ON THE SUPERCONDUCTIVITY OF SOLUTIONS OF SODIUM IN AMMONIA *)

In March–July 1946, brief communications by R. Ogg¹ were published on his discovery of the superconductivity of solutions of sodium in ammonia, observed at temperatures from 90 to 190° K. Ogg’s works attracted general

) L. Giulotto, A. Gigli, Electrical conductivity of Na—ammonia solutions at low temperatures. Phys. Rev. 71*, No. 3, 211 (Febr. 1947).

attention, since the highest temperature at which superconductivity had previously been observed is approximately \(3^\circ K\) (NbN—NbH). Orr used solutions with concentrations from \(0.5N\) to \(2N\). In this concentration range, when the temperature is lowered to \(40\text{--}60^\circ C\), the solution separates into two phases differing in color. If, however, the solution is cooled very rapidly, as Orr did, it does not solidify, at least apparently, into two phases. With such rapid cooling the resistance of the solidified solution decreases by a factor of a thousand, from \(\sim 10^4\ \Omega\) in the liquid state to \(\sim 10\ \Omega\) in the solid. Such a resistance would be possessed by metallic sodium if it precipitated from the solution, forming a conducting channel. The deposition of sodium on the walls of capillaries containing a solidified solution was indeed observed. Orr, however, believes that the solidified solution is superconducting, and attributes the residual resistance to poor contacts arising from cracking during solidification. To prove the superconductivity of the solidified solution, Orr immersed a ring made of the solution in liquid oxygen in a magnetic field of the order of 1500 oersteds and then, after removing the ring from the magnetic field, measured its magnetic moment by the induction method. Only in 7 cases out of 200 were galvanometer deflections observed that made it possible to assume the presence of a current in the ring. The effect disappeared by itself after 1–2 min.

After the publication of Orr’s works, experiments were undertaken in various laboratories to check Orr’s experiments. In doing so, the majority of investigators did not confirm the results obtained by Orr. In the USSR, Shalnikov, Tumanov, and Sharvin² came to the conclusion that the content of the superconducting phase could not exceed \(0.1\%\). In this case the influence of cracking was excluded. Daunt, Desirant, Mendelssohn, and Börс³, using approximately the same technique, also obtained negative results. Burs, Kук, Pontius, and Zemansky⁴ came to the same conclusions; however, the possibility of cracking, which was not excluded in these experiments, makes them insufficiently convincing. Only Hoskins⁵ found a positive effect, though only in a very small number of cases.

Giulotto and Gigli (Institute of Physics, University of Pavia), having repeated Orr’s experiments, obtained negative results. Rapid freezing of the solution (with a concentration of about \(1N\)) occurred when a small quantity of solution fell into a vessel with liquid helium immersed in liquid air. The vessel was located between the poles of an electromagnet, whose field strength was \(\sim 500\) oersteds. A solid disk formed at the bottom of the vessel, the temperature of which after several seconds fell to the temperature of liquid air. After this the vessel was removed and placed near a magnetometer with a sensitivity of \(10^{-4}\) oersted. The duration of this operation did not exceed one second, and measures were taken to prevent rapid heating of the solid disk.

In all 20 experiments performed by the authors, no superconducting effect was discovered. The authors also measured, by means of the usual Wheatstone bridge circuit, the resistance of the solution, rapidly freezing it and then heating it. In doing so an extremely sharp drop in resistance was observed (at \(-80^\circ C\) for a \(0.7\) molar solution). The ratio of the maximum resistance in the liquid phase to the minimum in the solid phase is approximately 150, but in experiments with slower freezing or heating it reached still larger values. The phenomenon is reversible. The authors attribute the increase in resistance before solidification to the possible appearance of two phases. With very rapid freezing this increase in resistance was either very small or not observed at all. The authors consider their measurements of the resistance of the solid solution not very accurate, since they could not avoid the appearance of cracks in the disk. However, in other experiments in which, in the authors’ opinion, the influence of poor contacts was excluded, the resistance of the solid phase remained different from

zero. On the basis of their experiments the authors come to the conclusion that the change in resistance upon solidification of the solution is associated with an ordinary phase transition and cannot in any way be interpreted as the appearance of superconductivity. In most of the studies mentioned above, undertaken to verify Ogg’s effect, including those based on a more flawless methodology, the same results were obtained. Therefore we must apparently reject Ogg’s conclusion about the superconductivity of a solution of sodium in ammonia. Nevertheless, such a sharp drop in resistance upon solidification is of great interest and requires further study.

B. Geilikman

References

1. R. A. Ogg, Phys. Rev. 69, 243, 1946; 70, 93, 1946.
2. K. A. Tumanov, A. I. Shal’nikov, Yu. V. Sharvin, DAN, LVI, 33, 1947.
3. J. G. Daunt, M. Desirant, K. Mendelssohn, J. Birch, Phys. Rev. 70, 219, 1947.
4. H. A. Boorse, D. B. Cook, R. B. Pontius, M. W. Zemansky, Phys. Rev. 70, 92, 1946.
5. J. W. Hodgins, Phys. Rev. 70, 568, 1946.