Ellis A.M., Feher M., Wright T. Electronic and Photoelectron Spectroscopy: Fundamentals and Case Studies
Подождите немного. Документ загружается.
12 Photoelectron spectroscopy
105
The two most important properties of a photoelectron spectrometer are its resolution
and sensitivity. Although intrinsic factors do play a role, especially Doppler broadening, the
major factor affecting resolution is instrumental in origin. The main limitations on instru-
mental resolution are the dimensions of the analyser, the widths of both the entrance and
exit slits, as well as other factors, such as the presence of outside electric or magnetic fields
and local charges inside the spectrometer (e.g. from surface contamination). The resolution
can be improved by decreasing the entrance and exit slit widths, but this necessarily impairs
the sensitivity. Thus a trade-off between good sensitivity and acceptable resolution is nec-
essary, and the compromise that is normally taken yields a resolution in the 10–30 meV
range (∼80–240 cm
−1
). This is clearly much worse than that routinely obtained in optical
spectroscopy, and is such that it may even be a struggle to achieve full vibrational resolu-
tion for small molecules. Rotationally resolved spectra are not practical with conventional
photoelectron spectroscopy.
12.2 Synchrotron radiation in photoelectron spectroscopy
There have been a number of important experimental developments in photoelectron spec-
troscopy over the years. One of the most significant has been the widespread use of syn-
chrotron radiation. In fact synchrotron radiation has many other applications in science and
technology. Synchrotron radiation is produced from electron storage rings. In outline, a
burst of electrons is injected into a storage ring and confined to a near-circular path by a
series of magnets. The electrons, travelling at speeds close to that of light, generate intense
radiation as they accelerate around the ring and this radiation can be extracted for various
experiments. The construction of synchrotrons requires major financial investment. They
are essentially large particle accelerators and it is therefore only feasible to operate them as
central facilities. Experimental stations, known as beamlines, are located at various points
around the storage ring, as illustrated in Figure 12.2. The investigator travels to the syn-
chrotron to carry out experiments and will use the radiation output, together with any other
imported or permanent equipment, at one of the beamline stations.
The key properties of synchrotron radiation for photoelectron spectroscopy are: (i) it
is continuous over a wide wavelength range (10
−10
–10
−5
m); (ii) it is highly intense;
(iii) the radiation is plane-polarized. A specific wavelength is necessary for photoelectron
spectroscopy and so a suitable monochromator is placed in front of the spectrometer. The
plane-polarized nature of synchrotron radiation is important in angle-resolved work, i.e. in
studies where the intensity of electrons is measured at various angles relative to the plane
of polarization. Photoelectron angular distributions can provide important information on
photoionization dynamics.
12.3 Negative ion photoelectron spectroscopy
The most weakly bound electron in a singly charged anion has a binding energy equal to the
negative of the electron affinity of the atom or molecule. The electron affinity of an anion
is analogous to the ionization energy of a neutral species, but the former is normally much
106 Experimental techniques
Bending magnet
Main synchrotron
ring
Linac
Beamlines
Booster
synchrotron
Figure 12.2 Schematic illustration of a synchrotron radiation source. Electrons, produced by ther-
mionic emission from a heated cathode, are injected into the storage ring after acceleration up to very
high speeds by a linear accelerator (linac). A series of bending magnets situated at various positions
along the ring force the electrons to adopt a roughly circular path. As the electrons traverse the bending
magnets their acceleration produces light emission that can be exploited in a beamline. Experimental
stations are located at the end of each beamline.
smaller than the latter. For example, the highly electronegative Cl atom has a first electron
affinity of only 3.7 eV, and this is large by the standards of most atoms and molecules. One
can immediately see that if photoelectron spectroscopy is attempted on anions, relatively
low photon energies can be used to remove an electron. Indeed, visible lasers, such as a
selected single line from an argon ion laser, are often used as the light source.
Apart from the light source, the essential components in negative ion photoelectron
spectroscopy are similar to those in conventional photoelectron spectroscopy. The only
other experimental difference concerns the production of anions, which can be achieved
in a number of ways. The most common source is a gas electrical discharge, which will
make a variety of species including neutrals, cations, and anions. The low photon energy
will only remove electrons from anions, and so the presence of neutrals and cations will not
cause any complications in the spectroscopy. Other sources have used dissociative electron
attachment (an electron attaching to a neutral molecule followed by rapid dissociation to
form a neutral fragment and an anionic fragment) or sputtering of anions off solid surfaces
by energetic particle bombardment.
Negative ion photoelectron spectroscopy provides similar information to conventional
photoelectron spectroscopy with one important difference. While traditional photoelec-
tron spectroscopy is performed on neutral atoms or molecules and yields their ionization
energies together with spectroscopic constants of the corresponding cations, negative ion
photoelectron spectroscopy provides electron affinities and spectroscopic constants of the
neutral species. Case Study 15 considers a specific example in some detail. There are many
12 Photoelectron spectroscopy
107
examples of neutral molecules that were first studied by negative ion photoelectron spec-
troscopy. In some cases the spectroscopic constants obtained were subsequently used to
guide the search for higher resolution optical spectra of these molecules.
12.4 Penning ionization electron spectroscopy
Penning ionization electron spectroscopy is somewhat similar to photoelectron spectroscopy.
The principal difference is that instead of using ultraviolet photons to ionize a sample,
ionization is brought about by collisions with metastable excited atoms. Metastable species
are atoms or molecules in excited states that have long lifetimes, sometimes as long as
several seconds, because the transition back to the ground state is forbidden by one or
more optical selection rules. For spectroscopic purposes, metastable noble gas atoms are
used, especially Ne (
3
P
2
), which has an energy 16.62 eV above ground state Ne. Metastable
noble gas atoms can be produced in a carefully controlled electrical discharge or by using
electron impact with a high energy (80–100 eV) electron beam. In both cases, care must be
exercised to remove charged particles from the metastable atom beam and to prevent light
from the discharge reaching the ionization region. On collision with the sample molecules
their excess energy is used for ionization:
M + Ne
∗
→ M
+
+ Ne + e
−
The electrons are then analysed according to their kinetic energies as in photoelectron
spectroscopy. The similarity in experimental conditions is such that it is possible to perform
Penning and photoelectron spectroscopy in the same apparatus under identical conditions.
12.5 Zero electron kinetic energy (ZEKE) spectroscopy
ZEKE spectroscopy was introduced in 1984 and has developed into an important spec-
troscopic technique. It is an example of threshold photoelectron spectroscopy, so-called
because the aim is to photoexcite molecules to a specific energy level of the ion (an ioniza-
tion threshold), which will produce electrons with very low (or, in principle, zero) kinetic
energy. A tunable light source is necessary for threshold photoelectron spectroscopy.
The basic idea is shown in Figure 12.3. When the photon energy exactly matches the
energy difference between a specific level of the neutral and a specific level of the ion,
excitation to that energy level of the ion must produce electrons with zero kinetic energy by
conservation of energy. However, the ion may also end up in lower energy levels of the ion (if
any are available), and the emitted electron will therefore take up the excess energy.
3
Conse-
quently, some electrons will be produced with zero kinetic energy and others with non-zero
kinetic energies. We will call the former ZEKE (pronounced ‘zee-kee’) electrons. At other
wavelengths, where the photon energy does not precisely match a neutral–ion energy level
3
The probability of populating the various vibrational levels in the ion is, to a good approximation, governed by
the Franck–Condon principle.
108 Experimental techniques
Cation energy
levels
Neutral molecule
hn
hn¢
ZEKE electron
Figure 12.3 Basic principles of threshold ionization. In the process shown on the left, electrons can
be produced with both zero and non-zero kinetic energies. However, on the right the energy mismatch
between the photon energy and the separation between ion and neutral molecule energy levels means
that only electrons with non-zero kinetic energies are produced.
separation, photoionization will still take place but electrons with zero kinetic energy cannot
be produced. Consequently, if it were possible to preferentially detect ZEKE electrons as
a function of the wavelength of the tunable light source, then an ion←neutral excitation
spectrum could be obtained. This is the basic idea of threshold photoelectron spectroscopy.
Threshold photoelectron spectroscopy was around for some years before what is now
known as ZEKE spectroscopy was introduced. Many clever schemes for discriminating
between zero and non-zero kinetic energy electrons were developed, and some of these ideas
were subsequently employed in ZEKE spectroscopy. However, while much important work
was carried out with the pre-ZEKE forms of threshold photoelectron spectroscopy, there
were a number of problems, a notable one being complications caused by autoionization.
4
The introduction of ZEKE spectroscopy in 1984 combined the use of pulsed lasers
for ionization together with a delayed pulsed electrical field method for detecting ZEKE
electrons. The basic idea is simple and is illustrated in Figure 12.4. Suppose a tunable pulsed
laser is capable of ionizing a molecule. As its wavelength is scanned it will move in and out
of resonance with ion←neutral transitions. Any ZEKE electrons produced will be stationary
whereas non-ZEKE electrons (i.e. moving) will drift rapidly out of the original ionization
volume. It is possible to discriminate between ZEKE and non-ZEKE electrons by applying
a pulsed electric field across the ionization volume. This is achieved by sandwiching the
ionization volume between two conducting plates. If the electric field is initially zero, and is
then pulsed on shortly after the laser pulse, then electrons will be attracted towards the more
positively charged plate. Furthermore, if there is a small hole in this positive plate then the
accelerated electrons can pass out of this region and go on to reach the detector. However,
4
Autoionization is a spontaneous ionization process that can occur for neutral molecules in excited electronic states
lying above the lowest ionization limit.
12 Photoelectron spectroscopy
109
Figure 12.4 Basic arrangement for a ZEKE experiment. This assembly is mounted inside a high
vacuum chamber. Non-ZEKE electrons are emitted in all possible directions and drift away from the
initial ionization volume. The voltage pulse, applied after a predetermined delay of c.1s, sends
both ZEKE and non-ZEKE electrons into the flight tube. The non-ZEKE electrons that are detected
are those moving towards or away from the flight tube along the central axis prior to application of the
voltage pulse: all non-ZEKE electrons with an off-axis velocity component drift too far away from the
ionization volume and when the voltage pulse is switched on are unable to pass through the aperture
in the middle plate. The non-ZEKE electrons that do reach the detector are easily distinguished from
the ZEKE electrons by their different flight times, as shown in the inset.
the movement of the non-ZEKE electrons means that these will start from different positions
in the inter-plate region when the electric field is pulsed on compared with ZEKE electrons.
Consequently, their arrival times will differ, as illustrated in the inset of Figure 12.4.
Most early ZEKE experiments used resonance-enhanced multiphoton ionization with
pulsed tunable dye lasers rather than single-photon ionization. However, there have now
been many single-photon ionization ZEKE experiments with VUV laser radiation generated
from specialized harmonic generation processes or from synchrotron radiation. In both cases
the time delay between pulsed laser ionization and application of the pulsed electric field is
normally in the region of 1 s. Although both ZEKE and non-ZEKE electrons are registered
by the electron detector, all non-ZEKE signals are subsequently discarded.
The great advantage of ZEKE spectroscopy is that electron energy analysis is not
required. Consequently, the primary cause of the low resolution in conventional photo-
electron spectroscopy disappears. In some of the most favourable cases a resolution as good
as 0.2 cm
−1
has been achieved, allowing rotationally resolved structure to be obtained.
This technique may be used for neutrals or anions, but for the former a much more robust
variant is possible and is described in the following section.
110 Experimental techniques
12.6 ZEKE–PFI spectroscopy
ZEKE–PFI is a variant of ZEKE spectroscopy and nowadays the distinction between the two
is often ignored. The role of the pulsed electric field in ZEKE spectroscopy is to accelerate
the electrons towards the detector once the ZEKE and non-ZEKE electrons have separated
in space. In ZEKE–PFI, the electric field actually causes ionization, hence the abbreviation
PFI for pulsed field ionization.
Molecules contain a largenumber of energylevels, known as Rydberglevels, close to their
ionization limit. A molecule in a Rydberg state has an electron in an orbital that is so diffuse
that it ‘sees’ the ionic core as virtually a point charge. Thus a Rydberg orbital is similar to
an orbital of atomic hydrogen. Just below the ionization threshold, within approximately
5cm
−1
, the lifetimes of molecular Rydberg states become quite long, sometimes tens of
microseconds, owing to interactions with small electric and magnetic fields present in the
apparatus. These Rydberg states can be ionized by application of a small electric field,
which pulls the Rydberg electron free from the ion core. This is the underlying principle of
ZEKE–PFI spectroscopy. Instead of using threshold ionization, the molecules are excited up
to Rydberg levels close to threshold by the pulsed laser(s). After a short delay, a microsecond
or so, a small electric field is pulsed on and the Rydberg states are field-ionized to produce
zero kinetic energy electrons. Of course direct ZEKE electron production is also possible,
but the ZEKE and ZEKE–PFI electrons can be distinguished by their arrival times at the
detector by application of a very small dc electric field across the ionization region prior to
the field ionizing pulse.
ZEKE and ZEKE–PFI spectra contain essentially the same information. There is also
another variant of the ZEKE method that has recently been introduced, mass analysed
threshold ionization (MATI), in which ions rather than electrons are detected. As the name
implies, this allows one to record a mass-selected ZEKE spectrum.
A detailed account of ZEKE and related spectroscopic techniques has been given by
M¨uller-Dethlefs and Schlag [1].
Reference
1. K. M¨uller-Dethlefs and E. W. Schlag, Angew. Chemie Int. Ed. 37 (1998) 1346.
Further reading
Further information on spectroscopic techniques can be found in the following books.
Laser Spectroscopy, 3rd edn., W. Demtr¨oder, Berlin, Springer-Verlag, 2002.
Photoelectron Spectroscopy,J.H.D.Eland, London, Butterworths, 1984.
Principles of Ultraviolet Photoelectron Spectroscopy,J.W.Rabalais, New York, Wiley,
1977.
Part III
Case Studies
13
Ultraviolet photoelectron
spectrum of CO
Concepts illustrated: vibrational structure and Franck–Condon principle; adiabatic and
vertical ionization energies; Koopmans’s theorem; link between photoelectron spectra
and molecular orbital diagrams; Morse potentials.
Carbon monoxide was one of the first molecules studied by ultraviolet photoelectron spec-
troscopy [1]. A typical HeI spectrum is shown in Figure 13.1.
1
The spectrum appears to
be clustered into three band systems. The starting point for interpreting this spectrum is to
consider the molecular orbitals of CO and the possible electronic states of the cation formed
when an electron is removed.
13.1 Electronic structures of CO and CO
+
Any student familiar with chemical bonding will almost certainly be able to construct a
qualitative molecular orbital diagram for a diatomic molecule composed of first row atoms.
Such a diagram is shown for CO in Figure 13.2. The orbital occupancy corresponds to
the ground electronic configuration 1σ
2
2σ
2
3σ
2
4σ
2
1π
4
5σ
2
. The σ MOs actually have σ
+
symmetry but it is not uncommon to see the superscript omitted. Since all occupied orbitals
are fully occupied, the ground state is therefore a
1
+
state and, since it is the lowest
electronic state of CO, it is given the prefix X, i.e. X
1
+
,todistinguish it from higher
energy
1
+
states of CO.
Consider the electronic states of the cation formed by removing an electron. If the electron
is removed from the highest occupied molecular orbital (HOMO), the 5σ orbital, then the
cation will be in a
2
+
state. Since this is expected to be the lowest energy state of the cation,
it is therefore labelled X
2
+
. Removing an electron from the 1π or 4σ MOs gives
2
and
2
+
states, respectively. From the orbital ordering in the MO diagram, our expectation is
that these two states are the lowest energy excited electronic states of CO
+
and so will be
labelled as the A
2
and B
2
+
states.
1
HeI radiation has a wavelength of 58.4 nm (≡ 21.2 eV) – further details can be found in Section 12.1.
113
114 Case Studies
20 19 18 17 16 15 14 13
Ionization ener
gy
/eV
Counts/second
(1p)
−1
(5σ)
−1
(4σ)
−1
Figure 13.1 HeI photoelectron spectrum of CO.
COC
2s
2p
2p
2s
5
σ
1
p
4σ
3σ
2p
6
σ
Energy
O
Figure 13.2 Qualitative molecular orbital diagram for CO. Only the valence orbitals are shown, i.e.
the 1σ and 2σ orbitals formed by overlap of the 1s orbitals of C and O have been omitted since they
are core orbitals and cannot be photoionized by HeI radiation.