Direct analysis of liquid and solid microsamples using a capacitively coupled microwave plasma atomic emission spectrometer

MISSING IMAGE

Material Information

Title:
Direct analysis of liquid and solid microsamples using a capacitively coupled microwave plasma atomic emission spectrometer
Physical Description:
Book
Creator:
Croslyn, Andrea E., 1972-
Publication Date:

Record Information

Rights Management:
All applicable rights reserved by the source institution and holding location.
Resource Identifier:
aleph - 30018738
oclc - 40833660
System ID:
AA00022903:00001

Table of Contents
    Title Page
        Page i
    Dedication
        Page ii
    Acknowledgement
        Page iii
        Page iv
    Table of Contents
        Page v
        Page vi
    List of Tables
        Page vii
    List of Figures
        Page viii
        Page ix
    List of acronyms
        Page x
        Page xi
    Abstract
        Page xii
        Page xiii
    Chapter 1. Introduction to plasmas
        Page 1
        Page 2
        Page 3
        Page 4
        Page 5
        Page 6
        Page 7
        Page 8
        Page 9
        Page 10
        Page 11
        Page 12
        Page 13
        Page 14
        Page 15
        Page 16
        Page 17
        Page 18
        Page 19
        Page 20
        Page 21
        Page 22
        Page 23
        Page 24
        Page 25
        Page 26
        Page 27
        Page 28
        Page 29
        Page 30
    Chapter 2. Instrumentation
        Page 31
        Page 32
        Page 33
        Page 34
        Page 35
        Page 36
        Page 37
        Page 38
        Page 39
        Page 40
        Page 41
        Page 42
        Page 43
        Page 44
        Page 45
        Page 46
        Page 47
        Page 48
        Page 49
        Page 50
        Page 51
        Page 52
        Page 53
        Page 54
        Page 55
        Page 56
        Page 57
        Page 58
        Page 59
        Page 60
        Page 61
        Page 62
        Page 63
        Page 64
        Page 65
        Page 66
        Page 67
        Page 68
        Page 69
        Page 70
    Chapter 3. Liquids analysis
        Page 71
        Page 72
        Page 73
        Page 74
        Page 75
        Page 76
        Page 77
        Page 78
        Page 79
        Page 80
        Page 81
        Page 82
        Page 83
        Page 84
        Page 85
        Page 86
        Page 87
        Page 88
        Page 89
        Page 90
        Page 91
        Page 92
        Page 93
        Page 94
        Page 95
        Page 96
        Page 97
        Page 98
        Page 99
        Page 100
        Page 101
        Page 102
        Page 103
    Chapter 4. Solids analysis
        Page 104
        Page 105
        Page 106
        Page 107
        Page 108
        Page 109
        Page 110
        Page 111
        Page 112
        Page 113
        Page 114
        Page 115
    Chapter 5. Conclusions and future work
        Page 116
        Page 117
        Page 118
    List of references
        Page 119
        Page 120
        Page 121
        Page 122
        Page 123
        Page 124
        Page 125
        Page 126
    Biographical sketch
        Page 127
        Page 128
        Page 129
Full Text










DIRECT ANALYSIS OF LIQUID AND SOLID MICROSAMPLES USING
A CAPACITIVELY COUPLED MICROWAVE PLASMA
ATOMIC EMISSION SPECTROMETER














By

ANDREA E. CROSLYN


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


1998






























This dissertation is dedicated to my parents, Robert and Karen, and my husband Michael,
who have always been my very best friends.














ACKNOWLEDGMENTS

My most sincere appreciation must go to Jim Winefordner, who has not only been

an awe-inspiring mentor, but also a good friend. The working atmosphere that Jim

creates with the respect he shows to others and his personal dedication to science made

graduate school a rich and wonderful experience. His bathroom jokes aren't bad, either.

Special thanks also go to Ben Smith, whose friendship, patience, and breadth of

knowledge were a constant source of inspiration.

My appreciation also goes to members of the Winefordner group, particularly

those students who contributed to the project with their hard work. These include

undergraduates Mary Jane Gordon, Lydia Burberry, Rachel McCusker, and Jason

Wallace, and graduate student Mike Shepard. In addition, my gratitude also goes to the

many visiting scientists who aided with their advice and suggestions, including Igor

Gornushkin, Oleg Matveev, Kobus Visser, Piet Walters, Alexei Podshivalov, and Nico

Omenetto. I would also like to thank Jeanne Karably and Donna Balkom for their

invaluable knowledge and assistance, and for their friendship. Finally I would like to

thank Wendy Clevenger, Leslie King, Bryan Castle, Ricardo Aucelio, and Gretchen Potts

for their provocative scientific discussion and the occasional lazy lab break. Their

friendship continues to mean a great deal to me.

The support staff to the chemistry department has been an invaluable asset to this

research. The department is truly lucky to have such skilled people working towards the









furtherment of its research. In particular the machine shop (Dr. Sam Colgate, Joe

Shalosky, Gary Harding, Mike Herlevich, Todd Prox, Dailey Burch), electronics shop

(Steve Miles, Larry Harnly, and Joe Carusone), and glass-blowing shop (Joe Caruso)

have all contributed to this work. Not only their skill but also their willingness to help

educate the graduate students is much appreciated.

These acknowledgements would not be complete without including the sincere

gratitude I have for my parents, Robert and Karen Rieffel, who have eaten peanut butter

and driven around on bald tires so that I could have everything I ever needed or wanted.

My appreciation also goes to my in-laws Jim and Becky Croslyn, who have loved me like

their own daughter and who supported my decision to drag their son here and there and

everywhere in pursuit of my career. This brings me to my soul mate and husband

Michael, without whose constant support and love I would be the graduate student

equivalent of a pile of rubble. His good-naturedness and personal integrity have always

been a source of comfort and strength.

Financial support for this work was provided by the National Science Foundation

(STTR), the Engineering Research Center (ERC) for Particle Science and Technology at

the University of Florida (National Science Foundation Grant No. EEC-94-02989), the

Industrial Partners of the ERC, and the Thermo-Jarrel Ash Corporation for donation of

the echelle spectrometer.

Finally, I would like to thank God for his many blessings.














TABLE OF CONTENTS
page


ACKNOW LEDGM ENTS ............................................................................................ iii

LIST OF TABLES ............................................................................................................ vii

LIST OF FIGURES ................................................................................................... v iii

LIST OF ACRONYM S ................................................................................................. x

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

CHAPTERS

1 INTRODUCTION TO PLASM AS .......................................................................... 1

M icrowave Plasm as ................................................................................................. 1
The Capacitively Coupled M icrowave Plasm a ....................................... 6
Diagnostics and Characteristics ................................................ 8
Gas Introduction ......................................................................... 9
Liquid Introduction .................................................................. 10
Solid Introduction .......................................................................... 11
The M icrowave-Induced Plasm a ........................................................... 12
The Surface-Wave / Surfatron Microwave Plasma .............................. 14
The M icrowave Plasm a Torch ............................................................ 18
Other Plasmas and Electrothermal Atomization Sources ................................ 20
FANES ................................................................................................. 20
FAPES ................................................................................................. 23
Direct Current Arc Atomic Emission Spectrometry ........................... 25
Inductively Coupled Plasma Atomic Emission Spectrometry ........... 25
M icrowave Radiation and Safety .................................................................... 27

2 IN STRUM ENTATION ................................................................................... 31

W aveguides ...................................................................................................... 31
Plasma Gases .................................................................................................... 44
Electrodes ......................................................................................................... 51
Power Supply, M agnetron and Filam ent Transform er ...................................... 55
Echelle Spectrom eter ....................................................................................... 59









Plasm a Conditions for M ultielem ent Analysis ................................................ 66

3 LIQUID S ANALY SIS .................................................................................... 71

Introduction ...................................................................................................... 71
Experim ental ................................................................................................... 74
Results and D iscussion ..................................................................................... 77
Single Elem ent Calibration ................................................................. 77
M ultielem ent Calibration .................................................................... 88
Real W orld Sam ples .............................................................................. 90
Sw eat ....................................................................................... 90
Plant M aterial ........................................................................... 96
Conclusion ........................................................................................................... 102

4 SOLID S ANALYSIS .......................................................................................... 104

Introduction ......................................................................................................... 104
Experim ental ....................................................................................................... 106
Results and D iscussion ........................................................................................ 112
Conclusion ........................................................................................................... 114

5 CON CLU SION S AND FU TURE W ORK .......................................................... 116

LIST OF REFEREN CES ................................................................................................. 119

BIOGRAPHICAL SKETCH ........................................................................................... 127














LIST OF TABLES


Table page

1 Temperatures and Electron Number Densities for Microwave Plasmas .............. 9
2 Maximum Microwave Radiation Levels ............................................................ 30
3 Final System Com ponents .................................................................................. 32
4 Original System Com ponents .............................................................................. 32
5 Experimental Waveguide Dimensions ............................................................... 41
6 Characteristics of Standard Rectangular Waveguides ......................................... 43
7 Hydrogen to Helium Ratios ................................................................................ 50
8 Optimal Plasma Conditions for Multielement Analysis ...................................... 70
9 Experim ental Param eters ..................................................................................... 76
10 Line Equations for Single Element Calibration ................................................... 78
11 Figures of Merit for Single Element Calibration ................................................ 80
12 Comparison of Detection Limits (ppb) with FANES, ICP-AES, and DCP-AES ..... 82
13 Figures of Merit for ICP-MS-I and Comparison of Axial and Lateral Viewing ....... 89
14 Line Equations for Sodium .................................................................................. 93
15 Calculated Sodium Values in Sweat .................................................................... 94
16 Comparison of Na Concentrations in the Sweat of Normal Adults ..................... 95
17 Measured and Certified Concentrations in NIST Plant Samples ........................ 97
18 Calibrated Weights for 1 jtL in the Drummond Pipette ......................................... 107
19 Linear and Log-log Equations for SPEX G-7 Elements ......................................... 113
20 Figures of Merit and Comparisons with DC Arc and ICP-AES ............................. 114














LIST OF FIGURES

Figure page

1 Electric Fields ......................................................................................................... 3
2 M agnetic Fields ...................................................................................................... 3
3 Electrom agnetic W aves ......................................................................................... 4
4 M icrowave Transm ission Lines .............................................................................. 4
5 Capacitively Coupled M icrowave Plasm a ............................................................. 7
6 Beenakker Resonator Cavity ................................................................................. 13
7 Surfatron ............................................................................................................. 16
8 Surfaguide ........................................................................................................... 17
9 M icrowave Plasm a Torch ..................................................................................... 19
10 FANES ...................................................................................................................... 22
1 1 F A P E S ....................................................................................................................... 2 4
12 DC Arc ...................................................................................................................... 26
13 IC P ............................................................................................................................ 2 8
14 Schem atic of Final Instrum ent ............................................................................ 33
15 Schem atic of Original Instrum ent ....................................................................... 34
16 Photograph of Instrum ent .................................................................................... 35
17 W aveguide Dim ensions ....................................................................................... 37
18 W aveguide Field Configurations ........................................................................ 39
19 In-Phase Coupling ............................................................................................... 41
20 Photograph of Plasm as ........................................................................................ 48
21 Cup Temperature versus Time at 6 LPM He, 0.070 H2iHe Ratio ....................... 49
22 Cup Temperature versus Flow Rate for Various Applied Powers ........................ 49
23 Cup Temperature versus Applied Power at Various H2/He Ratios ..................... 50
24 Evolution of the Electrode ................................................................................... 52
25 Schematic of Power Supply ................................... 56
26 Schem atic of M agnetron ..................................................................................... 58
27 Electron Flow in the M agnetron ......................................................................... 58
28 Echelle-CID Diagram .......................................................................................... 60
29 Cross Section of a CID Pixel ................................................................................ 61
30 CID Im age ........................................................................................................... 63
31 Zoom ed CID Im age ............................................................................................ 64
32 Subarray of the CID ............................................................................................ 65
33 Translation of Observation Height in the Lateral View ...................................... 67
34 Translation of the X Axis in the Lateral View ...................................................... 67
35 Translation of the Y Axis in the Lateral View ...................................................... 68
36 Translation of the X Axis in the Axial View ........................................................ 69









37 Translation of the Y Axis in the Axial View ........................................................ 69
38 Calibration of Arsenic .......................................................................................... 83
39 Calibration of Cadm ium ..................................................................................... 83
40 Calibration of Chromium ..................................................................................... 84
41 Calibration of Copper .......................................................................................... 84
42 Calibration of Iron .............................................................................................. 85
43 Calibration of Lithium .......................................................................................... 85
44 Calibration of M anganese .................................................................................. 86
45 Calibration of Lead ............................................................................................... 86
46 Calibration of Phosphorus .................................................................................. 87
47 Calibration of Strontium ..................................................................................... 87
48 Calibration of Zinc ............................................................................................... 88
49 M ultielem ent Calibration .................................................................................... 89
50 Calibration of Sodium .......................................................................................... 93
51 Correlation Plot by M atrix ...................................................................................... 101
52 Correlation Plot by Elem ent ................................................................................... 101
53 Drumm ond Pipette for Solid Samples .................................................................... 108
54 Calibration of Group 2A Elem ents in Graphite ...................................................... 109
55 Calibration of Group 6B 8B Elem ents in Graphite .............................................. 109
56 Calibration of Group lB 2B Elem ents in Graphite .............................................. 110
57 Calibration of Group 3A 5A Elem ents in Graphite ............................................. 110
58 Calibration of Group M etalloid Elem ents in Graphite ........................................... 111
59 Calibration of Group Nonm etals in Graphite ......................................................... 111














LIST OF ACRONYMS


AAS atomic absorption spectrometry

AES atomic emission spectrometry

ANSI American National Standards Institute

CID charge injection device

CMP capacitively coupled microwave plasma

ETV electrothermal vaporization

FANES furnace atomization nonthermal excitation spectrometry

FAPES furnace atomization plasma excitation spectrometry

GFAAS graphite furnace atomic absorption spectrometry

GFAES graphite furnace atomic emission spectrometry

GHz gigahertz

ICP inductively coupled plasma

kV kilovolt

LOD limit of detection

LPM liters per minute

mA milliampere

MHz megahertz

IP microwave induced plasma

MOS metal-oxide-semiconductor









MPT microwave plasma torch

NIST National Institute of Standards and Technology

nm nanometer

PDA photodiode array

pm picometer

ppb parts per billion

ppm parts per million

rms root mean square

RSD relative standard deviation

SMA simultaneous multielement analysis

SRM standard reference material

TE transverse electric

TM transverse magnetic

TJA Thermo-Jarrell Ash

VAC volts alternating current

W watts














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

DIRECT ANALYSIS OF LIQUID AND SOLID MICROSAMPLES USING
A CAPACITIVELY COUPLED MICROWAVE PLASMA
ATOMIC EMISSION SPECTROMETER

By

Andrea E. Croslyn

December, 1998

Chairman: Prof. James D. Winefordner
Major Department: Chemistry

The ultimate goal of the research is the trace multielement determination of

microsamples. This is accomplished using a capacitively coupled microwave plasma

atomic emission spectrometer. The plasma is used as an atomization source, and the

resulting emission is analyzed with an echelle spectrometer with a charge injection

device for detection. A sample cup is built into the electrode on which the plasma is

generated which holds the discrete amount of the microsample to be analyzed. These

amounts are generally five microliters for liquids or one-half milligram of solid material.

The system has been evaluated by a series of optimization studies for such

parameters as applied power, plasma gas flow rates, and electrode design. Following

system optimization, a series of aqueous calibration standards were run for single

elements to evaluate the performance of the system. Detection limits are generally in the

low part-per-billion range with precision of less than 10 percent. In addition, samples









such as sweat and some certified reference materials have been analyzed to determine

the applicability of the technique to real samples. Preliminary results for real samples

were promising, although the accuracy of the method worsens near the detection limits.

Preliminary work in the area of direct solids analysis is also promising, with linear

calibration and detection limits in the low parts-per-million range. However, emission

signal reproducibility is a problem with the solid samples, and some of the more

refractory elements in both solids and liquids can not be analyzed using this system.

This technique is unique because of its ability to perform simultaneous

multielement analysis on solid or liquid discrete microsamples without the need to

change the system or alter the conditions in any way. For liquid samples, the technique

has detection limits similar to or slightly higher than similar atomic emission techniques.

For solids analysis, few techniques at present have attempted analyzing microsamples

without the need for sample preparation, with the additional capability of multielement

analysis. The capacitively coupled microwave plasma atomic emission spectrometer is a

promising tool in the analysis of real microsamples.














CHAPTER 1
INTRODUCTION TO PLASMAS AND ATOMIC EMISSION



Microwave Plasmas

Microwave plasmas have been used as sources for atomic spectroscopy since the

1970s. Several common forms of this plasma source exist, including the microwave-

induced plasma (MIP), the capacitively coupled microwave plasma (CMP), the surface-

wave or surfatron plasma, the microwave plasma torch (MPT), and some other unique

designs. Microwave plasmas provide a less expensive alternative to the ICP in terms of

both initial and operational costs. Although still approaching the precision and detection

limits of the ICP, microwave plasma systems are continually being improved in terms of

power handling capabilities, efficiency of coupling, and in the analysis of solid, liquid,

and gas samples. The MIP offers the best analytical capabilities in terms of detection

limits and precision in the analysis of gas samples, where it has served particularly well

as an element-selective detector for gas chromatography. The surfatron has also

performed well under a wide variety of operating conditions. The CMP has proven to be

a robust atomization source which can handle all forms of sample introduction. The

MPT, while a relatively new source, is making great progress in the area of liquid sample

analysis. As these systems continue to become more refined, the microwave plasma may

play a more prominent role in commercial atomic emission spectroscopy. The following

sections will describe the primary microwave plasma systems. The capacitively coupled









microwave plasma will be described in greater depth since it was used exclusively in the

work described later.

There are many excellent reviews on different aspects of microwave plasmas used

as sources for atomic emission spectrometry. Reviews which include microwave and

other plasmas can be found in the Atomic Spectroscopy Update in the Journal of

Analytical Atomic Spectroscopy, 1-6 Analytical Chemistry,7-9 the yearly review series on

microwave plasmas by Dahmen,10-16 and others. 17-21

Microwaves are simply electromagnetic waves in the frequency range from 300

MHz to 300 GHz. The microwave region is unique in the electromagnetic spectrum

because at these high frequencies conventional wiring and electronics will not work due

to high lead reactances and long transit times. This can also be described as the skin

effect, which is a phenomenon in which high frequency (microwave) current travels on

the outer surface, or skin, of a metal rather than penetrating it.22 Therefore, microwave

systems transfer energy by means such as antennas, waveguides, and coaxial cables.

Microwave systems are used in a wide variety of fields including space telemetry, radar

communications, and microwave heating.23

Microwaves have both an electric field component and a magnetic field

component. The electric field is the force created when two electrons repel one another

(Fig. 1) and the magnetic field is the force on a moving charge due to other moving

charges (Fig 2). When these two forces act simultaneously in the form of waves, the

resulting waveform in Figure 3 is created. Characteristics which make one































Figure 1: Electric Fields23






*-Flux lines



* Wire
N







Direction
of current


Figure 2: Magnetic Fields23











Electric Field


Figure 3: Electromagnetic W,


> "-. Magnetic Field









aves23


Waveguide


Coaxial Cable


Microstrip


Figure 4: Microwave Transmission Lines23









Electromagnetic wave different from another include frequency, wavelength, impedance,

power density, and phase.

Microwaves are transmitted primarily through four forms. One way is

transmission through space using antennas. The other three forms, which find their use in

systems used in microwave plasma formation, are waveguides, coaxial cables, and

microstrips (Fig. 4). A waveguide is simply a hollow metal pipe or box through which

the microwaves travel from the source, or magnetron, to the electrode at the receiving

end of the waveguide. A coaxial cable consists of inner and outer conductors containing

an insulating material through which microwaves can travel. Finally, a microstrip works

similarly to a coaxial cable, with top and bottom planar conductors sandwiching an

insulator.23

A microwave plasma is formed when microwave energy is transmitted from the

source (or magnetron) through one of the transmission lines previously described to a

discharge tube containing the plasma gas. A plasma consists of a partially ionized gas

24
which is, on average, electrically neutral. Maximum power is transmitted to the plasma

by positioning the discharge tube where the plasma is to form at a point where the

electric field component is at a maximum. The discharge is then ignited by "seeding" the

plasma gas with electrons using a Tesla coil; in some cases, microwave heating of the

containment vessel will release electrons from the vessel walls and allow autoignition of

the plasma. The plasma is then sustained by collision of the electrons with gas atoms.

The free electrons are initially moving in an oscillatory motion in phase with the

microwave field. However, when the field changes phase quickly the electrons move out

of their oscillatory motion (or out of phase) and begin to collide with the surrounding









plasma gas atoms. The plasma is then self-sustaining if a given electron creates at least

one new electron by collision before it eventually is recombined.25

Microwave plasmas are categorized according to the way in which the microwave

energy is transferred from the source to the plasma. For example, the MIP transfers

microwave energy through coaxial cables while the CMP uses a waveguide. The specific

system designs of the MEP, CMP, MPT, etc. will be described in their respective sections.

Several helpful books containing information on microwave theory are

referenced. 22,23,26



The Capacitively Coupled Microwave Plasma

Since this microwave plasma system was used exclusively in the work to be

described later, this section will review the literature on the CMP from 1985 to present.

The capacitively coupled microwave plasma was developed by Murayama, Matsuno, and

Yamamoto at the Hitachi Central Research Center in 1968.27 A common CMP design is

shown in Figure 5. The CMP is generated by transmitting microwaves from the

magnetron through a rectangular waveguide to an electrode. The waveguide supports a

standing wave. Maximum coupling of the power from the magnetron to the electrode is

described in Chapter 2. The electrode is contained within a discharge tube and the

plasma forms at the tip of the electrode around which the plasma gas flows.

The CMP offers several advantages, including robustness (easily accommodating

both gaseous and liquid sample introduction) and operation at high powers. In addition,

the CMP has demonstrated promising results in the direct analysis of solid samples.

However, the CMP tends to be slightly less precise and suffers from a higher





























I-1/4 -*1


Coaxial Waveguide

Torch and Electrode


Adjustable
Screw


1" 1/4 X-


Figure 5: Capacitively Coupled Microwave Plasma








background than the MIP.28 Analytical figures of merit for the CMP will be provided in

Chapter 3.

Diagnostics and characteristics

Temperature and electron number density measurements for the CMIP have been

investigated; several values for this and the other microwave plasma systems are

summarized in Table 1. Spencer et al. examined spectroscopic plasma temperatures in a

high flow rate (> 6 LPM (liters per minute)) helium CMP and found little difference in

temperatures for aqueous versus organic solutions.29 Temperatures and electron number

density as a function of power, observation position, and solution uptake and carrier gas

flow rates were investigated for a helium/hydrogen CMP by Masamba et al.30 Hydrogen

in the plasma gas provided slight increases in the rotational and excitation temperature,

but reduced emission signals for elements introduced into the plasma by solution

nebulization.31 Bings and Broekaert found that the N2 CMIP provided lower detection

limits for several metals compared to Ar and air CMPs; however, these detection limits

were still one order of magnitude worse than those obtained with ICP-OES.32 Finally,

analytical performances were compared for a Pt-clad W-rod electrode and a tubular

tantalum electrode plasma torch by Patel and coworkers.33 The tubular tantalum

electrode was found to have improved signal-to-background, signal-to-noise and

precision, as well as improved detection limits over the tungsten rod electrode.









Table 1: Temperatures and Electron Number Densities for Microwave Plasmas

Plasma Torch Gas Power T,.t (K) Tec (K) ne xlO-14 Ref
W~. (W) (cm-3)
(mm)
MIP 4-5 Ar 85-100 2000-2700 34
IP 4-5 He 120-180 2200-2600 34
MIP 5 Ar 400-600 3110-3580 14200 35
MIP He 75 1492-1976 6100-6700 36
MIP 2 He 85 1270-1620 37
MIP 5 He 400 5.75 38
MIP 5 Ar 40 24 38
MIP 2 He 80 6-8 39
MIP 2 Ar 85 6100 4500 4-16 40
MIP 2 He 85 6100 8800 1-2.5 40
IP He 350 1600-2000 1 41
MIP Air 400-560 4533-4730 42
MIP Ar 100-200 600-950 43
MIP He 100-400 800-1500 43
MIP 02 100-300 1300-2500 43
MIP He 270 3000 500-3000 1.1-2.1 40
Surfatron 2 Ar 82 2000 2400 3-4 44
Surfatron 2 He 82 2000 3000 1 44
Surfatron 1 Ar 50 2200 5100 9 45
Surfatron 3 Ar 0-200 1500-2500 7000-8000 3-4 46
Surfatron He- 50-150 1850-2210 0.24-0.33 47
CO2
CMP He 700 1620 3430 4 29
CMP He- 700 1800-3000 2000-5000 4-9 30
H2 I-I_
CMP N2 600 2800-4300 4900-5500 0.01-1 32
CMP Ar 600 4800 32
CMi Air 600 4600 132


Gas Introduction

The CMP has been used in conjunction with GC separation for detection of

organic and inorganic compounds. Uchida and coworkers used GC-CMP-AES for








the determination of butyltin compounds with nanogram detection limits.48 In the same

study, organic and inorganic tin were determined using hydride generation and collection

in a cold trap prior to introduction into the CMP. Nanogram detection limits were also

found for nonmetals in organic compounds by Uchida et aL 49 Huang and Blades found

detection limits in the sub ng s-1 range for a variety of organotin compounds separated by

GC.50 Finally, the detection of trace levels of water in solid samples by evolved gas

analysis was studied by Hanamura et al.51 Absorbed water and chemically bound water

in a variety of solid samples could be distinguished and quantified; however, the method

proved less accurate than conventional methods and calibration curves needed to be

prepared on a daily basis.

Liquid Introduction

Solution nebulization sample introduction was investigated by Patel and

coworkers for the tubular electrode torch mentioned above.52 Detection limits for a

variety of elements were in the low and sub-jg ml"1 range with precisions of generally

one to two percent. Parts-per-billion detection limits were also found by Hwang et aL for

a high power (up to 1600 W) helium CMP.53 This system contained a graphite tube or

rod as the electrode, which exhibited lower emission background and no significant

contamination as compared to the metal rod electrode previously used. Aqueous and

organic fluorine and chlorine were also determined by the He-CMP.54 LODs for organic

fluorine and chlorine were 1 and 0.4 jig m1-1, respectively, while fluorine was undetected

and chlorine was only weakly observed for the aqueous solutions. A highly efficient

desolvation system was developed with pneumatic nebulization into a CMP by Uchida








and coworkers.55 Sensitivity and detection limits for manganese were improved over

CMP and ICP with conventional pneumatic nebulization.

Microsampling in a graphite cup contained at the top of the electrode was

investigated by Ali, Ng, and Ali.56 Plasma heating of the electrode vaporized the sample

with detection limits from 10-210 pg for a variety of elements and precision better than

12% RSD. Microsampling was also achieved using a tungsten filament electrode, onto

which samples could be injected before rapid vaporization and excitation in the plasma.57

Detection limits for 12 elements were below 100 pg, with a linear dynamic range of 3-4

orders of magnitude and precision better than 10% RSD. A tungsten cup electrode was

also studied for the analysis of metals in microsamples.58 Detection limits for Cd, Zn,

and Pb were in the low pg range for 10 gl samples with less than 10% RSD.

Finally, whole blood samples have been analyzed by the Winefordner group using

CMP-AES.59"61 The detection limit for lead in a 2 p.1 blood sample was 30 ng ml"1 with

precision better than 10% RSD.

Solid Introduction

Hanamura, Wang, and Winefordner analyzed hydrogen and oxygen in metals by

heating 1-2 g samples in a furnace under pressure and extracting the vapor in helium gas,

which was carried to the CMP.62 Steel samples were analyzed directly by Masamba et aL

by placing the sample in a cup cut into the top of the graphite electrode.63 Limits of

detection indicated a usefulness of this technique in the range of sub- jig g-1 to the

percent range for solid steels. Tomato Leaves (SRM 1573a) and Coal Fly Ash (SRM

1633a) Standard Reference Materials were analyzed directly by Ali, Ng, and

Winefordner64 using CMP-AES with 20% N2 / 80% He as the plasma gas. Detection









limits were reported in the nanogram and sub-nanogram range for a variety of elements.

The precision was 12-18% RSD for 5-10 mg samples. Analytical figures of merit for the

analysis of solids by the CMP are provided in Chapter 4.



The Microwave-Induced Plasma

The microwave-induced plasma is the most widely used microwave plasma.

Beenakker65 first reported use of this plasma operated in helium and argon at

atmospheric pressure in 1976 and as an element-selective detector for gas

chromatography in 1977.66 The MIP is formed by transmitting the microwaves from the

generator through a coaxial cable to a resonant cavity. Figure 6 shows a typical

Beenakker cavity design. The cavity is constructed from copper due to its high

conductivity. Twelve screws hold the removable lid tightly to the fixed bottom for good

electrical contact. The silica discharge tube through which the plasma gases flow

extends through the center of the cavity where the electric field strength is at a

maximum. A 1 mm copper wire loop extending inwardly from the cylindrical wall

serves to transfer the power to the cavity inductively. It is fixed to the cavity by a

connector and vacuum sealing kit which prevents arcing between the loop and the bottom

of the cavity. Tuning is accomplished by two finely threaded screws located in the

cylindrical wall opposite the coupling loop and in the bottom wall parallel with the

discharge tube.65










Tuning stubs


Discharge
- tube


Figure 6: Beenakker Resonator Cavity65


Copper
wire loop









The microwave generator is equipped with both a power output meter and a

reflected power meter. The cavity is tuned to minimum reflected power and the plasma

gas passed through the discharge tube. The plasma then autoignites or is ignited with a

Tesla coil.65

Several excellent reviews on the microwave-induced plasma as a source for

atomic emission spectroscopy have been written.28' 67, 68 Overall, MIPs are found to be

most often used as detectors for gas chromatography. Their high power densities make

them excellent atomization sources for metals and nonmetals alike.69 They also operate

easily at low power. However, IlPs do not accommodate liquid sample introduction

well and sometimes are even extinguished. In addition, optimization of the plasma with

frequency is not easy since the resonant cavity frequency is determined by the width of

the cavity.68



The Surface-Wave / Surfatron Plasma

A microwave plasma obtained through surface wave propagation was reported by

Moisan, Beaudry, and Leprince in 197570 and developed as a source for optical

spectroscopy.71 Surface-wave, surfatron, or surfaguide microwave plasmas rely on

propagation of microwaves along the boundary of a medium. If the surface-waves in a

gaseous medium are adequately energetic, a plasma can form and the surface-waves then

propagate and sustain the plasma simultaneously. An illustration of a surfatron is given

in Figure 7. The surfatron, which is responsible for launching the surface waves which

create plasma columns many times longer than the excitation structure (tens of

centimeters), is composed of two main parts. The first part is the coupler which transfers









the microwave energy from the coaxial cable to the plasma. The coupler can be moved

vertically; this movement in turn moves the end plate that controls the coupling of the

energy to the plasma. The second part is the excitation structure, which acts as an

extension of the coaxial transmission line and extends into a Faraday cage. The length of

the excitation structure, L1, is varied using a toothed rack and a pinion. A certain

frequency bandwidth (in terms of a maximum admissible reflected power) is associated

with any length L1. For a desired frequency, the length LI is set to an approximate value

before igniting the plasma with a Tesla coil. The depth of the coupler is then adjusted for

minimum reflected power (which is also the maximum plasma length). The length L1

can then be adjusted so that the impedance is matched to the characteristic impedance of

the transmission line, for a resulting reflected power of zero. Thus, no tuning stubs are

needed and the resulting plasma is azimuthally symmetrical.

Because the coaxial cable of this design is limited to about 1 kW at 915 MHz for

safe operation, the surfaguide was developed to permit use of higher microwave

powers.72 For a surfaguide launcher, the power which can be applied is only limited by

how efficiently the plasma tube can be cooled. The surfaguide design is illustrated in

Figure 8. Power is transmitted from the microwave generator to the plasma through a

tapered waveguide. A moveable plunger located at the opposite end of the input power

acts as a short circuit; tuning is accomplished by positioning the plunger so that there is a

minimum reflected power on the reflectometer. A Tesla coil is used to ignite the plasma.

A 1991 review of the design and physical principles of surface-wave plasma sources was

given by Moisan and Zakrzewski.73











Standard
coaxial connector

Coupler
End plate

Plasma


minD I NN~N~


L


Figure 7: Surfatron7


mm





Uff


114


ArI --


I "I


I ________________


NNI.,NN",


--i


f


I









Movable
short circuit
(plunger)

Signal Launcher Surfaguide
generator





Circulator Drcoal
Traveling wave waveguide
tubeamplifier rt-nition
2-4 Glz


Figure 8: Surfaguide71









Surface-wave microwave plasmas have the advantage of operation over a wide

range of parameters, including frequency, power, and flow rates.71 However, like the

MIP, this plasma cannot accommodate liquid introduction at low power.



The Microwave Plasma Torch

The microwave plasma torch was developed at Jilin University in 1985 and

improved by joint cooperation at Jilin University and Indiana University by Jin and others

in 1991. 14,7' The MPT works differently from other microwave plasmas in that an argon

plasma can be sustained at very low flow rate (10 ml min") and forward power (40-500

W). Under slightly different conditions an He or N2 plasma can be formed. Unlike the

Beenakker and surface-wave microwave plasmas, the MPT can more easily withstand

liquid sample introduction. Sample aerosol can be introduced into this plasma with or

without desolvation. Figure 9 illustrates a microwave plasma torch. The torch is similar

to the ICP torch, with three concentric metal tubes. The intermediate tube contains the

plasma gas. The sample and carrier gases are introduced through the central channel of

the torch and the sample is then vaporized and atomized in the plasma. The plasma does

not come in contact with the tip of the electrode and therefore does not suffer from

contamination from the electrode material. Microwave energy from the generator is

coupled to the torch through a cylindrical antenna which surrounds the intermediate tube

and is tuned by changing the distance from the top of the torch to the antenna (LI) and/or

the short circuit (L2). Madrid and coworkers have characterized the noise in an MPT,

which was found to be dominated by white noise below 100 Hz with discrete noise peaks
76
(presumably from argon flow fluctuations) in the region above 300 Hz.
















Vertically
sliding collar
To MWG


Antenna
Screw


IL
L 2







-, Plasma gas


i
Carrier gas & sample


Figure 9: Microwave Plasma Torch75









The MPT offers several advantages over other microwave plasma techniques such

as operation at low flow rates and forward power, ease of tunability, no contamination

from electrode material, and most important of all reduced sensitivity to introduction of

liquid aerosols. These attributes make the MPT a promising source in microwave plasma

spectrochemical analysis.75



Other Plasmas and Electrothermal Atomization Sources

Several other plasma techniques exist, aside from the microwave plasmas, which

have multielement capabilities and can be compared with the CMP-AES. These include

furnace atomization nonthermal excitation spectrometry (FANES), furnace atomization

plasma emission spectrometry (FAPES), direct-current arc atomic emission spectrometry

(DC Arc-AES), and inductively coupled plasma atomic emission spectrometry (ICP-

AES). Only emission techniques are featured here; atomic absorption and fluorescence

are not normally amenable to multielement analysis and will not be discussed. In

addition, graphite furnace atomic emission spectrometry is omitted in its original form

due to the improvement in figures of merit provided by furnace atomization nonthermal

excitation spectrometry and furnace atomization plasma emission spectrometry.



Furnace Atomization Nonthermal Excitation Spectrometry

A method coupling the efficient atomization of a graphite furnace with excitation

of a discharge was developed by Falk et al.77 Known as furnace atomization nonthermal

excitation spectrometry (FANES), or hollow cathode FANES (HC-FANES), this

technique provides the benefit of the nonthermal atomization and excitation processes








78
being controlled and optimized separately, although acting on the same volume. A

schematic of a FANES set-up is given in Figure 10. A 20 pL sample aliquot is pipetted

onto the wall of the graphite furnace through a removable lid. The sample is dried and/or

ashed with the lid off to remove vapors, then the furnace is sealed and pumped down to

1-5 torr of helium or argon. The low pressure discharge is then initiated, with the

graphite tube serving as the cathode. The furnace is then allowed to cool to ambient

temperature with the aid of a chilled water system, and the lid is removed for subsequent

sample injection.77

FANES offers the advantages of the highly efficient atomization source of

graphite furnace with simultaneous multielement capabilities. FANES exhibits high

sensitivity for a wide range of elements, a linear dynamic range of 5 6 orders, and ppb

detection limits.77 FANES is also free from analyte losses due to nebulization and

dilution of the sample by transport in the carrier gas, as is the case with dc arc and ICP.79

However, the detection limits of FANES are limited by fluctuation of the background,

which contains molecular bands due to impurities in the carrier gas and leaks in the

vacuum system.79

In addition to the FANES source just described, another system has been

developed which utilizes a small carbon rod running through the center of the furnace.

This rod serves as the cathode, and the graphite tube serves as the atomizer and anode;

the resulting plasma is a hollow anode discharge (HA-FANES). The HA-FANES

discharge differs from HC-FANES in that instead of the negative glow region being

uniform within the cathode, it forms a halo covering the entire length of the central

carbon rod cathode. Emission intensities in the HA-FANES are highest in the upper
























Quartz Window -


Graphite
- Contact Cylinders


- Quartz Window


Rotation ann for
changing graphite tube


Figure 10: FANES77









hemisphere of the negative glow region, where samples are deposited below the cathode

and to one side on the wall of the furnace.80,81 Detection limits appear comparable to

HC-FANES.



Furnace Atomization Plasma Emission Spectrometry

Like FANES, furnace atomization plasma emission spectrometry (FAPES) is an

adaption of graphite furnace which allows simultaneous multielement analysis

capabilities. In this technique, a plasma is formed inside a graphite furnace with a high

frequency antenna which runs along the furnace axis.82 Figure 11 shows a schematic of a

FAPES source. A graphite electrode 1 mm in diameter and 40 mm long is attached at

one end to a female RF connector and runs through the middle of a graphite furnace tube.

The applied RF frequency resonates between 27.02 and 27.32 MIz, and the power

applied is between 20 and 30 W. Helium is used as the discharge gas and flowed through

the furnace at about 10 mL/min. Drying, ashing, and atomization steps are employed

similarly to the conventional graphite furnace technique, but with the RF power being

applied midway through the ash cycle. The plasma autoignites immediately upon

application of the power.83 Although helium is mainly used as the plasma support gas,

argon has been found to form a stable plasma at higher frequencies.84

The FAPES technique and figures of merit are similar to FANES, with the

advantage of operation at atmospheric pressure. The low pressure operation of FANES

makes sample introduction more complex and increases analysis time. Detection limits

are in the picogram range with a linear dynamic range of 2 4 orders of magnitude.

Precision for hand pipetted replicate signals ranges from about 2 12 % RSD. 85













Graphite Furnace
H]ournn


Pneumatic Sampr,
Introduction Port


I


Quatt. Window



Graphite Rod




i aphite Furnace


V
To RF Filter and
Graphite Furnace
Power Supply


Figure 11: FAPES83









Direct Current Arc-Atomic Emission Spectrometry

The dc arc, which is the most common arc source used in atomic emission

spectroscopy, was developed as a method for spectrochemical analysis by Margoshes and

Scribner86 and Korolev and Vainshtein87 in 1959. One design of the dc arc, the

commercially available Spectrametrics Spectrajet, is shown in Figure 12.88 Many

variations of the dc arc exist. The plasma arc is created by passing current across a pair

of metal or graphite electrodes. Graphite is the most common electrode material because

it is inexpensive and available in high purity. Samples are generally solid powders,

chips, or filings which are introduced into the arc by vaporization from a cup shaped

electrode. The anode is usually the lower electrode which holds the sample.89 Liquids

can be analyzed by evaporating the solution to dryness before initiating the arc, or the

sample can be nebulized.88 Ignition of the arc is accomplished by bringing the two

electrodes into contact for a brief moment or through use of a low-current spark igniter.

The arc is then sustained by thermal ionization of the material between the electrodes.89

Dc arc-AES is known for its good detection limits, but poor precision due to arc

wander. Magnetic fields applied transversely to the arc90 and carbon ring electrodes9'

have been used to aid in stabilization. In addition, the dc arc suffers from selective

volatilization as the electrodes slowly heat, which also causes analyte signals to differ

depending on the matrix, and self-reversal is a problem.89



Inductively Coupled Plasma- Atomic Emission Spectrometrv

Inductively coupled plasma atomic emission spectrometry (ICP-AES) is probably

the most popular technique for trace spectroscopic analysis. Approximately ten times






















Sample and argon input
from spray chamber


Electrical
connection


Ceramic
bead
Water outlet

0.040 electrode
\o (cathode, negative)
ring 1



Electrical connection
on far side of post


Figure 12: DC Arc88









more papers have been published on the ICP than on the MIP or dc plasmas.92 The ICP

was first used for spectrochemical analysis by Fassel and coworkers at Iowa State

University and by Greenfield and coworkers in England in the 1960s. A schematic of the

ICP is provided in Figure 13. The ICP consists of a quartz torch surrounded by an

induction coil connected to a high-frequency generator, which is normally operated at 27

MHz and 1 5 kW of power (Figure 13a). Argon, the most common carrier gas, is

flowed though the center of the torch with the sample aerosol and is also used as an outer

sheath coolant gas (Figure 13c). The plasma is ignited with a Tesla coil. The induced

current in the torch, which is composed of ions and electrons, heats the support gas to

temperatures around 10,000 K. The plasma is stable and self-sustaining at this

temperature. The magnetic fields (H) and eddy currents (I) are shown in Figure 13b.89

The ICP has many advantages, including high temperatures, long residence times,

high electron number densities, a nearly chemically inert environment, and the absence

or near absence of molecular species. However, spectral overlap interferences can be

present, and start-up and operational costs are high. In addition, operation of this source

is not simple, requiring considerable training of the operator.89 Also, compared with

electrothermal atomization techniques such as FAPES and FANES, the ICP has losses

due to the nebulizer system and transport losses, as well as dilution of the sample in the

carrier gas.79



Microwave Radiation and Safety

Radiation safety should be a concern to anyone working near a microwave field. The

effects of microwave radiation on humans remains controversial.93 The American














Plasma


It
Induction coil


T Quartz tube

Ar


argon flow


Figure 13: 1CP89


Argon
tangentia
coolant flow









National Standards Institute (ANSI) has designated a maximum exposure level of 10

mW/cm2. This is generally used by industry and the military regardless of frequency,

pulsed or continuous radiation, partial or whole body exposure.94 However, exposure to

levels as low as 0.1 rW/cm2 has been linked to headaches and disruption of neural and

cardiovascular systems. High exposure levels have been linked to eye cataracts,

impotency in males, teratogenesis, and other health problems; many of the more serious

effects are due in part to the heat stress caused by microwave radiation. In Russia, the

standard is a more stringent 0.01 mW/cm2 for continuous exposure.95

Microwave radiation leakage has been found to be dependent upon the source

design and the carrier gas used in the microwave system. Radiation leakage increases as

the ionization potential of the gas decreases and the molecular weight increases; helium

was found to have the lowest leakage compared with neon, argon, and krypton. Also,

addition of molecular gas to the carrier gas has been found to reduce microwave

radiation leakage due to microwave energy being consumed in the molecular dissociation

process. 96

The microwave plasma used in this research was shielded with an aluminum

cage. In this work, it was occasionally necessary to make adjustements or monitor the

plasma with the door to the cage open. A microwave leakage detector from Narda /

Lockheed Martin Microwave (Hauppage, NY, model 8201, meter 8211, probe 8223) was

used to monitor exposure levels. Table 2 gives the maximum microwave radiation levels

with their positions when the cage door was open.








Table 2: Maximum Microwave Radiation Levels

Position Microwave Radiation Level
(mW/cm2)
At cage door (ovary level) 1.5

Bottom of torch < 15

Beside coaxial waveguide < 4

1 foot back from cage (ovary level) 1.0

2 feet back from cage (ovary level) 0.3


With the cage door closed, the maximum leakage was found to be _<1 mW/cm2, with the

maximum 1 mW/cm2 at the top of the front cage door and away from the operator. This

is well below the ANSI limit.

If there is a concern for safety, a lead apron can be worn by the operator. Lead

aprons and other safety products are available through Picker, International, Inc.

(Norcross, GA). Pregnancy lead aprons with extra lead in the region of the fetus are also

available.














CHAPTER 2
INSTRUMENTATION



The following sections describe the components of the microwave system used in

this work and the optimization of their designs. Nearly every component of the system

has been evaluated and redesigned since the beginning of the work presented in this

dissertation. Table 3 lists the final individual components and their place of manufacture,

and a schematic is given in Figure 14. Throughout this chapter references are made to

changes from the original system. Therefore, for clarity, Table 4 lists the original

components, with an accompanying system schematic provided in Figure 15. A

photograph of the final system is provided in Figure 16.



Waveguides


The waveguide functions to transport microwaves from the magnetron source to

the electrode which supports the plasma. The basic design of a waveguide is shown in

Figure 17. Several important features of the waveguide include the waveguide material,

the a and b dimensions, and the positioning of the magnetron probe and receiver (the

electrode in this case).








Table 3: Final Instrumental Components of the CMP


Component
High voltage dc power supply:
Model RHVS 5-3800
RS-232 hardware and software interface
Filament Transformer: #705-0086
Magnetron
Model NL10251-2 (2450 MHz, 1.6 kW)
Aluminum Waveguide
Coaxial Waveguide
Gas Mass Flow Controllers
Quartz Torch


Electrodes
Axial Viewing Mirror: #01MFG007-028
Echelle-CID Spectrometer
Power Supply PC
Spectrometer PC
Microwave Radiation Cage


Manufacturer
Bertan High Voltage; Hicksville, NY


Allied Electronics; Jacksonville, FL
National Electronics; Orlando, FL


Laboratory made
Laboratory made
Porter Instrument Co.; Hatfield, PA
Precision Glassblowing; Englewood,
CO
Laboratory made
Melles Griot; Irvine, CA
Thermo Jarrell Ash; Boston, PA
Insight, Inc.
Caliber, Inc.
Laboratory made


Table 4: Original Instrumental Components of the CMP


Component
High voltage dc power supply: Model 805-IA
Filament Transformer: #705-0086
Magnetron
Model NL10251-2 (2450 MHz, 1.6 kW)
Brass waveguide
Brass coaxial Waveguide
Rotometer gas flow regulators
Quartz Torch
Electrodes
Spectrometer: Jobin-Yvon HR 1000
1 m, 2400 grooves mm'1, linear
dispersion 0.5 nm mm-1
Photodiode array:
OSMA model IRY-1024G
Photodiode array software: ST 120, ver. 2.00
Spectrometric Multichannel Analysis
OSMA detector controller
Computer: 80286, 8 MHz
Microwave Radiation Cage


Manufacturer
Hipotronics; Brewster, NY
Allied Electronics; Jacksonville, FL
National Electronics; Orlando, FL


Laboratory made
Laboratory made


Laboratory made
Laboratory made
Instruments SA, Inc.; Metuchen, NJ


Princeton Instruments; Princeton, NJ


Princeton Instruments; Princeton, NJ


Princeton Instruments; Princeton, NJ
PCs Unlimited; Austin, TX
Laboratory made



















High Voltage
Power Supply


Filament
Transformer


Echelle CID
Spectrometer


Waveguide 0*


He&H2 -


Figure 14: Schematic of Final Instrument


Computer



















Coaxial Waveguide


-M agnetron

.-Waveguide


He/H
2


Figure 15: Schematic of Original Instrument


Filter












































CID Spech-c~iph M



Figure 16: Photograph of Instrument








The waveguide material plays a role in the efficiency of microwave transport.

The waveguide is constructed from metal, which conducts current along its surface.

Microwave frequencies cause an inductance to be set up in the conductor, and hence the

current will flow along the surface of the metal rather than in the center of it. The term

skin depth is used describe how far the current penetrates into the metal surface, and is

described as where the current has decreased to lie times its surface value.22 Since

current tends to flow along the surface of a metal conductor, the less conductive the

metal, the deeper the current will penetrate.23 Mathematically, the skin depth is

expressed as




2ff fc- p




where 8 is the depth of penetration of the surface current density (m), f is the frequency in

Hertz (Hz), cr is the conductivity in mhos per meter (Q'm1), and g. is the permeability in

Henrys per meter (Hm-1).26 Waveguides are generally made from aluminum, copper, or

brass. All waveguides used in this work were made from aluminum or brass. At the

frequency used in these experiments, 2.45 GHz, the skin depth is roughly 2 to 4 gtm

regardless of the metal used.23

The waveguide dimensions play a critical role in the transport of the microwaves.

The dimensions of the waveguide are also shown in Figure 17. For propagation of

microwaves in a confined system, several figures of merit are critical: the cutoff

frequency (fQ); the wavelength confined inside the waveguide (Xg); and the modes of






















DE-


Key () Electrode


O Magnetron Probe

Figure 17: Waveguide Dimensions





propagation (transverse electric and transverse magnetic). The cutoff frequency is the

frequency below which there is no wave propagation. This frequency is expressed as f, =

c / 2a, where c is the speed of light and a is the width of the waveguide. The wavelength

inside the waveguide is dependent on the dimensions of the waveguide and is given by



2 2

I 2a









where A, is the wavelength of the electromagnetic wave in an unconfined medium (in this

case, 2.45 GHz). The mode of propagation describes the configuration of the electric and

magnetic fields within the waveguide. Waves can propagate in the transverse electric

(TE) or transverse magnetic (TM) mode within the waveguide. The waveguide is

characterized by the mode which exists in the transverse plane, or the plane which is at

right angles to the propagation of the wave. Figure 18 shows the field configurations for

a waveguide in TE mode. The TE mode also has two subscript numbers which indicate

the order of the TE mode. The first subscript describes the number of half-wave patterns

existing in the a dimension, and the second subscript describes the number of half-wave

patterns in the b dimension.97 TE0 mode is the dominant mode in all rectangular

waveguides where a > b.98 For the TE10 mode to be the only mode to propagate in the

waveguide, the a dimension must be greater than one half but less than one wavelength

(kg), and the b dimension must be less than one half of the wavelength (Xg). So for the

waveguides used in this work, which are characterized by the TE0 mode, the electric

field component is only present in the transverse wave, with a one-half wave pattern in

the a dimension.97

Positioning of the magnetron probe and electrode within the waveguide is also

critical. Maximum coupling efficiency occurs when the probes are placed where the

electric field strength is at a maximum. This happens when the magnetron probe is

placed at kg/4 (or an odd multiple of Xg/4) from the rear end of the waveguide, and the

electrode at kg/4 (or an odd multiple of X/4) from the opposite end of the waveguide.98

Figure 19 illustrates the in-phase coupling.99 The microwave signal radiates to both the























lines of electric Force
lines of Mhgrieic Frce
0 Towardthe Cserver
X Away eth sver


Figure 18: Waveguide Field Configurations100


F D View




Tp View





Side view


Key








right and left of the magnetron probe. The signal which radiates to the left will travel one

quarter of a wavelength to the wall and then back, to total a half wavelength or 180

degrees. The signal will actually return a whole wavelength later back at the position of

the magnetron probe due to the 180 degrees of travel made by the wave and the180

degree reflection at the wall (360 degrees total). Therefore, the wave radiated to the left

of the magnetron probe will reinforce itself upon returning from the left of the probe.

Several waveguides with varying dimensions were evaluated for the most stable

plasma formation. Table 5 lists the dimensions of the waveguides tested and the material

with which they were made. All waveguides were built in-house, and dimensions were

measured from the inside of the waveguide. Waveguides #1 2 were available in the

lab; waveguides # 3 4 were designed and built to test plasma stability with the

associated dimensions. As shown in Figure 17, dimensions a and b refer to the length

and height, respectively, of the end of the waveguide. The c dimension is the distance

from the end of the guide to the center of the electrode, the d dimension is the distance

between the centers of the electrode and magnetron probe, and the e dimension is the

distance from the center of the magnetron probe to the opposite end wall. Dimensions in

parentheses are the lengths in millimeters converted to fractions of a wavelength.

Upon close inspection, the original waveguide (# 1 in Table 5) of the system

appeared to have dents and warps, particularly in the area surrounding where the plasma

formed. Since the plasma suffered from some flickering, it was suspected that the

warping of the waveguide over time due to the intense heat of the plasma could affect the

coupling efficiency from the magnetron to the electrode. Another waveguide with










1/2 wavelength of travel + 1/2 wavelength on reflection
rettms (3600) 1 wavelength later






















Figure 19: In-Phase Coupling99



Table 5: Experimental Waveguide Dimensions


Waveguide
#1 (original)


material
Ag-plated
brass


a (cm)
-9.6


b (cm)
-5.5


c (cm)
-5.55
(3/8 XS)


d (cm)
19.8
(1 1/4 Xg)


e (cm)
-1.97
(1/8 Xg)


#2 (adjustable) Al 8.85 4.30 4.24* -20.0 -2.3
(1/4 kg) (1 1/4 kg) (1/8 g)
#3 Al 16.5 8.25 3.3 13.2 3.3
(1/4 kg) (kg) (1/4 kg)
#4 Al 10.9 5.46 3.7 14.8 3.7
(1/4 Xg) (Xg) (1/4 Xg)
*This dimension is variable.








similar dimensions and with an adjustable end wall near the plasma end of the waveguide

(# 2 in Table 5) was available in the laboratory and was used to replace the original

(Figure 5). In addition to using the new "adjustable end" waveguide, two new

waveguides were designed and built. In designing the new waveguides, it was

discovered that both the original and "adjustable end" waveguides did not have the exact

dimensions predicted for maximum coupling efficiency. As stated earlier, maximum

coupling efficiency occurs when both the magnetron probe and electrode are placed at

Xg/4 from their respective end walls. In addition, if the electric field strength is at a

maximum at both of these positions, then it stands to reason that the magnetron probe and

electrode should be at a distance of one wavelength (or a whole multiple thereof) apart

for maximum coupling. Since neither waveguide #1 or #2 had these specific dimensions,

waveguides #3 and #4 were designed and tested to observe whether further improvement

in the stability of the plasma could be noted. The adjustable waveguide end wall was

positioned at 4.24 cm from the electrode, which obeys the X,/4 requirement for efficient

coupling. When the end wall was not close to the 4.24 cm position, the plasma suffered

from flickering and extinguishing. However, there is a small range over which the plasma

will remain stable, which is approximately 3 cm. After evaluation of the adjustable end

of the waveguide, the waveguides #3 and #4 were tested. The a and b dimensions of

these waveguides were found listed as used for commercial waveguides in The Handbook

of Microwave Measurements.97 Characteristics of several standard rectangular

waveguides are given in Table 6.100 Waveguide #4 produced a similar plasma to

waveguide #2 in both size and shape. A multielement-spiked graphite powder








Table 6: Characteristics of Standard Rectangular Waveguides

EIAa Physical Dimensions Cutoff freq Recommended
designation Inside, in cm (in) Outside, in cm (in) for air-filled freq range for
WRb Width Height Width Height waveguide in TE10 mode in
GHz GHz
770 19.550 9.779 20.244 10.414 0.767 0.96-1.46
(7.700) (3.850) (7.970) (4.100)
6500 16.510 8.255 16.916 8.661 0.909 1.14-1.73
(6.500) (3.250) (6.660) (3.410)
510 12.954 6.477 13.360 6.883 1.158 1.45-2.20
(5.100) (2.500) (5.260) (2.710)
430d 10.922 5.461 11.328 5.867 1.373 1.72-2.61
(4.300) (2.150) (4.460) (2.310)
340 8.636 4.318 9.042 4.724 1.737 2.17-3.30
1 (3.400) (1.700) (3.560) (1.860)
aElectronic Industry Association
bRectangular Waveguide
cCorresponds to Waveguide #3 in Table 4
dCorresponds to Waveguide #4 in Table 4

sample was used to quantitatively evaluate the plasmas, and the plasmas formed using

both waveguides #2 and #4 performed similarly based on emission signal intensities and

precision between runs. Either of these waveguides is effective in the capacitively

coupled microwave plasma formation. Although waveguide #3 had dimensions which

had been used commercially, it was not discovered until later that its size was not

effective for the 2.45 GHz frequency. The b dimension of the waveguide was not less

than one half of Xg, which means that modes other than the TE10 were propagating in the

guide and causing a complicated pattern of constructive and destructive interference.

Consequently, this waveguide performed poorly. The plasma failed to autoignite, and

when ignited manually with a Tesla coil, the plasma pulsed in height, up and down, in a

wave-like motion. The waveguide with the adjustable end wall was chosen for the rest of

this work.








In addition to the rectangular waveguide, a coaxial waveguide was used to aid in

propagation of the microwaves up the electrode to the site of plasma formation. The

coaxial waveguide, as shown in Figure 5, surrounded the torch and electrode and guided

the microwaves perpendicularly to their direction of propagation in the waveguide. Two

coaxial waveguides were tested, one of brass and one of aluminum. Each coaxial

waveguide was approximately 3.25" high, with a 2.5" base and a 1.5" hole through the

middle to accommodate the torch. Although each coaxial waveguide was made of

different material and had slightly different dimensions, their performances were similar

and they could be used interchangeably. Without the use of the coaxial waveguide, the

plasma would often fail to autoignite.



Plasma Gases

The plasma gases used exclusively in this work were helium and hydrogen. The

gas mixture was introduced tangentially through a quartz torch. On the original system,

simple variable area flowmeters were used to control the flow of these gases. The

flowmeters, which only provide an accuracy of 10% full scale, were replaced with

Hastings mass flow controllers to ensure an accurate and constant flow of the plasma

gases. The mass flow controllers provide a precision of better than 1% RSD in the flow

rate. In addition, a glass bead premix chamber was inserted between the mass flow

controllers and the torch to ensure homogeneity of the helium-hydrogen mixture.

Since it is desirable for the analyte signal to come off as quickly and completely

as possible, several parameters associated with the plasma gases should be optimized.








The plasma gas composition, the ratio of plasma gas components, and the flow rates and

applied power combinations were all evaluated for this system.

Most of the work on plasma gas composition was performed prior to this work.

Successful plasmas have been formed from a variety of gases, including helium, argon,

nitrogen, and air31' 101, 102 Helium and argon are preferred plasma gases since they give

no molecular background spectra. 103 Although argon is used most extensively in the ICP,

the use of helium has several advantages. According to Chan and Montaser, the

significant properties of a gas are the electrical resistivity, thermal conductivity, and

specific heat. 104 The lower specific heat of helium as compared to argon indicates that

less energy is required to heat the plasma to the same plasma gas temperature. In

addition, the higher ionization energy of helium (24.6 eV compared to 15.8 eV with

argon) suggests that excitation processes should be more efficient in the helium plasma,

which is particularly important in the analysis of nonmetals.18 Helium is not without its

drawbacks, however. It has a higher electrical resistivity than argon, which results in a

lower power transfer efficiency in the plasma. It also has a higher thermal conductivity,

which allows for faster heat dissipation towards the outer tube of the torch, causing torch

damage.04

Addition of hydrogen has been observed to aid in preventing the helium plasma

from "attacking" the torch. Addition of molecular species to inert gas in the ICP has

been observed to reduce the plasma size, which has been attributed to absorption of

plasma energy upon dissociation of molecular species.15, 106 In earlier CMP work by

Masamba, adding hydrogen was found to aid in preventing the plasma from attacking the

torch, therefore prolonging the life of the torch and reducing contamination of the plasma









by the torch material.'7 In this work, it has been observed that in addition to the

reduction in plasma size, the sample cup heated to much higher temperatures with

increasing amounts of hydrogen added to the plasma gas. This can be attributed to less

heat diffusing away from the plasma with the addition of hydrogen, and possible

entainment of air into the torch creating combustion between hydrogen and oxygen. In

addition, heat transfer from plasmas to solids (in this case, the cup) is increased with

electron transfer,108 and Masamba has shown an increase in electron number density in

the plasma with the addition of hydrogen.107

In addition to those studies on gas composition performed previously, a He-02

mixture was evaluated. In the analysis of solid samples, complete combustion of the

solid material does not occur under He-H2 plasma gas conditions. Small amounts of

oxygen were added to 12 LPM (liters per minute) of helium in hopes of more complete

combustion of the sample without degrading the torch or electrode. Unfortunately, even

adding oxygen in amounts less than 1 LPM degraded the tungsten cup and electrode

rapidly. At 1.3 LPM of oxygen a 0.7 mg sample of graphite powder was able to be

completely combusted in under 15 seconds. However, after only 10 runs, the electrode

was very noticeably eroded, making the He-02 mixture impractical for routine analysis.

The plasma shape is affected by the overall gas flow rate and the applied power.

Only a spherical plasma will form at low power (<350 W). The spherical plasma is small

and round in appearance, with a diameter of 1 1.5 cm. At higher powers (350-1200 W)

either spherical or cylindrical plasmas will form, depending on the total gas flow rate.

The cylindrical plasma is the diameter of the electrode (4 mm) and ranges in height from

about 1.5 cm at low flow rates (roughly < 5 LPM) to 3 cm and higher at the higher flow








rates. As the applied power is increased, lower and lower flow rates are needed to form

the cylindrical plasma. Eventually, only cylindrical plasmas will form (about 700 W and

higher). Figure 20 shows the cylindrical He-H2 plasma mixture at varying applied

powers, with the spherical plasma shown at 255 W and the cylindrical at higher powers.

Upon switching the system from lateral to axial viewing of the plasma, a study of

the plasma gas flow rates and applied power was performed. This study was used to

evaluate how combinations of power, total flow rate, and H2/He ratios affected the

heating of the cup. Optical temperature studies were performed to evaluate the

approximate thermal temperature of the cup in these studies. This was done using the

Omega Infrared Thermometer, Model OS3709. Figure 21 shows the temporal profile of

the cup temperature as the plasma is ignited. Usually by 30 seconds, the temperature is

nearly stable. The cup temperature increases with applied power. Figure 22 illustrates

the cup temperature increasing with both applied power and decreasing flow rates.

Powers above 1200 W produced increasingly robust plasmas which quickly attacked the

torch. Therefore, power settings lower than 1200 W are most practical. Optical

temperature studies were also performed which monitored the temperature of the cup

with varying H2/He ratios. Figure 23 shows the cup temperature increasing with both the

H2/He ratio and the applied power.

Time scans of silver, copper, and tellurium (2 ppm) were used to evaluate the

variation of the analyte signal with the H2/He ratio. These elements were chosen because

they were noted to have longer residence times in the plasma. Table 7 summarizes the

data. A power of 648 W was used for all of these studies.









48















































INC-









CNCd
Cd





4.1














vVV
V


V V TTV TV VTV V


1650-
1600-

1550-
1500-

1450-

1400-

" 1350-
C. 1300-
1250-
1-
1200-

1150-

1100-

1050


0 20 40 60

Time (seconds)


i 1
80 100


Figure 21: Cup Temperature versus Time at 6 LPM Helium, 0.070 H2/He Ratio


1750-

1700-

1650-

1600-

1550-

1500-

1450-

1400-

1350-

1300-

1250-


12flfl.


C


V


0


4 5


6

Flow Rate (LPM)


Figure 22: Cup Temperature versus Flow Rate for Various Applied Powers


V
V 00 0 0 0


.mnuu.nmun


605 W
865 W
1000 W
V 1140 W


mum









1800-



1600-



1400-


1200-


1 UUU I I I 1
600 700 800 900
Applied Power (V


-- H / He 0.028
SH/ He 0.048
H / He 0.070
v H/ He 0.114
I 1I 1I
1000 1100 1200
1)


Figure 23: Cup Temperature Versus Applied Power at Various H2/He Ratios


Table 7: Hydrogen to Helium Ratios

Trial He Flow Rate H2 Flow Rate H2 / He Copper (s) Silver (s) Tellurium (s)
(LPM) (LPM) Ratio ... ..........
1 10 0.22 0.022 23

2 0.35 0.035 50 50 19

3 0.59 0.059 35 16

4 1.03 0.103 21 12

5 1.47 0.147 34 11 7

6 7.5 1.47 0.196 22 9 6

7 9 2 0.222 19 8 5









It can be clearly noted that as the hydrogen to helium ratio increases, the

residence time of the analyte in the plasma decreases. This is expected since it has been

observed that the cup temperature increases with increased H2/He ratio. However, at the

last trial, where the H2/He ratio was 0.222, the plasma began melting the torch rapidly.

In conclusion, the cup temperature increased with increasing power, increasing

H2/He ratio, and decreasing plasma gas flow rate. The increased cup temperature

improved the atomization efficiency and reduced the analyte residence time in the

plasma. Plasma flow rates were chosen as low as possible to save costs while still

forming a stable and robust plasma. The hydrogen was mixed in the highest ratio so as to

heat the cup as much as possible without damaging to the surrounding torch. Normally,

1000 W of applied power is used to atomize and excite the sample. However, with some

of the more refractory elements powers of up to 1200 W were employed.



Electrodes

The electrode is used to direct the microwave energy out of the waveguide to the

point of plasma formation. The design of the electrode is crucial for effectively aiding in

the formation of a stable plasma. However, since it is more of an art than a science to

design an electrode that forms a stable plasma, the design of the electrode was

continually fine tuned. The evolution of the electrode is shown in Figure 24.

Since the ultimate goal of the research is the analysis of both solid and liquid

microsamples, the electrode was designed to contain a discrete amount of sample in a

tungsten cup at the top of a graphite or tungsten shaft. By preheating the sample at low

microwave energy before igniting the plasma, the solvent was evaporated from the















#6 #7


6 mn


57 mm












4 mm


1 Inn1


4 mm


Figure 24: Evolution of the Electrode


#10


50 mm


..'U


w









sample and the resulting residue could then be atomized and excited with the ignition of

the plasma. In the case of solid samples, the preheating step was used to remove any

moisture from the solid, thus avoiding sample sputtering from the cup upon plasma

ignition. The electrode in the original system was a graphite shaft, similar to a dc arc

electrode, with an indention at the top for sample containment (electrode #1). This

electrode suffered from memory effects due to solutions soaking into the graphite, and

also degraded over time from contact with the plasma. To minimize these problems, a 30

gtL tungsten cup was fitted into the top of the electrode for sample containment (electrode

#2). The electrode no longer had severe memory effects, but thermal heating of the cup

was inadequate due to the surrounding mass of graphite. Consequently, the transient

signal of the analyte would often last longer than 60 seconds.

The graphite shaft of the electrode was then replaced with a 1 mm tungsten wire,

with niobium cup holder at the top. A niobium spacer was placed several millimeters

down the rod to hold the electrode in the quartz torch (electrode #3). More rapid heating

of the cup resulted. The thin rod shaft provided the advantages of more efficient sample

atomization and excitation, quicker cooling between runs, and the electrode did not need

to be replaced since it did not degrade over time as did the graphite. Further

improvement was made by removing the niobium cup holder so that only the cup was

positioned at the top of the tungsten wire shaft (electrode #4). Although more efficient

atomization resulted from the new tungsten shaft design, serious problems stabilizing the

plasma arose. Instead of the plasma focusing in the center of the cup, the plasma would

often whip around the upper edge of the cup. The plasma also appeared thinner in








diameter. This could possibly be attributed to insufficient mass in the tungsten rod

available for properly conducting the microwave energy to the point of plasma formation.

The graphite shaft was then redesigned so that only the bottom of the cup came in

contact with the shaft (electrode #5). A piece of the thin tungsten rod used previously

was fitted through the cup and into the shaft to hold the cup in place. The whipping

observed with the tungsten rod shaft was no longer apparent. Further improvement to

this design was made by extending the tungsten rod up to about 1 mm above the cup

surface (electrode #6). This minimized plasma wandering. The resulting plasma was

very stable, and seemed to "focus" on the central tungsten pin. Unfortunately, this

electrode was only good for solids analysis since small amounts of liquid sample could

escape through the very small gap between the tungsten pin and the cup. Also,

atomization and excitation efficiency decreased from that of electrode #4 due to the

graphite pulling heat away from the cup and plasma.

The tungsten rod electrode was then redesigned to incorporate the "focusing" pin

used in electrode #6. In addition, the pin was pressure fitted to allow for liquid analysis,

as well (electrode #7). Whipping of the plasma around the cup edge continued to be a

problem. A graphite shaft was also fitted over the existing tungsten wire (electrode #8),

but plasma whipping still continued.

In evaluating the electrodes with tungsten "focusing" pins (#5-8), it was

discovered that part of the imprecision in the emission signal was due to slight moving in

the parts of electrode during sample introduction. A cup was then designed to screw

directly into the graphite shaft (electrode #9). This allowed for a very physically stable

electrode. In addition, the cup could be screwed in only part way, leaving a gap to help









the cup to heat more efficiently since not in direct contact with the graphite shaft

(electrode #10). When the cup was screwed all the way in to the graphite shaft, the

temperature approached 1150'C; with a 2.5 mm gap between the cup and the shaft the

temperature increased to 1400'C.

As mentioned previously, the electrode is positioned at X,/4 from the end wall of

the waveguide for maximum coupling efficiency. The penetration depth of the electrode

into the waveguide is another important parameter for effective coupling. The optimum

penetration depth of the electrode into the adjustable end waveguide is 14 19 mm.

Outside of this range, the plasma begins to diminish in size.



Power Supply, Magnetron. and Filament Transformer

The power supply, magnetron, and filament transformer work together to power

the plasma. Figure 25 provides a schematic for the electrical system. The filament

transformer heats the filament in the magnetron and steps the voltage from 110 V to 5

VAC.

On the original system, a manual dial controlled the power supply. The ramp rate

of the power supplied to the plasma was therefore only as reproducible as the human

operator. This led to a great deal of imprecision in the final measurement, both between

runs for a single operator and from one operator to another. The power supply was then

replaced with the Bertan High Voltage power supply (see Table 3), which was computer

programmable for both the voltage and current. The power supply had a 5 kV, 750 mA

maximum output (3750 W) and operated from 220 V input in the negative bias mode.

The stability for this power supply was 0.01% per hour, with 0.05% rms ripple noise.

















Magnetron is shown within the dotted lines
L1CI and L2C2: low pass filters
RI, R2, and R3: 5000 0, 5000 9, and 2500
0 ballast resistors in parallel
C3 and C4: capacitors
S1: on-off switch
F: fuse


Figure 25: Schematic of Power Supply








The magnetron consists of an anode and a heated cathode within a magnetic field.

The schematic of a conventional magnetron is shown in Figure 26; the magnetic field is

applied toward the plane of the paper. The anode is a metal block with cylindrical

cavities, the size of which determines the output frequency of the magnetic radiation

(2450 MHz in this case). The magnetic field causes the flow of electrons from the

cathode to the anode to be in a curved path rather than a direct one.99 The

electromagnetic field in the region between the cathode and anode causes electrons to

move in a wave-like fashion, around the inside surface of the anode. The circular cavities

act as individual cavity resonators and the traveling wave in between the cathode and

anode couples the fields from all the cavities together.109 The electrons flow in cycloidal

paths under the electric and magnetic forces as shown in Figure 27.1 One of the cavities

contains a small pick-up loop which extracts the microwave energy from the cavities as

the magnetron oscillates.99 When a negative potential is applied by the power supply, the

electrons are attracted to the anode and the resulting microwave energy is extracted out of

the magnetron and into the waveguide, where it is guided to power the plasma.

A preatomization step can be used to dry the sample. Typically, this requires

simple thermal heating with low microwave power without ignition of the plasma. A

preatomization of 48 W for 20 seconds will dry a 5 gL sample if the cup is cool. If the

plasma has already been fired several times, the cup will retain heat. Normally, waiting

60 seconds between runs will allow the cup enough time to cool so that the sample is not

sputtered from the cup due to excess heat. The sample will then generally dry within 15


















I
output


Figure 26: Schematic of Magnetron99


Anode


CaEoodepa

Electron pathi


Figure 27: Electron Flow in the Magnetron'0








seconds without the need for any additional applied power. Atomization and excitation

then occur when the power is raised above 48 W and the plasma autoignites.



Echelle Spectrometer

The detection system consists of an echelle spectrometer and charge injection

device (CID) detector. The instrument was manufactured commercially by Thermo

Jarrell Ash (TJA) primarily for use with a dc arc. The TJA instrument was modified by

replacing the dc arc source with the CMI.

The echelle spectrometer covers a wavelength range from about 190 to 400 nm.

It consists of achromatic focusing lenses, a 50 gim slit, collimator mirror, prism, grating,

camera mirror, and the CID detector (Figure 28). The prism, made of fused silica, breaks

the incoming light into orders while the grating (52.6 grooves/mm, 61.60 blaze angle) is

used for wavelength dispersion. The focal length of the spectrometer is 381 mm and the f

number is f/1. Echelle spectrometers use a grating with a larger blaze angle (> 450) to

achieve high resolution rather than a high groove density, as with conventional

spectrometers. They have the advantage of about an order of magnitude higher resolution

than conventional spectrometers and a larger wavelength range which is useful for

multielement analysis, but generally sacrifice luminosity due to the need for a small slit

width (< 1mm) so that interference between orders is avoided.1 Echelle spectrometers

are ideally suited for use with charge transfer detectors, since the resulting spectrum is a

set of parallel subspectra arranged in square."1
















Acustm s


Adqtedfrom TJA AtmCaW 20X0Hardware Guide


Figure 28: Echelle-CID Diagram


Slit


Plasim


Grting









The CID is comprised of a 512 x 512 array of pixels, each 25 gm square. The

camera is cooled to 420 C with an electrically operating refrigeration system, whose

probe hooks directly to the back of the camera. Cooling to low temperatures for a CID is

necessary to nearly eliminate dark current. A CUD works on the principle that as a photon

strikes the detector, charges are generated which are stored in metal-oxide-semiconductor

(MOS) capacitors. Unlike CCDs, CIDs use holes rather than electrons for charge carriers

in the semiconductor. Figure 29 shows the cross section of one pixel of the CID. An n-

type epitaxy (the photoactive part of the detector) is on top of the p-type substrate. The

collection gate is an electrode to collect the charge while the sensing gate is an electrode

to sense how much charge is collected. An insulator consisting of silicon dioxide and

silicon nitride layers separates the gates from the epitaxy. Charge collects under the

collection gate at the silicon/silicon dioxide interface, and is then collected and read by

the appropriate gates.




-V 0 V


sense gate ilicon nitride
collection gate silicon dioxide
________________ Z insulator
potential wellr epitaxy
n-type

substrate
p-type


Figure 29: Cross Section of a CID Pixel"2








A typical CID image is provided in Figure 30. This particular image is of a 25

element aqueous solution. The lower wavelengths are at the bottom of the array, moving

to the lower visible range at the top of the array. Boxes contained in the image are color

coded by the elements shown to the right. Other elemental lines are shown which have

no boxes, but the software will only allow eight elements to be selected at a time. Figure

31 is a zoomed view of the iron 259.9 rn and 259.8 nm lines, indicating the excellent

resolution of the echelle-CID. At this point in the array, the resolution is about 4.7 pm

per pixel. The boxes consist of a square block of pixels, the middle rows of which are

used for the analytical signal and the outer rows for background correction.

The software for the echelle-CID is broken into two sections. The first section is

the Research Mode, where images such as Figure 30 can be obtained. This gives useful

information on possible spectral overlap and allows the operator to calibrate the exact

location of the element boxes. The other section is the Data Collection Mode, which

simply gives background corrected intensities for the wavelengths selected. The

subarray for a particular element box can be monitored in this section to ensure that the

background is appropriately chosen for subtraction from the analytical signal. A typical

subarray is shown in Figure 32. The background selected for subtraction can be modified

for any subarray.

The precision of the spectrometer for low counts was evaluated. Using both Cd

and Fe hollow cathode lamps and a series of neutral density filters, the precision was

monitored as the light was attenuated. It was found that the measurement precision of the

echelle-CID was < 1% RSD for count rates greater than 3 cts/s; a large increase in the %

RSD began to be observed at less than 3 cts/s.



























ER Cu
Mo Ag
EM As
Eo Fe
a Ga
n] Mg
a Pb
E2 Zn


Figure 30: CID Image










































Figure 31: Zoomed CID Image








Subarray
SrUharray -i
Width:
Height-: EI"[I2F


Width-"

251 45 25-5 T
B-ackgrjO'und all5:79,

iJ Left: E1 Mp
:l night: Flaw: 1.84

eMap defaults ________________ :9 CoT-:19


Figure 32: Subarray of the CID








Plasma Conditions for Multielement Analysis

Multielement analysis using the CMP began after installation of the echelle-CID.

Prior to that time, primarily single element analysis was performed since the PDA

detector on the original system only had a 20 nm window. Previously, Pless performed

single element analysis on the original system for Mg, Ca, K, Na, Pb, Cd, Zn.'13 Several

parameters were studied and optimized for the greatest emission signal and/or signal to

noise ratio. These parameters included the drying power, atomization power, helium and

hydrogen flow rates, and the observation height viewed laterally in the plasma. The

helium and hydrogen flow rates and drying power were found to have little effect on the

elements studied. The optimal observation height at which the plasma was viewed

laterally varied from 0 5 mm above the cup surface. The optimal height was attributed

in part to the volatility of the element studied. In addition, the atomization power was

also found to have an effect on the signal. In general, there was a trend toward higher

signal intensity with applied power, followed by an eventual leveling off. Applied power

of 600 1000 W was found sufficient for the elements studied.

In this work, a multielement solution of Cd, Fe, Pb, Sn, and Zn was used to

optimize the position of the plasma with respect to the entrance slit of the echelle-

CID for lateral viewing of the plasma. Figures 33 35 show optimization for the lateral

view of the plasma. The observation height is optimal at about 3 mm above the cup

surface (Figure 33). The most intense emission is centered in the middle of the cup

(Figure 34). This can be observed in the figure since the cup is 4 mm in diameter, which

approximately corresponds to the 8 13 mm distance on the x axis. Within the focal

length, the emission is constant with distance of the electrode from the entrance slit of














280 -
260 -
240 -
220-
200-
C 180-
C 160-
140-
0) 120-
100
. 80
60
fl 40-
20-
0-
-20


A
Y V


56 57 58 59 60

Observation Height (mm)


* Cd
* Fe
A Pb
* Sn
* Zn


I 6I
61 62


Figure 33: Translation of Observation Height in the Lateral View


m Cd
a Fe
A Pb

v Sn
* Zn


* a


I
* 9 *


4 x x


... .. .. I" I I I
0.0 0.1 0.2 0.3 0.4 0.5

Micrometer Setting (inches)


Figure 34: Translation Across the X Axis in the Lateral View


0

U

A
S Y
I


170 -
160 -
150 -
140-
130-
120-
S11o-
-- 90 -

~80-
0 70-
(1 60-
50-
40-
" 30-
, 20-:
10-

-10






68


70- Fe
65 Cd
60 A Pb
55- Sn
50
>"45 Zn
40
35,
30
25- ,
o 20 *
15
>1 Ao- A A A A A A
(U" 5"
-5- Y Y V Y
-10
-15
-20 *
23.0 23.5 24.0 24.5 25.0 25.5
Distance from electrode to entrance slit (cm)



Figure 35: Translation Across the Y Axis in the Lateral View



the spectrometer (Figure 35). A multielement solution of As, Cd, Pb, and Zn was used to

optimize the emission signal for axial viewing of the plasma. Figures 36 and 37 show a

peak in the emission intensity in the x and y directions (looking down on the cup),

respectively, which correspond to the center of the cup. The x and y directions are

parallel to, and perpendicular to, the optical axis of the spectrometer, respectively.

Atomization power and time are the parameters with the most variability per

analyte. Therefore, for multielement analysis, these conditions were chosen to effectively

atomize the most refractory element in the analysis. Generally, the power was applied as

high as possible without doing damage to the torch or mirror. Table 8 provides the

settings for multielement analysis. These settings can be applied to both aqueous and

solid microsamples.







69


2400
2200
2000 A
As
-1800 -TCd
160 0 Cd
14oo- wPb
S1400
>, v Zn
1200 Zn
1000,
800

.o 600- 40=
.I_ 400 -
E :
w 200,
0
-200
I I " I I'
4 6 7 8 9 10
Relative Distance Across Cup Surface (mm)


Figure 36: Translation Across the X Axis in the Axial View







2400
-1-~ N As
2200 0-C
= oCd
2000
1800 Pb
W v Zn
1600 -V"
1400 If
n 1200 -

1000
800
0
m 600,
E 400-
w
200 -
0
-200i
3 4 5 6 7 8 9 10 11
Relative Distance Across Cup Surface (mm)


Figure 37: Translation Across the Y Axis in the Axial View






70

Table 8: Optimal Plasma Conditions for Multielement Analysis

Helium Flow Rate (LPM) 6.3

Hydrogen Flow Rate (LPM) 0.44

Drying Power (W) 48 W

Drying Time (s) 20

Atomization Power (W) 1000 W

Atomization Time (s) 15 -60

Time Between Runs (s) 60

Number of Runs 5-10

Sample Volume (jiL) 5














CHAPTER 3
LIQUIDS ANALYSIS

Introduction

Few techniques exist which are capable of simultaneous multielement analysis of

microsamples, particularly with the ability to analyze both solid and liquid forms. Some

techniques with these capabilities include FAPES, ICP-AES, and CMP-AES. The work

described here includes single element analysis, which aids in determining figures of

merit for single elements without the possibility of matrix or interelement effects, and

multielement analysis. In addition, preliminary work in the analysis of "real" samples

has been performed, including sweat and several digested NIST samples.

Elemental concentrations in a sample are determined through calibration of

standards of known concentrations. The calibration curve is a plot of the analyte

emission signal intensity versus its amount or concentration. At very high analyte

concentrations, the emission in the plasma can become self absorbed by atoms which are

not excited. This self-absorption of emitted radiation leads to a decrease in the expected

emission signal intensity, resulting in curvature in the calibration plot. The theory behind

these curves of growth will be discussed.

To understand the curves of growth, it is necessary to look at the factors which

affect the emission intensity. It is known that the intensity of a given transition is a

function of a number of variables according to classical theory. 114-116 These are concisely








outlined by Gornushkin et at117 in the following discussion. The spectral intensity is

defined as


I=c hc ng0 1(1-e-k(v)I)dv
A no go


where ox is a constant factor depending on the instrument; h is Planck's constant (J s); c is

the velocity of light (m s-1); X is the transition wavelength (in); nI, no, and gi, go are the

atom densities (cm-') and the statistical weights (dimensionless) of the upper (1) and

lower (0) states, respectively; k(v) is the absorption coefficient; and I is the absorption

pathlength (cm). This equation takes into account any reabsorption of the emission by

atoms in the ground state. The absorption coefficient, in turn, can be calculated by

a +" -2d

k d 7 (2),



where ko, a, and x are calculated as


ko =27c 2/3 e2 n0f b = z (3),
mec b V



a (AVN +AvL) l[-2_ A 1L- (4),
AVD Av.D



S V(5).
AvD








In these equations, e is the elementary charge (C), me is the electron mass (kg), f is the

transition oscillator strength (dimensionless), v0 is the frequency of the center line (Hz),

and AVN, AvL, and AVD are the natural, Lorentzian, and Doppler half-widths, respectively.

The Lorentzian and Doppler half-widths are defined as


AvL =- 2p [2RT +j]12. (6) and
zkT_ (M M)

Vo (8zkT) 2
AvD = -'---n2 i (7),



where p is the pressure of the perturbing species (Pa), R is the universal gas constant (J

mol"1 K'), Tis the absolute temperature (K), m and Mare the masses (kg) of the emitting

atom and the perturber, respectively, and cy is the collisional cross-section (M2).

Finally, the spectral line intensity given in Equation 1 can be written in terms of

the total absorption At, multiplied by a proportionality constant. The total absorption At

for a homogenous flame or plasma is defined as



At = 2z f(1 ek(x>)dx (8).
,An 2 0


Now a theoretical curve of growth can be constructed by creating a log-log plot of

At/2b versus nofl/b. This curve will have two asymptotes described by


A 2 l eo 2 njI as n0fl 0 (9), and
loi l mec b b









log t =log 2)2e a as --p --*oo (10).
L2bj I Mee b b


These equations give slopes of 1 and '/2, respectively. By multiplying a proportionality

constant to Equations 9 and 10, the theoretical emission curves of growth result.

Equation 9 describes the emission at low number densities and Equation 10 the emission

with self-absorption. Experimentally, these curves were obtained by plotting the

emission intensity (cts/s) versus the analyte concentration (ppm).



Experimental

Aqueous standards were obtained from Fisher Scientific (Fair Lawn, NJ), SPEX

Industries (Edison, NJ), Inorganic Ventures (Toms River, NJ), and High Purity Standards

(Charleston, SC). These standards were either 1000 or 10,000 ppm and prepared in low

percent concentrations of either hydrochloric (HCl) or nitric (HNO3) acid. Calibration

solutions were prepared by diluting these stock standards in deionized water from the

laboratory Milli-Q Plus water system (Millipore Corporation, Bedford, MA).

Before a calibration series is run, a number of steps are taken to prepare the

echelle-CID for data acquisition. First, a CID image is taken and the element maps for

all analytes are calibrated. Refer to Figures 30 and 31 for CID images. This calibration

is accomplished by moving the "box" for a particular line so that it is centered around the

emission spot. Next, a time scan is recorded to determine how long it takes for the

transient signal to return to baseline. If measuring for more than one element, the

element which takes the longest to vaporize determines the total integration time of the








run. Normally, the plasma can not be run for more than 45 s due to the heat from the

plasma damaging the torch and mirror. In addition to element map calibration and time

scans, a "method" is also created in the software to select which elements are to be

analyzed, which lines of those elements will be monitored, the number of runs, and the

integration time. Once these steps are taken, the system is ready for data acquisition.

For each solution in the calibration series, 5 10 runs are made. A 5 pL volume

of solution is pipetted into the tungsten cup. If the cup is already warm from subsequent

runs, the solution will often dry within 10 or 20 seconds (depending on the power and

duration of the previous run, and the length of time since the last run). If the cup is
"cold," a drying step of 128 W for 20 s can be used to evaporate the solvent. Once a few

runs have been performed, the cup will retain enough heat and a one minute delay time

between the extinguishing of the plasma and the sample injection of the next run is

generally used. If too little time elapsed since the previous run, the sample will splatter

as it is injected into the hot cup, or the plastic tip of the pipette can even warp from the

heat. Waiting too long between runs requires use of the dry step again. In general,

precision degrades dramatically if the runs are not reproducibly timed, and so careful

timing is kept for the entire experiment. After sample injection and drying, the plasma

power is increased and the plasma autoignites, thus atomizing and exciting the analyte

residue in the cup. Most often, power settings of 1000 W for 10 45 s are used for

aqueous samples. Table 9 lists the individual settings for the experiments discussed in

the next section. All calibration plots were taken while observing the plasma axially.

The software for the TJA echelle-CID automatically calculates averages, standard

deviations, and relative standard deviations for each set of runs. Calibration plots were









all prepared using Origin 5.0 (Microcal Software, Northampton, MA). Detection limits

were determined as 3ca/m, where a is the standard deviation of the blank and m is the

slope of the calibration curve.

Table 9: Experimental Parameters

Calibration He Flow H2 Flow # Runs Power Integration Time
Rate (LPM) Rate (LPM) (W) Time (s) Between
Runs (s)
As* 6.27 0.44 5 1000 30 60

Cd 6.27 0.44 5 1000 10 60

Cr* 6.27 0.44 5 1000 30 50

Cu 6.27 0.44 5 1148 60 60

Fe 6.27 0.44 5 1000 30 60

Li 6.27 0.44 10 1000 30 60

Mn 6.27 0.44 5 1000 30 60

Pb 6.27 0.44 5 1000 20 60

P* 6.27 0.44 5 1000 30 60

Sr 6.27 0.44 5 1000 30 60

Zn 6.27 0.44 6 1000 12 40

ICP-MS-I 10 0.15 10 650 15 90
(axial)
ICP-MS-I 7.74 0.12 10 430 60 60
(lateral)
Sweat 6.27 0.44 10 610 30 60

Plant 6.27 0.44 5 1000 30 60
Material* I
*Electrode #5 (Fig. 24) with Ta cup was used instead of electrode #10









Results and Discussion

Single Element Calibration

Single element calibration was simply used to evaluate the performance of the

system for comparison with similar techniques. Single element calibration plots were

prepared for a variety of elements. Generally, the linear range (without self-absorption)

extended from the low to mid-ppb range up to the low to mid-ppm range. At least two

lines for each element could be observed. Calibration plots are provided in Figures 38 -

48. Linear and log-log line equations for both the non-self-absorption and self-absorption

regions of the curves are provided in Table 10, and figures of merit provided in Table 11.

Plots for Sr and Cr include only the non-self-absorbing regions of the plots. In the case

of Sr, the detector became saturated before the self-absorption region was reached;

memory effects became a problem for Cr above the region plotted. Except for lithium

(670.8 nm), all the elements in Table 11 have their most intense emission lines in the UV

/ low visible range of the detector. Detection limits are in the low ppb or pg range, with

linearity between 1.5 2.5 decades and precision generally 2 4 % RSD. Precision

slightly worsened for those elements analyzed with electrode #5 with the Ta cup.

Memory effects were evident in the analysis of some elements, causing log-log slopes to

be higher than the theoretical values.

Table 12 provides a comparison of detection limits for CMP-AES with FANES,

ICP-AES, and DCP-AES. Although detection limits are slightly worse for CMP-AES,

the competing techniques have the advantage of either higher sample volumes or solution

nebulization. In the case of FANES, the sample volume was ten times higher than for

CMP-AES, which means the sample would be more concentrated in the plasma volume.









For ICP-AES and DCP-AES, solutions were nebulized; one would expect the detection

limits to be lower in these cases. In some instances detection limits for CMP-AES were

lower than those reported for DCP-AES.

Table 10: Line Equations for Single Element Calibration

Element / Linear Self-Absorption Log Equation Log Self-Abs.
Line Equation Eu _tion Equation
As197.2nm Y=10.4X Y=3.65X+37.0 Y = 1.16X+0.923 Y=0.574X+ 1.32

As228.8 nm Y=49.7X Y= 26.1 X+ 128 Y = 1.22X+ 1.56 Y=O.720X+ 1.89

As 234.9 nm Y = 25.5 X Y = 13.7 X + 63.2 Y = 1.24 X + 1.26 Y = 0.729 X + 1.59

Cd 214+4 nm Y= 16.3X Y = 3.45 X +74.8 Y = 0929 X 1.3' 6 Y =0528X + 152

Cd 226.5 nm Y =7.61 X Y = 1.86 X+3'2.8 Y =0.945 X +L102 Y =0,564 X + 1. 17

Cd 228.8n Y =194 X Y =66.6 X +75.1 Y = 0936 X+127 Y =0.578X +2,18

Cd361.Om Y =17,6 X Y =O.571X + 13.2 Y =1.06 X+ 0,788 Y = M25 X+ 0,785

Cr 206.1 n Y =5.27 X Y =1. 15 X +0.471

Cr 267.7 n Y =24.3 X Y = 1.20 X+ 1.05

Cr 283.5 n Y-=37.5 X Y = 1.21 X+1.23

Cr 284.3 urn Y=22.6X Y = 1.20X+ 1.05

Cu 219.2Znm Y =2.09 X Y = 86X + 5.88 Y = I 15X +0.301 Y =0.414 X+ 0.655

Cu 219.9nxm Y =14.7 X Y =3.22 X+35.8 Y = .I4 X+ 1. 16 Y =O0390 X+ 144

Cu 223.0 rn Y = 29.8 X Y = .24 X+ 72.9 Y =1. 12X + 1.46 Y =0.378 X+ 1.75

Cu 224.7 nm Y =5.70 X Y = .53 X + 16.0 Y =L14 X+ 0,742 Y =0.405 X + 1.09

Cu 324.7 nm Y=35OX Y =66.0 X +860 Y =1.08 X+2152 Y = 0353 X +2.83

Cu 327.3 nn Y= 155 X Y= 31A4X +394 Y = 1.07 X +2.16 Y = .362 X +2.49

Fe 238.2 mm Y =4.26 X IY= 1.81 X+21. 9 1Y =1. 11 X +0.545 Y =0.607 X+ 1.00









Table 10: Line Equations for Single Elements continued.


Fe239.5nm Y=4.46X Y=l.92X+23.1 Y = 1.20X+0.047 Y=0.611X+0.083

Fe 240.4 nm Y=2.56X Y = 1.09X + 13.8 Y = 1. 18 X + 0.264 Y = 0.605 X + 0.789

Fe259.8nm Y=2.33X Y=1.04X+9.64 Y = 1.27X+0.188 Y=0.656X+0.644

Fe259.9 un Y=8.40X Y=3.47X+46.2 Y = 1.16X+0.805 Y=0.595X+ 1.31

Li 274.1 nm Y =0.443 X Y = 162 X +28.2 Y = 1.01 X 0410 Y =0.560 X+ 0.528

Li323.2nm Y = 182 X Y =0.800X + 105 Y =1,28 X- 0,260 Y = 0,623 X +1.2

Mn 257.6 mn Y = 96.7 X Y = 15.8 X + 43.0 Y = 0.956 X + 1.99 Y = 0.483 X + 1.79

Mn259.3nm Y=82.7X Y=13.4X+36.9 Y = 0.957X+ 1.92 Y = 0.481X+ 1.72

Mn260.Snm Y=35.2X Y=5.68X+ 15.7 Y = 0.950X+ 1.55 Y = 0.480X+ 1.35

Mn 279.4 nm Y=556X Y = 74.8 X + 252 Y = 0.936 X + 2.75 Y = 0.436 X + 2.54

Mn 279.8 nm Y=307X Y = 44.0 X + 138 Y = 0.948 X + 2.49 Y = 0.451 X + 2,28

Pb 216.9 m Y = 19.1 X Y =5.30 X+ 4.9 Y =0.98i Xk+ 1.37 Y =0.53X + 160

?b 220i3m Y = 10.4 X Yi =.48 X + 126 Y = 0.987 X + 1.04 Y =. .517.X. + 1.60

Pb 261.4 n Y = 9.9 X Y =16.4 X +281 Y =0,993X + 1.82 Y =0.503 X+ 2,15

Pb 283.3 nm Y = 32 X Y=41.4X + 572 Y =1.01 X+2116 Y =O0555 X +2.44

P213.6mn Y=5.56X Y=3.54X+21.9 Y = 1.07X+0.685 Y = 0.782X+0.964

P214.9nm Y=3.94X Y=2.70X+ 12.2 Y = 1.05X+0.547 Y = 0.813X+0.779

Sr 407.7 nm Y= 175 X Y = .949 X +2.29

Sr 421,5 nm Y =g4,5 X Y =0.974 X+ 1.95

Zn202.6nm Y=7.65X Y=l.65X+339 Y = 1.10X+0.752 Y = 0.430X+ 1.86

Zn206.2nm Y=9.45X Y=2.81X+382 Y = 1.12X+0.814 Y = 0.533X+ 1.77

Zn213.8nm Y=76.9X Y=17.0X+918 Y = 0.849X+2.03 Y = 0.482X+2.45

Zn 334.5 mn Y= 11.4 X Y = 4.64 X + 408 Y = 1.09 X + 0.943 Y = 0.660 X + 1.62






80


Table 11: Figures of Merit for Single Element Calibration

Element / Line LOD (ppb) LOD (pg) LDR (orders) Precision at mid-curve
_(% RSD)
As 197.2 nm 60 300 1 8.8

As 288.8 rnm 30 150 1 9.8

As 234.9 nm 30 150 1 10

Cd 214im 8 40 2 2.5

Cd 226.5 nm 60 300 2 2A4

Cd 228,8 nm 1 5 2 1.9

Cd '16 1,0ni 11 55 2 2,3

Cr 206.1 rm 240 1200 1.5 11.7

Cr 267.7 nm 200 1000 1.5 12.0

Cr 283.5 r 200 1000 1.5 12.0

Cr 284.3 wn 200 1000 1.5 11.8

Cu 219.2 nm3 10 50 1.5 2.4

Cu.219.9 nm 30 150 1.5 1.4

Cu 223.0 nni 10 50 1.5 1.4

Cu 224,7 un 25 125 1.5 2.0

CU 324,7 um 3 15 2.5 1.5

Cu 327.3 nun 1 5 2.5 1.3

Fe 238.2 nr 300 1500 2 6.0

Fe 239.5 nm 250 1250 2 7.2









Table 11: Figures of Merit for Single Element Calibration continued.

Fe 240.4 nm 200 1000 2 7.8

Fe 259.8 rn 290 1450 2 7.3

Fe 259.9 rn 250 1250 2 6.0

Li 274.I n 150 750 2 1.6

Li 323.2 in 1000 5000 1 2.7

Mn 257.6 mn 2 10 2 2.5

Mn 259.3 nn 1 5 2 2.4

Mn 260.5 rn 2 10 2 1.9

Mn 279.4 rn 0.6 3 2 2.3

Mn 279.8 rn 0.8 4 2 2.0

Pb216.9 n 4 20 1.5 5,6

Pb 220,3 nm 10 50 2 3,4

Pb 261.4 n 10 50 1.5 4.9

Pb 283.3 m 2 10 1.5 4,6

P 213.6 urn 32 160 1.5 6.5

P 214.9 rn 540 2700 1.5 5.1

Sr 407.7 n 49 245 1.587

Sr 421 .5 un 47 235 1.5 8.7

Zn 202.6 n 26 130 2.5 3.7

Zn 206.2 urn 24 120 2.5 3.5

Zn 213.8 mn 3 15 2 3.7

Zn 334.5 urn 26 130 2.5 4.6









Table 12: Comparison of Detection Limits (ppb) with FANES, ICP-AES, and DCP-AES

Element CMP-AESa FANESb ICP-AESO DCP-AESc

As 30 2 45

Cd 1 0.02 0.07 0.5
(5 pg) _(I.1 pg)
Cr 200 0.08 0.08 1
(1000 pg) (4.0 pg)
Cu 1 0.02 0.04 2
(5 pg) 1_2 pg)
Fe 200 0.09 0.09 3
(1000 pg) (4.5 pg)
Li 150 0.0004
(750 pg) (0.02 pg)
Mn 0.6 0.01 0.5

Pb 2 0.06 1 23
(10 pg) (3 pg)
P 32 15 75

Sr 47 0.002 2

Zn 3 0.04 0.1 2
((.5 pg) (2_p)
a Using 5 pL sample volumes and compomised multielement conditions.
b Using 50 pL sample volume.79
'Using pneumatic nebulization.89


In addition to the detection limit comparison provided here, Chapter 1 outlined

other figures of merit for these techniques, including linear dynamic ranges (FANES 5 -

6 orders,77 ICP-AES and DCP-AES 4-5 orders89), and precision (FANES 7% RSD,118

ICP-AES and DCP-AES 1 10 % RSD9). The CMP-AES technique is competitive in

detection limits and reproducibility, but does not have the range of linearity provided by

the other techniques.












228.8 nm
234.9 nm

197.2 nm


I lO
1 10
Concentration (ppm)

Figure 38: Calibration of Arsenic


.8 nm


.4 nm
.5 nm
.0 nm


0.1 1 10 1
Concentration (ppm)

Figure 39: Calibration of Cadmium


1000-.



100



10


10000


1000












283.5 nm
-267.7 nm
-284.3 nm

-206.1 nm


1000,



100-


10,














1000.


100-


10-






0.1


... .....~uf ~ h I


0.1 1 10
Concentration (ppm)

Figure 41: Calibration of Copper


324.7 nm
327.3 nm

223.0 nm
219.9 nm
224.7 nm
219.2 nm








100


1 10
Concentration (ppm)

Figure 40: Calibration of Chromium














259.9 nm
239.5 nm
238.2 nm


100-




10










0.1-















100-




10









0.1


323.2 nm


274.1 nm


10 100

Concentration (ppm)

Figure 43: Calibration of Lithium


1000


'I I I I I "I I
0.1 1 10

Concentration (ppm)

Figure 42: Calibration of Iron













0.4 nm
).8 nm
1.6 nm
.3 nm
).5 nm


'' 1 . . I' ' I' ' ' 1 '
0.01 0.1 1 10
Concentration (ppm)


Figure 44: Calibration of Manganese


283.3 nm
261.4 nm
216.9 nm
220.3 nm


Figure 45: Calibration of Lead


1000



100



10


1000-



100-



10-






0.1


. 1 . , 1 I I I' I I I '' 'I j I '."
0.01 0.1 1 10 100
Concentration (ppm)














100-






10.


1 10
Concentration (ppm)

Figure 46: Calibration of Phosphorus


1000-





100


.,213.6 nm
-214.9 nm















10








407.7 nm

421.5 nm


1 10
Concentration (ppm)

Figure 47: Calibration of Strontium