Plasma spectrochemistry with a fourier transform spectrometer

MISSING IMAGE

Material Information

Title:
Plasma spectrochemistry with a fourier transform spectrometer
Physical Description:
ix, 135 leaves : ill. ; 29 cm.
Language:
English
Creator:
Manning, Thomas Joseph John, 1961-
Publication Date:

Subjects

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

Notes

Thesis:
Thesis (Ph. D.)--University of Florida, 1990.
Bibliography:
Includes bibliographical references (leaves 127-133).
Statement of Responsibility:
by Thomas Joseph John Manning.
General Note:
Typescript.
General Note:
Vita.

Record Information

Source Institution:
University of Florida
Rights Management:
All applicable rights reserved by the source institution and holding location.
Resource Identifier:
aleph - 001638766
notis - AHR3706
oclc - 24160927
System ID:
AA00003751:00001

Full Text










PLASMA SPECTROCHEMISTRY WITH A
FOURIER TRANSFORM SPECTROMETER







BY






THOMAS JOSEPH JOHN MANNING


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


UNIVERSITY OF FLORIDA


1990


UNLYiRSliY "' '"";L Uil;ii3






















To my family; mother, father, brother, sister,
grandparents; whose support, love, patience,
and caring made this possible.












ACKNOWLEDGEMENTS


I would like to express my extreme gratitude to the

following people whose help dedication, friendship, honesty,

and loyalty made this work possible.

My family played a tremendous role in my recent success.

My mother has been a source of inspiration with her hard work

(teaching, Clemson) and working through her lung problems.

She always asks me the question, "and what does the much

mean," I wish I could say much more to express my gratitude

and love. My father has dedicated his life to his family, and

through his encouragement and prodding, my life has progressed

at a rate which he can take much of the credit for. My

parents have blessed me with good health, a good education,

good morals, and the ability to dream and create with the best

of them.

My sister has long been a stabilizing factor in the

family and her husband, George, a welcome addition. Patrick

is young, cute, impressionable, and has a great arm. Jim, my

big brother, has turned out to be a good brother, someone I

can count on and talk to when needed.

My grandparents, particularly my grandfather who recently

passed away, are sorely missed. I can picture them talking to

iii












friends, on a stoop or in a hall way, in Brooklyn. Bragging

about their grandson, the doctor. I have come to realize that

obtaining a goal of this nature is done as much for others

that you care about, so that they are proud, as it is a

selfish goal to satisfy your curiosity about what you are and

what you can be.

Susan and Jeanne have become friends over the years. I

realize my jokes were not always tactful, but they were meant

to be funny. They were an outlet when the possibility existed

for group politics to get out of hand.

Dr. J. D. Winefordner helped fund my salary for part of

my stay at UF and gave me a the chance to discuss my ideas and

experiments with him and other members of the group.

Doug Hof and Byron Palmer of the Los Alamos Fourier

Transform Facility have become more than just collaborators

since I first started working at LANL in 1987. I would have

left UF out of frustration if it were not for the outlet they

and the LAFTS provided. They let me carry out my ideas and

publish my work so I could establish creditability that group

politics never gave me the chance to establish. Doug and

Byron showed great patience and friendship in allowing me to

come back again and again to the LAFTS and for this I am

truly grateful.






Lots of friends have been made in the lab who have helped

me; Steve L., Tye B., Ali A., Kin N., Paul J., Andres C., Lynn

P., Alicia O., Martin A., Jorge V., Barbara K., Leigh Ann F.,

Ed V., Ben W., Eric G., Tony M., Ramme, Nancy S., Edison B.,

Dennis H., Scott S., Alain B., Wellington M., and En Yu Shoa

are the ones I can name right now. Jill Angus of UNM and LANL

has helped me a great deal. I hope our good relationships

keep up for years to come.

Arlene has recently come to play a major role in my life.

Her love, caring, beauty, and outdoor lifestyle are a welcome

addition to my life. I hope our relationship will be a long,

full one that others will be envious of.

I thank God for Los Alamos, Byron, Doug, and my parents

for sticking by me and helping me work through some tough

situations. Without others, my accomplishments would be

hollow if they could have been achieved at all.













TABLE OF CONTENTS


page

ACKNOWLEDGEMENTS ..........................................iii

ABSTRACT...................................................... viii

CHAPTER

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

2 HIGH RESOLUTION FOURIER TRANSFORM SPECTROMETER
TO IDENTIFY THE ROTATIONAL STRUCTURE OF
2 g 2 Zu N2(0,0) IN A HELIUM INDUCTIVELY
COUPLED PLASMA
Introduction.................................. 6
Experimental.................................10
Results and Discussion.......................12
Conclusions............................... ...14

3 A LOW FLOW ATMOSPHERIC PRESSURE
NEON INDUCTIVELY COUPLED PLASMA
Introduction.................................. 31
Experimental....................... ... ........31
Discussion................................... 33
Conclusions.................................... 34

4 A NONFLOWING VARIABLE GAS INDUCTIVELY COUPLED
PLASMA AS A LIGHT SOURCE FOR HIGH
RESOLUTION SPECTROSCOPY
Introduction................ ............. 35
Experimental................. .................... 36
Results......................................37
Discussion....................................38
Conclusions.................................... 41

5 OBSERVATIONS OF LINE SHIFTS AND LINE PROFILES
IN AN INDUCTIVELY COUPLED ARGON PLASMA
Introduction...................................47
Experimental.................................... 50
Results/Discussion..........................52
Conclusions............................... .... 55






6 A VARIABLE BANDPASS FILTER FOR ULTRAVIOLET
AND VISIBLE FOURIER TRANSFORM SPECTROSCOPY
Introduction...... ..........................84
Experimental................................ 86
Results................ ....................... 88
Conclusions.... .............................88


7 AN INTENSITY, WAVENUMBER, AND RESOLUTION
STANDARD FOR VISIBLE FOURIER
TRANSFORM SPECTROSCOPY
Introduction.................................96
Experimental.................................97
Discussion...................................97
Conclusions...................................100

8 A SYSTEM TO STUDY NOISE IN FOURIER
TRANSFORM SPECTROSCOPY
Introduction.................. .................104
Experimental................................ 106
Results......................................107
Conclusions.................................109


9 RELATIONSHIP BETWEEN RESOLUTION AND
SIGNAL-TO-NOISE RATIO'S IN FOURIER
TRANSFORM SPECTROSCOPY ........................119

10 CONCLUSIONS........................ ...........109

REFERENCES.................................................... 128

BIOGRAPHICAL SKETCH.. ..................................135


vii















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


PLASMA SPECTROCHEMISTRY WITH A
FOURIER TRANSFORM SPECTROMETER


By

Thomas Joseph John Manning

August 1990


Chairman: James D. Winefordner
Major Department: Chemistry


This dissertation can be interpreted as being two

dimensional. The first dimension uses the Los Alamos Fourier

Transform Spectrometer to uncover various physical aspects of

a Inductively Coupled Plasma. The limits of wavenumber

accuracy and resolution are pushed to measure line shifts and

line profiles in an Inductively Coupled Argon Plasma. This is

new physical information that the plasma spectroscopy

community has been seeking for several years. Other plasma

spectroscopy carried out includes line profile studies, plasma

diagnostics, and exact identification of diatomic molecular

spectra.


viii












The second aspect of the dissertation involves studies of

light sources for Fourier Transform Spectroscopy. Sources

developed use an inductively coupled plasma (ICP) power

supply. New sources (neon ICP, closed cell ICP, and helium

ICP) were developed and new methods to enhance the performance

and understand a Fourier Transform Spectrometer were studied

including a novel optical filter, a spectrum analyzer to study

noises, and a standard to calibrate and evaluate a Fourier

Transform Spectrometer.













CHAPTER 1
INTRODUCTION


This dissertation will introduce new applications and

studies that couple a Fourier Transform Spectrometer operating

in the ultraviolet and visible region of the spectrum to

various plasma and discharge sources.

Fourier Transform Spectrometers are built for three

reasons; intensity precision and accuracy, high resolving

power, and wavenumber accuracy and precision. The throughput

advantage and multiplex advantage also make the Fourier

Transform Spectroscopy (FTS) attractive but are not the main

reasons the worlds most powerful ultraviolet and visible

Fourier Transform Spectrometer's have been built.

The first advantage is intensity precision and accuracy

in the line profile. This aspect is utilized less than the

other two in this research but is used in temperature

measurements (e.g. Boltzmann methods involving diatomics

(rotational temperature) and atoms (excitation temperature)),

line profile studies, isotope ratio determinations, etc. The

second advantage is resolution or high resolving power. This

advantage was used to look at rotational levels of molecules,

physically resolve line profiles, and identify isotope shifts

and hyperfine structure of atomic lines. The final advantage






2

is the excellent wavenumber accuracy and precision. This

facet of FT-UV/VIS spectroscopy was utilized to measure line

shifts in an Inductively Coupled Plasma (ICP), and better the

existing wavenumber accuracy of a well studied N2+ transition

in ultraviolet. A summation of each chapter follows.

The second chapter is a continuation of my master's

thesis. The Helium Inductively Coupled Plasma described in

this paper was first developed for nonmetal atomic emission

spectroscopy at Los Alamos National Lab (LANL). The source

proved to be very stable (low flicker noise component)

compared to other flowing systems, had very little background

and is sufficiently energetic to be used as a light source.

This project was the first published work using the Los Alamos

Fourier Transform Spectrometer. It clearly demonstrated that

the instrument, a 13 million dollar national facility, was

functioning at a high level. It also demonstrated the He-ICP

is a good high resolution light source. The N/,(0,0)

transition was resolved with higher wavenumber accuracy (>1

part in 107) than had been previously been reported in the

literature.

This same torch was used to generate a low flow,

atmospheric pressure Neon Inductively Coupled Plasma. Chapter

three is a brief outline of the first Neon Inductively Coupled

Plasma described in the literature. This was originally

generated for inductively coupled plasma-mass spectrometry

studies, but the price of neon proved too expensive to








continue this research.

Chapter four deals with a nonflowing, variable gas

inductively coupled plasma as a light source for high

resolution spectroscopy. This work was recently issued a

patent by the United States Patent office. It generated

plasmas with 10 different gases (He, Ne, Ar, Kr, Xe, H,, N2,

CO2, C2H2, 02) over a wide range of pressure (0.01-100 torr).

Work was also initiated to demonstrate the static, low

pressure source could be interfaced to a mass spectrometer

with several advantages over the traditional ICP-MS. For

Fourier Transform Spectroscopy (FTS) research, the static

environment is important to minimize flicker noise in the

source. Varying the gas allows us to vary the Doppler

temperature and electron density as well as the amount of

background light flux.

Chapter five is a physical analytical study. This

chapter represents the largest project undertaken; the

measurement of line shifts and line profiles of light emitted

from a argon inductively coupled plasma. There are small

shifts in wavenumber position between low pressure sources (no

collisional broadening, no shift) and atmospheric pressure

sources. Line profiles were deconvoluted to measure

broadening contributions from Doppler and collisional

processes in the atmospheric pressure argon ICP.

Approximately 180 lines with a signal-to-noise ratio (SNR)

greater than 100 were measured with the LAFTS. This pushed






4

the wavenumber position and resolution capabilities of the FTS

technique. Argon, iron, calcium, barium, and strontium

emission line profiles from an inductively coupled argon

plasma were recorded at sub-Doppler resolution.

The project involving line shifts concluded with the

development of a optical filter that allowed variance of both

bandpass and wavenumber position. It was designed for

ultraviolet and visible Fourier Transform Spectroscopy in

conjunction with the ICP. Its purpose was to reduce the

amount of unwanted light entering the FTS This filter is

discussed in chapter six.

The Re I (575 nm) line has both hyperfine structure (HFS)

and isotope shifts (IS). This structure was proposed as a

standard to determine the quality of spectra acquired from a

FTS. The reason for doing this is simple. Instrument

companies producing instruments have been selling bad

instruments, that is Fourier Transform Spectrometers that do

not come close to meeting their published specs (often by

orders of magnitude). The people purchasing the instruments

typically do not know much about FTS and are not aware of the

problems. By proposing a standard for wavenumber position,

intensity precision, and resolution, it will allow people to

judge the quality of the instrument they are about to buy or

to judge the performance of the instrument at any given time

during its lifetime. It can also serve as a calibration

standard when fine tuning the instrument. Chapter seven









describes this standard.

Chapter eight covers noise studies in Fourier Transform

Spectroscopy using a relatively simple arrangement. The

output of the light source is chopped by a mechanical

chopper. Chopping simulates the constructive and destructive

interference in an interferometer. When the beam is blocked

by the chopper, this simulates destructive interference, and

when light is passed, this simulates constructive

interference. The beam is chopped at a rate (Hz) similar to

that produced by constructive and destructive interference in

a Michelson interferometer. This work indicates that a shot

noise limited case has not been achieved by analytical

spectroscopists, despite their claims.

Analytical spectroscopists typically use a ratio of

signal and noise intensities to calculate the signal-to-noise

ratio in FTS. This method fails at poor resolutions when the

line profile is broadened by instrumental effects decreasing

the amplitude (peak height) of the line profile. Provided our

integration times are equal, there should be identical signal-

to-noise ratios for high and low resolutions alike. This

research has found this not to be the case. At poor

resolutions, the loss of signal amplitude, caused by

instrumental broadening, negatively affects the SNR. Many

analytical studies have been carried out at poor resolutions

and LODs based on amplitude SNRs giving low values. Many of

these studies claim to have achieved shot noise limited cases.







6
Work in chapter seven gives conclusive evidence that

analytical chemists have not achieved a shot noise limited

spectra.












CHAPTER 2
USE OF A HIGH RESOLUTION FOURIER TRANSFORM SPECTROMETER
TO IDENTIFY THE ROTATIONAL STRUCTURE OF THE B E/ X2Z
TRANSITION OF N/(0,O) IN A HELIUM
INDUCTIVELY COUPLED PLASMA



Introduction

The use of Fourier Transform Spectrometers in the near,

middle and far-infrared spectral region is common in many

analytical labs. Since a wealth of atomic information lies in

the visible and ultraviolet regions, there has been a great

deal of effort by various groups worldwide to build Fourier

Transform Spectrometers that operate over this region of the

spectrum. In 1975 Horlick and Yuen (1) reviewed

spectrochemical analysis with a Fourier Transform Spectrometer

(FTS). More recent reviews by Faires (2) and Thorn (3)

discuss theory, instrumentation, and applications of a FTS in

Atomic Emission Spectroscopy (AES). Stubley and Horlick

(4,5,6), Marra and Horlick (7), Faires (8,9), Faires, Palmer,

Engleman, and Niemczyk (10), and Faires, Palmer, and Brault

(11) carried out pioneering work in FT-ICP-AES. Various

aspects of the LAFTS have been previously reported (12,13).






Los Alamos Fourier Transform Spectrometer

The Los Alamos Fourier Transform Spectrometer (LAFTS),

housed in a 3360 ft2 building, is the highest resolution

visible and ultraviolet FTS in the world. Table 1 lists some

of its characteristics. The instrument is currently operating

with single pass optics allowing for a maximum resolution of

0.0026 cm' (0.023 pm at 300 nm). The interferometer is housed

within a 14 foot by 7 foot vacuum tank (30 mTorr achieved) and

utilizes two entrance ports for experiments. The FTS sits on

a 4 foot by 10 foot NRC vacuum compatible optical table

isolating the instrument from vibration.

Figure 1 shows the folded design of the instrument. The

LAFTS utilizes a double sided interferogram which improves the

signal-to-noise ratio and helps eliminate phase problems. The

A/D converter is composed of seven 16-bit A/D converters,

scale amplifiers, and an 18-bit D/A converter for calibration.

Parsons and Palmer (13) give detailed characteristics and

research plans for the LAFTS.

For inductively coupled plasma-atomic emission studies

(ICP-AES) the advantages of the FTS are intensity precision,

wavenumber accuracy, and high resolution. Provided electronic

noise is minimized, the limiting noise in the FT-ICP system

will be the plasma. Several papers (14-18) have discussed

sources of noise in Argon ICPs.








Helium Inductively Coupled Plasma

Argon is the most common plasma gas used in Inductively

Coupled Plasmas for elemental analysis. Abdallah and Mermet

(19) compared temperatures of helium and argon in ICP and MIP

systems. Abdallah et al. followed with a paper (20) utilizing

a 50 MHz Helium-ICP at atmospheric pressure for analytical

studies. Seliskar's laboratory (21-24) conducted work with a

reduced pressure torch that is adaptable to a commercial ICP

unit. Chan et al. (25-27) developed and tested a tangential

flow, low flow, atmospheric pressure torch that is constructed

entirely of macor and adapts readily to commercial units.

Montaser and Van Hoven (28) give a critical review on mixed

gas, molecular gas and helium ICPs. Due to the high

ionization potential of helium (24.6 eV) versus argon (15.8

eV), it has achieved improved limits of detection for elements

with relatively high ionization potentials (Cl 12.9 eV, Br

11.8 eV) (25). He-ICPs have fewer spectral lines than Ar-

ICPs in the desired regions of AES decreasing the chance for

overlap (26). Tan et al. (29) recently introduced a laminar

flow He-ICP torch and demonstrated its use in the analysis of

some nonmetals.



N,+(0.0)

In 1933 Coster and Brons (30) published detailed work on

the emission spectrum of N2/. Diatomic nitrogen (NN,N2*) has

been utilized as a rotational temperature indicator in RF






10

plasmas (31-39) and various discharges (40-46). Most papers

use the unresolved vibrational bandhead or the partially

resolved rotational structure. The use of Raman scattering

from N2 as a temperature indicator has also been demonstrated

by several groups (47-50). Gottscho et al. (51) give an

account of perturbation effects within the B2Z u XZ,

transition of N2/. Herzberg (52) gives an excellent discussion

of both the theoretical and experimental applications of the

transition. Montaser and Van Hoven (28) give references for

additional uses of N2 and N2+ in plasma diagnostics. Nitrogen

is used in plasmas for metal surface treatment (53), synthesis

of nitrogen oxydes (54), and CO, CO2, and gas lasers (55).

The use of N2 in sub-doppler resolution studies is discussed

in this paper.

The first negative band of N2 has several advantages for

rotational temperature determinations in plasmas. The (0,0)

bandhead (391 nm) is in an accessible region of the spectrum.

The Q branch (AJ=0) is forbidden decreasing the number of

rotational lines present relative to N2(0,0), N2(0,1), and

OH(0,0) (56), which are used as rotational temperature

indicators. At moderate resolution (<0.25 cm') the

rotational structure can be adequately resolved.

This chapter presents a low flow, laminar flow,

atmospheric pressure helium ICP torch. Nitrogen is seeded

into the helium plasma and the emission from N92 is analyzed

using the Los Alamos Fourier Transform Spectrometer.








Experimental

Helium inductively coupled plasma torch

The torch is optimized for low flow (2-6 L/min),

simplicity in design, and adaptability to a commercial unit.

Chan et al. (25) present a low flow (7 L/min), tangential

flow, macor torch. Two problems arise with this design.

First, macor (Corning, Horseheads, NY), a heat resilient,

machinable ceramic, is extremely brittle and time consuming to

work with. Second, the spiral design allows the plasma to

wrap around the torch at low flow rates increasing the chance

for damage to the macor torch. An attempt to run a mixed

argon-helium plasma also proved detrimental to the macor.

These two problems were overcome by a) reducing the amount of

macor and constructing the torch largely from teflon and b)

making the torch laminar flow and keeping the plasma above the

macor tip.

Figure 2 shows the diagram of the torch used in this

study. The plasma gas enters the torch through two side ports

spaced 180 apart. Two ports are utilized to give an even

distribution of gas through the eight symmetrically spaced

channels that carry the plasma gas from the teflon channels

through the bores in the macor tip and into the plasma. The

sample is carried up through the center of the teflon and

macor pieces and enters the plasma region through a 2/1000

inch aperture. The torch handles 10,000 ppm Na solution with

no clogging problems. This system uses a fritted disk






12

nebulizer (11). It has higher efficiency (25% with 0.3 L/min

gas flow), lower cost, and a decreased noise power spectrum

(14) versus the ultrasonic nebulizer. Helium does not work

with a concentric nebulizer due to its relative lack of

viscosity versus argon. An extended quartz tube was used to

keep impurities (N2, H20 etc.) from the atmosphere out of the

plasma. It has one or two wraps of teflon tape around its

base for an airtight fit with the base. A four turn, gold

plated (57) coil was used to generate the 27.12 MHz RF field.

The central teflon piece and the teflon base (2" diameter) fit

tightly together. The torch is supported on a teflon stand in

the ICP box. The tesla coil is attached, via an ignition

wire, to one of the side ports with a copper clip. Quartz

tubing (1/4" diameter) is inserted into the two side ports.

A 12/5 balljoint is inserted into the sample uptake port. The

macor tip has a 0.5 inch diameter and a 0.5 inch height. It

fits tightly with the teflon channel (see Figure 2). Under

normal operating conditions this fit is sufficient to seal the

tip from outside contaminants and keep the sample and plasma

gas separate. As an added precaution, the outside of the

macor-teflon connection is wrapped with teflon tape. The

macor to teflon connection can be fused with chemical etching

techniques or a high temperature quartz based cement can be

used as a sealant. Impedance matching utilized has been

previously described in the literature (58,59).








Inductively coupled plasma

This study employs a 27.12 MHz RF Plasma Products (Cherry

Hill, NJ) HFP 5000 D model Inductively Coupled Plasma unit.

The ICP operating conditions are outlined in table 2. The

nitrogen is 99.9995% certified pure and was obtained from

Union Carbide-Linde Division (Albuquerque, NM).

Los Alamos Fourier transform spectrometer

The parameters of the LAFTS used for this study are as

follows: The spectral output is from 14000 cm1 to 40000 cm".

The detectors utilized are RCA 1P28s. The beamsplitters are

aluminum coated quartz and no optical filters are used. The

resolution chosen (0.09 cm') is sub-Doppler for nitrogen (19).

The number of points taken is 889,219, which is zero filled to

2 (1,048,576) points for transformation. The data

acquisition time for the 20 scans is 1 hour 17 minutes and the

computer transform time is 42 minutes. The spectrum is

unapodized (boxcar). Atmospheric pressure in the LAFTS (7300

ft above sea level) is 593 Torr, and the tank temperature is

21C.



Results and Discussion

Table 3 gives the line positions (0.01 cm') of the P and

R branches observed in the Helium Inductively Coupled Plasma.

The accuracy of the (0,0)band data given by Coster and Brons

(30) is updated by 0.1 to 0.2 cm .

Figure 3 shows part of the (0,0) spectra. The P (1-38)






14

and R(1-10) branches are identified. The pattern is

consistent for larger branch numbers not shown. The doublet

splitting (J-1/2,J+1/2) is evident for both branches (P >16,

R >10). The alternation of line intensity (even > odd) is

unmistakable for both branches. Doublet splitting and even-

odd alternation of line intensity are treated theoretically by

Herzberg (52).

Figure 4 gives the Fortrat diagram for the P and R

branches in the He-ICP. The P branch folds over itself at

25540.10 cm1 giving the vibrational bandhead. The K value

used for all branches that are doublet split is J-1/2. At

higher K values (R >40, P>44) only the even branches are

detectable from this source (He-ICP).

The method to calculate the rotational temperature from

a well resolved structure is taken from Abdallah and Mermet

(19). This method allows for relative line intensity

measurements (as opposed to absolute) for either the P or R

branches. For the P branch,


Log I = f(K"(K"-l)) (1)
K"

and the R branch,

Log gl = f((K"+l)(K"+2)) (2)
K"+1

where a = 1 (even number lines), 2 (odd number
lines)
f = -Bhc/kT = (-2.983/T)
I = experimental line intensity
K" = quantum number of lower state
T = temperature (kelvin)






15

Plotting the left side of equation 1 or 2 (for varying

values of I) versus the right side (increasing values of K")

of the respective equation yields a straight line with a slope

proportional to 1/T. Rotational temperatures for the He-ICP

using the parameters and data listed above are R(odd, 1700 K),

R(even, 1680 K), P(odd, 1830 K), and P(even, 1660 K).

Various papers (35,38,39) outline similar Boltzmann

methods to calculate rotational temperatures in plasmas. This

method allows for four rotational temperatures (P (even,odd),

R(even,odd)). Thorne (60) claims that the Boltzmann method for

N2/(0,O) is suitable for measurements under 5000 K. Cabannes

(39) claims it is allowable for temperatures under 10000 K.

Kornblum and de Galen (61) have interpreted the resulting

measurement as the gas kinetic temperature.



Conclusion

A new He-ICP torch has been demonstrated. The torch has

also been used to generate a neon ICP (62) and is being

investigated as an ionization source for helium and neon

ICP-MS studies. The He-ICP is not being pursued for atomic

emission spectroscopy.

The use of N2 as a diagnostic tool is common. This paper

allows the spectroscopist to easily identify and use the

structure for rotational temperatures. The line position

accuracy of the rotational levels is improved against those

currently in the literature.



























Figure 1. Optical Elements: a) cat's-eye reflectors b) beam
splitter and carousel c) 8-inch turning mirrors d)
collimator optics e) output mirrors Light enters from
either the end or side port and is collimated by d. There it
goes to the upper beam splitter, b, and is divided. These two
beams go to the turning mirrors c and to the primary mirror of
the cat's-eye reflector, a. In the cat's-eye reflector the
beam is focused down onto the secondary mirror and reflected
back to the primary where it is made parallel again. The
effect of the cat's-eye reflector is to reflect the beam back
along a path that is parallel but displaced. From the cat's-
eye reflectors the beams return to the 8-inch mirrors and then
to the lower beam splitter. There the beam is recombined and
interferes with itself. The light then travels to the output
mirrors, e, that focuses the beam and to the flat mirrors that
direct the beam to the detectors.







17





















d
d
end input __--j )) c


e c h~U

~I~KCS .0






















Figure 2. Diagram of low-flow, atmospheric pressure He-ICP
torch. The top inset shows the macor tip and teflon central
channel interface in greater detail. The bottom inset shows
the eight plasma gas holes surrounding the sample gas inlet on
the macor tip.

































DOWNWARD.EXPA~N0DE
VIEW OF MACOR TIP

0
O0
PLASM o O 0
GAS
0 0


SAMPLE
INLET






















Figure 3. The four panels show a portion of the rotational
structure of the N2*(0,O) band. The top x-axis is angstroms
and the bottom is wavenumbers (cm').





























8e+008



6~e008



4e+008



2e+008



0


WAVENUMBER























5e1008


4e+008



2e+008


2e+008



le+008



0


WAVENUMBER


Figure 3--continued



























3e+008



2eM
(O 2e-008
z
I-.
z

le+008


WAVENUMBER


Figure 3--continued


























4e+008



3e+008


2e+008



le+008



0


WAVENUMBER


Figure 3--continued


























Figure 4. Fortrat diagram of P and R branches. At 25540.10
cm the P branch (P12, P13, P14) folds giving the
bandhead.



































*
*
G
G
E
E
Q




.
Q

Q
E

E
Q






* 0
* E


* 0
* 0
* 0
* 0
* 0

40
Qm
B0


25700


13


n BRANCH
* P BRANCH


25900
25900


26100
26100


WAVENUMBER


n


. n .. ..


25500








Table 1. Los Alamos FOurier Transform Spectrometer
Specifications


Spectral Ranges:


Region Range

UV/VIS 200 nm-l.l um


NEAR IR
FAR IR


0.8-5.5 um
4-20 um


Beamsplitters

quartz/aluminum

CaF-GaP
KC1-Ge


Detectors

photomultiplier,
si PIN diode
InSb
HgCdTe


Resolution and Operating Parameters:


Optical Path
Difference
Resolution

Sample Rate



Number of points

Maximum Transform
Size


2.5 meters (single pass)
5.0 meters (double pass)
0.0026 cm (single pass)
0.0013 cm (double pass)
5000 Hz not digitally filtered
40000 Hz with digital filtering
reduced to 5000 Hz throughput

>4,000,000 points

22 (4,194,304)









Table 2. Helium inductively coupled plasma operating
conditions


Source Unit
Frequency
Forward Power
Reflected Power
Plasma Gas Flow
Nebulizer Flow
Sample Uptake Rate
Viewing Height above
load coil
Aperture to FTS


RF Plasma-Therm 5 kW unit
27.12 MHz
1.5 kW
<5 W
4 L min'
0.5 L minm
0.3 mL min"
14 mm

7 mm









Helium inductively coupled plasma.


P BRANCH


K Value
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
16
17
17
18
18
19
19
20
20
21
21
22
22
23
23
24
24
25
25
26
26
27
27
28
28
29
29
30


J(+1/2)
















J-1/2
J+1/2
J-1/2
J+1/2
J-1/2
J+1/2
J-1/2
J+1/2
J-1/2
J+1/2
J-1/2
J+1/2
J-1/2
J+1/2
J-1/2
J+1/2
J-1/2
J+1/2
J-1/2
J+1/2
J-1/2
J+1/2
J-1/2
J+1/2
J-1/2
J+1/2
J-1/2
J+1/2
J-1/2


WAVENUMBER(cm')
25562.36
25558.76
25555.54
25552.62
25549.96
25547.66
25545.67
25543.81
25542.47
25541.37
25540.59
25540.11
25540.11
25540.11
25540.59
25541.15
25541.37
25542.16
25542.47
25543.54
25543.85
25545.14
25545.44
25547.12
25547.35
25549.40
25549.65
25552.20
25552.59
25554.82
25555.09
25557.99
25558.28
25561.46
25561.76
25565.24
25565.56
25569.32
25569.66
25573.71
25574.06
25578.42
25578.76
25583.41









Table 3--continued.


J+1/2
J-1/2
J+1/2
J-1/2
J+1/2
J-1/2
J+1/2
J-1/2
J+1/2
J-1/2
J+1/2
J-1/2
J+1/2
J-1/2
J+1/2
J-1/2
J+1/2
J-1/2
J+1/2
J-1/2


R BRANCH


J(1/2) Wavenumber (cm')
25570.30
25574.77
25579.51
25584.57
25589.93
25595.61
25601.58
25607.85
25614.44
25621.33
J-1/2 25628.41
J+1/2 25628.56
J-1/2 25635.89
J+1/2 25636.12
J-1/2 25643.67
J+1/2 25643.90
J-1/2 25651.76
J+1/2 25651.99
J-1/2 25660.16
J+1/2 25660.39


25583.77
25588.70
25589.06
25594.31
25594.67
25600.21
25600.56
25606.41
25606.76
25612.88
25613.22
25619.59
25619.99
25626.96

25634.12
25634.39
25640.85
25641.80
25648.44
25656.19
25668.43
25676.62
25685.82


K Value
0
1
2
3
4
5
6
7
8
9
10
10
11
11
12
12
13
13
14
14









Table 3--continued.

15 J-1/2 25668.86
15 J+1/2 25669.09
16 J-1/2 25677.79
16 J+1/2 25678.09
17 J-1/2 25687.18
17 J+1/2 25687.40
18 J-1/2 25696.72
18 J+1/2 25697.02
19 J-1/2 25706.62
19 J+1/2 25706.88
20 J-1/2 25716.78
20 J+1/2 25717.08
21 J-1/2 25727.27
21 J+1/2 25727.58
22 J-1/2 25738.05
22 J+1/2 25738.38
23 J-1/2 25749.15
23 J+1/2 25749.53
24 J-1/2 25760.53
24 J+1/2 25760.86
25 J-1/2 25772.21
25 J+1/2 25772.55
26 J-1/2 25784.17
26 J+1/2 25784.54
27 J-1/2 25796.46
27 J+1/2 25796.82
28 J-1/2 25809.02
28 J+1/2 25809.39
29 J-1/2 25821.89
29 J+1/2 25822.26
30 J-1/2 25835.04
30 J+1/2 25835.41
31 J-1/2 25848.48
31 J+1/2 25848.86
32 J-1/2 25862.21
32 J+1/2 25862.58
33 J-1/2 25876.22
33 J+1/2 25876.59
34 J-1/2 25890.44
34 J+1/2 25890.85
35 25905.36
36 J-1/2 25919.96
36 J+1/2 25920.27
37 J-1/2 25934.16
37 J+1/2 25935.11












CHAPTER 3
GENERATION OF A LOW FLOW ATMOSPHERIC PRESSURE NEON ICP


Introduction

A 27.12 MHz Neon Inductively Coupled Plasma (Ne-ICP) has

been generated in this laboratory. Argon has long been used

as the plasma gas for an ICP. Abdallah and Mermet (19) and

Chan et al.(25) have generated atmospheric pressure helium

plasmas in inductively coupled plasma (ICP) units. Walters et

al. (63) generated a reduced pressure (0.5-14 torr) 144 MHz

neon ICP for electrodeless discharge lamp studies. This is

the first known report of an atmospheric Ne-ICP. The torch

design has been used for helium ICP studies and is described

by Manning et al.(64). The neon ICP is being investigated as

a possible atomization source for adaptation to a commercial

inductively coupled plasma-mass spectrometry instrument.



Experimental

The Ne-ICP is generated with a 27.12 MHz RF Plasma

Products HFP 5 kW unit (Cherry Hill, NJ). The forward power

for the ICP is 500 W with less than 5 W reflected (58). The

low flow (3 1/min), laminar flow, atmospheric pressure torch

is described by Manning et al.(64). It utilizes an extended

quartz sleeve to minimize impurities (N2, HO, etc.) from the






33

atmosphere entering the plasma. The nebulization system

employed is the fritted disk nebulizer (11). The sample is

pumped to the fritted disk with a standard peristaltic pump.

The flow rate of the plasma gas is 3 L/min with a minimum of

2 L/min and a maximum of 7 L/min possible. The nebulizer gas

flow rate is 0.3 L/min with a sample uptake rate of 0.3

mL/min. The nitrogen is seeded in at a rate of 5 mL/min. The

neon gas is supplied by Union Carbide-Linde division gas

distributors (Albuquerque, NM).

Los Alamos Fourier Transform Spectrometer

The Los Alamos Fourier Transform Spectrometer (13), a

national facility for high resolution atomic and molecular

spectroscopy, is used in this experiment for medium resolution

(0.07 cm') spectra in the near IR, visible, and ultraviolet

(8500 cm' to 33372 cm) Aluminum coated quartz beamsplitters

are employed and the detectors utilized are silicon PIN

diodes. The number of points taken for each run is 953,863

and the transform size is 220 (1046000) points. The run in

which only water is introduced into the plasma consists of 10

scans (39 minutes data acquisition time) and took 16 scans in

the run introducing both water and N2 (64 minutes data

acquisition time). The number of scans is limited by the

current availability of research grade neon. The cost of neon

is the limiting factor.








Results and Discussion

Neons' ionization potential (21.56 eV) is higher than

that of argon (15.76 eV) but lower than helium (24.587 eV).

This should offer better detection limits for various

nonmetals (eg. F, Cl, Br, etc.). The neon plasma as a source

does not have the luminosity of the argon ICP and is similar

to the helium ICP in this respect. The plasma is stable with

powers as high as 1.5 kW. Introducing sample reduces its

stability. With an aerosol entering the plasma, it could not

operate at powers greater than 600 W. Above this threshold

the plasma either extinguished or formed numerous filaments

and produced a non-homogeneous plasma. The plasma operated

with powers as low as 150 W. Luminosity appears linear with

power.

An attempt to measure the rotational temperature using

N2+(0,0) (391.44 nm) and OH(0,0) (306.4 nm) is made. The N2

bandhead has a SNR of 4.2 and the R branches used in

diagnostics are not discernible making reliable measurements

impossible. The OH(0,0) band is not present at all in the

spectrum. The Hg line is used to measure the electron number

density (65) at 8x10'3 cm The fact that both species

(OH,N2+), which are strong UV emitters in plasmas and

discharges, do not have a readily detectable emission rate,

coupled with the low electron density measurement (8x10'3 cm3)

and the low input power (500 W), give reason to believe the

Ne-ICP in this setup is cool compared to the argon and helium







35

plasmas (19) run at higher powers. Bands from the first

positive system (B3Z, A3.) of N2 are obvious in the near IR

and visible. This transition (10,000 cm'-17,000 cm') requires

less energy for excitation than OH(0,0)

( >31,000 cm').



Conclusion

An atmospheric pressure neon ICP is generated. In the

current setup it operates reliably at 500 W forward power. The

Ne-ICP is being pursued as an atomization source for a

commercial ICP-MS setup. A greater amount of diagnostics is

not feasible due to the expense of running a Ne-ICP. A single

Ne line (585 nm) was used to determine the Doppler

temperature. The value (1058 K) was obtained by deconvoluting

the line profile with a Voigt profile.

Using a Ne-ICP as an ion source for ICP-MS studies should

offer the advantage of "freeing" various mass values that are

not accessible with argon as the plasma gas.












CHAPTER 4
A NON-FLOWING, VARIABLE GAS INDUCTIVELY COUPLED PLASMA AS
A LIGHT SOURCE FOR HIGH RESOLUTION SPECTROSCOPY



Introduction

Ten self-contained 27.12 MHz Inductively Coupled Plasmas

(ICP) are generated in a static torch. The torch offers

spectroscopists a closed system, low pressure (0.01 to 100

Torr), variable power (currently up to 1 kW) plasma. Doppler

temperatures (Na D, 589.0 nm) measured are approximately 5

times that of electrodeless discharge lamps (EDLs) or Hollow

Cathode Lamps (HCLs) operating under normal conditions. It has

improved noise considerations verses the flowing ICP

traditionally used.

Plasmas are generated using helium, neon, argon, xenon,

krypton, hydrogen, nitrogen, oxygen, carbon dioxide, and

ethylene. In some cases, particularly xenon, the closed

system minimizes the consumption of expensive gases. In the

case of hydrogen, it minimizes the potential hazard of working

with the gas in a flowing system. Ideas for absorption,

ICP-MS, and lasing action in a static plasma are discussed.

This torch generates a low noise, high power plasma that can

be used for a variety of spectroscopic studies.








Experimental

Figure 5 shows the torch design utilized in this study.

The quartz tube has a 17 mm OD, 15 mm ID, and a 14 cm length.

A quartz window is fused to the viewing end of the tube. In

some circumstances, using grounded clips helps the plasma

symmetrically fill the entire tube. End on viewing increases

the luminosity of the source. The four turn coil is gold

plated 1/8 inch copper tubing. The ICP instrument is a RF

Plasma Products HFP 5000 D (5 kW) unit. Often the plasmas are

initiated by introducing the field (RF ON) and slowly

increasing the power. In instances where this is not

possible (typically >10 Torr), a tesla coil discharge is sent

through one of the grounded clips to ignite the plasma. A

vacuum rack equipped with a liquid nitrogen cold finger, a

roughing pump, and a diffusion pump is utilized for evacuating

the cell and filling it with the desired gas. The cell is

evacuated to 105 Torr, back filled with the desired plasma

gas, evacuated to 10.5 Torr, and filled to the desired

pressure.

The cell is demountable and easily inserted into the ICP

coil. A 12/5 balljoint connects the rack and the torch. As a

precaution, nitrogen is blown across the quartz torch for

cooling purposes. Pressures given are the fill values at room

temperature. Powers recorded are the forward minus the

reflected values.

The Los Alamos Fourier Transform Spectrometer is used to






38

record the spectra at high (0.039 cm') resolution with 5

scans. The detectors are Silicon PIN diodes, and the

beamsplitters are aluminum coated quartz. The spectral window

transformed is from 8500 cm' to 40000 cm'. 2,052,056 points

are acquired and zero filled to 2,097,152 (221) points.

Additional parameters of the LAFTS are supplied by Manning et

al (64) and Parsons and Palmer (13). All gases used are

certified research grade (>99.995 %).



Results

Figure 6 shows the Na D (589.0 nm) line profile in a Xe-

ICP. Sodium (and lithium) impurities are present in most of

the spectra obtained. They are attributed to impurities in

the quartz. The sodium D line is used for the determination

of the Doppler temperatures in each plasma. Values ranged

from xenon (7 torr, 250 W) at 3700 K to both carbon dioxide

(2.8 torr, 250 W) and oxygen (6 torr, 550 W) at 2400 K. These

values are approximately 5 times that of HCLs or microwave-

EDLs, assuming a normal current or power input. To avoid

having the sample absorb on the cooler ends of the torch, the

length of the torch may have to be reduced so the tube length

fits entirely within the coil. Adding small quantities of a

desired species (e.g., typically in the 1-10 microgram range)

allows the torch to be used as a stable reservoir of atomic

lines. The same is possible for molecules stable in the

plasma environment.






39

The low pressure inert gas (He, Ne, Ar, Kr, Xe) plasmas

seeded with trace amounts of N2 for diagnostic measurements

exhibited varying degrees of continuum emission in the near IR

and visible. Xenon is the most prevalent with approximately

15 prominent continuum emission bands. The signal to

background measurements of the emission bands ranged from 3 to

15 and are typically 2-3 nanometers wide. The bands are found

between 625 and 880 nm. They are semicircular in appearance.



Discussion

The analysis of atomic and molecular species stable in a

static plasma utilizes some of the capabilities of the UV/Vis

Fourier transform spectrometer (particularly the LAFTS). In

particular the intensity precision, wavenumber position

precision and accuracy, and resolution capabilities are

partially realized.

There is no reason to believe this cell could not

generate other plasmas. Atomic species which are gaseous at

low pressure (mercury, radon, volatile salts), diatomic gases

(NO, CO, C12, Fl2, etc.), polyatomic gases (NO,, SO2, C2H,, other

hydrocarbons, etc.) should readily form low pressure plasmas

in this setup. The ethylene-ICP and other hydrocarbons

systems may be of interest to those studying the diamond

synthesis mechanism in a RF-plasma environment (66). Closed

cell hydrogen-ICP minimizes the dangers of working with a

flowing system. This hydrogen-ICP allows scientists to study






40

atomic interactions in a system similar to the surface of

cooler stars.

Fusing quartz windows to both ends of the plasma tube

allows a variety of absorption measurements to be conceived.

For example, intracavity absorption (67) with an ICP is

possible with the advantage of a longer path length, smaller

sample size, and an increased residence time per atom verses

the traditional flowing ICP torch.

A closed system ICP is currently being investigated as an

ion source for mass spectrometry. One end of the tube is

directly interfaced to the mass spectrometer with the flow

regulated by a gated valve. The sample is injected through the

syringe port (0.1 to 1 microliter) before ignition. The

solvent evaporation is enhanced with the vacuum leaving only

sample in the quartz tube. With plasma ignition, the sample

sputters into the plasma. We have found this to work quite

well for the nonmetals and softer metals. The variable gas

static plasma for ICP-MS should offer many advantages over the

flowing argon system. Expensive gases (e.g., neon, krypton,

xenon) can be routinely used due to low gas consumption.

These inert gases may be used to free mass ranges not

accessible with argon (e.g., sample overlaps with ArN ArH*,

ArOH+, etc.). Provided experimental conditions are identical

(power, pressure), the temperature and electron density of the

inert gas plasma (He, Ne, Ar, Kr, Xe) will have an inverse

relationship with the ionization potential. That is, the






41

higher the ionization potential is, the lower the temperature

and electron density will be. Thus by varying the inert gas,

the temperature and electron density range can be varied. The

closed cell allows for small sample sizes. It reduces some of

the hazards associated with radioactive samples in a flowing

system. The pressure differential between the plasma and the

mass spectrometer is reduced by at least two orders of

magnitude. Problems associated with the skimmer/orifice

degradation are reduced. The sample waste associated with a

flowing system is minimized. Nebulization is eliminated. In

some situations it may be possible to use nonpolar samples and

solvents. With evaporation/dehydration techniques, the

formation of hydrides, hydroxides, and oxides will hopefully

be reduced by applying this procedure. The low pressure cell

operated here can be converted to a fireball (68) at higher

powers. The fireball resembles the atmospheric pressure argon

ICP in appearance. This fireball could not be sustained for

extended periods of time or the quartz would glow red and

eventually melt. If a continuous vacuum is applied that

maintains a constant pressure, the fireball can be maintained

for extended periods of time.

Eckert (69) gives an excellent theoretical discussion of

elemental detection in a static induction plasma. His

calculations predict a limit of detection for magnesium (279.6

nm line) of less than 10"15 g in a nonflowing argon-ICP. With

small modifications to the torch, the closed cell ICP may have






42

analytical usefulness. This analytical research would be

similar to work previously carried out with microwave

electrodeless discharge lamps.



Conclusion

The static plasma outlined in this note describes an atom

and molecule reservoir for spectroscopic studies (70). It is

also being pursued as an ion source for mass spectrometry

(70). No graphical data is presented to support the claim

that noise factors in the static plasma are improved over the

flowing system. We hope the reader follows the reasoning that

with the lack of turbulence induced by the gas flow, and no

nebulization system, many of the sources of noise inherent in

a flowing system are either reduced or eliminated in a static

plasma. The intensity variation as a function of time is at

least an order of magnitude improvement over the flowing

system under normal operating conditions.

With various adaptations to the cavity, the cell could be

used for a variety of emission, absorbance, and fluorescence

measurements, both analytical and physical in nature.

























Figure 5. Closed cell ICP torch design.








44





















---- TO1
1FTSt


S-- STOPCOCK~



























Figure 6. Line profile of Sodium-D (589.0 nm) in Xe-ICP.

































2e+009



l. e+009

C-

Z
5-
z



0
o


WAVENUMBER












CHAPTER 5
OBSERVATION OF LINE SHIFTS AND LINE PROFILES
IN AN INDUCTIVELY COUPLED ARGON PLASMA


Introduction

Line shifts in analytical inductively coupled plasma's

(ICP) are a fundamental parameter that has not been studied in

any great detail. This paper will give an introduction to

line shifts measured and trends observed. For more detailed

treatments of pressure effects, including line shifts, the

reader is directed to Thorn (60) (overview), Alkemade et al.

(71) (in flames), Griem (72), and Breene (73) and references

within. Reviews by Konjevic and Weise (74) and by Konjevic

and Roberts (75) give numerous references on Stark broadening

and line shifts from a variety of sources.

Line shifts are important for a variety of reasons.

Analytical determination of elements in a complex sample is

often plagued misidentification of the species being studied.

In physical spectroscopy, ICPs are being used for the

measurement of transition probabilities. Knowledge of line

shift values are important to properly measure energy levels,

transitions, etc.. Whaling and Brault (76) used an ICP in

conjunction with the Kitt Peak Fourier Transform Spectrometer

to measure the transition probabilities of 1200 Mo (I) lines;






48

they introduce Mo as Mo(CO)6 eliminating any effects water has

on the spectrum. They observed line shifts of Mo as great as
-1
0.14 cm .

The Los Alamos Fourier Transform Spectrometer (LAFTS) is

an ideal instrument to measure small spectral shifts. Its

high resolution capabilities easily resolve physical line

widths emitted by hollow cathode lamps (HCL) (0.025 cm1) and

the inductively coupled plasma, (0.07 cm'). The wavenumber

precision of the LAFTS can be as great as 1 part in 108. The

LAFTS can acquire large spectral windows (9000 to 42000 cm')

at the required resolution to physically resolve the ICP

emission.

In studying line shifts, isotope shifts and hyperfine

structure must be considered to properly select a line.

Elements that clearly display either effect are avoided due to

the complexity introduced in attempting line profile studies.

The spectrum of a Pb emission line shows contributions from

three isotopes, 8Pb206 (24.1%), 82Pb207 (22.1%), and 82Pb208 (52.4%).

These isotope shifts are shown in Fig. 7. This spectrum was

taken from a low pressure HCL with a Doppler temperature of

approximately 500 K, resulting in Doppler broadened spectral

lines having negligible pressure (collisional) broadening.

Increasing the temperature to 6500 K and introducing

substantial amounts of collisional broadening, which simulates

ICP conditions, will cause the three isotopic lines to appear

as one asymmetric, broadened line. Measuring the line shift






49
becomes difficult because the individual lines have to be

deconvoluted before spectral shift measurements are possible.

There will be uncertainty introduced into the calculation by

the deconvolution which may affect the validity of the shift

measurement.

Figure 8 shows the hyperfine structure of tantalum. 73Ta 81

is the predominant isotope occurring naturally (99.998%) so

isotopic effects are negligible. With a nuclear spin of 7/2,

73Ta181 clearly exhibits hyperfine structure. Kuhn (77) gives

an excellent treatment of hyperfine structure and isotope

shifts. Ta emission in an ICP has significant Doppler and

pressure broadening of the lines giving asymmetric line

profiles with 8 convoluted hyperfine elements making shift

measurements virtually impossible.

Argon has one predominant isotope (lAr40,99.60%) which has

no nuclear spin making it an ideal choice for line shift

studies. Figure 9 shows a line profile typical of Ar I

emission at 420 nm from the ICP. Iron has one dominant

isotope (26Fe56, 91.72%), which has no hyperfine structure;

Fig. 10 shows a line profile of Fe I at 259.9 nm. The effect

of the lower abundance isotopes of Fe on the line center

position (shift measurement) and the full width at half

maximum (FWHM, a-parameters; where a = 0.86*(I/AD) are

considered negligible for this study.

Kato et al. (78) measured profiles of 16 spectral lines

stemming from 8 elements emitted by an ICP and measured with






50

a pressure-scanning Fabry-Perot interferometer. The results

their study are similar in absolute magnitude of some lines

but occasionally differ in sign. They did not consider the

effects of hyperfine structure and/or isotope shifts as

carefully as we did. Boumans and Vrakking (79) published work

on high-resolution spectroscopy using an echelle spectrometer,

which refers to line shifts in the ICP in connection with the

wavelength calibration of the spectrometer.



Experimental

The Los Alamos Fourier Transform Spectrometer (LAFTS) is

utilized in conjunction with a 27.12 MHz RF Plasma Therm

(Cherry Hill, NJ) ICP. Papers by Manning et al. (64) and

Parsons and Palmer (13) give details of the LAFTS. For all

studies, the viewing height above the coils is 12 mm. The gas

flow rates are 12 L/min (outer), 0.8 L/min (intermediate), and

0.8 L/min (nebulizer). For ICP studies, the LAFTS was operated

at 0.07 cm' resolution with a free spectral range of 33005.2

cm For the HCL studies, the LAFTS was operated at 0.02 cm

resolution with a free spectral range of 40015.2 cm1. The

detectors used were PMTs (1P28) or silicon pin diodes. The

silicon pin diodes were utilized for work involving Ar

emission in the near IR.

The output of the light source (ICP or HCL) is focused

with unity magnification on the FTS spectrometer aperture.

The f number of this spectrometer and the ICP/HCL optical







51
setups are matched. The output of a neon filled Al HCL is

impressed on the ICP emission using a neutral density filter

(0.2) before the focusing optics. Figure 11 shows an outline

of the entrance optics and the light sources. The Ne I line

at 585.2 nm from the HCL is used as a calibration standard for

all runs.

All emission lines used in this study (HCL and ICP) have

signal to noise ratios (SNR) greater than 100. This

requirement reduced the number of lines available between the

two sources. Some strong lines in the HCL were not always as

prevalent in the ICP and vice versa. Ionic lines for iron and

argon detected in the ICP were not observed in HCL emission or

had low signal-to-noise ratios, reducing the accuracy and

precision of the wavenumber measurement.

Six (Fe (1.1, 1.5, 1.9 kW), Ar (1.1, 1.5, 1.9 Kw))

Doppler temperatures were calculated from the ICP emission

line profiles using a Voigt fit. This Doppler temperature was

then used for all calculations involving those particular

conditions. For example, an Fe I line with a high SNR is

chosen from the 1.1 Kw ICP data. Voigt profiles are generated

until one is identical to the Fe I (ICP) profile. This

allowed us to estimate the contributions from Gaussian

(Doppler broadening) and Lorenztian (collisional broadening)

components. The Gaussian component is then used for all line

profile calculations involving Fe emission from the ICP at 1.1

kW. We measured the Gaussian and Lorentzian contributions for






52

one Ar and one Fe spectral line at each power level (1.1, 1.5,

1.9 kW) giving a total of six Doppler temperatures.

The Hp line is used to estimate the electron density in

the analytical zone viewed by the LAFTS. These electron

densities applied only to the analytical zone; the region

where the Fe, Ba, Ca, and Sr emission predominantly occurs is

studied. Ar emission is not predominantly from the analytical

zone but from the outer tangential flow zone. No electron

density calculations were made for the latter zone.



Results and Discussion

The results are separated into three groups. First, Fe

and Ar emission lines in the ultraviolet and visible at three

separate power settings (1.1, 1.5, 1.9 kW) were measured.

Second, Ar emission in the near infrared and far visible was

measured. Third, prominent ion lines of Ca, Sr, and Ba at 1.1

kW were measured.

The first section refers to line shifts as a function of

power for 28 Fe lines and 21 Ar lines in the ICP. The power

dependence (1.1, 1.5, 1.9 Kw) is shown to increase Doppler

temperature, electron density, and shift. The iron data are

refer to the analytical zone of the ICP, whereas the argon

data refers to the outer tangential flow. For the argon and

iron data, the line position is given in Angstroms and

physical parameters (line shift, width, etc.) are given in

wavenumbers (cm'). Line positions are given in Angstroms







53
because the references used for transition assignments list

them in this manner. Most research measuring line shifts

report their results in cm'.

Table 4 gives the Doppler temperature (6500 K),

transitions types, wavelengths, line shifts, Doppler and

Lorentzian contributions to the line profile, and the a-

parameter for 21 argon lines at 1.1 Kw. Transition

assignments for the Ar-term configurations are taken from Li

and Humphrey (80). A single argon line at this power is fitted

with a Voigt profile to approximate the Doppler contribution

to the line profile. The resulting Doppler temperature of

6500 K is then used to calculate the Doppler half-width and a-

parameter for all argon lines at this power and the Lorenztian

contribution is calculated with an iterative method:


Av (1/2) *AV + ( (1/4)*Av + Av ) (3)


where Av, is the total line width, AvL is the Lorentzian

component, and AvD is the Doppler component.

All line shift values given are shifted towards the

red(positive values). Negative values are blue shift. Tables

5 and 6 give similar data for Ar at 1.5 and 1.9 kW

respectively.

Figure 12 shows the relationship between , the average

distance of the electron from the nucleus and the average

shift for the 6p-4s, 5p-4s, and 4p-4s transitions of argon at






54

each of the three powers. The values for 6p, 5p, and 4p

levels are estimated using



= (1/2)a[3n2 1(1+1)] (4)



where a, is the Bohr radius, n is principle quantum number,

and 1 is the azimuthal quantum number.

In Tables 7,8, and 9, give the results for Fe at 1.1,

1.5, and 1.9 kW, respectively. For each table (power level),

a Doppler temperature and electron density are calculated.

The electron density was calculated using the Hp line (486.1

nm). The transition configuration, wavelength of transition,

line shift, FWHM, Doppler and Lorentzian contributions to the

line profile, and the a-parameter for each line profile

studied have been measured. The iron concentration was 1000

gg/ml for all runs. The Fe transition assignments were taken

from Fuhr, Martin, and Wiese (81).

In Table 10, the spectral characteristics of 25

additional argon lines obtained with the ICP (1.1 Kw) in the

near-IR and far visible are also given. A large number of

intense Ar lines emitted from the ICP and the HCL occur in

this region. It contains exclusively 4p-4s transitions.

In Table 11, the final set of data presents line shifts

of calcium, barium, and strontium in a 1.1 Kw ICP are given.

These lines are used to investigate the relevance of line

shifts in analytical spectroscopy. We consider each element







55

to have minimal isotope effects (20Ca4 (96.97%), ,,Sr88 (82.56%),

s6Ba'3 (71.66%)); there are no hyperfine structures on the major

isotopes. The electron density and Doppler temperature are

considered to be the same as those in table 4 (Fe at 1.1 kW).

Emission from these ions are typically blue shifted.



Conclusion

The purpose of this chapter is to report the results of

measurements of shifts in line positions that are a function

of pressure broadening, in particular, Stark broadening. Fe

shows red shifts are on the order of 0.01-0.03 cm'. Line

shift studies on Ca, Sr, and Ba resulted in shifts of the same

absolute value as those of Fe but containing a number of blue

shifts. Line shifts for Ar in the visible and ultraviolet are

equivalent to or greater than those for those for Fe. The

shifts for 6p-4s transition is largest, followed by the 5p-4s

transitions and the 4p-4s transitions. We also show that line

shifts for Ar increase with power. Line shifts are shown to

increase as the average distance of the electron from the

nucleus radius increases provided the lower (ground) state is

the same for all three transitions.

From our measurements, line shifts should have little

consequence to the analytical spectroscopist. The data from

Ca, Ba, Sr, and Fe shows that the average shift is on the

order of 0.01-0.03 cm'. Line positions of this accuracy are

not readily measurable by most analytical systems.

























Figure 7. Only one of Pb's three isotopes has any hyperfine
structure (,2Pb) In the ICP, this isotope structure would
be convoluted into one asymmetric line profile making line
shift measurements of each line extremely difficult. (Top x-
axis is A and the bottom is in wavenumbers, cm').






























1.50



I-
S1.00
Z

Z

0.5








27140 27141

WAVENUMBER





















Figure 8. The hyperfine structure of ,7Ta81 taken from a low
pressure HCL. The higher pressure and temperature of the ICP
would make line selection difficult. (top x-axis is A and the
bottom is cm')










































WAVENUMBER




















Figure 9. Typical line profile of Ar I emission from the ICP.
(top x-axis is A and the bottom is cm )









61












42p2. 42p1 42p0






1.50





1.00




U,
Z 05





0.0(
S .. 0 ..-----------


23795 23800 23805

WAVENUMBER




















Figure 10. Line profile of Fe emission (1000 ppm) from
analytical zone in ICP. (Top axis is in A and the bottom is
cm )




































WAVENUMBER




















Figure 11. Diagram outlining the optical configuration of
light sources (ICP and HCL) and optical components outside
LAFTS.














NEUTRAL
DENS I TY


LENS


0


FH Ce
sT crro4df.L


ICP
HCL
















Figure 12. The relationship between , the average atomic
radius, and the shift in the line position is demonstrated for
1.1 kW (*), 1.5 kW (+), and 1.9 kW (A). The value is for
the upper level and should be multiplied by the constant a,
(Bohr radius) for the correct approximation.







67








1.0



,--0.8


E
, 0.6

7-
U-


LLJ
Z

0.2



0.0
20 30 40 50 60
RELATIVE ENERGY LEVEL RADIUS






68

Table 4. The transition, wavelength of transition, 1, line
shift, Al, full width at half maximum (FWHM), AXT, Doppler
halfwidth, AAD, Lorentzian halfwidth, AXL, and the a- parameter
of Ar emission from a 1.1 kW ICP. The Doppler temperature is
6500K. Positive shift values are red-shifted. Below each
general transition (6p-4s, 5p-4s, 4p-4s) the average shift and
standard deviation is tabulated.



Transition X AXA AXT AAD AL a
(A) (cm) (cm ) (cm') (cm )



6p[1.5]2-4s[1.512 3555. 0.3446 0.6335 0.255 0.53 1.73

6p[0.5]o-4s[1.5], 3607. 0.3136 0.6385 0.251 0.54 1.7

AVERAGE SHIFT (6p-4s) 0.3291 (0.0219)



5p[0.5],-4s[1.5]2 3950. 0.1242 0.4405 0.229 0.32 1.16

5p'[1.5]2-4s[l.5], 4045. 0.1038 0.3972 0.233 0.26 0.93

5p[1.5]2-4s[1.5]2 4159. 0.1154 0.4270 0.218 0.31 1.18

5p[1.5]1-4s[1.5]2 4165. 0.1104 0.4385 0.218 0.33 1.26

5p'[0.5]1-4s'[0.5]o 4183. 0.1383 0.4451 0.217 0.34 1.30

5p[2.5]2-4s[1.512 4191. 0.0957 0.3788 0.216 0.25 0.96

5p'[1.5],-4s'[0.5]0 4192. 0.0984 0.3849 0.216 0.26 0.99

5p[0.5]0-4s[1.5], 4199. 0.1243 0.4011 0.216 0.28 1.08

5p[2.5]3-4s[1.5]2 4201. 0.1063 0.5264 0.216 0.44 1.69

5p[0.51]-4s[1.5]2 4252. 0.0807 0.3280 0.213 0.19 0.74

5p[1.5]2-4s[1.5], 4267. 0.1417 0.3746 0.212 0.25 0.97

5p[1.5],-4s[1.5], 4273. 0.1046 0.3578 0.212 0.23 0.90

5p[2.5]2-4s[1.5], 4301. 0.0976 0.3545 0.211 0.23 0.91


5p'[1.5]2-4s'[0.5], 4334. 0.0972


0.4147 0.209


0.31 1.23









Table 4--continued.

5p'[0.5]i-4s'[0.5], 4336. 0.0773

5p'[1.5],-4s'[0.5]i 4346. 0.1284

5p[0.5]o-4s'[0.5], 4511. 0.1179

AVERAGE SHIFT (5p-4s) 0.1095


0.5441 0.209

0.4080 0.209

0.3613 0.201

(0.0182)


4p'[0.5],-4s[1.5]2 6967. 0.0179 0.1698 0.130

4p'[1.5]2-4s[1.5]2 7069. 0.0176 0.1610 0.128

AVERAGE SHIFT (4p-4s) 0.0177 (0.0002)


0.46

0.30

0.25


0.07

0.06


1.83

1.19

1.03


0.44

0.39






70

Table 5. The transition type, wavelength of transition,l,
line shift, Al,, full width at half maximum (FWHM) ,AlT, Doppler
halfwidth, AAD, Lorentzian halfwidth, AIL, and the a-parameter
of Ar emission from 1.5 kW ICP. The Doppler temperature is
6800 K. Positive shift values are red-shifted. Below each
general transition (6p-4s, 5p-4s, 4p-4s) the average shift and
standard deviation are tabulated.



Transition X AX. AXT AA, AlL a
(A) (cm ) (cm ) (cm ) (cm1)


6p[1.5]2-4s[1.5]2 3555 0.5442 0.7807 0.261 0.69 2.20

6p[0.5]0-4s[1.5], 3607 1.178 0.3333 0.257 0.13 0.42

AVERAGE SHIFT (6p-4s) 0.8610 (0.4481)



5p[0.5]i-4s[1.5]2 3950 0.1622 0.5703 0.235 0.47 1.66

5p'[1.5]2-4s[1.5], 4045 0.1383 0.6131 0.229 0.53 1.92

5p[1.512-4s[1.5]2 4159 0.1703 0.4494 0.223 0.34 1.27

5p[1.5],-4s[1.5]2 4165 0.1357 0.6745 0.222 0.60 2.24

5p' [0.5],-4s' [0.5]o 4183 0.1715 0.5064 0.222 0.41 1.54

5p[2.5]2-4s[1.5]2 4191 0.1700 0.4185 0.221 0.30 1.13

5p' [1.5],-4s' [0.5]o 4192 0.1578 0.4338 0.221 0.32 1.20

5p[0.5]0-4s[1.5]i 4199 0.1970 0.5339 0.221 0.44 1.66

5p[2.5]3-4s[1.5], 4201 0.1528 0.4406 0.221 0.33 1.24

5p[0.5],-4s[1.5]2 4252 0.0811 0.4219 0.218 0.31 1.18

5p[1.5],-4s[1.5], 4267 0.1708 0.4779 0.217 0.38 1.45

5p[1.5],-4s[1.5], 4273 0.1606 0.4202 0.217 0.31 1.19

5p[2.5]2-4s[1.5]1 4301 0.1562 0.5351 0.216 0.45 1.74


0.1219 0.8084 0.214 0.75


5p'[1.512-4s'[0.5], 4334


2.41






71

Table 5--continued.

5p' [0.5],-4s' [0.5], 4336 0.1350 0.5421 0.213 0.44 1.77

5p' [1.5]1-4s'[0.5], 4346 0.1797 0.5264 0.213 0.44 1.71

5p[0.5]0-4s' [0.5], 4511 0.1979 0.4068 0.206 0.30 1.21

AVERAGE SHIFT (5p-4s) 0.1564 (0.0283)



4p' [0.511-4s[1.5]2 6967 0.0204 0.1801 0.133 0.08 0.499

4p' [1.5]2-4s[1.5]2 7069 0.0305 0.1801 0.131 0.09 0.571

AVERAGE SHIFT (4p-4s) 0.0254 (0.0071)






72

Table 6.The transition type, wavelength of transition, l,line
shift, AA,, full width at half maximum (FWHM), AIT, Doppler
contribution, AID, Lorentzian contribution, AIL, and the a-
parameter of Ar emission from a 1.9 kW ICP. The Doppler
temperature is 7100 K. Positive shift values are redshifted.
Below each general transition (6p-4s, 5p-4s, 4p-4s) the
average shift and standard deviation are tabulated.




Transition I Als AIr AID AIL a
(A) (cm') (cm') (cm') (cm')


6p[1.512-4s[1.5]2 3555 0.6432 0.5962 0.266 0.48 1.49

6p[0.5]0-4s[1.5]i 3607 1.165 0.3068 0.263 0.08 0.25

AVERAGE SHIFT (6p-4s) 0.9040 (3680)



5p[0.5],-4s[1.5]2 3950 0.2409 0.5762 0.239 0.48 1.66

5p'[1.5]2-4s[1.5], 4045 0.1869 0.5261 0.234 0.42 1.49

5p[1.5]2-4s[1.5]2 4159 0.2064 0.5011 0.227 0.40 1.46

5p[1.5],-4s[1.5]2 4165 0.1894 0.4608 0.227 0.35 1.28

5p'[0.5],-4s'[0.5]0 4183 0.2152 0.7554 0.226 0.69 2.53

5p[2.5]2-4s[1.5]2 4191 0.1635 0.4865 0.226 0.38 1.40

5p'[1.5]1-4s'[0.5]o 4192 0.1587 0.4689 0.226 0.36 1.32

5p[0.5]0-4s[1.5], 4199 0.2129 0.8689 0.226 0.81 2.98

5p[2.5]3-4s[1.5]2 4201 0.1765 1.2200 0.225 1.18 4.35

5p[0.5],-4s[1.5]2 4252 0.1812 0.4081 0.223 0.29 1.08

5p[1.5]2-4s[1.5]1 4267 0.2300 0.5035 0.222 0.40 1.50

5p[1.5]1-4s[1.5], 4273 0.1630 0.5839 0.222 0.50 1.87

5p[2.512-4s[1.5]i 4301 0.1952 0.4673 0.220 0.36 1.36








Table 6--continued.

5p'[1.5]2-4s'[0.5], 4334 0.1699 0.5565 0.218 0.47 1.79

5p'[0.5]i-4s'[0.5]1 4336 0.2405 0.7283 0.218 0.66 2.51

5p'[1.5],-4s'[0.5], 4346 0.1662 0.4559 0.218 0.35 1.33

5p[0.5]0-4s'[0.5] 4511 0.2759 0.6060 0.210 0.53 2.09

AVERAGE SHIFT (5p-4s) 0.1983 (0.0335)



4p'[0.5],-4s[1.5]2 6967 0.0261 0.1967 0.136 0.10 0.612

4p'[1.5]2-4s[1.5]2 7069 0.0261 0.2098 0.134 0.13 0.807

AVERAGE SHIFT (4p-4s) 0.0261 (0.000)






74

Table 7. The transition type, wavelength of transition, X,
line shift, Als, full width at half maximum (FWHM) AX,
Doppler contribution, AAI, Lorentzian contribution, AAL, and
the a-parameter of Fe emission from a 1.1 kW ICP. The Doppler
temperature is 6000 K and the electron density 1.20 x 10 cm .
Positive shift values are red-shifted. For this table, the
average line shift is 0.0185 cm' with a standard deviation of
0.0062 cm The concentration of iron aspirated into the ICP
was 1000 Ag/ml.



Transition X AIA AXI AAD AAL a
(A) (cm ) (cm ) (cm ) (cm1)


aD z4F II 2755.73 0.0452 0.3394 0.274 0.12 0.364

a5D ySF I 2966.90 0.0181 0.3200 0.254 0.12 0.392

a5D y5D I 2994.43 0.0220 0.3144 0.252 0.11 0.363

a5D ysD I 3021.07 0.0232 0.3091 0.249 0.11 0.366

a5D ySDO I 3047.60 0.0201 0.3265 0.248 0.14 0.470

a5D ySDO I 3059.09 0.0172 0.3474 0.247 0.17 0.573

asD zP I 3440.99 0.0154 0.2646 0.219 0.08 0.303

a5F Z3G I 3565.38 0.0190 0.2605 0.212 0.09 0.353

a5F zSGo I 3581.19 0.0194 0.2601 0.211 0.09 0.355

a F zsGo I 3608.86 0.0168 0.2599 0.209 0.09 0.357

aSF zG" I 3618.77 0.0199 0.2849 0.208 0.13 0.518

a5F z2Go I 3631.46 0.0175 0.2555 0.208 0.09 0.360

a F z'G I 3647.84 0.0183 0.2784 0.207 0.12 0.482

a D zFo I 3719.93 0.0127 0.2447 0.203 0.08 0.328

a D z5F I 3745.56 0.0112 0.2470 0.201 0.08 0.330

a5D z25F I 3748.26 0.0119 0.2472 0.201 0.08 0.330

aSF ySF I 3758.23 0.0187 0.2466 0.201 0.08 0.331









Table 7--continued.

a5F ySF I 3763.79 0.0179 0.2460 0.200 0.08 0.332

a3F y3D I 3815.84 0.0221 0.2477 0.197 0.09 0.378

a5F y5D I 3820.43 0.0184 0.2574 0.197 0.11 0.463

a5F ySD I 3825.88 0.0194 0.2439 0.197 0.09 0.374

a5F y5D0 I 3834.22 0.0165 0.2437 0.197 0.09 0.380

a5D zs5D I 3859.91 0.0122 0.2368 0.196 0.08 0.340

a5D z25D I 3886.28 0.0113 0.2324 0.194 0.07 0.299

a3F y3F I 4045.82 0.0192 0.3011 0.186 0.19 0.847

a3F y3Fo I 4063.60 0.0222 0.2263 0.186 0.08 0.358

a3F z5G" I 4383.55 0.0148 0.2173 0.172 0.08 0.386

a3F zSGO I 4404.75 0.0165 0.2151 0.171 0.08 0.388







76

Table 8. The transition type, wavelength of transition, X,
line shift, AAs, full width at half maximum (FWHM), AXT,
Doppler, AAD, Lorentzian AX,, and the a-parameter of Fe
emission from a 1.5 kW ICP. The Doppler temperature is 6200
K and the electron density is 2.05 x 105 cm Positive shift
values are red-shifted. The average shift value for this
table is 0.0188 cm' with a standard deviation of 0.0066 cm'.
1000 gg/ml Fe solution was aspirated into the ICP.



Transition I AX, AAh AXD AXL a

(A) (cm') (cm") (cm') (cm')



a4D Z4FO II 2755.73 0.0431 0.3615 0.273 0.15 0.455

a5D y5F I 2966.90 0.0173 0.3290 0.254 0.13 0.425

a5D y5Do I 2994.43 0.0255 0.3349 0.252 0.14 0.462

a5D ySD I 3021.07 0.0148 0.3115 0.250 0.11 0.366

a5D ySDo I 3047.60 0.0277 0.3251 0.247 0.14 0.470

a5D y5D I 3059.09 0.0124 0.3262 0.247 0.14 0.472

a5D z5aP I 3440.99 0.0214 0.2827 0.219 0.11 0.417

a5F zSGO I 3565.38 0.0218 0.2766 0.212 0.10 0.384

a5F z5Go I 3581.19 0.0209 0.2731 0.211 0.11 0.434

a5F zSG I 3608.86 0.0199 0.2716 0.209 0.11 0.437

a5F z'5G I 3618.77 0.0198 0.2689 0.209 0.11 0.439

a5F z2Go I 3631.46 0.0188 0.2786 0.208 0.12 0.480

aF z5Go I 3647.84 0.0160 0.2639 0.207 0.10 0.402

a5D Z 5F I 3719.93 0.0134 0.2963 0.203 0.16 0.656

a5D z2Fo I 3745.56 0.0146 0.2497 0.202 0.09 0.371

a5D z5F" I 3748.26 0.0097 0.2575 0.201 0.10 0.431

a*F y5F I 3758.23 0.0232 0.3186 0.201 0.19 0.787









Table 8--continued.

a5F y5F I 3763.79 0.0132 0.2509 0.201 0.09 0.373

a3F y3DO I 3815.84 0.0205 0.2625 0.198 0.11 0.462

a5F y'D I 3820.43 0.0181 0.2820 0.197 0.14 0.589

a5F y5D I 3825.88 0.0191 0.2542 0.197 0.10 0.422

aSF ySD I 3834.22 0.0196 0.2476 0.197 0.09 0.380

ta5D z'5D I 3859.91 0.0108 0.2429 0.195 0.09 0.382

a5D zD O I 3886.28 0.0231 0.2651 0.194 0.12 0.514

a3F y3Fo I 4045.82 0.0137 0.2445 0.186 0.10 0.446

a3F y3Fo I 4063.60 0.0115 0.2707 0.186 0.14 0.627

a3F z25G I 4383.55 0.0146 0.2208 0.172 0.09 0.434

a3F z5Go I 4404.75 0.0211 0.2529 0.171 0.14 0.679







78

Table 9. The transition type, wavelength of transition,l,
line shift, Als, full width at half maximum (FWHM), Al,,
Doppler contribution, AAD, Lorentzian contribution, AAL, and
the a-parameter of Fe emission from a 1.9 kW ICP. The Doppler
temperature is 6500 K and the electron density is 2.85 x 1015
cm Positive shift values are red-shifted. The average shift
for all lines listed in this table is 0.0176 cm with a
standard deviation of 0.0132 cm'. The concentration of Fe
aspirated into the ICP was 1000 Ag/ml.


Transition


a'D z4F I

a5D ySFO I
asD y5Do I


a5D ys5D I
a5D ySDO I

aSD yDo I
a5D zyP0 I


a5F z23G I

aF zSGo I

aF zSGo I

a5F zSG I

a5F z5G I

aF zSGo I


aD zF" I

a5D z5F I

aD zIF" I

a5F y5F" I

a5F y FO I


(A)

2755.73

2966.90

2994.43

3021.07

3047.60

3059.09

3440.99

3565.38

3581.19

3608.86

3618.77

3631.46

3647.84

3719.93

3745.56

3748.26

3758.23

3763.79


AA
(cm)

0.0390

0.0249

0.0544

0.0116

0.0226

0.0157

-0.0109

0.0305

0.0221

0.0188

0.0105

0.0386

0.0101

0.0164

0.0210

0.0140

0.0188

0.0193


AXT
(cm )

0.5696

0.3240

0.4185

0.3210

0.3560

0.3155

0.3246

0.2661

0.2872

0.2592

0.2641

0.2796

0.2652

0.2617

0.3310

0.2507

0.3103

0.2912


A1D
(cm )


AIL
(cm )


0.284 0.42 1.23


0.264

0.262

0.259

0.257

0.256

0.228

0.220

0.219

0.217

0.217

0.216

0.215

0.211

0.209

0.209

0.207

0.208


0.11

0.25

0.11

0.17

0.11

0.16

0.09

0.12

0.08

0.09

0.11

0.09

0.09

0.20

0.08

0.17

0.14


0.36

0.79

0.35

0.55

0.35

0.58

0.34

0.45

0.31

0.34

0.42

0.35

0.35

0.79

0.32

0.68

0.56









Table 9--continued.

a3F y3Do I 3815.84 0.0135 0.3395 0.205 0.21 0.85

a5F y5D I 3820.43 0.0176 0.2575 0.205 0.09 0.36

a5F y'5D I 3825.88 0.0152 0.2599 0.205 0.10 0.41

a5F y5D I 3834.22 0.0266 0.4480 0.204 0.35 1.42

a5D z25D I 3859.91 0.0089 0.2895 0.203 0.15 0.61

a5D z5D" I 3886.28 0.0081 0.2760 0.202 0.13 0.53

a3F y3Fo I 4045.82 0.0118 0.2508 0.194 0.10 0.43

a3F y3Fo I 4063.60 0.0076 0.2509 0.193 0.10 0.43

a3F z2G" I 4383.55 0.0165 0.2426 0.178 0.11 0.51

a3F zG I 4404.75 -0.0102 0.2338 0.178 0.10 0.47







80

Table 10. The transition type, wavelength of transition, 1,
line shift, Als, full width at half maximum (FWHM), A4T,
Doppler contribution, AA,, Lorentzian contribution, AAL, and
the a-parameter of Ar emission from a 1.1 kW ICP. The Doppler
temperature is 6500 K. Positive shift values are red-shifted.


Transition


1 AXA AXT
(A) (cm ) (cm )


A(c AX(c
(cm ) (cm )


4p[0.5],-4s'[0.5]o

4p[2.5]2-4s' [0.5]1

4p[0.5],-4s[1.5],

4p[1.5],-4s'[0.5]1

4p[1.5],-4s'[0.5],

4p[0.5],-4s [1.512

4p[1.5],-4s'[0.5]o

4p' [1.5],-4s' [0.5],

4p[2.512-4s[1.5],

4p' [1.5]2-4s' [0.5],

4p' [0.5],-4s' [0.5],
4p[2.5],-4s[1.512

4p[1.5],-4s[0.5],

4p[2.5]2-4s[1.5]2

4p[1.512-4s[1.5],

4p'[1.5]i-4s'[0.5]o

4p'[0.5],-4s'[0.5]o

4p[1.5]1-4s[1.512

4p[1.5]2-4s[1.5]2

4p[0.5]0-4s[1.5],


10472. 0.0159

9787. 0.0172

9660. 0.0121

9356 0.0050

9227. 0.0189

9125. 0.0042

8670. 0.0268

8523. 0.0151

8427. 0.0016

8410. 0.0164

8266. 0.0183

8117.-0.011

8105. 0.0217

8016. 0.0216

8008. 0.0248

7950. 0.0254

7726. 0.0269

7725. 0.0230

7637. 0.0463

7515. 0.0280


0.1365

0.1934

0.1389

0.1963

0.1999

0.1419

0.1506

0.2092

0.1677

0.2250

0.2125

0.1833

0.1686

0.1602

0.1652

0.1610

0.1658

0.1710

0.1683

0.1851


0.0865

0.0926

0.0938

0.0969

0.0983

0.0994

0.1046

0.1064

0.1076

0.1078

0.1097

0.1117

0.1118

0.1131

0.1132

0.1140

0.1173

0.1174

0.1187

0.1206


0.08

0.15

0.08

0.15

0.15

0.07

0.08

0.16

0.10

0.18

0.16

0.12

0.10

0.08

0.09

0.08

0.08

0.09

0.09

0.11


0.77

1.35

0.71

1.29

1.27

0.59

0.64

1.25

0.77

1.39

1.21

0.89

0.74

0.59

0.66

0.58

0.57

0.64

0.63

0.79







81

Table 10--continued.

4p'[0.5]o-4s'[0.5], 7505. 0.0301 0.2297 0.1208 0.17 1.17

4p'[1.5]2-4s[1.5], 7386. 0.0224 0.1785 0.1227 0.10 0.65

4p'[0.5]1-4s[1.5], 7274. 0.0255 0.1806 0.1246 0.10 0.68

4p'[1.5]2-4s[1.5]2 7069. 0.0241 0.1756 0.1301 0.09 0.57

4p' [0.5],-4s[1.512 6967. 0.0260 0.1783 0.1301 0.09 0.57







82

Table 11. Line shifts, Av, of Ca, Sr, and Ba in 1.1 Kw ICP.
The wavenumber positions, v, for hollow cathode lamp
(standard, nonshifted) verses the ICP emission. The Doppler
temperature is 6000 K and the electron density is 1.20 x 10
cm for the ICP. Wavenumber positions given are calibrated
against a Ne I standard for absolute values. Negative values
are blue-shifted.


ELEMENT


v (HCL)
(cm')


v (ICP)
(cm' )


Avs
(cm )
(cm)


Barium II
Barium II
Calcium II
Calcium II
Strontium II


20261.5370
21952.3718
25191.5197
25414.4307
24516.5792


20261.5479
21952.3975
25191.5241
25414.4147
24516.5870


-0.0109
-0.0257
-0.0044
0.0160
-0.0078












CHAPTER 6
A VARIABLE BANDPASS FILTER FOR ULTRAVIOLET/VISIBLE
FOURIER TRANSFORM SPECTROSCOPY



Introduction

Fourier Transform Spectroscopy (FTS) has several

advantages over grating and prism spectrometers. Included in

these are wavenumber precision and accuracy, intensity

precision, and resolving power. Detrimental to the FT

technique is the multiplex disadvantage (7,82), which arises

from shot and/or flicker noises at detected wavelengths other

than the analytical wavelength hitting the detector causing a

decreased signal-to-noise ratio compared to a single slit

scanning dispersive spectrometer system. This disadvantage is

more prominent in FT-UV/VIS then in FT-IR studies. In FT

studies involving a continuum source, this disadvantage can

become severe. The Inductively Coupled Plasma (ICP), Direct

Current Plasma (DCP), and absorbance techniques utilizing a

white light are examples of light sources that can magnify the

multiplex disadvantage.

In our research (83), the Los Alamos Fourier Transform

Spectrometer is used to measure Stark shifts in wavenumber

positions of atoms and ions from a 27.12 MHz atmospheric

pressure argon Inductively Coupled Plasma (ICP). The line






84

positions of atoms and ions in the ICP are compared to the

unshifted line positions of the same species in a commercial

hollow cathode lamp (HCL). To achieve adequate wavenumber

positions (1 part in 4x10'), the lines analyzed need good

signal-to-noise ratios (SNRs). Both the HCL and ICP emit

numerous spectral lines. The ICP, being a hotter source, has

considerable ion line emission. The HCL, with Doppler

temperatures typically 1/10 that of the ICP (6000 K vs 600 K)

has considerable atom line emission, particularly the fill gas

emission in the near infrared. In such measurements of Stark

shifted line positions, we find that the data set available is

small for two reasons. First, the difference in source

emission intensities between the same lines emitted from the

ICP and the HCL reduces the number of lines available for

comparison. Second, the multiplex disadvantage will

substantially reduce the SNR of all lines.

A single filter with two features is desired. First,

the ability to vary the bandpass of the filter over a wide

range. Second, the ability to select the wavenumber region

observed over a wide range. Optical filters do not meet our

specifications for two reasons; the bandpass allowed rarely

matches the needs desired, and the transmission of these

optical filters often increases in the near IR resulting in

excessive shot noise from spectral lines/bands in this region.

Our sources (HCL, ICP) use inert gases having strong emission

in the near IR region, resulting in a serious deterioration






85

of the SNR for lines observed in the ultraviolet and visible.

Hirschfeld and Chase (84) and Hirschfeld and Milanovich

(85) summarized several techniques that could be used to

overcome problems associated with excess light in FT-Raman

experiments. One of these is the use of a nonscanning

subtractive dispersive double monochromator. This filtering

approach has been shown to be a valuable tool for micro-Raman

work (85). We have adapted this filter for use as a bandpass

filter in FT-continuum source work. We demonstrate the

feasibility of this filter using a white light source.



Experimental

Figure 13 shows the configuration of the experiment. A

Bomem DA-3 is used at low resolution (30 cm') for this study.

The gratings (Gl, G2) are manufactured by Instrumental SA.

Both are blazed for 700 nm (1200 grooves/mm, 25 mm x 25 mm).

F numbers of the input lens (L3) and FTS are matched. The

white light source (W.L.), gratings, lens-1 (LI) and lens-3

(L3) are held constant for all experiments. The mirrors

(M1,M2), lens-2 (L2) aperture-2 (A2) are varied for selective

purposes. L1 serves to collimate the light source. Al (5 mm

diameter) and A3 (3-10 mm diameter) are circular apertures.

For Al, we use an aperture as opposed to a slit to maintain

some of the throughput advantage of the FTS for emission

studies. Since we are looking at relatively large bandpasses

(> 500 cm'), the loss in spectral resolution of an aperture






86

verses a slit is not significant. G1 and G2 are mounted back

to back. A2 serves to block unwanted light and to let through

a preset band. A2 is rectangular; A2 and L2 are moved

laterally to select both wavenumber region and bandpass. Ml

and M2 are matched concave mirrors, where M1 transfer the

range of nearly collimated radiation selected by A2-L2 and M2

transfers that range of nearly collimated light on to the

surface of G2 which further disperses and transfers the band

of nearly collimated radiation into L3-A3 where A3 is the

aperture to the FTS.

The bandpass can be adjusted by changing the diameter of

A2 and/or the distance of A2 from GI. Narrower bandpasses can

be achieved with smaller apertures (Al, A2) at greater

distances (A2 from Gl). Decreasing the apertures (Al, A2)

decreased the photon flux entering the FTS and therefore

decreases the multiplex disadvantage. Moving A2-L2

perpendicular to the nearly collimated light reaching M1

resulted in a change in the central wavelength of the band

passed.

The distances between optical components varied slightly

depending on central wavenumber and bandpass selection.

Approximate dimensions are as follows: L1-A1, 1 cm; Al-Grating

1, 11 cm; Grating 1-A2, 12 cm; M1-M2, 26 cm; Grating 2-L3, 12

cm.








Results

Figure 14 shows two spectra (25 scans) of the filtered

white light source with a 2000 cm1 bandpass. The strong

narrow line in both is the He-Ne (6328 A) laser leakage and is

used as a visual reference for alignment of the optics Al-G1-

A2-L2-Ml-M2-G2-L3-A3. For this demonstration the central

wavenumber shifts from approximately 15000 cm1 to 16000 cm'.

Figure 15 shows a narrower bandpass (approximately 1000

cm') that varied from 16500 to 18700 cm' in four windows. The

gratings performance fell off above 20000 cm' and visual

alignment of the optics in Figure 13 below 13000 cm' becomes

difficult by eye; an infrared sensor is needed to see the

radiation. This justified the range (13000-19000 cm')

demonstrated here.



Conclusion

A bandpass filter which varied both wavenumber position

and bandpass and provide a limited spectral range is

demonstrated. The bandpass filter is being developed the

filter for FT-UV/VIS source studies. The filter removed

unwanted light which affected the quality of broad band

multiple line spectra. Removal of the prominent lines of neon

or argon appearing in the 600-900 nm range in hollow cathode

lamp (HCL) emission is also of interest to us. These lines

are often the most intense in HCL emission studies. Optical

filters have two disadvantages which often arise in studies of






88

this sort. First the bandpass of the optical filter may not

match the desired need, especially in the ultraviolet.

Second, transmission of these filters often increases in the

NIR (above 900 nm) and allows the fill gas emission

(particularly Ar) to enter the FTS. Our aim is to replace

optical filters with a simple filter that allows variation in

both position and bandpass of the filter.

We should note that a variety of derivations of this

filter are possible. For example, replacing M1 and M2 with

parabolic mirrors would allow the removal of L2. Ml and M2

would collimate and reflect the radiation to G2 eliminating

the need for L2.



























Figure 13. Outline of bandpass filter. (W.L. = white light,
A = aperture, G = grating)














I Mirror 2


/2


Grating



Grating 1


White
I Light

Al Lens 1


ljrror 1


Lens 2
Lens 2


Lens 3
Lens 3


Ow





















Figure 14. 2000 cm" bandpass of filtered white light source
centered at 15000 cm1 (top) and 16000 cm' (bottom). Sharp line
(6328 A) is He-Ne laser leakage from Bomem FTS alignment
system.