Realization of the 3He Vapor-Pressure Temperature Scale and Development of a Liquid-He-Free Calibration Apparatus

The ³He vapor-pressure temperature scale was realized using an apparatus based on a continuously operating ³He cryostat at the National Metrology Institute of Japan (NMIJ), National Institute of Advanced Industrial Science and Technology (AIST). The cryostat has two operational modes: a ³He circulation mode and a 1 K pot mode. The ³He circulation mode can be used for ³He vapor-pressure measurements below 1.6 K, and the 1 K pot mode can be used for measurements above 1.3 K. Either mode can be selected for measurements from 1.3 K to 1.6 K. The realization of the ³He vapor-pressure temperature scale in this study fully covers its defined temperature range from 0.65 K to 3.2 K in the International Temperature Scale of 1990. The latest realization results are presented in this article. In addition, a liquid-He-free calibration apparatus was developed. It does not require liquid helium as a cryogen, which usually entails cumbersome handling and periodic refilling. The apparatus was designed for the calibration of capsule-type resistance thermometers from 0.65 K to 24.5561 K (the triple point of neon). The cooling system of the apparatus consists of a commercially available pulse-tube refrigerator and a ³He Joule?CThomson (JT) cooling circuit developed at NMIJ/AIST. The pulse-tube refrigerator is used in a pre-cooling stage and cools the apparatus to approximately 5 K. The ³He JT cooling circuit is used to cool the apparatus from 5 K to below 0.65 K. Since the ³He JT cooling circuit is a closed circuit, the apparatus can run continuously with only simple maintenance required. The basic characteristics of the apparatus are described.     

Balloon-borne He cryostat for millimetre bolometric photometry

A cryostat including a ³He refrigerator has been designed, built, coupled to a multiband far infrared (FIR) photometer and successfully operated on a balloon flight. Here the cryostat, which is able to cool three bolometers to 0.3 K and one to 1.5 K, is described. The total holding time of the cryostat exceeds 5 days, being limited by exhaustion of the main ⁴He and N₂ baths. The autonomy of the ³He cryostat is more than 50 h, and the recycling time is ≈3 h. The performance of the system in the ARGO 1993 flight is also reported.     

Measurements of the Superfluid Joule–Thomson Refrigerator Using High Concentration 3He–4He Mixtures

The superfluid Joule–Thomson refrigerator (SJTR) uses a liquid superfluid ³He–⁴He mixture to provide cooling below 1 K. Performance measurements of the SJTR using 5% and 11% ³He concentration mixtures are reported. High concentration operation shows higher cooling powers at high temperature. Ultimate temperatures are seen to increase with increasing concentration due to a pinching of the temperature defect in the recuperative heat exchanger. This pinching effect is due to the variation of the heat capacity of the ³He–⁴He mixture with temperature and concentration and is discussed in detail and design changes are suggested to mitigate it.     

Experimental study of staging method for two-stage pulse tube refrigerators for liquid He temperatures

The first two-stage pulse tube refrigerator, providing a lowest temperature of 2.23 K and a cooling power of 370 mW at 4.2 K, employed a parallel arrangement of the two pulse tubes with phase shifters located at room temperature¹. With the aim of increasing the COP at liquid ⁴He temperatures, three modified staging methods were tested in this paper. All refrigerator versions operate with the same two regenerators as already used in the first two-stage setup¹ and also the same 6 kW He-compressor combined with a redesigned G-M rotary valve. The best performance is achieved with a parallel arrangement two-stage refrigerator by introducing proper negative DC flow and impedance tubes. So far the highest cooling power achieved on the second stage at 4.2 K was 0.5 W. With a heat load of 20 W at 67 K on the first stage, the second stage can provide a cooling power of 0.42 W at 4.2 K. Details of the design of the different refrigerators and a comparison of their performance are presented.     

A small helium-3 pulse-tube refrigerator

A new three-stage pulse-tube refrigerator (PTR) is developed by scaling down a previous PTR by 50%. The new system is small in size and weight, capable of operating using little input power, and uses a small amount of working gas and regenerator material. In addition to that the system is flexible and convenient for modifications. The volume of the low-temperature part of the new PTR (pulse tubes + regenerator) is as small as 0.28 l. With ³He as a working fluid a no-load temperature of 1.73 K is reached and a cooling power of 124 mW at 4.2 K is realized.     

A multipurpose He refrigerator

We introduce a mini ³He refrigerator, operating at ∼300 mK starting from 4.2 K without pumping on the main ⁴He bath. The innovative idea is that the present one is suitable for a very fast operation; for its use, it is sufficient a storage ⁴He Dewar. In this way we drastically reduce the time required to cool it down, because there is no need for a classic cryostat. This prototype is particularly aimed for all those operations in which it is necessary to test a large number of samples that do not require long duration measurements at low temperature.     

Condensation stage of a pulse tube pre-cooled dilution refrigerator

In our article, experiments with a pulse tube (PTR) pre-cooled dilution refrigerator (DR) are presented, where an upgraded ³He condensation stage has been tested. The DR had a ³He flow rate of up to 1.1 mmol/s. The ³He gas entering the refrigerator was first pre-cooled to a temperature of ～50 K at the first stage of the PTR. In the next cooling step, the ³He was run through a recently installed heat exchanger, which was attached to the regenerator of the second stage of the pulse tube cryocooler; at the outlet of this heat exchanger the temperature of the ³He was as low as ～4 K. Due to the non-ideality of the helium gas, the second regenerator of a two stage PTR has excess cooling power which can be made use of without affecting the base temperature of this stage, and it is this effect which was put to work, here. Finally, the ³He was further cooled in a heat exchanger, mounted at the second stage of the PTR, before it entered the dilution unit of the cryostat.The installation of a heat exchanger at the regenerator of the second stage of the PTR is especially important for the construction of DRs with high refrigeration capacities; in addition, it allows for a plain design of the subsequent Joule–Thomson (JT) stage, and herewith facilitates considerably the construction of “dry” DRs. The condensation rate of the ^3,4He mash prior to an experiment was increased. The pressure during condensation could be kept near 1 bar, and thus a compressor was no longer necessary with the modified apparatus.     

Development of a thermodynamic model for a cold cycle 3He-4He dilution refrigerator

A thermodynamic model of a ³He-⁴He cold cycle dilution refrigerator with no actively-driven mechanical components is developed and investigated. The refrigerator employs a reversible superfluid magnetic pump, passive check valves, a phase separation chamber, and a series of recuperative heat exchangers to continuously circulate ³He-⁴He and maintain a ³He concentration gradient across the mixing chamber. The model predicts cooling power and mixing chamber temperature for a range of design and operating parameters, allowing an evaluation of feasibility for potential ³He-⁴He cold cycle dilution refrigerator prototype designs. Model simulations for a prototype refrigerator design are presented.     

Phase separation in solid mixtures of helium-3 and helium-4

Phase separation temperatures have been determined in bcc³He-⁴He mixtures as a function of³He concentration and melting pressure from measurements of changes in the X-ray lattice parameter and Bragg peak shape. A new rigid tail dilution refrigerator cryostat was used to study³He-⁴He crystals with³He concentrations of 0.10, 0.20, 0.30, 0.45, 0.60, and 0.70 and melting pressures between 3.0 and 4.3 MPa. The phase separation temperatures determined are in good agreement with regular solution theory and give little support for an asymmetry in the coexistence curve expected from a Nosanow-type model and reported from previous experiments using other signatures of phase separation. At a given concentration, differences in phase separation temperatures determined from slow cooling and warming data, respectively, are as much as 25 mdeg, but this is less than half the differences reported from previous experiments. A bcc-hcp transformation was seen in a crystal with 10%³He at aboutT=0.3 K for a melting pressure of3.7 MPa.     

Closed-Cycle Joule–Thomson Cryocooler for Resistance Thermometer Calibration down to 0.65K

A closed-cycle Joule–Thomson cryocooler for resistance thermometer calibration has been developed. It consists of a Gifford–McMahon mechanical refrigerator and a closed-cycle ³He Joule–Thomson expansion circuit that utilizes the isenthalpic expansion of ³He for cooling. The developed cryocooler can reach temperatures as low as 0.6K and can operate for months with a simple procedure. The typical cooling power of the cryocooler is 1mW at 0.65K with a molar flow rate of 160μmol ·s⁻¹ through the ³He Joule–Thomson circuit. The possible mechanical vibration level experienced by the resistance thermometers was measured with a laser vibrometer. It was confirmed that the maximum acceleration level is 0.1m· s⁻² and will not cause a problem for thermometer calibration.     

A He pulse tube cooler operating down to 1.3 K

Regenerative cryocoolers that employ ⁴He as working fluid can only reach a lowest temperature of about 2 K. This limitation can be overcome by the use of ³He as working fluid. Here we report on the performance of a two-stage pulse tube cooler that consists of two parallel stages with independent gas circuits. The pressure oscillation in each stage is generated by means of a separate compressor in combination with a rotary valve. With ⁴He in both stages, the minimum no-load temperature of the 2nd stage was 2.23 K, and cooling powers of 50 W at 53 K and 380 mW at 4.2 K were simultaneously available at electrical input powers of 4.54 and 1.45 kW to the 1st and 2nd stage, respectively. Using ³He as working fluid in the 2nd stage, a minimum stationary temperature of 1.27 K has been achieved, which is, up to now, the lowest temperature obtained by regenerative cryocoolers. At an electrical input power of 1.3 kW, the 2nd stage provides a cooling power of 42 mW at 2.0 K and 518 mW at 4.2 K. With ³He, at the same operating condition, the cooling power at 4.2 K was found to be larger than with ⁴He.     

Superfluid Stirling-cycle refrigeration below 1 kelvin

A new method for cooling below 1 K, the superfluid Stirling cycle, uses the gaslike thermodynamic properties of the³He solute in a superfluid³He-⁴He solution. The first prototype superfluid Stirling-cycle refrigerator cools to 0.6 K from a starting temperature of 1.2 K, with cooling powers at the lowest temperatures of a few tens of microwatts. The cycle works in both classical-gas and Fermi-gas regimes.     

Development of a low-temperature He compressor for superfluid He-He mixtures

The development, testing and modeling of a “compressor” that is capable of increasing the concentration of the ³He component of a liquid superfluid ³He-⁴He mixture is discussed. This compressor was developed to drive refrigeration cycles for cooling below 1 K. The compressor design and performance testing is described in detail. The compressor was operated at 1.2 K and ³He molar flow rates of 130 μmol/s were achieved. Compression ratios in excess of 6 were also demonstrated. The theoretical models presented are used to estimate the expected efficiencies of the compressor as well as the effect of the ⁴He component on the power required to drive the compressor.     

Uniform temperature cooling power of the superfluid Stirling refrigerator

Uniform temperature cooling power measurements of a superfluid Stirling refrigerator are presented for³He-⁴He molar concentrations of 5.9%, 17% and 36% and for temperatures between 0.37 K and 1.4 K. The results are compared to an ideal Fermi gas model and to a more general thermodynamic model. The Fermi model agrees well with the 5.9% concentration data; however, the more elaborate model is needed for higher concentration mixtures.     

Development of a 3He Refrigerator for Possible Experiments of Solid 4He on a Small Jet Plane

The gravity on the Earth (g _E) has not been taken seriously to mask the fundamental phenomena on quantum solids, though there are some important studies on the critical phenomena of superfluid ⁴He under microgravity. We are planning to investigate the effect of gravity on the equilibrium shape of solid ⁴He. Since we had a chance to do such an experiment on a small jet plane through the ground based program by JAXA, we got started to construct a cryostat which could cool down as low as 500?mK and meet severe restrictions of experiments on a jet plane. The main part of the refrigerator was a usual ³He-evaporator pumped by a scroll pump. A?small GM refrigerator was installed to provide 4?K stage. 1?K pot was also put in which was also pumped by another scroll pump to condense ³He gas and sample ⁴He. The cryostat was designed to have two optical windows to be able to observe solid ⁴He under microgravity. In the test flight for the refrigerator, the minimum temperature of 690?mK was kept during the entire flight of two hours in which 7 to 8 times parabolic flight was performed. Each parabolic flight includes about 20?seconds microgravity and 20?seconds 1.5 to 2.0?g _E period before and after the microgravity. We did some preliminary experiments with bcc solid ⁴He under microgravity. The crystal remained stuck to the bottom of the sample cell even in the 20 seconds microgravity condition.     

Theoretical Models for the Cooling Power and Base Temperature of Dilution Refrigerators

³He/⁴He dilution refrigerators are widely used for applications requiring continuous cooling at temperatures below approximately 300 mK. Despite of the popularity of these devices in low temperature physics, the thermodynamic relations underlying the cooling mechanism of ³He/⁴He refrigerators are very often incorrectly used. Several thermodynamic models of dilution refrigeration have been published in the past, sometimes contradicting each other. These models are reviewed and compared with each other over a range of different ³He flow rates. In addition, a new numerical method for the calculation of a dilution refrigerator’s cooling power at arbitrary flow rates is presented. This method has been developed at CERN’s Central Cryogenic Laboratory. It can be extended to include many effects that cannot easily be accounted for by any of the other models, including the degradation of heat exchanger performance due to the limited number of step heat exchanger elements, which can be considerable for some designs. Finally, the limitations of applying the results obtained with idealized thermodynamic models to actual dilution refrigeration systems are discussed.     

Characterization of White LEDs at Cryogenic (4 K) Temperatures

In order to capture sample preparation images inside a ³He cryostat via a fibre-optics bundle, a low-temperature white light source was required. High-intensity white LEDs were thus characterized by studying I(V) curves and extracting forward voltage drops and operating temperatures at 77?K, 40?K and ～6?K. Four models were identified as fully functional down to 6?K and images of in-situ sample cleaving in a non-optical ³He refrigerator were obtained.     

Proposed designs for a “dry” dilution refrigerator with a 1 K condenser

Recent development of “dry” dilution refrigerators has used mechanical cryocoolers and Joule-Thomson expansion stages to cool and liquefy the circulating ³He. While this approach has been highly successful, we propose three alternative designs that use independently-cooled condensers. In the first, the circulating helium is precooled by a mechanical cooler, and liquified by self-contained ⁴He sorption coolers. In the second, the helium is liquefied by a closed-cycle, continuous flow ⁴He refrigerator operating from a room temperature pump. Finally, the third scheme uses a separate ⁴He Joule-Thomson stage to cool the ³He condenser. The condensers in all these schemes are analogous to the “1-K pot” in a conventional dilution refrigerator. Such an approach would be advantageous in certain applications, such as instrumentation for astronomy and particle physics experiment, where a thermal stage at approximately 1 K would allow an alternative heat sink to the still for electronics and radiation shielding, or quantum computer research where a large number of coaxial cables must be heat sunk in the cryostat. Furthermore, the behaviour of such a refrigerator is simplified due to the separation of the condenser stage from the dilution circuit, removing the complex interaction between the 4-K, Joule-Thomson, still and mixing chamber stages found in current dry DR designs.     

3He/4He dilution refrigerator with high cooling capacity and direct pulse tube pre-cooling

In the article, a ³He/⁴He dilution refrigerator (DR) is described which is pre-cooled by a commercial two-stage pulse tube refrigerator (PTR); cryo-liquids are not necessary with this type of milli-kelvin refrigerator. The simple design of the condensation stage of this so-called dry DR is novel and explained in detail. In most dry DRs the circulating ³He gas is cooled by a two-stage PTR to a temperature of about 4 K. In the next cooling step, the ³He flow is cooled and partially liquefied in a Joule–Thomson circuit, before it is run to the dilution refrigeration unit. The counterflow heat exchanger of the Joule–Thomson circuit is cooled by the cold ³He gas pumped from the still of the DR. In the DR described here, the heat exchanger of the Joule–Thomson stage was omitted entirely; in the present design, the ³He gas is cooled by the PTR in three different heat exchangers, with the first one mounted on the first stage of the PTR, the second one on the regenerator of the second stage, and the third one on the cold end of the second stage. The heat load caused by the ³He flow is mostly absorbed by the first two heat exchangers. Thus the ³He flow presents only a small heat load to the second stage of the PTR, which therefore operates close to its base temperature of 2.5 K at all times. A pre-cooling temperature of 2.5 K of the ³He flow is sufficiently low to run a DR without further pre-cooling. The simplified condensation system allows for a shorter, compacter and more economical design of the DR. Additionally, the pumping speed of the turbo pump is no longer obstructed by the counterflow heat exchanger of the Joule Thomson stage as in our earlier DR design.