Nonmetal selective liquid chromatographic detection utilizing ultrasonic nebulization and membrane desolvation with a helium microwave -induced plasma

INFORMATION TO USERS
This manuscript has been reproduced from the microfilm master. UMI films
the text directly from the original or copy submitted. Thus, some thesis and
dissertation copies are in typewriter face, while others may be from any type of
computer printer.
The quality of this reproduction is dependent upon the quality of the
copy submitted. Broken or indistinct print, colored or poor quality illustrations
and photographs, print bleedthrough, substandard margins, and improper
alignment can adversely affect reproduction.
In the unlikely event that the author did not send UMI a complete manuscript
and there are missing pages, these will be noted. Also, if unauthorized
copyright material had to be removed, a note will indicate the deletion.
Oversize materials (e.g., maps, drawings, charts) are reproduced by
sectioning the original, beginning at the upper left-hand comer and continuing
from left to right in equal sections with small overlaps.
Photographs included in the original manuscript have been reproduced
xerographically in this copy.
Higher quality 6” x 9” black and white
photographic prints are available for any photographs or illustrations appearing
in this copy for an additional charge. Contact UMI directly to order.
ProQuest Information and Learning
300 North Zeeb Road. Ann Arbor, Ml 48106-1346 USA
800-521-0600
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
ABSTRACT
Name: Debashis Das
Department: Chemistry and
Biochemistry
Title: Nonmetal Selective Liquid Chromatographic Detection Utilizing
Ultrasonic Nebulization and Membrane Desolvation with a Helium
Microwave-Induced Plasma
Major: Chemistry
Degree: Doctor of Philosophy
Approved by:
Date:
cu
7 A^/oi
Dissertation Director
NORTHERN ILLINOIS UNIVERSITY
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
ABSTRACT
Atomic emission spectroscopy (AES) detection for chromatographic
separations offers excellent selectivity and sensitivity and provides elemental
information. The helium-microwave induced plasma atomic emission spectroscopy
(He-MIP AES), suitable for detection o f nonmetals (Cl, Br, I, S, P), has been
successfully employed in conjunction with gas chromatography. However, with liquid
chromatography (HPLC), the performance o f MIPs is severely affected due to solvent
loading.
Membrane desolvators provide high analyte transport and desolvation
efficiencies with most solvents. This dissertation focuses on application o f the
ultrasonic nebulizer (USN)-membrane desoIvator-He MIP system as a detector for
reverse-phase HPLC.
Transport o f the volatile organic analytes was maximized by temperature
optimization of the membrane desolvator. Temperature in the 50°C range provided
efficient desolvation while minimizing analyte loss due to transport through the
membrane. Modifying the solvent with dilute H2S04served to maximize analyte
transport by reducing the volatility o f the solvent The inability o f USN with a fixed
frequency (1.4 MHz) automatic transducer to nebulize certain HPLC solvents (aqueous
solutions o f 10 -70% methanohwater and acetonitrile:water) was remedied by
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
frequency optimization using a manual transducer controller.
Separation and detection of a mixture o f 2,6 dichlorobenzamide (DCB) and 4
chlorobenzamide (CB) using a 60% aqueous solution o f methanol was achieved by the
HPLC-MIP AES detector. The detection limits were 36 and 47 ng/s for DCB and CB,
respectively. The selectivity of the HPLC-MIP AES system was demonstrated with the
detection o f vitamin B,2 in a mixture o f vitamin B„ B6, and Bl2. Superior sensitivity of
the plasma AES detection over UV detection systems was exhibited via separation and
detection o f a mixture o f P-glycerophosphate and triphenylphosphene. UV inactive Pglycerophosphate could be detected using the P emission line at 253.7 nm by MIP
AES. Detection limits were 3.3 ng/s. Usually difficult by plasma spectrometry,
detection o f high molecular weight compounds could be achieved. Greater than 10 kb,
DNA samples from salmon testes were successfully detected.
MIP AES with membrane desolvation is a viable detection technique for
reverse-phase HPLC. However, studies to further improve the detection limits are
necessary.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
NORTHERN ILLINOIS UNIVERSITY
NONMETAL SELECTIVE LIQUID CHROMATOGRAPHIC DETECTION
UTILIZING ULTRASONIC NEBULIZATION AND MEMBRANE
DESOLVATION WITH A HELIUM MICROWAVE-INDUCED PLASMA
A DISSERTATION SUBMITTED TO THE GRADUATE SCHOOL
IN PARTIAL FULFILLMENT OF THE REQUIREMENTS
FOR THE DEGREE
DOCTOR OF PHILOSOPHY
DEPARTMENT OF CHEMISTRY AND BIOCHEMISTRY
BY
DEBASHIS DAS
DEKALB, ILLINOIS
AUGUST 2001
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
UMI Number 3023682
UMI*
UMI Microform 3023682
Copyright 2001 by Bell & Howell Information and Learning Company.
All rights reserved. This microform edition is protected against
unauthorized copying under Title 17, United States Code.
Beil & Howell Information and Learning Company
300 North Zeeb Road
P.O. Box 1346
Ann Arbor, Ml 48106-1346
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Certification:
In accordance with departmental and Graduate School policies,
this dissertation is accepted in partial fulfillment o f degree
requirements.
rtation Director
H o tg
ANY U9E Of MATWM CONTAIN®
ua m UUST K DULY ACKNOWliDGtD.
rw AUTHOR'S P&M1SSJON
K^TAtNEC
W ANY PORTION JS TO II
°*
INCIUO® IN A PUBLICATION
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
ACKNOWLEDGMENTS
I would like to thank my research advisor, Dr. Jon W. Carnahan, for his
valuable guidance and support through the years, both as a research advisor and as a
personal friend. I am deeply indebted to him for the knowledge I have acquired under
his tutelage. My gratitude to all the faculty and staff members in the Department of
Chemistry and Biochemistry, Northern Illinois University, for their immense help and
support. A special thanks to my dissertation committee members and some of the
technical staff members: Dr. David Ballantine, Dr. Lee Sunderlin, Dr. Narayan
Hosmane, Dr. David Changnon, Dan Edwards, Larry Gregersen, and Charlie Caldwell,
without whose assistance the research could not have been carried out.
Many a thanks to my friends, Pamela Keating and Gary White, for their
relentless support through thick and thin, for which I will always cherish their
friendship.
Finally, my deepest gratitude to my mother, Ms. Ranjita Das; father, Dr.
Hrusikesh Das Berma; brother, Debjit Das; and all my relatives for believing in me and
for their endless love and affection.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
DEDICATION
To Maa, Daddy and Pintu, with gratitude
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
TABLE OF CONTENTS
LIST OF TABLES.................................................................................................
Page
vii
LIST OF FIGURES...............................................................................................
ix
Chapter
1. BACKGROUND.......................................................................................
1
1.1
Introduction....................................................................................
I
1.2
Advantages o f AES Detection for HPLC Systems .........................
4
1.3
ICP, DCP, and He-MIP as Excitation Sources ...............................
S
1.3.1
1.4
1.5
1.6
He - MIP Nonmetal Excitation Phenomenon ....................
7
Liquid Sample Introduction D evices...............................................
11
1.4.1
Pneumatic Nebulization....................................................
12
1.4.2
Ultrasonic Nebulizer ........................................................
13
Desolvation Devices .......................................................................
15
1.5.1
Cryogenic Desolvation......................................................
18
1.5.2
Hydraulic High-Pressure Nebulizer Desolvation...............
18
1.5.3 Membrane Desolvation System s........................................
20
Research Objective .........................................................................
23
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
vi
Chapter
Page
2. HELIUM MICROWAVE INDUCED PLASMA-USN-MEMBRANE
DESOLVATOR ATOMIC EMISSION SYSTEM
CHARACTERIZATION...........................................................................
25
2.1
Introduction....................................................................................
25
2.2
Fate o f the Analyte as It Passes Through the System.......................
27
2.3
System Specifications ....................................................................
29
2.3.1
He-MIP Plasma System ......................................................
29
2.3.2
Sample Introduction System .............................................
31
2.3.3
Aerosol Desolvation S ystem .............................................
32
2.3.4 O ptics.................................................................................
32
2.3.5 Data A nalysis.....................................................................
33
2.3.6 Reagents.............................................................................
33
2.4
Experimental ..................................................................................
34
2.5
System Optim ization......................................................................
37
2.5.1
Inorganic Analytes in Water .............................................
37
2.5.2
Inorganic Analyte in Methanol .........................................
43
2.5.3
Characterization o f the System with Organic A nalytes
44
2.6
Examination of Background Emission with Membrane
Desolvation ...................................................................................
53
2.7
Examination o f a 500 W Plasma for D etection...............................
62
2.8
Summary ........................................................................................
68
3. INTERFACING HPLC WITH THE MIP AES USING USN AND
MEMBRANE DESOLVATION................................................................
69
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
vii
Chapter
page
3.1
Introduction....................................................................................
69
3.2
Experimental ..................................................................................
71
3.3
Separation Using HPLC-MIP AES .................................................
73
3.3.1
Effects of Anlyte on Background.......................................
76
3.3.2 Investigation of Analyte Loss in the S ystem ......................
78
3.4
Nebulization o f LC Solvents with USN...........................................
87
3.5
Assessment ofFixed-Frequency Transducer B ehavior....................
94
3.6
HPLC-MIP AES Using Solvent M odifiers......................................
100
3.7
Summary .........................................................................................
104
4. HPLC-MIP AES APPLICATIONS............................................................
106
4.1
Introduction.....................................................................................
106
4.2
Examination o f Br, I, S, and P Em ission..........................................
107
4.3
Sensitivity o f MIP-AES Detectors....................................................
114
4.4
HPLC-MIP AES Applications..........................................................
119
4.4.1
Detection of DNA .............................................................
119
4.4.2
Detection of Vitamin Bi2 ...................................................
124
4.4.3
Detection of Nucleotides...................................................
132
Summary .........................................................................................
136
5. CONCLUSIONS AND FUTURE DIRECTIONS......................................
137
REFERENCES .....................................................................................................
140
4.5
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
LIST OF TABLES
Table
page
1. Unoptimized Conditions Used for Preliminary Experiments ...............
35
2. Optimum Operating Conditions for Inorganic and Organic
Analytes.....................
52
3. USN Nebulization Efficiency o f Binary Solvent Mixtures (v/v)
With Automatic and Manual Transducer Controllers..........................
90
4. Nebulization Frequency o f Water:methanol and Water:acetonitrile
Mixtures Using USN with Manual Transducer....................................
92
5. Nonmetal Detection Limits Using MIP AES.........................................
113
4
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
LIST OF FIGURES
Figure
page
1. Energy diagram for nine nonmetals. The horizontal lines represent
ionization energies o f Ar and He. The dots indicate the energies of
the ion states of the respective nonmetals. The energies are calculated
as the energy o f ionization plus the energy o f ion excitation................
10
2. Comparison o f primary pneumatic (a) and primary ultrasonic
(b) volume distributions for water. .....................................................
16
3. Ultrasonic nebulizer-membrane desolvator...........................................
17
4. (a) Cryogenic desolvation device used with the ultrasonic nebulizer and
(b) desolvation set-up used witth the hydraulic high-pressure
nebulizer...............................................................................................
19
5. Schematic for the MIP-USN-membrane desolvator atomic emission
system
...........................................................................................
26
6. Fate o f the liquid droplet as it passes through the USN-MIP AES
system...................................................................................................
28
7. CETAC MDX-100 tubular membrane desolvator.................................
30
8. (a) Spectrum obtained from aqueous solution of 500 ppm Cl as MgCL
without (b) and with membrane desolvator. ........................................
36
9. Cl signal as a function o f plasma gas flow rate.......................................
38
10. Cl signal as a function o f USN gas flow. .............................................
40
11. Cl signal as a function o f membrane desolvator countercurrent gas
flow ra te s .............................................................................................
41
12. Cl calibration plot under optimized conditions........................................
42
13. Cl signal as a function o f membrane desolvator countercurrent gas
flow ......................................................................................................
45
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
X
Figure
page
14. Inorganic and organic Cl signal vs membrane desolvator
temperature .........................................................................................
47
15. Organic Cl signal vs membrane desolvator tem perature.....................
48
16. Cl signal vs membrane desolvator countercurrent gas flow ....................
SO
17. Cl S/N vs membrane desolvator countercurrent gas flo w ........................
SI
18. Five hundred ppm Cl signal from 2,6 dichlorobenzamide in methanol
without membrane desolvator...................................................................
54
19. Five hundred ppm Cl spectrum from 2,6 dichlorobenzamide in
methanol....................................................................................................
55
20. (a) Image of the original plasma (nothing is being nebulized),
(b) image o f the plasma when methanol is being nebulized with
membrane desolvator and (c) image of the plasma when methanol is
being nebulized without the membrane desolvator. A 500 W
generator was used for this study..............................................................
56
21. Methanol background with and without membrane desolvator
58
22. Plasma background spectrum at different membrane desolvator
temperaturess while methanol is being nebulized......................................
60
23. Plasma gas vs Cl S/B..................................................................................
63
24. USN gas flow vs Cl S/B.............................................................................
65
25. Membrane desolvator countercurrent gas flow vs Cl signal. 500 ppm
Cl as DCB in methanol used. Plasma operated at 400 W. USN gas
flow was set at 1.8 L/m in..........................................................................
66
26. Plasma power vs Cl S/B.............................................................................
67
27. System schematic for the HPLC-MIP A E S ..............................................
72
28. Chromatogram o f DCB..............................................................................
74
29. DCB calibration plot using HPLC-MIP AES.............................................
75
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
xi
Figure
30. CB calibration plot using HPLC-MIP AES..........................................
page
77
31. (a) DCB chromatogram when monochromator is set for 479.S Cl
emission line and (b) DCB chromatogram when the monochromator is
set at 479.6 nra, off the Cl emission line................................................
79
32. Cl calibration plot using DCB with direct solution nebulization
80
33. Cl calibration plot using DCB at reduced USN “U” tube temperature .
82
34. Cl calibration plot from 2,6 dichlorobenzamide dissolved in
methanolic 1% H2S04, 1% CH3COOH, and 1% HN03..........................
84
35. Cl S/B from 500 ppm of DCB in 0.01 and 0.1% H2S04 in methanol
solution..................................................................................................
86
36. (a) Chromatogram of DCB and CB using 100% methanol as mobile
phase and (b) separation of DCB and CB using 70% methanol as
mobile phase. Flow rate was 0.8 m L/m in............................................
88
37. Plot o f percent methanol in the solvent versus Cl S/B from solutions
o f500 ppm Cl as DCB solution.............................................................
93
38. Plot of percent methanol versus density with respect to water. All
density values were taken from reference 133......................................
95
39. Plot o f percent methanol versus corresponding surface tensions
97
40. Plot o f percent methanol versus corresponding viscosity relative to
w ater....................................................................................................
98
41. Plot o f percent IPA versus corresponding viscosity with respect to
w ater.......................................................................
99
42. System schematic o f the HPLC-MIP AES showing post column
addition o f dilute H2S04 .......................................................................
101
43. Chromatogram showing Cl peaks from 30 |ig
2 ,6-dichlorobenzamide and 30 pg 4-chIorobenzamide ......................
102
44. Calibration plots from 2 ,6-dichlorobenzamide and
4-chlorobenzamide................................................................................
103
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
xii
Figure
45. One ppt Br, MIP AES spectrum from aqueous KBr03 .......................
page
108
46. One ppt I, MIP AES spectrum from aqueous KI03 .............................
109
47. One ppt S, MIP AES spectrum from aqueous (NH4)2S04 ...................
110
48. One ppt P, MIP AES spectrum (213-216 nm) from aqueous H3P 04 . . .
I ll
49. One ppt P, MIP AES spectrum (252-257 nm) from aqueousH3P 04 . . .
112
50. LC-MIP AES chromatogram o f P-glycerophosphate (500 ppm P) and
triphenyl phosphene (500 ppm P)..........................................................
115
51. LC-MIP AES calibration plots for GP and TPP....................................
117
52. Chromatogram of GP (500 ppm P) and TPP (500 ppm P) using the
UV-Vis detector. .................................................................................
118
53. Structure of tetranucleotide...................................................................
120
54. Diagram of a double-stranded DNA.....................................................
121
55. LC-MIP AES chromatogram obtained from 500 ppm DNA, extracted
from salmon testes.................................................................................
123
56. UV picture o f electrophoretic bands from the molecular weight
strands (in slots 2,3, and 12) and the DNA sample from salmon testes
(in slots 5 ,7 ,9 ,1 0 ) on the agarose gel...................................................
125
57. Molecular structure o f (a) vitamin B„ (b) vitamin B6 and
(c) vitamin Bl2. ...................................................................................
127
58. Chromatograms obtained from vitamin Bl2 (10 ppm P) using
(a) UV-Vis detector and (b) MIP AES..................................................
128
59. Chromatogram from a mixture o f vitamin Bt (10 ppm), vitamin B6
(10 ppm), and vitamin B,2 (10 ppm P) using (a) UV-Vis detector and
(b) MIP AES..........................................................................................
130
60. UV-Vis spectrum o f vitamins (250-400 nm )B l,B 6, and B 1 2 ............
131
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
xiii
Figure
61. Molecular structures o f (a) cytidine S' monophosphate disodium salt,
(b) uridine S' monophosphate in free acid form, and (c) guanosine S'
monophosphate disodium s a lt.............................................................
62. Chromatogram o f a mixture o f CMP, UMP and GMP using the
UV-Vis 4etector at 254 nm....................................................................
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
page
134
135
CHAPTER 1
BACKGROUND
1.1 Introduction
In recent years plasmas have been extensively used as ionization and excitation
sources for atomic emission (AES), fluorescence (AFS), and mass spectrometric (MS)
systems. Among the different types of plasmas, inductively coupled plasmas (ICP),
microwave-induced plasmas (MIP), and direct-current plasmas (DCP) are most
common. Some of the important properties which make these sources desirable are
high temperatures (2000 -10000 K), few interferences, the capability for simultaneous
multielement analysis, high sensitivities, low detection limits, and large linear
calibration ranges.1'5 These traits make plasmas ideal sources for trace metal analyses
for solid, liquid, and gaseous samples. While maintaining many o f the aforementioned
properties, helium MIPs have the unique property of producing intense nonmetal line
emissions.**7 Hence, the source warrants investigation for determinations o f organic
analytes. MIP atomic emission spectroscopy (AES) as a detector for various
chromatographic techniques has been investigated by many authors.8*10 Helium
MIP-AES is ideally suited for interfacing with gas chromatography (GC)11'17because
helium is used both as the MIP support gas and the GC mobile phase.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
However, due to the inherent poor solvent tolerance o f plasmas, direct organic
solution nebulization often causes complications such as carbon deposition on the
plasma torch, significant spectral interferences, reduction of analytical signals, or even
plasma extinguishment.1*'27 To avoid these solvent-related problems, the analyte is
often desolvated before being introduced to the plasma. O f the many available
desolvation techniques, membrane desolvation has shown great promise.2*'34
Although He MIP-AES systems possess desirable characteristics for the
detection o f nonvolatile analytes, high-performance liquid chromatography (HPLC)
interfaces have seen limited success. This approach is hindered by two fundamental
drawbacks. First, typical LC mobile-phase solvents (methanol, acetonitrile, THF, etc.)
are not compatible with optimal plasma performance. Additionally, liquid mobile
phases do not lend themselves to efficient analyte transport. The problem due to
solvent introduction can be minimized with the use of various desolvation devices.
These devices remove a significant portion o f the solvent from the analyte stream
before introduction into the plasma.
Mason et al.3Ssuccessfully used a moving-band sample introduction system to
interface an LC with the MIP. Operation o f this interface occurs in five steps. In the
first step the eluent from the HPLC is deposited on a moving polyimide band (3.2 mm
wide and 0.0S mm thick moving at a speed o f 2.5 cm/s), hi the second step the
solvent evaporates as the band passes under an infrared heater. Third, the effluent
passes into a vacuum region where residual solvent vapors are removed. The dry
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
3
analyte is then is earned to the vaporization chamber in the fourth step. In this step,
flash vaporization of the analyte is achieved by a nichrome heater. The vaporized
analyte is directed to the plasma for detection. The fifth step involves cleaning o f the
band for the next run as it passes under the clean-up heater, where the residual analyte
is removed. While solvent removal is efficient, the system is difficult to operate and
maintain. Excellent detection limits could be achieved with this interface ( between 31
and 97 ng for Cl). However, the analytical reproducibility (relative standard deviation
[RSD] achieved -12%) is somewhat poor.
Preliminary successful studies with a flat-sheet membrane desolvator interface
have been carried out by Akinbo and Carnahan.34,36 About 99.9% o f the solvent could
be removed using these devices. The Cl detection limits achieved with the LC-MIPflat-sheet membrane desolvator were between 2 and 7 Jig, about one order of
magnitude higher than that obtained with the moving-band interface. On the other
hand, the reproducibility obtained was much better (RSDs o f 2.5 - 6.7%). Analyte
transport problems caused calibration plots to exhibit non-zero intercepts.36
In this dissertation, an extensive study o f the development o f detection systems
based upon the combination of HPLC, ultrasonic nebulization (USN), membrane
desolvation, and a He MIP will be discussed. Detection limits, linearity o f calibration
plots, and precision will also be presented. Particular effort is directed toward
enhancing analyte transport. Various applications to demonstrate the analytical
performance o f this detector are illustrated.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
4
1.2 Advantages of AES Detection for HPLC Systems
The commonly used detectors for HPLC systems are based upon UV
spectrometry and refractive index measurements. These detectors utilize the bulk
physical and chemical properties of the analyte molecule as a basis for detection. In
general, detectors o f this type lack exceptional selectivity or sensitivity.
Conversely, AES detectors are inherently selective because, when excited, all
elements emit radiation at their characteristic wavelengths. Also, the emission features
are very narrow ( line widths are typically in the range o f 0.01 nm or less) compared to
UV absorption bands. Hence, if the chromatographic column does not separate two
or more analytes o f interest, the detector may be able to distinguish and quantify them
by monitoring emissions from unique atoms that constitute the molecules. Another
AES advantage for HPLC detection is the ability to provide elemental content
information. Under ideal conditions, part per trillion detection limits can be achieved
by AES. Plasma AES systems are sensitive to almost all types o f compounds present
in complicated matrices. With all these advantages, the suitability o f a helium
microwave-induced plasma atomic emission system for reverse-phase HPLC detection
is worth exploring.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
5
1JICP, DCP, and He-MIP as Excitation Sources
Direct-current plasma, inductively coupled plasma, and microwave-induced
plasma are the commonly used plasma sources for atomic emission spectroscopy. The
typical DCP is produced by a DC discharge between a three-electrode array. The
plasma is sustained by a flow o f argon over the electrodes. Ar is also used for sample
nebulization. Typical combined gas flow rates are about 8 L/min.37 DCPs are
operated at power levels between 500 - 700 W3*with a voltage of 10 to 50 V. The
excitation temperatures can reach 6000 K in DCPs. Some o f the attractive properties
o f DCP are its ability to handle organic solvents as well as aqueous solutions with high
solid content and its low power consumption. DCPs are effective excitation sources
for most metals. However, intense molecular background from the graphite electrodes
is a problem and sample volatilization is often not complete because o f the short
residence times o f the analyte in the plasma.
The ICP is the most extensively used plasma source for atom excitation and
ionization.39 The plasma is maintained by a radio frequency field (between 5 to 50
MHz) interacting with a 12 to 18 L/min argon flow. The typical powers at which ICPs
are operated are 1 to 1.5 kW. The plasma temperature at the hottest region is
between 9000 to 10,000 K. Because of their high thermal energies, ICPs are excellent
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
sources for almost all metals.40 The analyte residence times in the plasma are fairly
long (2-3 ms). ICPs have very high electron densities; hence, ICP performance is not
significantly affected by the addition of easily ionized elements. Detection limits of
part per trillion levels are often achieved under ideal conditions. However, ICPs do
not possess sufficient thermal energy to produce UV or visible nonmetal atomic
emission.
Compared to all other plasmas, helium MIPs exhibit a unique characteristic of
producing intense atomic emission from nonmetals C, H, O, S, P, Cl, Br, I, and F41,43
in the 200 to 800 nm spectral range. MIP discharges most commonly are formed in a
quartz or a ceramic tube placed in the center of a Beenakker TM0I0 resonant cavity.
Power is supplied to the cavity via a coaxial cable from a 2.4S GHz microwave
generator. Impedance matching o f the plasma system may be done with the help of
tuning stubs and MIPs are most commonly operated at powers from 20 to 500 W.
Typical He gas flow rates for sustainance o f the plasma are from 1 to 15 L/min. Like
ICPs, a microwave plasma is initiated by providing “seed” electrons from a spark. In
the microwave field the electrons gain sufficient momentum and ionize the support gas
by collisions. More and more electrons and ions are formed in a cascade, and stable
atmospheric pressure plasma temperatures range from 2000 to 5000 K. Compared to
other types of plasmas, MIPs consume less gas.43 Since MIPs are usually smaller than
ICPs and impedance matching is somewhat more difficult, MIPs are less tolerant to
liquid sample introduction. The lower power plasmas (<120 W) are often extinguished
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
when significant amounts o f solvent are introduced. However, MIPs are very well
suited for gas phase sampling and have been successfully used in conjunction with gas
chromatographic systems.44 Because strong emission lines are observed for nonmetals
with MIPs, they are suitable for determinations of organics which are mostly
composed of nonmetals. However, their intolerance to solvents, especially organic
solvents, must be addressed.
1.3.1 He - MIP Nonmetal Excitation Phenomenon
Examples of intense nonmetal ion emission in He-MIPs have been reported in
the literature. However, calculations o f nonmetal electronic-state emissions using
thermodynamic and kinetic theory do not suggest strong nonmetal emission from
MIPs.7 Published detection limits o f GC-MIP for Br at the 478.6 nm ion emission line
have been reported to be 106 pg and that o f Hg at the 2S3.7 nm atom line is 60 pg."
These masses translate to 8.0 x 10u Br atoms and 1.8 x 10n Hg atoms, respectively.
Calculations from the Boltzmann relationship (equation 1.1 below) using a He MIP
temperature o f5000 K and electron density of 5 x 1014cm*3 indicate that the ratio of
excited Br ions to ground-state Br atoms should be 4.1 x 10*20. Similar calculations of
the Hg atom excited-state to ground-state ratio yielded a value o f 1.7 x 10*4. Applying
transition probabilities, the detection limit photon flux was calculated to be 3.0 s*1for
Br and 2.5 x 1014s'1for H g45 Although the calculated number o f photons that should
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
be produced by Br is 14 orders o f magnitude less than that o f the Hg, the detection
limits observed were very similar.
This discrepancy between calculated and observed detection limits, and, thus,
emission intensities suggest phenomena other than thermal excitation o f atoms or ions.
A number o f theories have been proposed to explain this He MIP behavior. The most
accepted theory, however, is that of the charge transfer mechanism proposed by
Carnahan and colleagues.4*’47 As shown in equation 1.1, the theory is based upon
simultaneous nonmetal ionization and excitation by the interaction with ionized helium,
where N is a nonmetal.
N(0eV) + He\2A.5*eV) -> N+'(xeV) + He(0eV) + A £(24.58-xe^
(1.1)
This process is known as charge transfer. The energy defect (AE) represents the
energy difference between the ionization energy of He and the energy required to
promote the nonmetal from the ground-state atom to the excited-state ion. Small
energy defects enhance the likelihood of charge transfer occurrence.
Rapp and Francis4* stated that the ionization cross-section for one electron
transfer, such as that above where only one electron o f the species to be ionized is
affected, is maximized when the reaction is near resonant. The ionization cross-section
(for the effective rate constant) may be estimated as shown by equation 1.2.
0 = 0 .0 0 2 2 1 - ^ 1
(1 2 )
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
where o is ionization cross-sectional area, T is the square root o f the ionization
energy o f the element to be ionized divided by the ionization energy of hydrogen, v is
the velocity, a is the Bohr radius, AE is the energy defect, and h is the Planck’s
constant As an example, equation 1.3 shows the excitation o f Cl ion in He-MIP via
one-electron charge transfer. The difference between the ionization energy of He and
Cl+* is very small (AE = 0.01 eV). The most intense Cl emission line is that o f an ion
and is observed at 479.45 nm.
ClQs1 3p 5) + fle * (U 1) -
C r 'O s 1 3p s) * ffe (U 2) + 0.0leV (1.3)
Similar behavior is seen for other nonmetals. Figure 1 shows the energies o f all
electronic states for singly charged C, N, O, F, P, S, Cl, Br, and I. Values were
calculated by adding the element ionization energy to the energy of the specific ion
state. The energies are arranged vertically above the symbol for the element. The two
horizontal lines represent the ionization energies o f Ar and He. Close energy matches
are seen with He* for elements I, S, P, Cl, and Br. Strong ion emission lines in the UV
and visible spectral regions are seen for these elements in the He-MIP. C also
participates in the charge transfer process, but C* emits in the vacuum UV region. On
the other hand N, 0 , and i do not have any excited ion states close to the ionization
energy of He. In accord with the charge transfer theory, these elements do not show
strong ion emission lines in the He - MIP.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
10
130
C
Figure I.
N
O
F
P
$ Cl Br
I
Energy diagram for nine nonmetals. The horizontal lines
represent ionization energies o f Ar and He. The dots
indicate the energies of the ion states o f the respective
nonmetals. The energies are calculated as the energy of
ionization plus the energy of ion excitation. (This figure is
taken from Brandi and Carnahan.7)
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
11
1.4 Liquid Sample Introduction Devices
The sample introduction method plays an important role in the performance of
plasma-based atomic and mass spectrometric techniques.49,30 To utilize the full
atomization, excitation, and ionization powers of plasmas, suitable sample introduction
devices and methods must be used. Gaseous analytes are the ideal samples for plasmabased spectrometry because analyte transport is very efficient.51 In most cases, nearly
100% o f the gaseous sample can be introduced to the plasma. No desolvation or
volatilization is necessary; hence, most o f the plasma energy may be used for
excitation and ionization.32'34
For liquid sampling, pneumatic and ultrasonic nebulizers are commonly used.
These devices typically disperse the liquid sample into droplets o f a fine aerosol, which
are transported to the plasma by a carrier gas. With USN, more than 90% o f the
sample is typically discarded. That value is approximately 99% with pneumatic
nebulizers. Because o f the large amounts o f solvent associated with the analyte, much
energy is used for desolvation. Plasma energy is also used for isolation o f the analyte
from the residual solid analyte matrix. Matrix effects often diminish or interfere with
the analytical signal. These inherent difficulties with liquid sampling have prompted a
great deal o f research toward understanding the formation o f aerosols by various
nebulization devices as well as analyte desolvation before introduction to the plasma.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
12
In the following sections, various pneumatic55,36and ultrasonic nebulizer57,51 properties
will be discussed. Emphasis will be given to key aerosol characterization parameters.
1.4.1 Pneumatic Nebulization
Pneumatic nebulizers are operated most commonly in two basic configurations,
crossflow and concentric type arrangements. In the concentric arrangement the
sample solution passes through a capillary surrounded by a high-velocity gas stream
parallel to the capillary axis. In the crossflow type, the sample-carrying capillary is at
right angles to the gas stream. In both systems a pressure differential is created at the
sample capillary tip and the amount of liquid drawn through it is described by
Poiseuille’s equation (equation 1.4).
e= —
*
8q l
ci.4)
where Q is the flow rate o f the liquid, R is the capillary radius, P is the pressure
differential, TJ is the viscosity o f the liquid, and L is the overall length of the capillary.
The relationship between the aerosol droplet diameter and the solution properties in
conjunction with nebulizer parameters given by Nukiyama and Tanasawa59 is shown in
equation 1.5.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
13
585
v(o/p)w
+ 597[
*1 l 0-45 ( 1000^ \ 1.5
(op)'*
(1.5)
Q ".
where d0 - mean droplet diameter (pm), v = velocity o f gas (m/s), 0 = surface tension
(dyne, cm), p = density o f liquid (g/mL), T] = viscosity (poises), and QKqand
=
volume flows o f the liquid and gas, respectively.
1.4.2 Ultrasonic Nebulizer
With an ultrasonic nebulizer, the sample solution is typically introduced onto
an inert surface attached to a piezoelectric transducer. The transducer is driven by an
ultrasonic generator at frequencies o f200 kHz to 10 MHz. The ultrasonic generator
produces longitudinal waves that propagate from the surface of the crystal, through
the inert material (usually quartz or glass), toward the liquid-air interface. Aerosol
formation occurs when the amplitude of the wave becomes large enough to
significantly disrupt the surface o f the liquid film. The wavelength o f the surface wave
is given by Kelvin’s formula60,61 (equation 1.6):
(1.6)
P/ 2
where X is wavelength, 0 is surface tension, p is liquid density, an d /is ultrasonic
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
14
frequency. The average droplet size (D) is given by equation 1.7.
D = 0 .3 4 1 = 0.34
(17)
P/2
As is evident from equation 1.7, the average particle size is dependent on the liquid
properties and the frequency o f excitation.
However, from the atomic spectrometry point o f view, the average droplet
diameter is not very useful. It does not give any information about the mass
distribution. The Sauter mean diameter represents a value based on volume to surface
area ratio (equation 1.8),
i E(d3 An)
« = - L- — ~
s E(d2A«)
(1.8)
where d, is the Sauter mean diameter, d is the diameter of drops, and An is the
number of drops o f diameter d. This representation is an excellent approximation of
the mass median diameter. In plasma techniques, it is important to rapidly and
efficiently convert analyte within droplets into atomic and ionic species. A droplet
distribution with a smaller Sauter mean diameter means it has a larger surface area per
unit mass, which translates to easier atomization and ionization.
Recent studies show that droplets with diameters smaller than 8 iim contribute
significantly to the analyte signal produced with Ar ICP.56,62 Analytes within larger
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
15
droplets do not undergo complete desolvation-vaporization-atomization-ionization.
Larger droplets may actually pass through the plasma unaffected.63 To assess the
efficiency o f a nebulizer system, parameters like droplet size and mean diameter,
cumulative percent volume o f aerosol under a certain droplet size, span of a
distribution, axial velocity, radial velocity, droplet number density, and volume flux are
often examined. Figure 2Mshows the particle size distribution of a primary aerosol
(aerosol just after nebulization) from a pneumatic nebulizer and a USN using 100%
water. The volume percent of droplets less than 8 |im in diameter from USN is much
higher (roughly 9 times) than that observed for the pneumatic nebulizer.
The amount o f solvent converted to droplets by the USN was also about 8
times higher than that o f the pneumatic nebulizer.64 Hence, desolvation of the analyte
is necessary when USN is used. Typically, the aerosol from the USN spray chamber is
passed through a heated tube for droplet vaporization and subsequently through a cold
condenser for solvent vapor condensation to minimize solvent load. Such an
apparatus is shown in Figure 3. The approach is generally sufficient during aqueous
sample introduction. However, with organic solvents, additional desolvation is
requited.
1.5 Desolvation Devices
Various devices have been used for analyte desolvation. Some common
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
16
25
20
h 15
5
0
1
10
100
1,000
Diameter (jim)
Figure 2.
Comparison o f primary pneumatic (a) and primary
ultrasonic (b) volume distributions for water. Liquid uptake
rate and argon nebulizer gas flow rates were 1.0 mL/min
and 1.0 L/min respectively. Laser Fraunhofer diffraction
was used to obtain these data. (This figure is taken from
Tarretal.64)
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
fnrfCPgas
ana vapor
out
i
Argon
carrier
Sas
Plasma
fte te d
u-rube
Spray
chamber
If I
, / /
?£ * 0 elc ctric J
transducer
Figure 3.
Tj,
lVasf=
n
'a i“ ,fiD“CS
'C
San,.,,.
s°httion
? ,,hao'* * ./v ator. ^
® n ie /s
ReProduced
Wmpsm'i^ 0 „ 0fth
it
;; ;j
18
examples o f desolvation techniques are cooled spray chambers,6346 cryogenic
desolvation,22,23 heated - cooled condenser devices,67*71 and membrane desolvation.34,72
Some o f the other means o f desolvation used are jet separation,73 acid desiccation,74
and moving-band devices.31,35
1.5.1 Cryogenic Desolvation
Houk and coworkers73*77and Montaser et al.7* have done significant studies
with cryogenic desolvation devices for ICP MS. A schematic o f this device is shown
in Figure 4a. This device transports the aerosol several cycles alternately through
heated tube (140 tol50°C) and condenser (0 to -10°C) regions. This repeated cooling
(-80°C) and heating (140°C) eliminates most o f the associated solvent. The dry
analyte particulates are then transferred to the plasma for detection. This system is
equally efficient for both aqueous and organic solvents.
1.5.2 Hydraulic High-Pressure Nebulizer Desolvation
This nebulizer-desolvation (Figure 4b) device79,80 uses a high-pressure pump to
force the liquid through a 5 to 30 |im orifice. Aerosol is formed as this high-velocity
jet exits through the small opening and subsequently collides with an impact bead to
form an aerosol cloud. -The aerosol passes through a heated tube and two-stage
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
19
/
Second
cryocondenser
' (glass)
^
Heating
coils
(140°C)
To plasma
Aerosol
in
Absolute
ethanol
(-80°C)
Cooling
coiis
(a)
Heated
tube
Spray
chamber/
Coolant
To plasma
Liquid
condeaser
(25-30 °C)
Peltier
condenser
(15 to -40°C)
To waste * ■
u
Heat
sink
Coolant
in
Temperature controller
and power supply
0»
Figure 4.
(a) Cryogenic desolvation device used with the ultrasonic
nebulizer and (b) desolvation set-up used with the hydraulic
high-pressure nebulizer. (The figures were taken from
Montaser.40)
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
20
condensation chamber. The efficiencies o f condensation for aqueous and methanol
solutions were found to be 97% and 94%, respectively. Analyte transport efficiencies
were approximately 24%. Compared to pneumatic nebulization, order o f magnitude
improvements in detection limits were observed. However, with organic solvents,
small amounts of 0 2 must be added to reduce carbon deposition in the torch.
1.5.3 Membrane Desolvation Systems
Desolvation using membrane desolvators has shown great promise for both
plasma atomic emission and plasma mass spectrometric systems. Membrane
desolvators selectively allow the transport o f either analyte or associated solvent
across a semipermeable membrane. Typically, the wet aerosol originating from a
nebulizer is transported by the carrier gas into the central membrane channel. The
desolvator unit is maintained at a temperature higher than the boiling point of the
solvent (usually 150°C for water). The solvent goes into vapor form and diffuses
through the membrane to the outer channel. A sweep gas or a vacuum pump is used
to flush the solvent from the outer channel. The analyte, which usually has a boiling
point much higher than that o f the solvent, remains in a solid particulate form and does
not undergo membrane transport. These dry analyte particulates may then be directed
to the plasma.
With hydride,*1"*? carbon dioxide,** sulfur dioxide,*9 and ammonia*9,90
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
generation schemes, the volatile analyte undergoes membrane transport This mode
has also been used for plasma introduction o f volatile organic analytes such as phenols,
aldehydes, ketones, and carboxylic acids in aqueous matrices.91*96 However, for
aerosols originating from nebulizers, the solvent is usually converted to vapor form
and the vapor undergoes membrane transport This mode is more suitable for plasma
excitation sources and this discussion will be focused on such.
In terms o f types o f the membrane used, there are four major classes: porous,
nonporous, mixed nonporous-porous, and ion exchange. The mechanisms o f solvent
or analyte transport across these membranes are distinct from each other. For a
nonporous membrane, transmembrane transport occurs by permeation. Permeation
involves analyte or matrix solubilization followed by diffusion across the membrane.97
With porous-type membranes, however, the transport is strictly diffusion. The
desolvation efficiency and selectivity is governed by pore size, membrane thickness,
and its chemical characteristics (hydrophobicity or hydrophillicity).9* Some common
examples o f membrane materials are silicone,72,96,9* fluoroplastics
(polytetrafluoroethylene [PTFE], polyvinylidenedifluoride [PVDF]),99*105 cellulose,106*
101polypropylene,,09*UI and ion-exchange membranes"2 like Nafion."3*"5
Gustavsson and coworkers9M12 developed a planar nonporous silicone
membrane separator116,117for use with ICP-AES for both organic and aqueous
solutions. The aerosol was heated to form solid analyte particulates and solvent
vapor. This mixture was passed over the semipermeable membrane. The solvent
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
22
vapor permeated through the membrane because o f the negative pressure on the other
side maintained by a vacuum pump. Eighty to 100% desolvation was observed for
solvents like freon, chloroform, and water. Carrier gas flow rates were 0.5 >1.0 L/min
and solvent uptake rates were 0.5 tol.7 mL/min. Membrane desolvation in
conjunction with thermospray nebulizers has been used to introduce samples a into
plasma mass spectrometer by Montaser.40
A tubular microporous teflon membrane desolvator was used with an
ultrasonic nebulizer by Butto and Zhu.111 Diffusion o f solvent vapor across the
membrane was facilitated by employing a gas flow in a direction opposite to the flow
of analyte. Aqueous, organic solvents and petroleum products with boiling points
below 100°C have been successfully desolvated using this system. Bemer and co­
workers (in Montaser40) have used this system for oxide-ion reduction interference for
ICP-MS measurements. Tao and Miyazaki119have explored the application o f a
hollow non-porous polyimide membrane separator. This device was used with a
pneumatic nebulizer for ICP-MS applications. Significant decreases in MCT/M+ ratios
and ArCT signals indicated water loading reduction. Naflon membranes also have been
utilized by many researchers because o f their unique selectivity properties. The
hydrophilic sulphonic groups on the surface o f the membrane are extremely efficient in
eliminating associated aqueous solvent from the analyte. Yang et al.113have
successfully used this desolvator for thermospray ICP-AES applications and 99.9%
desolvation efficiency was observed for 1% (v/v) HN03. In addition to superior
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
23
desolvation efficiencies, these desolvators were found to be pulse dampeners,
eliminating plasma flicker often seen with thermospray nebulizers. Cairns et al."4 used
Nafion membranes with methanol and acetonitrile for use with HPLCICP-AES.
Desolvation efficiencies o f 89% were observed, which minimized baseline drift during
gradient elution. Akinbo and Carnahan34 investigated the use of a hydrophobic
nonporous polypropylene membrane for desolvation with ultrasonic nebulization for
MIP-AES applications. The desolvation efficiencies for both aqueous and organic
solvents were observed to be 99.99%.
Compared to other desolvation devices, the membrane separation devices are
advantageous for their superior desolvation efficiencies, ease of operation, inertness,
convenient optimization procedures, and ruggedness. However, design and choice of
membrane is extremely important for desirable desolvation efficiency for each
application of concern.
1.6 Research Objective
In this dissertation the analytical performance characterization and optimization
o f a microwave-induced plasma atomic emission system with ultrasonic nebulizer
sample introduction and membrane desolvation will be discussed. The application of
this system as a detector for reverse-phase HPLC is also addressed. For the sake o f
clarity the thesis has been divided into four parts.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
24
The second chapter discusses the characterization and optimization o f the MIP,
ultrasonic nebulizer, and membrane desolvator conditions for nonmetal (Cl, Br, I, S,
and P) atomic emission maximization in inorganic and organic analytes in aqueous and
organic solvents. Parameters such as plasma, carrier and countercurrent gas flows, as
well as plasma power are emphasized. The third chapter focuses upon organic analyte
transport studies through the MIP-ultrasonic nebulizer-membrane desolvator system.
Membrane desolvator conditions for low-boiling-point organic analytes and ultrasonic
nebulization problems for different HPLC solvent compositions are discussed. The
fourth chapter details applications o f reverse-phase HPLC-MIP AES. Figures of merit
o f this detector with respect to common LC detectors (UV and refractive index) are
presented. Applications of this system for detection of biologically active and
pharmaceutical products are demonstrated. The last section suggests steps that should
be taken to further enhance the performance of this system.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
CHAPTER 2
HELIUM MICROWAVE-INDUCED PLASMA-USN-MEMBRANE
DESOLVATOR ATOMIC EMISSION SYSTEM CHARACTERIZATION
2.1 Introduction
The primary objective o f this project is to utilize a membrane desolvator in
conjunction with a USN-He-MIP system for element-selective nonmetal detection o f
organic analytes present in aqueous and organic solutions. Once the system is
optimized it is intended to be used as a detector for reverse-phase HPLC. As
discussed in section l.l, organic solvents severely diminish the analytical
performance o f the plasma. Hence, desolvation of the analyte is imperative before
introduction to the plasma. Membrane desolvators have been effectively used with
aqueous solutions. In this chapter the efficiency o f the membrane desolvation for
organic analytes will be investigated. Also, the optimized operating conditions o f the
plasma, USN, and the desolvator will be presented.
Figure 5 shows a general diagram o f the MIP-USN-membrane desolvator
system. The sample solution was introduced to the USN with a peristaltic pump. A
helium carrier gas is used to drive the resultant aerosol through the heated tube-cold
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
26
Analyte Flow
Gas Flow
......
\ He.:
Coaatcr Carreat Gai
Mcatbraae
Desolvator
Moaockroaiator
Peristaltic
Parap
Ncbalbter
He ;►
Microwave
Gcaerator
Carrier Gas
Figure 5.
Schematic for the MIP-USN-membrane desolvator atomic emission
system.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
27
condenser combination of the USN. The solvent evaporation/condensation step
serves to partially desolvate the analyte. Passing the analyte stream through the
membrane desolvator further desolvated the analyte. The dry analyte particulates
were then directed to the plasma for detection. Emission from the plasma was
focused onto a monochromator for wavelength selection. The detection was done
using a photomultiplier tube (PMT). Data was acquired and manipulated by a
computer.
2.2 Fate of the Analyte as It Passes Through the System
Figure 6 depicts the evolving matrix of the analyte as it passes through a
conventional USN system from the point o f aerosol formation. The primary aerosol
from the USN is directed through a heated solvent vaporizaton tube followed by a
cooled solvent condenser tube. The partially desolvated analyte exiting the condenser
is introduced to the plasma. Thermal heat transfer from the plasma atomizes the
analyte and, subsequently, additional energy transfer in the MIP produces excited
atoms and excited ions. However, even after this partial desolvation of the analyte
stream, a sufficient amount o f solvent still remains to alter the plasma properties.
This is evidenced by intense OH emission bands in the case o f aqueous solvents. A
myriad o f molecular emission bands (CH, CO, CN, etc.) appear when organic
solvents are utilized. Hence, further desolvation o f the analyte stream is desirable. A
viable approach is the use o f membrane desolvation. A schematic diagram o f the
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
28
Thermal evaporation
followed by
condensation in USN
Thermal heat
transfer in plasma
M
*
Particulate
plus solvent
vapor
M
Free atoms plus
solvent vapor
\
Additional
energy transfer
in plasma
M* X
M*
X**
Excited atoms and ions and
solvent vapor chemistry
Figure 6.
Fate of the liquid droplet as it passes through the USN-MIP AES
system.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
29
CETAC MDX-100 membrane desolvator is shown in Figure 7. In the normal
configuration, the desolvator is placed between the outlet of the USN condenser and
the plasma. For aqueous sample mists, the membrane desolvator module is
maintained at a temperature o f 150°C. The wet analyte particulates enter the hot
central channel, encased by a microporous PTFE membrane. Ideally, residual solvent
is converted to the vapor form and diffuses through the membrane to the outer
channel. The driving force for this process is the osmotic pressure of the solvent
vapor. A helium countercurrent gas is used to flush the solvent from the outer
channel. The heavier nonvolatile analyte particulates do not undergo membrane
transport and exit the desolvator with a minimum amount of associated solvent. The
dry analyte is then introduced to the plasma for detection.
2 3 System Specifications
The experimental system was similar to that used by Akinbo and Carnahan36
with a number o f modifications. The reader should refer to the block diagram shown
in Figure S. A description o f important system components follows.
2.5.7 He-MIP Plasma System
A 120 W maximum power output, 2.45 GHz microwave generator (model
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
30
Partially
desolvated
analyte en terin g
m em b ran e
deso lv ato r
, ( OOOOOOOOQQQOO
.
• .
..
1
* •
». •
Sample inlet
(from USN)
D esolvated
analyte e x itin g
m em brane
d eso lv ato r
Sample outlet
(To H e-M IP )
|UCXXXX30CX300001f
H e a te r
Solvent
diffusion
Figure 7.
C ountercurrent
He gas + solvent
vapor
I
Countercurrent
gas(PureHe)
CETAC MDX-100 tubular membrane desolvator. Adapted from
literature - Cetac Technologies.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
31
MP6-4m-229, Kiva Instruments, Rockville, MD) was used. For some o f the
experiments a 500 W generator (model 420 B, Micro-Now Instrument Co. Inc.,
Chicago, IL) was used. For the lower power system the plasma was maintained in a 5
mm i.d., 7mm o.d., tangential flow quartz tube with a threaded teflon insert. For the
500 W plasma, an 8 mm i.d., 10 mm o.d. quartz tube was used. The torch was
inserted through the center o f a copper Beenakker TM010cavity with an inner diameter
o f 89 mm and a depth of 11 mm. Impedance matching was accomplished with a
three-stub tuner (model 1878B, Maury Microwave, Cucamonga, CA). The forward
power was 120 W and the reflected power was between 0 and 1 W.
2.3.2 Sample Introduction System
A Cetac U 5000 (Cetac Technologies, Omaha, NE) ultrasonic nebulizer with a
manual transducer (operated at 1.39 MHz) was used for solution nebulization. A
Rabbit peristaltic pump (Rainin Instruments, Wobum, MA) was employed for direct
solution nebulization of bulk samples.
For the HPLC studies, a dual-pump HPLC system (Dynamax, model SD200,
Rainin Instruments) and a C „ reverse-phase column (4.6 X 250 mm, 5|im particle
size, Microsorb-MV™ from Rainin Instruments) were used. A six port injector with a
20 JiL sample loop (Rheodyne, Cotati, CA) delivered the analyte to the column.
Eluates were directed to the USN via a 0.01 inch i.d. PEEK tube (Rainin Instruments,
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
32
Woburn, MA). The mobile-phase flow rate was maintained at 0.55 mL/min. For
certain experiments, 0.5 mL/min 0.1% H2S04 in methanol was mixed with the
effluent from the column through a post column “T” junction and the mixed solution
was directed to the USN.
2.3.3 Aerosol Desolvation System
Two-stage aerosol desolvation was utilized. The USN mist was initially
directed through the Cetac U5000 desolvator where it was heated to 140°C and then
condensed at -10°C with the water-ethyiene glycol-cooled condenser. The second
stage consisted o f a CETAC MDX-100 microporous tubular PTFE membrane
desolvator. Some o f the membrane dimensions are i.d = 0.5 cm, length = 2 m, pore
size = 1 |im .
2.3.4 Optics
A 10-cm focal length fused silica lens was used to focus the axial plasma
image on the entrance slit o f a 1-m focal length spectrometer (model 1000 M, Spex,
Edison, NJ). A 1200 grooves/mm grating blazed at 500 nm was used. Slit widths
were set at 10 |im . The photomultiplier tube (PMT) (model IP 28, Hamamatsu,
Middlesex, NJ) was biased at -850 V. For preliminary experiments, a low-resolution
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
33
0.35-m focal length GCA/McPherson (model EU/E-700, Acton, MA) spectrometer
was used. The grating had 1200 grooves/mm with a blaze wavelength o f250 nm.
2.3.5 Data Analysis
Signal from the PMT was acquired by a 486 DX, 33 MHz computer using
Spex autoscan software. Data integration time was 1 second per point. Quattro Pro*
software was used for data analysis. All measurements were obtained in triplicate.
Detection limits were calculated as the Cl concentration yielding a signal 3 times the
standard deviation o f the background.
2.3.6 Reagents
Greater than 99.99% pure He with a flow rate o f 12.3 L/min was used to
sustain the plasma. Reagent-grade MgCL (J.T. Baker Chemical Co., Phillipsburg,
NJ), 2 ,6-dichlorobenzamide (DCB), 4-chlorobenzamide (CB), and 2,4,6trichlorobenzoic acid (TCBA) (Aldrich, Milwaukee, WI) were used. Reagent-grade
methanol (Mallinckrodt Specialty Chemicals Co., Paris, Kentucky) was used for
making organic analyte-containing stock solutions and for the HPLC mobile phase.
Methanol and water mobile phases were boiled and sonicated for 30 minutes each to
remove dissolved gases.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
2.4 Experimental
For initial characterization, bulk samples were delivered to the nebulizer with
the peristaltic pump. Five hundred ppm chlorine was introduced as MgCl2in pure
water. Each sample was introduced with and without the membrane desolvator “in
line.” Characteristic Cl emission lines were monitored at 479.5,481.0, and 481.9 nm.
The operating conditions for this experiment are listed in Table 1. Effects o f the
membrane desolvator on aqueous Cl signal intensity were studied by comparing the
Cl signal at 479.5 nm, with and without the membrane desolvator. Figure 8a
illustrates chlorine emission without the membrane desolvator in place. As shown in
Figure 8b, use o f the membrane desolvator results in a 4-fold increase in signal.
Additionally, a significant decrease in background noise was seen when using the
membrane desolvator. A 100-fold improvement o f the detection limit, from 200 to 2
ppm, was seen with the membrane desolvator. Observed enhancement of signal with
membrane desolvator established the fact that solvent loading adversely affects the
analytical performance o f the plasma, and membrane desolvators can effectively
reduce this problem with aqueous solution nebulization.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
35
Table 1.
Unoptimized Conditions Used for Preliminary Experiments.
Plasma
Power
Plasma gas flow
120 watt
lOL/min
Peristaltic pump
Uptake rate
1.5 mL/min
USN
Carrier gas flow rate
Heated tube
Condenser temperature
0.5 L/min
140°C
5°C
Membrane desolvator
Countercurrent gas flow
Desolvator temperature
1 L/min
160°C
PMT
Voltage
- 900 volts
Monochromator
Focal length
Slit width
Integration time
-
0.35 m
100 pm
0.1s
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
36
100
Is "
|
40
20
479
460
461
482
483
Wavelength (nm)
(»)
100
80
1 60
?
40
20
479
480
481
482
483
WavatmgSi (nm)
(b)
Figure 8.
(a) Spectrum obtained from aqueous solution of 500 ppm
Cl as MgCl2 without and (b) with membrane desolvator.
Three characteristic Cl lines at 479.5,481.0, and 481.9 nm
are seen. All the peaks are normalized against the highest
peak in spectrum 8b.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
37
2.5 System Optimization
Three independent He gas sources were used in the system. The plasma was
sustained by the plasma gas. A carrier gas was used to drive the aerosol produced at
the USN through the system, hi the membrane desolvator, a countercurrent gas
flowing opposite in direction to the flow of the analyte is used in the outer channel to
flush out the solvent (Figure 6). The flow rates o f these three gases are critical for
optimal system performance. These flow rates were optimized for inorganic analytes
in aqueous and organic solutions and Anally for organic analytes in methanol and
acetonitrile solutions.
2.5.1 Inorganic Analytes in Water
Cl signals from aqueous MgCl: solutions were observed as each o f the flow
rates was varied. Figure 9 is a plot o f Cl signal as a function o f plasma gas flow rate
with the carrier gas flow set at 0.5 L/min. For this experiment the membrane
desolvator was not used. The greatest Cl signal was observed at a plasma gas flow
rate o f 12.5 L/min. Any further increase in the flow rate caused a reduction in signal,
probably due to a combination o f decreased residence time o f the analyte in the
plasma, analyte dilution, and plasma flicker due to nonlaminar gas flow through the
torch. Flow increases beyond 14.5 L/min resulted in extinguishment o f the plasma.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
38
a
§
120
, ----------------------------------------------------------------------------------------
100
|
I
gg : USN gas flow = 0.5 L/min
'35
O
60
o
z
40
20
10
11
12
13
14
15
Gas flow rate (L/min)
Figure 9.
Cl signal as a function of plasma gas flow rate. The USN gas flow
is set at 0.5 L/min. For this experiment the membrane desolvator
is off-line.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
39
From this experiment 12.5 L/min was determined to be the optimum flow rate for the
plasma gas.
The optimum carrier gas flow with and without membrane desolvator was
determined at a plasma flow rate of 12.5 L/min. Aqueous Cl signal was observed as
the carrier gas flow rate was varied. The results o f this experiment are determined
and shown in Figure 10. When the membrane desolvator was removed from the
system, the optimum flow rate was observed to be 0.5 L/min. Increasing the flow
rate beyond 1.75 L/min caused extinguishment o f the plasma due to excessive solvent
loading. However, with the membrane desolvator, the optimum carrier gas flow rate
was at 2.25 L/min. Any further increase in the flow rate caused plasma flicker and
extinguishment.
The membrane desolvator countercurrent gas flow rate was optimized in a
similar fashion at a plasma flow rate o f 12.5 L/min and a carrier gas flow o f 2.25
L/min. The results are shown in Figure 11. The Cl signal peaked at 1.25 L/min and
declined gradually with subsequent increases in the flow rate. While purely
speculation, it is possible that the decline in signal at flow rates higher than 1.25
L/min could be attributed to analyte collisions with the tube wall due to the
introduction of turbulence in the outer membrane desolvator channel.
A Cl concentration calibration plot was obtained with aqueous MgCL solution
under the optimized conditions. The calibration plot is shown in Figure 12. The plot
was linear with a correlation coefficient (r2) value o f0.99994. The slope and the
intercept o f the plot were calculated to be 0.26 counts/ppm and 5.01 counts
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
\
100 i Plasma gas flow = I2.S L/min
Countercurrent gas flow = 1
L/min with membrane desolvator
80 I
n
§>
I
10
i
S 60
.M
•
i
i
O 40 !
Z
Without MDX
20
-m -
With MDX
o
i0
Figure 10.
0.5
1
1.5
USN gas flow (L/min)
2
2.5
Cl signal as a function o f USN gas flow. The plasma flow rate is
set at 12.5 mL/min. Membrane desolvator countercurrent gas flow
rate is set at 1 L/min.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
41
,---------------
Normalized Cl signal
120
Plasma gas flow rate =
12.5 L/min
100
;
80
j
60
i
40
i
20
■
*
0
/
•
--------------------------
0
Figure 11.
0.5
1
1.5
countercurrent gas flow rate (L/min)
2
Cl signal as a function o f membrane desolvator countercurrent gas
flow rates.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
0
Figure 12.
50
100
150
Concentration in ppm
200
250
Cl calibration plot under optimized conditions. Cl introduced as
aqueous MgCl? solution.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
43
respectively. The Cl detection limit (DL) was calculated using the formula of
equation 2.1 and was found to be 1 ppm.
DL = 3 * C * ( | )
(2.1)
where C = analyte concentration, N = noise (standard deviation o f the background),
and S = background corrected signal. The detection limits achieved for aqueous Cl
solution were similar to those obtained by Akinbo and Carnahan using a flat-sheet
polypropylene membrane desolvator with a USN-MIP AES system.34,36
2.5.2 Inorganic Analyte in Methanol
The objective o f the project is to use the MIP AES system for reverse-phase
HPLC applications. Hence, investigation of the plasma behavior with organic solvent
nebulization is necessary.
For this experiment, a solution of MgCL in methanol was introduced to the
plasma without the membrane desolvator. To facilitate better desolvation o f the
analyte in the USN, the condenser was set at -S°C. Upon introduction o f methanol
without the membrane desolvator, the plasma changed from its usual color of pink
(due to H|T] 656 nm emission) to an intense blue (due primarily to C2 and CH
emission bands) and was extinguished within a minute. However, with the membrane
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
44
desolvator, the plasma could be sustained without extinguishment. Optimizations o f
gas flows in the system were also performed for the organic solvent solutions. The
optimum plasma and USN gas flow rates were similar to those observed for aqueous
solution nebulization. The optimal countercurrent flow was found to be 3.S L/min,
much higher than the 1.25 L/min optimal flow rate with aqueous solutions. Figure 13
shows Cl signal intensity from a methanolic 500 ppm MgCl2 solution in methanol as
the membrane desolvator countercurrent gas flow is varied. Using the (organic MgCL
solution) optimized conditions, Cl detection limits were 1 ppm. This value is the
same as that determined for the aqueous MgCL solution. These results show that
membrane desolvation is effective even with organic solvents.
2.5.3 Characterization o f the System with Organic Analytes
The system showed promising performance with inorganic analyte salts in
both aqueous and organic solvents. In this section, the behavior of the system with
organic analytes in organic HPLC-type solvents such as methanol-acetonitrile-water,
methanol-water, and acetonitrile mixtures are discussed.
A 500 ppm Cl solution as 2 ,6 dichlorobenzamide in methanol was introduced
with the membrane desolvator incorporated into the system. The spectrometer was
scanned from 470.0 to 483.0 nm to observe the three characteristic Cl lines at 479.5,
481.0, and 481.9 nm. However, no Cl signal was observed using optimum
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
45
55
Cl signal in arbritary units
50
Plasma= I2.S L/min
USN = 2.25 L/min
45
40
35
30
25
20
0.75
Figure 13.
1.25
1.75
2.25
2.75
3.25
Countercurrent gas flow (L/min)
3.75
4.25
Cl signal as a function of membrane desolvator countercurrent gas
flow. Five hundred ppm MgCl2 in methanol is used.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
46
conditions. Similar behavior with organic analytes was observed by Akinbo and
Carnahan.36 They solved the problem by adding 0.125 M NaOH to the analyte.
NaOH enhanced the transport o f the volatile organics. However, large amounts of
NaOH caused plasma flicker, which led to increased noise. Also, NaOH quickly
deteriorated the plasma torch. Hence, addition of NaOH is not a viable option for
routine analysis.
It was thought that the low-boiling-point organic analytes might volatilize as
they passed through the 150°C membrane desolvator and diffuse through the PTFE
membrane with the solvent vapor. To investigate these possible effects, the
temperature of the membrane desolvator was varied and methanol solutions
containing 500 ppm Cl as DCB and MgCU were introduced. Results are shown in
Figure 14. The inorganic chlorine signal increased as the desolvation temperature
was increased. High temperatures facilitate efficient desolvation and the inorganic
analyte exists as nonvolatile particulates. Contrary to the observations for the MgCl,
solution, at temperatures o f 90°C and greater, the DCB Cl signal decreased
significantly. At higher temperatures, it appears that the more volatile DCB is lost via
transport through the membrane, resulting in a decline in the organic Cl signal.
Similar trends were seen with 4-chlorobenzamide and trichlorobenzoic acid (Figure
15). These results indicate that a membrane desolvator temperature between 20° and
50°C provides the maximum signal for the organic analytes. However, temperatures
below 50°C caused plasma to flicker after 20*30 minutes of operation, due to less
efficient desolvation and the increased amount o f organic solvent being directed to the
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
47
70 , ------------------------------------------------------------------------------------------ -----------------
g0 i
•
50
Sig MgCI2
■'
Sig DCB
« 40 '
c
™
«
•
O 30
20 t
10
• ---------------------------------------------------
0
20
40
60
80
100
120
140
a
---160
Temperature in Centigrade
Figure 14.
Inorganic and organic Cl signal vs membrane desolvator
temperature.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
10
#___________________m
*
*
5 !
TCBA (mp -164)
'
■
" -
«
0
20
Figure 15.
30
40
50
60
MO temperature
70
80
Organic Cl signal vs membrane desolvator temperature.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
49
plasma. Hence, the desolvation temperature was set at S0°C for subsequent
experiments with organic analytes.
Under these conditions the effects of membrane desolvator countercurrent
flow rates on the signal were studied. Results are shown in Figure 16. Five hundred
ppm Cl as 2 ,6 dichlorobenzamide and 4-chlorobenzamide in methanol was
examined. Both analytes exhibited similar tendencies. The peak signal was observed
at a counter current gas flow rate o f 4.0 L/min. A plot o f Cl S/N versus
countercurrent gas flow is shown in Figure 17. The S/N increased gradually with
increases in countercurrent gas flow rates and peaked at 3.5 L/min, similar to what
was observed in the signal versus counter current flow plot (Figure 16). This
behavior can be attributed to increased analyte desolvation with increases in
countercurrent gas flow to a flow rate of 3.5 L/min. A decline was seen at 4 L/min.
Further increases in the countercurrent gas flow rate introduced carrier gas turbulence
resulting in increased background (baseline) noise. A countercurrent gas flow rate of
3.5 L/min was used for subsequent experiments. Complete optimization conditions
for organic and inorganic analytes are shown in Table 2. Detection limits determined,
based on concentrations yielding signal-to-noise ratios 3 times the standard deviation
of the baseline noise, for 2,6 dichlorobenzamide and 4 chlorobenzamide were 2 and
1.5 ppm respectively. Under these conditions, similar detection limits were observed
from DCB and CB in acetonitrile solutions.
These detection limits were comparable to those observed with the aqueous
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
50
60
Sig 2,6 DCB
50 !
Sig 4 CB
_40
ac
e»
M
a
30
20
10
Figure 16.
1.5
2
2.5
3
3.5
Countercurrent flow (L/min)
4
Cl signal vs membrane desolvator countercurrent gas flow.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
51
1200
S/N DCB
1000 i
! j ■S/N CB
800
§600
400
200
0.75
Figure 17.
1.25
1.75
2.25
2.75
3.25
Countercurrent flow (L/min)
3.75
4.25
Cl S/N vs membrane desolvator countercurrent gas flow.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
52
Table 2.
Optimum Operating Conditions for Inorganic and Organic Analytes.
Plasma gas
(L/min)
Carrier gas
(L/min)
Countercurrent gas
(L/min)
Membrane
desolvator
temp. (°C)
Condenser
temp. (°C)
Cl as aqueous
MgCl2 without
membrane
desolvator
12.3
0.8
—
—
5
Cl as aqueous
MgCI2 with
membrane
desolvator
12.3
1.8
1.4
160
5
Cl as DCB in
methanol with
membrane
desolvation
12.3
1.8
3.5
50
-10
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
53
inorganic Cl solutions. These results indicate that the efficiency o f solvent removal by
the CETAC MDX-100 membrane desolvator for both aqueous and organic solvents is
similar. For organic solvents, membrane desolvation is necessary to obtain signal.
Figure 18 shows a Cl spectrum from a 500 ppm Cl as DCB solution in methanol
without the membrane desolvator. No Cl signal is observed. Figure 19 is a 500 ppm
Cl spectrum from a DCB in methanol solution with the membrane desolvator under
the optimized conditions listed in Table 2. The three characteristic Cl emission lines
are clearly seen in Figure 19.
2.6 Examination of Background Emission with Membrane Desolvation
A series o f experiments was performed to examine general effects o f the
membrane desolvation and lower desolvation temperatures on the spectral
background.
Preliminarily behavior o f the plasma was visually inspected as methanol was
nebulized. Figure 20 shows pictures o f the plasma plume captured through a digital
camera. Figure 20a is the picture o f the He plasma without solution nebulization. As
seen in the figure, the plasma has a pink core with a blue outline. The pink coloration
is due to H(I) 656 nm emission and the blue emission at the edge is due primarily to C2
and CH emission bands. Figure 20b is the picture o f the plasma when pure methanol is
being nebulized through the membrane desolvator and the
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
479
Figure 18.
480
481
Wavelength (nm)
482
483
Five hundred ppm Cl signal from 2,6 dichlorobenzamide in
methanol without membrane desolvator. No Cl signal is seen.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
479
Figure 19.
480
481
Wavelength (nm)
482
483
Five hundred ppm Cl spectrum from 2 ,6 dichlorobenzamide in
methanol.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
56
a
Figure 20.
b
e
(a) Image of the original plasma (nothing is being nebulized), ( b)
image o f the plasma when methanol is being nebulized with the
membrane desolvator, and ( c) image o f the plasma when methanol
is being nebulized without the membrane desolvator. A 500-watt
generator was used for this study.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
57
desolvator temperature is set at the 50°C optimum temperature for organic analytes.
A slight increase in intensity of the blue outline coloration seen due to emission from
the methanol that is not eliminated by the desolvator. But the picture o f this plasma
appears similar to the original plasma, indicating that the desolvator efficiently
removes most of the organic solvent, even when operated at 50°C. Figure 20c is a
picture of the plasma as methanol is nebulized directly without the membrane
desolvator. The plasma color turned from pink to an intense blue. The plasma
became unstable under these conditions and the plasma torch was coated with carbon
immediately. This simple set o f studies implied effective removal of organic solvents
by the MDX-100 desolvator, even at the lower temperatures necessary for
semivolatile organic analytes.
To further corroborate the effectiveness the membrane desolvator, background
emission spectra from the plasma were analyzed. Figure 21 depicts the 200*600 run
spectrum during pure methanol nebulization with and without the membrane
desolvator “in-line.” The spectra are similarly scaled but have been offset on the yaxis for visual clarity. Without the desolvator, intense solvent-related molecular
emission is seen from C2, CN, and OH. Additionally, intense C(I) line emission is
observed at 247 nm. With the membrane desolvator in place and the temperature set
at 50°C as dictated by organic analyte considerations, the molecular emission bands
are reduced considerably and the only prominent features are those ofHe(I) lines at
501 and 587 nm. To examine the effects of temperature on the spectral background,
methanol was nebulized and 200-600 nm spectra were obtained with the desolvation
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
58
120
512.9
(C2)
100-
WrthMD
473.7
(C2)
436.5
306
(OH)
20
IC2)
388
(CN)
{CN)
-
200
250
300
350
400
450
500
550
600
Wavelength (nm)
Figure 21.
Methanol background with and without membrane desolvator.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
temperature set at SO, 100, and 150°C. These plots are shown in Figure 22. It should
be noted that, compared to Figure 21, the gain is greatly increased to accentuate the
spectral detail. All the plots in Figure 22 are normalized to the most intense 587-nm
He signal at a membrane desolvator temperature o f 150° C (Figure 22b). Figure 22a
is the emission spectrum from the plasma with no solution nebulization. The most
prominent features in this spectrum are emission lines from the He plasma gas. Figure
22b, 22c, and 22d show the methanol background spectra collected as the desolvator
temperature is varied. As the temperature is decreased from ISO to SOX, all carbonbased molecular emission bands from the solvent increase only slightly in intensity as
a result of reduced transport through the membrane. (C2and OH arise directly from
the solvent. It appears that CN arises from the combination o f carbon from the
solvent with N2either from atmospheric entrainment or helium contamination.) It is
important to note that, while decreasing the desolvation temperature somewhat
decreases the He(I) line emission intensities, the changes in He(I) line intensities are
only slight. This observation indicates that the changes in solvent transport from S0°
to 150°C may not significantly affect the “excitation chemistry” o f the plasma. The
required lower desolvation temperature does not appear to have a deleterious effect on
analyte excitation.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
60
100
Spccmm origan) p b s m
517
n it
4471
"
200
300
1
C(,)
495.7
400
WavMngtfi(nm)
600
500
22(a)
100
MOH iptcm m « MO tamp. 150
He
517
80
He
501.5
He
388.8
60
40
20 i
OH
306
a
01
388
CN
m
-4
200
22(b)
Figure 22.
250
300
He
447.1
436.5
a
H(l)
186
cm
495.7
473.7
C2
5119
Cl
563.6
-J
350
400
Wavelength (nm)
450
500
550
600
(Continued on following page)
Plasma background spectrum at different membrane desolvator
temperatures while methanol is being nebulized, (a) Original
plasma (nothing nebulized), (b) Membrane desolvator temperature
150°C. (c) Membrane desolvator temperature 100°C. (d)
Membrane desolvator temperature S0°C. The plots are normalized
to the He peak at 587 nm at membrane desolvator temperature of
150°C.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Figure 22 (continued)
100
MeOH background at M> tamp. 100
He
587
80
I
I
He
60
501.5
He
388.8
40
He
447.1
OH
306
20
CN
388
l
200
250
300
H(l)
486
C2
4365
CN
421
■-
ji
350
400
Wavelength (nm)
1 C2
j 512.9
C(l)
495.7
Cl
473.7
!
M
450
C2
563 6
1
600
550
500
22(c)
100
MeOH backgound M MOtemp. 50
He
587
80
60
He
5015
40
He
512.9
H(l)
486
388.8
OH
20 i
306
He
447.1
CN
388
Cl
O
CN
C2
CCD
495.7
473.7
C2
563.6
——
I ^1
200
250
300
350
400
Wavelength (nm)
450
500
550
600
22(d)
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
62
2.7 Examination of a 500 W Plasma for Detection
The low-power plasma (100 W) was used for the purpose o f preliminary
characterization of the membrane desolvator. Compared to low-power plasmas, highpower plasmas are more tolerant to solvents and provide higher thermal energy for
analyte excitation. Hence, a 500 W maximum power generator was used to
investigate the possible enhancement o f the analytical detection limits and
sensitivities. For these experiments, the 1-m Spex monochromator was used.
A series o f tests similar to those performed with the 100 W plasma were
carried out to optimize the USN carrier gas flow, membrane desolvator counter
current gas flow, and plasma power. A 500 ppm Cl solution of DCB in methanol was
used to study the effects o f these parameters on the S/B ratios. For this experiment,
the optimum USN flow (1.8 L/min) and countercurrent flow (3 L/min) from the lowpower plasma experiments were used. The minimum gas flow required to sustain the
plasma was about 11 L/min. However, at this flow rate, a heavy carbon deposition
from methanol was seen on the plasma torch, which led to arcing in the torch and
torch melting. This problem was resolved by increasing the flow rate to 15 L/min.
Figure 23 is a plot of Cl S/B ratio as a function of plasma gas flow rates. As the flow
rate was increased above 15 L/min, the Cl S/B ratio decreased steadily. A 10%
decrease in S/B was observed as flow rate was increased from 15 to 19 L/min. This
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
15
Figure 23.
16
17
Plasma gas flow (L/min)
18
19
Plasma gas vs Cl S/B. 500 ppm Cl as DCB in methanol used.
Five hundred W maximum power generator operated at 400 W.
USN gas flow was set at 1.8 L/min and the membrane desolvator
countercurrent flow at 3 L/min.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
64
decrease can be attributed to lower analyte residence time and analyte dilution with
increases in flow rates.
Figure 24 depicts USN gas flow rate optimization at various plasma gas flows.
The USN gas flow was varied with plasma gas flow rates o f IS, 16, and 17 L/min.
The optimum S/B ratio was observed at a USN gas flow o f 1.8 L/min for all of the
plasma gas flow rates investigated. This optimum USN flow rate was the same as
that observed for the 100 W plasma. The membrane desolvator countercurrent flow
was set at 3 L/min.
Countercurrent gas flow optimization was also done for different plasma gas
flows. Results are shown in Figure 25. Again, the optimum rate (3 L/min) was found
to be the same as what was obtained for the lower power plasma.
The high-power plasma generator gave an opportunity to vary the plasma
power over a large range (170 - 520 W) and maintain a stable plasma. The effect of
plasma power on the Cl S/B ratio is shown in Figure 26. It was expected that the S/B
ratio will increase with increase in plasma power as more thermal energy is generated.
As per expectation, the Cl S/B ratio increased (53%) as the plasma power was
increased from 170 to 520 W. Detection limits for DCB and CB calculated under
these optimum conditions were 1.2 and 1 ppm respectively. These detection limits
were very similar to those obtained with the 100 W plasma (2 and 1.5 ppm for DCB
and CB respectively).
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
1
Figure 24.
1.5
2
USN gas flow (L/min)
2.5
3
USN gas flow vs Cl S/B. Five hundred ppm Cl as DCB used in
methanol. Plasma was operated at 400 W. Membrane desolvator
counter-current flow was set at 3 L/min.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
66
Plasma
gas
flow
rate
S/B
15
-v -
16
■17
O'
0
Figure 25.
1
2
Counter current gas (L/min)
3
Membrane desolvator countercurrent gas flow vs Cl signal. Five
hundred ppm Cl as DCB in methanol was used. Plasma operated
at 400 W. USN gas flow was set at 1.8 L/min.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
100
Figure 26.
200
300
400
Plasma power (W)
500
600
Plasma power vs Cl S/B. Five hundred ppm Cl as DCB in
methanol was used. Plasma gas was set at IS L/min, USN gas at
1.8 L/min, and membrane desolvator countercurrent flow at 3
L/min.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
2.8 Summary
The membrane desolvator operated efficiently for both aqueous and organic
solvent desolvation. Results o f the background studies and the similar detection
limits achieved with aqueous and organic solvents support this conclusion. Plasma
gas, carrier gas, membrane desolvator countercurrent flow rates, and plasma power
conditions were key factors in obtaining maximum nonmetal emission signal. For
volatile organic analytes, operation o f the membrane desolvator at a lower
temperature was critical. Enhancement o f the signal using a 500 W generator was
investigated. Detection limits for DCB and CB were not significantly enhanced with
the higher power plasmas. These results indicate that enhancement of analyte
throughput and transport to the plasma and even better desolvation of the analyte call
for further investigation. Some o f these aspects are discussed in Chapter 3.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
CHAPTER 3
INTERFACING HPLC WITH THE MIP AES USING USN AND MEMBRANE
DESOLVATION
3.1 Introduction
As the toxicity and mobility, and hence the environmental and biomedical
importance, of an element is strongly dependent on the chemical form in which it
occurs, there is a growing interest for elemental speciation rather than total elemental
analyses. Typically, speciation involves two complementary analytical techniques:
isolation o f various species or forms o f the analyte and detection o f those individual
species. For separation, various techniques are available. However, chromatographybased procedures like GC, LC, SFC, and capillary zone electrophoresis (CZE) are
effective and commonly used. For detection, plasma-based mass spectrometric and
atomic emission techniques have been successfully used in conjunction with
chromatography. Plasma-based detectors exhibit superior sensitivity compared to
other types of detectors commonly used with chromatography. Also, as discussed in
section 1.2, plasma-based spectrometries provide additional selectivity to the
separation process.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
70
Various workers have interfaced plasma AES and MS systems with
chromatographic techniques for speciation and detection. Kim et al.120,12' coupled a
GC with Ar ICPMS for determination o f alkyl lead compounds (tetraethyllead,
trimethylethyllead, dimethyldiethyllead, methyltriethyllead, and tetraethyllead) in fuel.
Various mixtures o f organomercury, organotin compounds,122and
metalloporphyrins123have also been separated and quantified using this technique.
MIPs have also been used with GC for nonmetal selective detection, and GC-MIP
AES systems are commercially available. Bradley and Carnahan124 have used such a
system for oxygen selective detection for petroleum samples. Haloorganics in drinking
water were assayed by Quimby et al.l2s Becker and Colmsjo126used GC-MIP-AES to
determine 34 tri-, tetra-, penta-, and hexacyclic aromatic sulfur heterocycles. ICPs
have also been used with LC for separation of ionic, polar, nonpolar, neutral
compounds with low volatility, and low thermal stability. Houk and colleagues127,12s
successfully separated phosphorous-containing compounds (ortho, pyro, and
tripolyphosphate), organolead (trimethyllead, triethyllead) and organomercury
(methylmercury, ethylmercury, and phenylmercury) mixtures using reverse-phase ion
pairing LC-ICPMS. Dauchy et al.129have used reverse-phase HPLC-ICP MS to
separate mixtures o f organotin (di, mono, and tributyltin). Camara and coworkers130
used a similar approach to separate four selenium species (selenocystine,
selenomethionine, selenite, and selanate). Ion exchange chromatography has also been
used with ICPs for speciation o f Chromium (Cr [HI] and Cr [VI], Vanadium [IV]131
and Vanadium (V)) and rare earth elements.132
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
71
In this chapter, characterization o f a reverse-phase HPLC - MIP AES system is
discussed. Akinbo and Carnahan36 used a similar system for nonmetal selective
detection o f organics with reasonable success. However, they experienced transportrelated problems with low-boiling-point organic analytes. Also, the USN failed to
nebulize certain LC solvent mixtures, which limited the mobile phase options. These
problems and their solutions are addressed in this discussion.
3J Experimental
The schematic o f the HPLC-MIP AES is shown in Figure 27. This set-up is
similar to that shown in Figure 5, except a Rainin Dynamax HPLC was used for
solvent introduction instead of a peristaltic pump. A peristaltic pump (not shown) was
used to drain the excess solvent from the USN aerosol chamber. The specifications
for the HPLC system are provided in section 2.3.2. The solvent from the HPLC was
directed to the USN using a 0.01 inch PEEK tubing. Aerosol exiting the USN
nebulization/desolvation system was then transported to the plasma via the membrane
desolvator. The monochromator was set to detect the wavelength o f interest. Data
were collected throughout the chromatographic run. Typically, 20 |xL of sample
were injected into the LC column. Chromatograms were obtained by monitoring the
element-selective emission as the separated analytes exited the column and passed
through the plasma.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
72
Analyte Flow
Gas Flow
• He
Microwave
Generator
*••••••
Carrier Gas
Figure 27.
System schematic for the HPLC-MIP AES.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
73
For the HPLC experiments, the areas under the peaks were calculated by
summation of the difference between the baseline and each point on the peak. Twenty
five points on each side o f the peaks were averaged to calculate the signal at baseline.
Noise was calculated by taking the standard deviation of the baseline values from three
or more replicate runs. The detection limit was calculated by dividing the amount of
analyte by the width o f the peak. The width at half the height of the peak was
multiplied by 2 to calculate the peak width.
33 Separation Using HPLC-MIP AES
For preliminary studies, 20 microliters o f a methanol solution containing 5000
ppm of Cl in methanol as DCB was injected into the column. One hundred percent
methanol was used as the mobile-phase. The mobile phase flow rate was set at 1
L/min. At this flow rate the pressure developed in the LC system wasl.5 kpsi. The
monochromator was set to detect the 479.5-nm Cl emission line. The chromatogram
is shown in Figure 28. The retention time for the peak was 1.9 min and the peak width
0.41 minutes. As shown in Figure 29, a Cl calibration plot was obtained by injecting
DCB solutions in the mass range o f 20 to 120 pg. Areas under the peaks were
calculated as described in section 3.2. The plot is linear with a r value o f0.995 and
the slope o f the plot was 655.9 counts/ jig. However, the plot showed a significant
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Figure 28.
Chromatogram of DCB. Twenty [iL of 5000 ppm Cl as DCB in
methanol used. One hundred percent methanol at a flow rate of 1
L/min was used as mobile phase.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
0
Figure 29.
20
40
60
80
100
Mass of Cl in micrograms
120
DCB calibration plot using HPLC-MIP AES. One hundred percent
methanol at a flow rate o f 1 L/min used as mobile phase. Injection
volume was 20 |iL.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
76
negative y-axis intercept (-17812 counts, which corresponded to 27.4 micrograms of
Cl). Similar negative y-axis intercept behavior was also observed from the Cl
calibration plot o f CB obtained under the same conditions. The calibration plot from
CB is shown in Figure 30. The r2 value for this plot was 0.991. The slope was
calculated to be 433.4 counts/ |lg. The intercept was -14184.72 counts (corresponds
to 33 micrograms o f Cl). The detection limits obtained for Cl obtained from DCB and
CB were 2.9 |ig (14S.3 ppm) and 1.5 fig (78.8 ppm) respectively. The corresponding
relative detection limits calculated for DCB and CB were 96.9 and 35 ng/s.
The problem o f the substantial negative y-axis intercept was believed to be
caused by one or both o f the following reasons. Firstly the baseline drops when the
analyte elutes, which results in an erroneous reading when the analytical signal is
subtracted from the baseline (signal due to the solvent). Secondly a constant amount
o f analyte is lost in the system. To investigate these issues specific tests were
designed.
3.3.1 Effects o f Anlyte on Background
The monochromator was set at 479.5 Cl emission line and a chromatogram
was obtained by injecting 20 }iL o f 1500 ppm DCB solution. The other conditions
described in the previous section were maintained. To observe the effect o f the
analyte on the base line, the monochromator was set at 479.6 nm, away from the Cl
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
77
40000-
30000 —
0000 -
10000
0
Figure 30.
20
40
60
80
Mass of Cl (micrograms)
100
120
CB calibration plot using HPLC-MIP AES. One hundred percent
methanol at a flow rate o f 1 L/min used as mobile phase. Injection
volume was 20 |iL.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
78
emission line, and another DCB chromatogram was obtained. Figure 31 shows an
overlay of the two chromatograms. The chromatogram, shown in black (a) was
obtained from DCB when the monochromator was set at the 479.5-nm Cl emission
line and one shown in red (b) was obtained when the monochromator was o ff the Cl
line. As seen in chromatogram (a), DCB eluted between 1.8 and 2.2 minutes. During
the same time interval no decline in signal was observed compared to the baseline
signal in chromatogram (b). A decrease in signal between 1.8 and 2.2 minutes when
the monochromator was off the Cl line would have suggested baseline shift when the
analyte elutes. Hence, the possibility o f analyte interference with the background
signal being the cause for negative y-axis intercept in the calibration plots can logically
be eliminated.
3.3.2 Investigation o f Analyte Loss in the System
To investigate possible analyte loss in the system, the HPLC was replaced with
the peristaltic pump for solvent introduction. A number o f Cl calibration plots were
obtained using SO to 400 ppm Cl as DCB solutions in methanol. The plot is shown in
Figure 32. The plot was linear with a correlation coefficient (r2) o f0.9993. However,
consistent with HPLC-MIP calibration plots, these also showed a significantly negative
y-axis intercept. Similar behavior was observed with CB calibration plots. This
behavior suggested analyte loss either at the USN or in the membrane desolvator.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
79
On Cl line (a)
_9000
<0
«#
c
3
*
2
!t£
€
Off Cl line (b)
§7000
5000
Time (min)
Figure 31.
(a) DCB chromatogram when monochromator is set for 479.5 Cl
emission line and (b) DCB chromatogram when the
monochromator is set at479.6 nm, off the Cl emission line. One
hundred percent methanol at flow rates of 1 L/min was used.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
80
Cl signal in arbitrary units
500000
400000
•
300000
•
200000
100000
o
•
0
Figure 32.
100
200
300
Concentration (ppm)
400
Cl calibration plot using DCB with direct solution nebulization.
Conditions for organic analyte listed in Table 2 were used.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
After the aerosol is generated in the aerosol chamber, it passes through a
heated “IT’ tube (maintained at a temperature of 140° C) to a condenser. It was
suspected that at high “U” tube temperatures some o f the analyte might pass into the
vapor form and remain in that form even after passing through the condenser and,
hence, be lost through the semipermeable membrane in the desolvator. A series o f
calibration plots were obtained at different USN “U” tube temperatures. For all these
measurements the membrane desolvator was maintained at S0°C. The temperatures
were varied between 50° and 100°C. At temperatures above 70°C the plots obtained
were very similar to the ones obtained at the “U” tube temperatures of 140°C (Figure
32). The plot showed a response o f zero at about 40 ppm at these temperatures.
Improvement in the calibration plot behavior was observed when the USN was
operated at “U” temperatures between 50° and 70°C, the calibration plots are shown in
Figure 33. The r2values were 0.991 and 0.9995 for the plots at 50° and 70°C
respectively. Understandably, due to better aerosol transport at higher temperatures,
the sensitivities observed during operation at 70°C were about 3 times higher than
what was observed at 50°C. The plots showed a response o f zero at 26 and 18 ppm
at 50° and 70°C, respectively. Although not significant, this was an improvement over
what was observed at a “U” tube temperature of 140°C. However, operation with
“U” tube temperatures below 100°C resulted in severe condensation problems. After
about 20 minutes o f operation, the connecting tube between the USN and the
membrane desolvator was filled with condensed solvent, which resulted in plasma
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
82
Signal intensity (arbitrary units)
120000
100000
SO degree Celsius
80000
70 degree Celsius
60000
40000
20000
0
0
Figure 33.
50
100
150
200
250
Concentration in ppm
300
350
400
Cl calibration plot using DCB at reduced USN “U” tube
temperature. Solvent was introduced by direct solution
nebulization. Other conditions for organic analytes listed in Table 2
were used.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
flicker and poor measurement reproducibility. Hence, operation at lower “U” tube
temperatures was not a plausible solution to the nonideal calibration problem.
Akinbo and Carnahan3*noted similar behavior for organic analytes and utilized
0.125 M NaOH in methanol to enhance analyte transport and linearize the calibration
plot to a near-zero intercept. As discussed in section 2.S.3, the excess NaOH was
believed to encapsulate the analyte within desolvated NaOH particulates to avoid
volatilization as the analyte passed through the heated membrane desolvator.
However, large quantities of NaOH led to other problems such as significant
background emissions and reduced plasma torch lifetimes.
To enhance analyte transport, but avoid NaOH complications, solutions of a
number of other modifiers were characterized. Solutions containing 1% HN03,
CH3COOH, and H2S04 in methanol were examined and DCB calibration plots were
obtained. Results are shown on Figure 34. As with the pure methanol solvent, a
linear calibration plot was not obtained with the addition o f 1% HN03or CH3COOH.
However, the addition o f 1% H2S04 produced a linear plot with a correlation
coefficient o f0.9992 and an x-intercept corresponding to 1.3 ppm. The enhanced
calibration behavior with H2S04can be explained by improved analyte transport.
Sulfuric acid protonates the amine group of DCB to form less volatile species.
Additionally, low vapor pressure o f H,S04depresses the volatility o f the analyte.
Nitric and acetic acids have significantly higher vapor pressures. The minimum H2S04
concentration necessary was determined by preparing calibration plots with varying
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
84
1400001—
1
1200004tI
-T 1% H2S04
-
S joooo-^
3
e*
4-
a
^ooooi
7
•-
1%HN03
1% Acetic Acid
$0000+
SC
it
|oooo35
I
20000+
▼
▼
50
Figure 34.
100
150
200
250
300
Concentration (ppm)
350
400
450
Cl calibration plot from 2 ,6-dichlorobenzamide dissolved in
methanolic 1% H2S 04, 1% CH3COOH, and 1% HN03. Other
conditions are described in Table 2.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
85
amounts o f H2S04; 0.01% H2S 04 was found to be the minimum necessary
concentration to maintain the behavior. Calibration plots o f CB in 0.01% H2S04 also
yielded near-zero calibration intercepts. Additionally, calibration plots o f DCB in
various methanofrwater mixtures (80:20,70:30,60:40 [v/v]) with 0.01% H2S 04
exhibited near-zero intercepts.
Temperature optimization o f the membrane desolvator was performed for the
modified solvent. The membrane desolvator temperature was varied between 50° and
110°C and the Cl S/B was observed from 500 ppm DCB in 0.1 and 0.01% H2S04 in
methanol solutions. The results are shown in Figure 35. The S/B at 50° C was not
significantly different from that seen at 70°C. At these temperatures the S/B from the
0.01% H2S04-modifred solution was 33% higher than that o f the 0.1% H2S04modified solution. That 0.01% H,S04 in methanol is less viscous compared to the
0.1% H2S04 solution, leading to better nebulization, could be the reason for the
enhanced S/B with the dilute H2S04. Since there was no noticeable difference in the
signal between 50° and 70°C operation, the previously optimized membrane desolvator
temperature of 50°C was used for future experiments. Detection limits were obtained
with 0.01% H2S04 in methanol and the optimized plasma parameters are listed in the
bottom row o f Table 2. Detection limits as CB and DCB were found to be 1.0 and 1.1
ppm Cl, respectively. These values were not very different from those obtained from
Cl solutions with pure methanol (see section 2.7) and are comparable to those
obtained by Akinbo and Carnahan36 using a flat-sheet membrane desolvator.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
SO
60
70
80
90
Temperature in degrees Celsius
100
110
0.01% H2S04 -m - 0.1% H2S04
Figure 35.
Cl S/B from 500 ppm of DCB in 0.01 and 0.1% H,S04 in methanol
solution. Solvent was introduced by direct solution nebulization.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
87
3.4 Nebulization of LC Solvents with USN
Following attainment of calibration plot linearity, the peristaltic pump was
replaced by the HPLC and the separation of a mixture of DCB and CB was attempted.
The HPLC-MIP AES system was set up as described in section 3.3. Equal portions of
5000-ppm Cl solution as DCB and 5000-ppm Cl solution as CB were mixed. The
sample solvent was HPLC-grade methanol. Twenty |XL samples were injected into the
HPLC. One hundred percent methanol at flow rates o f 0.8 mL/min was used as the
mobile phase.
The chromatogram obtained is shown in Figure 36a. The retention times for
DCB and CB were 3.1 and 3.3 minutes, respectively. The peak widths were 30 s for
DCB and 45 s for CB. The detection limits calculated for the two compounds were
2.9 }ig (150 ppm) for DCB and 1.5 jig (79 ppm) for CB. With 100% methanol as the
mobile phase, complete baseline resolution could not be obtained. To achieve
complete separation o f the two peaks, a 70:30 mixture o f methanolrwater was used.
The resultant chromatogram is shown in Figure 36 b. However, the peaks were much
less intense compared to what was obtained with 100% methanol as mobile phase.
The detection limits obtained were 646 and 540 (ig for DCB and CB respectively, two
orders o f magnitude higher than those obtained with pure methanol as mobile phase.
The loss in signal intensity was due to inability of the USN to effectively nebulize the
70% methanol mobile phase. The decline in USN efficiency for this was noticeable by
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
88
40000
30000
20000
10000
•10000
4
Tme(min)
5
40000
30000
jg 20000
9
1
10000
3
o
0
-10000
0
1
2
3
4
5
6
7
Time((nin)
b
Figure 36.
(a) Chromatogram o f DCB and CB using 100 % methanol as
mobile phase and ( b) separation of DCB and CB using 70%
methanol as mobile phase. Flow rate was 0.8 mL/min.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
89
visual inspection of the mist within the aerosol chamber.
Akinbo and Carnahan36had experienced similar problems. The USN (CETAC
U500AT with automatic transducer) failed to nebulize certain water, methanol, and
acetonitrile binary and ternary mixtures that are used as HPLC mobile phases. This
problem severely impaired the ability to perform determinations requiring certain
solvent mixtures for separation.
The USN transducer consists o f a LiNb03 piezoelectric ceramic crystal. When
the oscillating voltage (at a frequency typically near 1 MHz) is applied, the crystal
alternately expands and contracts. This oscillation of the crystal induces a standing
wave on the crystal. This wave is transmitted to any liquid film that comes in contact
with the crystal. Systematic instabilities in the liquid surface cause it to break up into
fine aerosol. Most current commercial USNs are supplied with automatic transducer
controllers. These automatic controllers typically monitor the transducer at a fixed
frequency (1.4 MHz in the CETAC U50t vT USN) but automatically adjust the input
power for efficient nebulization of various liquids. Transducer operation at a fixed
frequency o f 1.4 MHz is effective for nebulization o f aqueous and pure organic
solutions. However, difficulties arise with mixed solution nebulization, as experienced
by Akinbo and Carnahan.36 This problem is illustrated in Table 3, as based upon visual
inspection o f the aerosol density in the aerosol chamber. As noted in the “automatic
transducer” column, 100% aqueous, methanol, acetonitrile, and isopropyl alcohol
solutions could be nebulized efficiently. Ninety and 80% solutions o f the organic
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
*
Table 3.
USN Nebulization Efficiency o f Binary Solvent Mixtures (v/v) with
Automatic and Manual Transducer Controllers.
Water:Methanol
Water:Acetonitrile
Water:Isopropyl alcohol
Automatic transducer
controller
(Fixed frequency)
Manual transducer
controller
100:0
Nebulized
Nebulized
90:10
Not nebulized
Nebulized
80:20
Not nebulized
Nebulized
70:30
Not nebulized
Nebulized
60:40
Not nebulized
Nebulized
50:50
Not nebulized
Nebulized
40:60
Not nebulized
Nebulized
30:70
Inefficient nebulization
Nebulized
20:80
Nebulized
Nebulized
90:10
Nebulized
Nebulized
0:100
Nebulized
Nebulized
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
solvents in water were also effectively nebulized. However, the other solvent mixtures
were either not nebulized at all or were poorly nebulized.
The automatic transducer controller o f the USN did not allow the transducer
frequency to be changed. However, a CETAC U5000 USN with manual frequency
and input power control was available and utilized. Various methanol, acetonitrile,
and isopropyl alcohol solutions in water were introduced to this USN and the
frequency was manually adjusted until aerosol formation was visibly noticed. By
adjusting the transducer frequency, all the solutions listed in Table 3 could be
nebulized. A frequency counter (Hewlett Packard, model S38S A) was connected to
the transducer power supply in parallel to monitor the frequency at which each solvent
was efficiently nebulized. The observations are listed in Table 4. The solvents that
could not be nebulized by the automatic transducer (10 to 70% methanol) required a
higher frequency (1.39 MHz) than the solvents that could be nebulized (1.33 MHz).
The frequency supplied to the transducer by the automatic controller was measured to
be 1.33 MHz, indicating that the frequency supplied by the automatic controller is not
appropriate for nebulization o f the 10 to 70% methanol mixtures. Thus, to accomplish
aerosol formation from all the solvent mixtures, the transducer frequency was set at
1.39 MHz with the help o f the manual controller. Cl S/B were measured as 500-ppm
Cl as DCB in 20 to 100% methanol solutions were introduced to the USN. The
results are shown in Figure 37. The Cl S/B gradually increased with increasing
amounts o f methanol in the solvent, even though the background emissions were
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
92
Table 4.
Nebulization Frequency o f Water:Metiianol and WatenAcetonitrile
Mixtures Using USN with Manual Transducer.
Water:methanol
Watenacetonitrile
Manual transducer
frequency (MHz)
100:0
1.33
90:10
1.39
80:20
1.39
70:30
1.39
60:40
1.39
50:50
1.39
40:60
1.39
30:70
1.38
20:80
1.33
90:10
1.33
0:100
1.33
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
20
Figure 37.
40
60
Percent MeOH
80
100
Plot of percent methanol in the solvent versus Cl S/B from
solutions of500-ppm Cl as DCB solution. Direct solution
nebulization was used for solvent introduction.
«
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
94
higher with the organic solvents. This shows superior nebulization o f the higher
methanol-containing solutions even after transducer frequency optimization.
However, approximately two times higher S/B with pure methanol compared to 20%
Cl solution translated to only a 2 ppm improvement (1 ppm versus 3 ppm) in the
detection limits of the two solutions. However, the capability to nebulize all mobilephase solvents provided an opportunity for a large variety of HPLC applications.
3.5 Assessment of Fixed-Frequency Transducer Behavior
To further investigate the behavior of the fixed-frequency transducer, physical
properties o f solutions such as the surface tension, density, and viscosity were
analyzed. Values for these parameters were obtained from the CRC Handbook of
Chemistry and Physics.133,134
Figure 38 is a plot o f density (kg/L) for the full range of water/methanol (v/v)
mixtures.133 The density was calculated by using equation 3.1.
(Cs + Cw)
1000
(3.1)
where D = density at 20°C (kg/L), Cs - anhydrous solute concentration (g/L), and Cw
= total water concentration (g/L). The density gradually decreases from 1 kg/L for
pure water to 0.8 kg/L for pure methanol. No unusual density profile for the 10 to
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
0
20
40
60
80
100
Volume % of MeOH
Figure 38.
Plot o f percent methanol versus density with respect to water. All
density values were taken from Weast and Astle.133
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
96
70% solutions o f methanol is observed, which would have indicated that density could
be a contributing factor for the inability o f the fixed-frequency USN to nebulize these
solvent mixtures. Hence, no correlation between the density and the USN’s peculiar
behavior can be drawn from the density plot o f water/methanol mixtures. A plot of
surface tension134(dynes/cm) versus percent methanol is shown in Figure 39. The
surface tension gradually decreases from 72.25 for pure water to 22.65 dynes/cm for
pure methanol. Again, like the density plot, no extraordinary surface tension profiles
for the unnebulizable solvent mixtures were observed.
Conversely, the viscosity133profile showed more interesting characteristics.
Figure 40 is a plot o f relative viscosity versus percent methanol in water. Relative
viscosity is calculated as the ratio o f the absolute viscosity of a solution at 20“C to the
absolute viscosity of water at 20°C. The relative viscosity increases from l(for pure
water) to 1.8 (for 40% methanol) and then gradually decreases to 0.6 for pure
methanol. The line perpendicular to the y-axis that extends from 10 to 70% methanol
shows that all the solvent mixtures with a relative viscosity o f 1.3 or greater were not
nebulized by the fixed-frequency USN. Hence, viscosity appears to be one o f the
contributing factors for the USN behavior. The viscosity profile o f isopropyl alcohol
(IPA):water mixtures was also examined. The plot is shown in Figure 41. Again, the
viscosity curve rises to a maximum and then decreases. The relative viscosity
increases from 1 (pure water) to 3.2 for 50% 1PA and then decreases to a relative
viscosity value o f 2.2 for 100% IPA. The solutions above the straight line connecting
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
0
Figure 39.
20
40
60
Volume % MeOH
80
100
Plot o f percent methanol versus corresponding surface tensions.
All surface tension values were taken from Weast and Astle.m
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
0
Figure 40.
20
40
60
80
Volume % of MeOH in water
100
Plot o f percent methanol versus corresponding viscosity relative to
water. All viscosity values were taken from Weast and Astle.133
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
0
Figure 41.
20
40
60
Percent IPA
80
100
Plot of percent IPA versus corresponding viscosity with respect to
water. All viscosity values were taken from Weast and Astle.133
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
100
10 and 70% IPA solution in Figure 41 could not be nebulized (refer to Table 3). This
observation is consistent with the H20:methanoI studies. The higher viscosity
solutions required alternate nebulization conditions. However, while the trend is
consistent for both sets o f solutions, the viscosity “cut-offs” vary. The viscosity trend
correlates well with the behavior, but the absolute viscosity does not. Obviously,
another factor is needed for a complete explanation.
3.6 HPLC-MIP AES Using Solvent Modifiers
A number of solvent compositions were examined for the separation o f DCB
and CB. Baseline resolution was achieved with a 60:40 (v/v) mixture of methanol and
water. The 0.SS mL/minute chromatographic eluate was post-column mixed with 0.55
mL/minute of 0.1% H2S04 in methanol and directed to the USN. A schematic of the
HPLC-MIP AES is shown in Figure 42. The plasma gas, the carrier gas, and countercurrent gas flow rates were those listed in Table 1 for organic analytes. Figure 43
shows the two Cl peaks obtained from the separation of a mixture containing 30 |ig
each of DCB and CB.
Calibration plots were obtained (Figure 44). The correlation coefficient
calculated for DCB was 0.9996 and 0.997 with CB. The relative standard deviations
(RSD) calculated for Cl with the system were between 5 and 8% for both DCB and
CB. The sensitivities were found to be 1227 counts/jig for DCB and 1045 counts/jlg
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
101
Analyte Flow
Gas Flow
.•*’ **• G ultrGrantQb
»•
DUe
|
tUnnac
N h te
•
a
a
a
a
a
a
a
• a.*i
*•••••*
O nirQ i
1
1
Modnk
Curator
Vf t ;
H bbO b
Figure 42.
System schematic o f the HPLC- MIP AES showing post column
addition of dilute H,S04.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
102
7000 —
6000
i
DCB
1
5000
i
£
§1000 !
1
Sooo
Q>
CO
2000
1000
0 —
0
Figure 43.
1
2
3
4
5
6
7
8
Time (min)
9
10
11
12
13
Chromatogram showing Cl peaks from 30 |ig 2,6dichlorobenzamide and 30 |ig 4-chlorobenzamide. Sixty:forty (v/v)
methanokwater was used as the mobile phase. The mobile-phase
flow rate was 0.55 mL/min. Mobile-phase modifier, 0.1% H,S04,
was pumped at a rate o f 0.5 mL/min. Other conditions are as
described in Table 2.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
35000
30000
&
2.6 DCB
1 25000
£*
2
|
a
4CB
20000
Z 15000
2
sa. 10000
•*
5000
2
Figure 44.
4
6
8 10 12 14 16 18
Mass of Cl in micrograms
20
22
24
26
Calibration plots from 2,6-dichlorobenzamide and
4-chlorobenzamide. Sixtyrforty (v/v) methanol:water was used as
the mobile phase. The mobile-phase flow rate was 0.5S ml/min.
Mobile-phase modifier, 0.1% H2S04, was pumped at rate o f 0.5
mL/min. Other conditions are as described in Table 2.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
104
for CB. It is likely that the 14.8% higher sensitivity for DCB was due to its higher
boiling point and reduced diffusion losses through the desolvator membrane, compared
to CB. Chlorine detection limits for DCB and CB were 110 ppm (36 ng/s) and 127
ppm (47 ng/s), respectively. These mass-flow detection limits were about 4 to 5 times
higher than those obtained with direct solution nebulization (10 ng/s) described in this
manuscript. The major contributions to increase in detection limits with HPLC are onand off-column dilutions as well as dilution due to post column addition o f H2S04
Using membrane desolvation with NaOH os the transport salt, Cl detection limits
obtained by Akinbo and Carnahan36 were between 74 to 200 ng/s, which are
comparable to detection limits presented in this study. However, detection limits
reported with LC-He-MIP using the moving-band interface were lower by about 2
orders o f magnitude (between 0.22 and 0.77 ng/s).3S With a moving-band interface
almost 100% of the analyte is transported to the plasma, whereas a much smaller
fraction is transported with the ultrasonic nebulizer.36 Additionally, superior
desolvation o f the analyte is achieved with the moving-band interface. However, the
membrane desolvation system provides better performance in terms o f ease of
operation and reproducibility.
3.7 Summary
Linearization o f the calibration plot was achieved using H2S04 as the solvent
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
105
modifier. Detection limits and precision obtained using this system were comparable
and in some cases better than those obtained by Akinbo and Carnahan.36 All HPLC
solvent combinations could be nebulized through the USN by using a manual
transducer control. This will provide unlimited mobile-phase choices for HPLC
separations. However, improvement is necessary to obtain detection limits
comparable to those obtained from GC-MIP systems. Enhancement o f analyte
throughput by utilizing other types o f nebulizers is one area that needs further
investigation. A detailed scheme for enhancement o f analyte throughput to the plasma
is provided in Chapter 5.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
CHAPTER 4
HPLC-MIP AES APPLICATIONS
4.1 Introduction
As discussed in Chapters 2 and 3, the MIP AES system combined with
appropriate optimized conditions for membrane desolvation, mixed solvent
nebulization, and solvent modifiers performed satisfactorily with chlorinated organic
compounds. With these positive results for the model chlorinated compounds, it is
appropriate to examine the potential of the system for a breadth o f sample types. In
this chapter, investigations are initiated with studies of other elemental analytes and
“real-world” type samples.
Emissions from some o f the other nonmetals (Br, I, S, and P) are detailed.
Comparisons with MIP AES and UV-Vis detectors for HPLC applications are
discussed. The performance o f the HPLC-MIP AES with biological and
pharmaceutical samples is demonstrated.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
107
4 2 Examination of Br, I, S, and P Emission
As discussed in Chapter 2, Cl line emission signals were used to optimize the
MIP AES system. Line emission from other nonmetals (Br, I, S, and P) was also
examined using the optimized conditions listed in Table 2. One part per thousand
(ppt) aqueous solutions of Br, I, S, and P as inorganic salts were introduced to the
plasma and the corresponding emission signals were observed. Figure 45 is a Br
spectrum of the 468 to 483 nm region using a solution of KBr03. Three distinct Br
ion lines were observed at 470.7,478.8, and 481.9 nm. The peak at 471.4 nm is
emission from helium and the molecular band at 473.7 is from C2. Characteristic
iodine peaks from the KI03-containing aqueous solution are shown in Figure 46.
Three iodine ion lines were seen at 516.3,534.0, and 534.7 nm. Figure 47 is a sulfur
spectrum from the (NH4)2S04-containing solution. Sulfur ion emission lines at 543.1,
543.5,545.6, and 547.6 nm are seen. Figure 48 and 49 show phosphorus atom
emission lines in the spectral regions o f 231 to 216 and 252 to 257 nm. A H3P04containing solution was used to acquire the P spectra. Characteristic P lines were
observed at 213.7,213.8,215.1,215.5,215.6,253.6,253.7,255.5, and 255.7 nm.
Table 5 shows the detection limits obtained for Br, I, S, and P using their most
intense emission lines. Br, I, and S detection limits were between land 3 ppm. These
limits of detection were similar to that observed for Cl. However, a P detection limit
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
108
100
, -----------------------------
470.7
80 >
478.8
I
60
(0
§ 40
o
i
481.9
Z
20
468
Figure 45.
470
472
474
476
478
Wavelength (nm)
480
482
484
One ppt Br, MIP AES spectrum from aqueous KBr03.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
510
Figure 46.
515
520
525
Wavelength (nm)
530
535
One ppt I, MIP AES spectrum from aqueous KI03.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
110
100
545.6
Normalized signal
80
60
40
1543.5
543.1
20
•20
542
Figure 47.
547.6
543
544
545
Wavelength (nm)
546
547
One ppt S, MIP AES spectrum from aqueous (NH^SO.,.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
213.72 ,: 1
215.48
:I ;
.215.60
0 213
Figure 48.
213.5
214
214.5
215
Wavelength (nm)
215.5
216
One ppt P, MIP AES spectrum (213-216 nm) from aqueous H3P04
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
255.68
20
253.58 1 ;1
I
0 252
Figure 49.
253
254
255
Wavelength (nm)
256
257
One ppt P, MIP AES spectrum (252-257 nm) from aqueous H3P 04.
Reproduced with permission o f the copyright owner. Further reproduction prohibited without permission.
113
Table 5.
Nonmetal Detection Limits Using MIP AES
Nonmetal
Analyte used
Wavelength (nm)
Detection Limit (ppm)
Br (II)
KBrOj
470.7
2
1(H)
KIOj
516.3
3
S(H)
(NH4),_S04
545.6
2
P(I)
h 3po 4
253.7
0.02
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
114
o f 20 ppb was observed. The lower detection limits obtained with P provided an
opportunity to perform some interesting LC-MIP AES applications using P-containing
compounds.
43 Sensitivity of MIP-AES Detectors
As discussed in section 1.2, one o f the attractive traits of MIP AES detectors is
its sensitivity to a myriad o f compounds, even when present in complex matrices. This
section compares the sensitivity o f the MIP AES to the commonly used UV-Vis
spectrophotometry for LC detection. The separation of a mixture of (3glycerophosphate (GP) and triphenylphosphene (TPP) is demonstrated. These
samples were obtained from Aldrich (Milwaukee, WI). Twenty |XL of a methanol
mixture containing 500 ppm P as GP and TPP in methanol was injected into the LC
column. The monochromator was set at the 253.7 nm P emission line. Eighty:twenty
methanokwater (v/v) was used as the mobile phase. The flow rate was maintained at
0.8 mL/min. Using a peristaltic pump, 0.1% H2S04 was post-column added to the
mobile phase through a “T” junction at a flow rate o f 0.8 mL/min. The resultant
0.05% H2S04 post-column solution was directed to the USN using a 0.01-inch PEEK
tube. The plasma, USN, and the membrane desolvator conditions were set in
accordance with the optimized conditions listed in Table 2.
The resultant chromatogram is shown in Figure 50. GP and TPP eluted at 1.6
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
115
70000 , -----------------------------------------60000
j
[
'
I
\
f 50000 I
3
fc*
I
240000 ;
'
.«
p o o o o
i
s
120000
_
I
■
*
10000
0
0
1
2
3
4
5
---------6
Time (min)
Figure 50.
LC-MIP AES chromatogram o f P-glycerophosphate (500 ppm P) and triphenyl
phosphene (500 ppm P). Twenty |iL injections were made. Retention times were
1.6 and 2.8 minutes for GP and TPP respectively. Eighty:twenty mixture o f water
and methanol at a flow rate o f 0.8 mL/min was used. The P line at 253.7 nm was
monitored.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
116
and 2.8 minutes respectively. The peak width at half height for GP was 6s and that for
TPP was 9s. The absolute detection limits were calculated to be 38 ng (1.9 ppm) and
30 ng (1.5 ppm) for GP and TPP, respectively. Relative detection limits calculated by
dividing the absolute detection limits by twice the peak width at half height were 6.3
and 3.3 ng/s for GP and TPP, respectively. Calibration plots for both phosphorus
containing compounds were obtained using 200 - 600 ng P solution mixtures (Figure
51). The calibration plots were linear with lv alu es ofO.9994 for GP and 0.9998 for
TPP. The plots showed near-zero intercepts. The sensitivities for GP and TPP were
85.9 and 153.8 counts/ng respectively.
A chromatogram for this mixture was also obtained using a UV-Vis detector
and is shown in Figure 52. The detector was set at 254 nm. Only one o f the two
compounds (TPP) is detected. Glycerophosphate does not contain any strong UV
absorbing chromophores; hence, it is not detected by the UV-Vis detector. A
glycerophosphate spectrum from 200 to 400 nm taken using a UV-Vis spectrometer
revealed very weak absorbance in the UV region. Usually, laborious and timeconsuming derivatization is necessary to make UV-inactive eluates compatible for UVVis detection. However, as shown, appropriate element selective detection with
plasma AES detection may allow many compounds to be easily and reliably detected.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
117
7000---------------------------------
Peak area (arbitrary units)
6000—
■
GP
•
TPP
5000I
4000
3000T
2000 —
1000
0
------------0
100
200
300
400
500
600
Mass of P (nanograms)
Figure 51.
LC-MIP AES calibration plots for GP and TPP. The r values for
the plots were 0.9994 and 0.9998 for GP and TPP respectively.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
0
1
2
3
4
5
6
Time (min)
Figure 52.
Chromatogram of GP (500 ppm P) and TPP (500 ppm P) using the
UV-Vis detector. Only the peak from TPP is observed (retention time 2.2
minutes). GP is UV inactive, hence is not detected by the detector.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
119
4.4 HPLC-MIP AES Applications
4.4.1 Detection ofDNA
High molecular weight compounds such as proteins, peptides, and nucleic
acids pose significant analysis difficulties for mass spectrometrists. Most mass
spectrometers do not possess high resolving power at higher mass ranges. With
plasma AES, the potential to obtain total elemental content information exists. The
performance o f the plasma system with a high molecular weight DNA sample from
salmon testes was examined.
The primary structure of a nucleic acid is a sequence of nucleotide residues
connected by 3-5' phosphodiester linkages. A tetranucleotide representing a segment
of single-stranded DNA is shown in Figure 53.I3S A DNA double helix is formed from
two antiparallel polynucleotide strands as shown in Figure S4.l3s Bases on opposite
strands are connected through hydrogen bonding forming a series o f base pairs.
Commonly, the length o f long DNA structures is measured in thousands of base pairs
or kilobase pairs (kb).
For detection o f the sample containing salmon DNA (Sigma Chemical Co., S t
Louis, MO) with MIP AES, the presence o f P atoms in the sugar-phosphate backbone
was utilized. A 500-ppm sample was prepared by dissolving 5 mg o f the DNA sample
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
120
5 ' end
'0—P=0
NH
i
Adenine (A)
= CH
— P=0
Guanine (G)
—p = o
>CH
SH
O—P=0
OH
H
3 'end
Figure 53.
Structure of a tetranucleotide. The
nucleotide residues are connected by
3-5' phosphodiester linkage. Figure
taken from Moran etal.135
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
121
Figure 54.
Diagram o f a double-stranded DNA.
Figure taken from Moran etal.l3s
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
122
in 10 mL o f tris-ethylene diamine tetraacetate ( EDTA) (10 mM tris +1 mM EDTA,
pH 8.0) buffer. Twenty |iL o f this solution was injected into the C „ HPLC column
and the 253.7 nm P emission line was observed. The mobile phase used was 80:20
methanokwater (v/v) at a flow rate o f 0.8 mL/min. The chromatogram obtained is
shown in Figure 55. The retention time for the peak was 1.6 minutes. Despite the
presence o f a large number o f carbon atoms in DNA samples, the plasma behaved
ideally. No plasma flicker or extinguishment was observed.
To determine the length o f the DNA strands present in the sample, gel
electrophoresis was performed. The agarose (SeaKem Leagarose, FMC Bioproducts)
and the buffer (tris-borate electrophoretic buffer, pH 8.0) solution was heated in a
125-ml Erlenmeyer flask until the agarose dissolved. The solution was then cooled to
60°C and 2 }lL o f 10-ppm ethidium bromide in water solution was added. The gel was
immediately poured into a mold. A 12-tooth comb was placed on the mold and the
solution was allowed to cool for an hour until a gel was formed. The gel contained
0.9% (w/v) of agarose.
Molecular weight standards were prepared by adding 1 (XL of Promega, lkb
DNA ladder with 1 }iL o f Promega loading dye and 4 pL o f water.136 The samples
were prepared by adding 1,2, and 3 [XL of a 5-ppm solution o f DNA to 1 |lL o f the
loading dye. Water was added to make the total volume 6 |iL . The comb was
removed from the gel and the framework from the gel mold was set in the
electrophoresis tank. The tris-borate buffer was added to the tank until the gel was
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
123
9000 ,
Signal (arbitrary units)
8000 ;
7000 |
i
; i
6000 !
5000 •
4000
3000
2000
2
3
Time (min)
Figure 55.
LC-MIP AES chromatogram obtained from 500 ppm DNA,
extracted from salmon testes. Injection volume was 20 |iL.
Mobile phase used was 80:20 mixture o f methanohwater at a flow
rate of 0.8 mL/min. P emission line at 253.7 nm was monitored.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
124
completely submerged. Using a micropipette, 4 }lL o f each sample and the standard
were added to the wells in the gel. Eighty volts were applied across the 8 cm gel for
80 minutes, using a Bio-Rad (model 500/200) power supply. Once the separation
was complete, a photograph of the gel was taken under UV light. The photograph is
shown in Figure 56.
Slots 2,3, and 12 were used for the standards. Sample was loaded into slots 5,
9, and 10. The samples and the standards migrated from left to right under the
influence of the applied voltage. The molecular weight standard used was a 14fragment DNA ladder, ranging from 0.25 tolO kb. Clearly separated bands across
slots 2 and 3 are seen in Figure 56. The bands seen closest to the slots 2 and 3 are the
10-kb fragments and the faint bands seen farthest away from the slots are the 0.25-kb
fragments. All the other fragments fall between the 10- and the 0.25-kb bands. The
sample bands across slots 5,9, and 10 are not well defined. The smear of bands
suggest the sample contains DNA strands o f various lengths. A heavy concentration is
seen closer to the slots, indicating significant presence o f greater than 10 kb pair
fragments.
4.4.2 Detection o f Vitamin Bn
Chromatographic detection schemes often lack selectivity, which may result in
overlapping chromatographic peaks. Often, separation and quantification o f only
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Figure 56.
UV picture o f electrophoretic bands from the molecular
weight standards (in slots 2,3, and 12) and the DNA sample
from salmon testes (in slots 5,7,9,10) on the agarose gel.
The direction of migration o f the bands is from left to right.
The lower molecular weight fractions move further to the
right compared to the high molecular weight fractions. No
band separation is observed with the samples.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
126
selected species in a mixture are necessary. However, extraneous components may
overlap with the chromatographic peaks o f the analyte. An example where one such
complication occurs is the analysis of vitamin B,2 in multivitamin tablets.
Multivitamins usually contain vitamins Bt, B6, BI2, and excipients. Since the human
body needs only 1 to 2 fig of BI2 aday, B,2 is typically present in ratios betweenlrlOO
to 1:1000, with respect to other vitamins in these formulations.
The most common chromatographic technique used for these separations is
reverse-phase HPLC.137 In isocratic modes, the peaks are not well separated.
Gradient elution methods are time consuming and may be less reproducible. In
.
situations such as this, a detector that can selectively detect the analyte o f interest is
desirable. The MIP AES system was examined for the separation o f related mixtures.
Of the three vitamins under discussion, only vitamin BI2 contains a phosphate
group. The reader is referred to Figure 57. Phosphorus selective detection by the
MIP AES detector was exploited. Chromatograms were obtained using both UV-Vis
and MIP AES. The solvent after exiting the UV detection system was directed to the
MIP AES. UV detection was done at 361 nm and the 253.7 nm P line was monitored
with MIP AES.
Figure 58 shows vitamin BI2 chromatograms when 20 |iL o f 886 ppm vitamin
B12 (20 ppm P) in a 50:50 mixture o f methanokwater. The mobile phase was a 1%
acetic acid solution in methanol at a flow rate o f I mL/min. Since the solvent goes
through the UV detector to the AES detector, the retention time with the latter is
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
127
NHj
«m
yCHg *cr
y _
HO.
CH90H
CH2OH
• HCI
IJ
V ^C W H
CHjV
HCI
CH,V
Cl^ - C -N H ,
C % C H ,- C - W ,
H^-C-Ofe
0 CH,
HjM-C-CHOfc1
tyi-c-c*
4
II
0
CH,
0
CH, C * C H ,- C - N *
p r t)
I
CH,
CH ,
„
I
CH
I
H
p
CI^OH
CH,
Figure 57.
Molecular structure of (a) vitamin B„ (b) vitamin B6and
(c) vitamin B,z. Figures were taken from Moran et al.135
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
-1000
--------------------------------------------------------------------------------0
1
2
3
4
5
Time(min)
b
Figure 58.
Chromatograms obtained from vitamin B,2 (10 ppm P)
using (a) UV-Vis detector and (b) MIP AES.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
129
about 40 s longer. The P detection limit calculated with the MIP AES was 1.2 ppm
(53 ppm vitamin B12).
Under the same conditions, 20 |lL of a mixture containing 10 ppm vitamin B,
and B6 and 20 ppm P as vitamin Bl2 in 50:50 water:methanol was injected. The
chromatograms obtained with the two detectors are shown in Figure 59. With the UV
detector the three co-eluting peaks are not resolved. The elution sequence was
vitamins B„ BI2, and B6by increasing order of retention. However, with the MIP AES
detector, only the P-containing vitamin B,2 is observed. Thus, the high selectivity of
the MIP AES detector is demonstrated.
From the absorption bands shown for the three vitamins in Figure 60, it is
clearly seen that selective detection o f vitamin Bl2 is not possible with UV-Vis
detection. Unlike the atomic emission lines, the absorption bands are broad and not
conducive for selective detection. However, for the MIP AES detector to be
applicable for the analysis of real multivitamin tablets (typically a 500-mg tablet
contains 100 |lg o f vitamin BI2), at least a 10-fold improvement in P detection limits is
essential. With the current P detection limits, a large number o f tablets must be
dissolved in small amounts of solvent for the resultant B,2 analyte concentration to be
detectable. As an example, a solution must contain approximately 25% multivitamin
tablets by weight for the phosphorous concentration from vitamin B,2 to be at the
detection limit.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
130
35
B6
A
! \
30
i \
£25
s
Bi
£ 20
B 121
h
A
i '
!
I 15
1 10
I
1
:
’■
!
;
;
i
5
i
0
---------2
3
Time(min)
5000
-----------------------------------------------------------------
I
4000 1
I
3000
|!
2000
1
3
£*
a
i
s.
i.9
10
1000
.1000
i.fc
—
--------------------------------------------------------------------------------
1
2
3
Time (min)
b
Figure 59.
Chromatogram from a mixture of vitamin B, (10 ppm),
vitamin B6(10 ppm), and vitamin B12(10 ppm P) using (a)
UV-Vis detector and (b) MIP AES. Three co-eluting peaks
are shown with the UV-Vis detection. With MIP AES, only
P-containing vitamin B12is detected.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
131
0.75---------------
-
- ---------- ----------
Absorbance (AU)
VrtB1
VitB6
Vit B12
0 .5 -
0.25 —
0
250
Figure 60.
300
Wavelength (nm)
350
400
UV-Vis spectrum o f vitamins (250-400 nm) B„ B6, and B12.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
132
4.4.3 Detection o f Nucleotides
Ion-pair chromatography is used to separate ionic compounds using a reversephase column. A counter-ion o f a charge opposite to that of the analyte is used in the
mobile phase. The counter-ion is thought to associate with the analyte to fonn an ion
pair. The counter-ion may contain a hydrophobic moiety which interacts with the
hydrophobic stationary phase o f the column to enhance analyte retention.
Ion-pair chromatographic methods are fast, reliable, and simple. The ion-pair
reverse-phase mode is routinely used for the separation o f bases and acids,
pharmaceuticals, amino acids, peptides, proteins, nucleic acid compounds, and for
separations of chiral analytes.13* Combes et al.139successfully used ion-pair reversephase HPLC to separate many fundamental biological compounds such as nucleotides
and nucleotide sugars. The nucleotides play a vital role in many biochemical reactions
in prokaryotic and eukaryotic cells and are important as energy storage molecules.
Both anion exchange and ion-pair chromatography have been used for the separation
o f these compounds. In this section, the application o f the MIP AES detection for
ion-pair chromatography is examined.
The separation of cytidine S' monophosphate (CMP), uridine S'
monophosphate (UMP), and guanosine S' monophosphate (GMP) is examined. The
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
133
molecular structures o f these compounds are shown in Figure 61. Pure samples o f
these nucleotides were obtained from Sigma Chemical Co. (St. Louis, MO). Before
attempting the separation, aqueous solutions o f the nucleotides were introduced to the
plasma through direct solution nebulization and the P signal at 253.7 nm was
observed. The P detection limits obtained were 20,34, and 47 ppb for CMP, UMP,
and GMP respectively.
For the LC separation, a solution of CMP, UMP, and GMP was prepared in
water. The final solution contained 50 ppm P from each o f the aforementioned
monophosphates. The mobile phase used was a 95:5 mixture of water and methanol
with a flow rate o f 0.55 mL/inin. The pH o f the mobile phase was adjusted to 4.8
using 100-mM solutions o f CH3COOH and CH3COONa. Since the signal from P was
monitored, commonly used phosphate buffers were avoided. Tetrabutylammonium
hydrogen sulfate (TBAHS) (Aldrich Chemical Company, Inc., Milwaukee, WI) was
used as the counter-ion. The concentration of TBAHS in the mobile phase was 8
mmol/L, about 5 times in excess o f the stoichiometric requirement. The tetrabutyl
portion of TBAHS provided the hydrophobic interaction with the C „ column and the
ammonium ion formed the ion pair with the anionic nucleotide. The separation is
shown in Figure 62, with the UV detector set at 254 nm. Three well-separated peaks
from CMP, UMP, and GMP were observed. The retention times for the analytes were
5.7,7.5, and 9.4 minutes for CMP, UMP, and GMP respectively.
This separation was conducted with the MIP AES under the optimized
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
ONa L
/ \
vOH rOH
c
Figure 61.
Molecular structures o f (a) cytidine S' monophosphate
disodium salt, (b) uridine S' monophosphate in free acid form,
and (c) guanosine S' monophosphate disodium salt.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
I
I
0.0
Figure 62.
I
r
" I
■ i ■
»
i
i
i
i
i ■' i — i
i
i 1
i
i
i
Tine (nm)
20.0
Chromatogram of a mixture of CMP, UMP, and GMP using the
UV-Vis detector at 254 nm. Retention times for CMP, UMP, and
GMP were 5.7,7.5, and 9.4 minutes, respectively. Mobile phase
used was 95:5 MeOH:H20 at a flow rate o f 0.55 mL/min. pH of
the mobile phase was 4.8. TBAHS was used as the counter-ion.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
136
conditions listed in Table 2. Phosphorus emission at 2S3.7 nm was monitored.
However, no signals from the nucleotides were observed with the plasma detector.
The plasma turned intense yellow due to the Na emissions from the CH3COONa in the
mobile phase. As discussed in section 3.3.2, excess Na could have perturbed the
plasma chemistry, and no signal was observed as a result. In future, different buffer
systems and their compatibility with the MIP detection system must be investigated.
4.5 Summary
Superior selectivity and sensitivity of the MIP AES over UV-Vis detection
systems were demonstrated. The ability o f the plasma system to handle high molecular
weight compounds such as DNA has also been shown. However, the inability of the
plasma system to produce signal from the nucleotides in the presence o f the acetic
acid/sodium acetate buffer is a concern. Systematic studies o f plasma performance in
the presence o f different buffer systems warrants further investigation.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
CHAPTER 5
CONCLUSIONS AND FUTURE DIRECTIONS
In this dissertation, characterization o f an MIP AES system for nonmetal
selective detection and possible applications as a detector for reverse-phase liquid
chromatography have been discussed. MIP AES with ultrasonic nebulization and
membrane desolvation was successfully used for nonmetal (Cl, Br, I, S, and P)
detection in organic analytes. Membrane desolvator temperature was successfully
optimized for maximum nonmetal signal from semivolatile organic analytes.
Linearization of the calibration plots was achieved through use of dilute H2S04 as a
solvent modifier. The inability o f USN with a fixed-fiequency (1.4 MHz) automatic
transducer to nebulize certain HPLC solvents (aqueous solutions of 10 -70%
methanolrwater and acetonitrilerwater) was remedied by frequency optimization using
a manual transducer controller. Coupling o f MIP AES in this configuration to a
reverse-phase LC system was accomplished for separation o f various biologically
active and pharmaceutically important mixtures. However, for LC applications, MIP
AES detection limits must be significantly enhanced for it to compete with the already
existing LC detectors.
As discussed in Chapter 2, despite optimization of various plasma, USN, and
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
138
membrane desolvator conditions, the detection limits did not improve significantly. It
appears that enhancement o f analyte throughput to the plasma using other types of
nebulizers could affect the MIP AES sensitivity. Akinbo and Carnahan34 have done
extensive research on analyte transport through USN-mebrane desolvator systems.
The USN has a high solvent uptake rate (0.S -1 mL/min) and it transports 5 to 10%
analyte into the plasma. At these flow rates the solvent effects in the plasma are
significant. To minimize these effects, a membrane desolvator is necessary. The use of
a membrane desolvator causes further loss of analyte and broadens chromatographic
peaks. However, if a high-throughput microflow nebulizer (uptake rate 20-50
|iL/min)140were used, reduced solvent effects might obviate the need for a membrane
desolvator. These nebulizers transport 50-60% of the total analyte. Hence, significant
improvements in detection limits can be achieved. The MIP AES with low-flow
nebulizers can be easily interfaced to micro-LC columns for separation and detection.
For LC applications, compatibility of the MIP system with different buffer
systems should also be assessed. As discussed in Chapter 4, the plasma system failed
to produce signal from the nucleotide samples in the presence o f the acetic
acid/sodium acetate buffer system. Also, the phosphate buffer could not be used for
phosphorus selective detection. Alternative buffer systems more compatible with
MIPs should be investigated to obtain the requisite pH for the separation.
Finally, for simultaneous elemental analysis, a CCD array detector or a
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
139
polychiomator should be used with the LC-MIP AES. Simultaneous elemental
detection would provide a quicker means to examine multi-element analysis for
empirical formula determinations and structure elucidation.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
REFERENCES
1.
L. Ebdon, S. Hill, R. W. Ward, Analyst 111, 1113 (1986).
2.
D. L. Haas and J. A. Caruso, Anal. Chem. 56,2014 (1984).
3.
J. P. Matousek, B. J. Orr, M. Selby, Prog. Anal. At. Spectrosc. 7,275 (1984).
4.
P. C. Uden, Chromatogr. Forum 1 ,17 (1986).
5.
P. C. Uden, Trends Anal. Chem. 6,238 (1987).
6.
K. Tanabe, H. Haraguchi, and K. Fuwa, Spectrochim. Acta36B, 633 (1981).
7.
P. G. Brandi, J. W. Carnahan, Spectrochim Acta 49B,
105 (1994).
8.
L. Ebdon, S. Hill, and R.W. Ward, Analyst 111,1113 (1986).
9.
J. C. Giddings, E. Grushka, and P. R. Brown, Advances in Chromatography.
(Marcel Dekker, NY, 1987), vol. 26.
10.
L. Zhang, J. W. Carnahan, R. E. Winans, and P. H. Neill, Anal. Chem. 63, 213
(1991).
11.
S. A. Estes, P.C. Uden, and R. M. Bames, Anal. Chem. 5 3 , 1829 (1981).
12.
N. P. Van Dalen, P. A. de Lezenne Coulander, and L. de Galen, Anal. Chim.
Acta. 94,1(1977).
13.
C. Bradley and J. W. Carnahan, Anal. Chem. 60,858 (1988).
14.
K. J. Slatkavitz, P. C. Uden, and R. M. Bames, J. Chromatogr. 355,117 (1986).
15.
M. Wu, M. Lee, and P.B. Farnsworth, J. Anal. At. Spectrom. 7 , 197 (1992).
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
141
16.
G. H. Webster, W. 0 . Doggett, and C. B. Boss, Anal. Chim. Acta 257,309
(1992).
17.
B. D. Quimby and J. J. Sullivan, Anal. Chem. 6 2 , 1027 (1990).
18.
A. W. Boom and R. G. Browner, Anal Chem. 5 4 , 1402 (1982).
19.
D. W. Hausler and L. T. Taylor, Anal. Chem. 53,1223 (1981).
20.
D. W. Hausler and L. T. Taylor, Anal. Chem. 53,1227 (1981).
21.
G. Kreuning and F. J. M. J. Maessen, Spectrochim. Acta 44B, 367 (1989).
22.
F. J. M. J. Maessen, P. J. H. Seeverens, G. Kreuning, Spectrochim. Acta 39B,
1171(1984).
23.
F. J. M. J. Maessen, G. Kreuning, and J. Balke, Spectrochim. Acta 4!B, 3
(1986).
24.
T. Brotherton, B. Bames, N. Vela, and J. Caruso, J. Anal. At. Spectrosc. 2 , 389
(1987).
25.
P. Barret and E. Pruszkowska, Anal. Chem. 5 6 , 1927 (1984).
26.
M. W. Baldes and B. L. Caughlin, Spectrochim. Acta 40B, 579 (1985).
27.
J. Su, K. Kavvagichi and A. Mizuike, Anal. Chim. Acta 152, 133 (1983).
28.
H. Tao and A. Miyazaki, J. Anal. At. Spectrom.10,1 (1995).
29.
J. S. Crain and J. A. Alvarado, J. Anal. A t Spectrom. 9,1223 (1994).
30.
A. Gustavsson, Spectrochim. Acta 42B, 113 (1987).
31.
L. Zhang, J. W. Carnahan, E. Winans, and P. H. Neill, Anal. Chem. 6 1,895
(1989).
32.
J. Qinhan, Z. Hangi, W. Ying, Y. Xianglin, and W. Yang, J. Anal. At. Spectrom.
9,851 (1994).
33.
A. Gustavsson, Spectrochim. Acta 43B, 917 (1988).
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
142
34.
O. T. Akinbo and J. W. Carnahan, Talanta 4 5 , 137 (1997).
35.
P. B. Mason, L. Zhang, J. W. Carnahan, and R. E. Winans, Anal. Chem. 65,
2596(1993).
36.
O. T. Akinbo and J. W. Carnahan, Anal. Chim. Acta. 390,217 (1999).
37.
J. W. Carnahan, K. M. Mulligan, and J. A. Caruso, Anal Chim. Acta 130,227
(1981).
38.
P. C. Uden, J. Chromatogr. A 703,393 (1995).
39.
J. D. Ingle Jr. and S. R.Crouch, Spectrochemical Analysis (Prentice Hall: NJ,
1988), p. 233.
40.
A. Montaser, Inductively Coupled Plasma Mass Spectrometry. (WILEY-VCH:
NY, 1998).
41.
R. D. Satzger, F. L. Fricke, and J. A. Caruso, Spectrochim. Acta 42 B, 705
(1987).
42.
P. G. Brown, T. M. Davidson, and J. A. Caruso, J. Anal. At. Spectrom. 3,763
(1988).
43.
Q. Jin, Y. Duan, and J. A. Olivares, Spectrochim. Acta 52B, 131 (1997).
44.
B. D. Qumby, P. C. Uden, and R. M. Bames, Anal. Chem. 51,875 (1979).
45.
M. Wu, J. M. Gehlhausen and J. W. Carnahan, J. Chem. Ed. 69,757 (1992).
46.
K. J. Jones and J. W. Carnahan, Spectrochim. Acta 47 B, 1229 (1992).
47.
J. W. Carnahan and G. M. Hieftje, Spectrochimica Acta 47B, 731 (1992).
48.
D. Rapp and W. E. Francis, J. Chem. Phys. 37,2531 (1962).
49.
G. M. Hieftje, J. Anal. Atom. Spectrosc. 11,613 (1996).
50.
J. A. C. Broekaert, R. F. Browner, and R. K. Marcus, Spectrochim. Acta 50B, 7
(1995).
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
143
51.
D. T. Heitkemper, K. A. Wolnik, F. L. Fricke, and J. A. Caruso, “Injection of
Gaseous Samples into Plasmas,” Inductively Coupled Plasma in Analytical
Atomic Spectrometry. (A. Montaser and D. W. Golightly, Eds., VCH, New
York, 1992), 2nded.
52.
S. Greenfield and A. Montaser, “Common RF Generator, Torches, and Sample
Introduction Systems,” Inductively Coupled Plasma in Analytical Atomic
Spectrometry. (A. Montaser and D. W. Golightly, Eds., VCH, New York, 1992),
2nded.
53.
A. Gustavsson, “Liquid Sample Introduction into Plasmas,” Inductively Coupled
Plasma in Analytical Atomic Spectrometry. (A. Montaser and D. W. Golightly,
Eds., VCH, New York, 1992), 2nded..
54.
R. F. Browner, “Fundametal Aspects o f Aerosol Generation and Transport,”
Inductively Coupled Plasma Emission Spectroscopy. (P. W. J. M. Boumans,
Ed., Wiley, New York, 1987), Prt II.
55.
B. L. Sharp, J. Anal. At. Spectrom. 3,613 (1988).
56.
B. L. Sharp, J. Anal. At. Spectrom. 3,939 (1988).
57.
V. A. Fassel and B. R. Bear, Spectrochim. Acta 41B, 1089 (1986).
58.
R. H. Clifford, P. Sohal, H. Liu and A. Montaser, Spectrochim. Acta 47B, 1107
(1992).
59.
S. Nukiyama and Y. Tanasawa, Trans. Soc. Mech. Eng. (Japan), Reports 4,5,
and 6, 1938(1950).
60.
R. J. Lang, J. Acoustical Soc. Amer. 34,6 (1962).
61.
V. A. Fassel and G. W. Dickinson, Anal. Chem. 40,247 (1968).
62.
M. P. Dziewatkoski, L. B. Daniels, and J. W. Olesik, Anal. Chem. 68,1101
(1996).
63.
R. F. Browner, A. W. Boom, and D. D. Smith, Anal. Chem. 54,1411 (1982).
64.
M. A. Tarr, G. Zhu, and R. F. Browner, Appl. Spectrosc. 45(9), 1424 (1991).
65.
R. C. Hutton, and A. N. Eaton, J. Anal. At. Spectrom. 2 ,595 (1987).
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
144
66.
G. Zhu, and R. F. Browner, J. anal. At. Spectrom. 3,781 (1988).
67.
C. Pan, G. Hhu, and R. F. Browner, J. Anal. At. Spectrom. 5,537 (1990).
68.
N. Jakubowski, I. Feldman, D. Stuewer, and H. Bemdt, Spectrochim. Acta 47B,
119(1992).
69.
R. S. Houk, V. A. Fassel, G. D. Flesch, H. J. Svec, A. L. Gray, and C. E. Taylor,
Anal. Chem. 52,2283 (1980).
70.
J. W. Lam and J. W. McLaren, J. Anal. At. Spectrom. 5,419 (1990).
71.
R. Tsukahara and M. Kubota, Spectrochim. Acta 45B, 581 (1990).
72.
A. Gustavsson, Spectrochim. Acta 43B, 917 (1988).
73.
A. Gustavsson, Spectrochim. Acta 42B, 111 (1987).
74.
J. Qinhan, Z. Hanqi, W. Ying, Y. Xianglin, and Y. Wenjun, J. Anal. At.
Spectrom. 9,851 (1994).
75.
D. R. Wiederin, R. S. Houk, R. K. Winge, and A. P. D’Silva, Anal. Chem. 62,
1155(1990).
76.
L. C. Alves, L. A. Allen, and R. S. Houk, Anal. Chem. 65,2468 (1993).
77.
L. C. Alves, D. R. Wiederin, and R. S. Houk, Anal. Chem. 6 4 , 1164 (1992).
78.
A. Montaser, H. Tan, I. Ishii, S-H. Nam, and M. Cai, Anal. Chem. 63,2660
(1991).
79.
H. Bemdt, Fresenius Z. Anal. Chem. 331,321 (1988).
80.
S. K. Luo and H. Bemdt, Spectrochim. Acta 49B, 485 (1994).
81.
G. E. Pacey, M. R. Streka, and J. R. Gord, Anal. Chem. 58,502 (1986).
82.
R. M. Bames and X ., J. V. mg, An.:!. At. Spectrom. 3, 1083 (1988).
83.
M. L. Magnuson, J. T. Creed, and C. A. BrockhofF, J. Anal. At. Spectrom. 11,
893 (1996).
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
145
84.
J. T. Creed, M. L. Magnuson, I. Chamberlain, and M., J. Sivaganesan, Anal. At.
Spectrom. 11,505 (1996).
85.
M. Yamamoto, M. Yasuda, and Y. Yamamoto, J. Flow Injection Anal. 2,134
(1985).
86.
M. Yamamoto, K. Takada, T. Kumamaru, M. Yasuda, S. Yokoyama, and Y.
Yamamoto, Anal. Chem. 59,2446 (1987).
87.
Z. Wang and R. M. Bames, J. Anal. At. Spectrom. 3,1091 (1988).
88.
T. Sunden and A. Cedergren, Anal. Chem. 56,1085 (1984).
89.
T. Wolfgang, Anal. Chim. Acta 291,305 (1994).
90. T. Kuwaki, M. Akiba, M. Oshima, and S. Motomisu, Bungeku Kagaku 36, T84
(1987).
91.
S. Ouyang, H. Y. Chen, and Y. Xu, Anal. Chim. Acta 337,165 (1997).
92.
R. C. Johnson, N. Srinivasan, R. G. Cooks, and D. Schell, Rapid Comm. Mass
Spectrom. 11,363 (1997).
93.
S. Bauer, T. Griffin, and J. Bauer, Int. J. Mass Spectrom. Ion. Processes 155,
107(1996).
94.
M. E. Cisper, A. W. Garrett, D. Cameron, and P. H. Hemberger, Anal. Chem.
68,2097(1996).
95.
I. Jelinek, Chemical Papers Chemicke Zvesti. 50, 131 (1996).
96.
A. R. Dongre and M. J. Hayword, Anal. Chim. Acta 327,1 (1996).
97.
K. Vlastimil, Critic. Rev. Anal. Chem. 23,323 (1988).
98.
K. Backstrom, A. Gustavsson, P. Hietala, Spectrochim. Acta 44B, 1041 (1989).
99.
M. Van Son, R. C. Schothorst, and G. den Boef, Anal. Chim. Acta 153,271
(1983).
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
146
100.
J. R. Clinch, P. J. Worsfold, and F. W. Sweeting, Anal. Chim. Acta 214,401
(1988).
101.
P. C. Hauser, S. S. Tan, T. J. Cardwell, R. W. Cattrall, and I. C. Hamilton,
Analyst 113,1551(1988).
102.
T. J. Aoki, Flow Injection Anal. 5,95 (1988).
103.
S. D. NikoliC, T. D. Jankovid, E. B. Molosavljevid, J. L. Hendrix, and J. H.
Nelson, Analyst 116,49 (1991).
104.
S. D. NikoliC, T. D. JankoviC, E. B. MolosavljeviC, J. L. Hendrix, and J. H.
Nelson, Fresenius, J. Anal. Chem. 342,98 (1992).
105.
M. R. Straka, G. Gordon, and G. E. Pacey, Anal. Chem. 57, 1799 (1985).
106. C. Xu, J. S. Patrick and R. G. Cooks, Anal. Chem. 67,724 (1995).
107. J. S. Cha, R. Li, and K. K. Sirkar, J. Membr. Sci. 119,139 (1996).
108. Z. Y. Zhang, J. Poerschmann, and J. Pawliszyn, Anal. Comm. 33,219 (1996).
109. D. H. Garg, W. Lenk, S. Berwald, K. Lukwitz, F. Simon, and K. J. Eichhom, J.
Appl. Polymer Sci. 60,2087 (1996).
110.
T. Kotiaho, F. R. Lauritsen, H. Degn, and H. Paakkanen, Anal. Chim. Acta 309,
317(1995).
111. T. Tryzell and B. Karlberg, Anal. Chim. Acta 308,206 (1995).
112. A. Gustavsson, P. Hietala, Spectrochim. Acta 45B, 1103 (1990).
113.
J. F. Yang, T. S. Conver, K. A. Koropchak and D. A. Leighty, Spectrochim.
Acta. 51,1491 (1996).
114.
W. R. L. Caims, L. Ebdon, and S. J. Hill, Fresenius’ J. Anal. Chem. 355,202
(1996).
115.
K. J. Leckrone and J. M. Hayes, Anal. Chem. 69,911 (1997).
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
147
116.
B. Magyar, P. Liennemann and H. Vonmmont, Spectrochim. Acta 41B, 27
(1986).
117.
J. W. McLaren, J. W. Lam, A. Gustavsson, Spectrochim. Acta 45B, 1091
(1990).
118. R. I. Butto and J. J. Zhu, J. Anal At. Spectrom. 9,90S (1994).
119. H. Tao and A. J. Miyazaki, Anal. At. Spectrom.l, 10 199S.
120.
A. W. Kim, M. E, Foulkes, L. Ebdon, S. J. Hill, R. L. Patience, A. G. Barwise,
and S. J. Rowland, J. Anal. At. Spectrom. 7,1147 (1992).
121.
A. Kim, S. Hill, L. Ebdon, and S. Rowland, J. High Resolut. Chromatogr 15,
665(1992).
122. A. Prange and E. Jantzen, J. Anal. At. Spectrom 10,105 (1995).
123. L. Ebdon, E. H. Evans, W. G. Pretorius and S. J. Rowland, J. Anal. At.
Spectrom. 9,939 (1994).
124. C. Bradley and J. W. Carnahan, Analytical Chemistry 60,858 (1988).
125. B. D. Quimby, M. F. Delaney, P. C. Uden, and R. M. Bames, Anal. Chem. 51,
875(1979).
126.
G. Becker and A. Colmsjo, Analytica Chimica Acta. 376(3), 265 (1998).
127.
S-J. Jiang and R. S. Houk, Spectrochim. Acta 43B, 405 (1988).
128.
S. C. K. Shum and R. S. Houk, Anal. Chem. 65,2972 (1993).
129.
X. Dauchy, R. Cottier, A. Batel, R. Jeannot, M. Borsier, A. Astruc, and M.
Astruc, J. Chromatogr. Sci. 31,416 (1993).
130.
M. A. Quijano, A. M. Gutierrez, M. C. Perez-Conde, and M. C. Camara, J.
Anal. A t Spectrom. 11,407 (1996).
131.
M. J. Tomlinson, J. S. Wang, and J. A. Caruso, J. Anal. At. Spectrom. 9,957
(1994).
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
148
132.
K. Kawabata, Y. Kishi, O. Kawaguchi, Y. Watanabe, and Y. Inoue, Anal. Chem.
63,2137(1991).
133.
R. C. Weast and M. J. Astle, CRC Handbook o f Chemistry and Physics. (CRC
Press Inc., Boca Raton, FL, 1982), 62nded., p. D-198.
134.
R. C. Weast and M. J. Astle, CRC Handbook o f Chemistry and Physics. (CRC
Press Inc., Boca Raton, FL, 1982), 62nded., p. F-35.
135.
L. A. Moran, K. G. Scrimgeour, H. R. Horton, R. S. Ochs, and J. D. Rawn,
Biochemistry. (Neil Patterson Publishers, Englewood, NJ, 1994), 2nded., p. 24l.
136.
J. Sambrook, E. F. Fritsch, and T. Maniatis, Molecular Cloning A. a
Laboratory Manual. (Cold Spring Harbor Laboratory Press, Plainview, NY,
1989), 2nded., p. 6.7.
137.
S. Wongyai, J. Chromatogr. A 870,217 (2000).
138.
A. Wemer, Chromatographia 31(7/8), 401 (1991).
139.
I. Meynial, V. Paquet, and D. Combes, Anal. Chem. 67,1627 (1995).
140.
M. Huang, H. Kojima, T. Shirasaki, A. Hirabayashi, and H. Koizumi, Anal.
Chim. Acta 413, 217(2000).
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.