Laser induced processes in molecular ions


Material Information

Laser induced processes in molecular ions
Physical Description:
xi, 193 leaves : ill. ; 28 cm.
Morgenthaler, Lee Nickey, 1952-
Publication Date:


Subjects / Keywords:
Lasers in chemistry   ( lcsh )
Laser spectroscopy   ( lcsh )
bibliography   ( marcgt )
theses   ( marcgt )
non-fiction   ( marcgt )


Thesis--University of Florida.
Includes bibliographical references (leaves 188-192).
Statement of Responsibility:
by Lee Nickey Morgenthaler.
General Note:
General Note:

Record Information

Source Institution:
University of Florida
Rights Management:
All applicable rights reserved by the source institution and holding location.
Resource Identifier:
aleph - 000088999
notis - AAK4379
oclc - 05691720
System ID:

This item is only available as the following downloads:

Full Text



Lee Nickey Morgenthaler


University of Florida



This research is the fruit of the effort of many

people. The instrument used was designed, fabricated, and

assembled by the workers in the electronics and machine

shops of the Chemistry Department of the University of

Florida. Without their efforts this project would not have

been possible. I received financial support from the

Chemistry Department, through the ever-smiling Dr. Gerhard

Schmit. I would also like to thank Dean Harry H. Sisler for

the financial support he provided in 1977-1978. I thank

the Tennesee Eastman Company for the financial support

provided during my final year of graduate school.

Many of my colleagues at the University of Florida made

a significant contribution to this work. Among them are

Dr. Dan Reo of ICAAS who assisted me in the use of the

curve fitting program, Dr. Milap Mathur of the Chemistry

Department who assisted in the "chemical" aspects of this

work and Dr. Ed Voigtman of the Fourth Floor who assisted

in the design of the computer system. Many thanks also to

the other members of Dr. Eyler's research group--especially

Ron Daubach, "Bad"Bob Doyle, and Tom Buckley--for many

helpful discussions.

Throughout this work I was sustained, both spiritually

and materially, by many good friends. I would especially

like to thank Aldo Atherton and Kelley and Janet St. Charles

for their assistance in times of need. My best friend

throughout this project was my wife Andi. She provided me

with an unfailing supply of love and very vocal encourage-

ment to complete the work.

Finally I would like to thank Dr. John Eyler for his

inspiration, assistance, encouragement, and friendship.


Chapter Page



ABSTRACT. . . ix


2 THEORY. . . 15

Introduction . 15

The Motion of Ion in the
Trapped Ion Cell . 17

The Pulsed Marginal
Oscillator Detector. ... 21

Derivation of the Expression for
the Photodissociation Cross Section. 26


Introduction . ... 37

The Vacumm System . .. 37

The Optical System . .. 40

The Digital Electronics. .. 46

Interfacing the Pulse icr With
a KIM-1 Microprocessor ... 51

Evaluation of the Performance of
the Computer Data System .. 65


Introduction . ... 72




Experimental .

Results .

Discussion .





Results. .

Discussion .





Results. .

Discussion .









. 152















. 127


. 137






: : : : : :

Results. .


Chapter Page

7 Discussion . 181
Conclusions . .. ... 186




E(x) -- Total energy at wavelength X in a laser pulse,

generally recorded as mV of joulemeter output, this

quantity times 2.2 is the energy in mJ.

E(x) -- E( ) in mV times x in nm, this quantity is propor-

tional to the photon flux in photons cm- s- nm-1 as

derived on p. 26.

f -- The fraction of the ion population actually in the

laser beam.

icr -- Ion cyclotron resonance mass spectrometry.

I.P. -- Ionization potential.

kL -- The rate constant for orbiting collisions intro-

duced on p. 81.

K'a -- The relative cross section for a photoinduced


m.o. -- Marginal oscillator.

N -- Mean collisional number (see p. 88).

N(A) -- The number of photons of wavelength A in a laser


P(A) -- Probability that an ion will undergo a photoinduced

process when irradiated at wavelength.

p(X) -- Photon flux at wavelength.

R -- The fraction of the ion population that disappears

in photolysis.

a(x) -- The absolute cross section for a photoinduced
process in cm

Z -- Collision frequency in s-1
Z -- Collision frequency in s


Abstract of Dissertation Presented to the
Graduate Council of the University of Florida
in Partial Fulfillment of the Requirements
for the Degree of Doctor of Philosophy



Lee Nickey Morgenthaler

June 1979

Chairman: John R. Eyler
Major Department: Chemistry

The pulsed icr pulsed laser technique is a means of

probing a number of physical processes in excited molecular

ions in the gas phase. A detailed spectrum, related to the

absorption spectrum, of the ion can be obtained at a time

when it has undergone anywhere from zero to fifty collisions

with the background gas. The spectra of collisionally

unstable ions may be studied before they decompose or

react and collision induced changes in apparently stable

ions can be followed spectroscopically.

Chloroethane molecular ion is an example of a col-

lisionally unstable species that has been investigated

successfully by this technique. The photodissociation

spectrum indicates that the first excited electronic state

of this ion is weakly bound or flat and that the second

excited state is purely repulsive with respect to scission

of the carbon-chlorine bond. In addition, there is a

strongly energy-dependent competition between two reaction

channels. This effect is not predicted by quasi equili-

brium theory, which indicates that the dynamics of the

decay processes are dictated by the potential surfaces of

the excited electronic states of the ion.

Several of the experiments presented here are attempts

to probe the dynamics of rearrangement processes that had

been discovered by the continuous icr technique. In the

first case, benzyl chloride, the results presented here

are not the same as those obtained in the continuous icr

experiment. A kinetic scheme that reconciles the two re-

sults is suggested. Based on the results presented here

one cannot be sure that rearrangement has occurred at all.

In the second example, cis trans isomerization in

the 1,3,5 hexatriene cation, emission spectroscopy indi-

cated that the structure of the ions formed from the two

isomers of the parent compound were different on the time
scale of 10- s. However, continuous icr photodissociation

spectroscopy indicated that the two structures were iden-

tical on the time scale of several seconds. The results

presented here indicate that the ions of the two isomers

are the same within 10 ms of ionization. No information

about the kinetics of the rearrangement process can be

obtained in this case.

The collisional relaxation of vibrationally excited

1,3,5 hexatriene molecular ion was also studied. The rate

of quenching by the parent molecule was observed to depend

on the ionization energy. Ions formed at the higher ioni-

zation energy were quenched on every nonreactive collision

with the parent molecule. This result is consistent with

other studies of similar processes in molecular ions. Ions

formed at the lower ionization energy were quenched less

efficiently by a factor of four--an unexpected result.

Quenching by argon atoms was also investigated and found

to be extremely rapid--about two orders of magnitude faster

than the corresponding process in neutrals.


The existence of charged particles arising from ordi-

nary chemical compounds was proposed by Arrhenius in his

doctoral dissertation, published in 1884. By 1900 con-

ductivity measurements on electrical discharges and flames

had established the presence of similar charged particles

in the gas phase. The study of gaseous molecular ions

began with the development of the mass spectrograph by

J.J. Thomson in 1907. Within a few years several investi-

gators reported evidence of chemical reactions between

gas-phase ions and the molecular samples from which they

were formed. A few studies of these reactions appeared in

the next forty years, but it was not until 1950 that the

study of ion-molecule reactions began in earnest.

In more recent times (beginning around 1960) the

investigation of ion-photon interactions was begun. This

field of study evolved from the study of ion-molecule

interactions both in terms of the motivation behind it and

in terms of the instrumentation used. Such studies have

yielded information about the energetic of electron

detachment from negative ions and about the dynamics of

photodissociation of both positive and negative ions.


In addition, the technique of photodissociation spectro-

scopy has shown promise of providing a sensitive probe of

the structure of gas-phase ions. The studies reported in

this dissertation yield information about the dissociation

of several molecular ions. Both the dynamics of the dissoci-

ation process itself and the structure of gas-phase ions

are probed. In addition, this work is one of the first

attempts to study such processes in a fully pulsed mode of

operation. The research presented here examines some of

the capabilities of the pulsed technique.

A number of experimental techniques have been de-

veloped for the study of ion-molecule reactions and ionic

decomposition. All of them were based upon mass spectrom-

eters that were modified to study the specific process of

interest. These modifications have included variable

pressure sources (1 ), photoionization sources (2 ),

various coincidence techniques (3 ), and various schemes

for kinetic energy analysis of fragments (4 ). A further

development was the use of two mass spectrometers in

series, so-called tandem mass spectrometers, to study the

ionic products of the reactions of a specific ion with a

neutral gas (5 ). Another type of mass spectrometer, the

ion cyclotron resonance (icr) mass spectrometer, also

proved to be very useful in these studies (6,7). The long

residence times and the low sample pressures that can be

achieved in these instruments make them ideally suited to

the investigation of gas-phase ion-molecule reactions.

Studies of ion-molecule reactions have had three

important impacts on chemistry as a whole. Studies of the

unimolecular decomposition of ions have played a major

role in the development of the theory of unimolecular

decomposition ( 8 ). This in turn has been an important

input to the transition state theory of chemical reactions

since the decay of the intermediate complex is treated as

a unimolecular decomposition.

Gas-phase ion-molecule reactions are often similar

to reactions hypothesized in liquid-phase chemistry. Many

reactions in solution are thought to proceed via ionic

intermediates which have very brief lifetimes and whose

existence in solution can only be inferred. Elaborate

theories relating the structure of these ionic inter-

mediates to their stability and chemistry have been

developed and inflicted upon generation after generation

of unsuspecting organic chemistry students. The study of

gas-phase ion-molecule reactions provides a means of

examining directly the chemistry of these legendary species

in the absence of "solvent" effects. It has provided a

test for many theories of reaction mechanisms (9 ).

Although it is a powerful analytical technique,

electron impact mass spectrometry has the serious drawback

that it only measures the mass-to-charge ratio of the ions

arising from a given sample. Many different compounds,

particularly structural isomers of complex molecules, yield

indistinguishable electron impact mass spectra. The third

major impact of studies of ion-molecule interactions is

a powerful technique for overcoming this difficulty:

chemical ionization mass spectrometry (10). In this

approach the sample is ionized in an ion-molecule reaction.

It may be hoped that a further elucidation of the proper-

ties of gas-phase ion-molecule interactions will lead to

new developments in this field.

Beyond the confines of chemistry itself, there are

many fields where the study of ion-molecule reactions may

lead to important new insights. Among them are aeronomy,

combustion technology, and radiation biology. Of course

the analytical technique of chemical ionization mass spec-

trometry is of importance in many more fields of investi-

gation than can be enumerated here.

A second class of gas-phase ion-particle interactions

that has aroused considerable interest may be called ion-

photon interactions. Principal among these are photodetach-

ment of negative ions and photodissociation of either

positive or negative ions. Such processes are of basic

physical and chemical interest. Studies of photodetachment,

for example, are thought to yield extremely accurate values

of the electron affinity of many molecular species (11).

The study of the photodissociation of positive ions has

yielded information about the location, stability, and

dynamics of the decomposition of excited states of molec-

ular ions. In addition, since a photon must be absorbed

for photodissociation to occur, the study of ionic


photodissociation provides a means of examining the

spectroscopy of gas-phase ions. Such information is very

difficult to obtain by spectroscopic techniques normally

employed in the study of neutral species.

The development of gas-phase ion spectroscopy has been

greatly accelerated by the advent of lasers. In general

ion densities are low in the gas phase and consequently

very intense light sources are required. Prior to the

introduction of laser light sources the spectral resolution

of studies of ion-photon interactions was limited by this

need for a high photon flux. Lasers with their high

intensity and spectral resolution are particularly well

suited to the study of gas-phase ion spectroscopy.

Instrumentation used to study ion-photon interactions

has been developed from that used to study ion-molecule

reactions. The simplest modification consisted of opening

a window into the ion source of a mass spectrometer and

irradiating the sample there (12,13). Such techniques may

be classified as "high pressure swarm" techniques using a

classification scheme introduced by Henchman (14). A

number of such studies have been described. They are,

with few exceptions, of principally historical interest.

For one thing, these techniques were not specific as to

which ion was in fact being dissociated or detached. Under

normal source conditions the daughter particles from the

photoinduced process might undergo as many as 100 collisions

prior to mass analysis. Since these species are almost by

definition high energy (in the chemical sense) species,

there is the possibility that these techniques may sample

secondary or higher-order products of reactions between

the daughter and neutral species. In addition, the spatial

distribution of the ion population in these instruments is

not known, making quantitative measurements of absolute

cross sections impossible.

Techniques based on the tandem mass spectrometer,

"beam techniques" in Henchman's scheme, have provided a

far more detailed picture of ion photodissociation and

photodetachment. Both crossed laser-ion beam (15) and coaxial

laser-ion beam (16) instruments have been constructed. In

both systems particle densities and spatial distributions

are reasonably well defined and absolute cross sections can

be measured with some confidence. Coaxial beams can be

Doppler-tuned to provide extremely high spectral resolution.

In crossed beam instruments the measurement of the angular

distribution and kinetic energy of photofragment ions and

photodetached electrons can be measured. Both of these

techniques provide extremely detailed data on the dynamics

of the photoinduced process. To date both have been

applied principally to simple ions.

Ion cyclotron resonance mass spectrometry can be

classified as a "low pressure swarm" technique. Such

instruments have been modified for studies of the photo-

detachment and photodissociation of gas-phase molecular ions.

Because the icr technique allows for irradiation and mass

analysis in the same chamber of the instrument and because

operating pressures are often so low that an ion undergoes

only a few collisions per second, some of the difficulties

encountered in the high pressure swarm techniques are mini-

mized in continuous icr mass spectrometry. However, this

is not always the case and must be considered on a case-by-

case basis. Unfortunately the spatial distribution of ions

in the icr cell is difficult to analyze (17) making the

determination of absolute cross sections nearly impossible.

In general, photodissociation and photodetachment studies

in icr instruments have involved a wider variety of ions

than those using other techniques. A number of studies of

large and complex ions have appeared (18,19).

The wavelength dependence of the cross section for

photodissociation or photodetachment, the photodissociation

or photodetachment spectrum, yields truly spectroscopic

information. The threshold for photodetachment can be

related to the electron affinity of the neutral species

(11). The onsets for detachment to excited states of the

neutral species can be identified by abrupt changes in the

photodetachment cross section with wavelength. The photo-

dissociation of gas-phase ions with visible light proceeds

via excited electronic states of the ion. This can be

confirmed by comparison of the photodissociation spectrum

to the photoelectron spectrum (20) of the parent neutral.

In the photoelectron spectrum the lowest ionization po-

tential corresponds to ionization to the ground state of

the ion. The higher ionization potentials correspond to

formation of excited states of the ion. The difference

between a given ionization potential and the lowest yields

the energy of the given excited electronic state of the

ion. This analysis is confirmed by the fact that almost

all photodissociation spectra that have appeared so far

correspond closely to the photoelectron spectra of the

parent molecules.

The photodissociation spectrum is related to the

absorption spectrum of the ion since photodissociation

involves the absorption of a photon. However, photodissoci-

ation also involves dissociation. Every photon that is

absorbed may or may not cause dissociation. As a result

the photodissociation cross section is only an lower limit

to the absorption cross section.

The most elegant and complete study of photodissoci-

ation of a molecular ion was, ironically perhaps, the first

to appear using the crossed ion-laser beam technique (21).

A fixed frequency laser was used to dissociate H2 and D 2

The fragment ion kinetic energy was obtained along with

the angular distribution of ionic photofragments. Such

results yield detailed information about the orientation

of the electronic transition moment and the lifetime of

the excited state. For H2 and D these results could

be compared to theoretical predictions (excellent agreement

was observed).

Moseley and his co-workers (16) have employed Doppler-

tuning of their coxial ion-laser beams to examine the

rovibronic structure of a long-lived (4nu) state of 02

This study highlights the extremely detailed information

that can be obtained in favorable cases by photodissoci-

ation spectroscopy.

The most important results obtained to date by icr

photodissociation spectroscopy have been related to the

identification of the structure of molecular ions in the

gas phase. Numerous ions of the same mass but arising from

different neutral precursors have been investigated.

According to Lineberger (11) these studies have provided

the most convincing evidence as yet available that the

structure of a molecular ion in the gas phase is related to

the structure of the parent molecule. This result is

important, for if it were not so, the comparison of gas-

phase ion chemistry to solution chemistry would have little

validity. A variation on this theme is provided by a

recent study of reactively produced chloropropene (C3H5CI )

molecular ions (22). Different isomers of C3H5CI identi-

fied by their photodissociation spectra, were shown to

arise from several ion-molecule reactions. Prior to this

study the actual structure of the products of an ion-

molecule reaction could only be inferred. Such investi-

gations promise to yield new and more detailed

information about the mechanisms of gas-phase ion-molecule


A second interesting area of investigation, currently

being pursued by steady state icr techniques, is multi-

photon dissociation of molecular ions (23). Because of the

low operating pressures employed in the continuous icr,

radiative decay (24) is often the only mode of relaxation

for excited ions. If radiative decay is slow, many photons

may be absorbed by an ion even at relatively low light

intensities. In addition to providing information about

the excitation and de-excitation of gas-phase ions, such

studies may lead to new insights into the phenomenon of

sequential photon absorption.

Pulsed icr mass spectrometry (25) was developed

somewhat more recently than steady state icr mass spectrom-

etry. Few studies of photodetachment or photodissociation

have been carried out with this technique (26,27).

Furthermore, none of these studies has attempted to exploit

the time resolution of the pulsed icr. Although several

investigators have used a technique that they call "time

resolved photodissociation spectroscopy" (28,29) they have

not obtained a time resolved photodissociation spectrum.

Rather they have measured the rate of disappearance of an

ion when irradiated with a continuous light source. The

term "exhaustive photodissociation" used in one report (29)

about this technique more accurately reflects the truth of

the matter.

In a sense the experiments reported below are

"technique" oriented rather than "problem" oriented. Each

is an attempt to use the time resolution of the pulsed icr

technique to obtain information that cannot be obtained in

the continuous icr. Each represents a different facet of

the instrument's capabilities. The first study (Chapter 4)

is essentially a calibration experiment. The photodetach-

ment of OH- has been well studied (30,31) and no new infor-

mation is reported here. The cross section for a photoin-

duced process is shown to vary slowly with ion residence

time between the first few and at least the fortieth col-

lision. Some effect of magnetic field strength and laser

alignment are observed and procedures are developed to

correct for them.

The study of chloroethane molecular ion presented in

Chapter 5 uses the time resolution of the pulsed icr to

investigate an unstable ion whose abundance is very low in

a steady state icr instrument (32). The photodissociation

spectrum yields interesting information about the excited

state potential surfaces of the ion. In addition, data on

the competition between dissociation channels were obtained.

A similar study of the butene molecular cation has been

reported by Riggin (33). His results are explained success-

fully by the quasi equilibrium theory (8 ) of unimolecular

decompositions. In contrast, the results obtained in the

study of chloroethane molecular ion reported here bear no

resemblance to the predictions of quasi equilibrium theory.

A study using the exhaustive photodissociation tech-

nique (29) suggested that the population of the parent ion

(m/e = 126) of benzyl chloride consisted of two molecular

structures one of which dissociated at a wavelength of

600 nm (30 percent of the population) and one of which did

not. The results presented below suggest that this analysis

is simplistic. The maximum percent dissociation observed

is at least 65 percent. Moreover, the apparent cross

section is unchanged for at least 2.5 s corresponding to

roughly 100 collisions. No evidence of rearrangement is

observed. A kinetic scheme consistent with all of these

observations (both pulsed and steady state studies) is


Another well studied molecular cation is that of 1,3,5,

hexatriene. Evidence obtained by emission spectroscopy (34)

indicated that there were two distinct isomers of the molec-

ular ion in the gas phase. One arose from the cis isomer

of the molecule and one from the trans. A study of ions

formed from both molecules carried out in a continuous icr

(35) indicated that no matter which neutral precursor was

used the structure of the molecular ion was the same. The

results presented below also suggest that the parent ion is

identical no matter which isomer of the parent molecule is

ionized. This indicates that the ion rearranges rapidly

after ionization.

A hot band at 646 nm was reported in both the emission

spectrum and the steady state photodissociation spectrum of

1,3,5 hexatriene molecular ion. This band was also ob-

served in the photodissociation spectrum reported in

Chapter 7. The decrease in the intensity of this band

relative to a transition from the ground vibrational state

of the ion with time was obtained. From this the rates of

quenching of the vibrationally excited ions in collisions

with the parent molecule and with argon atoms were obtained.

Although several measurements of the rate of comparable

processes have appeared based on indirect evidence, this is

the first direct measurement of the rate of V-T transfer in

a molecular ion. The results, in general agreement with

those obtained by indirect means (23,36,37), show collisional

quenching of vibrational excitation in molecular ions to be

very rapid compared to similar processes in neutral


In addition to the inherent value of each of these

results as a measurement of a basic physical process, these

studies outline the capabilities and limitations of the

pulsed icr instrument. The pulsed icr is not particularly

well suited to the study of the kinetics of unimolecular

decomposition. Many unimolecular reactions occur much

faster than the minimum time resolution (about 1 ms) of

the instrument. However, the pulsed icr is more useful

than the continuous icr for the study of the competition

between reaction channels because product ions can be

detected within a few ms of irradiation (before they

undergo any collisions). Thus, only the products of the

unimolecular decomposition of the excited ion are sampled.

Because of the narrow bandpass of the laser output, a

precisely known amount of excitation energy is applied in

this experiment. Subtle differences in the probability

of a reaction with excitation energy can thus be probed.

The ability to obtain a complete photodissociation

spectrum at a specific time after the formation of an ion

makes the study of the rearrangement of gas-phase ions

possible. Such processes have been shown to occur using

steady state icr photodissociation spectroscopy (29).

However, such studies yield little information about the

dynamics of the rearrangment process. Pulsed icr techniques

complement the steady state technique and offer considerably more

flexibility in the study of rearrangement processes.

The real forte of the pulsed icr is the study of col-

lisional processes because the time resolution of the

instrument is comparable to the collision lifetime of an

ion at normal operating pressures. The study of V-T energy

transfer in 1,3,5 hexatriene reported in Chapter 7 below

is one example of this capability. It may be hoped that

the study of ion-photon interactions will prove to be as

interesting as that of ion-molecule reactions. Studies

of ion-photon interactions in the gas phase should answer

some questions raised by the study of ion-molecule

reactions. Moreover, the study of ion-photon interactions

promises to yield considerable information about physical

and chemical processes in excited molecular species.



The instrument used in these experiments was a pulsed

ion cyclotron resonance (icr) mass spectrometer (25)

employing a trapped ion cell (38) and coupled to a flash-

lamp pumped dye laser in an intracavity configuration (39).

The specific mechanical, optical, and electronic layout of

the instrument is described in Chapter 3. In this chapter

the motion of ions in the trapped ion cell is considered;

the principle of mass analysis using a pulsed marginal

oscillator detector (40) is described. In the final

section the relationship between the quantities measured

experimentally and the physical observable of interest,

the relative cross section for a photoinduced process,

is derived (41).

Figure 2.1 is a schematic of a trapped ion cell. The

axes shown are referred to throughout the text and the

origin is labelled. ICR mass spectrometry is based on the

principle that a charged particle, in this case an ion,

in crossed magnetic and electric fields executes a circular

LlJ 0

motion perpendicular to the magnetic field the frequency of

which, wf, is given by:

wf = qB/m 2.1

where q is the charge of the ion, m is its mass, and B is

the magnetic field strength. If an oscillating electric

field of the same frequency as the ion motion is applied

perpendicular to the magnetic field,the ions absorb energy

from it and are accelerated to larger orbits. The power

loss from the oscillating circuit is proportional to the

number of ions of a given mass-to-charge ratio in the cell.

In the absence of the oscillating electric field the impo-

sition of the appropriate potentials on the walls of the

cell forms an efficient ion trap (38). The usefulness of

such a system for the study of ion-molecule reactions has

been reviewed recently (6,7).

The Motion of Ions
in the Trapped Ion Cell

The equations of motion for an ion in a trapped ion

icr cell have been obtained (17). The force, r, acting on

an ion with initial velocity v in crossed magnetic and

electric fields (B and t, respectively) is given by the

Lorentz equation:

F = q(t + v x 2). 2.2

Here, q is the charge on the ion. In the trapped ion cell

the potentials on the upper, lower, and end plates of the

cell (labelled as U, L, and E, respectively, in Figure 2.1)

are typically -1.0 V and those on the side plates (T) are

+1.5 V for positive ions. The polarity of these potentials

is reversed for negative ions. The magnetic field is

applied parallel to the z axis and the oscillating electric

field is applied along the x axis. The laser beam is

parallel to the y axis in the photochemistry experiments.

The potential on the side plates repels ions as they oscil-

late parallel to the magnetic field as described more

fully below.

The electric field can be approximated by a two-

dimensional quadrupole field for ions close to the origin

of the cell (17) and the magnetic field is constant over

the ion trap region. The motion of an ion trapped in such

a potential has three characteristic components. The most

important of these is the cyclotron motion in the x,y

plane. It has the characteristic frequency:

2 = (qB/m)2 4qVT/md2 2.3

where m is the observed cyclotron frequency of the ion

motion. The first term on the right hand side is the

ideal cyclotron frequency and the second term is a cor-

rection for the effect of the trapping potential, VT. The

separation between the upper and lower cell plates is d.

The second aspect of the ion motion is a rapid simple

harmonic oscillation parallel to the z axis about the

origin of formation of the ion. The frequency of this

motion is a function of the trapping potential.

The third component is a slow drift of the generating

center for the two oscillations described above. In theab-

sence of collisions and space charges, this drift causes the cross

section of the ion "cloud" to oscillate between being a

circle in the x,y plane and an ellipse in the x,y plane

with the long axis parallel to the y axis. The greater

the kinetic energy of an ion the larger the amplitude of

the drift motion.

In the experiment ions are formed initially in a band

along the z axis of the cell by the electron beam emitted

by a hot filament. The initial acceleration of the elec-

trons is due to the negative potential on the filament and

this is given as the nominal electron energy. However,

the trapping potentials affect the electron's motion so the

true energy of the ionizing electron is a function of its

position in the cell.

At the time of formation the generating centers for

the ion motion form a circular cylinder across the cell

parallel to the z axis. The frequency of the drift motion

is given by:

ad = n/2n 2.4

S = 2(VT V )( B) /2/a2B. 2.5

The constants a and B are related to the geometry of the

cell and a is the length of the cell parallel to the x axis.

The trapping potential is again given as VT and V is the

potential applied to the upper and lower plates of the cell.

Values of a and 6 have been reported for a cell similar to

the one used here (17). Substituting the appropriate

values of VT and Vo from our experiments yields a frequency

for the drift motion of approximately 190 Hz at 8000 G.

The frequency of the change in shape of the ion "cloud"

is twice the frequency of the motion. Thus, the time asso-

ciated with the change in shape is 2.7 ms at 8000 G and

less at lower fields. In all the experiments described

here ions are formed for at least 5 ms. Consequently, the

effects of the oscillation of the shape of the ion "cloud"

are cancelled out because ions are formed for at least two

periods of the oscillation. The resultant drift path is

best described as a quasi-ellipse in the x,y plane with

the long axis parallel to the y axis. Both of the fast

oscillations--the cyclotron motion in the x,y plane and

the simple harmonic parallel to the z axis--originate from

centers travelling this complicated drift path.

The Pulsed Marginal Oscillator Detector

In the pulsed marginal oscillator (m.o.) (40) detection

system the mass detection of ions is achieved by applying

a gated burst of potential to the upper plate of the cell.

This potential has the effect of changing the form of the

electric field, E, during the time the burst is applied

from the quadrupole potential of the static trap to:

t(t) = E1(sinwit) 2.6

The frequency of the applied potential is wi, El is the

amplitude, t is the time, and 1 is the unit vector in the

x direction. Solving the Lorentz equation with this new

potential yields a function for the ion velocity that is

time dependent. The kinetic energy, T, obtained from this

velocity, v, is:

T = 1/2 mv2 = [q2E /2m(wi- )2] sin2(1/2)( i C )t. 2.7
11 c

For the singularity, wi = wc, the straightforward appli-

cation of l'Hopital's rule yields:

T = q2E2t2/8m. 2.8

The power absorbed by a single ion is given by:

A(t) = dT/dt

= [q 2E2/4m(wi-c)] sin(( i-wc)t

A(t) = q E1t/4m,

if, (i = Wc.

In practical pulsed operation the detect pulse is usually

shorter than the collision lifetime of the ion so the

effect of collisions on the power absorption can be

neglected. The average power absorption per ion is given


= t/A(t)dt
o 2E/8m,
= q2E 2r/8m,



if, Wi = C .

Here, T is the ion's lifetime in the oscillating field.

For a sample of n ions:

= nq2
= nq E1r/8m.


Measurement of resonant power absorption is obtainedwith

a GCL circuit like the one illustrated in Figure 2.2 (40).

This circuit has a resonant frequency:











ro c


0 -






S = (LC)-1/2 2.14

where L is the inductance, C the capacitance, and the con-

ductance, G, represents losses due to nonideality of the

circuit. A constant amplitude radio frequency voltage is

provided to the resonant circuit through the feedback

resistor Rf. For pulsed operation this voltage is switched

on and off by the gated amplifier. At the resonant fre-

quency the admittance of the resonant circuit is:

V1 = V0/(l + GR ). 2.15

In this equation V0 is the rms voltage from the limiter and

V1 is the rms voltage across the resonant circuit. The

absorption by resonant ions in the cell represents a change

in the conductance of the circuit, AG. The corresponding

change in voltage, AV, is just:

AV = VO{I/(l + GRf) 1/[1 + (G + AG)Rf]} 2.16

= VIAG[Rf/(1 + (G + AG)Rf]. 2.17

If AV is small the incremental power absorption, A, due to

AG is given by:

A = V2AG. 2.18

If AG is small and Expression 2.18 above is substituted

into Expression 2.17 then:

AV = (A/V1)(Rf/1 + GRf). 2.19

If the expression for power absorption obtained above

(Equation 2.10) is substituted into Equation 2.19 the

instantaneous voltage change due to a single resonant ion

is obtained as:

AV = (q2E2t/4mV )(Rf/l + GRf) 2.20

The average value of the voltage change induced by an

ensemble of n ions is then:

= (nq2E2/8mV )(Rf/l + GRf). 2.21

In practical operation the output of the m.o. is integrated

by a switched integrator circuit as described in Chapter 3.

The final measured signal is thus and V1, Rf, and G are

constants of the analysing circuit. It can be seen that

the magnitude of an ion signal is directly proportional to

the number of ions of a particular mass-to-charge ratio in

the cell. The magnitude is also shown to be inversely

proportional to the ion's mass. This correction must be

applied routinely in pulsed icr work (25).

Derivation of the Expression
for the Photodissociation Cross Section

From this experiment one obtains a signal proportional

to the number of ions of a given mass-to-charge ratio in

the icr cell when no laser irradiation is applied, a signal

proportional to the number of such ions with the laser

flashing, and the total energy in each flash of the laser.

In this section the relationship between these measured

quantities and the quantity of interest, the relative

photodissociation cross section, is derived and discussed.

The probability that an ion will photodissociate at

wavelength x, P(x) is given by:

P(A) = 1 exp[-kt/p(A) o(X) dA] 2.22

where k is an overlap factor between the laser and the

ions, t is the laser pulse duration, p(A) is the flux of

photons of wavelength A and o(A) is the absolute cross-

section for photodissociation at A. The integration is

carried out over the bandwidth of the laser output.

If E(x) equals the energy in a laser pulse at wave-

length x then N(x), the number of photons of wavelength A

in the pulse, is:

N(A) = E(A) A/hc




p(x) = N(x)/at 2.24

where a is the cross-sectional area of the laser beam. The

number of ions, A, which dissociate from a sample B at wave-

length x is:

A = fBP(x) 2.25

where f is that fraction of the ion population actually

irradiated by the laser. The experimental fractional

dissociation, R, is then:

A/B = fP(A) = R. 2.26

Substituting 2.22 into 2.25 yields:

A = fB{1 exp[-kt/p(x)o(x)dx]}. 2.27

Since the bandwidth of the laser output at any grating

setting is small, ca. 0.3-0.5 nm, it can be assumed that

p(A) and o(x) are constant over the bandwidth of the laser

output. It may also be assumed that the bandwidth is

constant at all grating settings. Thus,

/p(A)c (A)dA = p(A)o(A) m

(m = constant) 2.28

and with these assumptions 2.27 becomes:

A = fB{1 exp[-kt p(x)o(A)m]}. 2.29

Substituting 2.23 and 2.24 into 2.29 yields:

A = fB{l exp[-(km/hca)E(x) Ao(A)]}. 2.30

Taking logarithms and rearranging yields:

-1n(l-A/fB) = (km/hca)E(x)A a(A). 2.31

Setting km/hca = K', substituting 2.26 into 2.31 and

solving yields:

K'o(x) = [In(f/f-R)]/XE(A). 2.32

The term k in Expression 2.22 above is loosely de-

scribed as an "overlap factor." As discussed in Chapter 1,

the laser ion interaction region in this experimental

apparatus is poorly defined. In our particular apparatus

this problem is compounded by the fact that routine mainte-

nance of the mass spectrometer necessitates frequent

realignment of the laser. The ion packet orbiting in the

trapped ion cell is probably considerably smaller than the

laser beam diameter, so k should be related to the fraction

of the laser beam that irradiates the ions. Since the ions

drift along the long axis of the cell k is also related to

thedegree of overlap between the laser beam and the ion

"cloud." As such it can be expected to change somewhat with

laser alignment. Qualitatively k can be thought ofas a

measure of the relative size of the laser-ion interaction


In the discussion above f is described as that

fraction of the ion population actually irradiated by the

laser; i.e., if 95 percent of the ion population is irradi-

ated, f = 0.95. However, a second factor may lead to an f

value less than unity. If some fraction of the ion popu-

lation is not dissociable then f will appear to be less

than 1.0. Figures 2.3 and 2.4 illustrate the behavior of

Equation 2.32 with changes in K' (2.3) and f (2.4).

An important pair of assumptions is introduced in

Equations 2.24 and 2.25. The quantity p(X) is set equal

to the total number of photons in the laser pulse divided

by the cross-sectional area of the beam and the duration

of the pulse. As such it is an average flux. Equating

this with the local flux through the sample is correct only

if the cross-sectional area of the beam is constant at all

E(x). If the area is a function of the geometry of the

lamp and optics, this will be true. If not, a serious

distortion of the intensity dependence may result. This

effect is illustrated in Figure 2.5. Here Curve A






LU d Lo LO Cn

w 0 0

. 0 0 0 b





I I i i I i
C nI-























S- C
cu 4




C) -0





10 20 30 40 50 6C
joulemeter reading / mV

Figure 2.5.

Illustration of Effects of Variation in Laser

Beam Profile on Measured R value (see text for



of 0.5-

represents the result if a is constant as assumed. In

Figure B, a is varied with E(A) according to the relation-


a = a0 + rE(A)

and a0 is set equal to 0.5; r is set equal to 0.01. The

physical interpretation of this is that the beam has a

diameter of 0.5 if the laser lases at all and that it

increases with increasing E(X). Thequantity R is calculated

from Equation 2.29, with K'o assumed to be 1 x 10-4, and

p(A) set equal to E(x)/a. The quantity R is plotted versus E(x).

observed. Figure C results if a is proportional to E(A).

Curve C is highly unrealistic. However, Curve B is barely

distinguishable in form from Curve A, but the quanti-

tative difference in apparent K'o is marked.

This model is chosen arbitrarily to demonstrate that

serious systematic errors might go undetected in these

experiments and to illustrate the difficulty of obtaining

reliable absolute cross sections in the icr apparatus. To

do so one would need to know precisely the size and spatial

distribution of the ion population in the icr cell at all

times. Further, one would need to know the size and

intensity profile of the laser beam for each shot. Given

the fluctuating nature of even the grossest experimental

variables there seems to be little hope of refining these


experiments sufficiently to yield accurate absolute cross

sections. One can only hope that systematic errors, if

important at all, are consistent enough from experiment to

experiment to yield reliable relative results.



The pulsed icr used in these studies was built largely

at the University of Florida. The flashlamp pumped dye

laser was purchased commercially and assembled in our

laboratory. The instrument consists of four main sections:

the vacuum system, the optics, the digital electronics, and

the computer data-logging system. Two of these, the optics

(39) and the digital electronics have been described in

detail (42). The vacuum system is standard for icr oper-

ation (43). A functional description of all three follows.

The final sections of this chapter are devoted to a detailed

description of the computer data-logging system and an

evaluation of its performance.

The Vacuum System

Figure 3.1 is a diagram of the icr vacuum system. The

entire system was fabricated in the shops of the Chemistry

Department of the University of Florida. All components are

..._ _________ I

> U



E, c
> -0

c 0 ra
0' U E >


3 CL
i )



-n 0) E O
S r-n .-C
S : fu -- E


l 0 C
J U 3 V)

o ~E) 0-
'0 4-O U

-4 m uj to 4 4-
L 4U 0 u c 4-

SE > 4-- -



stainless steel with the exception of the ionization gauge

which is glass. The entire system is wrapped with heating

tape and can be baked-out at temperatures exceeding 1000C.

Background pressures as measured on the Bayard-Alpert ion-

ization gauge (44) (factory calibration) are routinely less than
5 x 10- torr. During experiments the reaction chamber is

pumped using a 2-inch oil diffusion pump with a liquid

nitrogen trap. When experiments are not underway the

vacuum chamber is pumped by the Vac Ion titanium sublimation


Samples are routinely freeze-pump-thaw purified two to

three times on the inlet vacuum line before introduction

into the icr. The normal procedure for sample introduction

is to pump the foreline up to the sample bulb stopcock for
10 to 20 minutes. Typically pressures of 1 x 10 torr to
5 x 10 torr as measured by a cold cathode discharge gauge

are obtained. Valves H (Figure 3.1) are then closed and

the sample bulb stopcock opened. After several minutes the

valves at I (Figure 3.1) are closed. The sample is thus

trapped between Valves I and H. Samples handled in this

way are less likely to be contaminated by air leaks at the

connection between the metal inlet system and the glass

sample bulb (Figure 3.1.E) and usually provide samples

adequate for 4 to 8 hours of operation. With water samples

the sample bulb stopcock and Valve I must be left open to

maintain adequate sample pressure but this is the lone

exception to the normal procedure. While experiments are

in progress the sample pressure of one or two gases is

maintained using the control leak valves (Figure 3.1.F).

Usually 10 to 15 minutes is required for the sample pressure

to stabilize but once set they remain stable within a few

percent for a matter of hours.

The Optical System

The light source used in these studies is a flashlamp

pumped dye laser (45). The laser head is mounted in a box

made of copper screen to shield the m.o. detector from radio

frequency noise from the flashlamp discharge. Tuning over

the spectral region between 700 nm and 450 nm is achieved

by changing dyes and using a diffraction grating at one end

of the optical cavity. The grating is moved by a sine arm

drive. A diagram of the optical system is shown in

Figure 3.2.

In our instrument an intracavity technique (39) is

used to maximize the photon flux through the sample. The

optical cavity of the laser extends from A to B in

Figure 3.2 and the sample region is located at H. The

window surfaces on the coaxial flashlamp (D) and vacuum can

(C) are antireflectivity-coated to minimize intracavity

losses. A small portion of the beam is reflected off a

clean microscope slide (E) into a 0.5 m monochromator (G) to

a) E

a -

o E
u S.-
r-- O

E U c
r .- *-
( Q E n
0 E

(/> E m
3: 11 -


0) S- m3
C) Q- O E
a OU 5-
A O 0
a 0 3 C
i- ,-- 0 0
4-- E '- O E


0 >
(0 *r-
) O
o r a

4- 4-) S- *- 4-
o rn 5.- 4-' 3
5-. *r- C O
S- c o
E o) E rm

on u







-------- C

measure the wavelength. The monochromator is calibrated with

a helium-neon laser and a mercury lamp.

Once the wavelength has been established the Gen Tec (F)

(46) fast response joulemeter is placed to intercept the

reflected portion of the laser beam to obtain a measurement

of the laser intensity. The transient from the Gen Tec

(see Figure 3.3.A) is displayed on an oscilloscope. The

energy of the laser pulse is proportional to the peak

voltage of the transient. During an experiment this value

is read from the oscilloscope and recorded manually each

time the laser triggers.

To measure the energy output of the laser it was neces-

sary to determine the fraction of the laser beam diverted

by the microscope slide. This quantity was measured by

replacing the grating at A in Figure 3.2 with a 90 percent

reflecting mirror. Fifteen laser shots were fired with the

Gen Tec intercepting the light reflected by the microscope

slide and then fifteen more were taken with the Gen Tec

measuring the throughput of the 90 percent reflecting

mirror. This experiment was then repeated. The throughput

of the 90 percent reflectance mirror was taken as 10 percent

of the intracavity energy and the light reflected from the

microscope slide was compared to it. In the first run the

ratio of energy measured off the microscope slide to the

throughput of the mirror was 0.48 and for the second run

0.54. Thus, approximately 5 percent of the intracavity

Figure 3.3. Typical Transients Generated by the Experiment

A. Transient from Gen Tec

B. Transient from m.o.

40 mV

0 20 ms


200 mV -

0 -

0 3 ms

energy was reflected to the Gen Tec. This number and the

calibration data supplied by Gen Tec were used to estimate

the energy output per laser pulse.

Laser performance depended greatly on the dye used and

the quality of the optical alignment. The laser manu-

facturer claims an output pulse duration of 1 ps. The

bandpass was on the order of a few tenths of a nanometer

and could be measured to 0.4 nm with the monochromators

used. Energy output ranged from 40 mJ per pulse to over

200 mJ per pulse depending on the dye and the wavelength.

Variation in laser output from one shot to the next ranged

from 5 percent to 25 percent depending on dye, wavelength,

and lamp voltage. In those experiments where the laser

output was varied the output was changed both by varying

the lamp power supply voltage and inserting dirty micro-

scope slides into the laser's optical cavity at I

(Figure 3.2). It was determined that identical results

were obtained in a given sample at a give laser intensity

no matter which method was used to modify the laser output.

At least once in every experiment several experimental

cycles were run with the laser firing but the window of

the vacuum can was covered so that no light entered the

sample region. The ion signal was seen to be unaffected

by the laser discharge under these conditions.

The Digital Electronics

A one-of-a-kind digital timer system (42) was con-

structed at the University of Florida to control the for-

mation trapping and detection of ions with a series of

electronic pulses of specified widths and occurring at

preset intervals. The m.o. detector was assembled at the

University of Florida following the design of Mclver (40).

This detector is a fixed-frequency design; ions are brought

into resonance by changing the magnetic field strength.

The magnet is a Varian V-3400, 9-inch low impedance

electromagnet with a V-FR2500, 7-kw power supply and a

"Fieldial" Mark I controller. A set of regulated power

supplies was constructed to provide the d.c. trapping

voltages applied to the plates of the cell.

The first pulse of an experimental cycle is the precon-

dition pulse which initializes all the counters and "clamps"

them until the start pulse. The start pulse is the refer-

ence pulse for all delays and is the zero in all time

measurements. A crystal-controlled clock oscillator is

used to control the delay time between the start pulse and

the sample, grid, quench, detect, w2, and free pulses.

These delays and the widths of the grid, quench, w2, and

detect pulses are also controlled by the crystal clock using

a system of binary-coded-decimal thumbwheel switches on the

front panel of the control unit. The accuracy of the pulse

delays and widths is that of the clock oscillator or

10 ppm at constant temperature.

Ion formation occurs during the grid pulse. Electrons

ejected from a resistively-heated rhenium filament main-

tained at a preset potential between 0 V and -50 V are

prevented from entering the sample region by a grid biased

5 V negative of the filament. During the grid pulse the

grid is biased 5 V positive of the filament and electrons

are accelerated into the cell by the negative voltage on

the filament. Across the cell they are collected through

a hole in the side plate by a plate biased to about +12 V.

The electron current is amplified so that an electron
current of 1 x 106 A yields a signal of 1 V. Under typi-
cal operating conditions electron currents of 5 x 108 A
to 3 x 10- A were employed.

The w2 pulse is used to gate a radio frequency oscil-

lator for the double resonance ejection experiment (47).

Typically an Ailtech function generator was used in the

gated mode to apply a burst of radio frequency oscillation

to the upper plate of the cell. The frequency of this

oscillation is chosen to excite the cyclotron motion of a

particular ion to such an extent that it is driven from

the cell.

The free pulse is a +5 V pulse with a duration of

about 1 ms. A pulse with an amplitude of 180 V with a rise

time of 1.8 ps is required to trigger the laser. This

pulse is provided in the following fashion. The free pulse

(or, under computer control the result of a logic AND with

a computer address, see below) is passed through a optoiso-

lator chip. The optoisolator protects the control elec-

tronics from transients produced when the laser fires. The

output of the isolator is passed to a Schmitt trigger to

decrease the rise time. This fast pulse then triggers an

SCR circuit that provides the high voltage trigger for the


The detect pulse gates the amplifier in the limiter

branch of the m.o. circuit. As Mclver mentions in the

original article describing the design of this circuit

(40), the transient output of this detector (Figure 3.3.B)

contains two spikes caused by the mismatch between the

reference circuit and the cell circuit. The actual power

absorption transient appears between the spikes. These

spikes are eliminated by using a switched integrator and

delaying the switch pulse until after the first spike has

passed and turning the integrator off before the second

spike (42). A sample pulse from the control electronics is

triggered by the falling edge of the detect pulse to gate a

sample/hold module for the 2 ms required to read the output

of the switched integrator. The output of the sample/hold

module is displayed on a chart recorder or digitized for

computer logging as described below.

The final pulse of the experimental cycle is the

quench pulse. During this time a 10 V d.c. level

Figure 3.4. Sketch of Typical D.C. Output of Switched

A. Manual Operation

B. Computer Operation



B -Lr -rLr In -L

is added to the upper cell plate to sweep all ions from

the cell.

An experimental cycle can be initiated in three ways:

single-cycle trigger, fixed-rate trigger, and automatic

trigger. The single-cycle trigger is a pushbutton on the

front panel of the control electronics. The fixed-rate

trigger causes a cycle to occur once every 20 ms to 5 s in

a 1,2,5 sequence chosen by a front panel switch. The

automatic trigger causes a cycle to be triggered 175 ms

after the end of the quench pulse. In noncomputer experi-

ments the instrument was run in the single-cycle trigger

mode when the laser was flashing and in a fixed-rate mode

with a duty cycle of 2 s or 5 s when the laser was not

flashing. Under computer control the fixed-rate trigger

with a 2 s duty cycle was used exclusively.

Interfacing the Pulsed icr
with a KIM-1 Microprocessor

The next stage in the development of the pulsed icr/

laser instrument was the design and implementation of a

microprocessor-based digital data acquisition system. The

principal advantage of such a system is that data col-

lection and reduction is greatly simplified. A second

advantage is that data can be taken more quickly. Prior

to the implementation of computer control the duty cycle

of the icr was limited by the minimum duty cycle of the

laser--approximately 10 s. Using the computer to control

the laser it was possible to collect "laser off" ion level

data between laser shots and thus decrease the time needed

to run an experiment.

The first stage of interfacing is the manipulation of

the transient signals generated by the pulsed instrument

into signals sufficiently stable to be digitized. Since

each experimental cycle yields two pieces of data--an ion

level signal and a laser intensity signal--it is necessary

that each be routed to the analog to digital converter

(ADC) in its proper turn and then stored in computer memory

in a systematic manner so that it can be retrieved later

for further processing. The combination of hardware and

software used to accomplish these tasks is described in

this section. No mention is made of the higher level data

processing because the system to carry this out is still

under construction. One spectrum calculated from data

which had passed through the digital data system was

obtained by hand manipulation of the stored data simply to

test the reliability of the system.

The m.o. power absorption signal has already been put

into a form suitable for digitization by the switched

integrator-sample/hold system discussed above. The sample/

hold device was observed to maintain the same d.c. level

for several seconds, allowing rather leisurely digitization

of the ion intensity. The transient output of the Gen Tec

fast response joulemeter has to be passed through a peak

detector circuit. Since the droop rate on this circuit is

much faster than that of the sample/hold module, the laser

intensity has to be digitized rather soon after it is

measured. In addition, a reset pulse must be applied to

the peak detector on each cycle to return it to zero.

Once the data have been manipulated into a stable d.c.

level suitable for digitization, it is necessary to route

each signal in its turn to the ADC and on into computer

memory. The data routing is under computer control so the

computer must be "told" that the experiment is under way

and that data will soon be arriving. This is accomplished

by sending a pulse from the icr's digital timing system on

the interrupt request line (IRQ) of the microprocessor.

Upon receiving an interrupt the microprocessor jumps to

a preset location in memory and begins excecuting the

program contained there. From the computer's point of view

the experiment consists of long periods of "waiting around"

and periodic interrupts followed by a burst of activity

routing the analog signals into the ADC and filing the

digitized data in memory. In practice the same pulse is

used as the interrupt pulse and the laser trigger pulse.

The first stage of the program is a timed wait for the rise

time of the Gen Tec (25 ms) so that the computer does not

look for data before they are there.


The microprocessor used as the "brain" of the data

collection system is a KIM-1 manufactured by MOS Technology.

The central processing unit is a 6500 monolithic CPU chip.

As installed in our application the computer has either

4 or 8 kilobytes (K) of memory depending on the configu-

ration used. Contact between the computer and the outside

world is maintained through two eight-bit input/output

(I/O) ports and the IRQ. In this application only one I/O

port (the A port) is used.

The interface was custom built; a block diagram appears

as Figure 3.5. KIM's access to the interface is through

the A data port (PAD). PAD is treated like any other

memory address by the computer except that its contents are

determined by (input) or determine (output) the logic level

of a peripheral device. Whether a given bit of PAD is an

input or an output is determined by the contents of a

second memory location--the Port A data direction register


Since KIM needs to send and receive data along more

than eight individual lines, four eight-channel, three-bit

multiplexers are connected directly to the A port. In all

applications the lowest four bits of PAD, (PAO, PA1, PA2,

PA3) are used as outputs to control these mutiplexers.

Pins 4 through 7 of the A port are inputs and outputs as

the need arises. PAO is attached to the "inhibit" line of






VA 4-)

each of the four multiplexers. If the contents of PAO are

a logic high then no signals pass through the multiplexers;

only when the logic level of PAO is low do signals pass

into or out of the KIM. The three Pins PA1, PA2, and PA3

are connected to the control lines of each of the multi-

plexers. The number (from 0 to 7 binary) that appears on

these lines determines which of the eight channels of each

multiplexer is passed. The output line of each multiplexer

is connected to one of the four remaining pins on PAD.

Some of the time a control signal output from the KIM

cannot be maintained at the proper level for the entire

time that is required because the KIM must move on to do

something else. The system of four quad latches in the

interface is used to hold such signals when the computer

is needed elsewhere. Each latch has a "clock" line. When

the clock line is low the output is constant no matter what

the input. When the clock line is raised to a logic high

the output is the same as the input. If the clock line is

then dropped low again the new output is maintained no

matter what the input. In this way more than four external

logic levels can be maintained by the computer using only

four I/O lines.

A twelve-bit ADC is used to convert the analog signals

from the icr into digital data that can be read by the

computer. The ADC is operated in the free-run mode and is

thus restricted to certain timing requirements.

The minimum conversion time is 25 ms. During that 25 ms

the signal on the input line is ignored. Thus, to ensure

that a given signal is digitized it must be held on the

input line for about 30 ms. At some point (and possibly

twice) during that time the ADC will read the data and

commence digitizing it. When digitization is complete the

digital data are displayed on the output pins of the ADC.

These data are maintained there throughout the next con-

version cycle, i.e., for 25 ms. During this time they must

be read by the computer. Since in this application there

is plenty of time, data are routed through the ADC at a

rate controlled by the computer's internal clock and paced

to be sure that each set of data is held on the input pin

for at least 30 ms, collected by the computer 25 ms later,

and that one ADC cycle elapses before the next data point

is sent through.

Which of the two data points, laser output or ion

intensity, is digitized is controlled by the computer

through a one-bit, two-channel multiplexer. The "1" channel

sends the laser output signal from the peak detector to the

ADC and the "0" channel sends the ion intensity signal

through. These channels are selected by the computer

through the latches subject to the timing requirements

discussed above.

Finally, the computer controls the laser by applying

a logic level through a latch to one input pin of an AND

gate. The laser trigger pulse from the control electronics

is applied to the other input pin. The result of the logic

AND is passed to the laser trigger shaping circuit. If the

logic level from the computer is a "1" the laser trigger

pulse is passed and shaped to the proper specifications to

trigger the laser.


The software used to drive this interface and collect

the experimental data was written in 6502 assembly language.

No higher level language is available for the KIM and

assembly language is the language of choice for direct hard-

ware control. Timing for the experiment was provided by

the icr digital control electronics and it is necessary to

synchronize the operations of the computer with those of

the icr. The interrupt programming capabilities of the KIM

are ideally suited to this purpose. The interrupt is

"maskable"; that is, whether or not the computer responds

is under program control. It is also a "vectored" entity,

meaning that the response to an interrupt is under program

control. In addition, the "return from interrupt" allows

the program to return to the same point at which it was

interrupted, facilitating loop programming.

The program was designed to step-through the entire

sequence of a complete spectroscopic data point consisting

of three sets of measurements: the zero level of the m.o.

(m.o. level with the icr tuned so that no ion is in

resonance), the intensity of an ion signal without the

laser flashing, and the intensity of the ion signal with

the laser flashing. There is a main program which counts

the number of experimental cycles run and signals the oper-

ator when it is time to enter a new phase of the experiment.

There is a common program for all phases of the experiment

except that different values of various counters are used.

The main program initializes these counters and otherwise

consists of a series of jumps into and out of subroutines.

Such subroutine programming makes writing and debugging the

program easier because the work is broken down into more

manageable units and it makes the program relatively easy

to modify.

A "step-through" of the operation of a typical experi-

ment (as diagrammed in Figure 3.6) illustrates how the main

program works. Assuming that all adjustments have been

made to the laser and the icr, the operator powers up the

KIM and loads the program from a tape cassette. He then

punches "AD"--meaning address--"0200" and then "GO." The

computer lights up with "0005 OA." Address 0020 is

the start of the routine that loads a set of registers needed

by the program. These include setting the interrupt vectors

so that when the icr interrupts the computer it will

respond correctly. The routine sets the initial address

at which data are to be stored and initializes the "data

index" which controls the destination in memory of each

byte of data. The default values for the number of data

display lights "0005, OA"

operator punches "+, GO"

Flow Chart of Main Program

Figure 3.6.

points to be taken, and the "ratio counter"--the number of

laser off ion level measurements to be taken between each

laser shot--are also specified at this time.

The numbers "0005 OA" signify that Address 0005 con-

tains OA, which is the hexadecimal equivalent of 10. This

means that 10 measurements of the ion zero level are to be

made. The operator may change this number. Once the

operator punches "+" then "GO" the computer loads the number

of "zero" points to be taken into the "active shot counter."

This is the address where the number of points taken, the

number of interrupts, is counted by decrementing the

contents and taking a program branch when zero is reached.

The "laser trigger counter" is set so high that the laser

will not trigger during this phase of the experiment. The

interrupt is then enabled and the computer jumps to the

"wait" subroutine.

Once the specified number of interrupts has occurred

the computer lights up with "0009 OA" where OA, 10, is the

number of times the laser is to trigger. The operator now

"tunes in" the ion and turns up the laser. Upon the

pushing of "+" and then "GO" by the operator the computer

computes a new active shot counter from the ratio counter

and the laser shot counter. Should this value exceed "FF"

(256) an error message "FFFF FF" appears on the KIM's

display. The "trigger" counter is set so that the laser

will fire at an interval determined by the ratio counter.

The computer interrupt is then enabled and the computer

again jumps to the "wait" subroutine. It again services

the required number of interrupts and returns to display

"0005 OA." The operator makes the necessary adjustments

to the icr and continues.

The computer's response to an interrupt is the same

each time it is interrupted. As the flow chart (Figure 3.7)

indicates the computer first sets the two-channel multi-

plexer to pass the photodetector signal; it then waits for

the data to be digitized. When data are ready the computer

calls the "read ADC" subroutine which stores the data.

Data are stored at an address calculated by the computer.

The address is determined by the contents of Memory Lo-

cations 0001, 0002, and 0003. The address is given by the

contents of 0002 plus those of 0001 (low-order byte) and

0003 (high-order byte). Data from the photodetector are

read into the page determined by the high-order byte of

the "data destination register" (0003) and ion intensity

data are on the next page (a page consists of 256 memory

locations). A page of laser data is thus alternated with

a page of ion data beginning at an address specified by the


Once the photodetector data have been read and stored

the two-channel multiplexer is set to pass the ion signal.

While waiting for it to be digitized the computer adjusts

the data destination. Once the ADC has been read the data

destination is reset. The data index is then incremented




wait for data


M.O." subroutine

wait for data

set data destination

read ADC

Sreset data destitination

call "LASER TRIGGER" subroutine

return from interrupt

Flow Chart of IRQ Routine

Figure 3.7.

in anticipation of the next set of data, and if the

resulting value is greater than "FF" the high-order byte of

the data destination is incremented twice so that more than

256 bytes of data can be stored without stopping. Since

each experimental cycle produces 2 bytes of ion data and

2 bytes of laser data, whether the laser fires or not,

this means that more than 128 experimental cycles can be

run without stopping. The number of cycles that can be

run is limited by the amount of memory available to store


Finally, the laser trigger counter is decremented, and

if it reaches zero a logic high is put on one input pin of

an AND gate. On the next icr cycle the laser trigger pulse

from the control electronics will be passed through the AND

gate and on to the pulse shaping circuitry. The trigger

counter is derived from the ratio counter that the oper-

ator sets.

This program does not represent the final stage of

software development. One obvious improvement would be to

discard much of the laser output data recorded when the

laser does not fire, retaining only enough to calculate a

reliable baseline. This would streamline memory require-

ments considerably. One worthwhile feature of the program

is that the operator need only concern himself with the

contents of a few addresses: the data destination (0002,

0003), the laser shot counter (0009), the zero counter

(0005), the data index (0001), and the ratio counter (00E).

The operator need not specify any of these if he is satis-

fied with the default values provided by the program or he

may change them to suit his needs. If the operator wants

to know where the next byte of data will be stored, he need

only examine the data index and the high-order byte of the

data destination. They are the low- and highrorder byte,

respectively, of the address where the next byte of data

will be stored. This is especially useful when manual data

manipulation is to be employed.

Evaluation of the Performance
of the Computer Data System

Although the hardware and software for higher level

processing of the data stored in the KIM has not yet been

assembled and consequently the manipulation of computer-

logged data is still very inconvenient, it is worthwhile to

compare data that has passed through the digital data

system with data recorded in the normal manner. During the

study of the photodissociation of 1,3,5 hexatriene reported

in Chapter 7, the instrument was set up so that data could

be recorded simultaneously by hand and on the KIM.

Figure 3.8 is a plot of the spectrum obtained from the

computer data. Data from the same experiment processed in

the normal manner are presented in Figure 7.2.


Wavelength / nm



Figure 3.8. Photodissociation Spectrum of 1,3,5 Hexatriene

Obtained in Computer Controlled Experiment

0 I
,? ,0

There is very good agreement between the two spectra;

however, two of the points (indicated by arrows in

Figure 3.8) are markedly different in the computer analyzed

spectrum. The reason for this becomes apparent when

Figure 3.9 is examined. In a different experiment, the

intensity of the laser was measured both by hand and by

computer at a variety of laser intensities. A plot of the

decimal equivalent of the computer-stored data versus the

peak height of the Gen Tec as recorded manually clearly

shows that below 30 mV of Gen Tec signal the computer system

is nonlinear in laser intensity.

Ignoring the two data points calculated with erroneous

values for the laser intensity a more detailed analysis of

the data revealed good agreement between the two data sets.

Each spectral point is the average of ten experimental

cycles. Within each set of ten there was good agreement

between the computer-logged data and the data obtained

manually. In each of six sets studied the median, largest,

and smallest cross section were calculated from the same

"raw" data.

A comparison of the hexadecimal numbers in the com-

puter memory to the data obtained manually again reveals

good agreement. The two sets of laser data have already

been compared and shown to be linear over a limited range.

The scatter observed in Figure 3.9 is probably due to the

large reading error associated with reading the Gen Tec

transient from the oscilloscope screen. The relationship

0 0

0 00

0 0

0 0
0 0



Joulemeter reading / mV
(obtained by hand)

Figure 3.9.

Comparison of Laser Energy as Measured by Computer

and Manually

00 0


between ion intensity data displayed on the chart recorder

and the decimal equivalent of the numbers stored in the

computer memory is shown in Figure 3.10. Clearly, the

relationship is linear.

Since the results obtained from the computer-logged

data generally agree with those obtained manually it can

be concluded that the computer data routing system works.

Also, the linearity of the computer-measured ion intensity

suggests that the analog to digital converter works. The

nonlinearity in the computer-logged laser output data would

appear to indicate that the peak detector circuit is not

functioning properly.

Figure 3.10. Comparison of Ion Signal From Digital (Compu-
ter) and Analog Circuits



Analog Data / arb. units

1 r 1



Obviously, the preceding theoretical discussions do

not provide adequate justification for having faith in the

results of these experiments. A means of assessing experi-

mentally the behavior of the instrument under the influence

of the many experimental variables had to be found.

Ideally, one would hope to study an ion with no vibrational

energy modes, i.e., an atomic ion. However, several experi-

mental considerations favored the use of a "compromise"

ion: OH .

The cross section for the photodetachment of OH has

been shown to be large and nearly constant between 450 nm

and 630 nm (30)--a range encompassing most of the wave-

lengths of interest here. In addition, OH- is easily

formed from a readily obtained sample,water. The actual

process for the formation of OH is via Reactions 4.1 and


e- + H20 H" + OH 4.1

H + H20 OH- + H2 4.2

The first reaction is dissociative electron attachment to

water and is most efficient for electrons with ca. 7.0 eV

of translational energy. Reaction 4.2 is extremely effi-

cient, occurring on essentially every collision (48). The

use of an ion formed by such a scheme has the advantage

that Reaction 4.2 restricts to some extent the possibility

of internal excitation in the OH ion.

It must be noted that the interpretation of these

results requires the assumption that the photodetachment

cross section of OH is constant at any given wavelength

at all times and that all variations in the measured cross

section are due to instrumental effects. Two studies of

the photodetachment of OH- have been carried out in beam

instruments (30,31) and neither has revealed any effects

of internal excitation on the cross section. Furthermore,

at photon energies above the detachment threshold such

effects should be negligible. The photon energies used in

this study were all greater than 0.25 eV above the photo-

detachment threshold. It must also be assumed that studies

of a negative ion are valid for comparison to those of a

positive ion. Since the motion of a negative ion in the

icr cell is the same as that of a positive ion (except that

all oscillations have the opposite "handedness") (17) this

assumption seems reasonable.


The Effects of Variations
in icr Operating Conditions

The effect of magnetic field strength on the apparent

cross section for the photodetachment of OH- was studied.

The potential was maintained at +1.0 V on the end,

upper, and lower plates of the cell; the side plate trap-

ping potential was -1.5 V. The sample pressure was 2.8 x

10-6 torr of H20. Pulse Sequence 1 in Table 4.1 was used

and the irradiating wavelength was 580 nm.

In order to vary the magnetic field strength it was

necessary to vary the m.o. frequency. Three frequency

settings were used: 153.5 kHz (100 G/amu), 306.6 kHz

(200 G/amu), and 615.0 kHz (400 G/amu). These corres-

ponded to resonant field strengths of 1,727 G, 3,450 G,

and 6,850 G, respectively. The percent detachment of OH

versus laser intensity was measured at each field strength.

The effect of changes in the trapping, upper, lower,

and end cell plate potentials on the apparent photodetach-

ment cross section was also studied. A sample pressure

of 2.5 x 10-6 torr of water and Pulse Sequences 2 and 3 of

Table 4.1 were used. The m.o. frequency was 615.0 kHz

and hence the resonant field was 6,850 G.

Three sets of potentials were tested:

1. Side plate trapping potentials of -1.0 V--

upper, lower, and end plate potentials of

Table 4.1.

Pulse Sequences Used in the Study of
OH- Photodissociation

Pulse Grid Pulse Laser Pulse Detect Pulse
Sequence Delay Width Delay Width Delay Width
Number (ms) (ms) (ms) (us) (ms) (ms)

1 0 12 19 1 20 2.0

2 0 15 34 1 35 2.0

3 0 15 49 1 50 2.0

4 0 25 149 1 150 2.0

5 0 25 249 1 250 2.0

+1.0 V,

2. Side plate trapping potentials -1.5 V--

upper, lower, and end plate potentials of

+1.5 V, and

3. Side plate trapping potentials of -2.5 V--

upper, lower, and end plate potentials of

+2.0 V.

These potentials were chosen as typical of normal

operating conditions. Ion signal quality was independent

of cell plate potentials over this range. The use of po-

tentials much different than these led to poorer ion

signal-to-noise ratios. Percent detachment versus laser

intensity was measured at each set of trapping potentials.

In some samples the magnitude of the ion signal was

seen to vary considerably from one experiment to the next

due to irreproducibility of the emission current. Several

types of filaments were employed at various times in this

research. In one experiment the ion intensity was varied

by about a factor of four by changing the current through

the filament. The m.o. frequency was 615.0 kHz; sample

conditions, pulse sequence, and trapping potentials were

identical to those used to study the effect of magnetic

field strength. The apparent photodetachment cross section

was measured at both signal intensities.

The Effect of Time

One of the principal aims of this research was to

exploit the time resolution of the pulsed icr to obtain

information about the dynamics of changes in the structure

of molecular ions using photodissociation spectroscopy as

a probe. Time-resolved studies of the photodetachment of

OH were carried out to examine the nature and magnitude

of time-dependent instrumental artifacts.

Experiments were carried out at three sample

pressures: 2.5 x 10-6 torr H20, 5.0 x 10-6 torr H20, and
1.0 x 10- torr H20. In each case Pulse Sequences 1

through 5 in Table 4.1 were used. The m.o. frequency was

615.0 kHz so OH was in resonance at a magnetic field

strength of 6,850 G. The irradiating wavelength was 620 nm

and kiton red S dye was used. The dependence of the

percent detachment on laser intensity was measured in each

sample using each pulse sequence.

The Effects of Variations in Laser Parameters

This study of the variation in the apparent photo-

detachment cross section of OH" was carried out during a

period of several months. Results were collected with

several dyes and at several different laser alignments.

Great pains were taken to maintain the same icr oper-

ating conditions from experiment to experiment. The

pulse sequence, trapping potentials, and sample pressure

were the same as those previously described for the study

of the influence of magnetic field strength. The m.o.

frequency was 615.0 kHz for all of these experiments corres-

ponding to a resonant magnetic field strength of 6,850 G.

The conditions were chosen to provide a standard for

comparison to the data obtained with chloroethane as re-

ported in the following chapter (Chapter 5). The experi-

mental conditions, including field strength, were as close

to those used in the study of chloroethane as was feasible.


The Effects of icr Operating Conditions

The plot of percent disappearance of OH versus laser

intensity at three magnetic field strengths is shown in

Figure 4.1. Each point is the result of one laser shot.

Curve A was obtained at a field strength of 6,850 G;

Curve B at 3,450 G, and Curve C at 1,727 G. Clearly the

apparent cross section increases with increasing field.

However, the form of the dependence on field strength is

not obvious from this experiment, nor is it obvious from

the theory. Changes in the cell plate potentials and the

magnitude of the ion signal made no difference in the

measured cross section.

For each pressure listed above and for Pulse Sequences

1 through 5 listed in Table 4.1 the dependence of percent

detachment on laser intensity was measured. The results




















U ')


oj r-


L- r-



00 000



I I I i A i



jua43ipeaG IuaJead









obtained at 2.5 x 106 torr are shown in Figure 4.2.A,
those obtained at 5.0 x 10 torr are shown in Figure 4.2.B,
and those obtained at 1.0 x 10 are shown in Figure 4.2.C.

The data are plotted as laser energy versus percent detach-

ment. Each point represents the average of five laser

shots of approximately the same energy. A plot of the

percent detachment at a laser energy of 45 mJ versus detect

delay time is shown in Figure 4.3. Again Curve A refers

to the 2.5 x 10- torr sample, Curve B to the 5.0 x 10-6
torr sample, and Curve C to the 1.0 x 10- torr sample.

An indication of the physical basis of this effect

is obtained by comparing the percent detachment at 45 mJ

versus the "mean collision number," N as in Figure 4.4.

The rate constant for orbiting collisions between an ion

and a molecule or atom (49), kL, is given by:

k = 2n(ae2/ )I/2 4.3

where e is the charge on the ion, a is the polarizability

of the neutral species, and p is the reduced mass of the

collision partners. The frequency of orbiting collisions,

Z1, is:

1 = kLN.











C c







4i E

II 11

S- S.- S-
o 0) 0

4-1- 4-) 4-)
(0 tO (0

r- .- r-
o 0 0


a) a) a)

1I I1 If





< N









o o
-4 *-,


- 0


o o

4uaPQe4OGa 4uaaad









0 r-- r-- r-
S -. S.- 5-
) 0 o 0 0


4 0 0 0O

C X X x

(0 LO 0 0C

c co
0 0




I l
I I/


I < C
O -s C)
c (A
I- I i

O, O < m
D- -- I II

iuawqo3eaa 4uaOJad

Figure 4.4. Percent Detachment of OH- Versus
"Mean Collision Number"

[ = Data Obtained with
2.5 x 10-- torr Water

o = Data Obtained with
5.0 x 10-6 torr Water

A = Data Obtained with
1.0 x 10-6 torr Water





w 50
u \
ro \

20 40 60

Mean Collision Number

The number density of the neutral gas, N, is calculated

from the ideal gas law. The "mean collision number," Nc,


Nc = Zltd 4.5

where td is the detect delay time. It is then the number

of orbiting collisions an ion moving through an ideal gas

of the pressure of the neutral sample undergoes between

initiation of the experiment and photolysis and detection.

It has been shown that the trapping and drift potential

fields in the trapped ion icr cell accelerate the ion

only a small fraction above its thermal speed (17). It can

clearly be seen that the three curves obtained at the

three different pressures become nearly identical when

percent detachment is plotted against "mean collision

number." This suggests that the time-dependence of the

apparent photodetachment cross section is related to the

collisional equilibration of the sample. This appears

to occur very rapidly--the greatest change in percent

detachment occurs in the first four collisions.

Effects of Variation in Laser Parameters

Figure 4.5 shows two curves (A and B) obtained with

the same dye (rhodamine 6 G) at different wavelengths and

with the same laser alignment. Curve C is the result

obtained under identical conditions to those used to


\ \


I i I -1 1 11iII 1 I

4uaW424e0aQ ua2aid











Full Text
xml version 1.0 encoding UTF-8
REPORT xmlns http:www.fcla.edudlsmddaitss xmlns:xsi http:www.w3.org2001XMLSchema-instance xsi:schemaLocation http:www.fcla.edudlsmddaitssdaitssReport.xsd
INGEST IEID ESEL7SVY4_HFL1G0 INGEST_TIME 2012-03-09T01:27:12Z PACKAGE AA00003897_00001