Yesterday, I wrote about a refrigeration system used to cool a sensor array down to 50 mK. That’s 0.05 degrees above absolute zero. I mentioned that salt was used, and a commenter guessed paramagnetic salt. He was right. The technique is called adiabatic demagnetization refrigeration. The cooling is accomplished using an electromagnet and a hunk of salt.
An Adiabatic Demagnetization Refrigerator (ADR) is a kind of heat pump like the system in your household refrigerator, but unlike your refrigerator, it doesn’t run continuously. An ADR stores the heat that it absorbs and dumps it from time to time into a heat sink. The refrigerator in your kitchen uses the surrounding air (temperature about 300 K) as its heat sink. The ADR in the system that I worked on uses a bath of liquid helium (1.3 K).
The ADR stores heat in a “pill” of a paramagnetic salt. (Paramagnetic means that it is only weakly responsive to magnetism.) When the salt is subjected to an intense magnetic field, the magnetic domains will align with the field. When the field is removed, the domains will naturally tend to return to their random states, but that requires energy. That energy will be drawn from the surrounding material in the form of heat flowing into the salt pill.Image Credit: NASA
Eventually, the salt pill’s magnetic domains will have been fully randomized again, and it will no longer be able to remove heat from its surroundings. At that point, it is thermally reconnected to the heat sink, and heat flows from the pill to the sink. When the pill temperature rises to the that of the heat sink, the heat switch is opened, disconnecting the pill from the sink. The magnetic field is restored, and when the domains are fully realigned, the magnetic field is relaxed to restart the cooling process.
If the magnetic field is reduced quickly, the temperature of the pill will drop quickly. While testing the control system that our team worked on with a prototype ADR, I was able to achieve temperatures of about 20 mK by quickly forcing the magnet current to zero. By properly controlling the rate of discharge of the magnet, it is possible for the system in the ASTRO-H SXS to hold 50 mK (±1 µK) for about 100 hours. A great deal of the credit for that level of performance goes to the guys who designed and implemented the control algorithm. My part of the system was the analog electronics that measure the temperatures of the ADR and surrounding objects and provide the drive current to the ADR magnet.
The ADR magnet was an unusual device to power. The warmest temperature it could possibly see during operation is less than 10 K. That means the wire in the magnet is a superconductor with no electrical resistance. The ADR and sensor array are housed a dewar (a fancy name for a thermos bottle). The only electrical resistance in the system is in the cable between the control circuits and the outer layers of the dewar.
The SXS instrument and the ASTRO-H satellite are now in the integration and test phase. The schedule slipped somewhat because of earthquake damage to some of the facilities in Japan that are supporting the mission. Launch is now scheduled for 2015, so it will be a couple of years before I’ll be able to publish any #StarPorn from the SXS.