Cryogen-free low temperature and high magnetic field apparatus
S D Kaushik, Anil K Singh, D Srikala & S Patnaik
School of Physical Sciences, Jawaharlal Nehru University, New Delhi 110 067 Received 12 December 2007; accepted 15 January 2008
The importance of low temperature and high magnetic field measurements in pure and applied science research cannot be overstated. Traditionally these experiments have been carried out by evaporation of liquefied helium. This is a costly proposition, especially in our country, where maintaining liquid helium plants and the recovery lines has become persistent predicament. In this paper, the possibility of an alternative cost-effective technology based on two stage Gifford-McMahon closed cycle cryocoolers which is most ideally suited for small groups of researchers has been presented. The principle of operation and instrumentation details of a recently installed single compressor 1.6 K, 8 tesla cryocooler has been described.
Keywords: Liquid helium, Gifford-McMahon cycle, Regenerator, Cryocooler
1 Introduction
Temperature and magnetic field are the most crucial variables for the entire spectrum of pure science research. Particularly in material science and condensed matter physics, all characterization techniques are primarily centered on these facilities.
Studies over the widest range of temperature and magnetic field therefore, hold the key to discovering new science and functional materials. While measurements above the room temperature can be attained by controlled heating, most low temperature experiments are carried out using a variety of gases, prominently nitrogen and helium, in their liquid form.
The standard method is first to liquefy the gas by mechanical means, then to transfer it to the measurement set-up, and then to vapourize it over a small volume. The latent heat of vaporization during this process is obtained from the liquid itself leading to further lowering of temperature of the liquid phase of the gas. The time honored procedure involves liquid helium (He24
), that remains a gas at atmospheric pressure down to 4.2 K. By evaporating helium at low pressure, temperature down to 1.5 K can be easily achieved. For generating high magnetic field too, liquid helium is indispensable. Almost all high field magnets routinely used in research laboratories and industry are made up of Nb based superconductors (e.g. Nb-Ti, Nb3Sn). These materials need cooling down to 4.2 K to achieve superconducting state and sustain high currents in solenoidal geometry for up to 18 tesla field.
Therefore, the production, supply and storage of liquid helium are the most critical infrastructure requirement for a quality condensed matter program.
Current situation however has been discouraging on two counts. For over ten years now American Physical Society has been predicting severe shortage of helium gas by early 21st century. The cost of liquefier (with reasonable throughput and longevity) and its maintenance have skyrocketed in the recent past making it next to impossible for small groups to conduct low temperature measurements. Fortunately, with the advent of high wattage closed cycle refrigerators, the dependence on costly liquid helium to carry out quality low temperature experiments is a thing of the past. In this paper, we describe the principle of operation and successful installation of an entirely liquid cryogen-free low temperature high magnetic facility at the School of Physical Sciences, Jawaharlal Nehru University, New Delhi.
2 Principle of Closed Cycle Cryocoolers
The first patent for industrial production of liquid nitrogen by liquefaction of atmospheric gas was filed by Carl Linde (dated July 9, 1895). The invention was based on Joule-Thomson effect which states that any real gas undergoes a change in temperature while traversing from a high pressure region to low pressure region under isoenthalpic condition. The change in temperature, whether increase or decrease, is determined by change in internal energy of the gas when the average separation between the molecules of the gas increases under expansion2. Almost 65 years later, Kohler and Jonkers at the Philips company, were the first to implement a closed cycle based air liquefier, formally known as the Philips-Sterling cycle. The cooling was achieved by free expansion of gas rather than the throttling process of J-T effect.
This was based on compression of gas followed by transfer through a regenerator onto a space where the gas is expanded and cooled before returning to uncompressed state. The regenerator acts as a heat exchanger as well as a thermal seal between the warm and cold ends. The ‘Kohler’ design involved out of phase movement of two pistons with helium gas as the working substance3. The Gifford-McMahon (GM) cycle4,5, on which most of the present day sub-4 K cryogen-free refrigerating systems are based, is a straight forward modification of the Philips-Sterling method.
Single stage GM cryocoolers (~30 K), have been widely used as cryopumps and as radiation shield in MRI machines for decades. The schematic diagram of the GM cycle is shown in Fig. 1. The main parts of the GM apparatus are, a) compressor CP, b) displacer piston D, c) regenerator R, d) intake valve VI, and e) exhaust valve VE. The working material could be pure helium gas. The displacer is cased inside a cylinder
and its basic function is to displace a volume of gas through the thermal regenerator R. The regenerator maintains a large thermal gradient between the cold and hot ends of the cylinder and it is made up of materials with large molar heat capacity. The displacer is tight fitted to the cylinder through sliding seals that prevent gas flow through the radial space between the displacer and the cylinder. Effectively, the cold (C) and warm (W) volumes can be varied by the movement of the displacer but the total volume remains constant throughout the cycle. The opening and closing of the inlet and outlet valves are synchronized with the position of the displacer in the cylinder through a rotary drive mechanism. The four distinct steps of the GM cycles as shown in Figure 1 are:
(i) With the displacer at the cold end the intake valve VI is opened and the compressed helium gas fills the volume W; (ii) the displacer is moved to the warm end with VI open. In order to keep the pressure
(i) VI
VE
Load
W R
D
CP
VE VI
C
R D
(ii)
Load CP
R
VI
VE
C D
(iii)
Load CP
VE W
R
VI
D
(iv)
CP
Load
Fig. 1 — Schematic description of 4 stages of GM cycle, (i) intake valve is opened with exhaust valve closed. The displacer is at the cold end. High pressure helium gas from the compressor fills the warm end volume W, (ii) displacer is moved to the warm end with intake valve open and exhaust valve closed Helium gas is forced to move towards cold end through regenerator, (iii) intake valve is closed and exhaust valve is opened. This allows helium gas at cold end to undergo expansion that leads to cooling, (iv) displacer is brought back to cold end and the exhaust valve is closed
constant gas slowly flows into the cold volume C through the regenerator R; (iii) the intake valve is then closed and the exhaust valve VE is opened forcing the compressed gas at the chamber C to undergo expansion and consequent cooling; (iv) the displacer is moved towards the cold end to drive the remaining gas in the volume C. The exhaust valve is closed and the cycle is repeated.
The performance of the GM cryocooler depends to a great extent on the effectiveness of the regenerator material. In conventional two stage GM cryocoolers, usually lead (Pb) is used as regenerator and temperature down to 6 K can be achieved. However, the heat capacity of lead is negligible as compared to pressurized helium below 6 K and therefore, it is impossible to reach sub liquid helium temperatures with lead as the regenerator. Using rare earth alloy GdxEr1−xRh, Yoshimura et al.6 were the first group to reach 3.3 K by using a 3 stage GM cryocooler6. These compounds undergo a magnetic ordering transition below 20 K. Here the displacer has three stages, which means that it uses three different materials as regenerators depending on the temperature of the stage. In the process, Yoshimura et al.6 were able to produce liquid helium at the rate of 10 cc/h by condensing helium gas near the cold head. However, high price of the Rh based materials prevented large scale commercialization of this technique. Recent development in regenerator technology has enabled the use of Er3Ni/Er0.9Yb0.1Ni hybrid regenerator which is cost effective and can provide cooling power7 up to 1.5 W at 4.2 K. Now even sub 2 K cryocoolers are available using a variety of magnetic resonators. A typical design as described by Numazawa et al.6 uses spherical Pb particles in cupper mess as the first stage regenerator and a mixture of Pb, HoCu2, Gd2O2S, GdAlO3, and GdVO4 as the second stage regenerator8. The industry leader in this technology is the Sumitomo Heavy Industry (SHI) of Japan. Most of the cryogen-free high magnetic field systems built by Janis, Cryogenics, CryoIndustries, Cryomagnet, and Cryomech are based upon SHI compressors and displacers.
3 Cryogen-Free System at JNU
The low temperature high magnetic field facility installed at JNU can cool down to 1.6 K in temperature and can generate up to 8 tesla magnetic field without any use of liquid cryogen. With separate attachments the sample space can be cooled to 300 mK (helium 3) and heated up to 1000 K (encapsulated
oven). More advanced models with hybrid magnets can produce upto 18 tesla field. These models require two compressors, one for cooling the magnet and the other for cooling the sample space.
The system that is installed at JNU is a single compressor open-ended system purchased from M/s Cryogenics of UK. It involves a SHI compressor model CSW 71 in conjunction with a two stage displacer model SRDK 408D. The compressor is water cooled and requires continuous supply of chilled water (~15°C) at the rate 7 lit/min. The 3 phase power requirement is ~ 9 kW. The cooling powers at the first and second stage cold heads are 34 W @ 40 K and 1 W @ 4.2 K, respectively. The second stage cooling is shared between the sample space and the Nb-Ti superconducting magnet (Tc = 9 K). The first stage is connected to an ultra pure aluminium radiation shield. The magnet is latched to the second stage of the compressor. A condensation pot is attached to the magnet. When low pressure helium gas from a separate close cycle reservoir is brought in contact with condensation pot the gas liquefies and with suitable pumping, the base temperature of 1.6 K can be achieved. It is to be emphasized that the process is entirely liquid cryogen- free and during the run over the last two years we have not spent a single rupee on consumables like liquid nitrogen or liquid helium or helium gas. Further since the sample space is always cooled by the helium vapour, the vibrations due to mechanical movement of displacer do not interfere with the measurements.
Moreover, for entirely vibration-less environment, as mandated by STM and point contact measurement set-up, the advanced pulsed-tube displacer could be opted for.
The block diagram of the cryogen-free system is shown in Figure 2. The static and dynamic pressures in the compressor are 1.7 MPa and 2.5 MPa, respectively. Compressed helium from the compressor reaches the cold head via flexible high pressure hoses. The cold head is housed in a specially designed cryostat chamber that shields it from outside with vacuum and layers of radiation shield. The cryostat also includes a vertical column where the sample is inserted using a dip-stick (variable temperature insert or VTI), a superconducting magnet and a condensation pot. The temperature of the superconducting magnet always remains ~4 K except for small variation of the order of 0.5 K during charging and discharging of the magnet. The radial sample space is 30 mm and thefield homogeneity in
this region is 0.09% over 10 mm. The condensation pot is connected to a separate helium gas recycle consisting of a reservoir (~50 liter helium gas at ~3 psi pressure), an air filter and a dry pump. The helium gas gets liquefied locally at the pot and its vapour flow (and therefore the temperature) in VTI chamber is controlled by a needle valve. To reach 1.6 K, a pressure of 8 mbar is maintained in the VTI chamber.
Controlled heating using 25 W heater near VTI base dictates the temperature of the helium vapour at the sample space. Resistive temperature sensors are placed at various points in the cryostat such as the first stage, shield, second stage, magnet, condensation pot, and exchanger exhaust to continuously monitor the system parameters. The resistance (temperature) at these points is measured using a Keithley 2700 multimeter with a 10 channel scanner. The temperature in the VTI and on the sample holder is measured and controlled by Cernox sensors using a
Lakeshore 340 temperature controller. The Magnet power supply is cryogenic Model SMS 120 C. The maximum field of 8 tesla requires current flow of 108 Ampere. The magnet is fitted with a persistent switch that enables very stable homogeneous field.
Since we have purchased the bare system without the characterization attachments, we have the flexibility to design them as per our requirements. The various attachments already implemented at JNU include dc and ac, resistivity, Hall effect, magneto-resistance, dielectric constant, and RF penetration depth. We are currently trying to integrate a thermo-electric power and an ac susceptibility attachment into the system.
The interfacing software and automation have also been developed in house using Labview. The attachments and the sample can be taken out of the system without warming the system to room temperature (unlike the Quantum Design PPMS). This is achieved by an airlock valve.
Air lock valve
Stage2 Stage1
Needle valve
Cryostat VTI
Sample holder Condensation
pot VTI pressure
Superconducting magnet
Dry Pump
Magnet Power Supply SMS 120C Temperature Scanner
Keithley - 2700
16 pin Connector for sample characterization
probes Temperature controller
Lakeshore-340
Compressor Sumitomo
CSW-71
Helium reservoir
Fig. 2 — Block diagram of the GM cycle based cryocooler and cryostat. The primary helium closed cycle comprises of the compressor connected to the displacer through flexible hoses Separate helium gas close loop consists of a low pressure helium reservoir, a condensation pot, a needle valve and a dry pump. The position of superconducting magnet attached to the second stage of the cold head is also shown. The other interfaced instruments such as magnet power supply, Keithley multimeter with scanner, and temperature controller are also shown
Figure 3(a) shows the data for resistivity as a function of temperature with applied magnetic field ranging from 0 to 7 tesla on an ion irradiated MgB2
thin film9 (superconductor with Tc = 35 K). The field is applied parallel to the ab plane of the sample.
Similar measurement on NbSe2 (superconductor with Tc = 7.1 K) flake in the presence of 0 to 5 tesla field is
shown in Fig. 3(b). It is to be noted that MgB2 is an intermetallic superconductor and its 40 K resistivity (above transition temperature) even in the dirty limit is of the order of 30 µΩ cm. Evidently, the experimental data quality is not compromised because of the mechanical movement of the piston and due to the varying heat load during the isofield temperature scan.
4 Conclusion
To conclude, we have discussed the importance of closed cycle based cryogen-free low temperature and high magnetic field facilities especially in our country where maintenance of liquid helium plants has become a regular problem. We have also discussed the working principle of such an instrument and shared our experience on installation and running of such infrastructure. On the whole we believe that the technology of 4 K closed cycle cryocoolers is reliable and with constant improvement, it is going to replace the classical liquid helium based experiments sooner than later.
Acknowledgement
We acknowledge the funding from Department of Science & Technology, New Delhi, for the cryogen- free low temperature high magnetic field facility at JNU. SDK and AKS acknowledge the CSIR, and DSK acknowledges University Grants Commission, New Delhi, for financial support. We thank Indrajit Naik and J E Giencke for the samples used in this study.
References
1 Kaplan Karen H, Physics Today, 60 (2007) 31.
2 Staticstical thermal physics (F Reif, McGraw-Hill), 1985, p 175
3 Experimental techniques in low-temperature physics, G K White, (Clarendon Press Oxford), 1979 p 12
4 Gifford W E, US patent 2966035, 1960.
5 McMahon H O & Gifford W E, Adv in Cryogenic Engg, 5 (1960) 354.
6 YoshimuraHideto, Nagao Masashi, Inaguchi Takashi, Yamada Tadatoshi & Iwamoto Masatami, Rev Sci Instrum, 60 (1989) 3533.
7 Merida W R & Barclay J A, Advances in Cryogenic Engineering, 43 (1998) 1597.
8 Numazawa T, Kamiya K, Satoh T, Nozawa H & Yanagitani T, IEEE Trans Appl Suporcon, 14 (2004) 1731.
9 Kaushik S D, Patnaik S et al. Physica C, 442 (2006) 73.
22 24 26 28 30 32 34 36
0.0 5.0 10.0 15.0 20.0 25.0 30.0 35.0
H//ab
Resistivity (µ Ω cm)
Temperature (K)
7 T 0 T
0 1 2 3 4 5 6 7 8
0.0 0.2 0.4 0.6 0.8 1.0
5T 4T
3T 2T
1T 0.5T 0T
Normalized resistivity
Temperature (K)
Fig. 3 — (a) Magneto-resistance study of ion irradiated MgB2 film Resistivity is plotted as a function of temperature at different constant magnetic field from 0 to 7 tesla at 1 tesla steps. Magnetic field is applied parallel to ab plane of the sample, (b) Magneto- resistance study on NbSe2 flake plotted in y-axis is the normalized resistivity that was measured in the van-der Pauw configuration.
On applying field beyond 3 tesla, transition temperature of the sample decreased below 16 K
(a)
(b)
