Resonance line lasers as excitation sources for atomic spectrometry

MISSING IMAGE

Material Information

Title:
Resonance line lasers as excitation sources for atomic spectrometry
Physical Description:
xiv, 160 leaves : ill. ; 29 cm. * 300.
Language:
English
Creator:
Ayala, Norma Lourdes, 1960-
Publication Date:

Subjects

Genre:
bibliography   ( marcgt )
theses   ( marcgt )
non-fiction   ( marcgt )

Notes

Thesis:
Thesis (Ph. D.)--University of Florida, 1992.
Bibliography:
Includes bibliographical references (leaves 157-159).
Statement of Responsibility:
by Norma Lourdes Ayala.
General Note:
Typescript.
General Note:
Vita.

Record Information

Source Institution:
University of Florida
Rights Management:
All applicable rights reserved by the source institution and holding location.
Resource Identifier:
aleph - 001753121
notis - AJG6084
oclc - 26576230
System ID:
AA00003686:00001

Full Text










RESONANCE LINE LASERS AS EXCITATION SOURCES
FOR ATOMIC SPECTROMETRY
















By

NORMA LOURDES AYALA


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

1992


OF FLNOI2A L2lRARIES





























Copyright 1992

by

Norma Lourdes Ayala




























Dedicada a mis padres, Gregorio y Norma Lydia Ayala



"Lo que somos es el regalo que Dios nos hace, lo que Ilegamos a ser es
nuestro regalo a Dios."

unknown













ACKNOWLEDGEMENTS


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

and professional guidance that has made possible the completion of this dissertation.

I also want to thank all the members of the group, past and present, for their

friendship. In particular, I would like to mention Dr. Alicia M. O'Reilly, who made

my coming into the group less frightening, Ms. Nancy J. Mullins, roommate and

friend, and Dr. Giuseppe Antonio Petrucci. Their memories will be always in my

heart.

I would also want to thank Tye Ed Barber. His mechanical and technical

expertise, always admired and envied, made crossing the bridge from concept to data

possible in many of the experiments. His example of hard work and aim for

perfection has been a source of inspiration.

Finally, I would like to thank all the members of my family who on more than

one occasion have been the source of my strength, and especially my nephews,

Javier Elias and Luis Roberto, and my niece, Nicole, who can bring peace with their

smiles.













TABLE OF CONTENTS
page
ACKNOWLEDGEMENTS ................................. iv

LIST OF TABLES ....................................... viii

LIST OF FIGURES ...................................... x

ABSTRACT ............................................ xiii

CHAPTERS

I OPTICAL SOURCES ............................ 1

Introduction .................................. 1
Classification of Optical Sources ...................... 1
Conventional Optical Sources ..................... 2
Hollow Cathode Lamps (HCLs) .................. 2
Electrodeless Discharge Lamps (EDLs) .............. 6
Low Pressure Arc Lamps ..... ................... 9
Lasers as Excitation Sources ...................... 9
Dye Lasers .................................... 11
Excimer Lasers ................................ 15

II RESONANCE LINE LASERS ..................... 19

Introduction .................................. 19
Principles of Operation for RLLs ...................... 19
Photodissociation .............................. 27
Stimulated Emission ............ ...... ........... 28
Recombination ................................ 31
Control of RLLs Output Wavelengths ............... 32
GaI3 RLL Operation ........................... .. 36
Inl and TI RLLs Operation ...................... 42
Comparison of RLLs With Other Excitation Sources ..... 42


III LASER-BASED METHODS OF ANALYSIS ......... 48


Introduction ..................................













Atomic Fluorescence Spectrometry ................... 48
Laser Enhanced Ionization Spectrometry ............. 54

IV EXPERIMENTAL ............................. 57

Preparation of the Resonance Line Lasers ............ 57
Pump Source for the RLLs ....................... 63
Neutral Density Filters .......................... 63
Temperature Optimization of the RLLs .............. 64
Temporal Behavior ............................... 70
Spectral Output ................................ 71
Estimation of the Upper Value of the Linewidth
of the RLLs by the Absorption in a Metal Vapor
Filter Method .............. .................. 71
Standard Solutions ............................. 75
Laser Enhanced Ionization (LEI) .................... 75
LEI Signal Dependence on Applied Electrode
Voltage .................................... 79
Saturation Curves for LEI ........................ 83
Atomic Fluorescence ............................ 83

V RESULTS AND DISCUSSION .................... 89

Temperature Optimization of the RLLs .............. 89
Spectral Output ......................... ....... 89
Estimation of the Upper Value of the Linewidth of
the RLLs by the Absorption in a Metal Vapor
Filter Method ............................... 99
Temporal Behavior .............................. 111
LEI Signal Dependence on Applied Electrode Voltage ... 128
Saturation Curves for LEI ......................... 135
Evaluation of the Analytical Performance of RLLs ....... 140

VI CONCLUSIONS AND FUTURE WORK ............ 149


APPENDIX A INSTRUCTIONS FOR THE PREPARATION
OF THE RLLs AND THE METAL VAPOR
FILTERS .............................. 154













APPENDIX B CALCULATION OF THE METAL AND
METAL HALIDE VAPOR PRESSURE ...... 156

REFERENCES ......................................... 157

BIOGRAPHICAL SKETCH ................................ 160













LIST OF TABLES


Characteristics of Some Electrodeless
Discharge Lamps. ..............

Partial Listing of Wavelength Ranges
for a Rhodamine 590 Dye Laser. ....

Pump Sources for Pulsed Dye Lasers.

Characteristics of Several Dye Laser
Systems. .....................


Table 1-1.


Table 1-2.


Table 1-3.

Table 1-4.


Table 1-5.

Table 2-1.



Table 2-2.


Table 2-3.


Table 2-4.


Table 2-5.



Table 3-1.


page


Excimer Lasers Characteristics ...................

History of the Metal-Atom Photodissociation
Lasers Which Use Metal Halide Salts as the
Active Media. .................................

Effect of a Change in the Pump Source
for the CsI RLL Output. .........................

Effect of a Change in Halogen Partner
on the Na RLL Output ..........................

Operating Conditions for Some Metal
Triiodide RLLS. ...............................

Partial Atomic Energy Level List for
Al, Ga, In, and TI and the Dissociation
Energy of Their Halide Salts .......... ...........

Atomic Fluorescence Limit of Detection
for Ga, In, and T1 Using HCLs, EDLs,
and PDLs as Excitation Sources and a
Flame as the Atomizer. ..........................








Table 3-2.


Table 3-3.


Table

Table


Atomic Fluorescence Limit of Detection (LOD)
for Selected Elements............................

Limit of Detection for One Step Excitation
LEI for Ga, In, and T. ..........................

Conditions for the LEI Experiments. ................

Conditions for the LEI Signal Versus
Applied Electrode Voltage Experiments. .............

Conditions for the Atomic Fluorescence
Experiments. ................................

Operating Conditions for the Ga3, InI, TII RLLs. ......

Absorption Oscillator Strength (ft) Values
for Selected Atomic Transitions of Ga,
In, and TI.....................................

Results Obtained for the Effective Linewidth
of the Absorption Profile. ........................

Analytical Figures of Merit for the Gal3,
InI, and TII RLLs.............................

Comparison of LODs (ng/mL) for LEI ...............

Comparison of LODs (ng/mL) using several
atomic techniques ............................

Partial Atomic Energy Levels for the Alkali Metals
and the Dissociation Energy for Their Halide Salts. .....


Table 4-3.


Table

Table


Table 5-3.


Table 5-4.


Table 5-5.

Table 5-6.


Table 6-1.


52


56

78


82


87

98



110


112


127

147













LIST OF FIGURES


Figure 1-1:

Figure 1-2:


Figure 2-1.


Figure 2-2.



Figure 2-3.

Figure 2-4.

Figure 2-5.

Figure 4-1.


Figure 4-2.


Figure 4-3.

Figure 4-4.


Figure 4-5.

Figure 4-6.

Figure 4-7.


Schematic Diagram of a Hollow Cathode Lamp ........

Schematic Diagram of an Electrodeless
Discharge Lamp. ............................

Physical Processes Involved in the Operation
of a Resonance Line Laser. .......................

Periodic Table. Elements for Which Stimulated
Emission by Photodissociation of the Metal
Halide Salts Has Been Reported ...................

Partial Energy Level Diagram for Gallium ............

Partial Energy Level Diagram for Indium. ............

Partial Energy Level Diagram for Thallium ...........

Schematic Diagram for the Vacuum System Used for
the Preparation of the Cells. ......................

Schematic Diagram of the Experimental Set Up for the
Optimization With Respect to Temperature Experiments.

Oscilloscope Trace of the Energy Meter Output.. .......

Schematic Diagram of the Absorption in a Metal
Vapor Filter Experimental Set Up. .................

Schematic Diagram of the Experimental Set Up for LEI.

Diagram of the Trigger Circuit ................... ..

Schematic Diagram of the Experimental Set
Up for Fluorescence. ............................


page

4


8


21



30

39

44

46


60


66

69


73

77

81


85








Figure 5-1.


Figure 5-2.


Figure 5-3.


Figure 5-4.


Figure 5-5.


Figure 5-6.


Figure 5-7.


Figure 5-8.


Figure 5-9.



Figure 5-10.



Figure 5-11.



Figure 5-12.



Figure 5-13.


Optimization of the Ga RLL with
Respect to Temperature. ..........................

Optimization of the In RLL with Respect to
Temperature. ................................

Optimization of the TI RLL with
Respect to Temperature..........................

Spectral Output of the Ga RLL Showing
Stimulated Emission at 417.2 nm ..................

Spectral Output of the In RLL Showing
Stimulated Emission at 410.2 nm ...................

Spectral Output of the In RLL Showing
Stimulated Emission at 451.1 nm ...................

Spectral Output of the TI RLL Showing
Stimulated Emission at 377.6 nm ..................

Spectral Output of the TI RLL Showing
Stimulated Emission at 535.0 nm ...................

Transmittance of the Laser Line
Monitored at Selected Metal Vapor Filter
Temperatures. Gallium RLL .....................

Transmittance of the Laser Line
Monitored at Selected Metal Vapor Filter
Temperatures. Indium RLL, 410.2 nm line ............

Transmittance of the Laser Line
Monitored at Selected Metal Vapor Filter
Temperatures. Indium RLL, 451.1 nm line ............

Transmittance of the Laser Line
Monitored at Selected Metal Vapor Filter
Temperatures. Thallium RLL, 377.6 nm line .........

Transmittance of the Laser Line
Monitored at Selected Metal Vapor Filter
Temperatures. Thallium RLL, 535.0 nm line .........


91


93


95


97


101


105


107



114



116



118


120


122








Figure 5-14.

Figure 5-15.

Figure 5-16.


Figure 5-17.



Figure 5-18.


Figure 5-19.

Figure 5-20.

Figure 5-21.

Figure 5-22.

Figure 5-23.


Temporal Behavior of the Ga RLL. .................

Temporal Behavior of the In RLL. .................

Plot of LEI Signal Versus Electrode
Voltage (25 ppm Solution of Ga). ..................

Plot of LEI Signal Versus Electrode
Voltage (Air-Acetylene Flame,
20 ppm Solution of In)...........................

Plot of LEI Signal Versus Electrode
Voltage (10 ppm Solution of T). ...................

Saturation Curve for Ga LEI .................... .

Saturation Curve for In LEI (Air-Acetylene Flame). .....

Calibration Curve for Ga LEI. ......................

Calibration Curve for In LEI (Air-Acetylene Flame). ....

Calibration Curve for Tl LEI ...................


124

126


130



132


134

137

139

142

144

146













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

RESONANCE LINE LASERS AS EXCITATION SOURCES
FOR ATOMIC SPECTROMETRY

By

NORMA LOURDES AYALA

May 1992

Chairperson: James D. Winefordner
Major Department: Chemistry

Resonance line lasers (RLLs) were evaluated as excitation sources for atomic

spectrometry. Three RLLs, employing Gal3, InI, and Til, were constructed and

characterized. The Analytical figures of merit (AFOM) of the RLLs such as

spectral output, source irradiance, temporal behavior, and source lifetime were

determined. Also, the applicability of RLLs to spectrochemical methods of analysis

such as laser enhanced ionization (LEI) and laser induced fluorescence (LIF) in

flames was demonstrated. The limits of detection obtained for both of these

techniques using RLLs as excitation sources were in the parts-per-billion (ppb)

range.

Resonance line lasers are classified as line sources; they produce radiation at

fixed wavelengths. The wavelengths which are obtained correspond to atomic








transitions. RLLs are coherent sources and have peak powers in the order of kW

and irradiances of several hundreds of kW cm2. The effective combination of

narrow linewidth, high irradiance and coherence of a laser makes the RLL unique

as a true coherent line source.













CHAPTER I
OPTICAL SOURCES


Introduction

The ideal excitation source for atomic spectrometry is stable, intense,

available for all elements, tunable to different excitation transitions, and long lived.

These characteristics are not possessed by any single excitation source, so there are

a wide variety of sources to satisfy different needs. The following is a discussion of

the most commonly used excitation sources in atomic spectrometry.

Classification of Optical Sources

Spectrochemical methods of analysis such as absorption and fluorescence

require the use of an external radiation source. A radiation source might be

classified according to its spectral output, temporal behavior, irradiance or radiance,

stability and lifetime (1).

A source is classified as a line or continuum source depending upon how the

source profile or linewidth (AAs) compares to the absorption profile (AA.). If AA,

> AA.A, then the source is classified as a continuum source. If AA, < AXA, then the

source is classified as a line source. A continuum plus line source produces narrow

lines superimposed on a spectral continuum. If the source emits radiation

continuously with respect to time, it is called a continuous wave (cw) source. This

type of source is on all the time. If the source is intermittent with respect to time,








2
it is called a pulsed source (1). The irradiance of the source is defined as the power

per unit of area (kW cm'2), and the radiance of the source is the power per unit

area per unit of solid angle (kW cm'2 s'1).

Conventional Optical Sources

The term conventional source usually refers to a non-laser excitation source.

The following is a brief review of the most commonly used conventional line sources

for atomic spectrometry: the hollow cathode lamp (HCL) and the electrodeless

discharge lamp (EDL).

Hollow Cathode Lamps (HCLs)

A hollow cathode lamp (HCL) consists of a hollow cylinder made of or

coated with the element of interest, or an alloy of the element (1-3). The anode

usually consists of a wire. A pyrex or quartz cylinder enclose the anode-cathode

assembly. A few torr of a filler gas (200-1000 Pa), usually neon or argon, is placed

inside the envelope. A schematic diagram of a HCL is shown in Figure 1-1.

An external power supply is used to provide a potential difference of a few

hundred volts (150-300 V) between the anode and the cathode. This potential

difference causes ionization of the filler gas. Cathodic sputtering is produced when

cations strike the cathode, volatilizing the cathode material. The atoms inside the

vapor cloud, which is formed around the cathode, are excited by collisions with other

species present, such as electrons and filler gas atoms. The excited atoms relax by

emitting radiation at the characteristic lines of the element. Among the































Figure 1-1. Schematic Diagram of a Hollow Cathode Lamp.










Cathode

Anode

Window








5

disadvantages of HCLs is that not only are spectral lines due to the cathode

elements obtained, but also lines from the filler gas and from any impurity present.

This optical source has been developed extensively and is readily available. Single-

element as well as multi-element lamps are commercially available. Hamamatsu

(Bridgewater, NJ) offers 5 multi-element lamps (i.e., Na-K, Ca-Mg, Si -Al, Fe Ni,

Sr Ba).

HCLs operate at low pressure and low temperature, minimizing collisional

and Doppler broadening. HCLs are capable of producing very narrow atomic lines

on the order of 0.01 to 0.02 A (FWHM).

The lamp irradiance can be varied by changing the lamp current (10-50 mA).

Maximum, minimum, as well as optimum currents are suggested by the

manufacturer. The operating current changes the half width of the lines and the

lifetime of the lamp. Commercially available lamps include high intensity or boosted

output HCL, electrically modulated, and pulsed lamps. An increase in the lamp

operating current will increase the temperature of the HCL, which will result in

broader lines and also will reduce the lamp output intensity due to self-absorption

and self-reversal.

Self-absorption occurs when the concentration of the atoms around the

cathode is so high that emission is reabsorbed by the atoms. As a result of self-

absorption, the line profile is broader, and the peak intensity decreases. Self-

reversal occurs when there is a temperature gradient across the atomic cloud which

surrounds the cathode. The presence of a dip at the center of the profile is








6
indicative of this phenomena. Self-reversal has been described as a special case of

self absorption.

The lifetime of a given HCL is given by the product of the peak operating

current and the accumulated operating time. Hamamatsu Corp. lists lifetime-current

products to be around 3000 mA-hr for the As, Ga, and Hg lamps and 5000 mA-hr

for the other lamps (3).

Electrodeless Discharge Lamps (EDLs)

An electrodeless discharge lamp (EDL) consists of a sealed quartz tube which

contains a few torr of an inert gas and a small amount of the metal or metal halide

salt of interest (1-2, 4-8). An antenna or a waveguide cavity directs an intense RF

or microwave field on the lamp. If operating frequencies are of the order of 10-

3000 MHz, little self-reversal is observed. The discharge is started by the ionization

of the inert gas atoms using a Tesla coil. The external field accelerates the electrons

produced; the electrons acquire enough energy to maintain the plasma. The metal

or metal salt is vaporized by the heat produced. The metal vapor atoms are excited

by collisions with electrons and the emission spectrum of the metal is produced upon

relaxation. A schematic diagram of an EDL is shown in Figure 1-2 (4).

EDLs can have radiances 20 to 50 times greater than those of the HCLs. For

elements for which the HCL intensity is relatively low, the use of an EDL is

preferred (e.g., As, Se, Te, Hg). These sources have not been as extensively

developed as HCL. Table 1-1 shows the characteristics for some EDLs. HCLs and

EDLs have been used extensively in atomic fluorescence spectrometry.































Figure 1-2. Schematic Diagram of an Electrodeless Discharge Lamp (Taken from
Reference 4).











EDL


Glass
Wool *


Antenna




I-


Quartz
/Tube


Metal or
"*Metal Halide


Reflector

Microwave
Power Supply
Coaxial
Connector


23>









Low Pressure Arc Lamps

A low pressure arc lamp consists of a sealed glass or quartz tube filled with

metal vapor in which an arc discharge is formed between the two electrodes placed

inside the lamp (1). Low-pressure arc lamps are commercially available for the

following elements: Hg, Cd, Zn, Ga, In, TI, and the alkali metals. The most

commonly used low pressure arc lamp is the mercury lamp for which the main line

is the 253.7 nm line. These lamps are used to calibrate spectroscopic equipment.

Lasers as Excitation Sources

The characteristics of a laser as an excitation source are different from the

characteristics of conventional sources (1,9-10). Laser radiation is described by high

directionality, spectral purity, coherence, and high irradiance. The irradiance of a

laser is 4 to 10 orders of magnitude larger than the irradiance obtained with

conventional sources. The development of laser technology has led to the

improvement in limit of detection (LOD), selectivity, and sensitivity of techniques

in which the analytical signal depends on the source intensity. The use of lasers in

atomic fluorescence (AF) and laser-enhanced ionization (LEI) has made possible

the achievement of LODs in the sub-ng/mL range. Only two types of lasers will be

described in this section: the dye laser and the excimer laser.









Table 1-1. Characteristics of Some Electrodeless Discharge Lamps.

Element Ga In TI

Chemical form of metal iodide metal iodide metal
element used GaL/Ga Inlfn

Wavelength (nm) 412.2 410.5 377.6

Inside diameter 9.0 9.0 9.0
(mm)

Length (cm) 3.9 3.4 3.3

Volume (cm3) 2.0-2.4 2.0-2.4 2.0-2.5

Argon pressure 2.0 2.0 2.0
(torr)

Source radiance 3 x 10-3 1 x 10'3 4 x 104
(W cm-2 sr"1)

1LOD for AFS using 1.0 0.2 0.1
EDLs (ppm)


'LOD = Limit of Detection
AFS = Atomic Fluorescence Spectrometry
EDLs = Electrodeless Discharge Lamps
Taken from Reference 4.









Dye Lasers

A dye laser is a laser which uses an organic dye solution as the laser media

(9-10). A dye is a solution of a given organic molecule which possesses the property

of lasing (e.g., rhodamines, coumarins, fluoresceins, etc.). The wavelength range

that is available depends on the particular dye. The entire spectral region from 400

nm to 800 nm can be covered using a dye laser with several different dyes. No

single dye, however, can be tuned for the whole visible region of the spectrum. The

range of wavelengths that can be obtained from a given dye depends not only on the

pump source used but also on the solvent. Table 1-2 shows a partial tunability curve

for Rhodamine 590 dye (11).

Since dye lasers can be tuned across a wide range of wavelengths, additional

optics are required in order to select the particular wavelength of interest. The laser

cavity has a dispersive device, usually a grating, to select the wavelength range. The

grating is rotated in order to select the desired wavelength. Gratings, prisms,

interference filters, and etalons have been used for tuning purposes. Also, an

optical pumping source is required for the operation of dye laser systems.

Flashlamps, pulsed and continuous wave (cw) lasers have been successfully used for

pumping. Table 1-3 presents a list of commonly used pump sources for dye laser

systems.




















--o- -
~~V-4
kn S -S 00 T- ..0 I M -x
iaro # < K K a K
mCIoo c




















w %0 0 '0



Oar S t O W S I a \ S I 8
Wn Wlkono"W l" ) W



o00 So 2 22 1











Kinitr'^^'' '^i^i \o II^1


M
0





4)



.2
.!






W I





4a

0
a 0

Z I



. .9



S 4)
s I



nII 0
P gf
OBa-i
J+3 ,< t

0(














- Y -,


V b bbbe

lovE- i EsyT vs- 0 0 0 i< -l-l0
lir^^s k
(SM *O M W


I o 00



006


04
(U~b

0l ~;


-- o- -- o o --
0 0 0 000 0
4) (cc 03 _
0 1 d C <4. .d -= c
en en en enasse en in v4







93 m(m 00 -W mi -


.- V) '--'e V--4 .-o


QQ M z z00 I




0W) 00 t- 00 -it .o C .O'-4 I-



9 saS



N O M O b mOO^c O H N N
0000 00 a or- 0o o\
Q4
g v r~ zzzz'-~o~~~-










Table 1-3. Pump Sources for Pulsed Dye Lasers.


Pump Source Wavelength of the Pump Source
__(nm)


Excimer Lasers

ArF 193

KrF 249

XeCl 308

XeF 351

Copper vapor (Cu) 510

Copper vapor (Cu) 578

Nd-YAG 1060

frequency-quadrupled Nd-YAG 266

frequency-tripled Nd-YAG 355

frequency-doubled Nd-YAG 532


I








15
Being organic molecules, dyes do not necessarily dissolve readily in water.

Methanol, ethanol, benzyl alcohol, dimethylformamide, dimethylsulfoxide, p-dioxane,

toluene, and ethylene glycol have been used as solvents to prepare dye solutions.

The dye solution degrades with use, and usually, the dye solution is pumped through

a cuvette to minimize degradation or photodecomposition. The typical lifetime of

a dye solution is 10 to 100 hours of use. Table 1-4 shows the characteristics of some

dye laser systems (12,13).

The dye laser is the most frequently used laser in the visible region of the

spectrum. The major advantage of a dye laser system is its versatility, and this

versatility lies in its wavelength tunability. Wavelengths that are not available using

a dedicated laser can be obtained using a dye laser. The price to pay for this

versatility is complexity.

Excimer Lasers

An excimer ("excited dimer") laser is a type of gas laser which has been

available since 1975 (9,10). This type of laser is pumped by an electrical discharge.

Excimer lasers emit radiation in the UV region of the spectrum, and for this reason,

they are typically used in photodissociation studies and as pump sources for other

lasers such as dye lasers.

The ground state of an excimer molecule is dissociated thus, this type of

molecule can only exist in the excited state. Commercial excimer lasers are able to

handle several gas mixtures. The gas mixture used determines the output












Characteristics of Several Dye Laser Systems.


Pump Source Excimer, N2 Flashlamp Ion Laser
Nd:YAG__


Tunability Range1
(nm)

Average Power
(W)

Pulse Duration2
(ns)

Repetition Rates
(Hz)

Lifetime


Beam Diameter
(mm)

Beam divergence
(mrad)

Cost (K$)


300-1000


0.05-15


3-50


up to 10K


hrs to 2 months


2-10


0.3-6


4-100


340-940


0.25-50


200-4000


0.03-50


104 to 106 shots
per flashlamp

5-20


0.5-5


6-50


400-1000


up to 2


cw or ns to ps


cw or pulsed


hours


0.6-1.0


1-2


8-50


'Require change in dye solution to obtain the whole tuning range.
2cw = continuous wave laser
Taken from Reference 12 and 13.


Table 1-4.








17
wavelength which is obtained. Table 1-5 shows the characteristics for some excimer

laser systems (12,13,14).

In order to change the gas mixture, the old gas mixture is pumped out. The

laser cavity is passivated to remove contaminants, and the gas mixture is replaced

with the new gas mixture. As can be seen from Table 1-5 a more accurate name for

this type of laser is an exciplex laser since the molecule formed upon electrical

discharge is an excited complex.









Excimer Lasers Characteristics.


Excimer ArF KrF XeCI XeF
Wavelength 193 248 308 351
(nm)

Energy/Pulse 0.5 1.0 1.5 0.5
(J/pulse)

Repetition 1-1000 1-500 1-500 1-500
Rates (Hz)

Pulse 5-25 2-50 1-80 1-30
Duration
(ns)

Maximum 50 100 150 30
Average
Power
(W)

Lifetime 104-5x106 104-107 105-2x107 104-107
(shots/gas
fill)


Beam 2-6
divergence
(mrad in the
rectangular
beam)

Beam 2 x 4 to
dimensions 25 x 30
V x H (mm)1

Price (K$) 30-200

1Vertical x Horizontal dimensions
Taken from References 12,13,14.


Table 1-5.













CHAPTER II
RESONANCE LINE LASERS


Introduction

Lasers which use photodissociation as a mechanism to obtain stimulated

emission were first described by physicists about twenty years ago (15). A summary

of the history of metal-atom photodissociation lasers which use metal halide salts as

the active media is shown in Table 2-1 (15-33). Although these lasers are capable

of producing narrow linewidth radiation at atomic resonance transitions wavelengths,

they have found very little application in spectrochemical methods of analysis

(23,33,34). This is unfortunate because resonance line lasers (RLLs) have features

that could be exploited in analytical atomic spectrometry.

Principles of Operation for RLLs

Laser action by photodissociation can be described by the following physical

processes: photodissociation of the metal halide salt vapor, stimulated emission of

the excited metal atoms, and effective recombination of the photofragments back to

the metal halide salt. Figure 2-1 depicts the physical processes involved in the

operation of a resonance line laser (33,35).
































Figure 2-1. Physical Processes Involved in the Operation of a Resonance Line
Laser.



















Pho tod ---ssoc- on


"VA'


Recomb nat ion M+X


INTERNUCLERR DISTANCE


St imu I a ed
em i S ion









Table 2-1. History of the Metal-Atom Photodissociation Lasers Which Use Metal
Halide Salts as the Active Media (SE = Stimulated Emission).


1965 Zare & Herschback proposed the concept of metal-atom
(15) photodissociation lasers, but the unavailability of the
appropriate pump sources in the UV region of the spectra
prevented the immediate application of their ideas.
Year of Pump Source Active Observed SE Comments
Publication Media Wavelengths
(Reference) (nm)

1977 ArF InI 410.5 Lasing action of
(16) excimer laser 451.1 In is reported.
(. = 193 nm)

1978 ArF TII 377.6 TI RLL is
(17) excimer laser 535.0 reported for the
1st time.










Table 2-1--continued.


Year of Pump Source Active Media Observed SE Comments
Publication Wavelengths
(Reference) (nm)


ArF
excimer laser


Nal



KI







RbI










CsI


1979
(18)


Alkali-metal
halide RLLs
are reported.


_________________ __________________ I __________________ & _________________


589.0
589.6
1140.0

766.5
769.9
404.5
1250.0
1170.0
2720.0
3150.0

780.0
794.5
421.0
775.8
761.9
1370.0
2790.0
1530.0
1480.0
2290.0

852.1
894.3
455.5
3010.0
1360.0
1470.0
917.3
876.4
1010.0
1360.0
2950.0
4220.0










Table 2-1--continued.


Year of Pump Source Active Media Observed SE Comments
Publication Wavelengths
(Reference) (nm)
1979 ArF Gal3 417.2 nm First metal
(19) excimer laser triiodide RLL.
Two photon
mechanism is
involved.
1979 ArF AlI3 394.4 Multi-photon
(20) excimer laser 396.2 mechanism
is involved.
Gal3 417.2

InI3 410.5
451.1

BiI3 472.2
1980 ArF PbBr2 364.0 Two photon
(21) excimer laser 368.3 mechanism.
405.8
1980 ArF TIBr 377.6 Considerably
(22) excimer laser 535.0 lower
efficiency
than the
T__ RLL.
1980 ArF TIl 377.6 RLLs are 1st
(23) excimer laser 535.0 used for
Nal 589.0 analytical
589.6 purposes.
1140.0 (LEI
1138.2 experiment).
1980 ArF SnI2 380.1 Multiphoton
(24) excimer laser mechanism
SbI3 326.7 where two or
more photons
Gel4 326.9 are involved.









Table 2-1--continued.


Year of Pump Source Active Observed SE Comments
Publication Media Wavelengths
(Reference) (nm)
1980 ArF TIl 377.6 Construction
(25) 535.0 of laser system
with external
resonator.
1981 N2 TIBr 377.6 Two photon
(26) laser 535.0 mechanism
(A = 337 nm)
1981 KrF HgI2 435.8 Multiphoton
(27) excimer laser 546.1 mechanism.
= 248 nm
CdI2 480.0
508.6

ZnI2 472.2
481.1
1982 ArF Til 377.6 Miniaturization
(28) excimer laser 535.0 of the TII RLL.
SE from an
active volume
of 10-1 to 10.2
mm3 is reported.

1982 ArF HgBr2 404.7
(29) excimer laser
1983 N2,Xe,Kr,ArF TII 377.6 N2 and ArF
(30) flashlamps 535.0 flashlamps
for pumping are best suited
RLLs were to pump
evaluated. the TII RLL.
1986 ArF TII 535.0
(31) flashlamp










Table 2-1-continued.

Year of Pump Source Active Media Observed SE Comments
Publication Wavelengths
(Reference) (nm)
1988 4th harmonic PbCl2 374.0
(32) Nd:YAG
(A = 266.0
nm)
1990 ArF InI 410.5 The use of
(33) excimer laser 451.1 RLLs for
atomic
Til 377.6 fluorescence
535.0 is reported
for the
first time.










Photodissociation

Photodissociation of the metal halide salt vapor occurs according to the

following reaction

MX + hvpup -* M* + X

where

MX = metal halide salt

hvp, = photon energy of the pump source

M* = excited state of the metal atom

X = halide atom

First, the metal halide salt vapor absorbs radiation from the pump source.

The pump source has to provide enough energy to photodissociate the molecule

(MX) and excite the metal atom which, in most cases, means that

hvu,p Ea + ED

where

ED = dissociation energy of the metal halide salt

Ea = highest atomic level accessed upon photodissociation

The molecule can be photodissociated by either a one-photon or a multi-photon

process. The specific mechanism depends on the valence state of the metal of use.

The fact that the lasing mechanism involves photodissociation of the metal

halide salt in the vapor phase places at least two requirements on the type of salts

that can be used as active media. The first requirement is volatility since the metal

halide salt has to be heated to achieve an appropriate vapor density. Also, the salt








28
has to be thermally stable under the operating conditions of the laser, and its

dissociation energy has to be within the reach of the available pump sources. Up

to now, the metal halide salts which have been most successful have been the

iodides and bromides. These salts seem to provide a highly volatile, relatively

thermally stable, accessible, and inexpensive lasing media. By far, the most used

salts have been the metal iodides.

The pump source has to be able to meet the energetic requirements of the

photodissociation process which means that a pump source in the ultraviolet (UV)

region of the spectrum is needed. Excimer lasers, frequency-multiplied Nd:YAG,

and nitrogen lasers have been used as pump sources. Although the most used pump

sources have been lasers, flashlamps have also been used as pumping sources

(30,31).

Stimulated Emission

Figure 2-2 shows for which elements stimulated emission by photodissociation

has been reported. After photodissociation, the excited metal atom relaxes by

stimulated emission. Most photodissociation lasers described in the literature

involve optical transitions which are strongly allowed, and they can work without a

cavity. Since the photodissociation of the molecule is able to produce high inversion

densities, stimulated emission is observed from excited states of the metal. The

details of the stimulated emission observed for these lasers will be described in

Chapter 5 of this dissertation.

















0
.2



0

.0
.2

o


cU



.o.



a)





.- 4

0



a)

.I-

















0


E P P0- c3 9-

--MCC
=d ~ ~- _d


(o)


0





E-*
f-



c->





>~n


MCD


-I




E-
a^


52f


0 C C->
= C/-.3 ( IC C
C I-O CO o
MC


1111


A A I I I


C/') E- C


E-

















C 3




-o


-I

c-


.-


S-i


, D, ^42


^ss


5t
O



2;

E-

C7S




e7
P3








r>

cai


. -
--=










Recombination

After stimulated emission is observed, the metal and halogen atoms

recombine quickly to form the metal halide salt. This mechanism is highly effective

for metal-halide lasers, the parent metal-halide molecule is more strongly bound

than all the other compounds which are possible involving the metal or the halogen

(e.g. metal diatomics, halogen diatomics). When the metal halide salt is placed in

an unbuffered cell, the recombination is accomplished primarily by wall-stabilized

binary reactions. Under the conditions at which the cells are operated, the medium

can be considered collisionless during the duration of the laser pulse. When a

buffer gas is included in the cell, the recombination of the metal halide salt is

accomplished primarily by three-body collisions.

Since RLLs are sealed lasers, the recombination of the photofragments back

to the parent molecules is essential for the effectiveness of the system.

Recombination is a highly efficient step, and the recombination is done in a Is time

scale so that the parent molecule is formed before the next laser pulse. Based on

this time frame, it has been estimated that these lasers could be operated at

frequencies in the kHz range. This has not been confirmed experimentally due to

the lack of pump sources which can provide such repetition rates. Up to now, none

of the lasers reported in the literature has suffered degradation, and uses of more

than 107 shots have been reported (35).










Control of RLLs Output Wavelengths

With few exceptions, the highest atomic level accessed upon photodissociation

of a metal halide salt is given by the following formula,

Ea ~ hvppu ED (2-1)

where Ea is the highest atomic level accessed, hvpp is the energy provided by the

pump source, and ED is the dissociation energy of the metal halide salt (35).

Sometimes an energy gap does exist and lasing action from an atomic level which

is slightly above hvp.p ED is obtained. This is found in RLLs which work through

a two-photon photodissociation mechanism. When a multiphoton dissociation

process is involved, the first step of the multiphoton mechanism will not totally use

the energy provided by the pump source. This means that the photofragments can

be liberated with an excess vibronic, rotational, or translational energy. Because the

photofragments are liberated in an excited state, higher levels in the atomic system

can be obtained upon absorption of another photon. A description of this

phenomena will be given for the operation of the Gal3 RLL.

From equation (2-1), it can be seen that the two ways of controlling the

output wavelengths from an RLL are either a change in pump source or a change

in halogen partner in the metal halide salt.

A change in pump source will change the energy which is available for

photodissociation and access of atomic levels. An example of this is shown in Table

2-2. In this table, it can be seen that a change in pump source for the CsI RLL

dramatically changes the wavelengths for which lasing action is observed.








33

Table 2-2. Effect of a Change in the Pump Source for the CsI RLL Output (ED
= 3.4 eV).


Pump Source
excimer laser

Energy of the
pump source (eV)

E. ~ hvPP ED (eV)

Observed lasing
transitions

62P3 62S2,

62P -. 62S

72Ds5/ 62pl/2

72D5/ -+ 62P32

62P3/ -5 52D,3


ArF


6.4


3.0




852.1 nm

894.3 nm

1.36 jpm

1.47 ipm

3.01 jim

2.94 upm

917.3 nm

876.4 nm


KrF


5.0


1.6




852.1 nm

894.3 nm

Not observed

Not observed

Not observed

Not observed

Not observed

Not observed


Taken from Reference 35.


1 --








34

CsI has a dissociation energy of 3.4 eV. If an ArF excimer laser is used as

the pump source, the highest atomic level accessed will be approximately 3.0 eV

above the Cs atom ground state. This will provide lasing output at eight

wavelengths. If an KrF gas mixture is used for the excimer laser, the highest atomic

level accessed will be 1.6 eV above the Cs ground state. Excitation of the CsI RLL

with a KrF excimer laser will provide lasing action at only two wavelengths. The

judicious choice in pump source will allow a better selection of wavelengths and

more efficient use of the pump source energy.

For example, in a flame experiment, most of the atomic population is in the

ground state or in the first excited state; therefore, the KrF excimer laser will be the

preferred method of excitation since it only provides the atomic lines of most

analytical interest. This minimizes the possibility of spectral interference as well.

A change in the halogen partner of the metal halide salt can also determine

which atomic levels are accessed. A change in halogen changes the dissociation

energy of the salt. In Table 2-3, the effect of changing the halogen partner on the

sodium RLL wavelength output is shown when an ArF excimer laser is used for

photodissociation. When Nal is used as the lasing media, one photon from the ArF

excimer laser has enough energy to photodissociate and excite the Na atom so that

laser action is observed at three different wavelengths as illustrated in Table 2-3.

If NaBr is used as the salt, the ArF laser will dissociate the molecule but lasing










Table 2-3. Effect of a Change in Halogen Partner on the Na RLL Output.


Metal halide salt Nal NaBr


Dissociation Energy
(eV)

Pump Source
excimer laser

Energy of the
pump source (eV)

Ea ~ hvpp ED (eV)

Observed lasing
transitions

32P3/ -- 32S12

32P1/ -* 32S 1

42S2 32P3


3.05


ArF


6.42


3.37




589.0 nm

589.6 nm

1.14 Vm


_________________ 4 I


ArF


6.42


2.62




589.0 nm

589.6 nm

not observed


Taken from Reference 35.








36

action will only be observed on two lines since the residual energy (E) is not

enough to populate the 42S,1 level of the sodium atom.

Ga3 RLL Operation

Metal triiodides can also be used as the lasing media (19,20). Lasing on the

atomic transitions of the elements of group IIIA and VA has been reported using

metal triiodide salts. Table 2-4 shows the operating conditions for these lasers (20).

It is believed that lasing action takes place by a sequential two-photon

absorption mechanism. Ehrlich, et al. (20) have suggested the following mechanism

for the operation of the metal triiodide RLLs. Although the mechanism was

suggested for all the metal triiodides salts studied (Al, Ga, In, Bi), only the

energetic of the gallium RLL are considered for this dissertation. Figure 2-3 shows

a partial energy level diagram for the gallium atom (35). Table 2-5 shows a partial

atomic energy level list for Ga, In, and TI and the dissociation energy of their halide

salts (36-39).

The first step in the photodissociation mechanism of the gallium RLL is

either

Energy
Required
(eV)

GaI3 + hvpu, Gal + 21 2.8

or


Gal3 + hv1, -a Gal + 12









Operating Conditions for Some Metal Triiodide RLLS.


All3 Gal3 InI3 BiI3

Wavelength 394.4 410.5 472.2
(nm) 396.2 417.2 451.1

Pulse width 4.0 4.0 -- 3.2
(ns)

Operating 135 144 218 285
Temperature
(*C)


Taken from Reference 20.


Table 2-4.






























Figure 2-3.


Partial Energy Level Diagram for Gallium. The Atomic Zero Energy
Level is Displaced by the Dissociation Energy of the Monoiodide Salt.
The Dashed Line Represents the Energy of the ArF Excimer Laser
(6.42 eV), and the Dotted Line Represents the Dissociation Energy
of the Gal (3.46 eV).












S -- 6.20

417.2 nm
4 4.96
E
E >


0 2L
3 4 32 3.72
w

W 2 2.48 L



1 1.24










Table 2-5.


40
Partial Atomic Energy Level List for Al, Ga, In, and TI and the
Dissociation Energy of Their Halide Salts.


Atomic Energy Terms
Metal Dissociation
Halide energy Term Energy1
Salt (eV) Element Symbol (cm-1)

TICI 3.80 Ti 62P, 0

TIBr 3.38 62P3 7,793

TII 2.9 72S/2 26,477

InCI 4.5 In 52P,2 0

InBr 4.0 52P32 2,213

InI 3.35 62S, 24,373

GaCI 4.94 Ga 42P1 0

GaBr 4.35 42P32 826

Gal 3.46 52S,/2 24,789

AICI 5.1 Al 32p1/2 0

AIBr 4.5 32P32 112

All 3.8 42S/2 25,348


are measured from the state of lowest energy.


'The energy values









and then

Gal + hvpp -- Ga* + I 6.47

where Ga' represents the excited electronic state of the gallium atom.

The ArF excimer laser is able to provide 6.42 eV photons, and the

dissociation of the Gal3 by either of the suggested first steps does not require this

much energy. The remaining energy of the photon can excite the vibrational, and

rotational levels of the photofragments. The photofragments can be liberated with

high translational energy too. Osgood, et al. (20) have suggested that this excess of

energy provides the additional energy required for the excitation of gallium through

the last step.

At this time, it has been difficult to determine which first step is more

important for a given metal triiodide RLL. The absence of Iz fluorescence in the

case of the gallium RLL indicates that the first step of preference is

Gal3 + hvp. -* Gal + 21

The use of triiodides as lasing media has the advantage of lower operating

temperature since the triiodides have a higher vapor pressure compare to the

monoiodides. As can be seen from Table 2-4, an operating temperature of 218"C

was required for the InI, whereas a temperature of 330C was required for the InI

studied in this dissertation (see Table 5-1).

Triiodides salts are more susceptible to thermal decomposition though, and

they decompose into the monoiodide above certain temperatures. In the case of

InI3, this occurs at temperatures around 3000C. Gal3 decomposes at ~ 5500C which








42
is above the operating temperature used in this work. Because metal triiodide salts

are able to produce lasing action, the possibilities of laser media greatly increases.

InI and TII RLL Operation

The proposed mechanism to achieve stimulated emission by photodissociation

in the case of InI and TlI is the following

InI + hvpup In* + I

and

Tll + hvp + pumpI.

where In* and TI* represent the excited electronic state of the indium and thallium

atoms, respectively (35). The energetic of these two processes are depicted in

Figure 2-4 and Figure 2-5, respectively. When monoiodides are used as the active

media, lasing action is observed after the absorption of one photon. In this case,

again, the pump source should provide enough energy to photodissociate the salt

and excite the metal atom. Lasing action is observed at two wavelengths for the

indium and thallium RLLs, and it is produced by the following transitions

n 2S (n 1) 2Pl/2

n 2S1/ (n 1) 2P3/2

where n = 6 for indium and n = 7 for thallium.

Comparison of RLLs With Other Excitation Sources

Resonance lines lasers (RLLs) are classified as line sources, and they produce

radiation at fixed wavelengths. The wavelengths which are obtained correspond to

atomic transitions. RLLs have in common these two characteristics with





























Figure 2-4. Partial Energy Level Diagram for Indium. The Atomic Zero Energy
Level is Displaced by the Dissociation Energy of InI. The Dashed
Line Represents the Energy of the ArF Excimer Laser, and the
Dotted Line Represents the ED of the InI (3.35 eV).






44





5 6 2 1/2 6.20
451.1 nm

S4 4.96
E 410.2 nm >
0 ---.
\ \5 2P
3 3 3.72 (
...................... ................. IT
> LU
( 5 2 P Z
Cc 1/2 L
S2 2.48


1.24
1 1.24






























Figure 2-5. Partial Energy Level Diagram for Thallium. The Atomic Zero Energy
Level is Displaced by the Dissociation Energy of Til. The Dashed
Line Represents the Energy of the ArF Excimer Laser, and the
Dotted Line Represents the ED of the TIl (2.9 eV).













6.20



4.96



3.72



2.48



1.24








47
incoherent sources such as hollow cathode lamps and electrodeless discharge lamps.

RLLs are coherent sources and can have peak powers in the order of kW and

irradiances of several hundreds of kW cm-2. Because RLLs are coherent sources

and have multiwavelength emission, RLLs can be compared with dye lasers.

RLLs as well as dye laser systems require the use of a pump source. Dye

lasers are equipped with tuning elements for wavelength selection. RLLs, on the

other hand, are naturally locked to atomic transitions.













CHAPTER III
LASER-BASED METHODS OF ANALYSIS

Introduction


Atomic fluorescence spectrometry (AFS) and laser-enhanced ionization (LEI)

were chosen as the techniques to demonstrate the applicability of RLLs to

spectrochemical methods of analysis. The purpose of this chapter is to cover the

basic principles of both techniques, and present their analytical figures of merit.

Excellent reviews covering the theoretical and practical aspects of AFS and LEI as

analytical techniques are reported in the literature (40-45). The available

combinations of excitation sources, atomizers, experimental configurations, and

applications are immense. Only a very brief overview of the techniques is presented

here; it is not an exhaustive discussion by any means.

Atomic Fluorescence Spectrometry

Atomic fluorescence spectrometry (1) was proposed as an analytical technique

by Winefordner and Vickers in 1964. In atomic fluorescence spectrometry, radiation

is absorbed by the atoms, and the radiation promotes an electron to a higher lying

electronic state. Following excitation, the atom relaxes by the emission of light.

Fluorescence can be classified according to the electronic states involved in the

transition. The excitation process can be entirely radiative or can be collisionally

assisted. The same principle applies for the de-excitation process.

48








49

When the excitation and emission is monitored at the same wavelength, the

fluorescence is called resonance fluorescence. The term nonresonance fluorescence

applies when the energy levels involved in the excitation process and the emission

process are not the same.

Atomic fluorescence offers excellent analytical figures of merit that make it

very attractive and useful as a spectrochemical method of analysis. Atomic

fluorescence has multielement capabilities, and its linear dynamic range can extend

from 3 to 8 orders of magnitude. A precision from 0.5 to 5% can be obtained, and

the limits of detection are in the ng/mL range. Table 3-1 shows the atomic

fluorescence limit of detection for Ga, In, and 1T using hollow cathode lamps,

electrodeless discharge lamps and pulsed dye laser systems as excitation sources and

a flame as the atomizer (40). Table 3-2 shows atomic fluorescence limit of detection

for selected elements (40).

The basic instrumentation in atomic fluorescence include an excitation source,

an atomizer, a device for wavelength selection, and detection and signal processing

systems. Flames, plasmas, as well as electrothermal devices have been used as

atomizers. Line sources, continuum, ICP, and lasers are among the optical sources

which have been used as excitation sources. A way to distinguish the analyte

fluorescence from the background signal is needed. Usually, a monochromator is

used for selecting the emission line. The fluorescence signal is collected at angles


















S



N
a








W 31C a Wo 3



&

ad1
- .I







So o a c
(B S ^^^tf1 -7^-"'

1-


U l h^

o o
0 >00


















0

























%0 %0 00 % 00 o 0 0 o
o %







0o





Ewmme~m


4)
0 0





S4)
S i
R z



CO




II II


4)


.4)

I-6
0 a


aga2



II SI










Atomic Fluorescence Limit of Detection (LOD) for Selected Elements.


Element Excitation Fluorescence Cell' Excitation2 LOD
Wavelength Wavelength Source (ng/ml)
(nm) (nm)


Ag
Al

As
Au
Ba
Be
Bi
Cd
Ca
Ce
Co
Cr
Cu
Eu
Fe
Gd
Ge
Hg
K
Li
Mg
Mn
Mo


328
308.2
394.403
193.07
242.795
455.4
234.7
306.772
228.802
396.8
371.637
304.4
357.8
324.7
287.9
248.6
407.84
265.1
253.65
766.5
670.8
279.1
279.0
313.3


338
308.2
396.153
193.7
242.795
455.4
234.7
472.219
361.051
373.7
395.254
304.4
357.8
510.5
536.1
248.6
354.58
265.1
253.65
766.5
670.8
279.1
279.0
317.0


GF
GF
ICP
SAH
SOH
ICP
ICP
GF
GF
ICP
NA
GF
SAA
GF
GF
SAH
ICP
SNA
VC
ICP
ICP
ICP
GF
GF


PDL
PDL
PDL
EDL
HHCL
ICP
HCL
PDL
PDL
PDL
PDL
PDL
ICP
PDL
PDL
Xe300
PDL
EDL
HCL
HCL
HCL
ICP
PDL
PDL


0.002
0.4
0.4
0.1
5.0
0.9
0.8
0.016
0.000018
0.007
500.0
0.002
2.0
0.002
10
0.012
75.0
100.
0.001
3.5
0.3
0.2
0.0001
0.1


a ________________ _________________ 4 ________________ 1 ________________ & _____________


Table 3-2.









Table 3-2--Continued.


1AA = Air-Acetylene Flame
AH = Air-Hydrogen Flame
GF = Graphite Furnace
ICP = Inductively Coupled Plasma
NA = N20 -Acetylene Flame
SAA = Air-Acetylene Flame with Sheath
SAH Air-Hydrogen Flame with Sheath
SOH O2-Hydrogen Flame with Sheath
VC = Vapor Cell
SNA = N20 Acetylene with an Argon Sheath

2PDL = Pulsed Dye Laser System
FLDL = Flashlamp Pumped Dye Laser
EDL = Electrodeless Discharge Lamp
HCL = Hollow Cathode Lamp
HHCL = High-Intensity HCL
Xe300 = Xe Arc Lamp (300W)


Element Excitation Fluorescence Cell' Excitation2 LOD
Wavelength Wavelength Source (ng/ml)
(nm) (nm)
Na 589.0 589.0 GF FLDL 0.0028
Nb 406.11 428.45 ICP PDL 470.0
Ni 322.165 361.939 GF PDL 0.002
Pb 283.306 405.783 GF PDL 0.000005
Pr 406.13 405.65 ICP PDL 240.0
Pt 265.94 271.90 ICP PDL 4.0
Rh 369.3 369.3 ICP HCL 5.0
Ru 287.5 366.3 GF PDL 0.1
Sc 391.181 402.040 NA PDL 10.0
Se 196.03 196.03 SAH EDL 0.02
Si 288.158 251.433 ICP PDL 1.0
Sr 460.7 460.7 AA PDL 0.1
Te 214.3 214.3 SAH EDL 0.08
Ti 307.865 316.257 ICP PDL 1.0
V 268.796 290.882 ICP PDL 3.0
Zn 213.9 213.9 AH EDL 1.0








54
other than 180 to minimize the collection of scatter from the source. The most used angle

is 90 in which the observed fluorescence signal and excitation process are perpendicular

to one another. Only a small portion of the fluorescence signal is collected and detected.

The fluorescence signal is not directional, and it is emitted in a 4vr steradians cone of light.

A photomultiplier tube is the most commonly used detector, although multichannel

solid state detectors have also been used. The signal can be processed using a lock-in

amplifier or a boxcar average depending on the duty cycle of the fluorescence signal.

Laser Enhanced Ionization Spectrometry

Laser-enhanced ionization (LEI) spectrometry has been established over the last 15

years as a trace level analytical technique (1,44,45). LEI as well as AFS rely on radiative

processes for excitation. The difference between the two techniques is the detection

process. In LEI, the amount of charge produced after optical excitation is the parameter

which is determined. The signal appears as an increase in current flow.

After optical excitation, the system then relies on collisions to supply the rest of the

energy needed for ionization. The total ionization rate (dn1 dt) is proportional to the

energy difference between the ionization energy (ER) and the energy of the excited level

(E) accessed through optical excitation according to the following formula


dn a e--r (3-1)
dt



where T is the temperature (K) and K is the Boltzmann constant. The closer the excitation

process places the atom to the ionization continuum, the less energy has to be supplied by








55
collisional processes. Table 3-3 shows LODs for one step excitation LEI for Ga, In, and

TI using different excitation schemes.

The impressive analytical figures of merit (AFOM) of LEI as a technique can be

attributed to several factors. First, the detection process is electrical, and the signal

collection is close to being 100% efficient. The fluctuation in the dC component of the

signal which originates from the background noise is usually low. In an ideal scenario, the

limiting source of noise is due to the signal arising from the ionization which already exists

in the flame.

Second, scatter from the laser, flame emission background, and room lights present

no problem for the detection process since the detection does not rely on optical processes.

Third, LEI is a collisionally assisted process, so, collisional processes do not quench the

signal like in fluorescence but contribute to the LEI signal. The linear dynamic range for

LEI is typically four-to-five orders of magnitude, and the linearity of the calibration curve

is affected by a non-linear collection efficiency.























0 000 fn0
W) SS S 8- s



N C1 Qt 0 Cf Q d




00 o
VI)


0









=S -000 0













CHAPTER IV
EXPERIMENTAL


Preparation of the Resonance Line Lasers

The RLLs cells were made of quartz which will allow the transmission of

radiation in the UV region of the spectrum. GTE Sylvania SG25SC (Friedrick &

Dimmock, Inc., Millville, NJ) clear fused quartz tubing was used for the construction

of the cells. This type of quartz was developed for semiconductor applications, and

its purity is rated to be better than 99.99% silicon dioxide with a low hydroxyl

content. Its softening point is listed as 1835 K, and this type of quartz provides high

stability over a wide range of temperatures. The combination of high purity and low

thermal expansion coefficient makes it suitable for this particular application in

which several temperature conditions and heating rates are involved. The quartz

(model QT-26291) was cylindrical in shape and had an inside diameter of 26.0 mm,

an outside diameter of 29.0 mm, and a wall thickness of 1.5 mm. The quartz was

of high quality to avoid possible reactions of the metal halide salts with any

impurities present in the quartz.

The windows (ESCO Products, Inc., Oak Ridge, NJ) of the cell were made

of S1-UV-fused silica according to ESCO's standard grades. This type of fused silica

is at least 99.7 % silicon dioxide and offers transmittance to 185 nm and below. The








58
transmittance for 193 nm radiation is approximately 87%. The windows were 26.4

mm in diameter and 3.18 mm in thickness.

The metal halide salts used were of high purity (at least 99.99%) and were

obtained from Aesar (Johnson Matthey/Aesar group, Ward Hill, MA). The salts

were packed under an atmosphere of argon. The company provided material safety

data sheets for all the compounds bought. The special protection suggested in the

sheets was followed closely. Metal halide salts can be toxic and, especially for TII,

care was taken to avoid inhalation or skin contact. The use of laboratory coats,

plastic gloves, face mask, and safety goggles is highly recommended.

The salts were handled inside a polyethylene glove bag (Atmos Bag, Aldrich,

Milwaukee, WI) which was placed inside a fume hood. This procedure allowed the

transfer of the salts in a more isolated, controlled, and inert environment. An

analytical balance, a desiccator and accessories were put inside the bag. Special care

was taken to minimize contamination of the salts.

For the construction of the RLLs, the vacuum system shown in Figure 4-1

was designed and constructed. The main feature of the system is that pressures of

10-6 torr can be achieved. The instructions for the appropriate use of the vacuum

system are included in Appendix A.

After the cell was glass blown (Glass Shop, Department of Chemistry,

University of Florida), it was cleaned with chromerge, a mixture of concentrated

sulfuric acid (H2SO4, 17M) and chromium trioxide (Cr03). The cell was rinsed

several times with distilled water, and placed inside an oven to dry at 1200C for at































a)i









'CU


I-




a)

I-

a)


'.
0








U,
















a






CU














































































to

0


.4-








61
least one hour. After cooling down, the empty cell was placed in the vacuum

system; a tube furnace was used to heat up the cell to approximately 600"C as

measured by a K-type thermocouple digital thermometer (model 8528-10, Cole-

Parmer Instrument Company, Chicago, Illinois). The cell was under vacuum for no

less than 24 hours.

After the vacuum bake out procedure, the cell was placed inside the glove

bag. A flow of nitrogen was kept at all times inside the bag, and the flow of

nitrogen was increased every time the bag was opened.

A weighed amount of the salt was placed inside the cell (see Table 5-1). The

RLL cell was connected back to the vacuum system, and it was baked out under

vacuum for another 24 hours. Since the metal halide salts used in this project could

be thermally damaged, the heating temperature had to be monitored more carefully

than when the cell was empty.

Thermal decomposition was not such a big problem for TUl and InI, at least

for the temperatures that could be achieved with the furnace used. The gallium

triioxide, however, thermally decomposed at around 550C (19,20).

The vacuum bake out procedure was a critical step on the successful

preparation of these lasers. It was essential to remove the free halogen, free metal,

impurities, and water as much as possible. Unreacted metal would increase the

population of the metal in the ground state hence decreasing the efficiency of the

inversion mechanism. The free halogen could interfere since it would absorb

radiation from the pump source.








62
After the vacuum bake out of the RLLs which now contained the salt, the

desired amount of buffer gas was also placed inside the RLL, and then the RLL was

sealed. Argon (Research grade, 99.9995%, Alphagaz, La Porte, TX) was used as the

buffer gas.

Since the RLLs have to be heated to obtain the optimum vapor pressure, the

design of the tube furnace requires some discussion. In a regular furnace, the cell

windows are cooler due to heat transfer with the environment. The final oven

design was the equivalent to a three zone temperature furnace. Three furnaces were

positioned in a series, and this arrangement gave separate control of the

temperature in the parts of the cell that were more exposed to the ambient. With

this approach, the problem of the condensation of the metal halide on the cell

windows was solved. Each furnace had a separate control for the temperature

provided by a variable transformer.

Prior to doing experiments, the ability to maintain a given temperature and

temperature homogeneity was tested by placing the thermocouple thermometer at

several places on the furnaces. The temperature of the system could be maintained

within 5C. After a relatively good temperature homogeneity was obtained, two

K-type thermocouples were placed on the system to monitor and control the

temperature. This type of thermocouple was able to measure temperatures in the

range of -250*C to 13750C.

One thermocouple was attached to a digital meter, and the other

thermocouple was connected to a temperature controller (Gardsman by West, model








63

JP). The temperature of the cell was approximately 150C lower in the central

section of the cell compared to the region near the windows. Under these operating

condition, the species reservoir was at the walls of the center of the cells.

Pump Source for the RLLs

For all the experiments carried out for this dissertation, an excimer laser

(model 2110, Questek, Inc., Billerica, MA) using an ArF gas mixture was used as the

pump source. The repetition rate selected was 10 Hz, and the energy per pulse was

40 mJ. The pulse duration was listed to be from 15 to 25 ns according to

manufacturers specifications (14). The pulse-to-pulse stability of this laser is t5%,

typically.

Neutral Density Filters

Neutral density filters were used to attenuate the intensity of the laser beam

when necessary. The optical density (D) that a given filter is able to provide is

defined as


I
D = log,, T = -log1oT

where Io = incident power and

IT = transmitted power.

The transmittance of the filter can be calculated using the following formula

T = 10D (4-1)








64

Neutral density filters can be used in combination to produce higher

attenuation of the laser intensity. Since the optical density of the filters is additive,

the total density (DTot.) is given by


DT = E Di

where Di is the density of only one filter. A filter set (ORIEL Corporation,

Stratford, CT) was used for all the experiments for which calibrated attenuators

were needed. The filter set had filters of the following densities: 0.1, 0.2, 0.3, 0.4,

0.5, 0.6, 0.7, 0.8, 0.9, 1.0, 2.0, and 3.0.

Temperature Optimization of the RLLs

The schematic diagram of the experimental set up used for the optimization

with respect to temperature is shown in Figure 4-2. The beam of the ArF excimer

laser (model 2110, Questek, Inc., Billerica, MA) was focused on the heated cell.

Special care had to be taken not to focus the beam close to the windows of the cell

since the excimer laser beam was energetic enough to drill through the windows.

A quartz spherical lens 50.8 mm in diameter and with a 305.0 mm focal length was

used to focus the laser beam into approximately the middle of the cell. After the

furnace assembly, a colorless glass filter with a sharp cut in the UV (WG-280,

Corion, Holliston, MA) was used to filter the radiation from the excimer laser which

was not absorbed by the RLL.










*B
I-.

0
*0-

U)




0












U,













uII
'.4



a
4) a


a4)
'.4f












































LLh7








67

The output from the RLL was focused on a pyroelectric joulemeter (model

JS-05, Molectron Detector, Incorporated, Portland, OR). This calibrated detector

was able to detect input pulses with pulse widths from ps to 250 Ius and energies in

the nanojoule range. The meter is calibrated for termination on a 50 l resistor

according to manufacturer specifications. The output of the detector was a voltage,

and the peak amplitude of this voltage is proportional to the total energy contained

in the laser pulse. The output voltage was monitored using a digital storage

oscilloscope with a 100 MHz bandwidth (model 2232, Tektronix, Inc., Beaverton,

OR). An actual trace of the output of the energy meter can be seen in Figure 4-3.

The peak voltage was multiplied by the calibration factor (2.46 V/mJ) supplied by

the manufacturer in order to obtain the energy of the laser pulse.

The temperature of the RLL was systematically raised, and after waiting for

an arbitrarily selected time period of 45 minutes to one hour, the energy obtained

from the RLL was measured. The results obtained for the temperature optimization

of GaI3, InI, and TII are shown in Figures 5-1, 5-2, and 5-3, respectively. After this

optimization step, the operating RLL temperature was the one for which maximum

energy was obtained.

The InI and Tll RLLs provided lasing action at two wavelengths. In the case

of the InI RLL, two optical filters were used to distinguish between the 410.2 nm

and 451.1 nm lines of the In atom. The 410.2 nm line was selected with a bandpass

filter (P10-410, Corion, Holliston, MA) with the center of the band of transmitted

wavelengths located at 410 2 nm. The bandwidth of the filter was 10 nm centered












0

0:


0,


09
mr

V



V










.0?




o '










o

Or




































I""
E
..I..


(Aw) epni!IdwV


E jCI








70

around 410 nm and a 45% peak transmittance at 410 nm. The 451.1 nm line was

selected with a long pass filter (LG 450, Corion, Holliston, MA) with a 50%

transmittance at 450 5 nm. At the time of performing this experiment, the TII

RLL lines could not be distinguished with the optical filters available.

Temporal Behavior

The temporal behavior of the RLLs was studied using a fast silicon

photodetector (model ET-200, Electro-Optics Technology, Inc., Fremont, CA) with

a rise time of approximately 200 ns and biased with a 3V battery. This detector

cannot respond to 194 nm radiation The output of the detector was terminated in

a 50 0 resistor, and the voltage developed was measured with a digital scope (model

no. 54503 A, Hewlett Packard, Rockville, MD). This scope has a 500 MHz

bandwidth.

The output from the RLLs was attenuated considerably and did not impinge

directly on the photodiode to avoid permanent damage since the photodiode could

not handle the full power of the laser without destruction. The laser intensity was

attenuated until a peak output voltage of approximately 20 mV was obtained.

Calibrated attenuators were used to decrease the intensity of the laser. The neutral

density filters decreased the laser energy but did not change the temporal behavior

of the RLLs. The pulse width of the laser was determined as the full width at half

maximum (FWHM) of the scope trace.










Spectral Output

The background spectrum of the RLL was taken with a monochromator-

diode array assembly. The output of the RLL was projected on the diode array by

a 0.5 m spectrometer (SPEX Industries, Inc., Metuchen, NJ). The detector head

(model #5122A, Tracor Northern, Middleton, WI) had an image intensifier for low

light level applications and offered a spectral range of 350 nm to 850 nm with peak

sensitivity from 650 nm to 550 nm. The silicon photodiode array has 1024 elements

at 0.025 mm spacing. The data acquisition system used was the Tracor Northern

TN-6500. The output of the RLL was attenuated with neutral density filters before

the detector in order to avoid saturation and possible damage to the diode array.

Estimation of the Upper Value of the Linewidth of the RLLs
by the Absorption in a Metal Vapor Filter Method

Resonance line lasers are classified as line sources. The true spectral profile

of these lasers is complex, and it is characterized by Doppler broadening, collisional

broadening, and hyperfine structure (17,35). The profile can be studied under high

resolution conditions. Since a system of high enough resolution was not available

for a detailed line profile study, the following experiment was designed in order to

make an estimation of the linewidth of the RLLs. The experimental set up is shown

in Figure 4-4.

The experiment was a simple atomic absorption experiment in which a metal

vapor filter was used as the atom reservoir. This atom reservoir was selected


























4






0


-.








0

El.



4)






-Z


U)
U,


U)
i-














Ek
0 W



I.

.2








74
because it had been well characterized, and the width of the absorption profile of

the metal in the filter could be easily estimated using the formula provided by Ingle

and Crouch (1).

An atom cell for each RLL was constructed using the vacuum system

described before. The vapor filter consisted of a qualtz cell with the same

dimensions as the RLL cells. The preparation of the metal vapor filters was carried

out following the same procedure as for the preparation for the RLLs.

A known amount of the metal of interest was placed inside the filter cell

Nitrogen gas (300 torr) was used as a buffer to quench fluorescence. The vapor

filter cell was placed inside a tube furnace (model 55035, Lindberg, Watertown, WI)

which allowed the control of the atomic vapor filter temperature. Two quartz

windows were installed inside the furnace to prevent heat exchange with the

environment and deposition of the metal on the cell windows.

The gallium metal (Apache Chemicals, Inc., Seward, Illinois) was rated to be

99.999% pure. The indium metal (Aldrich Chemical Co., Milwaukee, WI) was

99.99% pure and the thallium metal (Aldrich Chemical Co., Milwaukee, WI) was

99.999% pure.

After increasing the temperature of the metal vapor filter, the transmittance

of the RLL beam was monitored. The absorption of the RLL beam as it travelled

through the atomic vapor filter was monitored using the same monochromator -

diode array assembly that was used to obtain the spectral output of the RLLs. For








75

the InI and Tll RLLs which produce stimulated emission at two wavelengths, the

transmittance of the laser line at each wavelength was monitored.

Standard Solutions

A series of standard solutions were made from a stock standard solution (atomic

absorption standard solution, Aldrich, Milwaukee, WI) by dilution with water obtained

from a Barnstead (Thermolyne Corporation, Dubuque, IA) nanopure water system.

These standard solutions were used for the atomic fluorescence experiments and the LEI

experiments as well.

Laser Enhanced Ionization (LEI)

The experimental set up used for the LEI experiments is shown in Figure 4-5

and the experimental conditions are shown in Table 4-1. A homemade water cooled

stainless steel electrode (7 cm long, 0.64 wide) was placed inside the flame. Since this

electrode served as the cathode, it was biased negatively using a high voltage power

supply (model 412B). The power supply was connected to a voltage regulator (model

MCR 1000, Sola, Elk Grove Village, IL).

Radio frequency pick up was a major source of noise in these experiments. The

5 cm slot burner head (Perkin-Elmer, Norwalk, CT) was shielded by a metal box (15

cm x 12 cm x 10 cm). The box had a rectangular orifice (6.6 cm x 1.5 cm) to allow

the flame to go through. The RLL beam was focused 10 mm below the electrode. The

electrode was placed 12 mm from the box, and the burner head was 3 mm below the

border of the box.









o"








4 1
4) I



II-



SII



* M
Si
ad


A1


0f-





. II
Q's

.a *













E
(So


EL
U (L
u cn
x (
LU --










Table 4-1. Conditions for the LEI Experiments.


Ga In In TI


Range of
standard
solution
concentrations
(Pg/mL)

Electrode voltage
(kV)

Flame
composition

Combustible flow
(L/min)

Air auxilliary
flow (L/min)

Nebulizing air
flow (L/min)

Boxcar settings

Delay (ps)

Gatewidth (pls)

# samples


0.202-101



1


air/acetylene


1.5


3.5


5.3




1.02

1

30


0.08-20



1


air/acetylene


1.5


3


5.5




0.6

0.4

100


0.032-40



1


hydrogen/air


5


6


5.7




0.4

0.6

10 or 100


0.099-20



1


hydrogen/air


5


6


6




0.270

2.5

30








79
The burner head served as the anode, and it was grounded through a 10 kfl

resistor. The resistor was used as a current to voltage converter. As the name of

the technique implies, the phenomena observed is an enhancement of a process

which is already occurring in the flame; a capacitor (220 pF, 5 kV) was placed prior

to the amplifier to filter out the DC component of the signal. An amplifier (model

Al, Thorn EMI, Gencom Inc., Fairfield, NY) was used prior to the boxcar average.

An amplification factor of 106 V/A was used for all LEI experiments because it gave

the best signal to noise ratio (S/N). The electrical connections between the

amplifier, the resistor, and the capacitor were made as short as possible to avoid RF

pick-up.

The signal from the amplifier was fed into the boxcar average (model SR

250, Stanford Research System, Palo Alto, CA) which was optically triggered using

a photodiode (FND-100Q, EG&G Judson, Montgomeryville, PA). The triggering

circuit is shown in Figure 4-6. The same triggering system was used for the atomic

fluorescence experiments.

The output from the boxcar was sent to the computer interface (model SR

245, Stanford Research System, Palo Alto, CA), and the signal was stored and

processed using a personal computer (20 MHz, Club American Technologies, Inc.,).

LEI Signal Dependence on Applied Electrode Voltage

The voltage applied to the electrode was raised in discrete steps, and the LEI

signal obtained when that particular voltage was applied to the flame was monitored.

The conditions for the experiments are shown in Table 4-2.


















































4)
4)
II



U





p2
H

0






,.j.
I-I




81







I




y 6^
jr -a








82
Table 4-2. Conditions for the LEI Signal Versus Applied Electrode Voltage
Experiments.


_Ga In Tl


Standard
concentration
(Pg/mL)



Flame
composition

Combustible flow
(LmAin)

Air auxiliary
flow (L/min)

Nebulizing air
flow (/min)

Boxcar settings

Delay (Is)

Gatewidth (its)

# samples


air/acetylene Iair/acetylene I hydrogen/air


1.5


3.5


5.3




200

2

30


5


5


5.7




270

2.5

30


I _________ I __________









Saturation Curves for LEI

The ability of the Ga and In RLLs to saturate optically the atomic absorption

transition was studied. In this experiment, the intensity of the RLL was

systematically decreased using calibrated attenuators or neutral density filters. The

filter were placed in front of the flame separately or in combinations to obtain the

desired degree of laser attenuation. The LEI signal obtained for a given laser

intensity was determined.

Atomic Fluorescence

The experimental set up for the fluorescence experiments is shown in Figure

4-7. The circular burner head was 10 mm in diameter. This home made burner

head contained a bundle of 75 capillaries (0.60 mm i.d.). An argon sheath

surrounded the flame. The sheath of inert gas reduced the flame instability due to

room drafts.

For the InI and TlI RLLs which have multiwavelength output, optical filters

were used to select the excitation wavelength for the analysis. The output from the

RLL was focused on the flame by a spherical quartz lens. The fluorescence signal

was detected at a 900 angle. The monochromator was placed perpendicular to the

laser beam, and the image of the flame was focused onto the slit of the

monochromator. A fast response time photomultiplier tube was used as a detector.

Calibrated neutral density filters were used to avoid the saturation of the detector.

A resistor was used as current-to-voltage converter. This voltage was sent to a












































0


:-






C,,


Jr.

0.









bO














C)




























o




0 I--


U..

1

FE


Io E
C
O


00)
os w(

<-


I-
0,


0>


- /

-.


a

16-


15
E
0
0








86
boxcar average (model SR250, Stanford Research System, Palo Alto, CA) which

was optically triggered using a photodiode (FND-100Q, EG&G Judson,

Montgomeryville, PA). The light needed for triggering the photodiode was provided

by a quartz beam splitter.

The output from the boxcar was sent to the computer interface (model

SR245, Stanford Research System, Palo Alto, CA), and the signal was stored and

processed using a personal computer (20 MHz, Club American Technologies, Inc.,).

The experimental conditions for the fluorescence experiments are summarized in

Table 4-3.