Applikation Plasmacleaner

Für nachfolgenden Vortrag anläßlich der Dreiländertagung 
für Elektronenmikroskopie in Innsbruck im September 2001 bedanken wir uns bei
Herrn Dr.rer.nat.habil. Bauer IFW Dresden
aufs Herzlichste

Institut für Festkörper- und Werkstofforschung, Dresden

H.-D. Bauer, IFW Dresden
K. Binder, Hebertshausen
W. Send, Karlsruhe

Plasma Cleaning
for Analytical TEM

Contamination by hydrocarbons


in the microscope free movable
hydrocarbon molecules, will be
linked on the surfaces, if there
hitted by electrons


  • vacuum oil backstreaming
  • vacuum grease and O-rings
  • microscope walls and specimen cartridge
  • specimen

Influencing factors

  • pretreatment
  • partial pressure
  • e current density
  • temperature of specimen and environment


  • heating
  • cooling
  • spraying slow electrons
  • irradiation with UV light
  • oxidation in O-plasma

  • The trouble with specimen
    contamination is as old as the EM y
  • The mechanism is, that ... (for ... see on top)
    molecules will be adsorbed and desorbed on surfaces,
    but linked ...
  • Sources of hydrocarbons can be ...
  • Even in the best and cleanest vacua, the
    main and ultimate source for
    contamination is the specimen itself,
    with the deciding role of its pre-treatment
    by specimen preparation.
  • Influencing factors are also ...
  • A series of methods are proved to be
    against contamination, as ...
    • heating the specimen, if compatible, up to some
      100 centigrades
    • cooling the specimen, to hinder the surface
    • cooling the environment, to decrease the partial
      pressure of contaminants.
  • Newer method = transformation of
    hydrocarbons into flying components by
    oxidation within a plasma, as last
    treatment just before inserting into the

For this purpose, we have a plasma

  • He works with an atmosphere of
    ordinary air or a mixing of Ar and O,
    with a pressure less than 1 mbar.
  • A coil on a glass cylinder generates a
    middle frequency magnetic field , as
    part of an electromagnetic swing circuit.
  • The induced electric field accelerates
    the charged particles in the atmosphere.
  • And so, an ICP burns between the
    cylinder and a shield grating.
  • In the environment, atoms and ions of
    Oxygen strike the specimen, with
    energies of 10 ...20 eV.
  • The specimen holder is placed
    • in the centre for intense treatment
    • or beneath for common use without
      obvious sputtering effect.

Here, a sight of our apparatus
consisiting of
  • turbo pump stand
  • reaction chamber with specimen holder
  • plasma generator at the top.
    Pay attention to a size-comparison with our
    table-lamp (in the corner)

Coarse contamination on metallic glas ZrAICuNIFe

contamination spots,
generated in the TecnaiF30
60 min HRTEM investigations

bild durch anklicken vergrößern

after 10 min cleaning

bild durch anklicken vergrößern

after 30 min cleaning

I show you here

  • as a TEM specimen a thinned foil
    of metallic glass after intense
    investigation during 60 min. in the
    Tecnai microscope.
  • On the specimen many typical
    contamination spots have been
    • We distinguish
      high contamination needles,
      produced by nanoprobe
      irradiation, and
    • flat contamination discs, by a
      wide irradiation, often with a thicker
    • the tracks come from scanning
  • After plasma cleaning during 10
    min. all spots are little reduced.
    See for instance in the marked region!
  • After 30 min. cleaning the flat
    cont. spots are killed, but from the
    needles distinct rests remain.
spots in STEM
modus by fixed electron probe
exposition 20s or 60 s
TEM modus, < 1 min.

bild durch anklicken vergrößern

after 10 min cleaning:

bild durch anklicken vergrößern

after 30 min cleaning:
further reduction
  • The behaviour of contamination
    needles was investigated in detail.

  • Here is a series of spots, generated
    by a fixed electron probe in STEM
    modus, with irradiation times of
    • 20s (top) and
    • 60 s (bottom).

  • After it, the spots have been
    observed in the TEM modus.

  • The right column was produced
    without liquid nitrogen in the
    cooling trap of specimen chamber.

  • The left column with cooling, the
    spots are some smaller.

  • Now, the specimen has been
    transferred into the plasma and
    treated for 10 min.
    All needles are reduced in

  • We tried to generate 6 new spots
    under the same conditions. As
    effect of plasma cleaning, the
    irradiated region remains empty.

  • As expected, after 30 min.
    cleaning the needles are further
bild durch anklicken vergrößern

TEM image, cross section
marked: area for EELS

Slow contamination
on TiN/AIO multilayer

bild durch anklicken vergrößern

without treatment
integration time
120 s

bild durch anklicken vergrößern

after cleaning
10 min.
integration time
120 s
Energieverlust in eV
  • The described experiments refer to
    high exposition doses, where the
    contamination layers are visible
    directly at the micrographs.

  • More frequently, by small
    irradiation doses very thin layers
    grow. Such layers are unvisible,
    the detection is only possible by

    The transparency presents an

  • The energy loss spectrum of
    marked region shows the elements
    N, Ti and O. Additionally, during
    the intentional long recording time
    of 120 s, also a C-edge arises,
    showing contamination.

  • After that, the specimen has been
    plasma cleaned for 10 min.

  • Now, in the loss spectrum, the C-edge
    is vanished, and no new edge
    can be generated during the 120 s.

Contamination on SiO film:

  • To interpret the results
    quantitatively, we follow the
    growth of contamination layers by
    thickness measuring also.

  • On a Si-oxide film we produce the
    contamination by STEM imaging.

  • With exposure times between 2
    and 6 s we write quadratic fields.
    one side of fields is thicker
    because of the beam reversal.

  • The thickness of border was
    measured by Low Loss EELS.

  • There is a well known relation
    • the scattered intensity and the
      zero loss intensity at the one
      hand and
    • the specimen thickness and the
      mean free path for electrons on
      the other hand.

  • The result is displayed in the
    • This is the relative thickness of the
      pure SiO2 film before the
    • With proceeding exposition the
      thickness increases
      proportionallyto the time. We
      calculate a growth velocity
      5 nm/min.
    • After plasma cleaning for 10 min.
      the velocity is unmeasurable small.

bild durch anklicken vergrößern

ESI: red = C, blue = O

bild durch anklicken vergrößern

after cleaning 15 min

bild durch anklicken vergrößern

after cleaning 15 min
  • Here, you can see the killing of the
    just discussed spot series,
    demonstrated by ESI:
    • red = Carbon
    • blue = oxygen (as containment
      of silicon-dioxyde)
    • C on O results in magenta.
  • The upper micrograph (without
    cleaning) proves ones more the
    fact that in contamination fields
    generally the center is thinner and
    all borders are thicker.
  • That is because the hydrocarbons
    have been delivered subsequently
    by surface diffusion from outside.
  • After cleaning 15 min. the fields
    are killed exceptly the corners,
    which are especially thick by
    electron probe reversal.

Plasma Cleaning: Destruction of Carbon films

TEM: ceramics-particles on carbon

bild durch anklicken vergrößern

generation of a needle
by nanoprobe during 20 s

bild durch anklicken vergrößern

after cleaning 1.5 min:
no new generation during
100 s inside of maked region

bild durch anklicken vergrößern

after cleaning 4.5 min:
evident reduction

bild durch anklicken vergrößern

after cleaning 10.5 min:
considerable destruction
  • We know, the plasma cleaning is
    enable to remove the hydrocarbon
    layers by oxidation.
  • In this context, the question is of
    interest, what happens, if carbon
    films are used in the TEM as
    specimen support?
  • the answer is given by the
    following experiment.
  • With an electron nanoprobe a cont.
    needle is generated by shifting the
    probe during 20 s.
  • After plasma treatment for 1.5
    min. the needle is reduced. No new
    needle can be generated even
    during 100 s. We tried it inside of marked
  • After cleaning for 4.5 min. the
    needle is further reduced, but the
    supporting film is nearly
  • Only after plasma treatment for 10
    min. the supporting film is
    destructed considerably.
  • Conclusion: Also carbon films can
    treated in the plasma cleaner,
    however careful!


  • By the used Plasma Cleaner with air or Ar/O2 during 3
    to 30 min.
    • existing hydrocarbon contamination layers can be
      removed without evident sputtering,
    • development of new contamination layers can be
      hindered effectively.
  • If the specimen consists of carbon, the Plasma cleaning
    should be restricted to 5 min.