In 1929, Mott was appointed a lecturer in the
School of Physics at the
University of Manchester. He returned to Cambridge in 1930 as a Fellow of and lecturer at
Gonville and Caius College, Cambridge, and in 1933 moved to the
University of Bristol as Melville Wills Professor of Theoretical Physics. In 1948, he became Henry Overton Wills Professor of Physics and Director of the Henry Herbert Wills Physical Laboratory at Bristol. In 1954, he was appointed
Cavendish Professor of Physics at Cambridge, a position he held until 1971. He was instrumental in the painful cancellation of the planned
particle accelerator because of its very high cost. He also served as Master of Gonville and Caius College, 1959–1966. His early works were on the theoretical analysis of collisions in gases, notably the
collision with spin flip of an electron against a hydrogen atom, which would stimulate subsequent works by André Blandin and
Jun Kondo about similar effects between
conduction electrons, as well as magnetic properties in metals. This sort of activity led Mott to writing two books. The first one, which was edited together with
Ian Sneddon, gives a simple and clear description of quantum mechanics, with an emphasis on the
Schrödinger equation in
real space. The second describes atomic and electronic collisions in gases, using the rotational symmetry of
electronic states in the
Hartree–Fock method. But already in the middle of the 1930s, Mott's interests had broadened to include solid states, leading to two more books that would have a great impact on the development of the field in the years prior and after
World War II. In 1936,
Theory of the Properties of Metals and Alloys (written together with H. Jones) describes a simplified framework which led to rapid progress. The concept of
nearly free valence electrons in metallic alloys explained the special stability of the
Hume-Rothery phases if the
Fermi sphere of the
sp valence electron, treated as free, would be scattered by the
Brillouin zone boundaries of the atomic structure. The description of the impurities in metals by the
Thomas Fermi approximation would explain why such impurities would not interact at long range. Finally the delocalisation of the
valence d electrons in
transitional metals and alloys would explain the possibility for the
magnetic moments of atoms to be expressed as fractions of
Bohr magnetons, leading to
ferro or
antiferromagnetic coupling at short range. This last contribution, produced at the first international conference on magnetism, held in
Strasbourg in May 1939, reinforced similar points of view defended at the time in France by the future Nobel laureate
Louis Néel. In 1949, Mott suggested to
Jacques Friedel to use the approach developed together with Marvey for a more accurate description of the
electric-field screening of the impurity in a metal, leading to the characteristic long range charge oscillations. Friedel also used the concept developed in that book of virtual bound level to describe a situation when the atomic potential considered is not quite strong enough to create a (real) bound level of symmetry e ≠ o. The consequences of these remarks on the more exact approaches of cohesion in rp as well as d metals were mostly developed by his students in Orsay. The second book, with
Ronald Wilfred Gurney,
On the Physical Chemistry of Solids was more wide-ranging. It treated notably of the oxidation of metals at low temperatures, where it described the growth of the oxide layer as due to the electric field developed between the metal and absorbed oxygen ions, which could force the way of metallic or oxygen ions through a disordered oxide layer. The book also analysed the photographic reactions in ionic silver compound in terms of precipitation of silver ions into metallic clusters. This second field had a direct and long lasting consequence on the research activity of John (Jack) Mitchell. Mott's accomplishments include explaining theoretically the effect of light on a
photographic emulsion (see
latent image). His work on oxidation, besides fostering new research in the field (notably by J. Bénard and
Nicolás Cabrera), was the root of the concept of the
band gap produced in semiconductors by gradients in the distribution of
donor and acceptor impurities. During
World War II, Mott joined the "Army Cell" of
radar researchers. He was put in charge of getting the Army's
GL Mk. II radar working in the presence of serious calibration problems that caused the measurements to change as the antenna tracked its targets. He solved this problem by designing a large metal wire mat that was built around the radars to provide a very flat reference plane. During the war Mott worked on the role of plastic deformation in the progression of fracture cracks. When he returned to Bristol after the war, his having met and hired
Charles Frank enabled the two of them to make considerable advances in the study of
dislocations, with the help of others such as
Frank Nabarro and
Alan Cottrell. Bristol became an important centre of research in this topic, especially at the end of the 1940s. If Mott only produced early and somewhat minor contributions to that field, notably on alloy hardening with Nabarro and on the topology of a dislocation network lowering the apparent elastic constants of a crystal, there is no doubt that Mott's enthusiasm played its role in the three major steps forward in the field by Frank on crystal growth and plasticity and later, in Cambridge, by
Peter Hirsch on thin film
electron microscopy. At the same time, however, Mott gave a lot of thought to
electronic correlations and their possible role in
Verwey's compounds such as nickel oxides which could switch from metals to nonmetallic insulators under various physical conditions - this is known as the
Mott transition. The term
Mott insulator is also named for him, as well as the
Mott polynomials, which he introduced. == Personal life ==