can't really go into each one of these people and explain to you in detail what each one of them has done.

00:28:08 On each of the slides, I've tried to write the name of the person who actually did the work.

00:28:14 I would, however, like to mention a few people in particular.

00:28:17 Bob Cava, Bertram Badlog, and Bruce Vandover have really been key people in all aspects of this work.

00:28:24 Another Bell Labs person that I'd like to just give my gratitude to again is Val Cook,

00:28:31 who's not involved in superconductivity at all, but she's really got this symposium organized,

00:28:36 believe me, over the protestations of all the speakers who were meeting out.

00:28:45 On this slide, I've summarized again, and this is a slightly different picture than art used,

00:28:49 of the structures of these cuprate compounds.

00:28:52 On the left-hand side is the perovskite structure.

00:28:56 Basically, this structure is just a cubic structure in the ideal form of octahedra sharing corners

00:29:05 with a large A ion in the middle.

00:29:07 The general formula is ABO3.

00:29:10 The compounds listed under that happen to be superconductors,

00:29:14 but in this compound and in the K2NiF4 structure, believe me,

00:29:18 chemists have worked long and hard on these materials, and there are literally hundreds of examples known.

00:29:23 I think based on what we now know, a lot of that needs to be reexamined.

00:29:30 The K2NiF4 structure is very closely related to the perovskite structure.

00:29:36 You can see that what happens is we have cleaved at these oxygen bonds here

00:29:44 so that the layers have come apart.

00:29:47 Now we have planar layers of octahedra here,

00:29:52 which are now totally separated from the next layer, which is also offset by half a unit cell.

00:29:59 Now the A ions, the potassium ions in the K2NiF4 structure, are in between those layers.

00:30:06 Actually, only half of them are shown in this slide for clarity.

00:30:10 Again, believe me, there are hundreds more of these kind of compounds that are known.

00:30:15 Based on the early IBM work and the Japanese phase identification of the K2NiF4 structure

00:30:27 as the superconducting component in the lanthanum barium copper oxide,

00:30:32 we started substituting all kinds of things into these compounds.

00:30:36 Within a week after we knew about that work,

00:30:41 Bob Cava had submitted a paper on the lanthanum strontium,

00:30:45 where Tc had gone up another 10 degrees,

00:30:47 which was the biggest jump anybody had ever seen in superconductivity at that time.

00:30:53 We were trying to substitute all kinds of other things at the same time.

00:30:57 We had many people trying to substitute different things,

00:31:00 different transition metals, different A ions.

00:31:02 Most of those didn't lead to too much in the way of superconductivity,

00:31:06 although there are many stories to be cleaned up and retold there some other day.

00:31:11 One of the things that might seem obvious to substitute in that K2NiF4 structure

00:31:16 is yttrium or smaller rare earths for the lanthanum.

00:31:20 In fact, Art already mentioned to you that when you do that,

00:31:23 you don't get the same structure.

00:31:25 In fact, Bob Cava had substituted a small amount of yttrium

00:31:29 for lanthanum in that K2NiF4 structure.

00:31:34 Tc went down.

00:31:35 We were very surprised when Paul Chu and his coworkers

00:31:39 reported superconductivity in the 90-degree range in yttrium barium.

00:31:44 What we did as soon as we heard about that,

00:31:48 we made a little ternary phase diagram

00:31:51 listing all of the compounds that we knew of that we could find in the literature

00:31:55 that had the elements of yttrium, barium, copper, and oxygen.

00:31:59 Those are all the ones listed there,

00:32:01 and the little circle there that says Chu,

00:32:04 that's the composition that Paul Chu reported.

00:32:09 You'll hear more about this kind of phase diagram later from Tom Mason.

00:32:13 He's going to talk in quite detail about this.

00:32:16 But in our very first reactions,

00:32:20 we identified that a major fraction of what was present in Paul Chu's composition

00:32:24 was this Y2BaCuO5,

00:32:26 which is a compound which had been made a few years ago by Riveau in France.

00:32:31 Realizing that that was a major constituent of his phase,

00:32:35 then we knew that the actual superconducting phase

00:32:38 had to be on the opposite side by the lever rule up in this area.

00:32:43 Having made a whole bunch of compounds right around the phase diagram,

00:32:47 zeroing in on that region,

00:32:49 within a couple of days we had zeroed in on this composition,

00:32:53 which we formulated as 2-1-3.

00:32:57 I call these the Los Angeles phases, for those of you who know the area code,

00:33:01 instead of 1-2-3.

00:33:03 But anyway, that ratio of metals forms a single phase compound,

00:33:09 which Meissner Effect, by Bertram Badlock, showed was a bolt superconductor.

00:33:14 I've left the oxygen off here,

00:33:16 because at this time we didn't know how much oxygen it contained.

00:33:19 Now, I just want to digress and show you a little chemistry here.

00:33:23 These are actually synthesized compounds,

00:33:25 and here's the way you make them.

00:33:27 You take either mixtures of oxides or oxide precursors, such as carbonates,

00:33:32 mix them together in the proper proportions,

00:33:34 grind them up in a mortar and pestle,

00:33:36 heat them in a furnace,

00:33:38 do this two or three times until you get a homogeneous mixture,

00:33:41 and then what's turned out to be somewhat important in many of these

00:33:45 is a final treatment in pure oxygen.

00:33:50 Now, what's important to know here

00:33:52 is what is the actual formal oxidation state of copper.

00:33:56 Looking at it from a parochial viewpoint,

00:33:59 we assign all the variable valence to the copper.

00:34:05 In actual fact, I think the molecular orbital calculations

00:34:09 sort of indicate that it's equally on oxygen as well as copper,

00:34:13 so that we can view this as somewhat as oxygen 1-

00:34:17 just as easily as copper 3+.

00:34:19 But at any rate, the way one does that is to reduce these things

00:34:22 in a TGA with hydrogen,

00:34:24 measure the weight loss according to the reaction that's listed there,

00:34:29 and if you do that, you come out with a number very close to 7

00:34:33 for the amount of oxygen in this, which works out to copper 2.26.

00:34:38 In the lanthanum strontium copper oxide case,

00:34:41 about 2.15 was the copper oxidation state.

00:34:46 Now, from our initial reactions on pure phase material,

00:34:51 we were able to index a powder unit cell

00:34:54 based on the lattice parameters shown here,

00:34:57 which appears to be very closely related to a triple perovskite cell.

00:35:03 We were also aided by getting some small single crystals

00:35:07 in one of our very first reactions,

00:35:09 and we got these parameters off that crystal as well,

00:35:12 which also helped enormously.

00:35:14 The triple unit cell we formulated as a rising, as shown here.

00:35:20 The perovskite unit cell is shown on the left,

00:35:23 and barium is in the middle.

00:35:25 The oxygens are on the midpoint here,

00:35:27 and copper is at the corner of the unit cells.

00:35:30 Now, if we just stack three of those on top of each other,

00:35:34 and now yttrium and barium are much different in size,

00:35:37 and so they will order in a structure like this,

00:35:41 and that's where the tripling of the unit cell comes.

00:35:44 Also, an ideal perovskite would have nine oxygens,

00:35:48 whereas we've analyzed this to have seven oxygens,

00:35:51 so it's got to have a lot of missing oxygens.

00:35:53 So where is the oxygen going to be missing?

00:35:55 Well, since yttrium is a lot smaller,

00:35:58 we expected this plane to be completely empty of oxygen

00:36:01 so that the smaller yttrium could be accommodated.

00:36:05 Now, that takes us down to eight oxygens,

00:36:09 and we really had no preconceived notions

00:36:11 where the other missing oxygens might be.

00:36:14 But as I mentioned, we were fortunate enough

00:36:16 to get some small single crystals out of one of our first reactions,

00:36:20 and we determined the single crystal x-ray structure on that,

00:36:23 which is shown here.

00:36:25 And basically, this is the structure that I showed you,

00:36:30 and the extra missing oxygens were in this plane,

00:36:34 where we had two different oxygen sites,

00:36:39 and we had close to 50% occupancy on each of those.

00:36:43 We'll get back to this a little bit later.

00:36:46 Now, what's important when we're going to start

00:36:48 trying to do some chemical substitutions in this

00:36:50 is to know what the sites of the different ions are like,

00:36:53 what might we substitute.

00:36:55 The yttrium and the barium sit on quite different sites in this structure.

00:36:58 The yttrium sits in an eight-coordinate site,

00:37:01 which is basically cubic, looks like a fluorite-type structure,

00:37:05 or a site in something like cubic zirconium.

00:37:09 The barium is much larger,

00:37:12 and it sits in a very traditional perovskite site.

00:37:15 It's 12-coordinate,

00:37:17 keeping in mind that two of those oxygens are statistically missing.

00:37:21 The bond distances are much different.

00:37:24 But copper is somewhat more interesting.

00:37:27 We had two different kinds of copper in the structure.

00:37:30 The two planes closest to yttrium

00:37:33 look like this copper 2 on the left.

00:37:36 It's basically a square plane of copper

00:37:38 with distances of 1.95,

00:37:41 which is in the regime where

00:37:45 Art said, at least in the K2-NiF4s,

00:37:48 that the things weren't conducting.

00:37:51 We have a fifth oxygen at about 2.4 angstroms away,

00:37:55 and that's about the same distance as the axial oxygens are

00:37:58 away from copper in the K2-NiF4 structure.

00:38:02 We can look at this as either 4- or 5-coordinate.

00:38:05 Take your pick.

00:38:07 But the other copper in the basal plane

00:38:10 where there were some missing oxygens,

00:38:12 at first glance has an octahedral coordination,

00:38:16 except that it's a very funny octahedral.

00:38:18 In the axial, there's 4 oxygens at 1.93 angstroms

00:38:23 and 2 oxygens at shorter distances, 1.8.

00:38:28 Now, we would expect the axial oxygens to be longer,

00:38:32 so what we actually postulated,

00:38:34 keeping in mind that half of the oxygens in this plane are missing,

00:38:38 is that what we really had was square planar geometry around that copper,

00:38:42 which could be oriented around the copper

00:38:45 either in 2 mutually perpendicular directions.

00:38:50 In fact, I don't have a slide,

00:38:53 but several groups have now done neutron powder diffraction studies of this

00:39:01 and have seen that, in fact,

00:39:04 in those samples, the oxygen is completely missing from one of these

00:39:10 and completely occupies the other,

00:39:12 so that, in essence, what we have in the structure

00:39:15 are 2-dimensional layers of copper oxides here

00:39:19 and 1-dimensional chains running in this plane.

00:39:26 And I think we still don't know

00:39:31 what of those features are important and critical to the superconductivity.

00:39:36 Now, I just want to go into the properties just a little bit here.

00:39:41 Here's the magnetic susceptibility that a chemist might see for this material.

00:39:48 Above Tc, the compound is paramagnetic,

00:39:52 and the paramagnetism comes from 2 components.

00:39:55 There's a temperature-independent portion,

00:39:58 which metals generally have, a polyparamagnetic term,

00:40:02 and from that one can derive something about the density of states,

00:40:05 which is this gamma term here.

00:40:08 And there's another component, which a basic Curie-Weiss law,

00:40:11 and the value one gets is 0.3 Bohr magnetons per copper,

00:40:18 is how much localized moment there is in this structure.

00:40:22 Now, Art mentioned to you that these compounds

00:40:25 are often close to the borderline between localization and metallic behavior,

00:40:29 and that certainly seems to be what's happening here.

00:40:34 Now, remember we have 2 different kinds of copper in the structure.

00:40:38 I don't think we know yet where the local moment resides in the structure,

00:40:43 whether it's on particular coppers,

00:40:45 whether it's an awful lot to be on defects.

00:40:49 This would correspond to about one-sixth of all the copper being localized copper 2+.

00:40:57 The other feature, which you'll probably notice here,

00:41:00 is that as you go below Tc, this thing suddenly goes totally diamagnetic.

00:41:05 There is a sudden drop that would be very difficult for anyone to miss.

00:41:14 Now, this table shows a number of compounds that we've made,

00:41:18 which we've obtained as single-phase materials now,

00:41:21 where we've substituted ions on these various sites.

00:41:28 Several different rare earths can be substituted for the yttrium.

00:41:31 Most of the smaller rare earths, we can make mixtures of those.

00:41:38 A couple here, one in particular that deserves a little bit of mention,

00:41:42 which isn't in that category, is barium lanthanum.

00:41:46 What I've formulated here is Ba2La.

00:41:49 Now there, we get a cell which appears to be cubic.

00:41:52 My guess is on closer inspection it won't really be cubic,

00:41:55 but it appears to be somewhat different.

00:41:58 Here, we don't have that great size difference in ions,

00:42:01 so we might not get the same kind of ordering effects,

00:42:06 which lead to the tripled unit cell.

00:42:12 I guess no talk like this would be complete

00:42:14 without showing a few resistances of some of these materials.

00:42:18 This is sort of what it looks like.

00:42:20 If you measure the resistivity of these materials as a function of temperature,

00:42:25 this is only four of those materials.

00:42:27 They don't go at exactly the same temperature,

00:42:29 but there isn't a big difference within that one structure type.

00:42:35 Here's a table if you're really interested in numbers.

00:42:38 I think I already told you what was important.

00:42:41 You can make a whole bunch of compounds,

00:42:43 and r equals zero changes somewhat, but it doesn't change enormously.

00:42:48 I think what we now have is a large series of compounds

00:42:53 which people can start to do some systematics on

00:42:56 and try to really understand these materials.

00:42:59 The ones where I've got an asterisk here,

00:43:01 these are materials that we know aren't single phase in these measurements.

00:43:09 One of the things that's very important, as was mentioned earlier,

00:43:12 is what kind of critical fields can these materials take.

00:43:15 Let me just, for reference, tell you that the critical field

00:43:19 at helium temperatures for niobium-10 is around 30 tesla,

00:43:24 and magnets operate at around 18 tesla.

00:43:29 The latest materials in this line that we've measured,

00:43:35 and there's results on the yttrium compound alone from Chu,

00:43:40 which was around 20 tesla at 77 degrees for the pure yttrium compound,

00:43:46 and that's about what we get as well.

00:43:49 As we get more and more europium, the critical field gets much higher.

00:43:54 I wouldn't say we totally understand that at the moment,

00:43:58 but let me just tell you that above liquid nitrogen temperature,

00:44:02 we have critical fields here that are higher than they need

00:44:07 to make the same strength magnets that people today make at 4.2 degrees.

00:44:14 Now, I just want to show you, I mean, as Art mentioned,

00:44:18 these compounds are only five weeks old at the oldest,

00:44:23 but there are people already starting to think about applications and materials.

00:44:27 These are some materials produced by Dave Johnson, a ceramist at Bell Labs.

00:44:32 This is a precursor to a tape.

00:44:37 This is powder of the superconducting material during yttrium,

00:44:41 which has an organic binder, and it's flexible.

00:44:43 It's like an organic polymer.

00:44:46 After that's fired at high temperature, it forms a somewhat stiffer ceramic.

00:44:52 It can be bent about as much as is shown here

00:44:55 and is then as R equals zero above 90 degrees.

00:45:03 Sung Ho Jin and his colleagues at Bell Labs have made wire.

00:45:08 Here's a coil about 10 inches long, which has R equals zero above 90 degrees.

00:45:16 This has been made by taking a metal tube

00:45:19 and filling it with superconducting powder and swaging it down,

00:45:23 standard metallurgy techniques,

00:45:28 and then doing some heat treatments on it after that.

00:45:32 This slide shows you what that wire looks like in cross-section.

00:45:37 This is about a 50-mil wire in this particular case.

00:45:41 Here's the superconductor, and here it is looking at it lengthwise.

00:45:49 I'd like to just compare a couple of things that I think are key differences

00:45:55 between what we are now calling the low-temperature materials

00:45:59 and the higher-temperature materials.

00:46:03 The lanthanum-strontium system has an extensive single-phase region.

00:46:07 You can vary the lanthanum-strontium ratio pretty much at will.

00:46:13 You only get bulk superconductivity over a very narrow range.

00:46:17 You can see superconductivity almost anywhere you look,

00:46:20 but you really only have bulk superconductivity in a very narrow range,

00:46:24 which is difficult to control.

00:46:26 In the yttrium-barium case, because of the size difference of the ions,

00:46:32 you essentially have a line phase.

00:46:34 There is no variability in the yttrium-barium-copper ratio,

00:46:38 which I think is actually a nice advantage to these materials,

00:46:42 because if you screw up, you're going to have a bit of a second phase,

00:46:48 but nothing really is going to happen to the good stuff.

00:46:52 This actually, I think, is somewhat changed now.

00:46:55 I was going to say that the yttrium-barium is not very stoichiometric in oxygen,

00:46:59 but based on the latest neutron structure that I was describing,

00:47:03 7 is the ideal stoichiometry,

00:47:05 and that actually is sort of the best superconducting composition.

00:47:10 There still may be more variability,

00:47:12 but this is a question that needs further examination in both systems.

00:47:16 Another difference here, which we need to follow up on the importance,

00:47:20 is that in the lanthanum-strontium case,

00:47:22 all the copper in the structure is equivalent,

00:47:24 whereas in the yttrium-barium, you really have two different kinds of copper,

00:47:28 and the importance of that needs to be looked at further.

00:47:34 Now, I only have one more thing that I want to do,

00:47:37 and that's I want to leave you with a visual impact,

00:47:40 some things which these still photographs, slides, can't possibly do justice to.

00:47:45 I have a videotape, which was made by Doug Osheroff,

00:47:50 who's a low-temperature physicist at Bell Labs,

00:47:52 which demonstrates two experiments,

00:47:55 one of which is very easy to do in the laboratory,

00:47:58 and I think people who teach a lab course,

00:48:00 this would be a great undergraduate laboratory experiment

00:48:04 to really let people know they can make a superconductor

00:48:08 and then they can look at some of its properties.

00:48:11 It's an example of the diamagnetism and how you can see it.

00:48:16 And if we can start the tape, is there somebody running it?

00:48:22 How do we start it?

00:48:27 What you're going to see, the first segment goes by rather quickly.

00:48:31 What you're going to see is a pellet of the superconducting compound.

00:48:36 It's being chased around by a magnet on the end of a stick,

00:48:40 and I just don't know how to do justice to this in a slide.

00:48:44 Look in.

00:48:53 Look, Ma, no scrapes.

00:49:00 Now, the fact that that can sit still above this

00:49:04 demonstrates that there are persistent currents in this superconductor.

00:49:08 You cannot do this with magnets.

00:49:12 I was really amazed when I saw how stable this was.

00:49:15 You can shove this all around, and it comes back to very stable positions.

00:49:19 This is the principle of magnetic levitation trainings,

00:49:22 even though I think this magnet looks more like a space shuttle.

00:49:27 I think this is really something that left a visual impact on me,

00:49:32 and I hope it will on you.

00:49:35 Thank you very much.

00:49:56 Thank you.