Periodic Reporting for period 1 - NewCooler (Continuous sub-mK refrigeration by demagnetization of atomic nuclei)
Reporting period: 2023-03-01 to 2024-08-31
A substance in thermal equilibrium with the pervasive cosmic microwave background would reach a temperature of 2.73 kelvin. However, scientists can far surpass this base temperature of the universe. Starting in the 19th century, liquefaction of “permanent” gases was demonstrated, culminating with the liquefaction of helium under atmospheric pressure at 4.2 kelvin in 1908. Once it was produced in sufficient quantities, the temperature of liquid helium could be further reduced by pumping its vapor as it boiled, yielding temperatures of order 1 kelvin. The reduction of thermal agitation at such low temperatures yielded the discovery of startling new phenomena including the flow of electrons without resistance (superconductivity) and the flow of liquid helium without viscosity (superfluidity).
Since the vapor pressure of liquid helium decreases precipitously as its temperature is decreased below 1 kelvin, pumping becomes an impractical cooling technique and different methods are required to achieve temperatures in the millikelvin range and below. Access to temperatures near 10 mK was greatly simplified with the appearance of cryogen-free (“dry”) dilution refrigerators. These cryostats do not use a liquid helium bath for pre-cooling, so no external helium supply is required. Dry refrigerators typically have much more experimental space and are easier to automate than their conventional counterparts. They have facilitated rapid advances in fields such as superconducting quantum technology, astrophysics and materials science, and have led to the growth of the cryogenics industry around the world.
Starting with the discovery of superfluid 3He in the 1970s, cooling condensed matter well below 10 mK has been productive, yielding fascinating discoveries up to the present day. Areas of focus include nanomechanical resonators, low dimensional electron systems, amorphous solids, electronic transport in nanostructures and dark matter searches.
Sub-mK temperatures in condensed matter are obtained using adiabatic nuclear demagnetization. The refrigerant of a nuclear demagnetization refrigerator (NDR) is a metal containing nuclei with non-zero spin. First, a magnetic field is applied to the refrigerant while it is in contact with a precooling dilution refrigerator at ~10 mK, causing the spins to align with the magnetic field. Then the refrigerant is thermally isolated from the precooler and the applied field is decreased. The thermal isolation implies that the degree of alignment of the spins remains constant. Thus the temperature of the spins ideally decreases in proportion to the field. Once the desired temperature is achieved, heat leaks are balanced by further demagnetization.
Until now, NDR have had to return to the precooler temperature for recycling when the magnetic field reaches zero. The autonomy of such NDR is particularly limited when pre-cooled by dry dilution refrigerators, partly because of their vibrations. Thus researchers working at sub-mK temperatures have not fully benefited from cryogen-free technologies.
Since the vapor pressure of liquid helium decreases precipitously as its temperature is decreased below 1 kelvin, pumping becomes an impractical cooling technique and different methods are required to achieve temperatures in the millikelvin range and below. Access to temperatures near 10 mK was greatly simplified with the appearance of cryogen-free (“dry”) dilution refrigerators. These cryostats do not use a liquid helium bath for pre-cooling, so no external helium supply is required. Dry refrigerators typically have much more experimental space and are easier to automate than their conventional counterparts. They have facilitated rapid advances in fields such as superconducting quantum technology, astrophysics and materials science, and have led to the growth of the cryogenics industry around the world.
Starting with the discovery of superfluid 3He in the 1970s, cooling condensed matter well below 10 mK has been productive, yielding fascinating discoveries up to the present day. Areas of focus include nanomechanical resonators, low dimensional electron systems, amorphous solids, electronic transport in nanostructures and dark matter searches.
Sub-mK temperatures in condensed matter are obtained using adiabatic nuclear demagnetization. The refrigerant of a nuclear demagnetization refrigerator (NDR) is a metal containing nuclei with non-zero spin. First, a magnetic field is applied to the refrigerant while it is in contact with a precooling dilution refrigerator at ~10 mK, causing the spins to align with the magnetic field. Then the refrigerant is thermally isolated from the precooler and the applied field is decreased. The thermal isolation implies that the degree of alignment of the spins remains constant. Thus the temperature of the spins ideally decreases in proportion to the field. Once the desired temperature is achieved, heat leaks are balanced by further demagnetization.
Until now, NDR have had to return to the precooler temperature for recycling when the magnetic field reaches zero. The autonomy of such NDR is particularly limited when pre-cooled by dry dilution refrigerators, partly because of their vibrations. Thus researchers working at sub-mK temperatures have not fully benefited from cryogen-free technologies.
The recently proposed continuous nuclear demagnetization refrigerators (CNDR), which are based on multiple nuclear demagnetization stages, would allow ultra-low temperature researchers to take full advantage of cryogen-free technology, making it much easier to reach T<1 mK. Our calculations implied that we could construct a CNDR with a cooling power of tens of nW at 1 mK [D. Schmoranzer et al., Cryogenics 110, 103119 (2020)]. However, building a CNDR is challenging, in part, because the two nuclear stages must be linked by superconducting heat switches that have very low thermal resistance when closed. This thermal resistance limits the rate at which the CNDR can be cycled and consequently limits its cooling power. We had demonstrated a heat switch whose thermal resistance when closed is five times better than the previous state of the art and, unlike the latter, does not rely on a cyanide-based plating process [J. Butterworth et al., Review of Scientific Instruments (2022)]. An essential fabrication step was the efficient removal of the native oxide of the aluminum superconducting element followed by gold deposition without breaking vacuum.
The ERC PoC project NewCooler was dedicated to building on these achievements to construct the first CNDR. Our successful use of aluminum in our heat switch motivated us to experiment with aluminum nuclear refrigerant. This material is more abundant and workable than the brittle rare earth-based conventional refrigerant PrNi5. Furthermore, the minimum temperature of a few microkelvin that can be achieved with Al is far below that of PrNi5, which is limited to 0.4 mK due to nuclear magnetic ordering. The first step was to demonstrate that an aluminum nuclear demagnetization refrigerator in "single-shot" mode would meet our design criteria for the CNDR. With the support of the ERC PoC project NewCooler, we achieved this and reported the results in M. Raba et al., Physical Review Applied 22, 024027 (2024). The results included measurements of the minimum temperature, heat leak and thermal time constant of the refrigerator. We confirmed that our design is effective, efficient and sustainable, avoiding toxic substance such as cadmium solder used in previous nuclear refrigerators.
Following the demonstration of the "single-shot" refrigerator, we focused on the design and construction of the continuous system, which is presently being manufactured.
The ERC PoC project NewCooler was dedicated to building on these achievements to construct the first CNDR. Our successful use of aluminum in our heat switch motivated us to experiment with aluminum nuclear refrigerant. This material is more abundant and workable than the brittle rare earth-based conventional refrigerant PrNi5. Furthermore, the minimum temperature of a few microkelvin that can be achieved with Al is far below that of PrNi5, which is limited to 0.4 mK due to nuclear magnetic ordering. The first step was to demonstrate that an aluminum nuclear demagnetization refrigerator in "single-shot" mode would meet our design criteria for the CNDR. With the support of the ERC PoC project NewCooler, we achieved this and reported the results in M. Raba et al., Physical Review Applied 22, 024027 (2024). The results included measurements of the minimum temperature, heat leak and thermal time constant of the refrigerator. We confirmed that our design is effective, efficient and sustainable, avoiding toxic substance such as cadmium solder used in previous nuclear refrigerators.
Following the demonstration of the "single-shot" refrigerator, we focused on the design and construction of the continuous system, which is presently being manufactured.
We recently reported the performance of a new aluminum nuclear demagnetization refrigerator, whose design simultaneously maximizes the thermal conductance along the nuclear stage and minimizes eddy current heating. It is optimized for powerful continuous cooling when part of the future dual stage NDR, which will maintain temperatures below 1 mK indefinitely. Its high thermal conductance yields rapid thermalization, so that the system can be rapidly cycled, maximizing its cooling power. Furthermore, its low susceptibility to eddy currents yields a low parasitic heat load. We expect this innovation to broaden the field of microkelvin physics, accelerating the rate of discovery and increasing its technological potential.
We have taken the necessary steps to protect the intellectual property generated by this project. Future work will include demonstration of a prototype CNDR and commercialisation.
We have taken the necessary steps to protect the intellectual property generated by this project. Future work will include demonstration of a prototype CNDR and commercialisation.