Introduction

The symposium was held to honour his life-long contributions to polarised neutron scattering, and in particular the three-dimensional generalisation of the polarisation analysis (Spherical Neutron Polarimetry) and his pioneering contributions towards the polarisation of wide-band neutron beams with 3He spin filters.

After the construction of the high-flux reactor of the Institut Laue Langevin (ILL), a series of advances occurred in Europe. P.J. Brown and J.B. Forsyth discovered that the (111) reflection of the Heusler alloy Cu2MnAl is a better candidate for polarising a neutron beam, and the production of good single crystals was undertaken successfully by the CEA-Grenoble, the CNRS and the ILL. In the early 1970s, J. Schweizer and F. Tasset were in charge of the construction of a polarised neutron triple-axis spectrometer of the first generation, and a few years later, P.J. Brown and J.B. Forsyth built a diffractometer dedicated to the determination of magnetisation distributions.

At the beginning, the flipping system was of the type proposed by C.G. Shull: a Larmor frequency coil  in a magnetic field. F. Mezei brought from Budapest a second type of flipper (Mezei flipper) which consists of a rectangular coil into which the polarisation precesses around the field. There is no frequency to tune but it is still wavelength and stray field dependent. F. Tasset proposed a third type of flipper (Cryoflipper) which is made of two opposite fields separated by a thin Meissner sheet: the field changes sign very abruptly and there is no wavelength and stray field dependency.

F. Mezei demonstrated in 1972 that by the Spin Echo method, very small velocity changes could be observed independently of the velocity spread. He also had the idea of supermirrors and pointed out the idea of a generalised polarisation analysis. With the successful development of supermirrors at the ILL, O. Schaerpf reconstructed in 1992 the diffuse scattering instrument D7 in order to accommodate polarisation analysis. The trick here was to use a triple Helmholtz coil at the sample position to bring the incident polarisation along any of the three directions X, Y or Z, and to analyse the scattered beam in the same direction (XYZ Method or three-directional polarisation analysis). From the measured six cross-sections, one can separate the nuclear, nuclear spin incoherent and magnetic contributions. This is a generalisation of the technique discovered in 1969 by R.M. Moon, T. Riste, and W.C. Koehler (one-directional polarisation analysis).

Following the pioneering experiment of H. Alperin who demonstrated in 1973 that the generalised or three-dimensional polarisation analysis could be realised by connecting two different guide-field directions onto a zero-field sample chamber, F. Tasset built in 1989 an apparatus at the ILL to determine the direction of the scattered polarisation vector for any given incident polarisation and any scattering angle. Using his expertise with superconducting screens developed when building the Cryoflipper and in spite of very little support, he constructed a compact Cryogenic Polarisation Analysis Device (Cryopad) which takes advantage of the Meissner shields to properly define the magnetic field and zero-field regions crossed by the incident and scattered neutron beams.

For the first time, all the components of the complicated expression of the scattered polarisation vector could be measured at any scattering angle, which immediately provided unique information on magnetic structures. The first experiments performed with P.J. Brown and J.B. Forsyth were very successful and motivated F. Tasset to pursue the development. F. Tasset proposed to call the technique Spherical Neutron Polarimetry (SNP) so that no confusion is made with the technique introduced at Oak Ridge by R.M. Moon, T. Riste, and W.C. Koehler.

In 1996, F. Tasset constructed a second-generation Cryopad allowing the polarisation vector of the incident beam of neutrons to be set in any direction and the magnitude and direction of the scattered polarisation vector to be measured to a precision of 3 degrees for any given momentum transfer. The results were presented to the scientific community at the PNCMI international workshop organised at ILL in 1998 and the Spherical Neutron Polarimetry technique was finally recognised by the experts. 

About 5 years later, three copies of a third-generation Cryopad had been constructed and implemented at ILL and JAEA . A non-cryogenic version called MuPAD was also constructed at FRM II by R. Gähler, M. Janoscheck et al. This version is less precise and less reliable but helped people understand the benefit of Meissner screens. Indisputably, F. Tasset has contributed a lot to the adoption of the vectorial property of the neutron polarisation. Today, ILL builds a fourth copy of Cryopad for TU-Aachen (FRM II) and discussions have started with HMI for a fifth unit.

These zero-field polarimeters are used to investigate complex magnetic structures, measure magnetisation distributions of antiferromagnets, parity-violation effects and even the neutron electric dipole moment. Recently, E. Lelièvre-Berna and C. Pappas have also applied SNP to neutron spin echo spectroscopy (Polarimetric Neutron Spin Echo). Only with this technique was it possible to reveal unambiguously the existence of a new type of magnetic phase in MnSi featuring chiral fluctuations.

Because of the low intensity available on polarised neutron instruments, the search for new methods for polarising neutron beams has always been active. The use of gaseous 3He spin polarisers was first introduced by F. Tasset at Allevard on the 31st of August 1989 when he presented a talk entitled “The polarised 3He gas filter: a promising method for hot neutron beams” at the ILL workshop “Proposals for the 3ème souffle”. The principle of this filter is based on the enormous difference in the absorption cross-sections for neutrons with spin parallel and antiparallel, difference which exists for a very broad band of wavelengths.

In the second half of 1989, F. Tasset invited T. Chupp at ILL. Then they met at ENS and ILL for preparing an experiment for which the Scientific Committee allocated 3 weeks of beam time. In October 1990, they performed this experiment by optical pumping of Rb vapor and Rb-3He collisions at room temperature with the help of Harvard University. F. Tasset presented the results at the ICNS conference (Oxford) in 1991: the filtering of the neutron spin was becoming a reality.

After obtaining some financial support from the European Commission, F. Tasset went into a collaboration with Mainz University whereby a filling station was built for the ILL comprising the optical pumping and the  compression of the gas. In this case, the gas was polarised by optical pumping of 3He atoms excited by an electric discharge at about 1 mbar and collisions of excited 3He - 3He nuclei. In 1996, he tested successfully this new device on the diffractometer D3 at ILL. Several experiments were then very successfully carried out with up to 55% 3He polarisation.

In 2000, F. Tasset started the construction of a second-generation filling station financed by a UK/EPSRC grant (now STFC). In a few years, the maximum 3He polarisation reached 70% and the production rate were multiplied by 5. Today, this station daily supplies several instruments with spin filter cells filled with 3He gas polarised at 80%. This new technique has been adopted by the international neutron community and the construction at ILL of new filling stations has started for neutron facilities located in UK and Australia. F. Tasset’s pioneering contributions toward the polarisation of wide-band neutron beams with 3He spin filters have changed the face of the neutron world.

E. Lelièvre-Berna (ILL)