Resonance Photoionisation Mass-Spectrometer for detection of ultra-trace amounts of Krypton in extraterrestrial samples
Krypton isotope structures in extraterrestrial material record the physical properties of s-process sites and the irradiation history of the early solar system.
The isotopic structure of s-process krypton in presolar SiC grains is of particular interest because it is affected by two branching points (where an isotope’s lifetime against neutron capture in the s-process site is comparable to its β-decay half-life). The branching ratio at 79Se is sensitive to temperature as well as neutron flux since the nucleus has a low lying excited state, while that at 85Kr depends only on neutron flux. Potentially, analyses of krypton in individual SiC grains thus allow both temperature and neutron flux of s-process to be investigated [1].
Cosmic ray spallation reactions cause a characteristic enrichment of specific noble gas isotopes in minerals [2-3]. A record of the irradiation environment of the early solar system may be accessible through analysis of the Kr system in chondrules.
Determination of elemental Kr/Xe ratios will allow aqueous and igneous processes to be traced and, after neutron irradiation, halogen geochemistry to be studied.
To address these issues we are developing an ultrasensitive (capable to measure ~1000 atoms) instrument for Kr isotopic analysis.
Experimental Procedure:
A Laser Resonance Ionization Mass Spectrometry (RIMS) method [4] is used for determination of isotopic ratios of krypton and elemental Kr-Xe ratios. The method is based on selective resonance excitation of atoms from level to level by pulsed lasers (pulse width ~10 ns) into a highly lying states followed by ionization. Created ions are separated according to mass by a time-of-flight (TOF) mass spectrometer. The experimental procedure includes the following steps:
- Laser step-heating or laser probe–releases sample atoms into the mass spectrometer volume.
- Condensation of atoms at localized cold spot (<60 K) [5].
- Desorption of condensate with 1064 nm Q-switched Nd:YAG laser, 10Hz duty cycle.
- Photoionization by a pulsed laser system - dye lasers pumped by a Q-switched Nd:YAG laser.
- Time of Flight Mass Separation.
- Ion detection by electron multiplier.
Fig. 1 Experimental setup and Kr photoionization scheme. Atoms are resonantly excited from the ground state by 116.5 nm vacuum ultra violet (VUV) light generated using four-wave non-linear mixing in Xe. A second resonant step uses 558.1 nm dye laser pumped by second harmonic (532 nm) of pulsed Nd:YAG laser. Fundamental 1064 nm light of the same YAG is used for ionization step.
VUV light generation by four-wave mixing in Xe/Ar mixture
The interaction of four coherent optical fields through the third order non-linear susceptibility known as four-wave mixing is well studied [6]. The application of this process for the resonant photoionization of Kr has been demonstrated at the Institute of Rare Isotope Measurement, University of Tennessee [7]. In our experiments two beams of 252.6 nm (obtained after doubling of 505 nm dye laser beam in BBO crystal) and 1507.3 nm (mixing of 632 nm dye laser beam with 1064nm of Nd:YAG laser in Lithium Niobate crystal) interact while passing through the cylindrical gas cell (19x180 mm). As a non-linear medium a Xe-Ar mixture is used. Because Xe is negatively dispersive and Ar is positively dispersive near the 116 nm spectral region [8], the intensity of generated VUV light will peak if a phase matching condition is met. This condition is:, ∆k=2k(252.6nm)+k(1507nm)- k(116.5nm)=0, where k(λi) is a wave vector depending on the refractive index. In out experiment, the partial pressures of Xe and Ar gases in the mixing cell (and thus the wave vectors) are varied until VUV generation is observed in an NO cell. Once VUV generation has been observed, the beam can be tuned to Kr resonance by changing the wavelength of IR laser and simultaneously tuning the Xe/Ar mixing ratio to the optimum for the new wavelength.
Fig.2 Four-wave mixing scheme and theoretical tuning range of VUV. Due to an allowed Xe transition into 5p5 7S [1/2]1 level VUV is absorbed at ~117 nm.
Kr and NO ionization chambers
Two cylindrical ionization chambers (70x126 mm) (referred to as “the cells” hereafter) have been developed for off-line spectroscopic measurements of Krypton and monitoring of VUV generation. They are made of standard conflat vacuum parts. Inside each cell there is a glass tube with two plane stainless steel electrodes (20x90mm). The cells and mass spectrometer are separated by MgF2 viewports. After pumping both chambers down to 10^-3 mBar the first one is filled with 1 mBar of krypton. Another cell is filled with 10^-1 mBar of nitric oxide (NO). The NO gas has a high ionization cross-section near the 116 nm spectral region [9]. When the VUV is generated in the Xe-Ar cell, NO gas becomes ionized. The charge collected at the cathode is amplified by a charge sensitive amplifier. The potential difference 70 V is chosen in such a way that both cells are working as proportional counters. The ionization signal is registered by Kr cell as the laser beams are tuned into resonance with atomic transitions used in Kr excitation scheme.
Figure 3 shows Kr and NO cells’ ionization signals detected under different experimental conditions: a) when the mixing cell is empty (no VUV generation occurred) no signal is detected by the Kr cell. The signal registered by NO cell is also negligibly small and is mainly due to ionization by 252.6 nm beam; b) when the generation cell is filled with the gas mixture that has the right Xe-Ar ratio (VUV is generated), a huge signal is detected by NO cell. A krypton signal appears after Kr cell is filled with ~20 Torr of gas, which denotes saturation of the first VUV transition of the photoionization scheme (provided the input wavelengths and Xe/Ar ratio are matched to generation of the correct wavelength). The filled Kr cell absorbs most of the VUV radiation so the NO signal intensity decreases; c) Kr ionization is enhanced after the second 558.1 nm laser beam is added into the photoionization scheme.
Fig.3 Kr and NO cells’ ionization signal detected under different experimental conditions. See the text for explanations.
Krypton and Xenon high sensitivity TOF spectra of air samples
The photoionization scheme described above has been successfully applied for measurements of krypton in our TOF mass spectrometer. The development of the mass spectrometer benefits from its prototype ‘RELAX (Refrigerator Enhanced Laser Analyser for Xenon)’ [10] that has successfully been used for determination of isotopic ratios of Xe in various extraterrestrial samples [11-12] Having optimized the experimental conditions we observed stable generation of VUV beam. The green and IR beams were aligned into the ion source to complete the ionization process. Small aliquots of air containing 10^6atoms of krypton were introduced into MS. The cold finger was operated at 60 K. The VUV power density was estimated <1 μJ/mm2. The power densities used for second and third ionization steps were 1 mJ/mm2 and 2 mJ/mm2 respectively.
Measured TOF spectra (Fig. 4) are the signal sum from 1000 laser shots that corresponds to ~2 min. collection time. The krypton photoionization yields decrease as second and third lasers are excluded from the scheme. No krypton ionization occurs without VUV generation (1507 nm laser is blocked). The remaining peaks at the spectrum d) belongs to the carbon clusters ionized during desorption from the cold finger.
Fig. 4 TOF spectra of krypton measured a) using all photoionization beams. In this spectrum the 84Kr peak corresponds to 5.7x10^5 atoms and the 80Kr peak to 2.3x10^4 atoms. b) with 1064 nm beam blocked, c) with 558 and 1064 nm beams blocked, and d) using only 252.5 nm beam; all beams including 1507.3 nm were blocked (no VUV generation).
For measurements of xenon atoms another ionization scheme was employed. The atoms are excited from ground state by two-photon absorption (252.5 nm, as used for frequency mixing) followed by the same wavelength ionization. Because of high power density required for saturation of two-photon transition the possibility of using the second laser to increase the ionization efficiency was investigated. Using Xe reference cell (PXe=100 Torr) two resonant transitions (558.25 nm and 562.21 nm) starting from (5p5(2P03/2)6p, J=2) level excited by two-photon absorption from ground state were found. In spite of gain in efficiency archived (the Xe signal increased by factor of ~4) it was still too low to be used for Xe sample analysis in our current configuration. The required power density has been reached by UV beam focusing into an ion source with 250 mm lens. We will eventually use a separate laser system in the same laboratory for simultaneous analysis of Xe and Kr isotopes.
Fig. 5 TOF spectrum of xenon ionized by focused 252.5 nm beam.
References:
[1]Lewis et al.(1992) Inst.Phys.Conf.Ser.No128,27-30.
[2]Eugster et al. (2002) MAPS 37, 1345-1360.
[3]Polnau et al. (2001) GCA 65, 1849-1866.
[4] Letokhov V.S. (1987) Laser photoionization spectroscopy. Academic Press, New York.
[5] Hurst G.S. et al., J. Appl.Phys., 55(5) (1984), 1278-1284.
[6] N.Bloembergen (1981) noble lecture.
[7] Thonnard et al. (1992) Inst.Phys.Conf.Ser. No128, 27-30.
[8] R.Mahon et al. QE-15, (1979) 444.
[9] K.Watanabe, F.M.Matsunaga, and H.Sakai, (1967) Appl. Opt. 6, 391.
[10] Gilmour J. D. et al. (1991) Measurement Science & Technology 2(7), 589-595.
[11] Gilmour J. D et al. (2006) Meteoritics & Planetary Science 41(1), 19-31.
[12] Gilmour J. D (2005) Geochimica Cosmochimica Acta 69(16), 4133-4148.
Appendix:
The TOF MS with VUV generation cell and ionization chambers attached:
The laser System:
The gas delivery system (under the MS table):
For further information e-mail: Ilya.Strshnov@manchester.ac.uk