With this PhD thesis project, we propose to significantly advance the nuclear refrigerator (NR) technique in
at least two ways: First, in a new generation of network NR, improve material heat leaks (sample holder),
microwave filtering and NR design to facilitate cooling of nanosamples below 1 mK, learning from the
already running experiment. Also, thermometry needs to be improved, for which we propose to use a
SQUID based Johnson-noise thermometer. Then, with a CBT, GaAs quantum dots or other thermometer,
ultra-low temperatures need to be demonstrated. Finally, when sufficiently low temperatures have been
reached, the nuclear-spin phase transition can be investigated in low-density high mobility 2D electron
gases (available to us from Loren Pfeiffer, Princeton) using a quantum point contact as Overhauser field
detector. Sample nanofabrication can be done in the SNI clean-rooms in-house. These experiments will
be done in close collaboration with the theory group of Daniel Loss.

Second, in a technological advance, we propose to use a cryogen-free dilution refrigerator (CFDR)
for precooling the NRs. We note that so far, no nuclear cooling whatsoever has been demonstrated
on a cryogen-free system. To function efficiently, the CFDR is required to run well below 10 mK,
while keeping vibration amplitudes to a very low level. This is a very challenging task, and several
stages of vibration damping and decoupling are being investigated in order to reduce the intrinsic coldhead,
valve and motor vibrations of the pulsed tube system (in collaboration with BlueFors). The
ultimate aim of such a system is to make microkelvin temperatures available without requiring a steady
supply of (liquid) Helium – a limited, non-renewable resource – thus significantly reducing the cost of
operation and tremendously increasing the space available for experiments at low temperatures. Further,
the system needs only electrical power and cooling-water, potentially making microkelvin temperatures
widely available to almost any lab in the world. This gives the potential for commercialization of NR
systems.