Magnetically-enhanced miniature glow discharge sources for atomic emission spectroscopy

MISSING IMAGE

Material Information

Title:
Magnetically-enhanced miniature glow discharge sources for atomic emission spectroscopy
Physical Description:
xi, 174 leaves : ill. ; 29 cm.
Language:
English
Creator:
Raghani, Anil R
Publication Date:

Subjects

Subjects / Keywords:
Atomic emission spectroscopy   ( lcsh )
Glow discharges   ( lcsh )
Chemistry thesis, Ph. D
Dissertations, Academic -- Chemistry -- UF
Genre:
bibliography   ( marcgt )
non-fiction   ( marcgt )

Notes

Thesis:
Thesis (Ph. D.)--University of Florida, 1995.
Bibliography:
Includes bibliographical references (leaves 166-173).
Statement of Responsibility:
by Anil R. Raghani.
General Note:
Typescript.
General Note:
Vita.

Record Information

Source Institution:
University of Florida
Rights Management:
All applicable rights reserved by the source institution and holding location.
Resource Identifier:
aleph - 002056872
oclc - 33815361
notis - AKP4892
System ID:
AA00002045:00001

Full Text













MAGNETICALLY-ENHANCED MINIATURE GLOW DISCHARGE SOURCES FOR
ATOMIC EMISSION SPECTROSCOPY















4*


ANIL


R.RAGHANI


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
































To my family for their love and inspiration.















ACKNOWLEDGMENTS

There are many people who have their contribution to this work. I would like to


sincerely express my gratitude to my research advisor Prof James D.


Winefordner for his


freedom


to work


independently


training


to think


innovatively.


"Politeness


embellishes wisdom" is a proverb in Sanskrit and Prof Winefordner is a true reflection of

this proverb and I will always remember this. I wish to thank Dr. Jerome E. Haky of

Florida Atlantic University who constantly encouraged me to pursue a doctoral degree


when I was his graduate student for my M. S.


degree.


I am fortunate to be a JDW group member and many thanks go to all the group


members, past and present,


who helped me bring about the completion of this work. I


would like to say many thanks to Dr. Ben Smith for his support and expert advice.


appreciate with many thanks the support extended by Chester Eastman and Dailey Burch

of the Machine Shop for their willingness to help.

There were many family moments that were missed while studying abroad and my

parents sacrificed my presence to those moments because they believed the importance of


my education. I am eternally grateful to my parents for their love and support.


I wish to


thank my brothers and sister for their encouragement and keeping me apprise of the
















TABLE OF CONTENTS


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

LIST OF FIGURES.................................. ...................................................................... Vl

LIST OF TAB LES .... ..... .................................. ........................... .... ..... .......................... ix
LSIT .


ABSTRACT.................................................................................................. ................. x

CHAPTER


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


fuow is~chia1Ie~rgel1,r rnertia *********** *****...... ....... .....


PRINCIPLES OF OPERATION OF GLOW DISCHARGE............................... 11


General Description and Properties ........ ....... .......................... ..................
Cathodic Sputtering ..P.. ..... ... .............. .... ........ .... . .. .............................
Ionization and Excitation Processes .................................................. ......


Hollow Cathode Effect...


MAGNETICALLY-COUPLED GLOW DISCHARGE SOURCES ................... 37


Literature Review ........... ...


Theory of Glow Discharge Influenced by Magnetic Field ........................
Scope of D issertation .... ..... ........... .................. .. ..... ........... .... .... .


SPECTROSCOPIC EVALUATION OF A MAGNETICALLY-BOOSTED


MICROCAVITY HOT T.OW CATHOTE nDRTSCHARGPr


47










Emission of the Sputtered Material ........................... ................ ... .... .76
C conclusions. .. .. ............ .......... .... ....... ...................................... ............... 83

5 CYLINDRICAL MAGNETRON GLOW DISCHARGE FOR AES .................... 84

Design of a Cylindrical Magnetron Source........ ..... ..... ......... ................. 84
Charge Transport Processes in the Microcavity HCD................................ 88
Limitations of the Quartz Chamber Design............................................. 95
C conclusions ...... ..... .......... ........................... ................... .......... ............ 98
6 PLANARMAGNETRON GLOW DISCHARGE .. ................ ................. ...... ... 100

Strategy for Design of Magnetron Glow Discharge Source................... 100
Description of Planar Magnetron Design.... ............................ ............... 1 01
G general Plasm a Features ........ ...... ....................... ...................... ..... ...... 108
Measurements of the Magnetic Field Strength............ ........................... 113
Voltage Current Characteristics of the Planar Magnetron
G low D ischarge................... .................... .................. ..................... 116
Atomic Emission of the Cathodic Material............................................. 125
Optimizationof the DischargeParameters.............. ........................... 129
Comparison of the Calibration Curves and Limits of Detection
of Som e Elem ents.......... ................................................. ... ............. 140
Simultaneous Multi-elemental Analysis of Nanoliter Samples................ 154
C onclusions. ...... ..... ................... ................... ....................... ..... ............... 155

7 CONCLUSIONS AND FUTURE WORK ......... ............................ ....... ...... .. 160

Future D direction ... ............................................................................... 163
R E FER EN C E S ............................ ..................... ............................................... 166

BIOGRAPHICAL SKETCH ............................................ ................................ 174
















LIST OF FIGURES


FIGURE


Schematic of a simple glow discharge device............................ ............ ..... .3

Voltage-current characteristics of a glow discharge ................................... 13

Various regions min a glow discharge........................................................... 16

The sputtering processs............................................................................... 20

Collisional processes mi a glow discharge ......................................... ............. .. 26

Hollow cathode effect ............................................................................... 32

Cross-section of a hollow cathode discharge.... .......................................... 34

Magnetron glow discharge sources used in thin film
deposition technology .......................................................................... 39


(a) Motion of a charged particle in a magnetic field (b) motion of a
charged particle in a mutually perpendicular magnetic and
electric field ............................................................................


Schematic diagram ofthe quartz glow discharge chamber.......................... 49

M icrocavity hollow cathodeeee .... .................. ..... .................. ................... 52

Schematic diagram of the experimental set-up ........................................... 54

Modified design of quartz glow discharge chamber.................................... 58

amnI atr rraal' rail an. netffa~ nnank^b4rn ne A 1il a













4-8

4-9

4-10

4-11

4-12


Effect of magnetic field on aluminum atomic emission ............................... 68

Radial profile of aluminum microcavity HCD ............................................. 71

Effect of magnetic field on magnesium ion emission................................... 73

Effect of magnetic field on argon ion emission ............................... ................. 75

Variation of electron temperature with magnetic field strength................... 81

Schematic diagram of cylindrical magnetron glow discharge for AES ........ 86

Dependence of atomic emission of magnesium on pressure........................ 90

Relation between the emission intensity and ambipolar


diffusion coefficient ..........


......................... .. 97


Stainless steel chamber design for planar magnetron glow discharge ........ 104

The cathode assembly for the magnetron glow discharge......................... 106

M agnetron glow discharge .... i.................. ................. ........ ..... ....................... 110

Glow discharge without the magnetic field............................................... 112

Magnetic field profile on the cathode surface.................................. ......... 115

Side view of the planar magnetron glow discharge source........................ 118

Current-voltage characteristics of the glow discharge in the presence
and in the absence of the magnetic field............................................. 120


Current-voltage characteristics of the magnetron glow discharge


at various pressures .........


Atomic emission of aluminum from the glow discharge with and
without the magnetic field ....................................................


- A


111111111111











6-12


6-13



6-14


6-15


6-16


6-17


6-18


6-19


Optimization of discharge current and pressure from atomic emission of
magnesium solution residue from the magnetron glow discharge........ 135

Comparison of temporal profiles of atomic emission of 100 ng of
boron residue from the glow discharge with and without the
m agnetic field ............. .......... ............................................................. 139

Calibration curves of magnesium from the glow discharge with and
without the m magnetic field.................................................................. 143

Calibration curves of silver from the glow discharge with and
without the magnetic field ...... .................. .............................. .. .... .... 145

Calibration curve of boron from the glow discharge with and
without the magnetic field................... . .. ... .. .. ..... .......... .. ................ 147

Calibration curve of europium from the glow discharge with and
without the magnetic field........ ........................................................... 149

Calibration curve of copper from the glow discharge with and
without the magnetic field ..... ........... ..................................... ............ 151

Comparison of the atomic emission spectra for the multi-elemental
analysis from the glow discharges with and without the
m agnetic field .................................................................................... 157
















LIST OF TABLES


Glow Discharge Processes. ...................... ...... ..... .... .. ............ .................... 27


Excitation Potential of Rare


Relationship between Intensity Maxima and Energy of Upper Level........... 82

Relation between the Ambipolar Diffusion and the Discharge Pressure ...... 94

Details of the Planar Magnetron Cathode Assembly................................. 107

Comparison of Limits of Detection (3bzank) for the Glow Discharge


with and without the Magnetic Field.....


Comparison of Limits of Detection (3ao ) for the Glow Discharge
with and without the Magnetic Field................................................


Gases ................... ................... ................... .... 29


152
















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


MAGNETICALLY-ENHANCED MINIATURE GLOW DISCHARGE SOURCES FOR
ATOMIC EMISSION SPECTROSCOPY


By

Anil R. Raghani


August, 1995


Chairman: Prof James D.


Winefordner


Major Department: Chemistry

Applications of low pressure glow discharge (GD) sources are well documented


for spectrochemical analysis of solids as


well as


liquids.


Although these


sources are


extremely useful for chemical analysis by atomic emission


spectroscopy,


the emission


intensity output is rather limited relative to other atomic emission sources.

There have been numerous efforts to boost the emission signal from GD sources.

In this dissertation, three magnetically-enhanced GD sources were designed and evaluated

with respect to their spectroscopic characteristics. In all three designs, the source volume

was limited to less than 100 mL in order to minimize the dilution of atomic vapor of the











magnesium were found to increase with increasing magnetic field strength, reaching a


maximum for each analytical line.


The attainment of the maxima was attributed to the


increase in the electron


temperature and


radial


diffusion


electrons with increasing


magnetic field strength.


compact


cylindrical


magnetron


MCHCD


was


designed


using


permanent


magnets. An enhancement up to a factor of three was observed for magnesium atomic

emission. The enhancement factor of the atomic emission in the presence of the magnetic

field was found to be linearly dependent on reduction in the ambipolar diffusion coefficient

at lower pressures.


A planar magnetron GD source with a chamber volume of about 60 mL


was


designed and evaluated for the analysis of less than a microliter of aqueous samples. Limits

of detection for magnesium, silver, boron, europium and copper ranged from 3 to 40 times

lower than the source without the magnetic field when the measurements were made


under the compromised discharge conditions


each type of source.


The improved


detection limits for the magnetically-boosted GD source was attributed to the increased

current density of the discharge due to the formation of a plasma ring in the presence of


magnetic field.


Finally,


magnetically-enhanced


planar


cathode GD


source


simultaneous multi-elemental analysis of discrete aqueous samples was shown to be more

sensitive than the discharge without the magnetic field.















CHAPTER 1

INTRODUCTION

Glow Discharges in Analytical Chemistry


Glow discharges have been used for many years for fundamental spectroscopic

studies; they have also been used extensively for analytical applications in atomic emission


spectroscopy (AES)'4


, atomic absorption


spectroscopy (AAS),"5


atomic fluorescence


spectroscopy (AFS)93and resonance ionization.14 More recently, glow discharge sources

have also served as ion sources in glow discharge mass spectroscopy (GDMS).s"s'


The glow discharge devices are reduced pressure (1


discharges formed between two electrodes.


-10 torr), inert atmosphere


When a high electrical potential (300-1000 V)


is applied between a cathode and an anode, a breakdown of the enclosed gas results,

forming positively charged ions and electrons. A schematic of a typical glow discharge is


shown in Figure


In analytical applications of a glow discharge, a solid sample is


usually used as the cathode. If the sample is a liquid solution, it


is deposited on the


cathode surface and its dry residue is analyzed.

Low pressure glow discharges have several advantages over atmospheric pressure

electrical discharges such as arc and inductively coupled plasmas. First, inert atmosphere
































Figure 1


: Schematic diagram of a simple glow discharge.






3





















-Negative Glow

- Cathode (Sample)










Consequently,


spectral


chemical


interference


from


these


molecular


gases


greatly reduced.1 Second, glow discharges are operated at lower currents which help to


sharp


spectral


lines


with lower background


radiation."


Sharp


spectral


lines


particularly desirable in order to reduce the spectral interference.


Third, atomization in a


glow discharge is governed by a sputtering process. As a result, analyte along with its


matrix


volatilized


simultaneously


making


these


discharges


prone


to matrix


interferences2


Fourth, unlike the atmospheric electrical discharges, glow discharges do


not attain local thermodynamic equilibrium (LTE). As a result, self-absorption is greatly


reduced in case of a glow discharge.


The reduction in self-absorption causes improved


sensitivity of these sources in analytical atomic spectroscopy.


Finally, glow discharge


sources are quite suitable for a wide variety of sample analyses. Conducting solid samples

are directly analyzed without a need for their dissolution, and nonconducting samples can

be made conducting by mixing the samples with a conducting matrix such as fine copper


powder.


Recently, Marcus and coworkers extensively studied a radiofrequency based


glow discharge source for the direct analysis of non-conducting solid samples.242 Further,

these discharges are very useful for the analysis of the samples which are available in

limited quantities, as is often the case with biological or forensic samples; this occurs

because glow discharge sources do not require much sample preparation, and thus can be


used


directly


analyze


small


volume


liquid


samples.29-31


Nonmetals,


which


challengino tn nther exc.itatinn cniirere ra.n hp pncilv mpanwnir hv olrnw r ieharoA crirera










which is most commonly employed in AAS and AFS.


One of the factors hampering a


wider acceptance of this source in atomic emission spectroscopy could be the somewhat


limited


emission


intensity


output


achievable


when


discharge


is operated


under


conditions which fully preserve its basic merits.


The most recent trends in research on


glow discharge sources in atomic spectroscopy is towards improving upon the emission

intensity output in AES and to increase the atomic population of analytes in AAS and

AFS.33


Historical Develooment


observations


phenomena


glow


discharge


back


to the


nineteenth century. The names of Hittorf, Geissler and Crookes are well known to plasma

physicists. A milestone in the evolution and development of low pressure discharges was


reached by


Paschen who constructed the Geissler tube.


He was able to study the


structure of the helium spectrum thoroughly and found that his results agreed well with


Bohr'


theory.


1929,


Schiller"3


investigated


hyperfine


structures and


later found


applications of hollow cathode glow discharge sources for the analysis of small samples.


Later on, Frerichs36


Takahashi"7


clarified the mechanism of some basic processes


leading to excitation in


glow


discharges.


Sawyer3"


added


to the


knowledge of


glow


discharges. In 1947, McNally et al. 9 recognized the analytical potential of glow discharge

for the application of halogen determination, after which there was a revival of interest. A

cpnnqidprahlp amnMint nc~fr orr-h ,17na f,-,,asA ,nr +1,0 ,nko+, ,mnU 1 int ,.. vii7tliA40 a-.4








6

Until this time, the glow discharge was not conceived as a technique for trace analysis in


AES.


This is mainly attributed to the stringent need for a pair of demountable electrodes,


which serve as the analytical sample in solids analysis. Interconnection of the electrodes to


the glow discharge chamber to obtain proper vacuum conditions is often


time consuming


and thus impractical for routine analysis. Despite these impracticalities, glow discharge


sources were quite popular for trace analysis by


AES in the former USSR.


1968,


Grimm


developed a glow discharge source for the analysis of both conducting and non-


conducting samples. By the time glow discharge sources were capturing the interest of

analytical atomic spectroscopists, other electrical discharges were being developed rapidly

because of their high sensitivities even though the latter sources were relatively expensive,


for example,


the inductively coupled plasma.


This competition from other discharges


resulted in a declining interest in glow discharge sources for atomic emission spectroscopy


in the


1960s and


1970s.


However, in the


1980 and


1990s, increased interest in glow


discharges in atomic and mass spectrometry was rekindled.


For example,


Winefordner'


group investigated the application of a glow discharge source as an ideal atom reservoir in


1994.


miniature


glow


discharge


source


was


designed


laboratory


measurements of rare earth elements and limits of detection up to the attogram range were


approached


laser


excited


atomic


fluorescence


glow


discharge


spectroscopy.42


Throughout these years of research efforts, it is now believed that glow discharge is a

anriwiTno /.Q0hl0l atP ont-lann A+* a-^4an'4^* at








7

Boosted Glow Discharge Sources in AES


As stated in the previous sections, the intensity output of a glow discharge source

for atomic emission studies is limited when it is to be used for trace elemental analysis.

Major efforts have been recently afforded to the development of the techniques which


enhance the atomic emission intensity.


These techniques are known


as boosted-glow


discharge techniques.


The two main approaches to boost the analytical atomic emission


signal are either by modifying the geometry of glow discharge source or by applying an

external source where the interaction of an external field on these discharges result in


enhancement in the emission intensity.


The latter are commonly known as operation-


modified glow discharge sources.


In geometry modified glow discharges,


the diameter of the


hollow


cathode is


reduced to less than


2 mm. It was observed by Czakow" that the emission intensity is


inversely proportional to the diameter of the hollow cathode glow discharge.


This type of


glow discharge is referred to as a microcavity hollow cathode (MCHC) glow discharge.

An empirical relation between the atomic emission intensity and the diameter of the hollow

cathode is given by the following equation:45


where I


is the intensity of the emitted radiation, d is the diameter of the hollow cathode


and k is a constant.








8

discharge has been shown to be an excellent atomic emission source for simultaneous

multi-elemental analysis of nanoliter volumes of discrete samples with the detection limits


ranging from sub-picograms to femtograms.46


In another application of MCHC,


renal


fluids were analyzed for biologically important elements such as sodium, potassium, etc.47


Other


geometry-modified


HCD


emission


intensity


boosting


have


been


investigated.


These


are conical


bottom,4"


spherical


bottom,49


stepwise


variable


inner


cavity,50 etc.


The concept behind these geometry-modified glow discharges is to increase


the stability of the glow discharge.


In the operation-modified glow discharge sources,


various types of external fields


have been used for boosting the emission intensity. Another approach to enhance the


emission


intensity


been


supernmposmg


microwave


discharge


on the


glow


discharge in which the ground state atomic population in conventional glow discharge

sources was further excited by applying the microwave discharge as a supplementary


source.s5153 Caroli et al.54


have successfully shown the application of the microwave-


boosted HCD for the analysis of sediment residue.

The emission intensity of an analyte in a glow discharge is proportional to the


applied


discharge current.


When


the d.


current


is replaced


current


pulse,


noticeable


unprovement


in the


spectral


intensity


was


reported


since


higher


current


amplitudes can be applied without overheating the cathode.


Overheating of the cathode


resIlts in thermal atnmihatinn nfths rthnthr mnatnvaI ra.lCnin a;l, nl^.^A4n ^..k.na--








9

employed a capacitively coupled radiofrequency-boosted hollow cathode discharge for


AES.


A radiofrequency voltage burst was followed by each current pulse; in this way, the


unexcited


atoms


during


current


pulse


were


fully


excited


subsequent


radiofrequency burst. Intensity enhancement up to three orders of magnitude was attained

without sacrificing the sharpness of the spectral lines.

A secondary discharge of high current density was placed between the cathode and

the anode by Van Gelder.57 This configuration is termed as an auxiliary discharge boosted


glow discharge.


atoms.


One of the excitation processes is by collisions between electrons and


This approach results in an increase in the electron density in the vicinity of the


cathode which in turn results in higher excitation of the analyte atoms giving rise to

increased intensity of the atomic emission.

A high rate of sample vaporization can be obtained using a tubular graphite furnace


as a hollow cathode.


The furnace is heated at a rate of up to 2000 K s'


, thus forming an


instantaneous atomic cloud. Since the furnace also acts as a cathode, the analyte atoms

experience excitation as in conventional glow discharge with the added advantage of a


higher


emission


intensity.


technique


known


as Furnace


Atomic


Non-thermal


Excitation Spectrometry (FANES), was developed by Falk5' and coworkers.


Banks et al.59'60


designed a glow discharge device in which the flow of the fill gas


was directed onto the sample cathode surface through six channels.


This device is known


aR a iect-naeittPA lO nw ditrhnaro Atnmnn PmiEcinrn intpnictv nnhnrhamYant lnr 4t, a fatr t










of magnetic field


was accomplished


applying the magnetic


either parallel


perpendicular to the cathode axis. In each case, the electron residence time was lengthened


resulting


m more


atoms


being


excited


giving higher


emission


intensities.61


intensity enhancement technique is presented in detail in Chapter 3.

In all of the above boosting methods, compromises are made to obtain the highest


enhancement


factor


atomic


emission.


Some


these


factors,


which


are traded-off


include, instrumentation complexity against analytical sensitivity, cost of instrument versus


sensitivity,


etc. Thus far, no ideal boosting technique has been developed,


which further


warrants


investigations


in this


area


to consolidate


expand


analytical


capabilities of boosted glow discharge sources for atomic emission spectroscopy.
















CHAPTER


PRINCIPLES OF OPERATION OF GLOW DISCHARGE

General Description and Properties


Stability


and type of a low pressure electric discharge is dependent on the voltage


and the current attained by the discharge. At the breakdown potential Vb of


gas, a very low current (<


the enclosed


flows between the electrodes and this discharge is


known as


Townsend discharge shown in Figure


Electrons emitted by the cathode


move to the anode, and during this passage they produce positive ions which are attracted


to the cathode.


Townsend discharge,


there is a steady-state relationship between


positive ions formed and the discharge current.


This discharge is entirely dependent upon


the external ionization source (the applied voltage) and extinguishes upon the removal of


electrical potential to the electrodes.


In other words, the


Townsend discharge is not a


self-sustaining discharge.


Townsend discharge is followed by a transition region, leading


to the normal glow discharge where the voltage remains constant as the current is further


changed.


In the normal glow discharge,


although the current is increased the current


density remains constant since the area covered by this discharge increases proportionately

with current. After the cathode area is fully covered, further increase in current requires a

trac*^to ^0+h^X/A A.vkonfia.,^ .-.lZU :n ^^ ^MI n2'At tan-.aaat.,- ^j.l^J --^ -* .% .1



























Figure


Voltage-current characteristics of a glow discharge.





















Abnormal


Glow


Discharge


USSR 56 .psSSUOg U 635 mm



Li


U
U
U
U
U
S
U
S
S
U
S
U
U
U
S
U
S
U

*5 SSUeSS ems mu U* U mu Urns5m4
S
S
S
S
U
S
U
U
U
U
S
U


Normal Glow

Discharge


U U saw.us mum


III -II


10-10


10-8


10-6


Current (A)


Townsend

Discharge


*


-4










by thermionic emission, resulting in the arc discharge.


analytical


spectroscopists,


abnormal


glow


discharge


or simply


glow


discharge, is of importance since the current attained by this discharge is sufficient to

produce atoms from the analytical sample, as will be explained in the subsequent sections.

The glow discharge is characterized by the existence of eight different spatial regions as


shown in Figure


63 Electrons emitted from the cathode are driven to the anode by the


electric field. Before reaching the anode, electrons cause ionization and excitation of the

fill gas atoms depending on the energy associated with the electron at each point of the

electron pathway to the anode. Near the cathode surface, electrons are accelerated in a

strong field executing very few ionizing collisions in their initial acceleration stages due to

the low energy attained. As electrons move farther from the cathode where the electric

field is weaker, they are now more energetic facilitating more ionization of the atoms with

simultaneous electron multiplication. The two regions observed as a consequence of these


processes are known as the Aston dark space and the cathode layer, respectively.


farther from the cathode, the electric field becomes even weaker, the electrons which did

not dissipate their energy in the preceding regions now have higher kinetic energy giving


more ionization of the atoms.


This effect forms a distinct region known as the cathode


dark space (CDS), also sometimes referred to as the Crooke's or Hittorf'


dark space. The


CDS carries the entire cathode potential fall of the discharge." A fraction of the electron

population escaping from the CDS enters the brightest zone called the negative alnw




























Figure


Various regions in a glow discharge.

















Positive Column


Faraday
Dark Space


Cathode


Layer


Aston
Dark S


(+)


Anode Glow (Ciode)


Amde










unexcited atoms,


strike the cathode surface producing secondary electrons which are


helpful in sustaining the discharge.


There are two main groups of electrons that enter the negative glow:


(i) fast


electrons,


which retain


their


energy,


slow


electrons,


which are


energy


electrons produced in the CDS having lost some of their energy due to inelastic collisions


with


discharge


particles.


energy


of the


slow


electrons


insufficient


cause


ionization of the fill gas atoms but their energy is enough to excite these atoms upon


collision.


The excited atoms cause the appearance of a characteristic negative glow.


the other hand, the fast electrons penetrate the negative glow causing more ionization, and

their energy is dissipated with the distance they travel through the negative glow. As the


fast electrons exit the negative glow,


they undergo recombination with


positive ions.


However, farther from the cathode and after the negative glow region,


the field rises


slowly, lowering the probability of recombination, resulting in the Faraday dark space.


Electrons, pulled out of the negative glow by this relatively weak field, may


gain sufficient


energy to excite and ionize atoms. At this point, the positive column is developed. After

leaving the positive column, electrons are attracted to the anode while positive ions are

repelled by the anode; consequently, a negative space charged is generated, forming the

anode fall. Electrons moving through the anode fall gain energy as they are accelerated.

These energized electrons have sufficient energy to excite and ionize gas atoms near the


anode nivina an anode mlow.


Of all these regions.


the negative slow has the highest










negative glow directly gives spectrochemical information about the cathode. Advantage of


this fact is taken in atomic spectroscopy. In atomic emission spectroscopy,


the emission from the negative glow,


one measures


whereas in atomic fluorescence spectroscopy the


negative glow is irradiated by an external line source such as a hollow cathode lamp or a

laser and the atoms contained in the negative glow are excited to a higher energy level. On

radiative de-excitation, the fluorescence is measured. Similarly, a light passing through the


negative


glow is absorbed by atoms


in the negative


glow.


The fraction


of the


light


absorbed


is measured


m atomic


absorption


spectroscopy.


negative


glow


possesses ions from the cathode, which may be collected by a mass spectrometer in glow

discharge mass spectroscopy.


Cathodic Svutterine


Atomization in a glow discharge is by a nonthermal process known as sputtering.

When a positive ion strikes the cathode (analytical sample), the following processes are

possible: (i) implantation of ions into the cathode; (ii) reflection of ions from the surface of

the cathode, leading to recombination or loss of ion energy; (iii) ejection of electrons from

the cathode surface, in order to sustain the glow discharge; and (iv) ejection of the atoms


from the cathode,

sputtering process.


which is sputtering of the cathodic material. Figure 2-3 represents the

The atomic cloud formed as a result of sputtering is representative of


the bulk of the cathode. Less than 0.8 % of the sample is thermally vaporized, while nearly

all of the atomization in a dlow discharge is hv snutterino 6s



























Figure


:The sputtering process.






















Primary
Ion


Sputtered Particle
(ion or neutral)


Cathode


Implanted
Ion


Escape Depth
(3-4 A


Primary Ion
Penetration Depth







21

Sputtering is a direct result of the process when an energetic specie imparts its


kinetic


energy


to the


surface


atoms


cathode


through


elastic


collisions.


bombarding species loses its energy to atoms in its vicinity producing what is known as a


collisional


cascade.,"'6


An atomic specie is ejected


when


energy


released


collisional cascade is greater than the surface binding energy of the cathode; the ejected


particles are transformed into gas phase.


The minimum energy required to eject atoms


from the cathode is referred to as the sputtering threshold,


which is usually of the same


magnitude as the heat of sublimation of the cathodic material and usually four times the

binding energy of the cathode.68 The sputtering threshold is dependent on the nature of the

target material and also on the fill gas employed in the discharge device. In addition to

atoms, sputtering also produces secondary electrons, ions (positive and negative), photons


and atom clusters from the cathode surface.


Over 90 % of the sputtered particles are


ground state neutral atoms.4

An empirical parameter characterizing the sputtering process is the sputtering yield

or sputtering efficiency, which is defined as the number of the atoms ejected per incident

ion. In practice, the sputter yield, S, is expressed as:69


WNe
Mi t


(2-1)


where W is the weight of the sample (cathode) lost (g), N is the Avogadro'


number


(mole'), e is the electron charge (C), M is the atomic mass of the target material (g), i' is










The ion current is related to the discharge current i as follows:


(2-2)


where y is the number of electrons released per incident ion, and it is approximately 0.1

for argon, the fill gas most commonly used in the glow discharge devices.

The factors which contribute to the sputtering yield are the energy of the incident


ion, angle of incidence,


type of the fill gas,


and the nature of the target material.


complete expression showing the relationship between all these factors is given below:67


3cmyntE:


(2-3)


K (mti + m)2 U


In this equation, E is the energy of the incident ion (eV), mi and m, are the masses of the

incident ion and the target atom (g), respectively, Uo represents the surface binding energy

(eV) and a is a monotonically increasing function of the ratio m/mr and is equal to 0.17 if


the ratio is 0.1 and increases to 1.4 for m/mr


The sputtering yield increases linearly with energy of the incident ion once the


sputtering threshold is attained by the bombarding ion.


The energy of the incident ion is


determined by the external applied voltage.


Thus,


the higher the applied voltage,


higher is the incident ion energy. Usually, the ion energy is considered equal to the applied

voltage. At a very high energy of the incident particle on the cathode surface, the particle








23


sputtering. Xenon would thus give the highest sputtering efficiency; however, owing to its

exorbitant cost, argon is preferred as a fill gas.

The sputtering yield tends to vary only slightly with target temperature, but rises


sharply


when


temperature


cathode


nears


vaporization


point.


bombarding energies, the sputter yield decreases with increasing target temperature.


is due to annealing of the surface species producing more strongly bound atoms.68 Because

of continuous sputtering of the cathode, it tends to get hot which results in inefficient

sputtering. For this reason, the cathode is often cooled in a glow discharge device.


The gas pressure also has a tremendous effect on the sputtering yield.


At higher


gas pressures, the sputtered atoms do not diffuse away from the cathode due to their

shorter mean free path at these pressures. Consequently, the sputtered atoms diffuse back

to the cathode, thus giving less target material loss. It has been estimated that 90 to 95%

of the sputtered atoms redeposit in glow discharge sputtering. The secondary electrons

generated during the sputtering process recombine with the incident ions resulting in an

inefficient sputtering. Also, if the sputtered species ejected from the cathode is a positive

ion, it is attracted back to the negatively charged cathode, resulting in lower values of

sputtering yield. A detailed review of the glow discharge sputtering can be found in the


literature.69'71


Ionization and Excitation Processes


Clnllkinnnl


nrnrcse! e ars rsannnaKhls fnr the


fnmr~nntnn C;~;A o4 rtA


pyF;~prl








24

only a few are significant with respect to ionization and excitation in the glow discharge.

The two primary processes responsible for ionization and excitation are summarized in


Table


Collisions with electrons. Based on the energy associated with the electrons and

their origin, the electrons present in the negative glow are classified in three categories.

First, the primary electrons are high energy electrons (20-25 eV) that are produced from

the cathode surface during ion bombardment and gain energy as they are accelerated


through


cathode


dark


space


toward


negative


glow.


Second,


secondary


electrons with energy ranging from


to 10 eV originate from ionizing collisions or from


prunary


electrons


which


have


their


energy


through


collisions.


Third,


slow


(ultimate) electrons are low energy electrons (0.05-0.6 eV) that result from primary and

secondary electrons when these latter electrons are thermalized to the plasma temperature

through numerous collisions with other plasma species. Electron number densities on the


order of 106


10'-108


and 109-10"


have been estimated for the primary, secondary,


and ultimate electrons, respectively.74 Whether collisions with electrons result in ionization

or excitation of the species in the glow discharge depend on the energetic of the colliding


electrons. Ultimate electrons do not produce excited or ionized species,


while secondary


electrons,


having moderate


energy


have


sufficient


energy


cause


excitation


of the


colliding species but do not have enough energy to produce ionization of the gas or


os uttered atoms


..N~v t v-b .-


The excited atoms can underon further enllisionn with nther el~ntrrnn


bftb l, $gL, *


I


































Figure 2-4


: Collisional processes in a glow discharge.



























IRecombination


26














oi/




Ionization I.
_______H.__* l & ********* if t-


ast Atoms


*
I Negative Ions
* *
C*


Electrons


Radiatio


Positive Ions


Elastic Collision


Excitation


Ionization


Heat


Electrons










Table


Glow Discharge Processes


Primary Ionization and Excitation Processes


Electron Impact
M+e" -- M


+ 2e

+e"


Penning Collisions
Me+Am" -- M


+ A +e"

+A


MO+ Am


Secondary Processes


i. Nonsymmetric Charge Transfer


Symmetric (Resonance) Charge Transfer


+ X%0W)


X(fat) + Xr(dow)


Dissociative Charge Transfer


MX + A'


Associative Ionization


M+A.


Mt"


Photoionization and excitation


M + hv

M +hv


Cumulative Ionization


MO+e-


4 ~ ~~ 4 -r 4 t


+ 2i


MO+e~


MO + A''


+ AO


+ Ao


MO + A'


Xt~cfi~t:


+~-t-AP


.. r










Collisions with excited species.


The metastable atoms are the atoms in the energy


level which have a low propabality of radiative deactivation.


These long-lived metastable


atoms can produce an ionized specie or another excited species by the process known as


Penning ionization as shown in Table


If the energy of the metastable atom is greater


than the ionization potential of the neutral species, then it can easily cause ionization.

Similarly, if the metastable atoms have energy greater than the excitation potential of the


species with which they collide then they produce excited species in the discharge.


metastable energies of most rare gases used in glow discharge spectroscopy are higher

than most elemental ionization and excitation potentials (Table 2-2). As can be seen from


Table


the heavier rare gases have lower excitation potentials. However, heavier rare


gases are preferred for their better sputtering efficiency,


as explained in the previous


chapter. On the other hand, lighter rare gases have higher excitation potentials. Therefore,


a compromise is always made in glow


choosing a rare gas.


discharge atomic emission spectroscopy when


A gas should have good sputtering efficiency and good excitation


potential. In glow discharge atomic emission spectroscopy, argon is most often used since


it has acceptable sputtering and


excitation efficiency and low cost.


Hollow Cathode Effect


The electrical and spectroscopic characteristics of


a glow discharge with a planar


cathode can be significantly altered by using a cathode in the form of a hollow cylinder.

Tin. omiorn imntanoix tic ha a n fniuinA in nrraaco b, 9'_ nrr/lara nV moarnnhnAa inc t hvr









Table


: Excitation Potentials of Rare Gases


Rare Gas Energy (eV)



He 19.82



Ne 16.62



Ar 11.55



Xe 8.32








30

was employed and an anode was placed opposite to the cathode." The distance between

the two arms of the cathode was gradually decreased and its electrical and spectroscopy

characteristics were noted. Each arm of the cathode formed a negative glow when their


separation


greatest.


distance


between


the two


arms


decreased,


negative glows of the corresponding arms


began to merge into one another.


This is


illustrated in Figure


The emission intensity increased as the separation between the


two arms of the U-shaped cathode decreased. Generally, in hollow cathode devices used


today, a cylindrical hollow tube with either one or both ends open, is used.


The hollow


cathode discharge retains all the properties of a planar cathode glow discharge and the

positive column, Faraday dark space and anode glow are outside the cavity of hollow

cathode discharge devices.

The cathode layer in a HCD is circular along the inside wall of the hollow cathode


(Figure 2-6).


The cathode sheath region (Aston dark space, cathode glow, and cathode


dark space) is much narrower in HCD compared to a planar cathode GD.


The electric


field decreases with distance away from the inside cathode surface and the negative glow


at the center of the hollow cathode has the least electric potential.


In HCD,


the CDS


carries almost the entire cathode fall potential as it does in the case of a planar cathode

GD. However, since it has to carry this potential through a much smaller distance, the field


gradient is much steeper,


which results in ions having a higher average energy as they


imninnfe on the lirfnapr nfthle r-thbndP AC a radilt th
































Figure


: Hollow cathode effect.






32




MP

d (+)








C(+)




/mb

































Figure 2-6


: Cross-section of a hollow cathode discharge.






















Cathode Wall















Negative Glow


Hittorf's Dark
Space















Cathode Layer










center of the hollow


cathode and the intensity maximum


does not change its spatial


position with discharge current.

The mean free path of electrons in a hollow cathode is on the order of the diameter


of the cathode cavity.8


As a result, the high energy electrons cross the negative glow,


without losing their energy through collision and they reach the opposite side of the inner

wall of the hollow cathode. Each time an electron passes through the negative glow it

excites the species present in the hollow cathode. As the electrons reach the opposite wall,

they are bounced back due to the higher cathode fall potential. This causes the electrons to

make multiple passes through the electrodes between the opposite sides of the wall. In

other words, the species are excited a number of times through collisions in the hollow

cathode, as opposed to the case with a planar cathode where the electron passes through


the negative glow just once and


it is eventually thermalized


to the plasma gas


temperature.

In the planar cathode glow discharge, one of the processes by which electrons are

emitted from the cathode surface is photoemission. Photons have a very small probability

of reaching the planar cathode surface, depending on the cathode surface area. Electrons

produced as a result of photoemission in the planar GD, therefore, do not contribute to the

excitation and ionization processes in the glow discharge. In the hollow cathode, photons

are confined inside the cavity of the cathode, and thus they have a higher probability of

striking, the cathode surface The likelihood nfnhntnna qtrilrino the nathnds inrrmaro frnnm










photoemission.


The electron number density in a HCD is,


therefore,


higher than the


electron number density in a planar GD.

Plasma species also strike the cathode surface in HCD and they have a higher


residence time inside the hollow cathode.


These species are excited as they pass through


the analytical zone. The escape rate for the rare gas and analyte species is less than 2% of

that for a planar cathode glow discharge system.82
















CHAPTER 3
MAGNETICALLY-COUPLED GLOW DISCHARGE SOURCES


Literature Review


In chapter 1,


a brief outline of the methods for the enhancement of atomic emission


intensity output from a glow discharge source was given. One of the methods by which


the emission


intensity


can be enhanced


is by


applying a magnetic


to the


glow


discharge.


Magnetically-coupled


plasmas


have


been


studied


extensively


the term


'magnetoplasma'


is frequently used for the plasmas which are modified by an external


magnetic field. Physicists and engineers have studied magnetoplasmas extensively and

have exploited them for plasma confinement.'345 Vacuum engineers have found a number

of practical applications for these devices as vacuum pumps,'6 pressure gauges,7 and very


high efficiency sputter-type vapor sources for thin film fabrication.",


The glow discharge


sources with magnetic field employed in thin film deposition technology are known as


magnetron glow discharge sources. Figure 3-1


for these applications.


shows different magnetron sources used


One of the aspects of the present study is the designing of a


magnetron glow discharge source for atomic emission spectroscopy.


Since


1980,


there


been


a growing


interest


analytically


useful



























Figure 3-1


: Magnetron glow discharge devices used in thin film technology.












Cathode


Permanent
Magnets


Pole Piece


(a) Planar Magnetron


Cathode


Anode


ExB drift


ExB drift


Cathode






Anode

Field Lines


(b) Cylindrical Magnetron






40

magnetic field.

In 1978, Grimm9 presented a magnetic field glow discharge source with a planar


cathode which was subsequently used by


Kruger


et. al.9 The source incorporated a


permanent magnet which caused the electrons in the plasma to execute a spiral motion

about the magnetic field lines leading to enhanced collisional excitation, sputtering and


light output as compared to a conventional GD source.


These authors further evaluated


the performance of this source for the analysis of mild and stainless steel and compared the

results to a source without magnetic field.91 The analytical performance of their boosted


was


respectable


analysis


steel,


however,


difficulties


were


encountered in the analysis of alloyed steels owing to their differing magnetic properties.


Ferreira et al.


discharge sourc


made line profile measurements of radiation from the magnetic field glow

e. They showed that Zeeman splitting of spectral lines is present in this


type of source. Sacks and coworkers carried out extensive investigations on the effect of


magnetic field on planar cathode glow discharge.


93-9


The sputtering characteristics of six


different alloys and metals were studied by these researchers under a wide range of fill gas


pressures.


The mass lost as a result of sputtering was highest for the highest atomic


number species.


They also studied the excitation and


ionization characteristics of the


planar magnetron glow discharge device, and it was found that ionization increases with


decrease in the discharge pressure."


other work,


preliminary


studies of


analytical


applications of a magnetron glow discharge was reported by these authors.1(" Recently,








41

to three times increase in the emission intensity for copper and chromium atomic emissions

from various alloys resulted.102

The influence of a magnetic field on a hollow cathode discharge was first identified

and partially studied from a physical point of view about 30 years ago.103-105 Schrenck et

al.06 investigated magnetic field effects on the emission of hollow cathode lamps using

cathodes made of either ferromagnetic (Fe), paramagnetic (Cr, Mg), or diamagnetic (B)

material. A visual increase in the emission intensity was observed for boron as cathode.


Magnesium,


being slightly paramagnetic at room temperature,


showed little effect.


substantial increase in the intensity was observed for Mn, Ni and Cu when using an iron

hollow cathode under a magnetic induction of 340 gauss. Later, Rudnevsky et al.10o-109

investigated spectrochemical aspects of the HCD in magnetic fields with transversal and

longitudinal orientations with respect to the cathode axis. Trivedi et al.n0 used an external

magnetic field to alter the radiative properties of a demountable hollow cathode lamp with

a center post cathode and of several commercial hollow cathode lamps. Another version

of a center post cylindrical magnetron glow discharge was devised by Tanaguay et al.11 min

which the radiative and electrical properties of the discharge were studied. A similar center


post cylindrical magnetron glow


discharge used as an ion bombardment furnace was


developed by Sacks et al. 2for the analysis of liquid solution residues,


picogram level detection limits.


which resulted in


The influence of a rotating magnetic field on a HCD was


investigated hv Pavlnvic and nnhmrcavlivwiv113


tha Pomitnirn ;ntn:i;tnao nit


fill naaan rrtnn







42

Theory of glow discharge influenced by magnetic field

Plasmas are moderately conductive and, therefore, are only weakly perturbed by

the application of electric fields. However, the application of a magnetic field can have

very significant effects. A moving charged particle in a magnetic field is subject to a force


=qi;


(3-1)


where q is the charge on the particle (C), v is the velocity (m/s), and B is the magnetic

field (gauss).


When


magnetic field


uniform


no electric


present,


charged


particles drift along the field lines with speed v1 which is unaffected by the magnetic field,

and the particles orbit along the field lines with a gyro or cyclotron frequency given by

following equation


eB
- = 1.76


'B rad/s


(3-2)


and the gyro or Larmor radius is expressed as


my
=-(-)
q'B


/2m(KE)


(3-3)


The motion of the particle in this situation is a helix as shown in Figure 3-2a.


Equations 3-4 and 3


give useful forms of the relationship


for the cyclotron


frequencies of electrons and singly charged ions, respectively:


= 2.8 GHz/kG


(3-4)








43

It is interesting to note that the cyclotron frequency of an electron is about 7.3 x 104 times

greater than that of a singly charged argon ion.

When magnetic and electric fields are uniform and the electric field is parallel to


magnetic


field,


particles


are freely


accelerated


helix


pitch


increases


continuously.


When the electric component is perpendicular to the magnetic field, a drift


of speed given by


p1
= 10' --
B


cm/s


(3-6)


develops in a direction perpendicular to both the electric and magnetic field vectors and


combines with the orbiting motion as shown in Figure 3-2b. This is the E


xB drift.


For a


particle created at rest in uniform and perpendicular


fields,


the trajectory


becomes the cycloid generated by a circle of radius given by equation 3-3.


Scooe of the Dissertation


With the previous success of the microcavity hollow cathode discharge for the

measurement of trace metals in sub-microliter samples in this laboratory, it was decided to


design a simple and inexpensive


boosted


glow


discharge


source for


atomic


emission


spectroscopy for these measurements and compare the results with the results obtained

with a conventional glow discharge source. If the volume of the discharge chamber is kept

as small as practically possible, then there is a substantial improvement in the emission






























Figure 3-2 : (a) Motion of a charged particle in magnetic field; (b) motion of a
charged particle in a mutually perpendicular magnetic and electric field.







45













B B





iU
SI

E







tExB

(a) (b)







46

enhancing the emission intensity output because of its simplicity of use and low cost.

Three designs of the magnetically-boosted glow discharge devices were constructed. In

chapter 4, a magnetically-coupled microcavity hollow cathode discharge is evaluated to

investigate the possibility of its application for the analysis of microliter volume solution


residues.


The magnetic field was applied using an electromagnet.


The effect of varying


magnetic field on the atomic and ionic emissions from the microcavity HCD are studied. A

theoretical model based on semicorona equilibrium is presented to explain the observed


effect of magnetic field on the emission intensities. In chapter

compact microcavity hollow cathode discharge are discussed.


results obtained with a


The design is based on a


cylindrical magnetron sputtering source. Small permanent magnets are used in this design

in order to keep the size of the device as small as possible. In this magnetron design, the

charge transport processes which are influenced by the presence of the magnetic field are


characterized on the basis of the atomic emission data.


In chapter 6, atomic emission


results obtained from a miniature planar magnetron glow discharge source are presented.

Spectroscopic results obtained from the sub-microliter volume residues of the aqueous

samples deposited on the cathode surface are compared with those obtained without a


magnetic


field.


Finally,


analytical


possibility


usmg


a planar


magnetron


glow


discharge


source


simultaneous


multi-elemental


trace


analysis


of less


microliter samples is explored.














CHAPTER 4
SPECTROSCOPIC EVALUATION OF A MAGNETICALLY-BOOSTED
MICROCAVITY HOLLOW CATHODE DISCHARGE

Design of the Discharge Device

In this part of the study, the microcavity design used previously in this laboratory

for the evaluation of its application in the measurements of microliter volume sample was


used.


A glow discharge chamber was made from a quartz outer tapered joint with an


optical quality window fused at one end of the joint. The outer tapered joint contained an


inlet and an outlet for the flow of the fill gas and vacuum connection (Figure 4-1).


inner joint with the electrode and the outer joint formed a demountable assembly which

held a tight vacuum necessary for a low pressure discharge. The total volume of the glow


discharge chamber was 70 mL. A stainless steel anode wire


2 mm in diameter was sealed


inside a 7/1


quartz inner tapered joint and


the anode was electrically grounded. Argon


gas was used as the fill gas and an on-line moisture trap removed moisture.


The flow of


the gas into the chamber was controlled using two needle valves connected in series.


gas pressure was monitored by two pressure gauges, one reading from 0 to


760 torr m


increments of 10 torr and another reading from 0 to 20 torr in increments of 1 torr. A

rotary vacuum pump (Varian Associates, Model SD-90, NJ) was used to pull the vacuum


and a ballast can was introduced between the vacuum pump and the chamber.


The ballast

































Figure 4-1


: Schematic diagram of the quartz glow discharge chamber.
























Argon In


Anode


Quartz
Window


To Vacuum


Circulating
Chilled Water











evacuating, the chamber was filled with argon gas to the desired pressure.


cathode


was


cooled


circulating


chilled


water through


a Tygon


wrapped around the neck of the chamber.

Microcavity hollow cathodes of 1.5 mm diameter and 6 mm depth were machined


from high purity (> 99.99%) metal rods 6 mm in diameter.


These dimensions of the


nmicrocavity hollow cathodes were previously found to be optimum for solution residue


analyses.14 For the magnesium studies, NIST standard reference material 124b was


used


which contained 4.54 % magnesium in aluminum alloy. The cathodes were inserted into a


ceramic sleeve and


epoxied into a quartz inner tapered joint (Figure 4-2).


The cathode


was connected to the negative terminal of a high voltage power supply (Hippotronics,


Model 803-330). The power supply used had the capability of delivering up to


3000 V and up to 400 mA. A ballast resistor of 29 kL was connected in series with the


cathode.


The use of the ballast resistor helped the formation of a stable discharge.


discharge could


initiated


instantaneously


by using


a laboratory-constructed


toggle


switch.

An electromagnet (Anac, Australia; Model 3470) was used for the application of

the magnetic field; the field strength could be varied by changing either the coil current of


the magnet or the gap between the poles of the magnet or both.


The microcavity HCD


was placed between the two poles of the magnet, and the field was applied perpendicular


f:aan- A i L -1.-.-*, i a a- j. ZA C l- --- A -_A --


~n ~n, nF cl\n m:nm na.:cl r

































Figure 4-2


: The microcavity hollow cathode.






























Ceramic Sleeve


Cathode


E ***:*


Microcavity
(1.5 mm in diameter
6 mm deep)


19/32 Quartz inner
taper joint



























Figure 4-3


: Schematic diagram of the experimental set up.












Magnet
Pole


To Vacuum


Magnet
Coil


Lens


If
I
I
CI
I
I -........ I


hole


Photo Diode
Array Detector


II !
FillGaJ
In .


Discharge
Power
Supply


Detector
Controller


Magnet


Computer











The measurements were made


by placing the


gaussmeter probe at


exactly the


same


position where the microcavity was to be positioned.

The emission from the microcavity was focused on the entrance slit of a 1 m focal


length grating spectrometer (Jobin-Yvon model HR1000; NJ


- 2400 grooves mmn1 linear


dispersion of 0


nm mm"') through a quartz piano-convex lens.


An iris aperture was


placed between the lens and the monochromator entrance slit to reduce the stray light


entering the spectrometer.


The monochromator slit width was 10 pm and the slit height


was maintained at 4 mm. A photodiode array detector (PDA) with a spectral range of 20


nm was used as the detector.


The detector contained 1024 pixels which gave a spectral


resolution of 0.02 nm per pixel. The diode array pixel height was 3 mm.


Therefore, each


observed spectral line intensity represented an average intensity over a thin vertical slice of

the discharge. One advantage of the PDA was that it could simultaneously monitor all the


emission within a 20 nm range.


This is particularly useful for the measurements of


trace


elements simultaneously by atomic emission spectroscopy.


The emission signal could be


integrated at increments of 33 ms over a desired period of time; also, it was possible to

accumulate a selected number of scans or both of these acquisition methods could be

applied simultaneously. Data were acquired with a personal computer and data acquisition


software (OSMA 1-120, Princeton Instruments, Inc


NJ, USA).


Radial profiles of the atomic emission from the microcavity were taken by moving
- 1 IA -^ Cte t .. .. *-a -. L- ". --- r ... .0 -











For solution analysis,


1000 ppm standard solutions (Fisher Scientific, NJ) were


used.


solution


was deposited


inside


the microcavity using


micropipet


Instrument, Co.,


CA, Model Pipetman P-2).


The solution was dried under a gentle stream


of nitrogen for a period of


minutes, and the solution residue was subjected to analysis.


The acquisition of data was begun, before switching on the discharge.

Solution Residue Analysis


Application of the magnetic field resulted in a very intense positive column and the


negative glow was seen to penetrate inside the microcavity.


The emission from inside of


the microcavity was brighter in the presence of the magnetic field. However, due to the

brighter positive column in the presence of magnetic field, a large background emission


from the positive column was present.


It was thus decided to reduce the size of the


positive column by decreasing the interelectrode distance, a common method adopted to

eliminate the positive column.11' In order to obtain the optimum signal to background

ratio of the analyte in the presence of the magnetic field as a function of interelectrode

distance, another quartz glow discharge chamber was designed. Figure 4-4 shows the new


design.


The distance between the anode and cathode


could be varied by inserting the


stainless steel anode deeper inside the chamber from the side arm.


electrode distance was found to be


The optimum inter-


2 mm without resulting in arcing of the discharge.


Figure 4-5 shows the temporally resolved emission spectra of a solution containing


a 4S


nI ~ 1. A. -


Ollainin
































Figure 4-4


: Modified design of the quartz glow discharge chamber.






















Argon in


Anode


Quartz
Window


Vacuum


--- ~4



























Figure 4-5


Temporally resolved emission spectra of a 100 ng copper solution residue.
































1/
t/.i
1/! I
I/II
I I
i/i
"'I I I
* *
III.'

-!
I
'iii'
I I
it,
I
/
I
I i/I I
A *
I If
I I
I ~ I
I,
I I I,!
I
*
ji I,'
*
I '!

I I~. -
I.. -


I.

* K! I
H
I

I 'I I I
'A
I
jj'i
I,
I I
i/Il
4/i i 37
I I?!
I jit''I I
F
I I
*
V
YH,
I I' I
* ~1~
I I

* *
I

.rL .
. .


. . .-.. -.. .-. 7 .... *. .- ... . -. .




I I A A A -I I. .1.A. A -
i i 1 i i

i i i I -


I

ii I--- *- - I- *- I- - -- - -
i i i II


*I I
i i i II i A -


. .. . ... .. . . ... ......... .. .. ... .. *. -.. .


r. I I I I
I. I I i
j . -.. ... . . i . i . . .-.. L .. . .- . -.. ... .
S I I I I I
r *



I I I I I
I A *

i .. .. ..... .. .. .. i ... ..... ........... T .. ..


I I AI t A -_
* $ ..i .. *. A.. A. . . -iA- A 1 .A .*. ... . . .. .. .
....t.. .. .......... a .......... .-..a.. ... .-. ._ -

L._ .-lA A... A.. : _.LAl k .-.A .-;*, -. .-$.


- A


323


324


325


326 327


Wavelength (nm)


16


14


12


10


80


60


40


2C


16
4a

MM








61


spectral window was employed in these studies, it was possible to simultaneously observe

the two resonance lines of copper at 324.75 and 327.40 nm. At the beginning of the


discharge (0


s in Fig. 4-5), no magnetic field was applied, and the line at 324.75 nm had a


good signal-to-background (S/B) ratio, whereas the line at 327.40 nm had a poor S/B. As

soon as the magnetic field was turned on (as shown by the arrow in Figure 4-5), the


intensity of the line at 324.7


increased by up to four times and the emission line at 327.40


nm began


appear


above


background.


Similar


results


were


reported


Rudnevskii et al."6 for a magnetically-boosted HCD of conventional


dimensions.


precision of measurements in the present work was, however, about 30 % for solution


analysis.


In order to optimize experimental conditions such as magnetic field strength,


current and pressure for solution residue analyses, it was decided to study the effect of

magnetic field on solid samples before further evaluation of solution sample residues.

The effect of magnetic field on the atomic emission of copper was studied by

applying a magnetic field parallel to the HCD axis where the discharge chamber was

placed at the center of the coil of the electromagnet. Up to 500 gauss of the magnetic field

on the discharge was obtained with this arrangement. However, there was no appreciable

effect of the magnetic field on the atomic emission was observed when the field was

applied parallel to the cathode axis.

Effect of Magnetic Field on Atomic Emission of Sputtered Cathodic Material

a afa a tO-- j.i. _-_-a a t -i *t- j. ..- -~. -A- : -_ -- t t.-^ -- .- 2-- *.-I..

































Figure 4-6


Effect of the magnetic field on copper emission.














35000



30000



25000


20000


S 15000


10000


5000



0


0 1000 2000 3000


4000


Magnetic Field (eauss'











with increasing magnetic field strength.


about


The emission intensity increased by a factor of


5 up to a magnetic field of 2500 gauss and further increase in the magnetic field


resulted in a decrease in the emission intensity.


by the magnetic field.


The background emission was unaffected


Precision (% relative standard deviation) for these measurements


was less than 2%. A possible mechanism reported by Pavlovic and Dobrosavljvic.n7
involved the negative glow inside the microcavity being pushed towards the inside walls of

the microcavity as the magnetic field is further increased. As a consequence, there would

be a reduction in the intensity at the center of the microcavity at higher field strengths.


However,


in the present study,


this possibility was ruled out by measuring the radial


profiles of the copper microcavity emission at 324.7


results are shown in Figure 4-7.


nm at different field strengths. The


The intensity of emission is a maximum at the center of


the microcavity at all magnetic field values, indicating that there is no localization of the

negative glow near the inside walls of the microcavity.

Aluminum is known to show a lower sputtering rate than copper, and therefore, is

less prone to self-reversal.69 Therefore, it was decided to examine the effect of magnetic


field on aluminum atomic emission.


Two resonance lines of aluminum atomic emission at


394.6 and 396.6 nm were chosen for this study.


Atomic emission fiom an aluminum


microcavity hollow cathode was observed under identical discharge conditions used for


the copper emission study; the results are shown in Figure 4-8.


Once again as with copper


* I .1 al.......1 .- --_-._-_---_------_-_

































Figure 4-7


: Radial profile of copper microcavity HCD at various magnetic field
strengths.
















10000


9000


8000


7000


6000


5000


4000


3000


2000


1000


0


-800


-600


-400 -200 0 200 400


Distance Across Diameter of the Microcavity (jim)

































Figure 4-8


Effect of the magnetic field on aluminum atomic emission.














12000


10000




8000


6000-


4000 -


2000


-A


Al (I) 394.15 nm
Al (I) 396.40 nm


Background


1000


2000


3000


4000


Magnetic Field (gauss)










Radial


profiles


aluminum


microcavity


HCD


once


again


showed


maximum emission intensity at approximately the center of the microcavity (Figure 4-9). A

similar study was conducted for magnesium ion lines by observing the emission from the


microcavity hollow cathode made from a NIST


sample of aluminum containing 4.54%


magnesium. The results for magnesium ion emission for the lines at 279.55 nm and 280.27

nm, shown in Figure 4-10, exhibit the same trends as for copper and aluminum emissions.


Popovici et al.61


studied the effect of magnetic field on the atomic emission of lead at


405.78 nm using different fill gases; the emission intensities did not change much after


reaching a maximum. However, no theoretical explanation was given.


Rudnevsky and


coworkers"6 have studied residues of the alkali metals deposited in a HCD in a magnetic


field;


observed


similar


pattern.


reduction


molecular


emission


background of CN was observed with increasing magnetic field strength. However, in our

case, the background was fairly constant near the emission lines of all the three elements.

Hollow cathode discharges are known to show self-reversal, and the decrease in

the emission intensity for the analytes studied here may be due to an increased sputtering

rate with increasing magnetic field. However, upon studying the effect of magnetic field

strength on argon ion emission at 404.29 nm (see Figure 4-11), a steady decrease in the

emission intensity with magnetic field strength was observed. This observation implies that


there is no increase in the ionization of the argon gas with increasing magnetic field.


nther wnrds it in nnqsihle that there is no increase in the snutterinm rates with increasing

































Figure 4-9


Radial profile of aluminum microcavity HCD.














14000



12000


10000



8000



6000



4000



2000


-600 -400 -200 0 200 400
flotanrP frnir (PfntAr rAfMlfrnnvit7 (umm1


600

































Figure 4-10


Effect of the magnetic field on magnesium ion emission.














18000


16000 -


14000-


12000 -


10000 -


8000-
-

6000-


4000-


2000-


Mg (II) 279.55 nm
Mg (II) 280.27 nm


1000


2000


3000


4000


Magnetic Field (gauss)

































Figure 4-11


Effect of the magnetic field on argon ion emission.















2000





1900





1800


1700


1600


1500


1000 2000 3000


4000


Maonetic Field (anul&<









Theoretical Validation of the Effect of the Magnetic Field on the Emission of the
Sputtered Material


Sen et al.n8


observed the variation of the spectral emission intensities of helium,


hydrogen and mercury as fill gases in glow discharges as a function of magnetic field

strength, and they reported similar variations in emission with increasing magnetic field for


the atomic lines of the fill gases used.


The emission maxima were attributed to two


discharge related factors that are simultaneously influenced by the applied magnetic field;

these factors are electron temperature and the electron number density along the axis of

the discharge." It was assumed in their studies that the excitation mechanism was mostly

by electron impact.119


Electron


temperature


increases


with


magnetic


field,


whereas


there


a radial


diffusion of electrons from the axis of the discharge with


increasing magnetic field.


order to confirm that a similar argument can be used to justify the trend observed in the

present case for the sputtered materials, the following mechanism is speculated:


The electron


temperature in a magnetic


field,


is related


to the electron


temperature without a magnetic field, T, as follows,118


B2


(4-1)


where B


is the magnetic field strength (gauss) and /, the pressure (torr).


is a constant


(torrgauss"2) dependent on the fill gas which is given by equation 4-2.


(eL)


(4-2)









The electron number density, nte


(cm"3) along the axis of the microcavity HCD in


the presence of the magnetic field is given as"


= n exp(-


eEC/2rB\
2kT, J
e


(4-3)


where ne is the electron density (cm'


) in the absence of magnetic field and


is the axial


electric field per centimeter (V/cm). TI

electron number density at a distance r


his equation implies that there is a reduction in the

from the axis due to the presence of the magnetic


field.


The intensity of the emission arising from a transition


semi-corona plasma, in the absence of a magnetic field, is given by12


hly ..


where h is Planck'


-> i, for an optically thin


SA) (exp-)
expkTj
win


(4-4)


constant (J.s), I is the path-length (cm), no is the number density of the


analyte atoms or ions in the ground state (cmn3), Aft


is the Einstein coefficient of transition


probability (s"'),


2.
m

Am
jm


is the summation of all the transition probabilities between


upper and lower levels, Xg

state g to the excited state


is the collisional excitation


rate coefficient from the ground


.1), U1 is the energy of the upper level (eV),


and v


the frequency of the emitted radiation (s1).

In the presence of a magnetic field, I, is given by'e


... 'a* a -


kfv






78

Taking the ratio of equations 4-4 and 4-5 and substituting for Ta from equation 4-1 and

maximizing the result with respect to B gives120


e2E2 r2


= 4CB UJ {,21+CI2 B ,2 c-1
^l~ax0)^ \l~l~Is


There is some degree of uncertainty involved in the theoretical calculation of the value for


using equation 4-6 since the values of


E and r


are not precisely known. Therefore,


for a particular gas at a constant pressure, it will suffice to prove that for a given set of

emission lines for an element,


BmaxUj ( + QCBm


/P2)-V


=c2


(4-7)


where C2 is a constant. In order to calculate C, from equation 4-2, one has to know the

collisional cross-section of electrons and fill gas atoms to calculate the mean free path.

Since the collision cross-section depends on the energy of the electron and that of the gas

atoms, its precise determination is difficult.

Sadhya et al.'2" proposed a method for calculating the value of C from equation 4-

1 by plotting the electron temperature in a magnetic field at various field strengths, the


slope of which gives CI.


Furthermore, the electron temperature T,


field was calculated from the intensity ratio of

thermodynamic equilibrium (LTE). Since a l


without the magnetic


two lines assuming the gas


ow pressure discharge is


was in local


not in LTE,


semicoronal equilibrium was assumed to prevail inside the discharge."9


Assuming another transition


i and taking the ratio of the intensity of the two


(4-6)










=(Uk


- U )/In


'"2l3i


(4-8)


assuming the excitation rate coefficients ,s and Xgk are the same due to the similarity of

excited states.

When the magnetic field is applied,


- U,)


)B(j')1
WB-1k}


(4-9)


where fy and fk are the oscillator strengths dimensionlesss) of the respective transitions

and TeB is the electron temperature in the presence of the magnetic field. It is important to

note that this equation is valid only for those emission lines which have the same ground

states because the statistical weights of the ground state are also affected by the applied


magnetic field as a result of Zeeman splitting.


The ratio of the statistical weights in the


presence of the magnetic field is unity if the two ground levels are same.


Two resonance


lines of copper (324.70 and 327.40 nm) were selected for the calculation of


electron


temperatures both with and without a magnetic field using equations 4-8 and 4-9. A plot


of (Te /T/,


versus


(where BX' is known as the reduced magnetic field) is shown


in Figure 4-12. It is evident from this plot that the electron temperature is constantly


increasing with applied magnetic field strength (where pressure is kept constant).


value of the slope of this plot, which


corresponds to C.,


gives 1.72 x 106 for argon which


is in the same range calculated for helium by Sadhya et. al.'20 Table 4-1 lists the values of

































Figure 4-12


: Variation of electron temperature with the magnetic field.































































0 2 4 6 8 10


2,.. 9 .9








Table 4-1


: Relationship between Intensity Maxima and Energy of Upper Level for


Various Analytes


Analyte Wavelength Energy of the Bm C2
(nm) upper level (gauss) (gauss.eV)
~_____________~(eV)_____________
Cu(I) 324.75 3.817 2500 2784

Cu(I) 327.40 3.786 2500 2761



Mg(II) 279.55 4.434 2400 3222

Mg (I) 280.27 4.422 2500 3225



A1 (I) 394.40 3.143 2300 2275

A1(I) 396.15 3.143 2300 2275


Bma values correspond to magnetic field strengths giving the maximum signals on Figures 4-6, 4-
8 and 4-10.








83

lines of a particular element, the closer are the maxima of the emission intensities that


occur when the field strength is increased as seen in Figures 4-6,


three elements chosen in this study, B,


4-8 and 4-10. For all


values are nearly the same for a given element.


For example, aluminum has the same energy of the upper levels for both lines at 394.40

and 396.15 nm and shows maxima for these lines at 2300 gauss.

Conclusions


It can be concluded from this study of the magnetic field effects on the emission


intensity


from


microcavity


HCD


two-field


dependent


factors,


electron


temperature and radial diffusion of the electron are both acted upon simultaneously and


opposite to each other by the applied magnetic field.


The electron temperature increases


with magnetic field, whereas there is an increase in radial diffusion of the electron with

increasing magnetic field. Therefore, a maximum in emission intensity does occur at a


certain field strength and this maximum is dependent upon the energy of the upper state of

the analytical line. Since there is no substantial advantage in using a very high magnetic


strength, it might be appropriate to use permanent magnets which tend to have


sufficient


to produce


positive


effect


magnetic


on atomic


emission


intensity."9


Use of a permanent magnet will facilitate a compact discharge design, which is


one of the goals of the present research. In the next chapter, a compact magnetically-

boosted glow discharge device is designed which used permanent magnets as a source of

maonattic field














CHAPTER


CYLINDRICAL MAGNETRON GLOW DISCHARGE PLASMA

Design of a Cylindrical Magnetron Source


With the preliminary


success of the enhancement of atomic emission intensities by


using an electromagnet, the second phase of the present research was to design a compact


magnetically-boosted microcavity


hollow


cathode discharge for


AES


evaluate its


spectroscopic performance by observing the effect of the magnetic field on the atomic


emission of the sputtered cathodic material.


Magnesium in aluminum alloy was chosen as


the model analyte for these studies. If there was substantial enhancement in the intensity

output of atomic emission in the presence of the magnetic field, then the application of this

design could be extended to the analysis of microliter volume aqueous samples. Permanent

magnets are ideal for this application since they usually have sufficient magnetic field to

affect the discharge.121 Therefore, neodymium-iron-boron permanent magnets were used

to apply a magnetic field which had a magnetic field of about 500 gauss on their surface.


Each magnet had a diameter of 6.


mm and thickness of 6.


rn,


The magnets were


embedded inside the cathode body surrounding the microcavity hollow as shown in Figure


This configuration is known as a cylindrical magnetron and it is widely employed in


thin film processing industries for rapid deposition of a film, as briefly discussed in chapter


























Figure 5-1


Schematic diagram of the microcavity cylindrical magnetron.



















Embeded
Magnets


Microcavity
Hollow Cathode


Side View


Cross-sect






87

on the magnets. A microcavity hollow cathode made from aluminum alloy (NIST # 5183)


containing 4.54


magnesium


was employed


used


to observe the effect


of the


magnetic field on the atomic emission of magnesium.


The magnets were placed opposite


to each


other


with


opposite


polarity


which


resulted


in a magnetic


passing


perpendicularly to the axis of the microcavity HCD.


attempted.


Various other configurations were


These included, (i) arranging the magnets along the axis of the microcavity


hollow cathode so as to have the magnetic field parallel to the axis, (ii) arranging four

pairs of the magnets such that two pairs give the magnetic field parallel and other two

pairs give the magnetic field perpendicular to the cathode axis, etc. However, the only


configuration, as shown in Figure


which resulted in observable enhancement of the


emission intensity, is discussed in the following paragraphs.

The same quartz glow discharge chamber and the experimental set up that was


previously


used


to evaluate


magnetically-boosted


microcavity


hollow


cathode


discharge was used here. A discharge current of 28 mA was applied.

Application of the magnetic field using the permanent magnets resulted in a bright


discharge within the microcavity compared to the one without the magnetic field.


There


was,


however, no observable effect of magnetic


on the positive column


of the


discharge, as opposed to the situation when the magnetic field was applied using the

electromagnet, discussed in the preceding chapter. This was due to the localized nature of

the magnetic field of the permanent magnets which were relatively closer to the discharge










Figure


shows the pressure dependence of magnesium emission at 285


from


magnetically-boosted


microcavity.


emission


intensity


magnesium in the absence of the magnetic field are shown by the dotted line. At higher

pressures (> 6 torr), there was no appreciable enhancement of atomic emission caused by


the presence of the magnetic field.

and down to 1 torr, there was a 2


When the pressure was gradually reduced below 6 torr


to 3 times enhancement of atomic emission intensity by


the presence of magnetic field. Below


torr of argon pressure,


the emission intensity


decreased with decreasing pressure in the presence of the magnetic field.


The emission


intensity without the magnetic field did not change as much as that in the presence of the

magnetic field when the pressure was decreased.

Charge Transport Processes in the Microcavity HCD


The observed dependence of the atomic emission intensity on


pressure in the


presence of the magnetic field


can be explained


considering the charge transport


processes inside the MCHC.


In a hollow


cathode,


the radial


charge transport is primarily by mobility


diffusion. The radial electron flux Fr, is given bym


AceErn


e


(5-1)


where p, and D, are electron mobility and electron diffusion coefficient, respectively; E, is

the radial component of the electric field, and e is the radial electron number density.

































Figure


Dependence of atomic emission of magnesium on the discharge pressure.