Making High-Temperature Superconductors

Copyright 2008-2009, Futurescience, LLC

What follows are the instructions for making ceramic superconductors in a high school science laboratory.  These were originally the instructions that were included in a Colorado Futurescience kit for superconductor fabrication.  We have not sold the fabrication kit for many years, so we decided to put these instructions on the web site.

Since we're making this information available without charge, we can assume no liability for the safety or reliability of the procedure.

This procedure was originally written in 1988 for high school science labs; but, of course, it can also be used at colleges and in other settings.  Individuals unaffiliated with any research laboratory or educational institution have also used this procedure successfully.  Unaffiliated individuals, however, may have difficulty obtaining the required chemicals.  None of the major suppliers of research chemicals will sell chemicals to individuals.  Colorado Futurescience no longer exists.  Futurescience, LLC also no longer regularly sells these chemicals, although we are willing to discuss making some of the necessary materials available on a custom-made basis to universities and to qualified high school science teachers.  We have put together custom kits with some of the harder-to-obtain materials for schools.  For the unaffiliated individual with a scientific background, one option for making these superconductors is to volunteer to assist a High School Science Department with this project.  Many schools are in need of qualified volunteers to assist with science projects.

Making yttrium-barium-copper-oxide superconductors requires the following items:

  • Yttrium Oxide
  • Barium Carbonate
  • Cupric Oxide
  • A Laboratory Furnace such as a Thermolyne Muffle Furnace or a converted pottery kiln.
  • Labware made of alumina.  The recommended labware is the 20 ml. CD-20 alumina dish made by Coors Ceramics.
  • An Oxygen Source
  • Liquid Nitrogen and a rare-earth magnet for testing and demonstrating the superconductors

(See the listing of sources of equipment and supplies for making superconductors.)

Important Safety Note:   Barium Carbonate is a toxic substance.  It can be safely handled with ordinary laboratory procedures. It is imperative, however, that (especially during the grinding and mixing of the chemicals) a good-quality dust mask is worn during the procedure.  It is also important to wear laboratory gloves, such as disposable latex surgical gloves, while working with these chemicals.

The process described here may appear rather formidable at first, but it is actually straightforward and typical of ceramic processes often used in scientific research.  The most time-consuming part of the process is gathering the necessary materials and equipment, since the materials and equipment are uncommon in a high school laboratory where experiments with ceramic materials are rare.

There are a number of methods of preparing yttrium-barium-copper-oxide superconductors.  The simplest is the one that has become known as the "shake-and-bake" method.  This involves a four step process.


  • 2.   CALCINATION (the initial firing)

  • 3.   THE INTERMEDIATE FIRING(S) (oxygen annealings)


Some of the details of the procedures are very flexible.  The number of intermediate firings and the length of the firings are largely up to the user.  In general, the more intermediate firings, and the longer the duration of the firings under oxygen flow, the better the superconductor.  But definite signs of superconductivity can usually be obtained without any intermediate firing at all.  In fact, if the initial mixing of the chemicals is sufficiently thorough, the intermediate firing is not necessary at all.

It is recommended that no more than half of your chemicals fired at a time so that all of your chemicals will not be ruined if there is a problem during a firing.  The most common problem encountered is overheating due to errors in the pyrometer on the furnace.  It is strongly recommended that the calibration of your temperature indicator be checked before the first firing.  If there are any doubts about the accuracy of your temperature indicator, an inexpensive pyrometric cone can be obtained from a pottery shop for a reasonably good calibration check.  See the section of the instructions on pyrometric cones.

See Using a Pottery Kiln as a Laboratory Furnace and Using Pyrometric Cones to Calibrate Furnace Temperature Indicators.

1. MIXING THE CHEMICALS.  The starting mix is a gray powder formed by thoroughly mixing yttrium oxide, barium carbonate, and cupric oxide.  There are a number of workable methods of mixing the powders.  The recommended procedure for modestly equipped laboratories is to first use a mortar and pestle to grind down any lumps or large particles in the chemical powders, then to shake the mix vigorously in a capped jar or stoppered flask for several minutes.  The ultimate purpose of this entire process is to combine the starting compounds at the molecular level.  A really thorough dry mixing of chemicals is a difficult process, but taking extra time and work during the initial dry mixing process will pay off later.  (Mixing in a sealed plastic bag is also a workable mixing method, especially if your starting chemical are lumpy, but it is best to shake them thoroughly in a jar at the end of your mixing process.)  Use a dust mask and avoid breathing any of the chemical powders.  The chemicals must be mixed in the proper proportions so that the atomic ratios of yttrium, barium, and copper are 1:2:3.  (For an explanation of how this is done, see the section on molecular weights.)  The proper weight ratios are:

  • Yttrium Oxide, Y2O3    11.29 grams
  • Barium Carbonate, BaCO3    39.47 grams
  • Cupric Oxide, CuO    23.86 grams

An important note about temperatures:   The furnace temperatures stated here are temperatures as they are usually measured on the temperature indicator of the furnace.  In most cases, the temperature sensor is located slightly higher in the furnace chamber than the sample you are firing.  Therefore, the indicated temperature will often be a few degrees higher than the actual temperature of the material.  If the temperature sensor in your furnace is at the same level as the material being fired, the maximum indicated temperature should never be allowed to exceed 980 degrees Celsius.

2. CALCINATION.  For the initial heat treatment, called calcination, the mix is heated at 925-950 degrees Celsius for about 18-24 hours.  (The 950 degree temperature works much better if this is within the rating of your furnace for continuous use.)  This first treatment may be done a crucible or evaporating dish made of alumina or of a good grade of laboratory porcelain.  It is best to use alumina unless you have a particular reason for using ordinary porcelain labware.  This first heat treatment forms the basic crystal structure of YBa2Cu3O6.5, and gets rid of the carbon dioxide from the barium carbonate.  (Barium carbonate is used instead of barium oxide because barium oxide of any reasonable purity is difficult to obtain.  Also, exposing barium oxide to air tends to quickly convert much of it to barium carbonate and barium hydroxide.)

The result of this first firing is a porous black or very dark gray clump.  The coloration should be fairly even.  An uneven green coloration is an indication that the powders are not as thoroughly mixed as they should have been, and that extra time and care should be taken to insure thorough grinding and mixing on subsequent steps.  The material will seem to shrink rather dramatically during the initial firing as it loses its carbon dioxide and becomes much denser than the original powder mix.

3. INTERMEDIATE FIRING(S).  The porous black clump is ground into a fine powder and placed in the furnace in an alumina dish.  A final furnace temperature of 925 to 975 degrees Celsius is recommended for the intermediate firings.  A temperature much higher than this will result in a material that is much harder to re-grind.  Temperatures above 1000 degrees Celsius may destroy the crystal structure.

After the mix has heated in the furnace for at least 18 hours at 925-975 degrees Celsius, begin a slow flow of oxygen into the furnace, and reduce the temperature slowly.  If you plan to test the sample for superconductivity after this firing, the cooling rate must be no more than 100 degrees per hour until 400 degrees Celsius is reached.  The rate of cooling from 400 degrees down to room temperature can be increased to about 200 degrees per hour.  If you do not plan to test for superconductivity after this firing, a cooling rate in excess of 100 degrees per hour may be used; however a cooling rate in excess of 250 degrees per hour is not recommended.  Do not remove the oxygen flow until the indicated furnace temperature has fallen below 400 degrees Celsius.

The material should be thoroughly re-ground in a mortar and pestle (or similar device) between each firing.  (If, after an intermediate firing, there is some green coloration in the resultant disk, it is important to take extra time and care in re-grinding and mixing the material before the next firing.)

Problems that occur in the mixing and grinding process in any of these steps are often due to hard, coarse particles being mixed in with the finely powder material.  An ordinary kitchen tea strainer can come in handy at this point to separate the coarser particles or lumps so they may be ground separately.  IMPORTANT:   If you an ordinary tea strainer, make sure it is made of a non-magnetic material, or make sure you are satisfied that none of the material in the sifter or strainer will contaminate the chemicals.  Even very small quantities of magnetic materials in the chemical mix can diminish or destroy the potential superconductivity.  (It is also for this reason that "ceramic grade" chemicals, which tend to have iron impurities, are not often usable for making superconductors.)  Shortcuts in grinding the materials, such as using an electric coffee grinder, often contaminate the compound with elements that destroy the superconducting properties.  Some contaminates will destroy superconductivity in very tiny amounts. To keep your chances of success high, grinding with a good-quality mortar and pestle is the best method.  This manual grinding can be an arduous process, but the results are worth the trouble.

4. THE FINAL OXYGEN ANNEALING.  The sample should be thoroughly reground, and the resultant black powder placed back in the alumina dish.   The thickness of the layer of loose powder in the dish should match the desired thickness of the final superconducting disk.  For this final firing, the powder should be as finely-ground and as densely-packed as possible.  Do NOT pack the powder into the dish by pressing on it from the top (as this can makes the superconductor tend to stick to the alumina dish).  Better results can usually be obtained by tapping the alumina dish with a pestle or a similar object so that the particles of the mix settle together in an evenly packed disk.

For this final heat treatment, heat the sample to between 950 degrees and 1000 degrees Celsius for about 18 hours.  The higher temperature is better, but be sure of the accuracy of your temperature indicator before getting too close to 1000 degrees.  Temperatures much above 1000 degrees risk decomposition of the crystal structure and the possibility of the material sticking to the alumina dish.  On the other hand, a final oxygen annealing at an indicated temperature of 950 degrees Celsius may yield a superconductor that will crack easily, but will otherwise be satisfactory.

Very Important: It is absolutely necessary that the cool-down take place very slowly and under adequate oxygen flow.  The rate of cooling must be no more than 100 degrees Celsius per hour, especially during the critical temperature region between 750 and 400 degrees Celsius.  Take special care to insure that the sample has access to plenty of oxygen, especially in during the cool-down from 900 to 300 degrees.  Brief interruptions in oxygen flow when the material is above 900 are unimportant, but continuous flow must be maintained during cool-down.  If the atmosphere in the furnace is not oxygen-rich while the sample is still above about 400 degrees, the material can lose vital oxygen from its crystal structure. After the furnace temperature reaches about 500 degrees, the rate of cooling can be increased.  (Often, a good superconductor can be made by using oxygen flow only during the cool-down period.)

During this final heat treatment, before the final cool-down begins, a superconductor that is more resistant to cracking during thermal stress can be produced by subjecting the sample to high-temperature thermal cycling.  To do this, vary the temperature between about 750 and 980 degrees at rates of change of about 200 per hour.  Then raise the temperature to about 980 degrees for an hour or more before beginning the final slow cool-down.  This thermal cycling is not a necessity at all.  In fact, if you started with a powder with fine particles of about the same size, this thermal cycling is not even very useful.  For many manually-ground powders, though, it will add significantly to the mechanical strength of the sample.

Additional hints and notes:


The rate of heating is not nearly as critical for the material as the rate of cooling.  However, it is best not to exceed 300 degrees per hour in order to avoid thermal stress cracks in the labware.  The manufacturers of laboratory alumina recommend not exceeding 150 per hour, but we have not experienced problems with faster heating rates.


The cool-down time during the oxygen annealing is very critical for making these superconductors.  If the rate of cooling is not strictly limited to less than 100 degrees per hour until the sample is below 500 degrees, the sample will probably not superconduct at liquid nitrogen temperatures.  It is especially important to drop the temperature very slowly through the 750 to 500 degree region.  There is evidence that much slower rates of cooling through this temperature range will produce an even better superconductor.

After the sample reaches a the 400 to 500 degree range, the rate of cooling can be increased.  The furnace may be opened and the sample removed from the furnace whenever the temperature falls below 200 degrees, and you believe that you can safely handle the hot sample.


Various types of oxygen supplies may be used as the source of the oxygen flow through the furnace.  One option is to use an acetylene torch setup with the acetylene tank removed.  The oxygen control on the torch provides an additional means of regulating oxygen flow. The tip can be removed from the torch, and plastic tubing clamped over the open end of the torch.  We have also used medical oxygen bottles for the oxygen supply.  The oxygen source is often the greatest problem in making these superconductors in a modestly-equipped laboratory, but there are a large number of options to choose from.  It is difficult to give exact instructions on the oxygen setup, since this will depend upon what type of oxygen source is most easily accessible to each individual user.  Oxygen is necessary only during the cool-down, and only a very slow flow of oxygen is needed.

The superconducting material can gain or lose oxygen easily when the material is above about 400 degrees Celsius.  Therefore, it is especially important to keep the sample in contact with plenty of oxygen during the cool-down.  Otherwise, it may lose much of the oxygen it has acquired. This material is an exceptionally good oxygen ion conductor.  The amount of oxygen that the superconductor absorbs is likely to be the main factor that determines the quality of your superconductor.

In most inexpensive furnaces, it is necessary to provide a rate of oxygen flow to the furnace chamber that actually wastes a considerable amount of oxygen.  But considering the relatively low cost of oxygen, this is usually more economical than the alternatives.  If you have a fairly air-tight furnace, the rate of oxygen flow need not be more than a few milliliters per minute.  Unfortunately, most inexpensive furnaces are not very airtight; and the more easily-available oxygen regulators are not capable of regulating at very low flow-rates.

If it appears that you may run out of oxygen before the firing is completed, turn off the oxygen flow intermittently while the sample is in the still above 900 degrees.  This material will regain any lost oxygen quickly while at high temperature, but adequate oxygen during the cool-down is vital.  If you run out of oxygen during cool-down, the material may not superconduct at liquid nitrogen temperatures.

Unless you have a fairly sophisticated oxygen and furnace system, the best regulator setting for your first attempt at making superconductor is usually about:

  • 1 liter/minute on a medical oxygen tank
  • or about 1 psi using a welding regulator.

These settings will use a considerable amount of oxygen during the oxygen annealings, but it is the least expensive alternative unless you are going to be making more than just a few superconductors.  Regulator settings of one-half of these values will provide more than adequate oxygen flow, but extremely low settings are difficult to maintain on ordinary regulators without constant human attention.

The best method of getting oxygen flow to the interior of your furnace will depend upon what type of furnace you have.  Many laboratory furnaces, such as Thermolyne muffle furnaces, are easily modified for oxygen flow by drilling a hole in the back of the furnace, just below the thermocouple.  If your furnace manual includes instructions for adding an additional thermocouple to your furnace, those instructions can be used as a guide for adding a clay pipestem to the back of the furnace.  A clay pipestem for this purpose can be obtained by disassembling an ordinary pipestem triangle such as those often used for heating crucibles over a flame.

For Thermolyne muffle furnaces such as the model 1300, first remove the metal plate from the rear of the furnace (by removing four screws and carefully sliding out the thermocouple).  Drill a 3/8" (10 mm.) hole in the metal plate about an inch below the hole for the thermocouple.

Re-assemble the furnace, and use the newly-drilled hole as a guide for drilling a 3/8" (10 mm.) hole about an inch into the refractory material.  Be very gentle in drilling into the refractory material, as this material is very soft.  If you're not careful, you can drill all the way into the furnace chamber in a fraction of a second.  You now have a hole that the supplied clay pipestem will slide into.

The next step is to use a 1/8" (3 mm.) drill bit to drill a hole the remaining distance into the furnace chamber.  Oxygen can flow through the hole in the clay pipestem, then through this 1/8" (3 mm.) hole, and into the furnace chamber.

Plastic tubing may be used to connect your oxygen tank to the clay pipestem on the back of your furnace.

If you plan on making a lot of superconductors, you may want to fit your clay pipestem on the inside of the metal back of the furnace.  This requires drilling the 3/8" (10 mm.) hole in the refractory material to a depth equal to the distance from the center of the clay pipestem to one end of the pipestem.

If you are using a pottery kiln for firing the superconductors, you can feed oxygen into the chamber through one of the viewing holes.  One way of doing this is to fit the clay pipestem into the ceramic plug that is used to plug the viewing holes in the kiln.  This can be done by carefully drilling the appropriate size hole in the ceramic plug.  The clay pipestem can be mounted securely in the hole with high-temperature cement such as muffler cement.  If you are doing the superconductor project in a school with an art department that teaches pottery, you may be able to persuade the pottery class to make a custom plug for feeding oxygen into the chamber.


The alumina dish may become stained with black superconductor after a few firings.  It may be restored to like-new condition by soaking it overnight in dilute (10 percent) hydrochloric acid.  (To avoid corrosive hydrochloric acid fumes from oxidizing objects in the room, make sure that even diluted hydrochloric acid solutions are not left uncovered.)  To avoid a situation where the superconductor tends to stick to the alumina dish, do not pack the powder into the dish before firing.  Leave the powder lying loose.  Also monitor the temperature carefully during firing to avoid overheating the material.


In most laboratory preparations of these superconductors, the superconducting powder is pressed into pellets or disks before the final oxygen annealing.  If you have the proper equipment, you may wish to press the superconductors into pellets or disks.  This has definite advantages over just sintering the loose powder.  In order to make pellets of this material that are solid enough to handle until you get to a furnace, add a small amount of distilled water to the powder before pressing, and mix in the distilled water in a mortar and pestle.


The current-carrying capacity, and the ability of these superconductors to levitate even larger magnets, may be increased by various methods.

In general, the more often you repeat the re-grinding and oxygen annealing, the better the superconductor will become.  There appears to be no danger of getting too much oxygen into the superconductors using ordinary equipment.

The use of higher temperatures will improve the quality of these superconductors.  All furnaces and kilns have a temperature gradient inside the chamber.  If the sensor for the temperature indicator is high in the chamber, the actual temperature at the bottom of the chamber may be much cooler.  This is a good reason for using pottery cones for a calibration check.

There are some potential problems when using these higher temperatures, though.  The material will come out of the furnace much harder and more difficult to re-grind.  Therefore, it may be better to wait until the last firing before using these higher temperatures.  After the last firing, a harder disk is desirable.

Higher temperatures induce a risk of melting these materials, especially if your temperature indicator is inaccurate.  (These materials can be processed by melting; but that is a different process than sintering.  Unintentional melting tends to make a mess inside the furnace, and may cause the material to stick to the alumina dish.)  Melting these materials is definitely something to be avoided unless you're deliberately experimenting in this area.

The other problem with higher temperature may be the limitations of your furnace.  Know the specifications of your furnace to avoid burning out the heating elements or doing other damage.  (We have operated lab furnaces rated for maximum temperature of 925 degrees Celsius at temperatures exceeding 1000 degrees Celsius for extended periods and still had the heating elements and thermocouples last for years.  But remember that you exceed the manufacturer's specifications at your own risk.)

Oxygen annealing for longer firing periods will also improve the superconductors.  Intermediate grindings, however, are important since the ultimate goal is to mix the constituent chemicals at the molecular level.

Grinding the material to a very fine powder is especially important just prior to the final annealing.  If the material is very finely ground, tapping the alumina dish with a pestle will reduce the porosity of the material and help to compensate for not pelletizing the material in a powerful press.  It is possible to overdo the grinding, though.  If an electric grinder is used, then the material is pressed into a pellet at high pressure, it can become difficult for oxygen to penetrate into the material.

There are indications that moisture will slowly destroy the superconductivity of this material.  For this reason, some people suggest coating the finished superconductor with a clear spray-on coating or varnish.  However, such a coating makes it impossible to re-fire the superconductor after re-grinding (as you may want to do if it should break).  Therefore, we recommend storing the finished superconductors in a sealed plastic bag with a small bag of desiccant to absorb moisture from the air.


The most foolproof test for superconductivity is the simplest.  This is the test for diamagnetism using small rare earth magnets made of samarium-cobalt or neodymium-iron-boron.  Use a very small rare-earth magnet at first.  Start with a rare-earth disk magnet about 6 mm. in diameter.  If you have made a good-quality superconductor, the magnet will levitate at least 3 mm. above the surface of the superconducting disk.  When cooling the disk with liquid nitrogen, cool the disk slowly by first pouring the cold gas on the disk.  (You can "see" the cold nitrogen gas above the liquid nitrogen because of the moisture it condenses from the air.)  When you begin adding the liquid nitrogen, add only a few drops at first.  Wait at least a minute after you first begin chilling the disk before you cover the entire disk with liquid nitrogen.

If the disk does crack, each piece will still superconduct.  If desired, you can re-grind the broken pieces and fire them in the furnace to put them back into one piece again.  This repair job is identical to the last oxygen annealing.  Be sure to supply plenty of oxygen to the furnace chamber during cool-down, though.  The oxygen supply on a re-fire is just as important as the earlier firing.

If you want to try an even more spectacular demonstration of diamagnetism, you can obtain a larger rare-earth magnet to see if your superconductor will levitate it.  (Some rare-earth magnets may still be available from Futurescience, LLC.)  Neodymium-iron-boron magnets are usually more economical than samarium-cobalt magnets.  The larger magnets provide a good test of the quality of your superconductor. With superconductors made with the techniques described here, we have levitated 1" neodymium bar magnets and 1/2" neodymium disk magnets, sometimes getting a levitation that was barely noticeable, and other times getting a levitation that was quite spectacular.  A superconductor with poor levitation can usually be improved by re-grinding it a giving it an additional oxygen annealing.

When a superconductor levitates a magnet, a magnetic mirror image is formed in the superconductor of the levitating magnet.  The magnetic mirror image insures that there is always a north pole induced in the superconductor directly below the north pole of the levitating magnet.  There is a south pole induced in the superconductor directly below the south pole of the levitating magnet.   This mirror image moves with the magnet as the magnet is moves, so that the disk magnet can be given a rapid spin without affecting its levitation.  In fact the magnet may continue to spin for quite a long time because its spinning encounters no friction other than the friction of air resistance.

The diamagnetism of the superconductor tends to get stronger as the temperature decreases.  One way to decrease the temperature is to decrease the air pressure around the liquid nitrogen cooling the superconductor.  Decreased air pressure reduces the boiling point of liquid nitrogen.  If you place the superconductor and liquid nitrogen in a bell jar and pump some of the air out, you may see the levitating magnet rise higher (although a very cold bell jar tends to accumulate a layer of frost).

ADDITIONAL READING (from the earlier days of High-Temperature Superconductors):

Magazine articles:

"Superconductivity seen above the boiling point of nitrogen," Physics Today, pp. 17-23. April, 1987.

"Superconductors," Business Week, pp. 94-100. April 6, 1987

"Superconductors," TIME, pp. 64-75. May 11, 1987

(various articles), Insight (Washington Times), pp.8-17. May 25, 1987.

"Superconducting", High Technology, pp. 12-18. July, 1987.

"High Tc May Not Need Phonons; Supercurrents Increase," Physics Today, pp. 17-21. July, 1987.

"Do-It-Yourself Superconductors," New Scientist. pp. 36-39. July 30, 1987. (An account of the experiences of the first high school class to make superconductors.)

"Superconductivity: Hype vs. Reality," Discover. pp. 22-32. August, 1987.

"Industry Warms to Superconductivity," New Scientist. pp. 56-61. October 22, 1987.

(Various articles), Journal of Chemical Education. pp. 836-853. October, 1987.

"A Superconductivity Primer," Nature. pp. 21-24. Nov. 5, 1987.

(Also of interest is the March, 1986 issue of Physics Today, which was a special issue on superconductivity written before the discovery of the new oxide superconductors.)

Scientific Papers:

Williams, et al. "High-Tc Superconductors: Selective Preparation and Characterization of Tetragonal and Orthorhombic (93 K Superconductor) Phases of YBa2Cu3Ox ," Inorganic Chemistry, 1987, 26, pp. 1834-36.

Panson, et al. "Effect of Compositional variation and Annealing in Oxygen on Superconducting Properties of Y1Ba2Cu3Ox," Physical Review B, pp. 8774-8777, v.35, no.6, June 1, 1987.

Tarascon, et al. "Oxygen and Rare-Earth Doping of the 90-K Superconducting Perovskite YBa2Cu3Ox ," Physical Review B, pp. 226-234, v.36, no.1, July 1, 1987.

Barnes, R.L. and Laudise, R.A., "Stability of Superconducting YBa2Cu3Ox in the Presence of Water," Applied Physics Letters, pp. 1373-1375, v.57, no.17, October 26, 1987

Return to the Futurescience Home Page.