APPLICATIONS OF MAGNETIC METHODS TO CHEMISTRY

By Linus Pauling

Sigma Xi, U.C.L.A., November 4, 1936.

It was with much hesitancy that I came here today to speak on magnetic methods in chemistry, in as much as you have among you a great authority on magnetism, Professor Barnett. I surmise on looking around, however, that Professor Barnett must have had a similar feeling of hesitancy about coming to listen.

On thinking over my subject I have realized that the applications I had in mind comprise only a small part of those existent. For example, one magnetic method much used by chemists of the 10^-4 th school, those who work with vacua instead of chemicals, consists in the magnetic breaker-offer. I am not going to talk about this, or the magnetic pump, or similar gadgets. I shall not mention even the magnetic method of attaining low temperatures, originated by a Norther colleague of yours, Professor Giauque, even though this advance in cryogenic technique, whereby the achievable low temperature around 1/100º rather than 1 K, is the greatest which has been made since the liquefaction of air, and even though the discovery, made in a chemical laboratory is of great importance for chemistry as well as for physics.

Instead of these, I shall talk about those methods which involve the determination of the magnetic susceptibility of a substance - of the magnitude of the magnetic moment induced in a substance by a magnetic field. The magnetic and electric properties show a peculiar difference. On applying an electric field to a substance it becomes polarized with the field, the positive particles being pulled by the field and the negative ones pushed, its ability to polarizability is positive. the magnetic moment induced in a substance by a magnetic field may, however, be either positive or negative, corresponding respectively to paramagnetism and diamagnetism, as was discovered 91 years ago by Faraday. A paramagnetic substance is attracted into a magnetic field, a diamagnetic substance repelled.

Weber-Langevin theory.

Larmor precession - diamagnetism. [Diagram illustrating diamagnetism] Orientation of orbits or spins - paramagnetism.

Diamagnetism has not been of much value to chemists. the extensive studies of Pascal led to the conclusion that in the main atomic susceptibilities can be added to give molecular susceptibilities, there being not much structural dependence. An exception is shown by graphite and other aromatic substances, which show an anomalously large diamagnetic susceptibility normal to the plane of the molecule, amounting to a 50-fold increase for graphite. This has been made the basis of an interesting auxiliary method of crystal structure investigation by Krishnan, a coworker of Raman. From the magnetic anisotropies of crystals information regarding the orientation of the molecules is obtained. I am hoping also that it can be used for structural organic work of a more old fashioned kind. There are some interesting substances called melon, melam, and melem, hydromelonic acid, cyameluric acid, etc., which were studied vigorously by all the old boys (Liebig, etc.) 100 years ago, their efforts at finding structural formulas being not crowned with success. These substances are obtained whenever almost anything is heated awhile - for example, Pharaoh's serpents, mercuric thiocyanate, burn to given melon; and by heating KSCN and SbCl₃ potassium hydromelonate, K₃C₉N₁₃, is obtained, as a substance stable at a red heat! the reaction may be 39 KSCN + 10 SbCl₃ → 30 KCl + 5 Sb₂S₃ + 12 CS₂ +3 K₃C₉N₁₃; actually CS₂ is given off. the structure suggested for these substances involves a new condense ring system, which would given a large anomalous diamagnetism. Mr. Redemaun of your Chemistry Department is working on this problem.

The most interesting field of application involves paramagnetism. Fortunately in most molecules and complexes we need worry only about the magnetic moment due to the spins of electrons, that due to orbital motion being quenched. In consequence from the measured paramagnetic susceptibility a statement can be made at once about the number of unpaired electrons in the molecule. Most molecules are diamagnetic: if the number of electrons is odd the molecule must be paramagnetic. This was verified over 10 years ago by G.N. Lewis and a student, Taylor, who studied a free radical.

The importance of the method lies in a correlation with stereochemistry. Bivalent nickel, and its even complexes, would contain 0, 2, 4, --- unpaired electrons. It is found that with four square bonds it has 0 (diamagnetic), and with four tetrahedral bonds or other ionic bonds; Ni(CN)₄ is of the first kind, Ni(NH₃ ⁺)₄ ⁺ of the second. the compounds of iron are especially interesting. For Fe111 we have φ = 6 covalent bonds, 3-4 covalent bonds?, 5 - ionic bonds. A recent study of

[Manuscript ends here.]