Title: Double resonance spectroscopy as a diagnostic tool and analytical technique for atomic spectroscopy
CITATION THUMBNAILS PAGE IMAGE ZOOMABLE
Full Citation
STANDARD VIEW MARC VIEW
Permanent Link: http://ufdc.ufl.edu/UF00102756/00001
 Material Information
Title: Double resonance spectroscopy as a diagnostic tool and analytical technique for atomic spectroscopy
Physical Description: Book
Language: English
Creator: Leong, Moi Bon, 1959-
Copyright Date: 1988
 Record Information
Bibliographic ID: UF00102756
Volume ID: VID00001
Source Institution: University of Florida
Holding Location: University of Florida
Rights Management: All rights reserved by the source institution and holding location.
Resource Identifier: ltuf - AFK9830
oclc - 19908597

Full Text
















DOUBLE RESONANCE SPECTROSCOPY AS A DIAGNOSTIC TOOL
AND ANALYTICAL TECHNIQUE FOR ATOMIC SPECTROSCOPY






BY







MOI BON LEONG


A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL
OF THE UNIVERSITY OF FLORIDA IN
PARTIAL FULFILLMENT OF THE REQUIREMENTS
FOR THE DEGREE OF DOCTOR OF PHILOSOPHY


UNIVERSITY OF FLORIDA


1988































Dedicated to my father, my brother, and my sister for thei-

undying support and trust in me. In memory of my mother who did not

live to see the growth of her son but will always be remembered by

him for the time she spent when she was here.
















ACKNOWLEDGMENTS


I would like to sincerely thank Dr. James D. Winefordner for his

guidance, inspiration, and encouragement during my three and one half

years in the finest spectroscopy group anywhere. It has been both a

learning experience and a great joy to work with such a distinguished

and friendly individual. I would like to thank you for the

opportunity.

I would like to acknowledge Dr. Benjamin Smith and Dr. Nicolo

Omenetto for their invaluable support and help as well as their

wisdom and guidance.

Lastly, I would like to thank the friends that I have made

during my quest, because without them it would not have been nearly

as fun and enjoyable. Of this host of friends, I especially thank

the clan for allowing me to eat with them. Thanks for the good

times, and I hope we remain friends and not strangers. I hope to see

you down the road: Brad, Tony, Benny, Jorge, Tom, Mark, Mike M.,

Mike R., Wellington, Richard, Chris, Joe, Ben, Andres, Doug, Leigh

Ann, Tiing, the secretaries (Jeanne, Chris, Susan, and Robin), and

Dave Berberich.















TABLE OF CONTENTS


Page

ACKNOWLEDGMENTS..................................................iii

LIST OF TABLES............................................ .......vi

LIST OF FIGURES... ..................... ....................... ...vii

ABSTRACT........................................................ .xii

CHAPTERS

1 INTRODUCTION............................................ .

Basic Principles of Two-Photon Methods..................... 1
Brief Review of Double Resonance Spectroscopy for
Atomic Species ...................................... 4
Intent of Dissertation................................... 5

2 FLUORESCENCE DIP SPECTROSCOPY FOR THE MEASUREMENT
OF ATOMIC PARAMETERS......................................8

Introduction to Fluorescence Dip Spectroscopy.............8
Experimental Facilities and Considerations ..............17
Results and Discussion...................................25
Conclusions............................................ 55

3 ATOMIC FLUORESCENCE AND IONIZATION MECHANISM FOR
LEAD IN AIR-ACETYLENE FLAME............................... 61

Basic Principles of Atomic Fluorescence and
Ionization Spectroscopies..............................61
Experimental Facilities and Considerations................64
Results and Discussion...................................69
Conclusions........................................... ..79

4 MEASUREMENT OF ATOMIC FLUORESCENCE FOR LEAD IN A
GRAPHITE TUBE ATOMIZER...................................81

Introduction to Graphite Tube Atomizers..................81
Brief Review of Atomic Fluorescence Using Graphite
Furnace Atomization....................................83









Experimental Facilities and Considerations...............85
Results and Discussion................................... 94
Conclusions................ ......................... .. 108

5 FINAL COMMENTS AND FUTURE WORK .......................... 111

APPENDICES

A GLOSSARY OF TERMS AND SYMBOLS ...........................114

B LIMITING CASES FROM FIGURE 2-4...........................116

C SAHA EQUATION...... ................... ... ........... 117

REFERENCES....................................................... 118

BIOGRAPHICAL SKETCH.............................................. 125















LIST OF TABLES


Table Page

2-1 Experimental Components for Fluorescence Dip
Spectroscopy........... ........ ....... ...... ...... .19

2-2 Experimental Parameters for Fluorescence Dip
Spectroscopy .......................................... 23

2-3 Comparison of Gate Widths with Respect to Slope,
Intercept, and r2 Correlation Coefficient..................30

2-4 Comparison of Relative (Ratio) of Atomic Parameters.......43

2-5 Comparison of Steady State Saturation Dip Parameter
with Respect to the Second Laser Excitation Wavelength....54

3-1 Experimental Components for Laser Enhanced Ionization
and Laser Excited Atomic Fluorescence Spectroscopies......66

3-2 Experimental Parameters for Laser Enhanced Ionization
and Laser Excited Atomic Fluorescence Spectroscopies......70

4-1 Experimental Components for Double Resonance Laser
Excited Atomic Fluorescence in a Graphite Tube
Atomizer............................ ...................... 87

4-2 Experimental Parameters for Double Resonance Laser
Excited Atomic Fluorescence in a Graphite Tube
Atomizer ............................................... 93

4-3 Peak Fluorescence Signals for 100 pg of Lead and
Related Noise Figures....................................103

4-4 Laser Excited Atomic Fluorescence of Lead in a
Graphite Tube Atomizer: Absolute Detection Limits
as Obtained by Single Step and Two-Step Excitation.......105















LIST OF FIGURES


Figure Page

1-1 Diagrammatic Representations of (a-c) Two-Photon
and (d) Double Resonance Excitation Schemes................2

1-2 Double Resonance Excitation Schemes Followed by
(a) Fluorescence or (b) Ionization Detection...............6

2-1 Generic Energy Level Diagram for Fluorescence
Dip Spectroscopy..........................................9

2-2 Cross-Sectional Area of the Atomizer Including
Excitation Beam and Fluorescence Geometry and
Prefilter and Postfilter Effects ..........................11

2-3 Rate Expressions for a Three-Level System.................13

2-4 Expression for the Steady State Population of the
First Excited State in Terms of the Radiative and
Collisional Rate Coefficients............................. 14

2-5 Theoretical Expressions for Fluorescence Dip
Spectroscopy for the Limiting Case of Optical
Saturation of the First Transition........................15

2-6 Block Diagram of Experimental Setup for Fluorescence
Dip Spectroscopy......................................18

2-7 Partial Energy Level Diagram for Sodium...................21

2-8 Scan of the Second Excitation Laser Through the
Upper Level Transitions of Sodium at 0.10 mJ per
Pulse with the Fluorescence Wavelength at 589.6 nm........26

2-9 Scan of the Second Excitation Laser Through the Upper
Level Transitions of Sodium at 0.73 mJ per pulse with
the Fluorescence Wavelength at 589.6 nm...................27

2-10 Scan of the Second Excitation Laser Through the Upper
Level Transitions of Sodium at 8.33 mJ per pulse with
the Fluorescence Wavelength at 589.6 nm...................28










2-11 Plot of the Relative Dip and the Relative Fluorescence
Intensity as a Function of Observation Height Above
the Load Coil.............................................. 31

2-12 Measurement of the Relative Fluorescence Intensity at
588.995 nm With and Without the Second Laser at
568.266 nm as a Function of the Second Laser Irradiance...33

2-13 Measurement of the Relative Fluorescence Intensity at
589.592 nm With and Without the Second Laser at
568.266 nm as a Function of the Second Laser Irradiance...34

2-14 Reciprocal Plot of Relative Dip vs Spectral Energy
Density with 2 ns Gate Width and 568.3 nm Transition......35

2-15 Reciprocal Plot of Relative Dip vs Spectral Energy
Density with 2 ns Gate Width and 568.8 nm Transition......36

2-16 Reciprocal Plot of Relative Dip vs Spectral Energy
Density with 4 ns Gate Width and 568.3 nm Transition......37

2-17 Reciprocal Plot of Relative Dip vs Spectral Energy
Density with 4 ns Gate Width and 568.8 nm Transition......38

2-18 Reciprocal Plot of Relative Dip vs Spectral Energy
Density with 600 ns Gate Width and 568.3 nm Transition....39

2-19 Reciprocal Plot of Relative Dip vs Spectral Energy
Density with 600 ns Gate Width and 568.8 nm Transition....40

2-20 Theoretical Reciprocal Plot of Relative Dip vs
Spectral Energy Density...................................41

2-21 Plot of Relative Dip vs Spectral Energy Density with
589.0 and 568.3 nm Excitation and Fluorescence
Detection at 589.0 nm...................................... 44

2-22 Plot of Relative Dip vs Spectral Energy Density with
589.0 and 568.3 nm Excitation and Fluorescence
Detection at 589.6 nm..................................... 45

2-23 Plot of Relative Dip vs Spectral Energy Density with
589.0 and 568.8 nm Excitation and Fluorescence
Detection at 589.0 nm.....................................46

2-24 Plot of Relative Dip vs Spectral Energy Density with
589.0 and 568.8 nm Excitation and Fluorescence
Detection at 589.6 nm.....................................47


viii










2-25 Plot of Relative Dip vs Spectral Energy Density with
589.6 and 568.3 nm Excitation and Fluorescence
Detection at 589.0 nm.....................................48

2-26 Plot of Relative Dip vs Spectral Energy Density with
589.6 and 568.3 nm Excitation and Fluorescence
Detection at 589.6 nm......................................49

2-27 Plot of Relative Dip vs Spectral Energy Density with
589.6 and 568.8 nm Excitation and Fluorescence
Detection at 589.0 nm.....................................50

2-28 Plot of Relative Dip vs Spectral Energy Density with
589.6 and 568.8 nm Excitation and Fluorescence
Detection at 589.6 nm.....................................51

2-29 Theoretical Plot of Relative Dip vs Spectral Energy
Density................................................. 53

2-30 Partial Energy Level Diagram for Palladium.................57

2-31 Measurement of Relative Fluorescence Intensity at
344.1 nm With and Without Laser Excitation at
565.5 nm as a Function of Laser Irradiance................58

2-32 Partial Energy Level Diagram for Calcium (II) Ion.........60

3-1 The Three Basic Pathways of Atomic Fluorescence:
a) Resonance Fluorescence, b) Direct Line
Fluorescence, and c) Stepwise Line Fluorescence...........62

3-2 Block Diagram of Experimental Setup for Laser
Enhanced Ionization and Laser Excited Atomic
Fluorescence Spectroscopies...............................65

3-3 Partial Energy Level Diagram for Connected Double
Resonance Excitation and Ionization of Lead...............67

3-4 Partial Energy Level Diagram for Connected Double
Resonance Excitation of Lead..............................71

3-5 Scan of the Second Excitation Laser With the
Fluorescence Wavelength at 239.379 nm and the First
Excitation Laser at 283.306 nm for Lead...................73

3-6 Scan of the Second Excitation Laser With the
Fluorescence Wavelength at 261.418 nm and the First
Excitation Laser at 283.306 nm for Lead...................74










3-7 Temporal Behavior of the Ionization Signal as
Recorded by the Oscilloscope With Full Laser
Irradiance in Both Beams (283.306 and 600.193 nm).........76

3-8 Temporal Behavior of the Ionization Signal as
Recorded by the Oscilloscope with the Laser
Irradiance at 600.193 nm Decreased by a 100-Fold
and the 283.306 nm Laser Irradiance Unchanged.............78

4-1 Graphite Furnace Designs..................................82

4-2 Block Diagram of the Double Resonance Laser Excited
Atomic Fluorescence in a Graphite Tube Atomizer...........86

4-3 Graphite Tube Atomizer Setup...............................90

4-4 Chart Recorder Tracings of the Furnace Emission
Noise at the Four Fluorescence Wavelengths
Investigated in this Work: a) 405.783 nm,
b) 261.418 nm, c) 239.379 nm, and d) 216.999 nm...........97

4-5 Boxcar Output for Laser Induced Noise into the
Detector System for Single Resonance Excitation
at 283.306 nm and Fluorescence Wavelength at
405.783 nm: a) 10% Transmission Neutral Density
Filter Placed Between the Laser and the Graphite
Furnace and b) 1% Transmission Neutral Density
Filter Placed as in a.....................................98

4-6 Boxcar Output for Laser Induced Noise into the
Detector System for Double Resonance Excitation at
283.306 and 600.193 nm and Fluorescence Wavelength
at 216.999 nm: a) Laser Operated at Full Power,
b) 10% Transmission Neutral Density Filter Placed
Between the First Laser (283.306 nm) and the Graphite
Furnace, c) 1% Transmission Neutral Density Filter
Placed as in b.......................................... 99

4-7 Boxcar Output for Fluorescence of 100 pg of Lead
With Excitation at 283.306 and 600.193 nm and the
Fluorescence Wavelength at 216.999 nm: a) Both
Laser Operated at Full Power, b) 10% Transmission
Neutral Density Filter Placed Between the First
Laser (283.306 nm) and the Graphite Furnace, and
c) 1% Transmission Neutral Density Filter Placed
as in b................................................ 102

4-8 Partial Energy Level Diagram for Disconnected
Double Resonance Excitation of Lead......................106











Comparison of Connected and Disconnected Double
Resonance Excitation of Lead..............................108















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

DOUBLE RESONANCE SPECTROSCOPY AS A DIAGNOSTIC TOOL
AND ANALYTICAL TECHNIQUE FOR ATOMIC SPECTROSCOPY

BY

MOI BON LEONG

April, 1988

Chairman: James D. Winefordner
Major Department: Chemistry

Double resonance spectroscopy is a technique whereby two lasers

(i.e., two-color) are used to excite two real and different atomic

transitions in a species of interest. The first step of the excita-

tion process is the absorption of photons by the atomic species to

the first excited state from the ground state. This is subsequently

followed by a secondary absorption of photons from the first excited

state to the second excited state. Once the atomic species have

reached the second excited state, there are several methods for

monitoring the events that occur, among these are ionization and

fluorescence measurements. In the present study, three different

measurement methods are used with three different atomization

cells. The measurements demonstrate the usefulness of double

resonance excitation processes as a diagnostic tool for flames,

plasmas, and possibly, graphite furnaces, as well as an analytical











method for the measurement of lead at femtogram levels in a graphite

tube atomizer.


xiii















CHAPTER 1
INTRODUCTION


Basic Principles of Two-Photon Methods

The development of pump lasers such as the nitrogen, excimer

(i.e., XeCl), and Nd:YAG (Neodymium:Yttrium Aluminum Garnet) together

with the organic dye laser has opened up the realm of analytical

atomic spectroscopy. Because organic dye lasers are not stand-alone

instruments, they are always coupled with a pump laser such as the

ones mentioned above. A majority of the analytical work to date has

thus involved the use of a pump laser in conjunction with an organic

dye laser. The addition of a second dye laser to this system has

provided the opportunity to perform two-color experiments.

Two-photon excitation methods usually involve a minimum of three

levels. These levels, however, are not necessarily real and well-

defined (i.e., virtual levels and ionization continuum). Figure 1-1

illustrates several types of two-photon excitation schemes. Two-

photon excitation taken in its basic form involves the utilization of

the same wavelength twice. The energy from the two photons excites

the atom from the ground state to a real excited state via a virtual

level that is midway between the ground and real excited states. The

real excited state must have the same symmetry and multiplicity as

that of the ground state for it to be a two-photon allowed transition

as shown in Figure 1-la. However, as shown in Figures 1-1b and 1-1c,










/7V


(a)


/


(b)


Z/ZZLZZ/
A


(c)


/7V


(d)


Diagrammatic Representations of (a-c) Two-Photon and (d) Double Resonance Excitation Schemes


Figure 1-1










the ionization continuum can be considered as a level. In both of

these cases, the second photon promotes the species of interest from

a real or virtual level to the continuum, but not a well-defined

level in terms of the specific state (e.g., symmetry and multi-

plicity) to which the species has been promoted. These cases have

been used to investigate atoms that are not easily accessible by one-

photon excitation schemes (1-24).

Double resonance excitation is a limiting case of two-photon

excitation (25). This excitation scheme incorporates three real,

distinct levels. The species of interest absorbs photons of a

discrete energy such that excitation occurs from the ground state to

the first excited state. Once the species are in the first excited

state, a second set of photons with a different and discrete energy

than that of the initial excitation is absorbed such that promotion

of the species occurs from the first excited state to the second

excited state as shown in Figure 1-1d. Thus, all further discussion

pertains to this specific excitation scheme.

To populate the first excited state, the wavelength of the first

laser is tuned to a resonance transition of the atomic species. This

first transition should be saturated; that is, there is an equal

population in the ground state and the first excited state due to the

laser irradiance. Once this has been attained, an increase in laser

irradiance should have no or little effect on the overall distribu-

tion between the two states, and hence, fluctuations in laser

irradiance have no effect. The second excited state is then

populated when the wavelength of the second laser is tuned to the










transition that couples the first and second excited states. Again,

saturation is desired but not always necessary because the laser

irradiance of the second laser can be varied to provide the optimal

performance characteristics of the experiment.



Brief Review of Double Resonance Spectroscopy
for Atomic Species

The technique of double resonance excitation by two lasers has

been growing primarily due to the development of more powerful

lasers. With this development, more work has gone into the investi-

gation of atoms in the ultraviolet (200 nm) to the near infrared

(-900 nm) regions (26-38). The more powerful pulsed lasers have

provided high pumping rates such that the rate of absorption exceeds

the radiative and/or radiationless deactivations (25) (e.g., colli-

sional de-excitation). The pumping rate of the first laser, that is,

the rate at which the first excited state is populated, must exceed

(i.e., beat) the deactivation processes from that first excited

state. If this does not occur, then there is not a sufficient number

of species in that first excited state to populate, to a great

extent, the second excited state.

A majority of the publications to date have been studies of the

high lying (Rydberg) states that are close to the ionization

continuum (26-28,32,35,36) as well as studies of the fine and

hyperfine structures of specific levels (29-31,33,34). More

recently, fluorescence and ionization measurements have appeared in

the literature with promising and exciting results (39-48).










Fluorescence monitoring from the second excited state to states below

the first excited state or even to the ground state as well as those

transitions that lie between the first and second excited states

(Figure 1-2a) have provided several interesting advantages. These

advantages include high selectivity, the detected fluorescence can be

blue-shifted relative to both pumping wavelengths (reduction of laser

scatter effects), and transition probabilities are generally greater

for those transitions originating from the second excited state than

the first excited state (39,40). Ionization measurements, e.g., two-

color laser enhanced ionization in flames, have provided some of the

best detection limits for several elements (41-48) in analytical

atomic spectroscopy. The primary advantage was that the atoms were

much closer to the ionization continuum than in a one-color situation

(Figure 1-2b), and hence, the collisional processes that promoted the

atoms into the continuum were significantly increased. In addition

to this, there was increased selectivity as well as increased

sensitivity by one to three orders of magnitude. Thus, double

resonance excitation schemes have opened up new avenues of

exploration in analytical spectroscopy.



Intent of Dissertation

In this dissertation, double resonance excitation schemes were

investigated in three different atomization sources [i.e., flames,

inductively coupled plasmas (ICPs), and graphite furnaces] to obtain

information about the atomic parameters of elements, mechanistic



























'I'





'I'


(a)


( b)


Figure 1-2 Double Resonance Excitation Schemes Followed by (a)
Fluorescence or (b) Ionization Detection





7



studies in laser enhanced ionization (i.e., temporal separation of

photoionization and collisional ionization), and the measurement of

ultratrace levels of lead by atomic fluorescence. Three different

techniques were used to gather this information, and they are

discussed individually. These excitation schemes have demonstrated

the usefulness and the viability for analytical atomic spectroscopy.
















CHAPTER 2
FLUORESCENCE DIP SPECTROSCOPY FOR THE
MEASUREMENT OF ATOMIC PARAMETERS


Introduction to Fluorescence Dip Spectroscopy

Principles of Fluorescence Dip

Fluorescence dip is, as the term implies, a dip in the

fluorescence. To better explain this phenomenon, it is best

illustrated with a generic energy level diagram (Figure 2-1). Atoms

are produced from the atomization of a liquid sample, converted to an

aerosol and then to a dry particle and finally to submicroscopic

species such that the species of interest are lying in their ground

state. The first excitation laser is fixed to a resonance transition

of the analyte such that the first excited state is populated. The

extent of this population should be in an optical saturation mode;

that is, both the ground and first excited states are equally

populated assuming gl = g2. The resulting fluorescence from the

first excited state, whether it be a resonant or nonresonant

situation, is monitored. The second excitation laser is allowed to

pass such that this laser beam arrives at the center of the

atomization cell temporally and spatially coincident with respect to

the first excitation laser beam. This second laser excites atoms

from the first to the second excited state.

A fluorescence dip occurs because the first excited state has

been depleted. This depletion is a function of the average laser






























































Figure 2-1 Generic Energy
Spectroscopy


Level Diagram for Fluorescence Dip


////////-///////-///////


\I/










irradiance of the second excitation laser. It is important to note

that the laser irradiance of the second excitation laser must exceed

the radiative and radiationless deactivation pathways from the second

excited state.

Theoretical Considerations

The theory, as proposed by Omenetto et al. (49), was the first

general treatment of the fluorescence dip as it pertained to atoms

and/or ions in atomic spectroscopy. Several assumptions were made:

1) the time behavior of the laser pulses was approximated by a step-

like function, i.e., with a rise time equal to zero; 2) the lasers

were spatially homogeneous in the fluorescence volume; 3) the atomic

vapor was dilute (no self-absorption, prefilter, and postfilter

effects); 4) the spectral bandwidths of the excitation lasers were

much greater than the absorption profiles; and 5) the rate equation

approach was considered valid; i.e., coherence effects were neglected

(49). More discussion of points 3 and 5 is presented below.

The atomic vapor is dilute if there are no self-absorption,

prefilter, and postfilter effects. Self-absorption occurs from the

reabsorption of fluorescence photons, within the excitation volume,

as they traverse the atom reservoir (50). Prefilter effects are

those regions in which the analyte of interest is present and is

illuminated by the excitation beam but the resulting fluorescence is

not viewed by the detector (Figure 2-2) (50). Postfilter effects are

those regions in which the analyte of interest is present but the

region is not illuminated by the excitation source (Figure 2-2) (50).








EXCI TAT I ON

LASER
L A S E R ......... :X






PREF I LATER R
REGION


POSTF I LT ER


REGION


FLUORESCENCE


Figure 2-2 Cross-Sectional Area of the Atomizer Including Excitation Beam and Fluorescence Geometry and
Prefilter and Postfilter Effects










The rate equation approach is considered valid if coherence

effects are neglected. Coherence effects are interactions between

the analyte of interest and the laser field. This usually occurs at

very high pumping rates (i.e., 1010 s-1 or greater). The high

pumping rates result in short interaction times (i.e., less than

100 ps). At such short times, the rate equation approach fails

because the atoms are oscillating between the first and second

excited states. Thus, the atoms are not exclusively lying in the

first excited state. At longer times, 1 ns or greater, the coherence

effects are negligible because they have been damped out.

The rate equations for the system illustrated in Figure 2-1 (see

Appendix A) are presented in Figure 2-3 (49). The time dependent

solutions for the population of the levels can be obtained after many

mathematical manipulations but are not shown here. However, the

conclusions from these solutions indicate that when the radiatively

induced rates significantly exceed the collisional rates, the

attainment of the steady state population of the final level is also

reached, and that when the first excitation laser is capable of

optical saturation, the steady state population of the ground and

first excited states depends upon the values of the radiatively

induced rate of the second excitation laser (49).

From the time dependent solutions and further mathematical

manipulations, it is possible to write an expression for the steady

state population of the first excited state in terms of the radiative

and collisional rate coefficients as shown in Figure 2-4 (see













dn ,(t)
= R23n, (t ) -(R3+R ,) n, ( t )
dt

dn, (t)
R12 n, (t) +R32 3n (t) -( R23+R2,) n (t )


n, (t) +n (t ) +n,(t )


= n


Figure 2-3 Rate Expressions for a Three-Level System


(1)



(2)


(3)








e,.p(V.i)Z +g b12P(~ )B23P(V23)

23P(l) [kl + B,,p(V) g'[Z2+1,.P(V, )g]+z3[z2+gB,,P(9 12


where the subscript means the second loser(2->3)is "on" as well


the first


loser(1l- 2)


Z3 = A+k.2+k,

Z2 = A+k.(

'- (92/93)


Figure 2-4


Expression for the Steady State Population of the First Excited State in Terms of the
Radiative and Collisional Rate Coefficients


(4)


and


4a)

4b)

4c)


4d)










Appendix A). Utilizing this equation, the limiting cases (i.e.,

absence of the second excitation laser; optical saturation of the

first transition; optical saturation of the second transition but not

the first transition; and optical saturation of both transitions) for

the steady state population of the first excited state can be derived

(see Appendix B). It can be shown that the atomic population is

redistributed among the three states according to their respective

statistical weights (39).

Knowing these equations, it is now possible to define the

relative steady state fluorescence dip, A', as shown in equation 5,

Figure 2-5, where (off) and (on) correspond to whether the second

laser is off [i.e., limiting case of absence of the second excitation

laser, B 23( 23) = 0] or on. This expression can also be written

explicitly as shown in equation 6, Figure 2-5 (see Appendix A). With

this expression, two graphical representations are possible. The

first is a reciprocal plot, i.e., (A')- vs [B p23 ) 1, and the

second is a conventional plot of A' vs B 23p(23). The reciprocal

expression of equation 6 is equation 7 in Figure 2-5 (see Appendix

A). It becomes apparent that a reciprocal plot as that described

above is not practical because it assumes that certain atomic

parameters are known (e.g., transition probabilities). However, the

quantum efficiency of the second transition is shown as equation 8 in

Figure 2-5 (see Appendix A) and the radiatively induced absorption

rate is listed as equation 9 in Figure 2-5 (see Appendix A). By

substituting equations 8 and 9 into equation 7, one obtains the

expression shown in equation 10 (Figure 2-5) (see Appendix A). A










n2 (of


- "2
(off)


A=3


23 ( 2 =


92 8rh'3 2
23


8irhv3
23
c55


SP (2 3 )


A' 9+ C-3-P )
9= +92 + +93 91 2 P (V2,) ]
91 '92 93 91 -92 8rhv3
-~ 2 3-


8 =


Y32


Figure 2-5


Theoretical Expressions for Fluorescence Dip Spectroscopy
for the Limiting Case of Optical Saturation of the First
Transition


(5)


A32+k32 +k31


(6)


(7)


(8)


(9)


(10)




(11)


(12)


(on)


f )
n2









plot of (A')1 vs [p(V3) would result in a straight line where

the slope can be used to evaluate the quantum efficiency, Y32. This

is possible because there are only constants left in the expression

for the slope.

A conventional plot of A' vs B2 3(v23) is also not practical for

the same reasons discussed above. With the substitutions previously

made, an expression (reciprocal of equation 10) is obtained as shown

in equation 11 (Figure 2-5) (see Appendix A). By plotting

A' vs p(v23), a full steady state saturation dip curve can be

obtained. Such a plot has two distinct features: the first is a

linear portion and the second is a saturation plateau. The inter-

section point of these two asymptotes constitutes the steady state

saturation dip parameter. The steady state saturation dip parameter

is defined as the value of the spectral energy density for which the

relative dip is equal to one half the maximum value of the relative

dip (plateau region). Once this value has been determined, the

quantum efficiency can then be calculated by equation 12 (Figure 2-5)

(see Appendix A) (49).



Experimental Facilities and Considerations

Instrumentation

A block diagram of the experimental setup is shown in Figure 2-6

and a detailed listing of the experimental components is described in

Table 2-1. A frequency doubled Nd:YAG laser (i.e., 532 nm) operated

at 30 Hz was used as the pumping source for a dual dye laser system.

The pump laser beam was split equally to pump each dye laser.




































50/50 BEAMS





POWER

SUPPLY










Figure 2-6


Block Diagram of Experimental Setup for Flutorescence Dip Spectroscopy










Experimental Components


Component


Nd:YAG laser


Dual dye lasers


RF generator


Impedance matching network


Glass concentric nebulizer


Torch


Spray chamber

0.3 m Spectrometer


900 Quartz prisms


Quartz lens

Neutral density filters


Photomultiplier tube


High voltage power supply


Boxcar average,
gated integrator

Computer interface


Computer


218











R 1547


226


SR 250


SR 245


PC-XT


Table 2-1
for Fluores


Model No.


YG 581-30


TDL 50


HFP 2500D


AMN 2500E


TR-30-C-3


T-1


cence Dip Spectroscopy


Manufacturer


Quantel International,
Santa Clara, CA

Quantel International,
Santa Clara, CA

Plasma Therm, Inc.,
Kresson, NJ

Plasma Therm, Inc.,
Kresson, NJ

J.E. Meinhard Assoc.,
Santa Ana, CA

Plasma Therm, Inc.,
Kresson, NJ

Laboratory constructed

GCA/McPherson Instrument.,
Acton, MA

Esco Products, Inc.,
Oak Ridge, NJ

Oriel Corp., Stratford, MA

Corion Corp., Holliston,
MA

Hamamatsu Corp.,
Bridgewater, NJ

Pacific Precision Instrum.,
Concord, CA

Stanford Research Systems,
Palo Alto, CA

Stanford Research Systems,
Palo Alto, CA

IBM Corp., Boca Raton, FL









Typical output irradiance from the frequency-doubled Nd:YAG laser was

250 mJ per pulse with a pulse duration of 12 ns. The first dye laser

was operated with a dye mixture of Rhodamine 590 and Rhodamine 610

(i.e., solvent is methanol) in a 4:1 ratio, respectively. The second

dye laser was operated with a pure Rhodamine 590 solution. The first

dye laser was tuned to either of the sodium (Na) atomic resonance

transitions (e.g., 588.995 nm or 589.592 nm) as shown in Figure

2-7. The second dye laser was tuned to either the 568.266 nm or

568.820 nm transition to probe the Na atom population in the first

excited state. The second dye laser could also be scanned across

these transitions to observe the dip shapes. Typical dye laser

irradiances were 6 mJ per pulse for the first dye laser and 0.03 to

7 mJ per pulse for the second dye laser. For the second dye laser,

the output irradiance was varied discretely with neutral density

filters and/or selective amplification. Selective amplification

meant that the dye laser could be amplified through the use of either

a preamplifier or amplifier cell or both simultaneously. A delay

line composed of prisms was arranged in such a manner that both laser

beams arrived at the center of the atomization cell at the same time

and in the same space, counterpropagating.

The inductively coupled argon (Ar) plasmas (ICPs) have been well

documented as an excellent atomization cell for atomic fluorescence

spectroscopy (AFS) (51-56). They possess several properties that

make them attractive for AFS. First, the high temperature and

chemically inert environment in Ar supported ICPs assure a high




















































589. 6 nm


U4 -1
/ /41450 cm






-1
34549 cm

















-1
16973 cm_
16956 cm






589. 6 nm


0 cm


Figure 2-7 Partial Energy Level Diagram for Sodium










degree of conversion of the analyte into free atoms (51,56).

Scattering of primary radiation by unvaporized analyte should,

therefore, be negligible. Second, on a comparison basis,

fluorescence quenching should be less in the predominantly Ar

environment than those in many other vaporization-atomization cells,

such as high temperature combustion flames (51,56,57).

The experimental operating conditions are listed in Table 2-2.

The ICP was operated at low powers (i.e., 500 W) to minimize as much

as possible the emission and/or ionization of Na. Sample solution

was introduced through a concentric nebulizer into a spray chamber.

The aerosol was then carried by argon into the center of the ICP.

The observation height of the fluorescence volume was 15 mm above the

load coil. This height is the most commonly used in atomic emission

measurements in the ICP, and it does not appear to degrade the atomic

fluorescence measurements as demonstrated in previous investigations

(58,59).

Laser excited atomic fluorescence of Na was detected at

588.995 nm or at 589.592 nm depending on the situation (i.e.,

resonance or nonresonance fluorescence transitions) at a right angle

to the laser beams using a 50.8 mm diameter, 111.6 mm focal length

quartz lens to image (1:1) the plasma onto the entrance slit of a

f/5.3, 0.3 m spectrometer. An iris diaphragm (38 mm diameter

opening) was placed directly after the lens to prevent overfilling

the grating. The entrance and exit slits were set at 50 um in order

to fully resolve the Na doublet. A neutral density filter of 1.0 was












Table 2-2
Experimental Parameters for Fluorescence Dip Spectroscopy


Forward RF power

Reflected power

Coolant Ar flow rate

Auxiliary Ar flow rate

Nebulizer Ar flow rate

Solution uptake rate

Observation height above load coil

PMT high voltage

Slit width

Slit height

Neutral density filter

External trigger rate

Gate delay

Gate width

Sensitivity

Input filter

Number of pulses averaged

Terminating resistor


500 W

0 W

15 L/min

0 L/min

0.7 L/min @ 32 psi

2.0 mL/min

15 mm

-900 VDC

50 im

10 mm

1.0

30 Hz

818 ns

2 ns, 4 ns, 600 ns

5 50 mV/V

>10 kHz

10 pulses

1.2 k2 (600 ns),
50 n (2 ns, 4 ns)









placed in front of the entrance slit. The output of the

photomultiplier tube was terminated either into a 50 0 or 1.2 kn

resistor depending on the gate widths employed. Signal averaging

from the output of the photomultiplier tube was performed with a

boxcar average, gated integrator. The averaged fluorescence signal

was inputed to an analog-to-digital converter which was interfaced

with a computer for data acquisition, analysis, and storage.

Measurement of the Fluorescence Dip

The fluorescence dip was measured in two ways. The first was by

scanning the second laser through the two upper transitions shown in

Figure 2-7. In this manner, the intensity of the dip as well as the

peak shape was observed. However, the overall measurement of the dip

was not made this way because of the relatively poor signal-to-noise

ratios. The alternative way to measure the dip was to tune the

second laser to one of the upper transitions. In this manner, it was

only necessary to block and unblock the second laser to obtain the

desired measurement as a function of laser irradiance.

There are four possible combinations of primary excitation and

fluorescence wavelengths, and for each combination of excitation and

fluorescence wavelength, there are two possible secondary excitation

wavelengths for which the dip will occur, i.e., there are eight

possibilities. For example, a possible combination is primary

excitation at 589.0 nm and fluorescence detection at 589.6 nm, and

secondary excitation at 568.3 nm (see Figure 2-7). In this study,

all possible combinations are investigated. In addition, there are

three different gate widths used for each possibility.










All the measurements pertaining to the dip were blank

corrected. A working solution of 120 mg/L of Na was prepared from a

2% stock solution which was diluted to the mark with deionized

distilled water (i.e., Barnstead water filtration system).



Results and Discussion

As shown in the energy level diagram (Figure 2-7), a resonant or

nonresonant fluorescence transition was monitored to observe the dip

in the fluorescence signal when the second dye laser was scanned

through the upper level transitions. This was shown clearly in

Figures 2-8 to 2-10. As can be seen, the dips became broadened as

the second laser irradiance was increased; that is, the peak shapes

(FWHM) were widened. This was primarily due to power broadening of

the second laser. Another outcome was that the peak of the dip did

not increase in a linear fashion with respect to the laser irradiance

of the second laser. This was explained by the condition of optical

saturation of the second transition. That is, increasing the laser

irradiance after a certain threshold had no appreciable effect on the

magnitude of the dip. The depletion of the first excited state was

between 50 and 80% depending on the irradiance of the second laser.

Gate widths of 2, 4, and 600 ns were used to note the effect.

The primary difference between the smaller (e.g., 4 ns) gate width

and the larger gate width (e.g., 600 ns), besides the time scale, was

that the pulse shape for the larger gate width was stretched through

the use of a 1.2 kn terminating resistor as compared to a 50 n












5.00


.-c

c-
S3.50


(c 3.00
4>-
0 568.3 nm
Q) 2.50 568.8 nm



267.5 568.0 568.5 569.0 569.5 570.0
Wavelength (nm)




Figure 2-8 Scan of the Second Excitation Laser Through the Upper Level Transitions of Sodium at 0.10 mJ
per pulse with the Fluorescence Wavelength at 589.6 Inm












5.00


4.50


>.,
*, 3.50
C3.00
C 3.00


o V f
.>2.50
0
Q) 568.3 nm 568.8 nm
1r 2.00


'S.5 568.0 568.5 569.0 569.5 570.0
Wavelength (nm)




Figure 2-9 Scan of the Second Excitation Laser Through the Upper Level Transitions of Sodium at 0.73 mJ
per pulse with the Fluorescence Wavelength at 589.6 nin












4.50


4.00


,~-3.50

>,
3.00
C

c 2.50


.> 2.00

-D 568.3 nm 568.8 nm
ry 1.50


'S67.5 568.0 568.5 569.0 569.5 570.0
Wavelength (nm)




Figure 2-10 Scan of the Second Excitation Laser Through the Upper Level Transitions of Sodium at 8.33 mJ
per pulse with the Fluorescence Wavelength at 589.6 nmn










termination for the smaller gate widths. The smaller resistance

termination allowed the integrity of the pulsed shape to be

maintained. The major result of the larger gate width was that the

intercepts of the reciprocal plots were greater than those employing

the smaller gate widths (Table 2-3). This was probably due to the

stretching of the signal pulse as mentioned earlier. The difference

between the 2 and 4 ns gate width was that the 4 ns gate provided

better signal-to-noise (S/N) and signal-to-background (S/B) ratios.

An explanation of this occurrence was that as gate widths became

narrower, the root mean square noise became greater. This was a

result of the reduced sampling efficiency and wider bandwidth. The

bandwidth was defined as approximately equal to 0.35 divided by the

selected gate width (60).

The effect of the observation height above the load coil was

investigated. As can be seen in Figure 2-11, the fluorescence signal

did indeed behave as a function of the observation height, but the

relative dip did not behave in an analogous manner. This was because

the relative dip required two measurements (i.e., fluorescence signal

with and without the second laser) so that the effect of observation

height was present in both measurements. However, the relative dip

was obtained by subtracting the fluorescence signal with the second

laser from the fluorescence signal without the second laser, and

hence, the effect of observation height was negated. An observation

height of 15 mm above the load coil was chosen because it provided

the greatest fluorescence signal.










Table 2-3
Comparison of Gate Widths with Respect to Slope,
Intercept, and r2 Correlation Coefficient


Wavelength
(nm) Gate Width Slopi
ex/fl/ex (ns) 10- J/m Hz Intercept r


589.0/589.0/568.3



589.6/589.0/568.3



589.0/589.6/568.3



589.6/589.6/568.3



589.0/589.0/568.8



589.6/589.0/568.8



589.0/589.6/568.8



589.6/589.6/568.8


600
4
2

600
4
2

600
4
2

600
4
2

600
4
2

600
4
2

600
4
2

600
4
2


0.00345
0.00264
0.00342

0.00391
0.00205
0.00297

0.00325
0.00357
0.00390

0.00458
0.00246
0.00336

0.00219
0.00150
0.00164

0.00158
0.00081
0.00155

0.00223
0.00148
0.00193

0.00198
0.00274
0.00179


2.499
1.574
1.483

2.540
1.403
1.732

2.475
1.748
1.973

2.502
1.230
1.582

2.244
1.465
1.384

2.603
1.429
1.599

2.382
1.608
1.680

2.630
1.163
1.725


0.990
0.986
0.996

0.965
0.996
0.987

0. 975
0.992
0.984

0.978
0.994
0.900

0.987
0.980
0.984

0.960
0.958
0.953

0.975
0.929
0.975

0.905
0.983
0.935














'


8.00




6.00
CL



* 4.00
0



2.00




0.00,


Height


Above


589.6/589.0
568.3 nm
600 ns gate
1 s1111111s 1111111


20 25 30
Load Coil (mm)


8.00

-r1

0
6.00 CD
0
CD
0
1
4.00


3
o)

2.00 <




.00


Figure 2-11 Plot of the Relative Dip and the Relative Fluorescence Intensity as a Function of
Observation Height Above the Load Coil


Il A u .llllnlill .ll. ll ll il. il.ai ...a.nii. i










The effect of monitoring resonance and nonresonance fluorescence

transitions was studied. This was shown in Figures 2-12 and 2-13.

The primary advantage of monitoring a nonresonance fluorescence

transition was that scattering (e.g., laser) was reduced. To realize

such a situation, the spectrometer had to have the capability of

resolving the Na doublet (i.e., 589.0 and 589.6 nm). The spectro-

meter was a 0.3 m focal length with a reciprocal linear dispersion of

2.6 nm/mm. Thus employing 50 pm slit widths (6X = 0.13 nm), the Na

doublet was completely resolved. As can be seen, the nonresonance

case (Figure 2-12) was not as noisy as the resonance case (Figure

2-13).

To extract information such as quantum efficiency and oscillator

strength, a slope method for the steady state population of the first

excited state is employed. The slope method employed the reciprocal

plots as described earlier. The plots are shown in Figures 2-14 to

2-19 with each figure containing the four possible combinations of

wavelength of first excitation laser and wavelength of fluorescence

in addition to the gate width and the wavelength of the second laser.

From theoretical considerations, the reciprocal plot was a

straight line as shown in Figure 2-20 with the slope that varied as

the quantum efficiency, Y32, varied. The reciprocal plots obtained

from these experiments showed that indeed there were differences

between the 568.3 nm and 568.8 nm transition probabilities. This was

clearly indicated by the slopes for these two upper level transitions

regardless what the lower level transitions (589.0 nm, 589.6 nm) were











6.00



5.00



-' 4.00
()
C
0)
-4-J
- 3.00



2.00



1.00


0.05
.0.05
-


0.03 mJ


0.09 mJ


0.15 mJ 0.2 mJ


- 30 s-


Figure 2-12 Measurement of the Relative Fluorescence Intensity at 588.995 nm With and Without the Second
Laser at 568.266 nm as a Function of the Second Laser Irradiance


Mfrl













3.40


3.20


3.00


>2.80
-4
I)
C 2.60
-4-

2.40
Q)

2.20 0.05 mJ 0.03 mJ 0.09 mJ 0.15 mJ 0.2 mJ


Cr 2.00 30 s


1.80






Figure 2-13 Measurement of the Relative Fluorescence Intensity at 589.592 nm With and Without the Second
Laser at 568.266 nm as a Function of the Second Laser Irradiance











10.00

3

8.00
1,4
: 2

6.00
Q)
4-j
_
Q) 4.00



2.00 5(1)589.0/589.0
S589.6/589.0
S589.0/589.6
S589.6/589.6

0 400 800 1200 1600 2000
1/Spectral Energy Density (x 10 8 J-1 m3 Hz)




Figure 2-14 Reciprocal Plot of Relative Dip vs Spectral Energy Density with 2 ns Gate Width and 568.3 nm
Transition











6.00


5.00


4.00
a
>*
3.00
-*

2.00

i 589.0/589.0
1.00 2 589.6/589.0
589.0/589.6
589.6/589.6

0 .0 0 . . o .. . . . . .. Iliii l ll i Ilii i l
00 400 800 1200 1600 2000
1/Spectral Energy Density (x 10 8, J-1 m3 Hz)



Figure 2-15 Reciprocal Plot of Relative Dip vs Spectral Energy Density with 2 ns Gate Width and 568.8 nm
Transition












10.00



8.00


1
6.00 4

> ~2

q) 4.00



2.00 I)l 589.0/589.0
2589.6/589.0
3 589.0/589.6
4 589.6/589.6

0.00i i I 1111i1 1i I i 1 111
0 400 800 1200 1600 2000
1/Spectral Energy Density (x 10 8, -1 m3 Hz)




Figure 2-16 Reciprocal Plot of Relative Dip vs Spectral Energy Density with 4 ns Gate Width and 568.3 nm
Transition











7.00

4
6.00


5.00
p5 3
)4.00
.>

3.00 2


2.00
S589.0/589.0
i 589.6/589.0
1.00 589.0/589.6
4589.6/589.6

0 0 0 . ..,,, i , , ,
0.. 400 800 1200 1600 2000
1/Spectral Energy Density (x 10 8, J-1 m3 Hz)



Figure 2-17 Reciprocal Plot of Relative Dip vs Spectral Energy Density with 4 ns Gate Width and 568.8 nm
Transition











8.00
S4
7.00
2
1
6CL600 3


5.00


4.00


3.00
(1 589.0/589.0
Si2) 589.6/589.0
2.00 : 3 589.0/589.6
; (4 589.6/589.6

........200 400 600 800 1000 1200
1/Spectral Energy Density (x 108 J -1 m3 Hz)




Figure 2-18 Reciprocal Plot of Relative Dip vs Spectral Energy Density with 600 ns Gate Width and
568.3 nm Transition











5.00
3,4
4.50
2

4.00


0)3.50

3.00


2.50
1 589.0/589.0
S589.6/589.0
2.00 3 589.0/589.6
1 589.6/589.6

1.5o0 0 06I. 40o6l 606I 8006 1000 1200
1/Spectral Energy Density (x 10 8, J m3 Hz)



Figure 2-19 Reciprocal Plot of Relative Dip vs Spectral Energy Density with 600 ns Gate Width and
568.8 nm Transition














80 -

70 -


*- 60 -
0

W 50 -

- 40 -

0 -


S20


10
6


1 3 5

1/Spec t r a I


10 15 20

Energy Densi t y ( x 10-12, -1 m3 Hz)


Figure 2-20 Theoretical Reciprocal Plot of Relative Dip vs Spectral Energy Density









(Table 2-3). It was evident that the gate widths and transitions

monitored did not appreciably affect the slopes of these plots. The

results indicated that the ratio between the average of the slopes

for each gate width of the two upper level transitions was

approximately two which was in good agreement with published results

of the ratio of the transition probabilities (Table 2-4) (61,62).

The calculation of an absolute quantum efficiency value using

equation 10 (Figure 2-5) and the measured slopes (Table 2-3) for

these two upper level transitions showed that the results were

between 0.001 and 0.004. However, the ratios of the quantum

efficiencies for the transitions showed that they were in good

agreement with the literature value (Table 2-4) (62). A possible

explanation for the such low values was that there were pronounced

ionization losses in this particular system; i.e., Na is an easily

ionizable element. The intercepts provide information about the

statistical weights of the levels. The statistical weight, g, is

equal to the number of superimposed levels, that is, the degeneracy

of a state. In the case where two states are mixing and act as a

single state, the statistical weight, g, is equal to the sum of the g

values for each state. The intercepts from the reciprocal plots were

in good agreement with the true value as shown in Table 2-4.

To extract the steady state saturation dip parameter, saturation

dip curves were plotted (Figures 2-21 to 2-28). The experimental

saturation dip curves showed a plateau region. The plateau region

was obtained by averaging the last four points that showed a tendency

to be constant. However, a linear portion was not as well-defined












Table 2-4
Comparison of Relative (Ratio) of Atomic Parameters


Atomic Parameter This Work Literaturea


Transition probabilities 1.90b 2.05
(568.8 nm/568.3 nm)
1.64c

1.97d


Quantum efficiency 1.90b 2.05
(568.8 nm/568.3 nm)
1.89c

1.96d


Statistical weights 2.48b 1.80
(gl +g2)/g3
1.45d

1. 65d


a Reference 62.
b 600 ns gate width.
C 4 ns gate width.
d 2 ns gate width.

















//


0.80





0.60

CL


0.40




0.20
0.20


//
0/'

p/


.I


10 -
Spectral Energy


10-'
Density


(J cm3 Hz)


Figure 2-21 Plot of Relative Dip vs Spectral Energy Density with 589.0 and 568.3 nm Excitation and
Fluorescence Detection at 589.0 nm


f//s
oA


2
4
600


0.09'


ns gate
ns gate
ns gate


-12


. ..I II


I a I aaI a


10-


m I


-/- '" '
i I
















/ A


--~-i
/4


/ / -
/ / <
ols/ /


/, 1
/


/.4

/


0.60






00.40
Q.)



ci)
0.20






0.00


10-"
Spectral Energy


1) 2 ns
2 4 ns
(3) 600 ns


10 -1
Density


(JAcm3 Hz)


Figure 2-22 Plot of Relative Dip vs Spectral Energy Density with 589.0 and 568.3 nm Excitation and
Fluorescence Detection at 589.6 nm


. I


gate
gate
gate


.. . .. .. . m... .1


10-"


a a a s e n a


* l *I* II


-- -- (3)



- -*, (3)












0.80





0.60

C_


S0.40

0


0.20





0.0o0-


-/- -- )
4^ **
//


//
/
*^


--- --*-,-. (3)


It-


/r


/1


10
Spectral


2 na
4 ns
600 ns


S
i:3)


-11
Energy


10 -"
Density


gate
gate
gate


(J cm3 Hz)


Figure 2-23 Plot of Relative Dip vs Spectral Energy Density with 589.0 and 568.8 nm Excitation and
Fluorescence Detection at 589.0 nm


a I I 1l I I


10-a


' '"""












0.70



0.60



0.50



) 0.40



) 0.30



0.20



0.10o
0


/
z:-'-z---- g3


/
/ f

/ +/
/ /

/+ p


S/
/ *




i a i


. I I I


10 -"
Spectral Energy


-- -- --4-u -- - -1


1) 2 ns gate
2 4 ns gate
3) 600 na gate




. I II I I I


10-1"
Density


(J/Acm3 Hz)


Figure 2-24 Plot of Relative Dip vs Spectral Energy Density with 589.0 and 568.8 nm Excitation and
Fluorescence Detection at 589.6 nm


10-


-- --


/














0.80






0.60


O5





0
) 0.40 -



4)
or_
0.20






0.0%0


---/-.---- (2)
/

/
/ --;/-- ----. (1)
/


r
/
/
/
/ -
F


- / -------*- (3)


I /-


Vr


. I


2 ns gate
4 ns gate
600 ns gate


S. II


10-" 10-1 10 *
Spectral Energy Density (J/ m3Hz)


Figure 2-25 Plot of Relative Dip vs Spectral Energy Density with 589.6 and 568.3 nm Excitation and
Fluorescence Detection at 589.0 nm


10-8


. * * i n * = = =


--


r


.
















/


/---,-


- -S*


--- (2)


--- (1)


S(3)
2 ns gate
4 ns gate
600 ns gate


...II . . I|


10-1
Spectral Energy


10 -
Density


I a I a aa I I


(JYcm3 Hz)


Figure 2-26 Plot of Relative Dip vs Spectral Energy Density with 589.6 and 568.3 nm Excitation and
Fluorescence Detection at 589.6 nm


1.00




0.80



L0.
- 0.60


0.40


0.20


/


A
/


4 -
-S


-ii


s- S


-Il


0.0'o


a i a m a m ...


I I ...l l


10 -













0.80





0.60

0C


) 0.40
e-j



0.20






0 oii


A7


- -,-4 (2)


C/
/ /


1, /
/ +/ /
/ *^*


I I uill


2
4
600


I I I a aal1 i


10Sl
Spectral Energy


I I I I l ill


10 -'1
Density


ns gate
ns gate
ns gate


I I I 1 1 I


(J cm3 Hz)


10-


Figure 2-27 Plot of Relative Dip vs Spectral Energy Density with 589.6 and 568.8 nm Excitation and
Fluorescence Detection at 589.0 nm


*
~













1.20



1.00



Q0.80



0.60
4-1

00.40



0.20



0.0o


~~~1
/4
/
/
4


-- - (2)

4
-' ( 1 )
4,


/ -* -


.o (1) 2
2- 4
3 600


I a I I aa il


10 -"
Spectral Energy


ne gate
ns gate
ns gate


i Ia I I 111 I I A I I I


10 -"
Density


(JAcm3 Hz)


Figure 2-28 Plot of Relative Dip vs Spectral Energy Density with 589.6 and 568.8 nm Excitation and
Fluorescence Detection at 589.6 nm


10-"


___


I I I 1 I I I









because the sensitivity of the measurement did not allow relative

dips below 0.1 to be measured. As can be seen from Figure 2-29, the

theoretical curve has a wide range of relative dip values. The

plotted curves showed only a small range of relative dips (0.1 to

0.8), but the linear portion is assumed so that an intersection point

can be obtained. The intersection of these two asymptotes provided a

graphical method for the calculation of the steady state saturation

dip parameter. Since the parameter had been defined on page 17, it

was then possible to determine the spectral energy density at one

half the maximum relative dip by graphical means. The figures shown

were for a primary excitation wavelength (589.0 or 589.6 nm), a

fluorescence wavelength (589.0 or 589.6 nm), and a secondary

excitation wavelength (568.3 or 568.8 nm) for three gate widths.

From these unextrapolated curves, it is evident that the saturation

dip parameter was different depending upon which upper level

transition was employed. The values for the steady state saturation

dip parameter are tabulated in Table 2-5. Calculations were per-

formed using equation 12 (Figure 2-5) to obtain the quantum efficien-

cies. Again the absolute values were low; however, the ratio of the

quantum efficiencies for the upper level transitions produced similar

values as that in Table 2-4. This served as a check for the recipro-

cal plots, and it showed consistency between the two types of plots.

From the published literature (61,62), it was clearly evident

that agreement was certainly good on a relative basis and fair on an

absolute basis (Table 2-4). On a relative basis, the ratios of the

transition probabilities and quantum efficiencies agreed well with
















-l10 -


.)
> 10

f 10


Cr 10'3


Spectral


Energy


Densi ty,


J/m3Hz


Theoretical Plot of Relative Dip vs Spectral Energy Density


10 -13 10 -12 10 -11 10 -10


Figure 2-29











Table 2-5
Comparison of Steady State Saturation Dip Parameter with
Respect to the Second Excitation Laser Wavelength


Steady State Saturation Dip Parameter
10-11 J/m3Hz
Wavelength, nm
ex/fl/ex 2 ns 4 ns 600 ns


589.0/589.0/568.3 2.20 1.77 0.93

589.0/589.6/568.3 1.78 1.46 1.75

589.6/589.0/568.3 1.78 1.90 1.69

589.6/589.6/568.3 2.76 2.16 1.46

589.0/589.0/568.8 1.05 1.78 0.63

589.0/589.6/568.8 1.47 0.88 0.77

589.6/589.0/568.8 1.18 1.44 0.80

589.6/589.6/568.8 0.89 0.63 0.63









the literature values. On an absolute basis, the g values for the

levels that were obtained through the intercepts of the reciprocal

plots were in good agreement. The absolute quantum efficiencies were

apparently low to what was expected but quantum efficiencies have not

been measured so it was uncertain whether the values were correct.

However, the results were important because it provided the

opportunity to investigate upper level transitions without knowing

much information about those levels. Indeed, if the transition

probability of one of the transitions was known, then other transi-

tions can be determined in a qualitative manner in terms of its

relative atomic parameters and its merits for atomic spectroscopy.



Conclusions

Although this was a first attempt to apply fluorescence dip

spectroscopy for the determination of atomic parameters, it was

moderately successful as it was first envisioned, and certain

conclusions were drawn. First, sodium was a good example, but it had

its drawbacks. Sodium had been thoroughly investigated because of

the wealth of information with which to compare the results of this

technique. However, sodium also was a very easily ionizable element

even at the low powers used in this investigation. There was

approximately 80% of the initial atomic population lost through

ionization without any benefit of laser excitation. This value was

obtained by the Saha equation (see Appendix C) using rough estimates

of the temperature and the electron number density in the ICP. A

more accurate value can be calculated using the Saha equation and the









electronic partition functions (63) if more accurate measurements

were made for the temperature and the electron number density. Other

investigators have calculated that 98 to 99.9% of the sodium atoms

are lost through ionization assuming a temperature of 7500 K and

local thermodynamic equilibrium (64,65). In addition to direct

ionization, there were losses due to laser excitation followed by

collisions into the ionization continuum, a trap. Second,

qualitative and some quantitative results were obtained, and

comparisons were made that proved the technique was viable for the

study of upper level transitions as well as a diagnostic tool for

plasmas and flames with the possibility of graphite tube atomizers.

Third, the first excited doublet states were strongly coupled with

one another by collisions. This meant that the doublet acted as a

"single" state. This was shown to be true because a dip was observed

for the resonant as well as the nonresonant fluorescence wavelength

situations. Fourth, the applicability to other elements for the

investigation of their atomic parameters was investigated.

Other elements such as palladium (Pd) and calcium (Ca) were also

investigated. The choice of palladium was made because it had an

abundant number of energy levels near the first excited state (Figure

2-30). The main purpose, then, was to monitor different fluorescence

transitions to note the observance of a dip, if any. There was the

primary dip observed for the connected states (Figure 2-31), but no

other dips were observed for several fluorescence transitions. This

indicated that the levels were not strongly coupled with one another.

























































292. 2 nm


// /A7236 cm


COLL ISI ONAL


I ON I ZAT I ON








E E
C C



L ,-




in I


.. -


-1
58562 cm 1
58448 cm-1


40771
39858
38812

36181

34069


1722
0094

7755
6564


-1
cm
cm
-1
cm
m-1
cm
-1
cm


c-1
cm
-1
cm
-1
c rn-1
cm


Partial Energy Level Diagram for Palladium


67236 c


Figure 2-30












1.00



0.80



4-' 0.60
(I)
cF
C


0.40


S14 mJ 8.3 mJ 0.73 mJ 0.10 mJ
0.20

30 a


0.00






Figure 2-31 Measurement of Relative Fluorescence Intensity at 34I.1 nm With and Without Laser Excitation
at 565.5 nm as a Function of Laser Irradiance









The investigation of calcium was entirely different. To circumvent

the losses due to ionization in the ICP, calcium ions (Ca+) were

chosen as the target because the singly charged ions were prevalent

and the possibility of creating the doubly charged ions was less than

1% (65). The second ionization potential was approximately 15 eV

above the first ionization potential. In fact, the second excited

state of the ionic transition was approximately 9 eV below the second

ionization potential. With all these possibilities, it would appear

that this would be an ideal situation for a dip. However, this was

not the case because the necessary wavelengths (Figure 2-32) obtained

from frequency conversion of the dye lasers were not able to saturate

the transitions as well as not "beat" the deactivation processes. In

addition, preliminary investigations of a fluorescence dip for lead

(Pb) were done in a graphite tube atomizer. These experiments, along

with the sodium dip experiments, demonstrated that fluorescence dip

spectroscopy is indeed a viable diagnostic tool for qualitative and

semi-quantitative information.

















tt


373. 7 nm


-~ -


396. 9 nm


393. 7 nm


52167 cm


370. 6 nm


-1
25414 cm
-1
- 25192 cm




396. 9 nm


0 cm


Figure 2-32 Partial Energy Level Diagram for Calcium (II) Ion


V I


rL















CHAPTER 3
ATOMIC FLUORESCENCE AND IONIZATION MECHANISM
FOR LEAD IN AIR-ACETYLENE FLAME



Basic Principles of Atomic Fluorescence and
Ionization Spectroscopies

Atomic fluorescence spectroscopy (AFS) is based on the

absorption of radiation by an atomic specie, thereby producing atoms

in an excited electronic state, and the measurement of the light

emitted when a fraction of the excited atoms undergoes radiational

deactivation (66) from the excited electronic state. Of the several

pathways for deactivation, the three most common are resonance

fluorescence, in which the same lower and upper levels are involved

in the excitation and deactivation processes (Figure 3-la); direct

line fluorescence, in which the same upper level is involved in the

excitation and deactivation processes (Figure 3-1b); and stepwise

line fluorescence, in which different upper levels are involved in

the excitation processes (Figure 3-1c) (67). In stepwise line

fluorescence, there are three steps. The first step is the absorp-

tion of photons to excite the atom to an excited level. This is

subsequently followed by collisional processes where demotion or

promotion of an electron to a nearby excited level occurs. The final

step is the deactivation process from the new excited level to the

lower level.











7


\(a)

(a)


\I /


(b)


(C)


Figure 3-1


The Three Basic Pathways of Atomic Fluorescence: a) Resonance Fluorescence, b) Direct Line
Fluorescence, and c) Stepwise Line Fluorescence


7FJ7;


I
I
I


=,:





63


For double resonance excitation, resonance fluorescence is not

possible because the transition from the second excited state to the

ground state is a forbidden one-photon transition (i.e., the

symmetries are incorrect). Thus, only direct and stepwise line

(e.g., anti-Stokes) fluorescence schemes are considered.

Laser enhanced ionization is based on selectively populating a

specific excited state of an analyte such that there is a decrease in

energy required for ionization from the excited state relative to the

ground state (41). Since the rate of collisional ionization

increases as the energy required for ionization decreases, it is

desirable to populate higher energy excited states (47). To this

end, double resonance excitation schemes are ideal for this situation

because they populate levels that are much closer to the ionization

continuum than one-step processes (Figure 1-2b). Signal enhancements

of 100-fold or greater are often achieved relative to that observed

when the one-step processes are used (47).

It is apparent that the combination of two-color laser enhanced

ionization and two-color laser excited atomic fluorescence is a

powerful technique with which to study ionization mechanisms in the

air-acetylene flame, traditionally the atomization cell of choice.

The two-color laser enhanced ionization was more useful for

diagnostic studies than the two-color laser excited atomic

fluorescence because the primary investigation was the overall

temporal features of the waveform for the electron pulse. However,

the atomic fluorescence technique proved useful in the determination

of the analytical wavelengths for the second transition as well as

for later experiments involving the graphite tube atomizer.












Experimental Facilities and Considerations

A block diagram of the experimental setup is shown in Figure 3-2

and a detailed listing of the experimental components is described in

Table 3-1. The laser system has been previously described in this

dissertation. There are two modifications of that laser system. The

first difference was that the second dye laser was operated with a

dye mixture of Rhodamine 610 and Rhodamine 640 in a 4:1 ratio,

respectively. The second difference was that the first dye laser was

frequency doubled to ultraviolet radiation through use of a potassium

dihydrogen phosphate (KDP) crystal. The first dye laser was tuned to

566.612 nm and then frequency doubled to 283.306 nm, the atomic

resonance transition of lead as shown in Figure 3-3. The second dye

laser was either scanned through or tuned to three possible

wavelengths (i.e., 600.193, 601.172, and 605.942 nm) (Figure 3-3).

Typical laser irradiances were 1 mJ per pulse for the ultraviolet

laser and 0.1 and 10 mJ per pulse for the second dye laser. For the

second dye laser, the output irradiance was changed discretely by

placing a neutral density filter at 2.0 in front of the laser beam.

The air-acetylene flame has been the stalwart for analytical

flame atomic spectroscopy over the last twenty-five years. The air-

acetylene flame has been studied in detail so that many of the

mechanisms, and thus components in the flame, are known and

understood (68). The primary limitations are the low temperature of

the flame (e.g., 25000K) and quenching of the fluorescence. To






































so/so
BEAMSPLITTER


Figure 3-2


Block Diagram of Experimental Setup for Laser Enhanced Ionization and Laser Excited Atomic
Fluorescence Spectroscopies









Table 3-1
Experimental Components for Laser Enhanced Ionization
and Laser Excited Atomic Fluorescence Spectroscopies


Component Model No. Manufacturer


Nd:YAG laser


Dual dye lasers


Frequency doubler


Slot burner, spray
chamber, and nebulizer

Stainless steel electrode

900 Quartz prisms


Quartz lens

0.35 m Spectrometer

Photomultiplier tube


High voltage
power supplies

Boxcar average,
gated integrator

Computer interface


Chart recorder

Digital oscilloscope


Plotter


YG 581-30 Quantel International,
Santa Clara, CA

TDL 50 Quantel International,
Santa Clara, CA

HP 50 Quantel International,
Santa Clara, CA

Perkin Elmer Corp.,
Norwalk, CT

Laboratory constructed

Esco Products, Inc.,
Oak Ridge, NJ

Oriel Corp., Stratford, CT

EU-700 Heath,a Acton, MA

R 1547 Hamamatsu Corp.,
Bridgewater, NJ

226 Pacific Precision Instrum.,
Concord, CA

SR 250 Standard Research Systems,
Palo Alto, CA

SR 245 Standard Research Systems,
Palo Alto, CA

D-5000 Houston Instrum., Austin, TX

2430A Tektronix, Inc., Beaverton,
OR

HP 7470A Hewlett Packard, Palo Alto,
CA


a Now GCA/McPherson.


















E, cm


59821


35287













10650


-------- ------------^WW^



0. 976 eV






E E E

C C rC

0) 1 0


o 0 0
(O (0 (0



E
-

co
0


(0


30 1 8

P, 8p









P -7s


p 2
P2-6p2



po-6p2


Figure 3-3


Partial Energy Level Diagram for Connected Double
Resonance Excitation and Ionization of Lead









reduce some of the quenching, an argon sheath has been used to

enclose the flame to prevent quenching from the surrounding air.

In these experiments, a 5 cm slot burner was used that did not

have the necessary design for a protective sheath, but this was not a

factor since quantitative measurements were not being made (i.e.,

limits of detection). The flame was mounted on a commercial atomic

absorption spray chamber. An approximate stoichiometric air-

acetylene flame was used for all studies. A water-cooled stainless

steel tube that was flattened to increase the surface area was used

as an electrode (i.e., cathode) with a negative bias voltage,

-1300 VDC. This electrode was immersed into the center of the Flame

such that the electrode was 21 mm above the burner head and that the

laser beam passed parallel to and 3 mm below the electrode, unless

otherwise noted. The burner head was grounded (i.e., anode) through

a 10 kn resistance so that it acted as a collector for electrons.

The ionization currents were capacitively coupled through 500 pF

to reduce flame background leakage currents and then were converted

to voltages using the 50 n termination mode in the digital oscillo-

scope. The waveform was acquired, displayed, and stored through the

oscilloscope.

Laser excited atomic fluorescence of lead was monitored at

several wavelengths (e.g., 216.999, 239.379, 261.418, or 405.783 nm)

at a right angle to the laser beams using a 50.8 mm diameter,

111.6 mm focal length quartz lens to image (1:1) the flame volume

onto the entrance slit of a f/6.8, 0.35 m spectrometer. This was

done to insure that the lasers were properly tuned to the correct









transitions. An iris diaphragm (31 mm diameter opening) was placed

directly after the lens to prevent overfilling the grating. The

entrance and exit slits were set at 1 mm to collect as much

fluorescence emission as possible. The output of the photomultiplier

tube was terminated into a 50 Q resistor. Data acquisition,

analysis, and storage were the same as that described earlier.

The experimental parameters are listed in Table 3-2. All the

measurements pertaining to the fluorescence and ionization signals

were done with a solution of 100 mg/L of Pb unless otherwise noted.

This solution was prepared from a 1 g/L stock solution which was

diluted to the mark with deionized distilled water.



Results and Discussion

Atomic Fluorescence Study

Although the fluorescence study was not done to obtain

analytical figures of merit, such as linear dynamic range and limits

of detection, atomic fluorescence did serve a very important purpose

in the outcome of the ionization studies. The fluorescence

monitoring of the double resonance as well as the single resonance

excitation schemes was important to the tuning of the lasers so that

the transitions under investigation were indeed the correct ones

being excited. To insure proper tuning, several fluorescence

wavelengths were monitored as shown in Figure 3-4.

Laser enhanced ionization is a very powerful technique, and

therefore, it was imperative that the centers of the atomic










Table 3-2
Experimental Parameters for Laser Enhanced Ionization
and Laser Excited Atomic Fluorescence Spectroscopies


Acetylene flow rate

Auxiliary air flow rate

Nebulizer air flow rate

Solution uptake rate

Scan rate of second excitation laser

Electrode potential

?MT high voltage

Slit width

Slit height

External trigger rate

Gate delay

Sensitivity

Input filter

Number of pulses averaged

Terminating resistor


1.2 L/min

7.5 L/min

2.8 L/min

4.5 mL/min

0.02 nm/s

-1300 VDC

-900 VDC

1 mm

1 cm

30 Hz

51 ns

10 mV/V, 50 mV/V

>10 kHz

3 pulses

50 n


















59281 cm-1
59281 cm


52412
51944
51917
51786
48189
46069
46061


-1
cm
c -1
cm-
c m-1
cm-1

cm-1
cm
-1
cm
C1


35287 cm-1


S1 ,10650 cm-1


7819 cm-1




0 c m-1
0 cm~


Figure 3-4 Partial Energy Level Diagram for Connected Double
Resonance Excitation of Lead


z / / / / / //












resonance line was being excited and not the wings of the atomic

lines. Excitation of the wings produced ionization signals. To

achieve proper tuning, it was best to monitor fluorescence

transitions so that fluorescence signals could be observed to

unequivocally prove that the laser excitation schemes were indeed

correct. This was done for the 283.306 nm excitation line by

monitoring the 405.783 nm. For the double resonance excitation

scheme, several fluorescence wavelengths were individually

monitored. As the fluorescence wavelength was fixed, the second

laser was scanned through the three upper level transitions as shown

in Figures 3-5 and 3-6. It can be seen that the most intense

transition was at 600.193 nm for both of these examples. This was

further verified by other fluorescence wavelengths which ranged from

216.999 to 373.995 nm.

Ionization Study

Ionization currents were produced through the double resonance

excitation scheme followed by collisional ionization or photoioniza-

tton as illustrated in Figure 3-3. Lead was first excited with the

ultraviolet radiation at 283.306 nm. It was then followed by the

visible radiation at 600.193 nm, in most cases.

Amplification was desirable for our experiments, but could not

be done because the available current-to-voltage amplifiers were too

slow in their response. The waveform collected was distorted from

the true waveform because of the speed with which these amplifiers































I ii I al a I I I I I I a a a i lI a iii a


602 603 604
Wavelength (nm)


605


Figure 3-5


Scan of the Second Excitation Laser With the Fluorescence Wavelength at 239.379 nm and the
First Excitation Laser at 283.306 nm for Lead


-0.00


-0.20


-0.40


CO
-0.60
-1-
C:

0V -0.80 '


0)-1.00


-1.2^


601


-


I


Mr


^^V^ ^^^,^^ I -.


n7-













0.00


-0.20



-+ -0.40
U)
C
4-,-
- -0.60
.>


- -0.80

1.00

-1.0o,


Wavelength (nm)


Figure 3-6


Scan of the Second Excitation Laser With the Fluorescence Wavelength at 261.418 nm and the
First Excitation Laser at 283.306 nm for Lead









operate (e.g., 20 MHz). This is compared with that of the digital

oscilloscope, 150 MHz. A voltage amplifier (X 10) was also tested,

but not used because it added too much noise to our system. There-

fore, it seemed that the best arrangement was also the simplest one

[i.e., signal was directly connected into a digital oscilloscope

(50 0 termination mode) without amplification].

Since the best arrangement was setup to collect and store the

waveforms for the two-color laser enhanced ionization of lead, we

wished to determine whether it was possible to distinguish temporally

the collisional ionization from the photoionization features of the

electron pulse. The observation of two features in the electron

pulse in Figure 3-7 was striking. It was clearly visible that a

fast, sharp, and intense component (FWHM -15 ns) and a slow,

broadened component (FWHM -200 ns) were there. The fast component

was believed to be the photoionization feature whereby ground state

lead atoms have absorbed three photons (i.e., 283.306, 600.193, and

600.193 nm) to reach the ionization continuum. The slow component

was felt to be the collisional ionization feature whereby ground

state lead atoms have absorbed two photons (283.306 and 600.193 nm)

and then, through collisions, have been elevated into the ionization

continuum.

Further investigation of the phenomenon was done. The 600.193

nm laser beam was attenuated with a neutral density filter of 2.0;

that is, the output irradiance was decreased by 100-fold and was

approximately 0.1 mJ. This lower energy was insufficient to cause













-J VVI -




z
Z
(I


Z
O






~ 100 0
T




TI ME


Figure 3-7 Temporal Behavior of the Ionization Signal as Recorded by the Oscilloscope With Full Laser
Irradiance in Both Beams (283.306 and 600.193 nm)









photolonization (Figure 3-8) but still able to produce a collisional

(i.e., ionization rate was lower then the radiative and quenching

rates out from the 7s level) ionization feature. When Figures 3-7

and 3-8 were compared, it was apparent that the photoionization

feature disappeared, but that the collisional feature was only

reduced by a factor of two to three. This indicated that the upper

8d levels were under optical saturation. When the second excitation

wavelength was changed to one of the other two transitions (e.g.,

601.172 or 605.942 nm), similar ionization signals were observed with

the only difference in the intensity. The intensity decreased as the

second excitation wavelength became more shifted toward the red as

shown in the fluorescence measurements (Figures 3-5 and 3-6).

Further experiments were performed as suggested by Turk of the

National Bureau of Standards (69) to verify the shape of the waveform

was real. Turk (69) felt the shape of the waveform was due to the

detection process rather than the production of ions. The additional

experiments included the effects of varying the bias voltage, the

position of the laser beam relative to the burner head and the

electrode, the variation of lead concentration, and the composition

of the flame.

The effect of negative bias voltage was studied. Voltages were

varied from -900 to -2000 VDC in increments of 100 V to note the

effect, but no effect was observed with the exception of intensity.

Since there was no effect, it was concluded that the waveform did not

vary with the voltage in terms of its shape and behavior. This is in

contrast to those results reported by Berthoud et al. (70).





































100
I I I


I I I I I


TIME


Figure 3-8 Temporal Behavior of the Ionization Signal as Recorded by the Oscilloscope with the Laser
Irradiance at 600.193 nm Decreased by a 100-Fold and the 283.306 nm Laser Irradiance Unchanged


100 mV










The position of the laser beam relative to the burner head and

the electrode was investigated. The shape and behavior of the

ionization waveform was identical with respect to the previous

experiments at a laser position centered between the electrode and

the burner (10 mm above the surface) except for a small decrease in

magnitude. Measurements taken with the laser beams very close to the

burner head (2 mm above the surface) showed a complete absence of the

slower pulse component possibly due to a very low degree of

collisional ionization within the flame reaction zone.

The effect of lead concentration and flame composition was

considered. No dependence of the ionization waveform shape on lead

concentration was noted. The concentration was varied from 10 to

500 mg/L. The normal air-acetylene flame that is used in everyday

circumstances is composed of compressed air and acetylene. Since

oxygen is the prime ingredient in compressed air for this type of

flame, a gas cylinder containing a mixture of oxygen and argon in the

ratio 1:4 was substituted for the compressed air. This flame has

provided the identical shape and behavior of the ionization waveform.



Conclusions

The ionization signal contained two distinct features in its

waveform. It was concluded that the photoionization and the

collisional ionization features were separated temporally as shown in

Figure 3-7. Several experimental parameters were varied to note if

the pulse shape and behavior changed, but none were observed. As

mentioned earlier, Berthoud et al. (70) noticed extreme variations in










their pulse shape as the voltages were changed. This effect was not

observed for these experiments.

However, these experiments were similar to those pointed out by

Broglia et al. (71) in their investigation of the optogalvanic signal

of uranium in a hollow cathode discharge lamp. By using a fast

detection system, they showed that the temporal shape of the

optogalvanic signal consisted of two components: a slow one, in a

time scale typical of the relaxation time of the discharge, and a

fast and more intense one, which occurred within the time scale of

their laser pulse (-5 ns). The fast component was observed only when

the difference in energy between the first excited state and the

ionization limit was less than the energy required to excite atoms

from the ground state to the first excited state, clearly pointing to

a direct photoionization process caused by nonresonant absorption of

a laser photon into the ionization continuum.

Much of the work in laser enhanced ionization spectroscopy was

and is done with electronic detection systems that were not as fast

(e.g., more than 50 ns in time scale) as the ones utilized here. The

combination of fast detection electronics and of the simultaneous

observation of the fluorescence and ionization signals seemed to be

the best method for elucidating the excitation-ionization processes

when atmospheric pressure flames are used as atom reservoirs.















CHAPTER 4
MEASUREMENT OF ATOMIC FLUORESCENCE FOR
LEAD IN A GRAPHITE TUBE ATOMIZER


Introduction to Graphite Tube Atomizers

Graphite tube atomizers are a part of a larger area known as

electrothermal atomizers or graphite furnaces. Electrothermal

atomizers have gained popularity through the years as an effective

and efficient method for the production of atoms for atomic

absorption spectroscopy (72-74). Besides graphite tube atomizers,

there are rods, cups, and filaments all of which are made of graphite

(Figure 4-1). Only in the last ten years have electrothermal

atomizers found a place in atomic fluorescence spectroscopy.

Graphite furnaces are electrically heated devices that are

heated through water-cooled contacts (e.g., copper, brass, or

stainless steel). The heating is usually performed in three stages

(drying, charring, and atomizing the sample). The drying stage,

-1100C, is used to evaporate the solvent as quickly as possible

without spattering the material inside the furnace (74). Charring,

350 to 12000C, is performed to remove volatile sample components at a

temperature as high as possible without any effect or loss of the

analyte (73). Atomization is accomplished by heating the furnace to

1600 to 30000C, with a very rapid heating rate (8000C/s or greater)

so that analyte is quickly vaporized and atomized (74).












2. 5 mm id


CUP


3. 5 mm


) ROD








TUBE


b0


Graphite Furnace Designs


-5 FD 0


Figure 4-1










Over the years, graphite furnaces have been modified to improve

the analytical performance. Among these modifications have been the

L'vov platform (75) and treatment of the inner surface of the

graphite furnace with coatings such as pyrolytic graphite (76),

tantalum coating (77) or lining (78), and palladium (79).

The L'vov platform is a small graphite platform that is placed

inside a graphite tube. The principle feature of the platform is

that its temperature lags behind that of the walls because it is

radiatively heated. This delay or lag in temperature causes the

vaporation of the analyte to be delayed until the atmosphere reaches

a high and quasi constant temperature. The platform results in a

reduction of matrix interference due to this constant atmosphere

temperature (74). Modifiers, such as the ones mentioned, generally

reduce or remove volatilization and vapor-phase interference by

allowing high enough pyrolysis temperatures to remove the bulk of

concomitants during the charring stage (80).



Brief Review of Atomic Fluorescence Using
Graphite Furnace Atomization

The vast majority of published literature employing graphite

furnaces has been in atomic absorption spectroscopy. However,

literature has appeared that has utilized graphite furnaces as

atomization cells for laser excited atomic fluorescence spectroscopy

(81-96). The graphite furnaces utilized have been either laboratory

constructed or commercially available models. The mating of laser

excited atomic fluorescence with graphite furnaces has provided









extremely low detection limits, in the range of sub-part per trillion

to parts per trillion level or subfemtogram to picogram levels in

terms of absolute amounts, for several elements, with linear dynamic

ranges reaching six orders of magnitude (89,91,92).

Graphite furnaces possess several properties that make them

attractive vaporization-atomization cells. First, the small volume

consumption, typically 2 to 10 microliters per measurement, makes it

well-suited for limited sample volumes as compared to the volume

consumption of flame and plasma systems. Second, solid as well as

liquid samples can be vaporized and atomized with no special adapta-

tions to the furnace. Third, the nebulization stage is eliminated

from the analytical sequence (75). Since the nebulization stage is

eliminated, there is no need for sample introduction problems, and

the sample is completely vaporized. Another advantage is that there

is unlimited sample vaporization time (greater than 1 s) as compared

to the limited vaporization time (less than 100 ms) for aerosol

particles (75). Fourth, many of the matrix interference have been

greatly reduced due to the advent of matrix modifiers as first

suggested by Ediger in 1975 (97). Fifth, graphite tube atomization

(90,94,95) is preferred to graphite cup or carbon rod atomization

despite the convenient optical geometry for observation of the

fluorescence emission. Vapor phase interference have existed with

the carbon rod because of the strong temperature gradient between the

atom formation and excitation zones. With graphite tube atomization,

the atoms are contained in a hot environment while being excited by










the laser and, therefore, much better analytical performance results

in the case of real samples (95).

Atomic fluorescence in graphite furnaces was first demonstrated

for Pb by Neumann and Kriese (81) in 1974. Subsequently, other

workers broadened the application to include a variety of elements,

e.g., Bolshov et al. (85) (Ag, Co, Cu, Eu, Fe, Ir, Mn, and Pb) in

1981, Human et al. (88) (Pb and Tl) in 1984, Goforth and Winefordner

(91,92) (Al, Cu, In, Li, Mn, Mo, Pb, Pt, Sn, and V) in 1986 and 1987,

and Dittrich and Stark (95) (Al, Ga, In, Ir, and V) in 1987. Much of

the work in this area has been performed by the Bolshov group in the

Soviet Union as evidenced by their numerous publications (82,84,85,

89,98-102). It is also worth noting that only a couple of the

published papers have delved into double resonance excitation schemes

(86,96).



Experimental Facilities and Considerations

Instrumentation

A block diagram of the experimental setup is shown in Figure 4-2

and a detailed listing of the experimental components is described in

Table 4-1. The laser system has been previously described in this

dissertation and is identical to the one used in the atomic

fluorescence and ionization studies with two differences. The first

difference was that the laser output irradiance of the first

excitation laser, 283.306 nm, was changed discretely by placing a

neutral density filter of 1.0 or 2.0 in front of the laser beam. The


















































Figure 4-2 Block Diagram of the Double Resonance Laser Excited At:)min Fluorescence in a Graphite Tube
Atomizer oa










Table 4-1
Experimental Components for Double Resonance Laser Excited
Atomic Fluorescence in a Graphite Tube Atomizer


Component Model No. Manufacturer


Nd:YAG laser YG 581-30 Quantel International,
Santa Clara, CA

Dual dye lasers TDL 50 Quantel International,
Santa Clara, CA

Frequency doublers HP 50 Quantel International,
Santa Clara, CA

Graphite tube and Laboratory constructed
timing circuit

Furnace power supply SCR 20-250 Electronics Measurements,
Neptune, NJ

Micropipette Brinkmann Instrum. Co.,
Westbury, NY

900 Quartz prisms Esco Products, Inc.,
and quartz lenses Oak Ridge, NJ

0.22 m Double monochromator 1680B Spex Industries, Inc.,
with scan controller CD2A Methucen, NJ

Neutral density filters Corion Corp., Holliston,
MA

Photomultiplier tube R 955 Hamamatsu Corp.,
Bridgewater, NJ

High voltage power supply 301 Bertan Assoc. Inc.,
Hicksville, NY

Amplifier 4131 Evans Assoc., Berkeley, CA

Boxcar average, SR 250 Stanford Research Systems,
gated integrator Palo Alto, CA

Computer interface SR 245 Stanford Research Systems,
Palo Alto, CA

Computer PC-XT IBM Corp., Boca Raton, FL




University of Florida Home Page
© 2004 - 2010 University of Florida George A. Smathers Libraries.
All rights reserved.

Acceptable Use, Copyright, and Disclaimer Statement
Last updated October 10, 2010 - Version 2.9.7 - mvs