Page 292 - Inorganic Mass Spectrometry - Fundamentals and Applications
P. 292

2 78





            The first  real  indication of the possible  utility of the rf powering  mode for “ana-
                                               in
            lytical”  chemical  applications  was  presented a  review article by  Coburn  and  Har-
            rison in 1981 [ 1 l]. In  a  way,  the  authors  provided  a  bridge  between  two  sides  of  a
            river:  Coburn  viewed  mass  spectrometry as  a  tool to understand funda~ental
            processes  occurring  in  plasma  deposition  sources,  whereas  Harrison,  whose  back-
                                                                  it
            ground  was  in  spark  and  plasma  source  mass  spectrometry,  viewed as  a  tool for
            spectroche~cal analysis.  At  that  time,  rf CD-MS was  simply  a  tool of physicists
            and  engineers,  but the ability  to  derive  composition-  and  depth-resolved  informa-
            tion  about  an insulating target  was  clearly  presented.  Soon after that publication,
            a  single  report  describing  the  use  of  rf powering for a  hollow  cathode  GD-MS
                                  ion
            source  [termed  an  ~ c ~ v i ~source (CIS)]  was  presented  by  Donohue  and  Har-
            rison  in  1975  [62]; in it mass  spectra  derived  from  metallic,  glass,  and  solution
                                                 of
            residue  samples  illustrated  the  possible  utility the  approach, The analytical  per-
            formance  (i.e., spectral characteristics,  sensitivity,  and  stability) of the  source  was
            found to be equivalent to or  better  than  that of the  much  more  fully  developed dc
            powered  hollow  cathode  sources,  suggesting  a  great deal of promise.
                 Despite  the  possible  utility  demonstrated  through  the  previously  cited  works,
            the  field of rf CD-MS lay  dormant for almost 15 years  until  a  paper by Duckworth
            and  Marcus  reintroduced  the  concept  in  1989  [26].
                                                    In that  paper, the operating  prin-
            ciples and  basic  design criteria for “analytical” rf GD-MS  sources  were  presented
            for a simple  diode  source  used for the  analysis of 0.5-inch-diameter  sample  disks,
            including  metals,  solid  glass,  and  compacted  (without  binder)  metal  oxide  pow-
            ders. In addition to illustrating the flexibility of the rf CD ion  source,  data  were
            presented  that  indicated  a  more  complex inte~elationship between  discharge  con-
                                                                 for
            ditions  and  ion  extraction  (sampling)  position  than  had  been  seen dc powered
            sources. The requirement to optimize  ion  sampling  position  and  discharge  condi-
            tions, along with  a  need  for efficient  sample  interchange,  has  led  to the  imple-
            mentation of  more  user-friendly  designs  based on 4.5-inch-diameter direct in-
            sertion  probes  (DIPs) [27].  Figure  7.7 illustrates the  general  approaches  used  in rf
                                           by
            GD-MS  source  design  as  implemented the Clemson  University  laboratory.  Use
            of DIPs provides  a  means of mounting  the  sample,  providing  electrical  contact,
            position optimi~ation, and  introduction  through  a  vacuum  interlock.  All  subsequent
                                                of DIP approach. On the basis of the
            rf  GD-MS  sources  have  employed  some  sort
            diameter of the probe, implement in^ sample  holders for pin-shape  and  small  disk
             samples is relatively straightfo~ard (Fig.  7.7a). A number of  groups  have  de-
                                 of
                                                                         to
             scribed  such  designs.  All these  designs  are  somewhat  different  in  respect the
            ion  volume  and  spectrometer  geometries  (as  well  as  investigators’  preferences);
            hovyever, the opti~zation of discharge  and  sampling  positions  are  remarkably  sim-
             ilk.
                 Given the wide  diversity of solid  sample  forms  and  the  particular  difficulty
             of machining oxide (e.g.,  glass  and  ceramic)  samples to a  fixed form, rf GD  ion
   287   288   289   290   291   292   293   294   295   296   297