close

Вход

Забыли?

вход по аккаунту

?

Examination of charge transfer in a helium microwave-induced plasma

код для вставкиСкачать
INFORMATION TO USERS
This manuscript has been reproduced from the microfilm master. UMI
films the text directly from the original or copy submitted. Thus, some
thesis and dissertation copies are in typewriter face, while others may be
from any type of computer printer.
The quality of this reproduction is dependent upon the quality of the
copy submitted. Broken or indistinct print, colored or poor quality
illustrations and photographs, print bleedthrough, substandard margins,
and improper alignment can adversely afreet reproduction.
In the unlikely event that the author did not send UMI a complete
manuscript and there are missing pages, these will be noted. Also, if
unauthorized copyright material had to be removed, a note will indicate
the deletion.
Oversize materials (e.g., maps, drawings, charts) are reproduced by
sectioning the original, beginning at the upper left-hand comer and
continuing from left to right in equal sections with small overlaps. Each
original is also photographed in one exposure and is included in reduced
form at the back of the book.
Photographs included in the original manuscript have been reproduced
xerographically in this copy. Higher quality 6” x 9” black and white
photographic prints are available for any photographs or illustrations
appearing in this copy for an additional charge. Contact UMI directly to
order.
UMI
A Bell & Howell Information Company
300 North Zed) Road, Ann Arbor MI 48106-1346 USA
313/761-4700 800/521-0600
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
ABSTRACT
Name: Patricia Grace Brandi
Department: Chemistry
Title: Examination o f Charge Transfer in a Helium Microwave-Induced Plasma
Major Chemistry
Degree: Doctor o f Philosophy
Approved by:
Date:
ion Director
NORTHERN ILLINOIS UNIVERSITY
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
ABSTRACT
With respect to the types of elements commonly determined by optical emission
spectrometry, helium plasmas possess a flexibility not available with argon-based plasma
sources. The versatility o f helium discharges is due to their ability to produce intense
nonmetal atom and ion emission in the ultraviolet, visible, and near inflared spectral
regions. This emission behavior is interesting since nonmetal ionization potentials are
large, ranging from 10-17 eV.
Evidence suggests that the mechanism responsible for the intense nonmetal ion
emission in helium discharges is charge transfer (CT). Within the focus o f this
dissertation, charge transfer is based upon near resonance energy transfer between ionized
helium and an excited state nonmetal ion. The effect o f the magnitude on the energy
defect (AE), or departure from resonance, must be considered. The CT reaction may be
written:
N ° (0 eV) + He+ (24.58 eV) -► N+* (x eV) + He° (0 eV) + AE (24.58 - x eV)
To gain more information regarding the chemistry of these systems, various energy level
ion populations o f chlorine, phosphorous, sulfur, and carbon in a helium microwaveinduced plasma are examined in this dissertation. These studies were done through the
integrated examination o f ion and atom line emission data, thermodynamic theory, and CT
theory. In addition, the relationship o f the number density o f ionized helium and the
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
population o f nonmetal ion emitting states was examined. This is done by doping the
plasma with argon, which has a lower ionization potential. A wide range o f elements were
examined in order to distinguish between ionization and thermal CT contributions to the
population o f ion electronic states.
The feasibility o f interfacing a helium microwave-induced plasma system to high
performance liquid chromatography (HPLC) with detection in the vacuum ultraviolet
(VUV) spectral region was investigated. By examining atomic emission in the VUV
spectral region, it was hoped that spectral interference from many o f these bands could be
avoided. Lastly, the conversion o f a pulsed 120 Hz - 3 kW - 2.4S GHz generator for
continuous wave output for microwave-induced plasma maintenance was detailed.
Reproduced with permission o f the copyright owner. Further reproduction prohibited without permission.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
NORTHERN ILLINOIS UNIVERSITY
EXAMINATION OF CHARGE TRANSFER IN A HELIUM
MICROWAVE-INDUCED PLASMA
A DISSERTATION SUBMITTED TO THE GRADUATE SCHOOL
IN PARTIAL FULFILLMENT OF THE REQUIREMENTS
FOR THE DEGREE
DOCTOR OF PHILOSOPHY
DEPARTMENT OF CHEMISTRY
BY
PATRICIA GRACE BRANDL
DEKALB, ILLINOIS
AUGUST 1996
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
UMI N um ber: 9703733
UMI Microform 9703733
Copyright 1996, by UMI Company. All rights reserved.
This microform edition is protected against unauthorized
copying under Title 17, United States Code.
UMI
300 North Zeeb Road
Ann Arbor, MI 48103
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Certification:
In accordance with departmental and
Graduate School policies, this
dissertation is accepted in partial
fulfillment o f degree requirements.
ation Director
K / 7/ n
Date
A N Y Llr-£ O c V > T f p . ‘ V
r.^,i j T*i>UD
HEN ■• •-- '
:
•„
V .
-.1
1H€ A'J”: •
it ANY
LNCLNQi-D iN A ’fiN.CAl iOK
^
to
‘ /" - I N ©
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
ACKNOWLEDGEMENTS
I would like to thank m y advisor, Jon W . Carnahan, for his guidance, participation,
and time. I also thank my committee, Professors David S. Ballantine, James E. Ennan,
and Lidia B. Viteflo, for their assisantance in the completion o f this work.
I would like to acknowledge Clarence L. Amow o f the Micro-Now Instrument Co.
(Skokie, IL) for his invaluable assistance in the modification o f the kilowatt-phis
microwave power supply. In addition, I appreciate the assistance of Charles Caldwell,
Larry Gregerson, and Dan Edwards o f Northern Illinois University.
Lastly, I thank my friends and family, without whom this would not be possible.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
DEDICATION
Dedicated with love to my parents, Rosemary and Thomas Brandi
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
TABLE OF CONTENTS
Page
LIST OF T A B L E S .....................................................................................................
viii
LIST OF FIG U R E S.....................................................................................................
ix
Chapter
1. EXAMINATION OF EXCITATION PROCESSES
AVAILABLE TO HELIUM DISCHARGES......................................
1
1.1 Introduction..................................................................................
1
1.2 Development o f the He MIP as an Excitation
Source for Spectroscopy..........................................................
3
1.3 Previous Examinations o f Excitation Processes
in the He M I P .........................................................................
11
1.4 The Charge Transfer Mechanism Postulate and
Experimental Evidence for Nonmetal
F.missirm..................................................................................
17
1.5 Direction of R esearch ...................................................................
24
2. THEORETICAL CONSIDERATIONS RELATING TO
CHARGE TRANSFER IN HELIUM
DISCHARGES ..................................................................................
27
2.1 Introduction..................................................................................
27
2.2 Single-Electron Charge Transfer...................................................
28
2.3 Multiple-Electron Charge Transfer................................................
33
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
vi
Chapter
Page
2.4 Energy Defect C onstraints............................................................
36
2.5 Theoretical Considerations............................................................
38
2.6 C onclusions...................................................................................
39
3. THERMODYNAMIC AND EXPERIMENTAL
EXAMINATIONS OF CHARGE TRANSFER
BETWEEN IONIZED PHOSPHOROUS, SULFUR
AND CHLORINE IN A HELIUM MICROWAVEINDUCED P L A S M A .........................................................................
40
3.1 Introduction...................................................................................
40
3.2 Determination o f Population R a tio s..............................................
41
3.3 Instrum entation.............................................................................
45
3.4 Results and D iscussion.................................................................
52
3.5 Conclusions....................................................................................
67
4. MANIPULATION OF THE CHARGE TRANSFER
PROCESS IN A HELIUM MICROWAVEINDUCED PLA SM A ..........................................................................
68
4.1 Introduction....................................................................................
68
4.2 Effect o f Doping on the Plasma C haracteristics..........................
70
4.3 Instrum entation............................................................................
71
4.4 Results and Discussion...................................................................
72
4.5 C onclusions...................................................................................... 109
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
vii
Chapter
Page
5. PRELIMINARY INVESTIGATIONS OF NONMETAL
DETECTION IN THE VACUUM ULTRAVIOLET
SPECTRAL REGION FOR HIGH PERFORMANCE
LIQUID CHROMATOGRAPHY..........................................................110
5.1 Introduction...................................................................................... 110
5.2 Instrum entation................................................................................114
5.3 Results and Discussion...................................................................... 115
5.4 C onclusions...................................................................................... 128
6. FUTURE CONSIDERATIONS...................................................................129
REFEREN CES............................................................................................................... 131
APPENDIX......................................................................................................................137
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
LIST OF TABLES
Table
Page
1. Number o f Nonmetal Electronic States Available for
One-Electron Charge Transfer with H e ................................................................
30
2. Number o f Electronic States Within Specific Energy
Defects Available for Muhiple-Electron Charge T ra n s fe r...................................
37
3. Energy Levels Available for Charge Transfer Within
± 1.50 eV o f Energy Resonance with Ionized H e liu m ..........................................
60
4. Thermodynamic and Experimental Energy Ratio
Calculation R e s u lts ...............................................................................................
61
5. Cross Section and Overpopulation R esults.............................................................
64
6. Emission Lines Examined in the Argon-Doping Experiments
.............................
74
7. Energy Defects for Single-Electron Charge Transfer Between
Ionized Helium and Argon for Several Elem ents...................................................
75
8. Thermodyanamic and Experimental Energy Ratio
Calculation Results for the N onm etals...................................................................... 107
9. Cross Section and Overpopulation Results for the N onm etals................................ 108
10. Summary o f Calibration Data for the Solvent
Loading E xperim ents............................................................................................... 126
11. Summary o f Detection L im its...................................................................................127
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
LIST OF FIGURES
Figure
Page
1. Electric Field Strength in the TMqjq C a v ity ......................................................
5
2. Tangential Flow Microwave Induced Plasma T o rch ............................................
7
3. Microwave Induced Plasma TMq ^q Resonator C a v ity ......................................
9
4. Charge Transfer Process o f Magnesium with A rgon............................................
19
5. Energy Diagram Chart for Nine N onm etals.........................................................
22
6. Cross Section vs. Energy Defect for Possible
Single-Electron Charge T ransfers........................................................................
31
7. Plot o f Distance vs. Energy for Zinc-Helium C o m p lex es...................................
34
8. Block Diagram o f the VUV S e t-u p .....................................................................
46
9. Detail o f the Plasma Interface...............................................................................
50
10. Partial Energy Level Diagram for Phosphorous...................................................
54
11. Partial Energy Level Diagram for S u lf u r ............................................................
56
12. Partial Energy Level Diagram for C h lo rin e .........................................................
58
13. Effect of Argon-Doping on the Atomic and Ionic Emission
Intensities for M agnesium.....................................................................................
76
14. Effect o f Argon-Doping on the Atomic and Ionic Emission
Intensities for C a lc iu m ........................................................................................
78
15. Effect o f Argon-Doping on the Atomic and Ionic Emission
Intensities for M anganese.....................................................................................
80
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
X
Figure
Page
16. Effect o f Argon-Doping on the Atomic and Ionic Emission
Intensities for Z in c ...............................................................................................
82
17. Effect o f Argon-Doping on the Atomic and Ionic Emission
Intensities for Phosphorous..................................................................................
84
18. Effect o f Argon-Doping on the Atomic and Ionic Emission
Intensities for S u lf u r ............................................................................................
86
19. Effect o f Argon-Doping on the Atomic and Ionic Emission
Intensities for C h lo rin e.........................................................................................
88
20. Effect o f Argon-Doping on the Atomic and Ionic Emission
Intensities for C arbon............................................................................................
90
21. Effect o f Argon-Doping on the Experimental Energy Population
Ratios for Magnesium and C alcium ......................................................................
92
22. Effect o f Argon-Doping on the Experimental Energy Population
Ratios for Manganese and Z in c ............................................................................
94
23. Effect o f Argon-Doping on the Experimental Energy Population
Ratios for Phosphorous, Sulfur, Chlorine, and C a rb o n ......................................
96
24. Partial Spectrum of Carbon in the Vacuum Ultraviolet
Spectral R egion.........................................................................................................101
25. Partial Energy Level Diagram for C arb o n................................................................104
26. Spectrum o f the Pure Helium Discharge in the UV-Vis
Spectral R egion.........................................................................................................112
27. Effect o f Solvent Loading on the Background Emission
in the VUV Spectral R e g io n ...................................................................................116
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
xi
Figure
Page
28. Effect o f Solvent Loading on the Chlorine Emission
Lines in the VUV Spectral R e g io n ......................................................................... 119
29. Calibration Plot for Cl 1 134.724 n m ...................................................................... 121
30. Calibration Plot for Cl I I 107.105 n m ...................................................................... 123
31. Unmodified pw Power Supply Circuit D iagram ...................................................... 140
32. Modified Power Supply D iagram ............................................................................ 143
33. RMS and P-P RSD o f the Modified Power Supply Monitored
at the Forward Power M e te r...................................................................................146
34. KiP-MIP Power T rain............................................................................................... 149
35. Real-Time Output Waveform o f the Modified Generator
and the Photomultiplier Output Signal o f the N 1 149.26 nm
F.missinn T.m e...................................................................................................................... 151
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
CHAPTER 1
EXAMINATION OF EXCITATION PROCESSES AVAILABLE
TO HELIUM DISCHARGES
1.1 Introduction
Plasma-based systems have become standard excitation sources for routine
atomic spectrometric analyses. These discharges are very energetic and are able to
provide high atomization efficiencies, high degrees o f excitation, and low detection
limits for a variety o f elements. In particular, the argon inductively coupled plasma
(ICP), simultaneously introduced by Fassel and coworkers (1, 2) and Greenfield and
coworkers (3) in the 1960s, has had widespread usage. Not only has the ICP been
used with aqueous nebulization techniques, it has also been used in many hyphenated
systems, including inductively coupled plasma mass spectrometry (ICP-MS).
However, with respect to the types o f elements commonly determined by
optical emission spectrometry, helium plasmas possess a flexibility not available with
argon-based plasma sources. Helium-based sources have been used in the
determination and analysis o f metals, nonmetals, and metalloids. The versatility of
helium discharges is due to their ability to produce intense nonmetal atom and ion
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
emission in the ultraviolet (UV) (4), visible (Vis) (5,6) and near infrared (NIR) (7-10)
spectral regions. Helium discharges have been developed as powerful detectors for
compounds separated by gas chromatography (11-18). As well, progress in nonmetal
determinations has been significant in the areas o f aqueous nebulization (19-22), direct
solid sampling (23), liquid chromatography (24,25), and supercritical fluid
chromatography (26-29).
The behavior o f nonmetals in helium discharges, particularly the production of
emitting ionic species, is o f particular interest. Nonmetal emission in argon plasmas is
weak and arises from predominately neutral atoms (30-32). However, in higher
electron density helium plasmas (1 0 ^ - 1 0 ^ e" / cm3), nonmetal emission lines are
often intense. These lines are principally from neutral species for a few elements, but,
the most intense emission for many nonmetals arises from positively charged ions (5, 6,
33-35). This ion emission behavior is interesting since nonmetal ionization potentials
are large, ranging from 10 - 17 eV (36). Further, the upper states o f the observed
nonmetal ion transitions are o f high energy (19 to 29 eV above the ground state).
Populating these high energy ion states is not thermodynamically favorable and high
intensity nonmetal ion emission is not expected.
Unlike the argon inductively coupled plasma (Ar ICP), for which much
mechanSiic information has been gathered (34, 37-51), fundamental examinations of
analytical helium plasmas have not been as extensive. The bulk o f the previous helium
plasma fundamental investigations have focussed on the determination o f plasma
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
parameters such as temperatures and electron densities (52-54). A few papers have
addressed nonmetal ionization and excitation processes (5, 33, 55-58). Because intense
emissions from many nonmetals in helium plasmas arise from nonmetal ions,
mechanistic examinations o f these discharges must include appropriate explanations for
the extensive production o f these high energy species. Comprehensive mechanistic
studies which examine the population of thermodynamically unfavored high energy ion
states o f nonmetals in terms o f their observed ionization phenomenon are not available.
1.2 Development o f the Helium MIP as an Excitation Source for Spectroscopy
An excellent review o f the refinement o f microwave supported discharges is
given in an article by Zander and Hieftje (59). Microwave plasma sources have
historically been operated at a frequency o f2450 MHz due to the availability o f medical
diathermy units. These generators are generally operated at low powers o f 25 to 200
W with helium as the plasma support gas. Coupling between these sources and the
plasma is most commonly done with an electrodeless resonance cavity, although
surface-wave devices (surfatrons) have also been used (60). The transference o f power
from the microwave generator to the plasma requires that the impedances o f the
generator, transfer lines, and the plasma be matched. This can be accomplished
through the use of tuning stubs connected to the cavity or to the transmission line or
through the use of a slotted waveguide tuner.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
The cavity most commonly used is the Beenakker TMq^q resonance cavity.
This device was first described in 1976 (61), with modifications to the original design
being presented in subsequent papers (62-67). The designation TMmn^ refers to the
transverse magnetic modes which are established in the cavity in the angular, radial, and
longitudinal directions, respectively. Therefore, the TM qjq cavity has only one
maximum in the radial direction (Figure 1). In the center o f the cavity, the radial
symmetric standing electromagnetic wave has a maximum The microwave wave
pattern originates from the coupling antenna.
The microwave torch which contains the plasma is positioned axially at the field
maximum The tangential flow torch and the TM q jq cavity used in this work are
shown, in Figures 2 and 3, respectively. The helium plasma support gas is guided
through the threaded teflon insert and forms an annular flow of gas. To enable the
ignition o f a plasma, a source of electrons is introduced into the gas. These initial
electrons are accelerated in the magnetic field and collide with neutral gas atoms,
resulting in collisionally induced ionization. Successive collisions between the electrons
released through ionization and ground state gas atoms result in the production o f
further ionized species. This process continues until a steady state discharge is
maintained; the power absorbed by the plasma is equal to the power lost by the
plasma. Once the plasma is ignited, the described gas flow design maintains the
discharge within the torch away from the walls to protect the quartz tube from the
plasma.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Figure 1. Electric Field Strength in the TMq^q Cavity
Adapted from Reference 61.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
W ////////A A A
*h
V ///////////A
Y/A/A/AAA/AAA/
i
_
Coupling Loop
I&
19
'5
Si
3
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Figure 2. Tangential Flow Microwave Induced Plasma Torch
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
8
Plasma
QtnrtzTube
Tfaaded Teflon Insert
0-Rings
/
He
Ceramic Tube
Teflon Spacer
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Figure 3. Microwave Induced Plasma TMq^q Resonator Cavity
(Top) Axial Cavity View
(Bottom) Side View
Adapted from Reference 67.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
10
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
11
The development o f the helium microwave-induced plasma (He MIP) has
facilitated research in the direction o f routine nonmetal determinations. While much of
the work with microwave-induced plasmas has been done with low to moderate power
systems (50 - 500 W), systems with high power outputs have shown significant
promise. These high-power systems are operated at powers in the range o f 1.2 to 2.2
kW and are called kilowatt-plus microwave induced plasmas (KiP-NDPs). In principle,
the larger plasmas formed by these high power units should allow analyte to undergo a
greater degree o f desolvation, atomization, and excitation. These discharges are
sufficiently robust to allow the direct introduction o f solids (23).
1.3 Previous Examinations o f Excitation Processes in the He MIP
Any complete examination into the production o f the excited ionic nonmetal
species in helium discharges requires that the mechanism explain the population o f high
energy states o f nonmetal ion species. The participants in this mechanism must be of
adequate concentration in the discharge as well as possess sufficient energy to populate
the ionic energy levels of nonmetals. In the helium discharge, eight major reactive
species exist: free electrons (e~), ground state helium atoms (He), excited helium
atoms (He ), excited helium ions (He
+
), excited diatomic metastable helium (H ^ X
+
diatomic helium ions (He2 ), helium ions (He ) and photons. The superscript m
indicates a metastable species and the superscript * indicates an electronically excited
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
12
species. Carnahan and Hieftje present an overview o f these species which examines
their populations and energies (56). Previous mechanistic examinations have focussed
on the role o f several o f these species.
In one o f the first papers on nonmetal excitation, Houpt postulated the
production o f nonmetal excited ions to be a sequential three-step process (33). This
process involved nonmetal ionization by helium dimer ions, neutral metastable dimers,
and atomic metastable species. Subsequent ionization to the nonmetal +2 state was
thought to occur as a result o f charge transfer with monoatomic helium ions. Finally,
nonmetal +2 ion-electron recombination was said to produce electronically excited
singly charged nonmetal ions. The entire process can be described as follows:
Step 1: Ionization to th e+1 state
N
+ He2+ -► N+ + 2 He
N
+ He2m
N+ + 2 He
+
N
+ Hem -c N+ + He +
e
e"
Step 2: Ionization to the +2 state
N+ + He+ -► N2+ + He
Step 3: Recombination
N2+ + e ' -> N+*
For the ionization reactions involving the metastable states (Penning ionization),
the potential energy o f the metastable helium species must be equal to or greater than
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
13
the first ionization energy o f the nomnetal For reactions involving monatomic and
dimer helium ions, near energetic resonance with nonmetal ionization is required.
The overall weakness of this postulate is that it implies sequential reactions
between species o f inherently low concentrations. Because reaction rates are
multiplicative functions o f rate constants (collisional cross sections) and species'
concentrations, each additional sequential step results in a smaller population o f the
product. Second, in only a few cases does dimer helium ion resonance occur for
ionization to the nonmetal ground ion state. Last, the second step o f this scheme relies
upon the collision o f two positively charged species, an event that coulombic repulsion
would probably prevent.
Beenakker proposed a two-step nonmetal ionization-ion excitation process (5).
After nonmetal ionization by a ground state neutral nonmetal collision with a high
energy electron or a metastable helium atom, the ion is excited by potential energy
transfer from diatomic metastable helium This process may be depicted as follows:
Step 1: Ionization
N + Hem -*> N+ + He + e'
N + e ' -► N+ + 2 e”
Step 2: Excitation
N+ + He2m -»• N+* + 2 He
The metastable atom, which has a potential energy o f 19.73 eV, has sufficient energy to
ionize all nonmetals considered. Beenakker speculated that a significant population of
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
14
electrons can have energies as high as the ionization potential of helium (24.58 eV).
This supposition indicates that electron energy distributions do not follow Maxwellian
statistics. Adherence to Maxwellian relationships implies the average electron energy is
0.43 eV at 5000 K. If electron energies are indeed described by this distribution, very
few electrons would have sufficient energy to participate in the described, ionization
process. While Huang et aL (68) observed non-MaxweOian behavior o f electrons using
a helium discharge with a microwave plasma torch, the deviations are not so large as to
produce more than minute populations o f electrons with energies o f a few electron
volts.
Beenakker noted that ion emission is seen for S, P, Cl, Br, and L All o f the
excited states o f the observed nonmetal ion emission lines have energies from 12.3 to
15.9 eV above the ground state nonmetal ion. Therefore, these nonmetal ion excited
states could be populated by exothermic reactions with the metastable molecule.
Additional support for this postulate is found in that the author did not observe ion
emission for C, N, O and F. The lowest excited energy levels o f these ions are greater
than 15.9 eV. Because o f these energy range matches, it was felt that ion excitation
involving the helium metastable molecule folly explained the observed ion emission.
However, like Houpt's scheme, this mechanism dictates sequential collisions o f species
that are o f low populations in the plasma. Again, kinetic considerations make this
scheme unlikely.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
15
Bauer and Skogerboe also proposed that the most favored process leading to
nonmetal ion excitation is sequential ionization and excitation (55). Ionization was
thought to occur by collisions with monatomic and diatomic metastable helium or
electrons. The dominant ionization process was not specified due to lack of knowledge
concerning the rates. In agreement with Beenakker, the most likely second step was
considered to be coOisional excitation with the helium dimer. The entire scheme is as
follows:
Step 1: Ionization
N + Hem -► N+ + He + e"
N + He2m
N + e*
N+ + 2 He + e '
N+ + 2 e"
Step 2: Excitation
N+ + He2m -* N+* + 2 He
Again, the assumption is that sequential reactions involving metastable species
known to be o f low concentration or non-Maxwellian electrons must occur. As
previously stated, basic kinetic arguments eliminate the likelihood o f two-step
metastable reaction. Major deviations from Maxwellian behavior are required to
explain ionization through collisions with electrons. These mechanisms dictate
significant deviations from thermodynamic behavior to ionize elements such as S or CL
Electron temperatures in the range o f 130,000 to 150,000 K are required.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
16
Both Beenakker and Bauer and Skogerboe mentioned, but discounted as
unimportant, charge transfer in hefitun plasmas. Both discussed simultaneous
ionization and excitation by means o f charge transfer, described as fellows:
N + He+ -► N+* + He
N + He2+ -+ N+* + 2 He
For this process to occur, there must be near resonance between the potential energy of
the helium ionic species and the energy needed to simultaneously ionize and excite the
nonmetal atom Beenakker states
helium ions. .
. . no evidence is found for reactions involving
In Bauer and Skogerboe's analysis the energy resonance requirement
was met only for three upper level energy states o f the iodine + 1 ion. Therefore, it
was assumed that this process did not occur to an appreciable extent.
However, in both charge transfer discussions the only nonmetal ion states
considered were those o f the upper energy levels o f the observed nonmetal ion
emission lines. Charge transfer to excited ion states and subsequent thermal population
o f energetically nearby upper levels o f observed transitions were not considered.
Consequently, while their discussions can be used to discount direct excitation to the
upper level o f the observed transition, the examination criteria cannot be used to
eliminate the general charge transfer process between nonmetals and ions of the plasma
gas.
Risby and Talmi commented briefly on nomnetal excitation processes in the
helium plasma, suggesting that the only helium species present in significant number
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
!
I
densities are He, He , He2 , and He (58). They suggested that species such as
He+ ,H e2+ , or He2m will not have appreciable concentrations. However, they did
not comment on ion emission. These statements regarding species concentrations
contradict and discount the previously cited mechanisms. It should be noted that the
spectrum o f He2m has been observed in helium surfatron plasmas, microwave plasma
torches and microwave-induced plasmas (69). Monatomic and dimer helium ions in
helium discharges have been observed by plasma mass spectrometry (70).
1.4 The Charge Transfer Mechanism Postulate and Experimental
Evidence for Nonmetal Emission
Evidence suggests that the mechanism responsible for the intense nonmetal ion
emission in helium discharges is charge transfer (CT). The general CT reaction can be
written:
N ° + G+ -► N+* + G°
where G is the plasma gas and N is the species undergoing CT. The superscripts zero
and plus indicate the atom and ion forms. The asterisk indicates an electronically
excited state. In general, charge transfer is farihated in the case o f near resonance
energy transfer between ionized helium and an ion state o f the species undergoing CT.
Several examinations o f the population of emitting excited ion states have been done
for nonmetals and metals.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
18
Studies which have examined CT in analytical plasmas have been performed
with magnesium as the probe element in an argon inductively coupled plasma (ICP).
Burton and Blades examined CT involving the ion energy states of magnesium which
are in near resonance with ionized argon, the 3d
2
3d
2
2
and 4s S ^ energy
levels (37). An energy level diagram illustrating this process is given in Figure 4. The
applicable energy levels are labeled with their term symbols. The emission o f
magnesium was studied at several RF powers and data was sampled at various heights
above the load coil of the ICP. It was shown that these energy levels o f magnesium
were overpopulated as a result o f charge transfer with ionized argon when examined in
terms of the energy level populations calculated from local thermodynamic equilibrium.
These same energy levels were examined by Farnsworth, Smith, and Omenetto
(71). A two-step laser enhanced ionization process was used to severely deplete the
population of the atomic magnesium ground state energy level Simultaneously, ion
emission lines which originated from energy levels populated by Ar-CT from this
ground state were monitored. Emission from these ion lines decreased significantly as
the ground state was "laser-depopulated." Therefore, it was concluded that charge
exchange from ground state magnesium played an important role in the population of
the emitting ion states. However, magnesium is highly ionized in the ICP (98 %) in the
absence o f CT due to thermal effects alone (72) and the overall effect on magnesium
ionization chemistry is minimal
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Figure 4. Charge Transfer Process o f Magnesium with Argon
Adapted from reference 37.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
20
+o
2
* ------------------Tm,m
2
+*___________ Q
Energy
.
Mg
» •«
Sm
Mg+_--------------- 2S1/2
At
Mg
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
21
Carnahan and Hieflje (56) examined nonmetal ionization and excitation for
chlorine. They noted that all of the states involved in ion emission lie 13.38 - 19.17 eV
above the ground state chlorine ion, or 26.39 - 32.18 eV above the ground state
chlorine atom. The 3s3p
53
PQ Cl II state resides 11.58 eV above the ground state ion,
or 24.59 eV above the ground state atom. (The ionization potential o f Cl I is 13.01
eV.) It was postulated that the chlorine atom undergoes charge transfer
ionization/excitation upon collision with He+. The energies of the species provide near
resonance: He+ has 24.58 eV to give in charge transfer. This reaction with outer
principal quantum number electron configurations may be written:
Cl (3s2 3p5) + He+(ls) -► Cl+* (3 s3 p 5) + He (Is 2).
Excitation from the 3s3p^ Cl II state to the observed 26.39 - 32.18 eV upper energy
states requires only the energies commonly attributed to collisional processes. The
authors went on to describe the populations of the observed states in terms o f
Boltzmann distributions.
To determine if charge transfer is important for other nonmetals, Jones and
Carnahan investigated the energy levels o f several nonmetals as related to the energy of
ionized helium (57). Figure 5 shows the energies o f all the electronic states for nine
singly ionized nonmetals. The energies for the ion states o f each element are plotted
just above the respective symbol. The two horizontal lines show the ionization energies
for argon and helium. Examination of this figure and recalling that the most intense ion
lines are for iodine, bromine, chlorine, sulfur, and phosphorous in helium discharges
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Figure 5. Energy Diagram Chart for Nine Nomnetals
Reprinted with permission from reference 57.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
23
>V
V-J
N
>I)
r*
in
cn
o
o
o
(^3) uioit 3ms punoiS sqi 3ivoq« uot jo X3J3ug
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
24
supports the likelihood that CT is important. The ion states o f these five elements
exhibit good energy matches for CT in helium discharges. None o f the other elements
+
4*
have close energy matches with either Ar or He , with the exception o f the 2s 2p
22
P
state o f carbon and He+. The authors went on to detail states postulated by CT in the
helium discharge to explain the observed spectra.
1.5 Direction o f Research
The ability of helium discharges to produce intense nonmetal atom and ion
emission, an ability not shared with the more commonly used argon-based plasma
system, presents an interesting opportunity. The trace determination o f nonmetals with
a helium-based plasma system can be done with the same ease that the determination of
metals can be done with argon discharges. However, to take full advantage o f the
flexibility of helium discharges, it is necessary that the very behavior which makes them
so useful be understood.
This dissertation examines the production o f excited nonmetal species, a
process which is not thermodynamically favored due to the high ionization potentials of
nonmetals. This was done in terms o f the charge transfer reaction
N° + He+ -► N+* +H e°
where N is a nonmetal species. This process requires that the energy of the ionized
helium be in near resonance with that of the ionized excited nonmetal species.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
25
To examine the contribution o f charge transfer for the overpopulations o f
excited chlorine, phosphorous, and sulfur ion energy levels, an integrated
thermodynamic-experimental procedure was undertaken. Specifically, the
thermodynamically predicted populations were compared to those determined
experimentally and were in turn examined in terms o f charge transfer ionization cross
sections.
Also, the plasma chemistry was manipulated through the addition o f small
amounts o f argon. Since argon has a lower ionization energy than helium (15.76
versus 24.58 eV), the addition o f argon to a helium plasma would cause the
suppression of helium ionization. Since ionized helium is a reactant in the charge
transfer process, the production o f excited ionized species would also be suppressed.
The effect o f argon doping on the emission intensities of several 2A elements, metals
and nonmetals was examined.
Lastly, while examining the vacuum-ultraviolet (VUV) spectral region during
these experiments, it was clear that nonmetal line emission was intense and molecular
emission was minimal. These characteristics invite the examination o f the system for
nonmetal selective detection in nonaqueous media. Initial investigations were
performed to examine the feasibility o f interfacing a high performance liquid
chromatography (HPLC) system to a kilowatt-plus helium microwave-induced plasma
(KiP-MIP) with detection in the VUV spectral region. By examining atomic emission
in the VUV spectral region, it was hoped that spectral interference from intense
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
molecular bands could be avoided. The effect of solvent loading o f the plasma
analytical performance was examined.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
CHAPTER 2
THEORETICAL CONSIDERATIONS RELATING TO CHARGE
TRANSFER IN HELIUM DISCHARGES
2.1 Introduction
Within the focus o f this work, charge transfer is based upon near resonance
energy transfer between ionized hefium and a ground state nonmetal atom to produce a
ground state helium atom and an excited state nonmetal ion. Since exact energy
resonance between ionized helium and the excited nonmetal ion does not occur, the
effect of the magnitude o f the energy defect, or departure from resonance, on the
likelihood o f charge transfer occurrence must be considered. Considering the energy
defect (AE), the nonmetal ionization reaction may be written:
N(OeV) + He+ (24.58 eV) -► N+*(xeV ) + He°(0eV ) + AE (24.58 - x eV)
where x is the energy of the nonmetal ionization phis the energy o f the excited ion
above the ion ground state. However, slight variations in the total energies may be
seen due to the kinetic energies of the colliding particles. The reaction may be slightly
endothermic or slightly exothermic. Theoretical considerations dictate that the reaction
may occur in either case, but with electron configuration restrictions.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
28
la terms o f electronic structure, there are two kinds o f charge transfer (73, 74).
There is charge transfer involving only the transfer o f a single electron without
rearrangement o f the remaining electronic configuration o f the nonmetaL In contrast,
there is charge transfer which involves the transfer of one electron concomitant with
"electronic rearrangement" o f the nonmetaL The latter is designated a muhipleelectron charge transfer. Sections 2.2 and 2.3 detail these processes.
2.2 Single-Electron Charge Transfer
In single-electron charge transfer, an electron is removed through ionization
without change to the remaining electron configuration, as is the following case with
chlorine. In this example, a single electron is removed horn the 3s orbital:
Cl(3s2 3p5 2 P°) + He+ ( l s 2S) -* C l(3s3p53P°) + He (Is2 1S) + 0.01 eV.
Note that the energy defect for this process is very small, 0.01 eV.
Rapp and Francis stated that the ionization cross section for a one electron
charge transfer is maximized when the reaction is near-resonant and slightly exothermic
(73). The cross section can be described:
a
= 3.4435
2 V4
——
a2 u 4
V
(2-1)
where o is the ionization cross sectional area, y is the square root o f the ionization
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
29
energy o f the element to be ionized divided by the ionization energy o f hydrogen, v is
the velocity, a is the Bohr radius, and <■>is the energy defect, AE, divided by (h/2n).
Therefore, there exists an inverse fourth power correlation between the charge transfer
cross section and the energy defect. The cross section is particularly important as the
rate o f reaction is directly proportional to o
1/2
Table 1 lists the elements and the number o f electronic states that are available
for one-electron charge transfer within specific energy defect ranges. In this table,
multiple j values o f the same term state are counted as a single state. The number of
states available for single-electron charge transfer cannot be used as a direct indicator
to predict the contribution o f the single-electron charge transfer process for the
production o f emitting ionic species, but their existence does provide the necessary
pathway. O f these elements, P and Cl obey the conditions for single-electron charge
transfer. The specific states available for single-electron charge transfer for P and Cl
are 3s 3p^ ^S° and 3s 3p^ ^P°, respectively. In Figure 6, a plot o f the log of the
relative cross section as a function of energy defects is shown for possible one-electron
transfers within a large AE range. The relative positioning for the resonant energy
levels in terms o f their energy defects from perfect resonance should serve as a
predictor for the extent o f charge transfer ionization/excitation.
Reproduced with permission o f the copyright owner. Further reproduction prohibited without permission.
30
Table 1
Number o f Nonmetal Electronic States Available for One-Electron Charge
Transfer with He+
LP. (eV)
States within
+0.5 to 0 eV
States within
0 to -0.5 eV
States within
-0.5 to -1.0 eV
c
11.26
p
11.0
—
1
—
Cl
13.01
—
1
—
1
1
—
—
—
K
Br
4.339
11.84
2
1
—
—
This table considers elements 1-57, 72-76 and 78-83.
Data taken from reference 36.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Figure 6. Cross Section vs. Energy Defect for Possible Single-Electron Charge
Transfers
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
32
Relative Cross Section (o)
104 -
Kr
Ar
1
0
1
2
3
4
5
Energy Defect (AE)
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
6
33
2.3 Multiple Charge Transfer
la multiple-electron charge transfer ionization, one electron (or possibly more)
is promoted to a high-energy oibital as, simultaneously, an electron is removed. An
example and qualitative explanation o f this reaction with nonmetals is described below:
S (3s2 3p4 3P) + He+ (Is 2S) -► S+ (3s2 3p2 4s *P) + He (Is2 1S) - 0.56 eV.
As the ground state sulfur and the ionized helium species approach, the orbitals of the
sulfur are perturbed. In effect, the kinetic energies of sulfur and ionized helium are
converted to potential energy o f the orbitals. In terms o f electronic (potential) energy,
this reaction cross section is maximized when the reaction is slightly endothermic with
respect to the potential energies o f the unperturbed species.
To better understand this process, it is useful to consider an example from the
literature (74). Figure 7 is a plot o f intermolecular distance versus potential energy for
the Zn - He* and Zn+*- He complexes. These species play an important role in the
primping mechanism for metal vapor lasers. At large distances, the sum o f the potential
energies o f He+ and Zn° is equal to the ionization energy o f helium. Since the two
species do not interact at large distances, this portion o f their potential energy curves is
essentially flat. As Zn° and He* approach (top curve o f the Zn
state), the potential
energy drops slightly due to a weak interaction between the ionized helium and the
induced polarized zinc species. As the distance between He+ and Zn° decreases, the
sum of the potential energies o f these species becomes equal to that o f the He-Zn
Reproduced with permission o f the copyright owner. Further reproduction prohibited without permission.
Figure 7. Plot o f Distance vs. Energy for Zinc-Helhim Complexes
Adapted from reference 74.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
35
Energy
Za# ( s 0 ) + He+
+ * 2
Zn
( P ) + He
o
3/2
, + (2D ) + Be 0
Zn
5/2
>
Distance
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
36
2
2
4complexes (lower 2 curves for the P and D states o f Zn ). At the intermolecular
+
+*
distance where the energies of the Zn - He and the Zn - He "complexes" are
equivalent, the potential energy curves o f these species "cross" and charge transfer may
+*
occur. In this case, the Zn -He complex is repulsive. Repulsion results in the
separation o f the moieties and the formation o f a neutral helium and an electronically
excited zinc ion.
Table 2 lists the elements and the number o f electronic states available for
charge transfer within given endothermic energy defect ranges. O f these elements,
boron, phosphorous, sulfur, arsenic, selenium, bromine and iodine have ionic electronic
energy level configurations which obey the conditions for two electron charge transfer.
While the number of states available for two-electron charge transfer cannot be used as
a tool for predicting the contribution of the two-electron charge transfer process for the
production o f emitting ionic species, their existence does provide the necessary
pathway.
2.4 Energy Defect Constraints
Differing effective energy defect ranges have been used in previous
examinations o f charge transfer (57, 75, 76). Tumer-Smith, Green and Webb (75)
determined relative cross sections for several metal atom / noble gas ion reactions in a
low pressure microwave induced gas discharge. The largest reaction rates were found
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
37
Table 2
Number o f Electronic States Within Specific Energy Defects Available for MultipleElectron Charge Transfer
Element
LP. (eV)
Be
B
P
S
K
Cu
Zn
Ga
As
Se
Br
Rb
Ag
Cd
Te
I
Xe
9.320
8.296
11.0
10.357
4.339
7.724
9.391
6.00
9.81
9.75
11.84
4.176
7.574
8.991
9.01
10.454
12.127
6.106
n
0 to -0.5 eV
3
2
6
2
—
5
1
8
—
2
1
1
2
4
2
4
1
6
-0.5 to -1.0 eV
1
—
4
2
1
7
4
—
1
2
1
—
6
7
2
4
2
3
This table considers elements 1-57, 72-76 and 78-83.
Data taken from reference 36.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
38
for metal ion energy levels within 0.1 to 0.4 eV o f the ionization potential o f the noble
gas reactant, hi further work by Green and Webb (76), large cross sections were
calculated for processes with energy defects up to 2 eV. Jones and Carnahan (57)
showed that nonmetal ionization in helium discharges correlated well with an energy
defect of £ 0.3 eV. In all o f these studies, endothermic versus exothermic charge
transfer was not considered.
It should be noted that the energy defect refers to the difference in potential
energy existing between two species at rest. However, at a temperature o f 5000 K, the
average energy of particles in a discharge is 0.43 eV with a velocity profile in the form
o f a Maxwellian kinetic energy distribution. As this energy is not accounted for in the
calculation o f o, the cross section for individual particles may only be estimated.
However, the calculated cross section serves satisfactorily for initial approximations.
2.5 Theoretical Considerations
A state of thermodynamic equilibrium exists when each instance of energy
transfer is balanced by the reverse process (77). Included in this definition are all forms
of energy transfer as well as all types o f energy. Gaseous plasma systems, including the
argon inductively coupled plasma (Ar ICP) and the helium microwave induced plasma
(He MIP), are not in thermodynamic equilibrium. However, a restricted form of
thermodynamic equilibrium, local thermodynamic equilibrium (LTE) can be achieved in
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
39
conditions where energy transfer occurs primarily by collisions between particles rather
than by the absoiption or emission o f radiation. In this state, several relationships can
be used to characterize these systems for initial approximations: the Maxwell and
Boltzmann equations describe energy distributions and the Saha equation describes the
degree o f ionization (59). While these relationships cannot be taken to be absolute
descriptors of the energetic state o f particles in the helium microwave induced plasma,
they can be used as a first approximation for the examination o f the behavior o f
introduced analyte.
2.6 Conclusions
While charge transfer can occur as a one-electron or a muhiple-electron
process, the exothermic one-electron process is believed to be a more favorable process
and more likely to occur, hi this study, the aim was to examine a likely and
mathematically characterizable pathway towards the thermodynamically unfavorable
population of highly energetic nonmetal ion energy levels. Therefore, the one-electron
charge transfer process was the primary mode examined.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
CHAPTER 3
THERMODYNAMIC AND EXPERIMENTAL EXAMINATIONS OF CHARGE
TRANSFER BETWEEN IONIZED PHOSPHOROUS, SULFUR AND CHLORINE
IN A HELIUM MICROWAVE-INDUCED PLASMA
3.1 Introduction
Charge transfer between monomer helium ions and nonmetals has been
postulated to be the source o f high nonmetal ion populations in high electron density
discharges. Due to the high ionization potentials of nonmetals, the thermodynamically
predicted populations of ion energy levels are expected to be exceedingly small.
However, the spectra o f many nonmetals include intense emission lines arising from
positively charged ions. Therefore, the upper ion energy levels involved in these
transitions can be said to be overpopulated with respect to thermodynamic predictions.
hi this study, various energy level ion populations of chlorine, phosphorous and
sulfur in a helium microwave-induced plasma were examined in terms of single electron
charge transfer. This was done through the integrated examination o f ion and atom line
emission data, thermodynamic theory, and charge transfer theory. Energy level
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
41
populations were calculated from thermodynamic theory and compared to those
obtained from experimental observations o f resonance and nonresonance transitions.
The resonance emission lines from phosphorous, sulfur and chlorine were
examined. These transitions involve ground state atom and ion energy levels. To
examine these highly energetic transitions, wavelengths in the vacuum-ultraviolet
(VUV) spectral region from 90 to 200 nm were monitored. Emission lines in the
ultraviolet-visible (UV-Vis) spectral region at wavelengths between 200 and 850 nm
were monitored to examine nonresonance transitions. Monitoring UV-Vis emission
lines allows examinations o f higher energy level populations. Energy level population
calculations were performed on these data as well as on UV-Vis emission data obtained
by Tanabe et al. (78).
3.2 Determination of Population Ratios
Thermodynamic Relationships In the absence o f charge transfer,
thermodynamics should fully describe the populations o f nonmetal ion and atom
electronics states. I f the plasma is not in local thermodynamic equilibrium,
thermodynamic measurements and the resulting implications may be somewhat in error.
However, in most cases, these manipulations should be sufficient for a first
approximation for the behavior o f analyte in the helium discharge. The information
needed for the thermodynamic calculations was taken from energy level transition
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
42
probability tables published by the National Bureau of Standards (36, 79). In this
study, several thermodynamic relationships were used. The Saha equation defines the
fraction o f ionized species:
(3-1)
*xP
where N- is the number density for species i, Ejon is the ionization potential in eV, T is
the temperature in K,
is the statistical factor for species i and Pe- is the electron
density (pressure) in atmospheres. Superscripts +, o and - indicate positively charged,
ground state, and a negatively charged species, respectively. Using this equation and
experimental values for T and P -, the thermodynamically-predicted ion-to-atom ratio
may be calculated.
The Boltzmann equation defines the population ratios o f excited species to the
corresponding ground state:
f Nx°*)
= ®x°* eEt*•o . / tT
^ ^
where E ^ * is the energy difference between the ground state atom and the atom in
some excited state and k is Boltzmann's constant. The superscript * indicates an
excited species. The corresponding equation may be written for excited and ground
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
state ions. The Boltzmann equation may be used in conjunction with the Saha equation
to determine the population ratio for any atom or ion energy level The combination of
equation (3-1) and the appropriate forms o f equation (3-2) allows the calculation o f the
ratio o f excited ions to excited atoms in the specified states.
N
x
'
N X ° * ' calc
(3-3)
r
V
*xr
- (E_o. - E, •)/ kT
N X ° ' Saha 8 X+ 8 X ° *
Line Intensity Relationships To correlate the observed emission to the
population o f the emitting energy level, the following expression can be used:
(3-4)
Xx°* = Ax°* h V * N x* V
where I is the measured intensity of species i, A- is the transition probability o f species
i, h is Planck's constant, v is the frequency and V is the observed volume. Again, an
analogous expression may be written for the ionized species. Combining the equations
for the atomic and ionic species and rearranging yields:
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
44
(3-5)
(
lx°* Ax+# V *
Using this equation with the experimental intensities and the corresponding values for
A- and u •, the ratio o f the ion-to-atom energy level populations may be calculated.
Data Interpretation. In the absence o f secondary mechanisms such as charge
transfer, the calculated thermodynamic excited ion to excited atom population ratios
should be in agreement with those obtained from the experimental measurements
(Equation 3-5) within the errors o f the experiment and the transition probabilities. If
the population ratios determined by experimental means are much greater than those
predicted by thermodynamic theory, the excited ion energy level is said to be
overpopulated. The overpopulation ratio is defined as follows:
Overpopulation Ratio = (N x~ )/(N xo «)exp
(3-6)
(Nx+*)/(Nx o »)calc
Plasma Diagnostics. In the presence o f an electric field, changes in the emission
line width and shape occur. By measuring the change in line width or shape, the
density o f free electrons can be obtained. Atmospheric pressure helium microwaveinduced plasmas maintained in TM q ^q devices have been found to have electron
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
45
densities in the range o f 5 x 1 0 ^ e" / cm^ (53, 54, 80, 81). Using the same plasma
system detailed in this study, an electron density o f 7 x 10
14
3
e" / cm was found by
examining the Stark line broadening o f the hydrogen p line at 486.13 nm (80). A
variation of the B oltzmann equation can be used to determine the excitation
temperature. An excitation temperature o f4600 K was determined through the use of
the nine-line method with iron as the probe element (80). Intermediate values o f
4600 K and 5 x 10
1st
*5
e ' / cm were used for calculations.
Conservation o f Spin. In this study, it was crucial to ensure that the monitored
analyte behavior accurately reflected the populations o f the emitting energy levels. The
effects of the magnetic field used to maintain the microwave induced plasma had to be
kept to a mmiiimm To ensure that this interaction did not occur, transitions between
energy levels that did not involve a change in multiplicity were used
3.3 Instrumentation
Microwave-Tndiiced Plasma The instrumental setup is the same as that utilized
by Alvarado and Carnahan (4). A block diagram o f the system is given in Figure 8.
The kilowatt-plus helium microwave induced plasma system (KiP-MIP) consisted o f a
3.0 kW, 2450 MHz microwave generator (Raytheon Company, C.AS. Division,
Waltham, MA), waveguides for power transmission (Gerling Laboratory, Modesto,
CA), a slotted tuner with a worm drive (Cober Electronics, Stamford, CT), a forward
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Figure 8. Block Diagram o f the VUV Set-up
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Microwave
Generator
IHe
USN
'MT
He
He
Computer
Acquisition.
Chart
Recorder
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
48
power meter (Gerling Laboratory, Modesto, CA), a waveguide to 1.625 in. coaxial
transition, and a 1.625 to 0.875 coaxial reducer (Andrew Corporation, Orland Park,
IL). The helium plasma was formed in a TMq^q resonator cavity with an inner
diameter o f 88.8 mm and a depth of 2 cm. The demountable plasma torch consisted of
a fused silica outer tube (8 mm id ., 10 mm o.d.), a doubly threaded teflon insert (8 mm
o.d., 4 threads / in.) and an alumina sample introduction tube (3 mm o.d., 2 mm id .)
VUV Spectral Region. Light dispersion was accomplished with a McPherson
(Acton, MA) model 234/302 VM 0.2 m focal length vacuum scanning spectrometer
with a holographic concave grating. The monochromator was mounted on a X-Y-Z
translation platform. Radiation was detected with a photomultiplier tube model 1P28
(Hamamatsu, Bridgewater, NJ) biased at -1000 volts in conjunction with a quartz
scintillator window (McPherson, Acton, MA). The active material o f the scintillator is
sodium salicylate. The scintillator window absorbs vacuum ultraviolet radiation and
emits fluorescent photons at a longer wavelength. However, the lack o f a wellcharacterized and sufficiently intense light source prevented direct calibration o f the
detector, hi work by Knapp and Smith (82), the quantum efficiency o f sodium
salicylate was examined in the VUV spectral range. It was found that the relative
quantum yield decreased by approximately 15 % between 165 and 95 nm. This
information was used for detector response normalization.
The plasma-monochromator interface was a water-cooled copper cone
mounted directly on the monochromator. The cone had a 2.5 cm diameter base with a
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
49
1.0 cm height and was constructed with a 9.7 cm diameter base. The sampling orifice
diameter was 2.0 nun This interface was modeled after the one used by Houk and
coworkers for VUV plasma work (83, 84) and inductively coupled plasma-mass
spectrometry. Vacuum ultraviolet radiation is absorbed by lenses, mirrors and filters as
well as by molecular nitrogen and oxygen. To avoid absorption by the latter two, the
spectrometer was purged with helium at a flow o f600 mL/min. A detail o f the plasmamonochromator interface is given in Figure 9. Through the use o f the X-Y-Z
translational platform, the monochromator was positioned so that the plasma impinged
on and flowed around the orifice. The purging helium exited the spectrometer through
the orifice o f the sampling cone, protecting the orifice from degradation.
UV-Vis Spectral Region. Intensities for emission lines in the UV-Vis spectral
region were taken experimentally with the previously described KIP-M3P and from the
tabulations by Tanabe et al. (78). Tanabe and co-workers utilized an atmospheric
pressure low power microwave-induced plasma system operated at 75 W with a helium
gas flow rate o f 0.5 L/min. These relative intensities have been verified numerous times
in many laboratories with TMq jQ-maintained helium discharges.
The kilowatt-plus helium-induced plasma system utilized for the VUV spectral
region examinations was also used with a UV-Vis monochromator. Dispersion was
accomplished with a 0.35 m focal length GCA/McPherson (Acton, MA) Model EU700 scanning monochromator. A 15.0 cm focal length lens was used to focus emission
Reproduced with permission o f the copyright owner. Further reproduction prohibited without permission.
Figure 9. Detail o f the Plasma Interface
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Sampling Cane
Torch
Monochromator
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
52
on the entrance slits. Radiation was detected with a Hamamatsu (Bridgewater, NJ)
photomultiplier tube model 1P28 biased at -900V.
Sample Introduction Sample introduction was accomplished with an Ultra 500
ultrasonic nebulizer with desolvation (CETAC Technologies, Ames, IA) operated with
a helium flow rate o f 1.0 L/min. The transducer frequency was 1.36 MHz, the
desolvation heater temperature was 180°C, and the condensor temperature was -5 °C.
The sample solutions o f phosphorous, sulfur and chlorine utilized were prepared with
analytical grade reagents and distilled, deionized water. A concentration o f 1000
pg/mL was used.
3.4 Results and Discussion
Data Summary. The populations o f discrete energy levels were examined from
the intensities o f emission lines in the VUV and UV-Vis spectral regions. Specifically,
emissions in the VUV spectral region were examined to monitor the behavior of the
intense resonance lines arising from phosphorous, sulfur and chlorine. The examination
of resonance lines, arising from transitions involving a ground energy state, allows the
calculation of excited ion/atom energy level population ratios from Equation 3-5. The
UV-Vis spectral region was probed to examine the populations of higher energy ion
and atom energy levels.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
53
Energy level diagrams for phosphorous, sulfur and chlorine are given in Figures
10-12. Included in these diagrams are the energy levels for several o f the transitions
in the VUV and the UV-Vis spectral regions, the ground state atom and ion energy
levels for each nonmetal, the energy level o f ground state ionized helium and the level
populated by charge transfer.
In Figure 10, the resonance lines at 177.499 nm (P I) and at 154.229 nm (PII)
are indicated. The energy o f each energy level referenced to ground state atomic He
and P is given to the left o f each level and the electronic configuration is given to the
right. Observed nonresonance lines at 253.565 nm (P I) and at 458.978 nm (P II) are
indicated. The charge transfer energy level, the energy level with the smallest energy
defect, is at 24.16 eV (3s 3p3 3S°).
In Table 3, the ion energy levels within 1.50 eV exothermic and endothermic of
energy resonance with ionized helium are given. The electron configuration
designation, energy defect (AE) and the type o f charge transfer indicated by the
electronic configuration are given. The ionization cross sections (o) are given for the
single electron charge transfer energy levels as calculated from Equation 2-1. For
phosphorous, there are two energy levels within this range, 3s 3p3 3S° (AE = - 0.405
eV) and 3s 3p3 lP° (AE = -1.35 eV).
The results from the thermodynamic and the emission line intensity calculations
are given in Table 4. The thermodynamically determined ion-to-atom energy level
+* tO*
populations, [N
/h r ]cajc, were calculated for each element using Equations 3-3.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Figure 10. Partial Energy Level Diagram, for Phosphorous
The energy o f each level is referenced to ground state helium and
phosphorous. These energy values are given to the left o f each level and
the electronic configuration is given to the right.
Reproduced with permission o f the copyright owner. Further reproduction prohibited without permission.
3s2 3p 4d b °
26.02 eV
24.56.eV______________________________ yf. _45R.978 nm____ fimnnd S tnteJfcioa----23-32 eV ------1------------------ 24.16 eV
3s23p4p T
18.62 eV
154.229 mn
7.00 eV _
, 2 , 2,. S ,
3s 3p 4s P
7.23eV .
253.565 nm
177.499 nm
, 2 , 3 4_o
OeV
3s 3p S
3s23p24s h
2.33 >v
T 3^3p32P
ftnw l s««4» A«n«n___
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Figure 11. Partial Energy Level Diagram for Sulfur
The energy o f each level is referenced to ground state helium and
sulfur. These energy values are given to the left of each level and
the electronic configuration is given to the right.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
3s23p24 p V
^ 0 . . » CV
2 4 i6 iV ______________________________ W
24 07 «V
545.381 n m ______ fltn m ri
T T T
3 ^ 3 p 4 sT
S ta te Tfe T o n ___
23.49 eV _________________
3s3p42F
125.953 nm
flw w iil Stn«» -t-1 Ton_____
9.22 eV
6.87 eV
131.657 nm
OeV
O nrm H
Hunt, A t«m
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Figure 12. Partial Energy Level Diagram for Chlorine
The energy o f each level is referenced to ground state helium and
chlorine. These energy values are given to the left o f each level and
the electronic configuration is given to the right.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
29.00 eV
26.41 eV
5-»
24.56 eV
3s 3p *P
3^ 3p3 4 p sP
i
479.454 am
V
o
GtomkLStatejje Ion_
3s 3p53P
107.105 om
13.0Le.V_
V 3^3p<3P
jGmnnd .StMfc±liap.____
11.96 eV
3s2 3p3 5p h '
452.621 nm
3s2 3p4 4s 2P
9.22 eV. T
9.22 eV
3s23p *4s ¥
134.724 nm
OeV.
jSmunLSttte Alum______
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
60
Table 3
Energy Levels Available for Charge Transfer Within ± 1.50 eV of Energy Resonance with Ionized Helium
<y(m2)
Species
Designation
AE(eV)
Type
PD
+1.39
+1.32
+1.19
-0.33
-0.405
-0.66
-0.67
-0.72
-0.96
-0.97
-1.02
-1.18
-1.20
-1.25
-1.35
Multiple
Multiple
Multiple
Multiple
pn
pn
3s2 3p 4d 3F°
3s 3p5s P
3s 3p 5s P
3s 3p 4p S
-i i 3 3co
3s 3p S
3s2 3p 4p !P
3s 3p3d *P
3s 3p4p D
3s 3p 4p J S
3s 3p3d D
3s 3p 4p P .
3s 3p 3d P
3s 3p 3d D
3s 3p4p D
, , 3 lno
3s 3p P
Single
Multiple
Multiple
Multiple
Multiple
Multiple
Multiple
Multiple
Multiple
Multiple
Single
225 x 10*28
sn
sn
sn
sn
sn
sn
sn
sn
sn
sn
sn
3s2 3p24p 2S°
3s2 3s2 4s’
3s2 3p2 3d
3s2 3p2 3d \
3s2 3p2 3d 4P
3s2 3p2 3d ^
3s2 3p2 3d 4D
3s2 3p2 4s "H*
3s2 3p2 3d 4F
3s2 3p2 4s 4P
3s3p4 h
+1.33
+0.84
+0.56
+0.42
+0.13
+0.01
-0.07
-0.16
-0.50
-0.56
-1.08
Multiple
Multiple
Multiple
Multiple
Multiple
Multiple
Multiple
Multiple
Multiple
Multiple
Single
5.87 x 10'28
cm
3s3p5 3P°
+0.00652
Single
5.10x10
pn
PH
pn
PD
pn
pn
pn
PH
pn
pn
pn
PH
M
<
A
A
/x
«
0
-X
1
*
1
<3
2.81 x 10"26
Reproduced with permission o f the copyright owner. Further reproduction prohibited without permission.
-19
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Table 4
Thermodynamic and Experimental Energy Ratio Calculation Results
Species
X.nm
Transition
[N ^/N 0* ] ^
Cl 11
Cl I
107.105
134.724
3p _ 3p
2 p .2 [)
4.72 x 10'11
Cl 11
Cl I
JCIII
479.454
452.621
479.454
452.621
5s° - 3p
2po _ 2po
5 s°- 5P
2po. 2po
tP ii
tP i
154.229
177.499
458.978
253.565
458.978
253.565
3p . 3qo
3go . 4p
3 p . 3qo
2po. 2p
3p . 3qo
2po _ 2p
SII
SI
SI
SI
SI
125.95
131.657
140.937
142.510
147.401
4go. 4p
3p . 3qo
3p . 3go
3p . 3qo
3 p . 3qo
if
3.90 x 10‘°
1.29 x io":
6.39 x 10"'
3.09 x 10
te n
PII
pi
ph
pi
[N+*/N**]
[N+*/N°+1exp / [N+*/N°+1calc
8.04
1.70 x 1011
1.23
1.06 x 1012
1.17
1.01 x 10
1.37 x 10
0.924
Z
6.75 x 10
|/
9.91 x 10"1
0.0813
8.21 x 101J
0.0856
8.64 x 10
0.461
0.191
25.9
10.6
1.18 x 10^
1.47 x 10^
4.05 x 10,1
3.42 x IO'
0.0336
to
9.25 x 10*
0.0350
9.64 x 10
1.15x10
-12
1.15 x 10
19
n
9.91 x 10
1
4p . 4[jo
545.388
SII
11
3s° - 3p
3.63 x 10
SI
527.891
4p _ 4[)0
545.388
tsil
1^
3s° - 3p
3.63 x 10
ts i
527.891
| K. Tanabe, H. Haraguchi and K. Fuwa, Spectrochim. Acta 36B, 119 (1981).
19
19
11
in
0\
62
-7
For phosphorous the theoretical energy level population ratio was 1.37 x 10 for the
VUV lines and 9.91 x 1 0 "^ for the UV-Vis transitions. Using Equation 3-5 the
+*
experimental energy level population ratio, [N
tO*
/V r ]
, was determined for the
VUV and the UV-Vis data. This ratio was 0.924 for the VUV lines and was 0.0813
and 0.0856 for the experimental UV-Vis and Tanabe et al. (78) data, respectively. The
experimental energy level population ratios found from the experimental data are
significantly larger than those determined from thermodynamic theory. The
6
13
overpopulation ratios were 6.74 x 10 for the VUV data and 8.20 x 10 and 8.64
x 10
13
for the experimental UV-Vis and Tanabe et al. data, respectively.
In Figure 11, the energy level diagram for sulfor, the resonance lines at 131.657
n m (S I)a n d a t 125.953 nm(S II) as well as the nonresonance lines at 527.891 nm (S
I) and at 545.381 nm (S II) are indicated. The charge transfer energy level is at 23.49
eV (3s 3p4 h ) .
In Table 3, there is one energy level within 1.50 eV of energy resonance with
ionized helium, 3s 3p4 ^F (AE = -1.08 eV). From Table 4, the theoretical energy level
population ratios, [N+ /N° ]
VUV data and was 3.63 x 10
• jjj
[N
13
, ranged from 3.09 x 10"^ to 3.90 x 10'^ for the
for the UV-Vis transitions. The experimental ratio,
£
/N° ]
, ranged from 0.191 to 25.9 for the VUV data and was 0.0336 and
0.0350 for the experimental UV-Vis data and for the Tanabe et al. (78) data,
respectively. The overpopulation o f the emitting ion energy level ranges from 1.18 x
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
63
IO3 to 4.05 x 107 for the VUV data and was 9.25 x 10*® and 9.64 x 10*® for the
experimental UV-Vis data and for the Tanabe et al. data, respectively.
In Figure 12, the energy level diagram for chlorine, the resonance lines at
134.724 nm (Cl I) and at 107.105 nm (Cl II) as well as the nonresonance lines at
452.621 nm (Cl I) and at 479.454 nm are indicated. The charge transfer energy level is
at 24.56 eV (3s 3p5 3P). From Table 3, the energy level within 1.50 eV o f energy
resonance with ionized helium for chlorine was 3s 3p
Table 4, the population ratio, [N+ /N°
P ° (AE = 0.00652 eV). From
was 4.72 x 10”** for the VUV data and
was 1.15 x IO’ 12 for the UV-Vis transitions. The population ratio, [N+ /N ° 3eXp.'
was 8.04 for the VUV data and was 1.23 and 1.17 for the experimental UV-Vis data
and for the Tanabe et aL data, respectively. The oveipopulation o f the emitting ion
energy level was 1.70 x 10** for the VUV data and was 1.06 x 10*^ and 1.01 x 10*7
for the experimental UV-Vis data and for the Tanabe et al. data, respectively.
Data Interpretation. Table 5 sum m arizes the overpopulations and cross section
data for phosphorous, sulfur and chlorine. In all cases, the overpopulations determined
from the UV-Vis data are similar for the three elements (in the range of 10
11
14
to 10 ).
While none o f the UV-Vis emission lines probes the charge transfer energy level
directly, all o f these states are energetically near the charge transfer levels. Apparently,
the energy levels involved in these transitions are overpopulated by collisional
excitation/deexcitation from the charge transfer energy leveL Phosphorous and sulfur
are similar in that the ionization energy of helium lies between the energy levels of the
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
64
Table 5
Cross Section and Overpopulation Results
Element
o 1/2(pm)
a
714
Overpopulation
VUV
1.71 x lO 11
Overpopulation
UV-Vis
1.06 x io}?
1.02x1012
P
0.167
6.75 x 106
8.20 x 10 }3
8 .64 x 10 13
S
0.024
1.18x10*
1.47 x 10*
4.05 x 107
3.42 x 107
9.25 x 10
9 .6 4 x 1 0 10
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
65
observed UV-Vis transition. W hile other factors play a role, there is a correlation of
• 2fi 2
the energy defect (-.405 eV for phosphorous and -1.08 for sulfur), o (2.81 x 10' m
for phosphorous and 5.87 x 10
28
2
m for sulfur), and the overpopulations for the
observed states (~ 1014 for phosphorous and ~ 1011 for sulfur). Smaller energy
defects dictate larger cross sections and overpopulations. The UV-Vis overpopulation
o f chlorine is large (10
12
), but smaller than might be expected when comparing the
values o f the energy defect and o to those o f phosphorous and sulfur. However, the
upper state o f the 479.454 nm emission line Hes 5.44 eV above the charge transfer state
while these energy differences are < 0.8 eV for phosphorous and sulfur. It is likely that
the greater overpopulations with phosphorous and sulfur are a result o f charge transfer
directly into the emission state manifold- However, with chlorine the UV-Vis lines
appear only after charge transfer and thermal excitation to the upper level o f the
emitting state.
For phosphorous and sulfur, the overpopulations determined from VUV data
(=10 ) are three to seven orders o f magnitude smaller than those determined from
UV-Vis data. The overpopulations for chlorine in the two spectral regions were
comparable but slightly higher (5x) in the UV-Vis spectral region. Examinations of the
energy level diagrams show the relative position o f the probed energy levels to the
energy o f the charge transfer level For phosphorous and sulfur, the charge transfer
energy level is not directly probed; the upper states o f the VUV lines are 4.4 - 6.0 eV
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
below the charge transfer level and the electronic states o f the UV-Vis transitions
bracket the charge transfer energy level
That the higher energy states producing UV-Vis emission are overpopulated to
a greater extent than the VUV states is an indication that charge transfer is "feeding"
ions into the "manifold" o f states in the energetic region o f the UV-Vis observable
states for phosphorous and sulfur. The addition o f a few ions to this sparsely populated
upper energy manifold will cause significant overpopulation ratios to be seen. As these
states decay to lower energy levels, this set of ions will also cause overpopulations for
the VUV states. However, because a larger number o f ions will exist in these
resonance states, the overpopulations will be less significant.
However, for chlorine the upper state o f the VUV line (3s 3p^ ^P°) is
populated directly by charge transfer ( AE = 0.00652 eV). Additionally, it should be
noted that the significantly larger value o f o
1/2
for chlorine (o
1/2
q
= 714 pm) should
contribute to charge transfer to an even greater extent. That the charge transfer cross
section is large for chlorine and that the upper state o f the observed emission (107.105
nm) is directly populated by the charge transfer reaction explains why the VUV
overpopulation measured for this line is significantly greater for chlorine.
Reproduced with permission o f the copyright owner. Further reproduction prohibited without permission.
67
3.5 Conclusions
The results o f these examinations are readily explained by charge transfer
theory. The charge transfer process is based upon the near energy resonance between
the energy o f ground state ionized helium and the energy of the ionized excited
nomnetal species. Examination o f Table 5 reveals that the square root o f the ionization
cross section is a predictor o f the overpopulation o f both the higher and lower energy
states, since the overpopulations increase as the ionization cross section increases. The
departure from resonance, AE, is a predictor of the extent o f the direct population of
the ion energy state by charge transfer from the ionized helium.
The high energy states are those which are probed through the UV-Vis spectral
data while the low energy states are examined through the VUV spectral data. From
Table 4, it can be seen that the high energy states are more overpopulated than the low
energy states. The high energy states have a greater degree o f overpopulation because
charge transfer produces excited state ions in this emission manifold. The low energy
levels are overpopulated from deactivation from the higher energy states populated by
the charge transfer process. Exceptions, however, are possible when charge transfer
occurs directly to the upper level o f an observed transition.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
CHAPTER 4
MANIPULATION OF THE CHARGE TRANSFER PROCESS IN A HELIUM
MICROWAVE-INDUCED PLASMA
4.1 Introduction
Iu the previous chapter, a correlation between the energy defect and the
predominance o f the charge transfer process for the population o f emitting ion energy
levels had been found. It was also shown that it was important to consider the energies
o f the emitting ion energy levels relative to the energy level populated by the charge
transfer process. In the case o f chlorine, where the examined VUV ion line originated
directly from the state populated by charge transfer, the overpopulation o f emitting
states was found to be several orders o f magnitude greater than the overpopulations
obtained for sulfur and phosphorous. The probed ion lines o f the latter two elements
originated from energy levels which were populated through collisional deexcitation
from the charge transfer energy leveL
To continue the mechanistic examination, the relationship o f the number density
o f ionized helium and the population o f nonmetal ion emitting states is examined. In
the charge transfer reaction
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
69
N ° + He+ -► N+* + He°
there exists a one-to-one stoichiometry between population o f the ionized excited
species, N+*, and the number density o f ionized helium. If charge transfer with ionized
helium plays an important role in the population o f excited ion energy levels, any
manipulation to suppress the population o f helium ions should lead to a decrease in the
population o f the emitting nonmetal ion states.
A convenient method to change the concentration o f ionized helium in the
plasma is to dope the plasma with small concentrations o f a substance, such as argon,
which has a lower ionization potential. The dopant win have a higher degree of
ionization than the helium plasma support gas and increase the electron density o f the
plasma. These "surplus" electrons can then recombine with the ionized helium,
reducing the concentration o f helium ions. Secondary effects o f the dopant substance
on the characteristics o f the plasma will be examined in the following section.
In the study described in this chapter, the effects o f doping the helium
microwave-induced plasma with argon were examined for a set o f elements whose
ionization potentials and charge transfer energy defects span a wide range o f energies.
The elements under examination included alkaline earth metals, transition metals and
nonmetals. This was done in order to distinguish between ionization and thermal
charge transfer contributions to the population o f ion electronic states.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
70
4.2 Effect o f Doping on the Plasma Characteristics
The effects o f the continual introduction o f a dopant are not limited solely to
changes in the degree o f ionization o f the plasma support gas. Among these effects are
changes in the energy transfer between the power source and the plasma and the
energetics o f the plasma. The energy o f the plasma is dependent not only on the
amount o f power delivered to the plasma, but also on the composition o f the plasma
and the apparatus used to generate and confine the plasma.
The TMqjq resonator cavity has been developed to enable the efficient transfer
of power from the microwave power generator to the plasma. Cavity dimensions and
tuning are typically devised for plasmas composed o f pure helium. When a substance is
introduced into the plasma, whether it be analyte or a dopant gas, the conduction of the
plasma is changed and power transfer between the microwave power generator and the
plasma may be altered as wefi.
In the study outlined in this chapter, the concentrations o f dopant argon added
to the helium plasma support gas were limited to two percent. This was done to avoid
microwave radiation leakage from the discharge and to minimize plasma "detuning"
effects. With the argon dopant concentrations used, no changes in microwave leakage
were noted.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
71
4.3 Instrumentation
The experimental setup is similar to that described previously. The kilowattplus helium microwave induced plasma system (KiP-MIP) consisted o f a 3.0 kW, 24S0
MHz microwave generator converted for the generation o f a continuous wave output
(Cober Electronics, Stamford, CT) and the appropriate components used for
microwave transmission. Details o f the microwave power generator modifications are
given in the Appendix. The helium plasma was formed in a T M q 1Q resonator cavity in
conjunction with a demountable plasma torch.
Detection in the VUV spectral region was accomplished with the same setup
described previously. A McPherson model 234/302 VM 0.2 m focal length vacuum
scanning spectrometer (Acton, MA) with a holographic concave grating was utilized.
Radiation from the scintillator was detected with a photomultiplier tube model 1P28
(Hamamatsu, Bridgewater, NJ) biased at -1000 volts. The spectrometer was purged
with helium before and during analysis. Data was acquired with a 16 bit muitifimction
board (AT-MIO-16X, National Instruments Corporation, Austin, TX).
Sample introduction was accomplished with an Ultra-5000 ultrasonic nebulizer
with desolvation (CETAC Technologies, Omaha, NB) operated with a helium flow rate
o f 1.0 L/min and a desolvation temperature o f 140°C. The condenser temperature was
-5°C. An auto-tuning power supply model ATX-100 (CETAC Technologies, Omaha,
NB) was used to drive the transducer.
Reproduced with permission o f the copyright owner. Further reproduction prohibited without permission.
The sample solutions o f magnesium, calcium, manganese, zinc, phosphorous,
sulfur and chlorine utilized were prepared with analytical grade inorganic salts and
distilled, deionized water. An analyte stock concentration o f 1000 pg/mL was used in
each case. Carbon was introduced into the plasma by means o f direct gaseous
introduction o f dimethyl sulfide.
4.4 Results and Discussion
Data SiimmaTv To compare the effects o f doping the helium microwave-
induced helium plasma on emission behavior, the intensities o f resonance atomic and
ionic emission lines were monitored as argon was added to the plasma. Measurements
were taken at argon concentrations ranging from zero to two percent. Using the
atomic and ionic emission intensities the experimental energy level population ratio,
(N
+ *
tO
*
/h r )
was calculated at each argon concentration using Equation 3-5
(3-5)
The species, wavelengths, and corresponding transitions and transition probabilities
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
73
(A-) which were monitored are given, in Table 6. Only transitions which did not involve
a change in electron spin were used for this study.
The energy defects for the nearest single-electron charge transfer processes
between ionized helium and argon for the dements examined in this study are given in
Table 7. Included in this listing are the first ionization potentials. Energy defects for
magnesnitn., calcium, and manganese with ionized helium could not be calculated
because the sum o f the first two ionization potentials is less than that o f the energy o f
ionized helium.
The results for the doping examinations are given in Figures 13-23. In Figures
13-20, the intensity trends for the individual atomic and ionic emission lines are given.
The data for the alkaline earth elements, calcium and magnesium, are given in Figures
13 and 14. The data for the transition metals, manganese and zinc, are given in Figures
15 and 16. The data for the nonmetals, phosphorous, sulfur, chlorine, and carbon are
given in Figures 17-20. For each figure, the atom and ion line intensities are quoted
relative to the intensities measured in the pure helium discharge. The atom line
emission data are represented by a * and the ion line emission data is represented by a
In Figures 21-23, the experimental energy level population ratio,
(N+ * /ISrO * )
t
, is plotted as a function o f added argon for each examined element.
The data for the alkaline earth elements, calcium and magnesium, are given in Figure
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
74
Table 6
Emission Lines Examined in the Argon-Doping Experiments
Species Wavelength (nm)
tA y C lO -'se c '1)
{Transition
Cl
156.10
1.5
2s2 2p2 3P - 2s 2p3 3D °
cn
133.53
6.0
2s2 2 p 2P ° - 2 s 2 p 2 2 D
Mg I
285.21
4.95
3s2 h - 3s 3p V
M gn
279.79
2.67
S s h - S ? 2? 0
C al
422.67
2.18
4s2 1 S -4 s 4 p 1P°
CaH
393.4
1.48
3s2 3p6 4s 2 S - 3s2 3p6 4p 2P°
Mn I
279.73
3.7
3d5 4s2 6 S -3 d 5 4 s 4 p 6P°
M nn
259.37
2.6
3d5 4s 7S - 3d5 4p 7P °
Z nl
213.86
7.09
3d10 4s2 XS - 3d10 4s 4p l P°
Z nn
202.55
3.3
3d10 4s 2S - 3d10 4p 2P °
PI
177.499
2.16
3s2 3p3 4 S° - 3s2 3p2 4s ^
pn
154.229
0.128
3s2 3p2 3P - 3s 3p3 3D °
SI
147.401
1.6
3s2 3p4 3P - 3p3 4s 3D °
sn
125.953
0.34
3s2 3p3 4S° - 3s 3p4 ^
Cl I
134.724
4.19
3s2 3p5 V * - 3s2 3p4 4s ^
cm
107.105
0.84
3s2 3p4 3P - 3s 3p5 3P °
{Transition information taken from references 36, 79, and 85.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
75
Table 7
Energy Defects for Single-Electron Charge Transfer Between Ionized Helium and
Argon for Several Elements
Element
Ionization
Potential (eV)
Energy Defect with
He+ (eV)
Energy Defect with
Ar+ (eV)
11.26
-1.32
-4.49
Mg
7.65
*
-8.09
Ca
6.11
*
-9.62
Mn
7.43
*
-1.50
Zn
9.39
-7.41
-6.34
P
10.98
-0.405
-2.08
S
10.36
-1.08
-2.33
Cl
13.01
0.00652
-1.27
C
* An energy defect could not be determined. The sum o f the first two ionization
potentials is less than that o f the ionization energy of He .
Reproduced with permission o f the copyright owner. Further reproduction prohibited without permission.
Figure 13. Effect o f Argon-Doping on the Atomic and Ionic Emission Intensities for
Magnesium
The absolute intensity ratio ofthe atom line to the ion line is 0.76:1 with the
pure helium discharge. The atom line emission data is represented by a *
and the ion line emission data is represented by a ■.
Reproduced with permission o fth e copyright owner. Further reproduction prohibited without permission.
77
«n
o'
'st
m
o'
(N
O
o
o
00
VO
CN
/ ^ is u s j u j
Reproduced with permission o fth e copyright owner. Further reproduction prohibited without permission.
o
Ar (L/min)
o'
Figure 14. Effect o f Argon-Doping on the Atomic and Ionic Emission Intensities for
Calcium
The absolute intensity ratio ofthe atom line to the ion line is 0.33:1 with the
pure helium discharge. The atom line emission data is represented by a *
and the ion line emission data is represented by a ■.
Reproduced with permission o fth e copyright owner. Further reproduction prohibited without permission.
Ca II
Ar (L/m in)
79
Aiisusjtq aAijBp'a
Reproduced with permission o fth e copyright owner. Further reproduction prohibited without permission.
Figure IS. Effect o f Argon-Doping on the Atomic and Ionic Emission Intensities for
Manganese
The absolute intensity ratio ofthe atom line to the ion fine is 0.S0:1 with the
pure helium discharge. The atom fine emission data is represented by a *
and the ion fine emission data is represented by a ■.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
81
Ar (L/min)
in
O
<N
O
o
m
CN
o
©
<N
A lisuaiuj aAtjBp-ji
Reproduced with permission o fth e copyright owner. Further reproduction prohibited without permission.
o
Figure 16. Effect o f Argon-Doping on the Atomic and Ionic Emission Intensities for
Zinc
The absolute intensity ratio o f the atom line to the ion line is 2.37:1 with the
pure helium discharge. The atom line emission data is represented by a *
and the ion line emission data is represented by a ■.
Reproduced with permission o fth e copyright owner. Further reproduction prohibited without permission.
83
in
o'
m
o’
Zn II
(N
O
o
o’
XlISU3;UJ 9A pB p-J£
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Ar (L/min)
o'
Figure 17. Effect o f Argon-Doping on the Atomic and Ionic Emission Intensities for
Phosphorous
The absolute intensity ratio o f the atom line to the ion line is 12.39:1 with
the pure helium discharge. The atom line emission data is represented by a
• and the ion line emission data is represented by a ■.
Reproduced with permission o f the copyright owner. Further reproduction prohibited without permission.
85
Ar (L/min)
in
O
<N
O
oo
O
10
o '
rf
<N
O
/^isuajiq 9AqBp"a
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
O
Figure 18. Effect of Argon-Doping on the Atomic and Ionic Emission Intensities for
Sulfur
The absolute intensity ratio o f the atom line to the ion line is 8.24:1 with the
pure helium discharge. The atom Hne emission data is represented by a *
and the ion line emission data is represented by a ■.
Reproduced with permission o f the copyright owner. Further reproduction prohibited without permission.
87
«n
Ar (L/min)
O
CO
1.50
<N
o'
•n
<N
o
o
o
m
>n
(N
AjISUSliq 9AUBp-iJ
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
o
o
Figure 19. Effect of Argon-Doping on the Atomic and Ionic Emission Intensities for
Chlorine
The absolute intensity ratio o f the atom line to the ion line is 7.57:1 with the
pure helium discharge. The atom line emission data is represented by a *
and the ion line emission data is represented by a ■.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
89
m
o'
«r>
<N
O
o
o
o
oo
V">
o'
CN
o'
XjISUSJUI
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
o
Ar (L/min)
O
CN
O
Figure 20. Effect o f Argon-Doping on the Atomic and Ionic Emission Intensities for
Carbon
The absolute intensity ratio o f the atom line to the ion line is 13.2 : 1 with
the pure helium discharge. The atom line emission data is represented by a
• and the ion line emission data is represented by a ■.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
hH
cn
o'
o
o
00
o'
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
o
o'
Ar (L/min)
91
Figure 21. Effect o f Argon-Doping on the Experimental Energy Population Ratios for
Magnesium and Calcium
+* tO*
The ratio o f the experimental energy level population ratios, (N /N )
o f calcium to magnesium is 2.59 with the pure helium discharge.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
,
93
Ar (L/m in)
o
<N
00
o
o
oo
o'
r~
(* oxn / * + xN)
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Figure 22. Effect o f Argon-Doping on the Experimental Energy Population Ratios for
Manganese and Zinc
The ratio o f the experimental energy level population ratios, (N
o f manganese to zinc is 3.06 with the pure helium discharge.
+ *
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
tO
*
fbr )
,
cn
2.00
CN
o
•n
•n
<N
o
o
o
m
o'
( * ° x N ./ * + x N )
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
O
o
A t (L/min)
95
Figure 23. Effect of Argon-Doping on the Experimental Energy Population Ratios for
Phosphorous, Sulfur, Chlorine, and Carbon
The ratio o f the experimental energy level population ratios, (N +* tO*)
o f phosphorous to sulfur to chlorine to carbon is 1.00 : 0.41: 0.33 : 0.18m
the pure helium discharge.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
in
o'
eo
O
GO
o
o
o
ON
m
oo
(*°XN / *+xN)
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
p
Ar (L/min)
o
98
21. The data for the metals, manganese and zinc, are given in Figure 22. The data for
the nonmetals, phosphorous, sulfur, chlorine, and carbon are given in Figure 23.
Data Interpretation. The group 2A elements examined in this study, magnesium
and calcium, undergo a high degree o f thermal ionization in a 4600 K plasma. Using an
electron density o f 5 x 1 0 ^ e~ / cm3 (80), magnesium is over 98% ionized and calcium
is over 99% ionized in a 4600K plasma. Contributions to the population of ion energy
levels from non-thermal mechanisms, including charge transfer, would be minimal.
Unlike elements such as the nonmetals which have high ionization potentials, the
atomic and ionic energy levels of the 2 A elements have appreciable populations
resulting from thermal processes. Populating these levels further by non-thermal
processes, including charge transfer, would change energy level populations by only a
slight degree.
The atomic and ionic emission intensities as a function o f argon doping for
magnesium and calcium are given in Figures 13 and 14, respectively. The intensities o f
both the atom and ion lines decrease, indicating that thermal energy transfer to the
analytes may be reduced by the addition of argon. Examination o f Figure 21 shows
minimal changes in the experimental energy level population ratio, (N
+ *
tO
*
/N^ )
,
across the range o f argon dopant concentrations studied. The energy defects for
magnesium and calcium are very large, -8.09 and -9 .6 2 eV, respectively. However, in
previous examinations o f charge transfer (7 5 ,7 6 ), large relative cross sections had been
determined only for energy defects ranging up to 2 eV. Charge transfer theory cannot
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
99
be used to compare the absolute contribution o f charge transfer with magnesium and
calcium towards the population o f the energy ratios. The slight variations seen in the
experimental energy level population ratios are likely due to changes in the plasma
energetics due to the addition o f the dopant gas.
The atomic and ionic emission intensities as a function o f argon doping for
manganese and zinc are given in Figures IS and 16, respectively. As with calcium and
magnesium, both the ion and atom intensities decrease for zinc with increased argon.
However, ion and atom line intensities both increased for manganese as the fraction of
argon was increased. The increase in the ion line intensity was greater than for the
atom line. Figure 22 shows the atom/ion ratio results o f the argon doping study for
manganese and zinc. The experimental energy level ratio, (N
+* tO*
/N^ )6XIJ, increased
substantially for manganese while the energy level population ratio decreased slightly
for zinc over the range o f argon dopant studied.
From Table 7, the energy defect for zinc with ionized helium is -7.41 eV and
the energy defect for manganese with ionized helium could not be calculated. The
energy defects for manganese and zinc with ionized argon are -1.50 and -6.34 eV,
respectively. The energy defect for manganese with ionized argon is substantially less
than that for zinc. Using an electron density o f 5 x 10
14 -
3
e /cm (80), manganese is
over 98% ionized and zinc is 49% ionized in a 4600K plasma.
A possible explanation for the behavior o f manganese may be found in charge
transfer theory. The relatively small energy defect for manganese with ionized argon
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
100
indicates a possible contribution o f charge transfer with ionized argon
M n° + Ar+ -► Mn+* + Ar°
for the production o f emitting ionic species, Mn
. Although manganese is thermally
ionized to a much greater extent than is zinc, Figure 22 suggests that charge transfer
contributes to the population o f the emitting energy levels. It is possible that the
increase in the ratio (Mn+ / Mn° ) is due to direct (or nearly direct) population of
observed ion states by charge transfer.
The atomic and ionic emission intensities as a function o f argon doping for
phosphorous, sulfur, chlorine, and carbon are given in Figures 17 - 20, respectively.
The atom and ion intensities for phosphorous decreased with increased argon. The
atom intensity for sulfur increased with increased argon. In each case, the ion line
intensities experienced a much greater absolute change than the atom line intensities.
The atom and ion line intensities for chlorine decreased significantly with increased
argon: the ion signal was quenched and could not be observed at 1.1% (0.30 L/min)
argon.
The carbon ion emission line at 133.53 nm has not been documented in the
literature. However, its presence has been suggested by Gislason (86). Its large
8
-1
-8
transition probability (6.0 x 10 sec ) and its corresponding long lifetime (6.0 x 10
sec) imply that the upper energy level o f the 133.53 nm transition (2s 2p
22
D) has an
appreciable population.
In Figure 24 is shown a spectral scan of a portion of the vacuum ultraviolet
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Figure 24. Partial Spectrum of Carbon in the Vacuum Ultraviolet Spectral Region
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
102
N I 141.19
N I 149.3
C l 156.10
0 1 130.4
C H 133.53
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
103
spectral region. The carbon ion emission line at 133.53 nm and the carbon atom
emission line at 156.10 nm are noted. In Figure 25 is given an energy level diagram for
carbon. The transitions at 156.10 nm and at 133.53 nm are noted as well as the energy
levels available for single electron charge transfer.
In Figure 20 is given the results for the doping examination for carbon. The
intensity trends for the individual atomic and ionic lines are given. The line intensities
are quoted relative to the intensities measured in the pure helium discharge. The atom
line intensity for carbon increased with increased argon. The ion line intensity
decreased significantly with increased argon.
The results o f the argon doping study for phosphorous, sulfur, chlorine, and
carbon are given in Figure 23. The energy level population ratio, (N
+* tO*
/N
)
,
decreased slightly for phosphorous (23%) and sulfur (37%) over the range o f a argon
dopant studied. The energy level population ratios for chlorine decreased to a greater
extent, 63%. The energy level population ratio for carbon decreased significantly over
the range of argon dopant studied, 96%. Using an electron density o f 5 x 10
14 e /
3
cm (80), phosphorous, sulfur, chlorine, and carbon are wealdy ionized in a 4600K
plasma, 0.76%, 3.26%, 0.0073%, and 0.85%, respectively.
The one electron charge transfer energy defects for phosphorous, sulfur, and
carbon with ionized helium are -0.405, -1.08, and -1.32 eV, respectively. The energy
defects for these elements with ionized argon are -2.08, -2.33 eV, and -4.49 eV,
respectively. However, the signal for the chlorine ion line (107.105 nm) was quenched
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Figure 25. Partial Energy Level Diagram for Carbon
The energy o f each level is referenced to ground state helium and carbon.
These energy values are given to the left of each level and the electronic
configuration is given to the right
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
24.58 eY.
25.00 eV
,2s2pJ P
23.24 eV
.2s2p2*S
Ground State He Ion
2s2p*b
20.56 eV
133.53 nm
1 0 6 eY________________▼ 2*2P
___ GtDinuLState+1 loa
7.96 eV
156.10 nm
T
2^2p V
Bfminil
Atnm
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
at an argon doping level o f 0.30 L/min (1.1%). The energy defect for chlorine with
ionized helium is exceedingly small, 0.00652 eV, and its energy defect for argon is
larger, -1.27 eV. The upper energy state o f the Cl I I 107.105 nm line, 3s 3p3 3P° is
directly populated by single-electron charge transfer with ionized helium. As the
concentration of argon increases, it appears that the chlorine charge transfer reaction is
circumvented, and the 3s 3p5 3P° level is no longer populated by non-thermal means.
Since the thermal population o f the ionic energy levels o f elements with high ionization
potential is slight, the removal o f a non-thermal pathway quenches ion emission
The results for the thermodynamic and emission line intensity calculations for
the nonmetals are given in Table 8. Analogous to the calculations performed in
Chapter 3, the thermodynamically predicted ion-to-atom energy level populations
(N ^ /N 0*^theo were ca^cu^ate^ f°r carbon using equation 3-3. The experimental
energy level population ratio, (N
/h r )
was determined using equation 3-5. The
values for the thermodynamically predicted populations and the experimental
-8
populations for carbon were 9.23 x 10 and 0.016, respectively. The overpopulation
ratio o f the emitting state for carbon is 1.73 x 103. The value o f the ionization cross
section, o, for the ionic energy levels for carbon, the 2s 2p2 ^P energy level (AE =
0.44 eV) and the 2s 2p2 2S energy level (AE = -1.32 eV) are 1.68 x 10"23 m2 and
1.29 x IQ'27 m2, respectively.
The overpopulations and cross section data for the nonmetals are summarized
in Table 9. The overpopulation determined for carbon is comparable to those
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
107
*
o
5
2
fi*
X
o
*
o
X
o
*
o
"©
X
o
f*
©
00
es
O
*
o
es
o
~©
X
no
" o
X
o
mi
o
mi
cn
CN
—*
r»
1
o
CN
’©
X
CN
r-*
X
in
CN
mm
'©
mm
X
in
mi
mi
* o
X
«n
r*
VO*
CN
ON
o
P'o
X
rJ
00
vO
00
cn
VO
in
00
o
o*
00
o
o
vO
X
cn
mi
~©
X
'©
X
»■«
On
On
vO
mm
'©
X
Ov
ON
o o o o
~©
X X X X
o o r^ w n e s
— ^ < t© T r
—I — ^ cn
X
in
CN
ON
VOON VO
vO
cn
cn
o
o
O O N -
v O v O t^ C ^
'o 'o 'o 'o
•N
^ ^
X X X X
O On O n On
O n CN c n O
cn —’ vo* cn
~©
X
s
ON
o
V
N
cn
o
o
*©
cn
mm
'©
X
cn
vO
cn*
X
cn
vo
cn
cn
o
X
cn
mi
vO
©
o ’
00
'©
X
cn
CN
ON
s
s
0
2
00
On
1
o
03
£
S3
o
e
o
e
es
s
en es i ■ i ■
■ 1o © o o
a , b u m s , ca a ,
en esv sesm es
, 2 l . o 9 . o oi 0 , 2 ' o ,
Q S,
2,
QwQQOo
&»cl,2,0.5,
QtTCks©« *»o■,n
« n n 'f i'SOn
- , es P
eo ,c«i ,eo ,
1o ■© 1o Q ■ I I I I Q IQ 8Li
a , co au 0, 0, a . Ma,a.a.a.a,caa,ca cn
cn
fo m e n ts roes
«
V
.2
CIfa
r^e^o—
-oo
—oo—
*
cn
O
O N V lN T i N
m mm
—ir~ o vO
*o
cn no*
•n-ONKwSKinK
r-^ oi es ©i es
c
CScn^^^*^CN^CN
cn <n
©
c«s
r-~
m
p
'
<n
KiMnmmm
—, —<tt n- tj- ——Ttcs^es
e
U
m t»- tj- — tt —
c
o
S
©s as oo i/i oo m
es ovt'K'Op'VO
es t t o\ m Os m
oo H oo e»i
v*<r>cn^©ooONooov
owoovwn^cnoocnoo
i)
a.
x
Ies
a
U
*©
e
«
o
SC
o
s
es
e
•o
es
S
es
-S'
s
es
e-
4?
Jx :
CO
es
§o>
o
cn
«
<
1
CL
AS
a:
00
fifi
NO
o
o
o
a.
co
—OCJ
U U U U + m
« •—
Q -a « c u eu ? + + +
cnc/i **
U U
H
U
c o co co co co c o co •*■+++
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
108
Table 9
Cross Section and Overpopulation Results for the Nonmetals
Element
o 1/2(pm)
Cl
714
Overpopulation
VUV
Overpopulation
UV-Vis
1.71 x 1011
1.06 x 10 J2
1.02 x 1012
P
0.167
6.75 x 106
8.20 x 10 J3
8 .64 x 10 13
S
0.024
1.18 x
1.47 x
4.05 x
3.42 x
9.25 x 10
9.64 x 1010
c
0.410
1.73 x 106
10^
lO*
10^
107
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
109
determined for phosphorous and sulfur in the vacuum ultraviolet (VUV) spectral
region. Similar to phosphorous and sulfur, the charge transfer energy level o f carbon
(2s 2p^ ^P) is not directly probed by the upper state o f the C II (133.53 nm) line; the
upper state o f the VUV line is 2.66 eV below the charge transfer level. Apparently, the
2s 2p
2^
D ion energy level is overpopulated by collisional excitation/deexcitation from
the charge transfer energy level There is a correlation between the energy defect for
carbon (0.44 eV), o
110
(0.410), and the overpopulation for the observed state (~ 10 ).
4.5 Conclusions
The doping of the helium microwave-induced plasma with argon revealed that
the concentration o f ionized helium affects the population o f ion energy states. In
addition, the magnitude o f the energy defect for single-electron charge transfer also
affects energy state populations. Of particular interest is that the chlorine ion emission
line at 107.105 nm was quenched by the addition o f argon. Also, the behavior o f
carbon in the helium plasma was consistent with charge transfer theory. The
manipulation o f the plasma chemistry in the helium microwave induced plasma further
indicates that charge transfer plays a significant role in analyte excitation for elements
with high ionization energies and small charge transfer energy defects.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
CHAPTER 5
PRELIMINARY INVESTIGATIONS OF NONMETAL DETECTION IN THE
VACUUM ULTRAVIOLET SPECTRAL REGION FOR
HIGH PERFORMANCE LIQUID CHROMATOGRAPHY
5.1 Introduction
While the helium microwave-induced plasma (He-MIP) has been marketed as a
detector for gas chromatography by Hewlett Packard, its use as a detector for high
performance liquid chromatography (HPLC) has not been as extensively studied. As
with other plasma-based systems, the main complication has been the introduction of
mobile phase into the plasma. Experimental approaches to the removal o f the mobile
phase have included the use o f a moving band interface, in which the eluent stream was
directed onto a moving band as a thin film (88, 89). The eluent then passes into a
series of vacuum and heating chambers in which much o f the mobile phase is removed
before thermal vaporization into the plasma. Also, work has been done with membrane
separator systems (90) where the eluent is passed through a membrane system in which
the solvent is selectively removed.
The use o f a plasma-based system as a detector for liquid chromatography
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
I ll
introduces difficulties which are not experienced with detectors which focus on the
bulk properties o f the eluent, e.g. ultraviolet-visible (UV-Vis) absorbance detectors.
Typically the eluent o f a HPLC system is largely organic in composition. The
introduction o f this eluent into the plasma changes the chemistry o f the discharge
significantly. Depression o f the plasma excitation temperature in the argon inductively
coupled plasma (Ar ICP) has been observed (91-93).
Boom and Browner attributed the temperature shift to the absorption o f radio
frequency (if) power by the organic solvents. When introduced by common
nebulization methods, aqueous solutions enter the plasma in the form of wet
particulates and organic solutions enter the plasma as vapor. These highly volatile
organic solvents load the ICP to a greater extent, enhancing their ability to absorb
radiation.
In addition to reducing the excitation temperature o f the discharge, the
introduction o f solvent leads to the production o f intense molecular emission bands.
This difficulty is experienced especially in the ultraviolet visible (UV-Vis) spectral
region. In Figure 26 is shown a spectrum o f a pure helium MIP obtained by Wang and
Carnahan which reveals the presence o f intense molecular emission bands and helium
atomic emission lines in the ultraviolet visible spectral region (94). O f particular note
are the intense bands at 306.4 nm (OH), 336.0 and 337.0 nm (NH) and the intense
helium atom lines at 388.9,402.6, 447.1,492.2, 501.6, and 587.6 nm. The presence of
intense molecular and atomic emission increases the likelihood o f spectral interference.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Figure 26. Spectrum o f the Pure Helium Discharge in the UV-Vis Spectral Region
Reprinted with permission from reference 94.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
O
(l)®H9Z8S
0 9G9S
f
o
CD
(|)®H9‘10S
(l)5R T 2B F S:
(l)HO*fr€fr
♦*N 8 IZf'
(l)OH 920fr\
(l)®H 6*88€ ^ - * l6g
HN 0*£G€
HN 0 9GG
O
o
N LL SG
HO V90Z
Wavelength
(|)®H I’M
(nm)
(l)H L*98fr—
ON (0ZG2 J®PO PU2) O ' W -
o
o
C\J
A i!S U 3 } U |
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
114
In this chapter are presented preliminary investigations into the feasibility o f the
interface of high performance liquid chromatography to the helium microwave-induced
plasma with detection in the vacuum ultraviolet (VUV) spectral region. By examining
atomic emission, in the VUV spectral region, it is hoped that spectral interference from
many o f these bands can be avoided. The effect o f solvent loading on the background
emission in the vacuum ultraviolet (VUV) spectral region is examined. In addition,
quantitative studies o f chlorine emission in a molecular band-free area of the VUV
spectral region are given.
5.2 Instrumentation
The experimental setup is sim ilar to that described previously. The kilowattplus helium microwave-induced plasma system (KiP-MIP) consisted o f a 3.0 kW, 2450
MHz microwave generator (Cober Electronics, Stamford, CT) and the appropriate
components used for microwave tr ansm ission. The helium plasma was formed in a
TMQ10 resonator cavity in conjunction with a demountable plasma torch.
Detection in the VUV spectral region was accomplished with the same setup
described previously. A McPherson model 234/302 VM 0.2 m focal length vacuum
scanning spectrometer (Acton, MA) with a holographic concave grating was utilized.
Radiation from the scintillator was detected with a photomultiplier tube model 1P28
(Hamamatsu, Bridgewater, NJ) biased at -1000 volts. The spectrometer was purged
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
115
with helium before and during analysis. Data was acquired with a 16 bit multifunction
board (AT-MKM6X, National Instruments Corporation, Austin, TX).
Sample introduction was accomplished with an Ultra-5000 ultrasonic nebulizer
with desolvation (CETAC Technologies, Omaha, NB) operated with a helium flow rate
of 1.0 L/min and a desolvation temperature o f 140° C. The condenser temperature
was -5 °C. An auto-tuning power supply model ATX-100 (CETAC Technologies,
Omaha, NB) was used to drive the transducer.
The sample solutions o f chlorine were prepared with an analytical grade
inorganic salt and distilled, deionized water. The effect o f organic solvent loading was
studied through the use o f methanol
5.3 Results and Discussion
The effect o f solvent loading on the background emission in the VUV spectral
region is given in Figure 27. The wavelength is indicated in units o f nanometers at the
bottom o f the figure. Aqueous solutions with methanol concentrations o f 0, 5,10,20,
30, and 35% were introduced into the plasma by means o f the ultrasonic nebulizer.
The intense lines in the VUV spectral region scan obtained with the introduction of the
0 % methanol solution are 113.41 nm (N I), 149.3 nm (N I), 156.0 nm (C I), 174.3 nm
(N I), and 193.1 nm (C I).
With the introduction of the 30 and 35% methanol solutions, the plasma
Reproduced with permission o f the copyright owner. Further reproduction prohibited without permission.
Figure 27. Effect o f Solvent Loading on the Background Emission in the VUV
Spectral Region
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
30 % Methanol
20 % Methanol
10 % Methanol
3 % Methanol
Water
100
120
140
160
110
200
Wavelength (nm)
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
118
became very unstable and adjusting the impedance matching with the slotted waveguide
tuner did not stabilize the discharge. Large emission bands were seen at 130-165 nm
and from 175 nm to longer wavelengths. As tabulated emission information in the
vacuum ultraviolet spectral region is sparse, the identity o f the band at 130-165 nm is
unknown. However, a number o f C I emission lines do reside in that region. The
emission band beginning at 175 nm is due to CO emission (78,95). The formation o f
this species in the discharge is highly favored as the carbon-oxygen bond is the
strongest chemical bond known.
Examination o f these background scans reveals the lack o f molecular emission
at wavelengths shorter than 135 nm. To examine the feasibility o f quantitative
determinations in this area, scans o f the VUV spectral region with the introduction o f
chlorine are given in Figure 28. The wavelength is indicated in units o f nanometers at
the bottom o f the figure. The Cl I line at 134.724 nm is obscured by a molecular
emission band at methanol concentrations greater than 5%.
However, the Cl II line at 107.105 nm is clearly resolved from surrounding
emission peaks. In Figures 29 and 30 are given calibration plots for the Cl 1 134.724
and the Cl I I 107.105 nm emission lines, respectively. The
values are 0.997 and
0.996, respectively. Detection limits are 9.4 and 31 ppm, respectively. A signal-tonoise ratio o f 3 was used for the determination o f the detection limit. A non-linearity
effect (roll-over) was seen at high concentrations. This is primarily due to self­
absorption, a condition where a significant number o f the emitted photons are absorbed
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Figure 28. Effect of Solvent Loading on the Chlorine Emission Lines in the
Spectral Region
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
120
II
1000 ppm Cl
30 % Methanol
1000 ppm Cl
20 % Methanol
\J
1000 ppm Cl
10% Methanol
V
1000 ppm Cl
5 % Methanol
\K
Q I 134.724 nm
cm
107.105 nm
1000 ppm Cl
u1
Water Blank
u
100
110
120
130
140
Wavelength (nm)
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Figure 29. Calibration Plot for Cl 1 134.724 nm
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
122
Signal
0 .0 8
0.07
-
0.06
-
0.05
-
0.04
-
0.03
-
0.02
0.01
-
-
0.00
0
50
100
150
200
250
300
Concentration (ppm)
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
350
Figure 30. Calibration Plot for Cl I I 107.105 nm
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
124
0.025
0.020
-
-
0
-
Signal
0.015
. 0
1
0
0.005
-
0.000
0
50
100
150
200
250
300
Concentration (ppm)
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
350
125
by ground state species existing in the cooler outer portion o f the plasma.
The effect o f solvent loading on analytical performance was examined by
obtaining calibration plots for the Cl 1 134.724 nm emission line at methanol
concentrations o f 0 and 5% and for the Cl I I 107.105 nm emission line at methanol
concentrations o f 0, 5, 10, and 20%. In Table 10 is given a summary o f correlation
coefficients and detection limits for these experiments. While a trend in detection limits
cannot be defined, the variation in detection limits is within the irregularities o f system
noise.
A summary o f detection limits for this work and literature values is given in
Figure 11. The abbreviations FIA and ICP refer to flow injection analysis and
inductively coupled plasma, respectively. The detection limits obtained from this work
are comparable to those obtained by Alvarado (87). However, the detection limits
obtained by Houk (84) using an inductively-coupled plasma are three orders o f
magnitude lower than those obtained by the microwave-induced plasma. In an ICP the
sample is introduced into the central channel o f the plasma while in an MDP the sample
travels on the outer portion of the plasma. This leads to a greater degree o f
desolvation, atomization, and excitation in the inductively-coupled plasma, leading to
lower detection limits.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
126
Table 10. Summary of Calibration Data for the Solvent Loading Experiments
Species
Methanol
Concentration
Correlation
Coefficient (r )
Detection
Limit (ppm)
9.4
Cl I
0%
0.997
cin
0%
0.996
Cl I
5%
0.98
2.9
cm
5%
0.98
11
cm
10%
0.96
18
cm
20%
0.98
30
31
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
127
Table 11. Summary of Detection Limits
Species
Technique
Nebulization
Detection
Limit
Reference
Cl I 134.724 nm
FIA-ICP-VUV
Pneumatic
50ppb
84
Cl I 134.724 nm
FIA-ICP-VUV
Ultrasonic
15ppb
84
C in 107.105 nm
MIP-VUV
Ultrasonic
10 ppm
87
Cl I 134.724 nm
MIP-VUV
Ultrasonic
3 ppm
87
C in 107.105 rnn
MBP-VUV
Ultrasonic
31 ppm
This Work
Cl I 134.724 nm
MIP-VUV
Ultrasonic
9.4 ppm
This Work
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
128
5.4 Conclusions
By examining atomic emission in the far VUV spectral region, spectral
interference from molecular bands can be avoided. However, the main complication to
the interface o f HPLC to plasma-based systems remains. The introduction o f organic
materials to the plasma enhances interfering molecular emission bands. The
development o f efficient schemes which will preferentially remove mobile phase from
eluted components is crucial.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
CHAFFER 6
FUTURE CONSIDERATIONS
In this work the overpopulations o f nonmetal ionic energy levels were examined
in terms o f charge transfer theory. The population, o f these highly energetic states was
examined in terms o f direct population o f the charge transfer energy level as well as
collisional deexcitation to other levels. A useful next step into these examinations
would be to construct more complete atomic and ionic energy level population "maps"
for several nonmetals with varying energy defects. By doing so, one can determine
which energy levels are populated in the helium discharge and gain a more complete
picture o f the charge transfer process.
An important method for the mapping o f energy level populations is through
the use o f a laser to manipulate the populations o f discrete energy levels. Similar to the
work by Farnsworth et al., a pulsed dye laser system can be used to manipulate the
populations of discrete energy levels. In a laser-based atomic fluorescence experiment,
a laser would be aimed through a plasma to sample near its most energetic region to
further excite an excited atomic or ionic energy level. In the reaction
N
*
+ hv -► N
**
*
**
where N and N are excited atomic or ionic species and the Energy^** >
E n erg y ^, transitions originating from the N
**
species can be examined. An
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
130
examination o f information gathered from several different laser-induced population emission events can be used to determine which species contribute to specific ion line
emission events.
hi addition to mechanistic examinations in atmospheric pressure systems,
population measurements in low pressure systems need to be examined. In low
pressure systems, collisional excitation and deexcitation processes are severely reduced,
resulting in population measurements which reflect to a greater extent the outcome of
single-step reactions, including charge transfer. Work in this direction includes work
by Horlick and co-workers in which a Fourier transform UV-Vis spectrometer was
used in conjunction with a glow discharge. In addition, work is currently taking place
to interface a helium microwave-induced plasma (He MIP) with a time-of-flight mass
spectrometer (TOF-MS).
Lastly, the development o f systems for the removal o f organic material from a
liquid stream is necessary to realize the interface o f high performance liquid
chromatography (HPLC) to the helium microwave-induced plasma (MIP). The use of
the MIP as a detector for liquid chromatography would allow for element selective
nonmetal detection, a more informative detection scheme than the more commonly
utilized bulk property detectors.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
REFERENCES
1. V.A. Fassel and R.N. Kniseley, AnaL Chem. 46, 1110A (1974).
2. V.A. Fassel and R.N. Kniseley, AnaL Chem. 46, 1155A (1974).
3. S. Greenfield, H. McD. McGreachin and P.B. Smith, Talanta 22, 553 (1975).
4. J. Alvarado and J.W. Carnahan, AppL Spectrosc. 47, 2036 (1993).
5. C.LM. Beenakker, Spectrochim. Acta 32B, 173 (1977).
6. K. Tanabe, H. Haruguchi and K Fuwa, Spectrochim. Acta 36B, 633 (1981).
7. J.E. Freeman and G.M. Hieftje, AppL Spectrosc. 39, 211 (1985).
8. J.E. Freeman and G.M. Hieftje, Spectrochim. Acta 40B, 653 (1985).
9. D.E. Pivonka, W.G. Fateley and R.C. Fry, AppL Spectrosc. 40, 291 (1986).
10. D.E. Pivonka, A.J.J. Schleisman, W.G. Fateley and R.C. Fry, AppL Spectrosc. 40,
766 (1986).
11. S.A. Estes, P.C. Uden, R.M. Bames, AnaL Chem. 53, 1829 (1981).
12. C. Bradley and J.W. Carnahan, AnaL Chem. 60, 858 (1988).
13. S.R. Goode, B. Chambers and N.P. Buddin, Spectrochim. Acta 40B, 329 (1985).
14. M.A. George, J.P. Hessler and J.W. Carnahan, J. AnaL Atom. Sped. 4, 51 (1989).
15. H. Haraguchi and A. Takatsu, Spectrochim. Acta 42B, 235 (1987).
16. B.D. Quimby and J.J. Sullivan, AnaL Chem 62, 1027 (1990).
17. B.D. Quimby, P.C. Dryden and J.J. SuDivan, AnaL Chem 62, 2509 (1990).
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
18. P.C. Uden, B.D. Quimby, KM . Bames and W.G. Elliott, AnaL Chim. Acta 101, 99
(1978).
19. S.P. Osbourne, Spectrosc. 7,37 (1992).
20. D.A. McGregor, K B . Cull, J.M Gehlhausen, A. S. Viscomi, M Wu, L. Zhang and
J.W. Carnahan, AnaL Chem. 60, 1089A (1989).
21. M. Wu, J.M Gehlhausen and J.W. Carnahan, J. Chem Ed. 69, 757 (1992).
22. S. Chan, H. Tan and A. Montaser, AppL Spectrosc. 43, 92 (1989).
23. J.M Gehlhausen and J.W. Carnahan, Anal. Chem. 6 3 , 2430 (1991).
24. L. Zhang, J.W. Carnahan, R.E. Winans and P.H . Neill, AnaL Chem 63, 212 (1991).
25. K G . Michlewicz and J.W. Carnahan, AnaL Lett. 20, 1193 (1987).
26. C.B. Motley and G.L. Long, AppL Spectrosc. 44, 667 (1990).
27. D.K Luffer, L.J. Galante, P.A. David, M. Novotny and G.M Hieftje, AnaL Chem
60, 1365 (1988).
28. K J. Skelton, Jr., P.B. Farnsworth, K E. Marlddes and M L . Lee, AnaL Chem 61,
1815 (1989).
29. L.J. Galante, M Selby, D.K Luffer, G.M Hieftje and M. Novotny, AnaL Chem 60,
1370 (1988).
30. S.J. Northway, K M Brown and KC. Fry, AppL Spectrosc. 34, 338 (1980).
31. G.F. Kirkbright, A.F. Ward and T.S. West, AnaL Chim Acta. 62 241 (1972).
32. KC. Fry, S.J. Northway, K M Brown and S.K Hughes, AnaL Chem 52 1716
(1980).
33. P.M. Houpt, AnaL Chim Acta. 86 129 (1976).
34. P. Brassem, F.J.M J. Maessen and L. DeGalan, Spectrochim Acta 31B 537 (1976).
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
133
35. C.LM Beenakker, P.W.J.M. Boumans and P J. Rommers, Philips Tech. Rev. 39 65
(1980).
36. C.E. Moore, Atomic Energy Levels as Derived From Optical Spectra, NSRDS-NBS,
VoL 1,2 and 3 U.S. Dept, o f Commerce (1966).
37. L.L. Burton and M W . Blades, Spectromchim. Acta 45B, 139 (1990).
38. J.W. Olesik, Spectrochim Acta 44B, 625 (1989).
39. D.C. Schram, LJ.M.M. Raaymakers, B. van der Sijde, H.J.W. Schendelaars and
P.W .J.M Boumans, Spectrochim Acta 38B, 1545 (1983).
40. G.M. Hieftje, G.D. Rayson and J.W. Olesik, Spectrochim Acta 40B, 167 (1985).
41. P.W .J.M Boumans, Spectrochim Acta 37B, 75 (1982).
42. M W . Blades and G.M Hieftje, Spectrochim Acta 37B, 191 (1982).
43. B.L. Caughlin and MW . Blades, Spectrochim Acta 39B, 1583 (1984).
44. LJ.M M Raaijmakers, P.W .J.M Boumans, B. Van Der Sijde and D.C. Schram,
Spectrochim Acta 38B, 697 (1983).
45. R.J. Lovett, Spectrochim Acta 37B, 969 (1982).
46. A. Goldwasser and J.M Mermet, Spectrochim Acta 41B, 725 (1986).
47. J.A .M Van Der Mullen, LJ.M M . Raaijmakers, A.C.A.P. Van Lammeren, D.C.
Schram, B. Van Der Sijde and H.J.W. Schenkelaars, Spectrochim Acta 42B, 1039
(1987).
48. L.L. Burton and MW . Blades, Spectromchim Acta 46B, 819 (1991).
49. K.P. Li, M Dowling, T. Fogg, T. Yu, K.S. Yeah, J.D. Hwang and J.D. Winefordner,
Anal. Chem 60, 1590 (1988).
50. K.P. Ii, T. Yu, J.D. Hwang, K S. Yeah and J.D. Winefordner, Anal Chem 60, 1599
(1988).
51. P.W .J.M Boumans and F.J. DeBoer, Spectrochim Acta 32B, 365 (1977).
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
134
52. J.M. Workman, P.G. Brown, D.C. Miller, C.J. Seliskar and J.A. Caruso, AppL
Spectrosc. 40, 857 (1986).
53. P.G. Brown, T.J. Brotherton, J.M Workman and J.A. Caruso, AppL Spectrosc. 41,
774 (1987).
54. S.R. Goode, N.P. Buddin, B. Chambers, KW . Baughman and J.P. Deavor,
Spectrochim. Acta 40B , 317 (1985).
55. C.F. Bauer and R.K. Skogerboe, Spectrochim. Acta 38B , 1125 (1983).
56. J.W. Carnahan and G.M. Hieftje, Spectrochim. Acta 47B , 731 (1992).
57. K.J. Jones and J.W. Carnahan, Spectrochim. Acta 47B , 1229 (1992).
58. T.H. Risby and Y. Tahni, CRC Crit. Rev. AnaL Chem. 14,231 (1984).
59. A.T. Zander and G.M Hieftje, AppL Spectrosc. 35, 357 (1981).
60. M Selby and G.M Hieftje, Spectrochim. Acta 42B , 285 (1987).
61. C.LM. Beenakker, Spectrochim. Acta 31B , 483 (1976).
62. C.LM Beenakker, Spectrochim. Acta 33B , 53 (1978).
63. C.LM Beenakker, B. Bosman and P.W.J.M. Boumans, Spectrochim. Acta 33B , 373
(1978).
64. J.P.J. Van Dalen, P.A. DeLezenne Coulander and L. DeGalan, Spectrochim. Acta
33B, 545 (1978).
65. D.L. Haas, J.W. Carnahan and J.A. Caruso, AppL Spectrosc. 3 7 , 82 (1983).
66. M Wu and J.W. Carnahan, AppL Spectrosc. 46, 163 (1992).
67. K B . Cull and J.W. Carnahan, AppL Spectrosc. 4 2 , 1061 (1988).
68. M Huang, D.S. Hanselman, Q. Jin and G.M Hieftje, Spectrochim. Acta 45B , 1339
(1990).
69. J.W. Carnahan, Q. Jin, T. Stam and G.M. Hieftje (unpublished data).
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
135
70. D.M. Chambers, J.W. Carnahan, Q. Jin and G.M Hieftje, Spectrochim. Acta 46B ,
1745 (1991).
71. P.B. Farnsworth, B.W. Smith and N. Qmenetto, Spectrochim. Acta 46B , 843 (1991).
72. R.S. Honk, Anal. Chem. 58, 97A (1986).
73. D. Rapp and W.E. Francis, J. o f Chem. Phys. 3 7 , 2631 (1962).
74. C. Melius, J. Phys. B: Atom. Molec. Phys. 7, 1692 (1974).
75. A R. Tumer-Smith, J.M. Green and C.E. Webb, J. Physics B: Atom. Molec. Phys.
6,114 (1973).
76. J.M Green and C.E. Webb, J. Physics B: Atom. Molec. Phys. 7, 1698 (1974).
77. S. Greenfield, H. McD. Mcgeachin and P.B. Smith, Talanta 22, 1 (1975).
78. K. Tanabe, H. Haraguchi and K Fuwa, Spectrochim. Acta 36B, 119 (1981).
79. W.L. Weise, MW . Smith and B.N. M ies, Atomic Transition Probabilities, NSRDSNBS, VoL II, U.S. Dept, o f Commerce (1969).
80. M Wu, PhD . Dissertation, Northern Illinois University (1990).
81. S.R. Goode and J.P. Deavor, Spectrochim. Acta 39B , 813 (1984).
82. R.A Knapp and A M Smith, AppL Optics 3 , 637 (1964).
83. R.S. Houk, V .A Fassel and B.R. LaFreniere, AppL Spectrosc 40, 94 (1986).
84. B.R. LaFreniere, R.S. Houk and V.A Fassel, AnaL Chem. 59, 2276 (1987).
85. R.C. W east and M J. Astle, CRC Handbook o f Chemistry and Physics, 62nd ed.,
(CRC Press, Boca Raton, FL, 1982).
86. E.A Gislason, University o f Illinois at Chicago, private communication (1993).
87. J. Alvarado, PhD . Dissertation, Northern Illinois University (1993).
88. L. Zhang, PhD . Dissertation, Northern Bllinois University (1990).
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
136
89. P.B. Mason, M S. Thesis, Northern Illinois University (1993).
90. O.T. Akinbo, Northern Illinois University, unpublished data (1996).
91. A.W. Boom and R.F. Browner, AnaLChem 54, 1402 (1982).
92. G. Kreuning and F.J.M J. Maessen, Spectrochim. Acta 44B, 367 (1989).
93. M W . Blades and B.L. Caughlin, Spectrochim. Acta 40B, 579 (1985).
94. Y. Wang and J.W. Carnahan, Anal. Chem. 65, 3290 (1993).
95. R. Mavrodineaux and H. Boitreux, Flame Spectroscopy (John Wiley & Sons, Inc.,
New York, NY, 1965).
96. K.B. Cull and J.W. Carnahan, AppL Spectrosc 42, 1061 (1988).
97. KJB. Cull and J.W. Carnahan, J. Microwave Power and Electromagnetic Energy 24,
151(1989).
98. J.R. Bames, Electronic System Design: Interference and Noise Control Techniques
(Prentice-Hall, Inc., Englewood Cliffs, NJ, 1987).
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
APPENDIX
CONVERSION OF A PULSED 120 Hz - 3 kW - 2.45 GHz GENERATOR
FOR CONTINUOUS WAVE OUTPUT FOR MICROWAVE INDUCED
PLASMA MAINTENANCE
A 1 Introduction
While much of the work with MEPs has been done with low to moderate power
systems (50 - 500 W), systems with high power outputs have shown significant
promise (4, 80, 96, 97). These systems are operated at powers in the range o f 1.2 to
2.2 kW and have been designated kilowatt-plus microwave induced plasmas (KiP MIPs). hi principle, the larger plasmas formed by these high power units should allow
the analyte to undergo a greater degree o f desotvation, atomization and excitation.
These discharges have proven to be sufficiently robust for the direct introduction of
solids (4).
The microwave power source used for plasma formation must be stable,
producing an output waveform with low ripple. The majority o f commercially available
high wattage microwave power supplies are designed for industrial heating
applications, a use which does not require a highly stable source. These high ripple,
Reproduced with permission o f the copyright owner. Further reproduction prohibited without permission.
138
pulsed wave (pw) sources are typically modulated at 120 Hz. While these generators
are capable o f maintaining MIPs, plasmas produced in this pw mode are dominated by
120 Hz noise. Sources used for plasma generation require a higher degree o f stability.
Higher power continuous wave (cw) sources are more desirable for this purpose.
For microwave sources in general, the line voltage is rectified and directed to
the magnetron cathode. The interaction between the cathode and an applied field
results in an oscillating electromagnetic field which exits the magnetron and is sent to
the application load. The waveform o f the input voltage to the magnetron directly
affects the degree o f the output ripple. Therefore, smoothing the waveform entering
the magnetron results in a smoothed output waveform. The input waveform can be
smoothed to the extent that a stable analytical plasma can be formed.
In this section, circuitry modifications are outlined which were made to a pw
2.45 GHz microwave variable power source to form a cw power supply suitable for
analytical plasma generation and maintenance. A Model S-3 (Cober Electronics,
Stamford, CT), 3 kW maximum power output, pw, 2.45 GHz supply was modified.
The stock microwave power source features variable power adjustment, forward and
reflected power meters, a water-cooled magnetron with interlock protection, and dc
overload protection circuitry for the anode power supply. The input waveform
entering the cathode was Ml-wave rectified.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
139
A.2 Discussion
TTnmndified pw Power Supply. A circuit diagram for the unmodified pw Cober
S-3 microwave generator can be seen in Figure 31. Abbreviations in brackets refer to
components in the figure. The 220 V line supply voltage (30 A) was directed to a
silicon-controlled rectifier power controller [1SCR] (Hamar Electronics, Inc.
Columbus, OH.) These silicon-controlled rectifiers operate as switches which conduct
when a gate pulse is received. The switch remains in a closed state as long as a
minimum current is maintained. Due to large ac spikes which can occur, SCR power
controllers can be quite noisy (98). This device controlled the RMS voltage entering a
transformer [2Tran] and the bridge rectifier [1REC].
A second transformer [ITran] is tapped from the line voltage, with the
secondary windings providing connection to d.c. overload protection circuitry, a
second power controller [2SCR], the filament transformer and cabinet door safety and
water coolant interlocks. This second power controller regulated the magnetron
filament voltage, reducing the cathode filament voltage as the anode current increased.
The internal positive <Lc. voltage source o f the 1SCR was connected to a series
o f diodes and in turn, to the manually adjustable microwave power controller. This
controller regulated the magnetron input voltage and the output microwave power.
The output terminals of the 1SCR were connected across a varistor [Var] and a
transformer [2Tran], The secondary windings o f 2Tran were connected to 1REC to
R eprod u ced with permission o f the copyright owner. Further reproduction prohibited without permission.
Figure 31. Unmodified pw Power Supply Circuit Diagram
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
I
a
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
142
provide full-wave rectification. One side o f the bridge was tapped to circuitry which
included a 5H filter choke, five varistors in series [2Var] and connections to the
microwave power metering circuitry. The other side o f the rectifier was tapped to the
magnetron with a 0.01 pF capacitor providing limited wave smoothing. The standby
magnetron filament voltage was 4.6 V. The magnetron output led to a rectangular
waveguide flange, size WR-584.
A helium microwave-induced plasma was maintained using the unmodified
power supply. However, the waveform modulation at 120 Hz caused the plasma to be
reignited and extinguished during each cycle o f the output waveform. The resulting
unstable plasma was audibly loud and unsuitable for analytical analysis.
Modifications to Construct the cw Power Supply. The primary goals o f this
work were to improve the control of the filament voltage and to smooth the supply
voltage to the magnetron. A simplified circuit diagram for the modified Cober S-3
microwave generator can be seen in Figure 32. In the original circuitry, the supply
voltage to the magnetron was controlled by SCR1. Ib is was removed and replaced by
a variac [VAC2]. The 2SCR in the original circuitry regulated the filament voltage
automatically, cutting back the supply voltage when the average anode current reached
550 mA. This SCR power controller was removed and replaced by a variac [VAC1]
which allows the operator to decrease the filament supply voltage manually. After the
unit has been in operation for several minutes, the supply voltage can be reduced,
allowing power reflected back into the circuitry by the load plasma to maintain the
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Figure 32. Modified Power Supply Diagram
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
144
00
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
145
filament cathode voltage.
H ie source voltage for the magnetron which was tapped from LREC was
directed through a pi filter. The pi filter is enclosed in the dotted line box in Figure 32.
The majority o f the filtering takes place across the 16 pF capacitor, which discharges
through the remainder o f the filtering circuitry. The 8 H, 300 mA inductor and 35 pF
capacitor provide additional smoothing. The 400 Q bleeder resistor provides a
discharge path for the capacitors and a fixed load for the power supply. The other side
o f the bridge rectifier was directed to a 5 H filter choke.
The results o f a study of the root mean squared (RMS) and peak-to-peak (P-P)
relative standard deviations (RSD) o f the output waveform for the modified Cober S-3
are shown on Figure 33. An oil-filled dummy load was used as the application load to
enable the study o f a wide forward power output range. The voltage was detected with
a connection to the forward power meter and was directed to an oscilloscope. The
RMS RSD was determined from the relation:
rm s
rsd
-
-g^ MS . 100%
where:
S (x r
°R M S* \
3"
xy
N - 1
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Figure 33. RMS and P-P RSD o f the Modified Power Supply Monitored at the
Forward Power Meter
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
147
100
P-P
RMS
500
750
1000
1250
1500
1750
2000
2250
Forward Power (W)
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
2500
148
The P-P RSD was determined from the relation:
P-P USD
a* p
- -? = £ . 100 %
(A-3)
X
where <*p_pis the waveform voltage range. The percent RM S and P-P R SD are
minimized at forward powers greater than 1400 W . The RM S R SD drops to < 5% at
powers greater than 1400 W.
For elemental analysis, the application load is a microwave-induced plasma. A
vacuum ultraviolet monochromator system as described in reference 4 was used to
optically monitor plasma behavior. A diagram o f the system is given in Figure 34. The
system utilizes rectangular waveguides (sizes W R -430 and W R -284) for power
transmission (Gerling Electronics, Modesto, CA), a slotted waveguide tuner (Cober
Electronics, Stamford, CT), forward and reflected power meters (Cober Electronics,
Stamford, CT), a waveguide to 1.625 in. coaxial transition, and a 1.625 to 0.875 in.
coaxial reducer (Andrew Corporation, Orland Park, IL). A TMq^q resonator cavity
with a diameter o f 88.8 mm and a depth o f 2 cm was used in conjunction with a
tangential flow plasma torch. The waveguide design is the same as that described by
Alvarado and Carnahan (4).
The real-time output waveform o f the modified generator at 1.6 kW forward
power as measured with an in-line antenna in the waveguide and the photomultiplier
output signal o f the N(I) emission line at 149.26 nm is shown on Figure 35. These
waveforms were obtained by using a 16 bit multifunction board (AT-MIO-16X,
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Figure 34. KiP-MIP Power Train
A
B
C
D
E
F
G
E-plane 90° Bend (WR-430)
Slotted Waveguide Tuner
Waveguide Reducer (WR-430 to WR-284)
Forward and Reflected Power Meters
E-plane 90° Bend (WR-284)
Waveguide (WR-284) to Coaxial (1.625 in.) Transition
Coaxial Reducer (1.625 to 0.875 in.)
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
150
Microwave
Power
Generator
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Figure 35. Real-Time Output Waveform o f the Modified Generator and the
Photomultiplier Output Signal o f the N 1 149.26 nm Emission Line
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
152
0.7
Generator Output
0.6
-
0.5
-
0.4
-
RSD (Detector Output at N I 149.26 nm) = 0.187%
0.3
-
0.2
-
0.1
-
0.3
-
0.2
-
0.1
-
0.0 i -------------------- ,-------------------- 1---------------------!
0.00
0.05
0.10
0.15
b 0.0
0.20
Time (sec)
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Detector Output Voltage
0.4
RSD (Generator Output) = 2.23 %
153
National Instruments Corporation, Austin, TX).
After ignition and tuning the system to minimize reflected power using a slotted
waveguide tuner, the plasma is stable throughout analysis. With an RMS RSD of only
2 %, the in-line generator waveform has been smoothed considerably. A stable
analytical microwave-induced plasma can be maintained The noise o f the N(I)
emission signal is minimal and does not fluctuate on the time-scale o f the generator
power supply. In effect, the "line-noise" o f the power supply is eliminated with the
present circuitry design and the system is plasma flicker noise limited; with regard to
the plasma system noise characteristics, the most desirable situation has been achieved.
A.3 Conclusion
A pw microwave source designed for industrial heating applications has been
converted into a cw source suitable for the generation o f an analytically useful He KiP
M3P. The resulting source has been demonstrated to be a reliable stable source with a
low ripple at powers greater than 1400 W. With this low ripple power supply,
discharges may be maintained which reflect minimal a.c. power supply characteristics.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Документ
Категория
Без категории
Просмотров
0
Размер файла
3 781 Кб
Теги
sdewsdweddes
1/--страниц
Пожаловаться на содержимое документа