Construction and use of a pulsed copper nuclear magnetic resonance thermometer

A pulsed nuclear magnetic resonance thermometer, constructed for temperature measurements below 0.5 K, is described. The nuclear free precession signal of copper nuclei is recorded with a gated low noise amplifier and with a phase sensitive detector gated to integrate the signal over an adjustable number of periods. The physical significance of the signal is discussed. A comparison of the resonance thermometer against a slurry type cerium magnesium nitrate (CMN) thermometer for the temperature region 10 mK to 100 mK is presented. The deviation Δ = T_NMR − T_CMN, representing the departure of the CMN thermometer from a Curie law behaviour, was measured as (0.5 ± 0.2) mK. In experiments with a nuclear refrigeration cryostat the resonance thermometer was calibrated against a nuclear orientation thermometer which provided an independent absolute temperature standard. At the low temperature end, from 1 mK to 10 mK, the linearity of the thermometer in T⁻¹ was confirmed by measurements of the nuclear spin-lattice relaxation time.     

The design and use of nuclear orientation thermometers employing54Mn and60Co nuclei in ferromagnetic hosts

The magnetic dipole interaction between the nuclear magnetic momentu and the hyperfine fieldH _hf in ferromagnetic metals leads to well-defined, evenly spaced nuclear energy levels which are accurately measurable by NMR techniques. When this is combined with the well-established anisotropic -ray emission properties of radioactive nuclei an excellent basis for absolute temperature measurement between 1 and 30 mK is obtained. Practical applications, however, require consideration of several experimental aspects. These are discussed in detail and graphs and tables are given in such a form that the data can be easily utilized. Experience gained with a nuclear orientation thermometer system employed in a nuclear refrigeration cryostat is discussed. A calibration of a pulsed copper NMR thermometer against a⁵⁴MnNi nuclear orientation thermometer is presented.     

The nuclear refrigeration of copper

Nuclear refrigeration experiments have been performed on samples containing 0.93 g-atom of copper. After demagnetization to a field of 800 G from initial conditions of 25 kG and 16 mK, lattice temperatures of less than 1 mK were reached. The temperature of the copper nuclei was measured with a pulsed nuclear magnetic resonance thermometer. The lattice temperature was then calculated from this value and the experimentally determined heat leak of 0.10±0.02 erg/sec. An attempt to measure the lattice temperature directly by observing the anisotropy in the angular distribution of the rays emitted by 1 µCi of ⁵⁴ MnZN was not successful. A number of solutions of the differential equations describing nuclear refrigeration have been obtained and were used to simulate experiments with various heat leaks and final magnetic fields. Calculations have also been fitted to the experimental results wherever possible.     

3He melting pressure thermometry

High-precision measurements of the³He melting pressure versus temperature have been made from 500 K to 25 mK using a⁶⁰Co nuclear orientation primary thermometer and a Pt NMR susceptibility secondary thermometer. Temperatures for the fixed points on the melting curve are: the superfluid A transition T_A = 2.505 mK, the A-B transition T_AB = 1.948 mK, and the solid ordering temperature t_n = 0.934 mK. These fixed points and a functional form for P(T) constitute a convenient temperature scale, based on a primary thermometer, usable to well below 1 mK.     

An intercomparison of a Cu nuclear susceptibility thermometer and an Sn99 mösbauer effect thermometer between 6 mK and 1 K

The cold-absorber Sn¹¹⁹ Mössbauer effect thermometer discussed by Parshin et al, has been compared with the steady-field Cu nuclear magnetic susceptibility thermometer. The Curie constant of the Cu thermometer was also determined from simultaneous measurements using a ballistic CMN thermometer. The sensors were installed in the mixing chamber of a dilution refrigerator capable of producing temperatures below 6 mK in the single cycle mode. The absorber never cooled below 7.8 mK due to heat leaks. Good consistency and reproducibility were established between 10 and 100 mK.     

Dilution Refrigerator for Nuclear Refrigeration and Cryogenic Thermometry Studies

This study explores the design and construction of an ultra-low temperature facility in order to realize the Provisional low-temperature scale from 0.9 mK to 1 K (PLTS-2000) in Japan, to disseminate its use through calibration services, and to study thermometry at low temperatures below 1 K. To this end, a dilution refrigerator was constructed in-house that has four sintered silver discrete heat exchangers for use as a precooling stage of a copper nuclear demagnetization stage. A \(^{3}\text {He}\) melting curve thermometer attached to the mixing chamber flange could be cooled continuously to 4.0 mK using the refrigerator. The dependence of minimum temperatures on circulation rates can be explained by the calculation of Frossati’s formula based on a perfect continuous counterflow heat exchanger model, assuming that the Kapitza resistance has a \(T^{-3}\) temperature dependence. Residual heat leakage to the mixing chamber was estimated to be around 86 nW. A nuclear demagnetization cryostat with a nuclear stage containing an effective amount of copper (51 mol in a 9 T magnetic field) is under construction, and we will presently start to work toward the realization of the PLTS-2000. In this article, the design and performance of the dilution refrigerator are reported.     

54Mn in Fe for ultralow temperature thermometry

Experimental investigations into the accuracy of nuclear orientation thermometry are reported.⁵⁴Mn in iron, a primary thermometer useful from 2–40 mK, gives consistent thermometry using both scintillation (3×3 in.-NaI) and solid-state (40-cc-Ge) detectors.⁶⁰Co-in-iron temperatures are consistent with the⁵⁴Mn temperatures above 10 mK, but are about 10% higher at 4 mK.¹²⁵Sb in iron registers a temperature that is perhaps 2 or 3% lower over the 4–15 mK region. Some of the techniques required for the utilization of this type of thermometry are discussed.Work performed under the auspices of the U.S. Atomic Energy Commission.     

Tin,a candidate for low temperature NMR thermometer?

The nuclear spin-lattice relaxation times for the tin isotopes Sn¹¹⁷ and Sn¹¹⁹ have been measured in the temperature range between 20 and 50 mK with a SQUID magnetometer by watching the longitudinal magnetization of the nuclear spins after the application of a rf pulse. The temperature was determined using a SQUID noise thermometer. The Korringa constants were estimated from the τ₁ values. The experimental device and the principal methods are described and the results and thermometer applications discussed.     

Magnetic thermometry to below one millikelvin with lanthanum-diluted cerium magnesium nitrate

Measurements are presented of the pressure and magnetic temperature coordinates of the phase diagram of liquid ³He using a powdered Ce_0.05La_0.95 magnesium nitrate thermometer in the shape of a right circular cylinder with diameter equal to height. The lowest magnetic temperature observed with the thermometer is 0.41 mK. The magnetic temperature of the intersection of the second-order line with the melting curve is 2.625 mK, and the critical temperature at zero pressure is 1.055 mK. The difference between T _c ^* and T _AB ^* extrapolated to melting pressure is 0.57 mK, in excellent agreement with the difference T _A-T_B found from measurements along the melting curve. The present data are compared with the La Jolla provisional absolute scale based on noise thermometry and zero sound. We make the assumption that the absolute temperature T and the present magnetic temperature T ^* are related by T = T ^* + and use the resultant scale to reanalyze a variety of experimental results that had been given in terms of the La Jolla provisional scale.Work supported by the U.S. Department of Energy under contract EY-76-S-03-0034, P.A. 143.     

The specific heat capacity and thermal conductivity of normal liquid3He

By observing the diffusion of a heat pulse along a 10-cm column of normal liquid³He with the aid of two vibrating wire thermometers, it has been possible to measure the heat capacityC and thermal conductivityK of the liquid in the temperature range fromT _C to 10 mK and at pressures of 0.21, 4.39, 9.97, 20.01, and 29.32 bar. By using a Pt NMR thermometer, an LCMN thermometer, and a³He melting curve thermometer calibrated using the melting curve given by Greywall in 1983, a temperature scale has been established and (1) it has been shown that this melting curve is consistent in the temperature range 5–22 mK with the Korringa law for the Pt thermometer with a Korringa constant of 29.8±0.2 sec mK, (2) departures have been observed from the Curie-Weiss law for LCMN at low temperatures, and (3) values of the superfluid transition temperature have been obtained that are about 4% lower than the Helsinki values. The measured heat capacities agree well with those of Greywall, but values ofKT are higher than those of Greywall and show more temperature dependence below 10 mK. The implications for the present results of the very different melting curve given by Greywall in 1985 are discussed in an Appendix.     

Viscosity of the3He-4He dilute phase in the mixing chamber of a dilution refrigerator

The viscosity of the dilute phase of a³He-⁴He solution has been measured using a vibrating wire viscometer situated in the mixing chamber of a dilution refrigerator. The viscosity was extracted from the damping of the resonator for temperatures between 3.7 mK and 100 mK. In the low-temperature Fermi liquid regime T²=48×10^–9 N sec m^–2 K². For temperatures less than 100 mK, the viscometer is a useful secondary thermometer that is not strongly dependent on the applied magnetic field.     

Thin film temperature sensor for cryogenic region with small magnetoresistance

Chromium nitride (CrN) thin films were fabricated onto Si wafers by RF magnetron sputtering equipment in pure N₂ gas. By adjusting the fabrication conditions, the film had negative temperature coefficient of resistance (TCR) below 300 K. It showed reasonable sensitivity between 300 and 1.8 K. The temperature resolution in the cryogenic temperature region was better than 1 mK. A good thermal cycle stability was observed. After 27 thermal cycles between 4 and 300 K, the coefficient of variation (CV) value was as small as 0.098% at 4 K, which corresponds to a 2.1 mK temperature shift. In addition, the thermometer was nearly insensitive to the magnetic field. The temperature shift due to magnetoresistance in a magnetic field of 9 T was less than 5 mK at 4 and 2 K. Therefore CrN can be an excellent choice of material for cryogenic temperature sensors under magnetic fields.     

Realization of the Triple Point of Water in Metallic Sealed Cells at the LNE-INM/CNAM: A Progress Report

A miniature metallic cell for the water triple point (TPW, temperature 273.16 K) was developed for capsule-type thermometer calibrations for realizations with adiabatic calorimetry techniques. The LNE-INM/Cnam previously developed a copper cell for the water triple point and the techniques for cleaning, filling, and sealing. On the basis of previous work, a new copper cell prototype for the TPW was developed and filled at the LNE-INM/Cnam. Measurements were performed using an appropriate calorimeter and a comparison block containing several thermometers. Preliminary results show a scatter of the temperatures measured at the phase transition of the order of 0.2 mK when measurements are repeated over a short-term period (1 month). A positive drift in the phase transition temperature of about 30μK·month⁻¹ was observed over several months. Studies are in progress to improve the cell, to reduce the reproducibility uncertainty to less than 0.1 mK and to have a phase transition with better temporal stability.     

Measurements of absolute temperature below 0.75 K using a Josephson-junction noise thermometer

In order to extend the international temperature scale of 1990, ITS-90, below its lower limit of 0.65 K, we have developed a temperature scale ranging from 6 to 750 mK. Values of absolute temperature are defined on this scale by an R-SQUID noise thermometer. A review is given here of the decade of our experience in the operation of this thermometer and in modeling its systematic errors. The reproducibility of noise temperature values was assessed using superconductive fixed points and the melting curve of³He. To assess the accuracy of the noise thermometer, it was compared with another absolute thermometer (based on nuclear orientation) at the lowest temperatures and with an internationally recognized scale above 0.5 K. The noise thermometer temperatures were used to calibrate two paramagnetic salt thermometers, the result being the construction of a temperature scale that is smooth to approximately 0.01%. On the basis of these comparisons, temperature values defined by the R-SQUID are deemed to be reproducible to within 0.1% and accurate to within 0.3%.Retired.     

3He melting pressure temperature scale below 25 mK

Using⁶⁰Co -ray anisotropy radiation as a primary thermometer, with a Pt NMR susceptibility secondary thermometer, we have made high-precision measurements of the³He melting pressure versus temperature from 500 K to 25 mK. Temperatures obtained for the fixed points on the melting curve are: the superfluid A transition T_A = 2.505mK, the A-B transition T_AB = 1.948 mK, and the solid ordering temperature T_N = 0.934 mK. We provide a functional form for P(T), which, with the fixed points, constitutes a convenient temperature scale, based on a primary thermometer, usable to well below 1 mK.     

Coulomb Blockade Thermometry in the Milli-Kelvin Temperature Range in High Magnetic Fields

We have investigated the usability of a Coulomb blockade thermometer (CBT) at low temperatures around 50 mK and in high magnetic fields up to 27 Tesla. The experiments performed extend previous investigations both to lower temperatures and higher magnetic fields. We show that CBTs provide an easy way of magnetic field independent thermometry in an up to now problematic temperature range below the applicability of vapor pressure thermometry and above the working range of nuclear orientation thermometry, which are established methods of thermometry in magnetic fields.     

Kapton Capacitance Thermometry at Low Temperatures and in High Magnetic Fields

A sensitive Kapton foil capacitance sensor, with size of 9.5 mm×4.5 mm, has been developed and used as a thermometer at ultra-low temperatures down to 1.2 mK and in high magnetic fields. There is no visible magnetic field dependence up to 15 T. The sensor was calibrated with ³He melting pressure thermometer (MPT) and vibrating wire (VW) viscometer. With the silver powder sintered heat exchanger sandwich-like design, the thermal relaxation time is as short as a few minutes at the base temperature. The low temperature (below 1.2 K) reproducibility has been tested and is satisfied within experimental errors.     

Nuclear specific heat of copper potassium tutton salt

The specific heat of copper potassium tutton salt has been measured down to 1 mK in order to investigate the hyperfine interaction of this copper compound. The magnetic ordering of the electronic system occurs at 29.5 mK in zero field. Below the transition temperature the electronic heat capacity decreases and the copper nuclear heat capacity of hyperfine splitting becomes dominant in the heat capacity of the compound. The nuclear heat capacity has a broad peak around 3.5 mK. The entropy of copper nuclear spin, which was calculated from the specific heat data, remains at 40% of ln(2I+1) at 1 mK.     

Magnetic thermal boundary resistance between enhanced nuclear spin sytems and liquid3He

The Kapitza resistance between TmVO₄ and liquid³He in the temperature range 1–40 mK was measured. The Kapitza resistance was proportional to T² below about3 mK. By adding a small amount of⁴He, the Kapitza resistance was enlarged. Applying magnetic fields up to about 70 Oe, the resistance also increased. The decreasing Kapitza resistivity with decreasing temperature was also observed between HoVO₄ and liquid³He. Its value, however, was two orders of magnitude smaller than that for TmVO₄. The experimental results can be explained by the magnetic coupling between enhanced nuclear spins of Tm or Ho and nuclear spins of adsorbed solid³He on the surface of the crystal. By using this model numerical calculation of the Kapitza resistivity was made.     

Search for superconductivity in lithium and magnesium

Recent theoretical work suggests that magnesium and lithium should be superconducting in the m°K temperature range. We have cooled samples of each of these metals to a temperature of 4 m°K, measured by a gamma-ray anisotropy thermometer. Although the magnetic field was less than 10^–2 Oe, no superconducting transitions were observed. The use of a nuclear orientation thermometer employing⁶⁰Co in single crystal (hcp) cobalt is described.Work supported in part by the U.S. Atomic Energy Commission, and in part by ARPA, U.S. Department of Defense.