|
Chapters:
|
MuSR /
Setup< Transverse field | Index | Calibrations > We shall describe briefly the GPS spectrometer at PSI, a typical versatile instrument at a continuous muon source, whose official page is here, and the MUSR spectrometer, its closer analogue at the ISIS pulsed source, whose official page may be found here Source: the accelerator is a synchrotron which accelerates protons to 590 MeV with a radio frequency cavities operating at 50 MHz. The beam is extracted continuously, with an efficiency close to 100%. The peak current is 2mA, corresponding to the power of 1 MW. This results in {$\approx 10^{16}$} protons s-1 with a fine time structure, bunched in bursts of {$2\,10^{8}$} particles each, repeating every 20 ns. The surface muon beam results from the extracted beam being directed onto a 5 mm pyrolythic carbon target (M). This is a transmission target: only a small fraction of the protons interact with it. Since one can accept muons only up to 50 kHz ({$5\, 10^4$} muons s-1 vide infra), the probability of having two muons from the same bunch is negligible ({$<10^{-6}$}) and we may consider the beam as continuous. Beamline: The instrument is fed by the πM3.2 beamline, after target M, which can be used either exclusively, or shared with another instrument (LTF). The proton beam has a waist (one standard deviation) of 10 mm at the target. The beamline includes a spin-rotator, i.e. a crossed fields device, providing an intense electric field at right angles with a magnetic field and with the beam direction. The two fields must stand in a specific ratio, of {$E/B=p\sqrt{1-\beta^2}/m\approx v$} for nearly non-relativistic surface muons. Therefore, when carefully tuned, the effect of these fields is to cancel exactly the force on the charge, only for particles of rest mass m and {$v/c=\beta$}, among all those with momentum {$p$} transported by the beam-line. This device serves two purposes:
This second use is required by precession experiments in large magnetic fields: the native 29 MeV/c surface muons exposed to a large transverse magnetic field at the sample would turn in narrow bends and never reach it. Instead spin-rotatated muons may be implanted in the presence of a large magnetic field parallel to their momentum and precess around it. The beamline must be retuned for each of these two settings (a total of six different typical conditions, with spin rotator on or off and instrument GPS, LTF or sharing). Spectrometer
The beam can feed a larger number of muons per second than can be dealt by a single instrument. The muon counter starts the clock when a muon is stopped in the sample. This condition is checked by coincidence scheme on the signal from all of the instrument detectors, for example there must not be a time-coincident hit on any of the positron detectors, which would indicate a scattered particle rather than a stopped muon. Only a fraction of the beam (10 mm waist, standard deviation) hits the muon counter (8 mm diameter). For samples smaller that this dimension a veto counter behind the sample may signal muons that did not stop in the sample, at the cost of reducing the active area of the B detector. In any case a strict requirement of continuous beam µSR is that only a muon is present in the sample at any given time to guarantee the correct attribution of the child positron to its parent muon, hence the correct determination of the individual muon lifetime. This implies that whenever the clock is started, a second muon hit before any positron requires the whole event to be discarded, and the probability of this event (the so called pile-up) is kept low by tuning the muon beam to no more than 50 kHz, for a measuring time T=5 µs (35kHz for T=10 µs). Source: the accelerator is a synchrotron which accelerates protons to 800 MeV in 10000 revolutions around the ring. Two bunches at opposite end of a diameter are accelerated together and extracted by a fast kicker magnet, delivering 4 μC of protons in two 100 ns long pulses. The entire acceleration process is repeated 50 times per second, so that the extracted proton beam (EPB) has a mean current of 200 μA is delivered to the targets. The EPB passes through a 10 mm thick pyrolythic carbon transmission target that produces two main muon beams, having the same duty cycle of the EPB: 50 double bursts per second, each formed by two pulses, 80 ns wide (FWHM), separated by 300 ns.
A common clock is started at each burst, triggered by an upstream Cerenkov detector. This implies that any precession frequency approaching {$\nu=\frac 1 {\Delta t_0}\approx 12$} MHz will be averaged to vanishing amplitude due to {$\sim 2\pi$} dephasing between early and late muons in the burst. Actually there is a large passband reduction already above 6MHz. Spectrometer
Longitudinal geometry When the axis of the two detector rings is longitudinal to the beam (i.e. muons come from the upper left corner of the sketch) each detector ring is either Forward or Backward, with respect to the initial spin direction. Compensation coils are tuned to reduce the stray fields down to few hundreds nT, for zero field operation. The maximum achievable field in this configuration is 0.25 T, along the beam direction. An additional small field of circa 2 mT can be provided by a separate set of coils, along an axis perpendicular to the beam, for calibration purposes. Transverse geometry Since there is no spin-rotator on MUSR (and the low passband limits anyways transverse fields to below a maximum of 40 mT), the transverse field geometry is achieved here by rotating the instrument around a vertical axis, so that the beam in the sketch would arrive, say, from the lower left corner. In both geometries every positron hitting a detector is a read-out signal for the common clock. No detector to define individual incoming muons is used. The beam profile can be controlled by means of upstream collimation, at a previous waist of the transported beam envelope (actually, the second before the sample position, since the one after that would be cut the momentum distribution, not the successive beam spot). The typical case provides 80% of muons within a sample of 30 mm diameter. Much smaller samples can be measured in the so-called fly-past mode, i.e. ideally nothing should intercept the muons that do not stop in the sample. This implies a sample suspended by very thin connections and a long tube extending the sample vacuum downstrem the spectrometer. < Transverse field | Index | Calibrations > |
