Page 21 - Inorganic Mass Spectrometry : Fundamentals and Applications
P. 21

when ions collected are compared to atoms loaded; the value of this ratio is, of
     course, element- dependent.^ For many elrements, samples of  1 nanogram or less
     can be  routinely analyzed provided a pulse-counting detection system is em-
     ployed.
          A new thermal ionization source developed by Olivares and coworkers at
     Los Alamos National Laboratory is much more efficient at production of  ions
     than  the more conventional filament configurations described previously [37].
     Designed originally for use in isotope separators, it has been modified for use in
     isotope  ratio  measurements using  a  quadrupole mass  spectrometer. Called  a
      thermal ionization cavity, it is designed around a tungsten tube that contains the
      sample and is heated by electron bombardment. A drawing of this source is given
     in Fig. 1.5. Because of the amount of heat generated the source must be water-
      cooled, making it more complex than conventional designs. Ionization efficiencies
      are  very  high  in  comparison to  those of  normal  thermal  ionization  sources.
     Efficiency for europium is 72%, for uranium 8%, and for thorium 2%, all figures
      an order of magnitude or more better than single-filament results. Typical preci-
      sion for isotope ratio measurements is quoted as 20.1%.




      The two most common types of detection systems used in isotope ratio measure-
      ments are Faraday cups and pu~se-counting electron multipliers, Multipliers are
      occasionally used in cu~ent-inte~ation mode but are not as precise as Faraday
      cups or as sensitive as pulse counting; there seems to be no compelling need for
      this mode of operation in thermal ionization, All these collectors, of course, are
      used  in many different kinds  of mass spectrometry. Faraday cups give better
      precision and require more sample than pulse counting; choice of collector is often
      dictated by the demands of the experiment (or, of course, by available equipment).
      Each type of collector has its strengths and weaknesses. Constructing a Faraday
      cup that is quiet and linear is not a simple task. Pulse counting, in. which each ion is
      counted as it arrives at the collector, requires fast electronics that must be very
      quiet; typical background counting rates in good systems are about one count a
      minute,  approximately the  cosmic  ray  background. An  example  of  a  pulse-
      counting detection system is shown schematically in Fig.  1.6. Multipliers are
      operated at negative potential for detection of positive ions;  2-4  kV is typical in
      pulse-counting applications. Each dynode is successively nearer ground than its
      predecessor, so the potential gradient across the detector causes secondary elec-
      trons to travel through it from the first dynode to the’last; transit times of 1 nsec or
      so are typical. Ions strike the first dynode of  an electron multiplier; each such
      impact generates two or more electrons, resulting in amplification of the original
      signal. The electron cascade passes from one dynode to the next, each stage
      effecting amp~fication of the original signal, with the pulse from the last dynode
   16   17   18   19   20   21   22   23   24   25   26