A new detection scheme for synchronous, high resolution ZEKE and MATI spectroscopy demonstrated on the Phenol·Ar complex

Article abstract of DOI:10.1016/S0009-2614(99)01193-8

We present a new detection scheme for high resolution mass analyzed threshold ionization (MATI) spectroscopy which utilizes fractional Stark state selective electric field ionization of high-n Rydberg states. The method represents a significant advance over previous approaches to MATI spectroscopy, which have been characterized by their low resolution compared to zero kinetic energy (ZEKE) photoelectron spectroscopy. In addition, the scheme can be used to synchronously obtain electrons and ions originating from the same Rydberg states, facilitating the acquisition of synchronous, high resolution ZEKE and MATI spectra. We demonstrate the technique by obtaining MATI dissociation spectra of the phenol·Ar complex, which provide dissociation energies of 364±13, 397±13, and 535±3 cm^-1 for the S₀, S₁ and D₀ states, respectively.

Full text of DOI:10.1016/S0009-2614(99)01193-8

17 December 1999
Ž
.
Chemical Physics Letters 315 1999 103–108
www.elsevier.nlrlocatercplett
A new detection scheme for synchronous, high resolution ZEKE
and MATI spectroscopy demonstrated on the PhenolPAr complex
)
Caroline E.H. Dessent, Stephen R. Haines, Klaus Muller-Dethlefs
¨
Department of Chemistry, UniÕersity of York, Heslington, York, YO10 5DD, UK
Received 17 August 1999; in final form 11 October 1999
Abstract
Ž
.
We present a new detection scheme for high resolution mass analyzed threshold ionization MATI spectroscopy which
utilizes fractional Stark state selective electric field ionization of high-n Rydberg states. The method represents a significant
advance over previous approaches to MATI spectroscopy, which have been characterized by their low resolution compared
Ž
.
to zero kinetic energy ZEKE photoelectron spectroscopy. In addition, the scheme can be used to synchronously obtain
electrons and ions originating from the same Rydberg states, facilitating the acquisition of synchronous, high resolution
ZEKE and MATI spectra. We demonstrate the technique by obtaining MATI dissociation spectra of the phenolPAr
1
complex, which provide dissociation energies of 364"13, 397"13, and 535"3 cmy for the S0, S1 and D0 states,
respectively. q 1999 Elsevier Science B.V. All rights reserved.
1. Introduction
demonstrated that the spectral resolution of peaks in
the ZEKE spectrum of benzene could be improved
by more than a factor of eight using FSSFI.
MATI spectroscopy was introduced by Zhu and
Johnson as a variation of the ZEKE method that
allows mass analysis of ions produced by pulsed
w x
Since its introduction in 1984 1 , ZEKE photo-
electron spectroscopy has advanced considerably to
become a high resolution technique that is widely
applicable for studies of molecules and clusters 2–6 .
One of the most significant innovations of recent
years has been the realisation that fractional Stark
w
x
Ž
. w x
field ionization PFI 8 . The technique is particu-
larly useful for cluster studies where it facilitates
unambiguous identification of the ionized species,
and can be used to follow cluster fragmentation
Ž
.
state selection via electric field ionization FSSFI
allows a noteable improvement in ZEKE spectral
w x
resolution 7 . In this approach, an offset pulse selec-
w
x
9–12 . The major drawback with the MATI tech-
tively ionizes the more fragile ‘red’ Stark states,
while the more resilient ‘blue’ Stark states survive
and are shifted down in energy upon application of a
nique has been obtaining spectral resolution that can
compete with ZEKE resolution since relatively large
separation fields and field ionization pulses have
been employed to separate PFI ions from ions spon-
w x
second pulse of opposite direction. Dietrich et al. 7
w
x
taneously produced by the laser pulse 13 . In this
Letter, we report a new scheme for the synchronous
detection of high resolution ZEKE and MATI spec-
)
Corresponding author. Fax: q44-1904-434527; e-mail:
kmd6@york.ac.uk
0009-2614r99r$ - see front matter q 1999 Elsevier Science B.V. All rights reserved.
Ž
.
PII: S0009-2614 99 01193-8

(
)
104
C.E.H. Dessent et al.rChemical Physics Letters 315 1999 103–108
tra. While a relatively high resolution approach to
MATI spectroscopy has been suggested previously
The ZEKE electron and threshold ion signals were
obtained using electric field pulses generated by an
w
x
Ž
.
13 , this is the first time that FSSFI has been used to
arbitrary function generator Le Croy, AFG 9100 ,
which produces pulses with rise times of less than 5
ns. The pulses used are described in detail in Section
3. Electrons were post-accelerated onto the electron
microchannel plate detector and ions were extracted
in the opposite direction into the reflectron time-of-
flight analyser with a high voltage pulse of 290
Vrcm applied 3 ms after the AFG pulse. ZEKE
spectra were recorded with an SRS boxcar integrator
while MATI signals were simultaneously detected by
a digitising oscilloscope. Laser frequencies were cal-
ibrated by comparison to iodine absorption spectra
maximize the experimental resolution. Our strategy
for obtaining high resolution MATI spectra can be
readily extended to allow the synchronous detection
of a high resolution ZEKE photoelectron spectrum.
We demonstrate the technique by presenting a MATI
w
x
dissociation 9,11,12 spectrum of the phenolPAr
cluster, to show that the approach is not only useful
in the high signal regime of molecular spectroscopy,
but can readily be used to obtain spectra of com-
plexes where lower signal intensities are the rule.
w
x
15 , to ensure an accuracy in the laser energy of
2. Experimental
"0.02 cmy1. All ZEKE and MATI spectra pre-
sented in this paper were obtained via the phenolPAr
w
x
The apparatus used has been described in detail in
S100 transition at 36 323 cmy1 16 .
w
x
a recent paper 14 . Briefly, phenolPAr clusters were
produced in a skimmed supersonic expansion by
passing argon gas over a heated phenol sample. The
excitation and ionization region consisted of a mutu-
ally perpendicular arrangement of the jet, two
counter-propagating frequency-doubled dye lasers
3. Results and discussion
3.1. Synchronous high resolution ZEKE and MATI
spectroscopy
Ž
Excimer pumped Radiant Narrowscan for S1 S0
excitation and Lambda Physik FL 3002 for D0 S1
The key to our method of improving MATI spec-
tral resolution lies in utilizing FSSFI which has
.
excitation and m-metal shielded extraction plates.
Ž
.
.
Fig. 1. Schematic of a typical field ionization pulse used in a FSSFI ZEKE experiment. The pulse is applied 3 ms after photoexcitation t0

(
)
C.E.H. Dessent et al.rChemical Physics Letters 315 1999 103–108
105
produced significant improvements in ZEKE spectral
w x
resolution 7 . The general procedure for effecting
Stark state inversion in ZEKE spectroscopy involves
Ž
.
the application of a short ;20 ns electric pulse in
the positive z direction, followed several hundred
nanoseconds later by one in the negative z direction.
Ž
In this Letter, negative and positive z direction are
used to indicate a field accelerating the electrons
.
onto or opposite to the microchannel plate detector .
A schematic of a typical pulse employed in experi-
ments in our laboratory is displayed in Fig. 1. In a
ZEKE experiment, photoelectrons produced by the
initial laser interaction will rapidly separate from the
neutral, long-lived Rydberg states in the ionization
region. This process is repeated upon application of
Ž
.
the first ionizing pulse positive z in the FSSFI
experiment, so that when the negative z pulse is
applied several hundred nanoseconds later only elec-
trons from original ‘blue’ Stark states are detected.
While this procedure readily produced high resolu-
tion ZEKE spectra, it will entirely fail as a scheme
for MATI spectroscopy since in FSSFI, PFI ions
produced by the positive z pulse would begin to
spatially disperse from the remaining Rydberg ex-
cited molecules in one direction only to be ‘re-
focused’ by the negative z pulse.
Fig. 2a displays the tailored pulse we have em-
ployed to combine FSSFI with MATI spectroscopy.
The excitation lasers fire at ts0, followed by the
Fig. 2. Pulse detection schemes and resulting ion time-of-flight
Ž .
spectra involved in
a the new high resolution detection scheme
Ž .
Ž .
application of a q1 Vrcm pulse 200 ns later ta .
and b low resolution MATI spectroscopy.
This pulse can be considered to be equivalent to the
separation pulse employed in a typical MATI experi-
Ž
.
ment Fig. 2b , but will also ionize the more fragile
‘red’ Stark states. The more resilient ‘blue’ states
survive, and the spontaneous ions spatially separate
from the neutral Rydberg states for 3 ms before the
electric field is rapidly inverted to y1 Vrcm for 20
ns at tb. Stark state inversion occurs, and the original
‘blue’ states are now ionized, producing high resolu-
tion ZEKE electrons and MATI ions. The polarity of
the electric field is now reversed again to continue
the spatial separation of spontaneous ions from both
the high resolution MATI ions, and the remaining
Rydberg states which will be ionized to produce low
resolution MATI signal by the substantial pulsed
the pulse sequence is presented with the pulse scheme
in Fig. 2a. This time-of-flight profile can be com-
pared to one obtained using a traditional MATI
separation pulse in Fig. 2b. Although the inverted
Ž .
pulse tb reverses the field experienced by the spon-
taneous ions, and therefore counteracts the spatial
separation achieved during the first 3 ms, the effect
is minimized by the shortness of this pulse. The
pulse displayed in Fig. 2a was tailored to optimize
the separation of the spontaneous, high resolution
and low resolution ion signals in the time-of-flight
spectrum and will be dependent on parameters such
as the experimental configuration, the mass of the
system being studied and the molecular beam veloc-
ity. The greater ion separation achieved by increas-
ing the magnitude of the pulse in the positive z
Ž
.
field 290 Vrcm used to extract ions into the
reflectron time-of-flight analyzer. The time-of-flight
distribution of ions produced at different points in

(
)
106
C.E.H. Dessent et al.rChemical Physics Letters 315 1999 103–108
direction must be offset against the lower ZEKE
resolution obtained for larger ionizing fields, and as
w x
demonstrated in the FSSFI ZEKE experiment 7 , the
ratio of the positivernegative z field must also be
balanced with regard to high resolution signal inten-
sity versus spectral resolution.
The high resolution MATI technique described
above can be straightforwardly extended to obtain
synchronous, high resolution ZEKE spectra in our
apparatus since electrons are extracted into a flight
tube perpendicular to the reflectron time-of-flight
w
x
spectrometer used to detect cations 14 . Examples of
simultaneously recorded ZEKE and MATI spectra of
phenolPAr are presented in Section 3.2.
3.2. MATI dissociation spectroscopy of phenolPArq
The synchronous ZEKE and MATI spectra
recorded using the new detection scheme are pre-
sented in Fig. 3. The spectra consist of two groups of
peaks with the lower energy region corresponding to
Ž
.
the adiabatic ionization energy IE region, while the
second group corresponds to excitation of the 6a
y1
Ž
.
intramolecular mode 518 cm ion internal energy .
Each group is composed of low frequency inter-
molecular vibrations, which have been analyzed as
part of a new investigation of the ZEKE spec-
Ž .
Ž .
b high resolution
Fig. 3. Synchronously recorded
a
ZEKE,
Ž .
MATI, and c low resolution MATI spectra of phenolPAr recorded
via the S100 band origin. The spectra are enlarged over the region
of the 6a intramolecular vibration, with the arrow indicating the
w
x
troscopy of phenolPAr 17 . The spectra are in good
agreement with a previous spectrum which was
w
xq
dissociation energy of phenolPAr . The adiabatic ionization
energy IE is labelled on the figure.
Ž
.
w
x
recorded over the IE region 18 .
The low resolution MATI spectrum presented in
Fig. 3c can be regarded as a rather typical example
and clearly illustrates the poor resolution of the
The tailored pulse scheme for obtaining high reso-
lution MATI spectra described in this work is fully
equivalent to the FSSFI pulse scheme used previ-
ously to obtain fully rotationally resolved ZEKE
spectra of benzene with linewidths of ;0.15 cmy1
Ž
method compared with ZEKE spectroscopy Fig.
y1
.
3a . The doublet feature centred at 29 cm
ion
internal energy merges into a single feature, while
the peak at 37 cmy1 ion internal energy is entirely
obscured compared to the ZEKE spectrum. How-
ever, the resolution of the MATI spectrum produced
w x
7 . This also represents the inherent resolution of the
present experiment since the TOF spectra shown in
Fig. 2 clearly illustrate that the high resolution MATI
signal can be resolved from the low resolution MATI
signal, and the linewidths in the high resolution
ZEKE and MATI spectra are equivalent. The spec-
tral resolution observed in the phenolPAr spectra
presented here is not limited by the inherent
ZEKErMATI resolution but due to the pump laser
selecting a group of rotational levels in the S1 state.
In future work, this high resolution MATI technique
will be used to partially resolve the rotational struc-
Ž
.
by ionization at tb Fig. 3b compares extremely
well with the ZEKE spectrum, a result which is
entirely expected since the two spectra arise from
ionization of the same Rydberg states. The FWHM
of the IE peak in the low resolution MATI spectrum
y1
Ž
.
Fig. 3c is 6 cm
while the FWHM of the corre-
sponding peak in the high resolution MATI spectrum
Fig. 3b is 3 cmy1. This value is indeed equal to the
Ž
.
Ž
FWHM of the IE peak in the ZEKE spectrum Fig.
3a which is also 3 cm
y1
.
.

(
)
C.E.H. Dessent et al.rChemical Physics Letters 315 1999 103–108
107
ture of the molecular cluster cation and to apply a
fitting procedure to determine the rotational con-
experiment illustrates that there is no reason to be-
lieve that the signal will always overlap between the
two channels due to energy transfer from electroni-
cally excited high n-Rydberg electrons to the vibra-
tional degrees of freedom of the ionic core, as sug-
w
x
stants 19 . This requires the determination of the
rotational state population in S1 as a function of
w
x
S1 S0 transition wavenumber 19 .
w
x
Inspection of the expanded spectra over the region
450–600 cmy1 ion internal energy reveals that while
the MATI signal ends abruptly above ;530 cmy1
due to dissociation of the ionic core, the ZEKE
signal continues. The high resolution MATI spec-
gested previously 12 . This observation, in the light
of the present results, can now be understood as an
artifact due to the lower-n Rydberg states employed
in that high field experiment. For those lower-n
Rydberg states the electron still interacts with the ion
core and the field can influence the electron-core
coupling.
w
xq
trum of phenolPAr is reproduced in Fig. 4 along
with the spectrum of its daughter fragment, phenolq.
The insert presents an expanded view of the dissocia-
tion region, illustrating that the last peak in the
parent channel occurs at 532 cmy1, while the first
w
xq
The phenolPAr
dissociation energy obtained
in this work is in line with dissociation energies of
w
x
other aromatic-noble gas complexes 11,12 . The
dissociation energies of the S0, and S1 states of
phenolPAr can be calculated using the cation bind-
ing energy and the red-shifts of the S1 S0 and
D0 S0 transitions of the complex relative to the
peak in the daughter channel occurs at 543 cmy1
.
Signal in the parent channel goes abruptly to zero at
535 cmy1, and since the breakdown of signal in the
parent channel has been shown to represent the
w
x
w
x
field-free dissociation threshold 12 , we are led to
phenol monomer 14 , giving values of 364"13 and
397"13 cmy1, respectively. The binding energies
illustrate that the interaction energy increases signifi-
cantly upon ionization of the phenol moiety due to
the introduction of the charge–dispersion interaction.
w
xq
assign the dissociation energy of phenolPAr as
535"3 cmy1. Note that unlike previous MATI
w
x
dissociation spectra 9,12 , there is no oÕerlap of
signal between the parent and daughter mass chan-
nels, probably due to the small electric field em-
ployed to ionize the Rydberg states in this experi-
Ž
. w
x
ment 1 Vrcm versus ;500 Vrcm 12 . This
4. Concluding remarks
w x
In their 1991 paper 8 , Zhu and Johnson re-
marked that while MATI and ZEKE spectra should
provide the same information, the higher voltages
necessary for MATI spectroscopy meant that elec-
tron detection should be preferred when the highest
optical resolution was required, while MATI would
be the method of choice when mass information was
required. The new detection scheme for obtaining
high resolution MATI spectra presented in this work
demonstrates that MATI spectroscopy need not be
regarded as a compromise between resolution and
mass information. In addition, the new scheme al-
lows high resolution ZEKE and MATI spectra to be
obtained synchronously, providing an experimental
approach for monitoring interactions between ZEKE
electrons and the cationic core.
w
xq
Fig. 4. MATI spectrum of phenolPAr
at mass channels 134
.
q
0
Žw
amu phenolPAr
xq
.
Ž
and 94 amu phenol recorded via the S10
band origin. The inserts show expanded plots of the dissociation
region with the dashed line indicating the position of the field-free
ionic dissociation threshold.

(
)
108
C.E.H. Dessent et al.rChemical Physics Letters 315 1999 103–108
w x
7
H.J. Dietrich, K. Muller-Dethlefs, L.Y. Baranov, Phys. Rev.
¨
Acknowledgements
Ž
.
Lett. 76 1996 3530.
w x
Ž .
8
w x
9
L. Zhu, P. Johnson, J. Chem. Phys. 94 1991 5769.
We thank the EPSRC for financial support of this
S.R. Haines, C.E.H. Dessent, K. Muller-Dethlefs, J. Chem.
¨
Ž .
Phys. 111 1999 1947.
Ž
.
work Grant No. Chemistry GRrL27770 .
w
w
w
x
Ž
.
10 J.D. Pitts, J.L. Knee, J. Chem. Phys. 109 1998 7113.
x
Ž
.
11 G. Lembach, B. Brutschy, J. Phys. Chem. 100 1996 19758.
x
12 Th.L. Grebner, P.v. Unold, H.J. Neusser, J. Phys. Chem. A
Ž
.
101 1997 158.
References
w
w
x
13 H.J. Dietrich, R. Lindner, K. Muller-Dethlefs, J. Chem. Phys.
¨
Ž
.
101 1994 3399.
w x
x
1
K. Muller-Dethlefs, M. Sander, E.W. Schlag, Chem. Phys.
¨
14 S.R. Haines, W.D. Geppert, D.M. Chapman, M. Watkins,
Ž
.
Lett. 112 1984 291.
C.E.H. Dessent, M.C.R. Cockett, K. Muller-Dethlefs, J.
¨
w x
Ž .
Chem. Phys. 109 1998 9244.
2
K. Muller-Dethlefs, E.W. Schlag, Angew. Chem. Int. Ed. 37
¨
Ž
.
w
w
w
x
1998 1346.
15 S. Gerstenkorn, P. Luc, Atlas du Spectre d’Absorption de la
w x
3
M.C.R. Cockett, K. Muller-Dethlefs, T.G. Wright, Annu.
¨
Molecule d’Iode, CNRS, Orsay, France, 1978.
Ž
.
x
Rep. R. Soc. Chem. Section C 94 1998 327.
16 E.J. Bieske, M.W. Rainbird, I.M. Atkinson, A.E.W. Knight,
w x
Ž
.
4
K. Muller-Dethlefs, M.C.R. Cockett, Nonlinear Spectroscopy
¨
for Molecular Structure Determination, Chap. 7. Blackwell
J. Chem. Phys. 91 1989 752.
x
17 S.R. Haines, C.E.H. Dessent, K. Muller-Dethlefs, manuscript
¨
Science, Oxford, England, 1998.
in preparation.
w x
Ž .
F. Merkt, T.P. Softley, Int. Rev. Phys. Chem. 12 1993 205.
w
w
x
Ž
.
5
w x
6
18 X. Zhang, J.L. Knee, Faraday Discuss. 97 1994 299.
Ž .
x
A. Held, E.W. Schlag, Acc. Chem. Res. 31 1998 467.
19 M.S. Ford, K. Muller-Dethlefs, manuscript in preparation.
¨