Appl. Phys. A 74 [Suppl.], S1498–S1501 (2002) / Digital Object Identifier (DOI) 10.1007/s003390201793 Applied Physics A Materials Science & Processing Modeling the neutron spin-flip process in a time-of-flight spin-resonance energy filter A.A. Parizzi 1,3 , W.-T. Lee 2 , F. Klose 2, ∗ 1 Argonne National Laboratory, Spallation Neutron Source Project, USA 2 Oak Ridge National Laboratory, Spallation Neutron Source Project, USA 3 Electrical Engineering Department, Federal University of Rio Grande do Sul, Brazil Received: 3 August 2001/Accepted: 12 March 2002 –  Springer-Verlag 2002 Abstract. A computer program for modeling the neu- tron spin-flip process in a novel time-of-flight (TOF) spin- resonance energy filter has been developed. The software allows studying the applicability of the device in vari- ous areas of spallation neutron scattering instrumentation, for example as a dynamic TOF monochromator. The pro- gram uses a quantum-mechanical approach to calculate the local spin-dependent spectra and is essential for op- timizing the magnetic field profiles along the resonator axis. PACS: 28.20.Cz; 29.25.Dz We have recently proposed to use a spin-resonance energy filter in a dynamic way, as a “pulse shaper” to control the wavelength-resolution in time-of-flight (TOF) experiments at spallation neutron sources [1, 2]. The setup is based on a method initially proposed by G.M. Drabkin for selecting the velocity of polarized neutrons at reactor sources [3]. The basic idea is to guide a polarized polychrome neutron beam through a magnetic resonator system. The resonator flips the neutron spins of a narrow band around a resonance en- ergy (differently from gradient RF-flippers, that flip broader bands [4]). The flipped fraction of the beam is separated from the remainder by a polarization analyzer – Fig. 1a. Figure 1b shows a possible implementation of the device. The resonator itself consists of a zigzag-folded, current-carrying foil that creates magnetic fields ±H per in a direction perpendicular to the neutron flight path. In addition, a longitudinal mag- netic guide field H o ( H o ≫±H per ) is applied throughout the device. The resultant field determines the local quantization axis and the Lamor precession frequency (see [1, 5–7] for details of the design and the required constraints between the fields to achieve a particular resonance energy). Pre- sented herein is a method of modeling the resonance filter by using a combination of analytical and numerical calculation methods. ∗ Corresponding author. (Fax: +1-630/252-4163, E-mail: FKlose@anl.gov) 1 Calculation of magnetic field profiles The magnetic fields ±H per that are generated by the current- carrying sheets have been calculated using an analytical semi- finite current sheet approach [8]. Figure 1c shows results for a typical resonator. In this example, the sheets are 5 cm wide and separated by 1.5 mm. The transverse field ±H per shows symmetric oscillations close to the midpoint of the resonator and end effects at both ends. The end effects, however, do not influence the spin flip process since the offset to negative ±H per values at the left end is compensated by higher values at the other end. In order to track the wavelength/arrival-time dependence for dynamic TOF applications of the energy filter, both ±H per and H o must be drifted according to the time structure of the source, resulting in currents proportional to 1/t [7]. Each individual neutron, however, needs to experience the static condition of its corresponding resonance field in order to achieve a proper 180 ◦ spin flip (the problem is the finite length of the resonator and the associated time that the neu- trons need to traverse it). In order to match both require- ments, we suggest using a magnetic guide field configura- tion with a linear gradient. The required relative gradient depends on the distance from the source to the center of the device [2,7]. On the other hand, when using the device in a static mode, i.e. tuned to a single resonance wavelength, a spatially uniform guide field is required. In order to facil- itate easy switching between both operational modes, it is advantageous to build the device with two sets of coils for H o , one for generating a uniform field and another for creat- ing a field with a gradient component – see Fig. 1d. To create the desired gradient, a certain fixed ratio between the cur- rents in the two sets of coils must be adjusted, and drifting both currents according to the time structure of the source maintains the optimal resonance condition for all incoming neutrons having the correct TOF/wavelength relation. In this way, the setup is suitable for constant wavelength (CW) static and TOF dynamic applications, allowing operation in differ- ent positions of the beam line and fine-tuning of the field gradient.