Mass spectrometer
General Design
The sector field mass spectrometers (SFMS) used for helium isotope and tritium measurement are commercial instruments (Fig. 11, VG 5400; Micromass; Manchester, UK). They are designed for measurement of all noble gases (He, Ne, Ar, Kr, Xe) and their isotopes. For our purpose, they are dedicated to helium isotope measurements. They are operated in a static mode, i.e., all pumps with the exception of two SAES getters are disconnected from the mass spectrometers during sample measurement. The SAES getters keep the background pressure low, especially that arising from hydrogen and HD without pumping the noble gases.
Fig. 11: Schematic diagram of the VG 5400
noble gas mass spectrometer
Vacuum system
During automated operation, the mass spectrometers are pumped between measurements of individual samples or standards by ion getter pumps (Varian, StarCell VacIon Plus 40) attached to the flight tubes by pneumatically operated valves (VG, CR 38 TU, and VAT, Series 57, respectively). The advantages of ion getter pumps are (1) the absence of any oil (low hydrogen and HD levels), and (2) a closed system in case of a power failure. An additional pumping line consisting of a turbomolecular - drag pump (Balzers, TMU 065; tritium mass spectrometer) backed by a rotary pump (Edwards, E2M2) is used for initial pumping after venting the system. An oven can be mounted over all vacuum parts for bakeouts at temperatures of up to 350 °C. A Penning gauge (Balzers, IKR 260) is used to monitor the pressure in the turbomolecular - drag pumping line (usually this pressure is below the detection limit of 2× 10-9 mbar).
Ion source
The VG 5400 uses a Nier-type electron impact ion source (Bright source). The electron emission is controlled by the trap current which regulates the filament current. To achieve the sensitivity required for high-precision helium isotope and tritium measurements, we have to use a relatively high trap current of 800 µA. The electrons emitted from the filament are forced onto circular trajectories by a magnetic field (source magnets) to increase the ionization probability of gas atoms in the source. The ions are extracted from the ionization cage and accelerated to 4.5 keV. The field of the separation magnet is controlled by a Hall probe through adjustment of the magnet current.
Ion detection
The 4He ion beam is measured with a Faraday cup as the voltage produced over a 1010 Ohm resistance. A 0.4 cm3 STP air standard produces a signal of about 2.5 V (measured by a 7-digit-voltmeter; Solartron Instruments, 7060 system), which is equivalent to 250 pA or 1.5× 109 4He atoms per second.
A channeltron (Galileo, 4869 EIC) operated in the ion counting mode is used for measurement of the 3He beam. The channeltrons produce a good plateau in the HV (Brandenburg, alpha III unit) versus count rate plot (Fig. 12). This plateau is typically located between 1.6 and 2.4 kV and its absolute position on the HV axis varies between individual channeltrons. The high voltage used for sample measurement has to be adjusted during the aging process of the channeltron (typically requiring increased high voltages). The ion beam produces a pulse in the channeltron (about 15 mV) which is amplified by about a factor of 10 in a preamplifier (Ortec, 9301), and another factor of 10 in the main amplifier (Phillips, 9650). The amplified signal is processed by using a discriminator (Phillips, 6930) and registered by a 100 MHz timer - counter unit (Ortec, 996). A 0.4 cm3 air standard results in a 3He count rate of about 2000 cps.
Fig. 12: Channeltron count rate versus high
voltage. Note extended scale for background count rate (right
Y-axis).
Before reaching the channeltron the 3He beam passes a velocity filter (WARP spell out the acronym filter from Micromass) to reduce the background by stopping scattered ions (tritium mass spectrometer only). This filter acts also as a protection for the channeltron. In case of high channeltron count rates, it blocks all incoming ions.
Tuning and peak shapes
The sensitivity of the mass spectrometers is mainly a function of the source parameters: repeller, electron energy, focus, beam center and source magnet position. The peak shape, on the other hand, is mainly determined by the position of the separation magnet. The position of the magnet can be adjusted in 3 dimensions. Additionally, there is an adjustable "pole shoe". The observed helium peak shapes (Figs. 13 to 16) are also a function of the ratio of the aperture to the peak width ratio, which determines the flatness of the peaks.
Fig. 14: 3He and HD/H3 peaks measured with
a channeltron shortly after a bakeout (0.4 cm3 air standard).
Fig. 15: 3He peak measured with a
channeltron (0.4 cm3 air standard). Same 3He peak as in Fig. 4; scale
for background count rate extended by a factor of 100.
Fig. 16 3He and HD/H3 peaks measured with a
channeltron 3 days after a bakeout
The 3He peaks are narrower by about a factor of three compared to 4He due to the different apertures (2 mm in front of the Faraday cup and 0.6 mm in front of the channeltron). In Figs. 14 and 16, the left peaks correspond to mass 3He (3.01603 AMU) and the right peaks are a combination of HD (hydrogen-deuterium molecule, 3.02193 AMU) and H3 (3.02348 AMU). The separation of the 3He peak from the other peaks with mass three is an essential feature of a mass spectrometer for 3He measurements. The scans plotted in Figs. 14 and 15 were produced a few days after a system bakeout following a change of both the filament and the channeltron. The HD/H3 peak decreases significantly over a period of about one week (Fig. 16 ) and reaches a fraction of a 0.4 cm3 STP air standard count rate after several weeks.