[3] Immediately thereafter he was hired to work at the Atomic Energy Research Establishment, Harwell by Brian Flowers, who was at the time head of the Theoretical Physics Division.
[5] Initially, Hubbard had been reluctant to write up his approach, explaining to Brian Flowers that other researchers had already addressed the same problem, via different although consistent methods.
His main research was on the electronic and magnetic structure of metals, but he also made more applied contributions including to gaseous plasmas and isotope separation.
[9] Walter Kohn described his contributions here as "the basis of much of our present thinking about the electronic structure of large classes of magnetic metals and insulators”,[2] and others have emphasized the importance of the techniques and insight Hubbard applied to the problem.
For instance, beginning in 1967 he wrote five papers on approximate band structure calculation,[10] and in the early 1970s much of his research was on critical phenomena,[11] following Ken Wilson’s seminal work.
While at Harwell he visited a number of American institutions, including a sabbatical at Berkeley (1958-1959), two summers at Brookhaven (1963 and 1969), and a semester at Brown University (Fall 1970).
[2] It was at Berkeley that he published a paper popularizing a method first derived by Ruslan Stratonovich for computing the partition function of many-body systems, now called the Hubbard–Stratonovich transformation.