Our basic explanation is split into several steps
|It is a question often asked: 'How is VPPEM different to PEEM'.
PhotoElectron Emission Microscopy (PEEM)
More information on how it works
The VPPEM optics is unique:
The angular image exiting from the VPPEM magnetic field aperture is
an unusual type of image, only a field emission microscope (FEM)
could be considered to have a similar form.
In our context, we define the term angular image in that the two-
dimensional (X,Y) spatial information from the sample is encoded as
a mapping onto a two-dimensional angular distribution.
The two dimensions being: θ the off-axis angle or inclination, and φ
the azimuthal direction around the optical axis. The angular image
coming out of the magnetic field aperture can be seen as the electron
optical equivalent of the image within a pinhole camera. Each ray from
a different point of an object comes through the pinhole at a different
angle, and is then projected onto a screen. A phosphor screen placed
in front of the magnetic field aperture would provide a similarly
projected image if the electrons were monochromatic.
The fact that the signal is encoded as an angle after leaving the
magnetic field has a large impact on the design of the electrostatic
section of the electron optics.
In comparison with the electron optics of the cathode lens class of
photoelectron emission microscope, PEEM we can see VPPEM is
very different to PEEM.
Although we are ultimately using the angular information, the image
coming out of the field aperture is actually a mixture of spatial, and
angular information. The result is that the CHA input lens does not
really ‘focus’ into the CHA. The input lens produces a non-Gaussian
somewhat caustic looking focus.
However, these apparent aberrations are in space not in angle.
Similarly, for the intrinsic chromatic aberration of the CHA. The CHA is
double focusing in θ and φ, however, the energy-dispersed image at
the exit of the CHA is not like the original pinhole image but is spread
out across the CHA output aperture in the dispersion direction.
We can still form a good image with the CHA output lens because we
are focusing on an angular image which is a virtual image at infinity.
The output lens is a telescope with the detector at the prime focus.
In contrast, a PEEM instrument that includes energy spectroscopy
must somehow compensate for the energy analyzer chromatic
aberration on the real image. This is done in one of two ways, either
to move the real image back away from the CHA, and use a virtual
image through the CHA, or to use a double CHA to compensate for
the image shift]. Both approaches are complex in the electron optical
design and implementation.
One issue is that the cathode lens microscope uses a large
accelerating voltage to form the image. A design is required to use
electrostatic decelerating lenses to inject the image into the CHA.
This is because the CHA requires a relatively low pass energy to give
sufficient energy resolution. Decelerating lenses do two things, one
is to increase the angle of the image, and thus increase aperture
size, and second, electrostatic decelerating lens are very aberrating.
The full solution to this problem is a remarkable series of
instruments PSMART PEEM /LEEM BESSY, and NanoESCA.
However, these instruments are significantly more complex than a
VPPEM design. Using an angular image at low voltage gives us the
advantage of simplicity in the design of the electron optics.
From a consideration of the electron optics, and the mechanism of
image formation, it can be seen that VPPEM does not lend itself
naturally to many of the important measurements that are the main
elements of PEEM investigations.
For example, the collimating action of the magnetic field rules out
measurement of the k-vector, and because of the saturating effect of
the magnetic field we cannot expect to see any useful information
using magnetic dichroism.
Also, as we have designed in a near normal incidence for the
photons onto the sample, using the polarization of the photons to
look at surface electron orbital orientation would require a large
We are using partial yield NEXAFS as the primary imaging signal. For
this the main task of the spectrometer is to discriminate low energy
electrons in the range 0f 0.5-1.5 eV to maximize the spatial resolution.
However, with a 50 mV energy resolution we should also be able to
image surface potentials using a series of images scanning the
sample voltage over the range of a few volts.
We can also image the directly emitted photon electron which in
principle can give us some form of depth profiling by changing the
detected emission energy, and thus the escape depth in the solid.
We should note that there are many aspects of the VPEEM optics that
have not been fully explored experimentally. These include space
charge, sample charging issues, sample voltage differences,
topographic effects, and sample related spatial resolution.
VPPEM will not be susceptible to image degradation due to space
charge effects, for although it has in effect a ‘back plane’ crossover of
the image at the CHA entrance, this is not a true crossover. This
crossover has a both spatial and angular size.
PEEMs have a contrast aperture in the focal back plane to restrict the
energy and angular spread of electrons, allowing tradeoff between
resolution and transmission. This crossover does introduce the
problem of space charge that ultimately degrades the image for high
brightness synchrotron photon sources. There is no such angle
limiting mechanism for the VPPEM. Electrons leaving the surface at all
angles are constrained to travel down the magnetic field lines, and are
then collimated as the field decays.
Sample charging will not effect VPPEM in the same way that it does
PEEM. Sample surface topography effects which are strong with
PEEM are relatively insignificant with VPPEM.
This is partly because of the large depth of focus, but also because
the surface is not immersed in a strong electrostatic field. One
currently unresolved issue is the effect of sample properties on the
The calculated special resolution is determined by the electron
energy and the magnetic field. From equation 3, for a 1 eV electron
energy and a 2 Tesla field we expect a 20-80% resolution of 1.5
micron. However, we observe much higher resolutions on some
samples, and this is still not understood.
Looking at the electron optics, we expect that VPPEM will have a place
in the analysis of engineering materials which are not flat planes,
good conductors, or even consolidated samples such as powder
The electron optics of VPPEM and PEEM have very different imaging
properties, they can be seen as interestingly similar, but ultimately