Design and performance of the Lamb-shift polarimeter

https://doi.org/10.1016/j.nima.2004.08.093Get rights and content

Abstract

A new compact low-energy polarimeter has been designed, developed and put into operation, based on the principle of Lamb-shift polarimetry. Here, we focus on ion-beam deceleration (14–35 kV to 500 V) and the polarimeter's magnetic field. The determination of the optimal setting of oven temperature and the DC gradient in the spin-filter will be presented and illustrated with some measurements. Further, performance of the polarized ion source will be shown by some typical measurements on proton and deuteron beams with the polarimeter.

Introduction

Since 1996 a polarized-ion source of the atomic beam type called POLIS is used as injector for the superconducting K=600 AGOR cyclotron at KVI. The maximum extraction voltage of the source is 35 kV. Both proton and deuteron (vector and tensor polarized) beams are delivered by the source. Until recently, the only way to determine the polarization was by means of in-beam polarimetry (IBP) [1] in the high-energy beam-line, based on elastic p+p or d+p scattering. Thus also for optimizing and tuning of POLIS, the IBP and the accelerator facility had to be used, which is obviously not very convenient. Therefore, we decided to build a Lamb-shift polarimeter (LSP) [2] following the design of the Triangle Universities Nuclear Laboratory [3]. A major difference compared to other LSP systems is that our polarimeter is not integrated in the ion source but is a stand-alone system, with a total length of 91 cm (see Fig. 1), connected to the low-energy beam line. In this article we will describe in more detail the deceleration lens system in front of the polarimeter, the magnetic field configuration and finally some recent polarization measurements.

Section snippets

Principle

In the LSP an electrostatic lens system decelerates the incoming longitudinally polarized beam of 14–35 keV to 500 eV. At this energy, the cross-section for production of neutral metastable atoms, through charge exchange with cesium vapor, is maximal. After neutralization, metastable atoms travel through a spin-filter system. The filter principle is based on a three level hyperfine resonance, between the 2S1/2 and the 2P1/2 sub-states. The filter needs three parameters to operate: a homogeneous

Deceleration

The LSP is located 2 m behind POLIS. In front of the LSP, a 90° electrostatic deflector guides the ion beam into the injection beam-line of AGOR. When a polarization measurement is requested, the deflector is moved upwards, enabling the ion beam to enter the LSP. In order to decelerate the ion beam to 500 eV, the LSP is placed on a high-voltage platform at a voltage of 500 V lower than the POLIS extraction voltage. To prevent the beam from diverging when it enters the LSP, a two-element

Constant magnetic field

Four main factors have to be taken into consideration when designing the magnetic field: (a) at the cesium charge-exchange cell the magnetic field B has to be much larger than the critical field Bc, with Bc=6.3mT for hydrogen and Bc=1.4mT for deuterium [4]; (b) no decrease is acceptable in the number of metastable atoms, in the selected hyperfine state, by unwanted quenching in the spin-filter due to imperfection of the homogeneity of the magnetic field. Therefore, a ripple of <±0.025mT was

The cesium charge-exchange cell

The cesium charge-exchange cell is a cesium recirculating system identical to the one of TUNL [3]. A reservoir filled with 10 g of cesium is heated to a preset temperature of about 200 °C. The cesium evaporating from the reservoir is directed through a chimney to the charge-exchange chamber. When the vapor hits the wall (T=60C), it condenses and flows back into the reservoir.

The neutralization probability for the incoming ions into a metastable state depends very sensitively on the cesium vapor

Peak width measurements

We have performed systematic measurements of the peak width as a function of the DC voltage of the spin-filter. The results are collected in Fig. 7, where the Cartesian coordinates of each pixel denote the magnetic field and the DC voltage. The gray scale represent the intensity of the signal. Clearly, the peak width and height decrease with decreasing DC voltage until the peak is lost in the background which increases with decreasing voltage. Below 50 V, the spectrum is almost completely

Polarization measurements

The LSP is in routine operation since its commissioning in September 2002. The design is such that it accepts beams up to 35 keV, till now LSP has been tested up to 25 keV. Our experience shows that the time needed to optimize POLIS has decreased significantly and at the same time higher polarization values are obtained. Some recent spectra are shown in Figs. 5 and 6.

For protons, a maximum pure vector polarization Pz=-0.912±0.01 is obtained while for the pure tensor polarization of deuterons a

Conclusions

In this article we show the successful application of the Lamb-shift polarimeter at ion beams with energies from 25 keV down to 500 eV. Several aspects of the LSP operation have been discussed. The optimal oven-temperature of 210 °C for the highest neutralization efficiency has been determined. We performed systematic measurements of the spin-filter resonance width as a function of the DC voltage to obtain the optimal setting for the peak width (VDC250V). The fact that the population of one

Acknowledgements

We would like to thank the TUNL group for sharing with us their experience of the LSP and, in particular, T. Clegg for his great help during the design and commissioning. This work has been supported by the Rijks Universiteit Groningen (RuG) and by the European Union through the Large-scale Facility program LIFE under contract number ERBFMGE–CT98–0125. It has been performed as part of the research program of the `Stichting voor Fundamenteel Onderzoek der Materie' (FOM), with support of the

References (4)

  • R. Bieber

    Nucl. Instr. and Meth. A

    (2001)
  • H.R. Kremers

    Nucl. Instr. and Meth. A

    (2004)
There are more references available in the full text version of this article.

Cited by (4)

View full text