|
Let’s begin with the well-known thermal fountain effect in superfluid He-4 (He II). Consider a tube whose one side is immersed vertically in He II and the other end sticks above the liquid surface. The bottom end of the tube is packed with fine powder so that the normal component of He II is inhibited from flow in or out of the tube. The superfluid can, however, flow through the small space between the powder grains since it can flow without friction. Such a packed powder is called a superleak. Now the upper side of the powder is suddenly heated by shining light on it. The superfluid component is instantly “attracted” towards the high temperature region and rushes through the “filter” of packed powder. The rushing superflow shoots up through the tube to form spectacular fountain.
|
|
|
Let’s close off the upper end of the tube so that superfluid is prevented from escaping. Instead of just flashing light pulse onto the end of superleak, let’s maintain a constant temperature difference between the He II bath and the liquid inside the tubing. In the steady state when superflow stops, there develops a pressure difference between inside the tubing and the bath. The applied temperature gradient is balanced by the pressure gradient.
|
|
|
An analogous effect occurs in 3He A1 superfluid phase but with applied magnetic field gradient. The superfluid acceleration in A1 phase can be created by gradients in pressure, temperature and magnetic field. Imagine repeating the same experiment as above by substituting temperature gradient with magnetic field gradient. Then we have the magnetic fountain effect. This effect occurs only in A1 phase.
|
|
|
A cutaway exposed schematic of magnetic fountain pressure cell is shown below. It is immersed in A1 phase liquid produced by a static field up to 8 tesla. The parallel channels represent the superleak (10 ~ 50 μm gap). Gradient in magnetic field is applied along the superleak. The induced fountain pressure is sensed by measuring the deflection of a flexible diaphragm shown in red. Four vent holes act to equalize the fluid pressure at the sensor and superleak. A vibrating wire loop viscometer is inserted for a temperature sensor.
|
The upper panel shows the applied field gradient as function of time. The lower panel shows the variation of induced magnetic fountain pressure. Note that the fountain pressure relaxes even when the applied field gradient is constant. See below for the temperature dependence of the relaxation time.
|
|
Unexpected Temperature Dependence of the Fountain Pressure Detector Response
Z in above plot is the measured deflection of the diaphragm. Note that Z > 0 only within A1 phase and that the relaxation time is a strong function of temperature.
|
Relaxation Time Extracted from Response at Left
These spin relaxation phenomena are currently under intense investigation.
|
|
Important conclusion from the analysis of relaxation time: A1 phase is conventionally thought to contain only one (majority) condensate pairs with the pair magnetic moment aligned parallel to the applied magnetic field. The rapid relaxation near Tc2 can be understood by assuming the presence of small amount of (minority) condensate pairs with magnetic moment aligned anti-parallel to the applied field.
|
The presence of minority spin pair condensate was predicted theoretically by H. Monien and L. Tewordt. See “Longitudinal Magnetic Resonance (NMR) in the A1 Phase of Superfluid 3He,” Journal of Low Temperature Physics 60, 323 (1985).
|
|
♫ Coming attraction: mini-spin pump experiment to induce spin polarization by “sucking in” spin-polarized superfluid A1 phase by moving the diaphragm with electrostatic force. Stay tuned.
|
