MSU

LIBRARIES

 

‘

 

RETURNING MATERIALS:
Place in book drop to
remove this checkout from
your record. FINES wi11
be charged if book is
returned after the date
stamped be1ow.

 

 

 

 

 

 

MICELLE STABILIZED ROOM TEMPERATURE PHOSPHORESCENCE
BY FLOW INJECTION ANALYSIS
AND
A DUAL BEAM FIBER OPTIC-BASED PHOTOMETRIC COMPARATOR

by

Paul Robert Kraus

A DISSERTATION

Submitted to
Michigan State University
in partial fulfillment of the requirements
for the degree of

DOCTOR OF PHILOSOPHY

Department of Chemistry

1988

ABSTRACT
MICELLE STABILIZED ROOM TEMPERATURE PHOSPHORESCENCE BY
FLOW INJECTION ANALYSIS AND A DUAL BEAM FIBER OPTIC-BASED
PHOTOMETRIC COMPARATOR
by

Paul Robert Kraus

A system is presented for routinely performing room temperature
phosphorescence (RTP) measurements of analytes in fluid solution. RTP spectrometry
has been used as a detection sySIem for flow injection analysis (FIA). This combination
utilizes the Speed and automation of FIA, and the inherent selectivity and sensitivity of
RTP. Results are presented to demonstrate the utility for the RTP/PIA configuration.

System components include a commercially available peristaltic pump, a rotary
sample injection valve, a flow-through fluorescence cell, and a spectrofluorometer as
the detector. Two spectrofluorometers were used: a commercially available instrument,
and a microprocessor controlled home-buflt instrument. All instrumentation used is
described in detail. Results are presented for the characterization Of the dominant
source Of noise under the conditions normally encountered in the RTP experiments.

The removal of ubiquitous molecular oxygen, an efficient triplet state quencher,
from the analytical solutions was the major problem encountered in this research.
Several methods for deoxygenating a flowing stream are evaluated for their
applicability to the RTP/PIA configuration. These methods include diffusion of oxygen
across a gas permeable membrane into an oxygen scavenging solution Of Cr(II), vacuum

deoxygenation through a gas permeable membrane, and deoxygenation through the

reduction of oxygen to water by sulfite ion; the latter was the most effective. The rise in
RTP to a steady-state intensity is Observed to be rapid after an initial lag time.

Calibration curves Obtained for naphthalene, 2-bromonaphthalene, and biphenyl
are Observed to be linear up to two orders of magnitude in concentration. The rise in
RTP for pyrene and for higher concentrations of biphenyl was Observed to follow an
exponential rise. Although a steady-state intensity was not obtained for pyrene, the
steady-state intensity was predicted from an exponential model. Results are presented
demonstrating the usefulness Of this technique.

An inexpensive and portable dual-beam photometric comparator has been
developed. It has an advantage over existing single beam systems in that direct
comparisons about a critical decision point can be made, and the effects of stray
radiation are minimized. A light emitting diode (LED) has been used as the light source
because of its fixed wavelength, narrow emission profile, low current requirements, and
low cost. The comparator has been designed for photometric determinations where it is
desirable to know if the analyte concentration is within a certain range. The sample
introduction system has been designed to accommodate microtiter wells such as those
often used in routine immunoassay procedures as the cuvettes. The unit has been
designed to be operated by non-technically oriented persons. It can be battery powered
for field use or operated from ac power.

DEDICATION

To my Grandparents,
my Parents, and

my wife, Theresa.

iv

ACKNOWLEDGEMENTS

During my accumulated time at MSU I’ve made many friends, perhaps too many
to acknowledge individually. However, several people do merit individual recognition.
First, I’d like to express my thanks to Professor Stanley Crouch for his help and for
teaching me that independent research can be a rewarding experience. I owe many
thanks to Azure Fleetfang and Panther Solecat (a.k.a. Gene Ratzlaff and Chas Patton)
for their assistance during my "formative years" when this all got started.

Pete Palmer deserves special mention for helping me to retain a much needed
level Of insanity, and for friendship above and beyond the call Of duty. I’ll never forget
our excursions at the RB. and discussions about the Hobbit/Lord of the Rings (l). Other
folks which deserve special thanks are Bob, Pete W., Adrian, Chris, Cheryl, Norm,
June, Jon and Karen to mention a few. The many other friends I’ve made while at MSU
will not be forgotten.

Many thanks tO Marty "Electric Bicycleman" Rabb for being patient with my
endless questions regarding electronic circuits and for the biking lessons. The men of
the machine shop, Russ Geyer, Dick Menke, and Deak Watters deserve
acknowledgment for construction of countless items and repairs.

On the lighter side, I’d like to thank Dr. Loran Bieber for his friendship and the
fishing trips during which he never failed to remind the crew to keep the rod tip up!
That’s what life is all about.

A very special acknowledgment is extended to my wife, Theresa, for her faith in
me and for putting-up with me through a very difficult time. Without Theresa I might
not have persevered!

And finally, to those people whose mere existence seemed to make life

miserable, I tip my "Social Statement Cap!" (with acknowledgments to Bob R.).

TABLE OF CONTENTS

Chapter

IV.

List Of Tables
List of Figures

INTRODUCTION
A. References

MOLECULAR LUMINESCENCE in SOLUTION
A. Absorption Of light
B. Fluorescence
Fluorescence Lifetimes
Quenching of Fluorescence
C. Phosphorescence
Phosphorescence Lifetimes
Oxygen Quenching
Heavy Atom Effect in Phosphorescence
D. References

ROOM-TEMPERATURE PHOSPHORESCENCE
A. Introduction to Room Temperature Phosphorescence
B. Solid Surface RTP
C. Sensitized RTP
D. Microcrystalline and Colloidal RTP
E. Cyclodextrin Enhanced RTP
F. Micelle-Stabilized RTP
G. Deoxygenation methods
Inert Gas Purging
Vacuum Degassing
Freeze-Pump-Thaw
Semipermeable Membranes
Chemical Scavenging
Chromiumal)
Zinc Column
Sodium Sulfite
l-I. References

FLOW INJECTION ANALYSIS

A. Principles of Flow Injection Analysis

B. Related Applications Of Flow Injection Analysis
StOpped-Flow in FIA
Luminescence Detection in FIA

C. References

vi

ii

viii

V. INSTRUMENTATION 55

A. Home-Built Microcomputer-Controlled Luminescence Spectrometer 55
Twin Bus Microcomputer System 56
Home-Built Luminescence Spectrometer 59
Microcomputer/ S pectrometer Interface 59
Experimental Determination of the Dominant Noise Source 67

B. Perkin-Elmer Model LS-S Luminescence Spectrometer 70

C. References 71

VI. LUMINESCENCE SPECTRA 72

A. Reagents 72

B. Fluorescence Spectra 74

C. RTP Spectra 76
MS-R'I'P Spectra 76
CD-RTP Spectra 85

D. Phosphorescence Lifetimes 90

E. References 95

VII. ROOM TEMPERATURE PHOSPHORESCENCE with FLOW INJECTION

ANALYSIS SAMPLE INTRODUCTION 96

A. Flow Injection System 96

B. Carrier Study 98

C. Deoxygenation of Flowing Streams 100
Chromium (II)/Zn(Hg) Oxygen Scavenging 100
Vacuum Deoxygenation 108
Sodium Sulfite Deoxygenation 110

D. RTP/FIA Calibration Curves 112

E. References 128

VIII. WHAT’S NEXT? 129

A. References 132

IX. PORTABLE DUAL BEAM FIBER OPTIC-BASED PHOTOMETRIC
COMPARATOR 133

A. Introduction 133

B. Experimental 135
Optical Arrangement 135
Analytical Manifold 135
Circuit Description 137

C. Results and Discussion 140

D. References 143

X. APPENDIX: FORTH SOURCE CODE LISTING 144

vii

Table
Table 5.1

Table 5.2
Table 7.1
Table 7.2

Table 7.3
Table 7 .4

Table 7.5

Table 7.6

Table 7.7
Table 7.8

Table 9.1

LIST of TABLES

lid;

Modules with their Principal Integrated Circuits

Effect of Radiant Power on the S/N

Comparison of Static and Flow Cell in Perkin-Elmer LS-S

Comparison of Static and Flow Cell in Home-Built
Instrument

Fluorescence Enhancement Factors

Results of Weighted Linear Regression for Naphthalene
Calibration Curve

Results of Weighted Linear Regression for
2-Bromonaphthalene Calibration Curve

Results of Weighted Linear Regression for Biphenyl
Calibration Curve

Parameters for Pyrene fit to Exponential Rise in RTP

Results of Weighted Linear Regression for Pyrene Calibration
Curve

Components of Comparator Circuit

viii

57
69
97

98
105

114

117

121
124

127
139

figure
Figure 2.1

Figure 3.1.
Figure 3.2.

Figure 4.1

Figure 4.2

Figure 5.1

Figure 5.2

Figure 5.3

Figure 5.4

LIST of FIGURES

Title

Generalized Jablonski diagram. A, absorbance; F,
fluorescence; P, phosphorescence; VR, vibrational relaxation;
IC, internal conversion; ISC, intersystem crossing; So, 31, and
S 2 are the ground and excited singlet states, respectively, and
T1 is the excited triplet state.

Simplified representation of a cyclodextrin molecule.

A) Simplified representation of a surfactant monomer.
B) Anionic micelle in aqueous solvent.
C) Reverse micelle.

Simplified representation of a flow injection system. The
reagent R 18 pumped through the manifold where the sample S
is injected by a sample injection valve. The product is
monitored by the flow-through detector PC. The product 18
then vented to wane W.

Laminar flow with molecular diffusion: (a) the sample plug,
(b) under conditions of laminar flow, (c) laminar flow with
molecular diffusion, ((1) Gaussian peak profile resulting from
molecular diffusion.

Block diagram of home-built luminescence spectrometer
showing orientation of the components. Key: 8, radiation
source; L, lens; C, optional chopper; Mex, excitation
monochromator; SC, sample chamber; PD, photodiode with
amplifier circuitry; Mem, emission monochromator; I/V,
current amplifier.

Schematic of transistor driver circuit used to amplify current
supplying power to chopper motor.

Timing diagram depicting the three types of phasing:

A, chopper signal lags source signal; B, chopper signal leads
source signal; and C, signals are in phase.

Photodiode amplifier circuit. R1=1 MO, R2 = 2. 2 M9, and
Rin- — 680 9.

ix

29

33

47

48

62

65

Figure 6.1

Figure 6.2

Figme 6.3

Figure 6.4

Figure 6.5

Figure 6.6

Figure 6.7

Figure 6.8

Figure 6.9

Figure 6.10

Figure 6.11

Figure 6.12

Excitation and fluorescence emission spectra of naphthalene.
A) excitation scan taken on Perkin-Elmer LS-5 with

Rem = 330 nm. B) emission scan taken on LS-5 with

hex = 276 nm. C) emission scan taken on home-built
instrument with hex = 280 nm. See text for explanation of
peaks above 450 nm.

Excitation and fluorescence emission spectra of
2-bromonaphthalene. A) excitation scan from LS-5 with
lam = 315 nm. B) emission scan from LS-S. C) emission
scan from home-built instrument.

Excitation and fluorescence emission spectra of pyrene.

A) excitation scan taken on home-built mstrument with
lem- - 395 nm. B) excitation scan taken on LS- 5 with. See
text for explanation 1n peak mtensities. C) emission scan
taken 011 LS 5, kex- - 335 nm.

Excitation and fluorescence emission spectra of biphenyl taken
on IS-S. A) excitation scan with lem- - 315 nm. B) emission
scan with lex- - 260 nm.

Fluorescence emission of naphthalene taken on home-build
instrument with he x: 280 nm and 16 nm bandpasses.
A)0.10MSDS, andB)0.10MSDS with 25mMTlNO3
added. See text for explanation.

Luminescence spectra of naphthalene in 0.10 M SDS with
25 mM 'I'lNO3 (dotted line), and deoxygenated solution (solid
line).

Phosphorescence spectrum of naphthalene 1n 0. 10 M SDS with
25 mM TlNO3 and 10 mM NaZSO3. Delay time after I-II—to-
LO transition of chopper 1s 10 us.

Rise in phosphorescence intensity for two concentrations of
surfactant. A) 0.15 M 30/70 Tl/Na DS. B) 0.10 M

30/70 Tl/Na DS. Phosphorescence intensity is in arbitrary
units.

MS- RTP spectrum of 2- bromonaphthalene 1n 0. 15 M
30/70 Tl/Na DS with 10 mM NaZSO3. A) home-built
instrument, and B) Perkin- Elmer L8- 5.

MS- RTP spectrum ofpyrene 1n 0.15 M 30/70 Tl/Na DS with
10 mM Na 8033. A) home-built instrument, and B) Perkin-
Elmer LS5

MS-RTP spectrum Of biphenyl in 0.10 M 30/70 TVNa DS with
10 mM NaZSO3.

CD-RTP spectrum of 2-bromonaphthalene in 0.01 M
B—cyclodextrin with 0.58 M DBE.

75

77

78

79

80

82

83

84

86

87

88

89

Figure 6.13

Figure 6.14

Figure 7.1

Figure 7.2

Figure 7.3

Figure 7.4

Figure 7.5

Figure 7.6

Figure 7.7
Figure 7.8
Figure 7.9
Figure 7.10

Figure 7.11

Figure 7.12

Figure 7.13
Figure 7.14

Phosphorescence decay of naphthalene. Data from, A) Perkin-
Elmer LS—5, B) home-built instrument. Curve through data
points represents the best fit to data as determined by the
single exponential decay of equation 61.

Phosphorescence decay of pyrene. Data from home-built
instrument. Curve represents best fit of data to single
exponential decay.

A) fluorescence peak profile of pyrene Obtained with water as
the carrier, and B) SDS carrier.

Characteristic fluorescence peak profiles obtained with
A) water, B) SDS, C) NaCl, D) NaBr, E) NaC2H302, and
F) LiCl.

Concentric tube dialyzer used for deoxygenation
with Cr(II).

FIA manifold used for deoxygenation with concentric tube
dialyzer

A) Fluorescence peak profiles obtained for pyrene by FIA.
Sample concentrations from left to right are 20, 15, 12, 10, 7,
5. 2, and 1 11M, respectively. B) Calibration curve using
average maximum peak intensity obtained from A.

Absorbance spectra of A) pyrene in 0.15 M SDS,

3) Chromium deoxygenating solution, and C) pyrene waste
solution showing contamination of pyrene sample by
chromium.

Vacuum deoxygenation apparatus.
FIA manifold used for RTP determinations.
Rise in RTP Observed for RTP/PIA system.

Rise in RTP for naphthalene observed with RTP/FIA system
for A) 140 11M, B) 100 uM, C) 30 11M, D) 10 11M,
E) 3 11M, and F) 1 11M solutions. Time is relative to injection

of sample.

Calibration curve for naphthalene by RTP/PIA. RTP intensity
is in arbitrary units.

Calibration curve for 2-bromonaphthalene by RTP/PIA. RTP
intensity is in arbitrary units.

Rise in RTP Observed for 100 11M biphenyl.

Fit of rise in RTP for 100 1.1M biphenyl tO an exponential
equation.

xi

92

94

99

101

103

104

106

107
109
111
113

115

116

118
119

120

Figure 7.15

Figure 7.16

Figure 7.17

Figure 7.18

Figure 9.1.

Figure 9.2.

Calibration curve for biphenyl by RTP/FIA. Intensity of
100 11M sample obtained from equation 7-4. RTP intensity is
in arbitrary units.

Rise in RTP for pyrene Observed with RTP/FIA system for

A) 200 11M, B) 100 1.1M, C) 30 M, D) 10 11M, E) 3 11M,
and F) 1 11M solutions. Time is relative to injection of
sample.

Fit of rise in RTP for pyrene samples to equation 7-4. See text
for explanation. Curves correspond to A) 200 11M,

B)10011M,C)3011M,D)1011M,E)3|.1M,andF)111M

pyrene sample concentrations.

Calibration curve for pyrene by RTP/FIA. RTP intensities are
obtained from fit of data in Figure 7.17 to equation 74.

Manifold of the Optical comparator showing configuration of
the source, fiber optic, and photodiodes. The grooved tracks
minimize scratches in the bOttom of the analysis cups.

Circuit schematic for optical comparator circuit. Light from

LED4 illuminates photodiodes PDl and PD2 through the
bifurcated fiber optic.

xii

122

123

125

126

136

138

CHAPTER I
INTRODUCTION

This dissertation presents work performed on two vastly different projects. The
main topic of this research is room temperature phosphorescence (RTP) spectrometry.
Phosphorescence is traditionally performed in glasses at 77 K to avoid collisional
deactivation Of the excited molecule. RTP was first reported for molecules adsorbed on
a solid support. Schulman and Walling [1,2] are commonly attributed with the
discovery of RTP. Since then, the pioneering work of Winefordner [3] has developed
RTP into a viable analytical technique. Recently, several Other RTP methods have been
reported which enable phosphorescence measurements to be made in fluid solutions.
These techniques can be automated easily because of the fluid sample matrix employed.
The research on RTP in this dissertation focuses on the development of an automated
system for performing RTP in fluid solution. Flow injection analysis is used for sample
introduction because of its versatility, speed, and the simplicity of the equipment
required.

Chapter II is a fundamental background of molecular luminescence in solution.
The chapter is divided into three main categories: absorption of light, molecular
fluorescence, and molecular phosphorescence. These are described to provide basic
principles which can then be assumed in subsequent chapters.

Room temperature phosphorescence is discussed in Chapter HI. A brief
historical survey traces the development of several RTP techniques. These techniques
are then discussed individually. Methods commonly used to deoxygenate solutions

prior to RTP analysis are then introduced.

1

Chapter IV describes the technique of flow injection analysis (FIA). The
underlying principles of FIA, sample injection, controlled dispersion, and reproducible
timing, are described. Two related applications of FIA, stopped-flow and luminescence
detection, are then discussed. This discussion provides a background for techniques
applied in later chapters.

The instrumentation used in this research is described in Chapter V. The
development of a home-built spectrofluorometer is described first. Studies performed to
characterize the dominant source of noise in the instrument are then discussed. A
Perkin-Elmer model LS-5 spectrofluorometer was also used in this research and is
described in this chapter.

Chapter VI presents specrra of the compounds used in this research.
Fluorescence spectra are obtained to determine the Optimum excitation wavelength for
each compound. RTP spectra of these compounds are also described. The spectra
obtained are presented to show the characteristic phosphorescence emission bands.
These spectra are then used to determine the emission wavelength which yields the
maximum phosphorescence intensity. Phosphorescence lifetimes Of selected
compounds are deterrrrined as a means to confirm the presence of RTP.

Experiments tracing the development of a system used for RTP with FIA sample
introduction are described in Chapter VII. Studies to determine the best carrier to use in
the flow system are discussed along with experiments to test various deoxygenation
methods. These studies resulted in the development of a system for performing RTP by
FIA. Ideas for future projects are given in Chapter VIII.

The second project presented in this dissertation is the development of a dual
beam fiber Optic-based photometric comparator. This work was performed as a
consulting project for a local business. This instrument was developed as an alternative
to optical comparison of reference and sample solutions. The instrument performs a

direct comparison of reference and sample solutions and indicates if the transmittance

of the sample is less than, greater than, or approximately equivalent to the reference.

The comparator was designed to be used by non-technically oriented persons.

CHAPTER I REFERENCES

Schulman, E.M.; and Walling, C., Science, 1972, 178, 53
Schulman, E.M.; and Walling, C., J. Phys. C hem, 1973, 77, 902

Paynter, R.A.; Wellons, S.L.; and Winefordner, J.D., Anal. Chem, 1974, 46,
736

CHAPTER II
MOLECULAR LUMINESCENCE IN SOLUTION

This chapter is a discussion of the photOphysical processes that lead to molecular
luminescence in solution. These are described here so that familiarity with these tepics
can be assumed in subsequent chapters. Three topics are discussed; absorption of light,
molecular fluorescence, and phosphorescence. The absorption of light by a molecule is
one way which the molecule can be excited to a higher energy level.

Fluorescence is a radiative transition from an excited energy level to the ground
state. Equations are derived which show the relationship between the fluorescence
intensity and the concentration of the analyte. The lifetime of a fluorescent molecule is
then described as are methods for determining fluorescent lifetimes. The effect of
quenchers on fluorescence and two mechanisms of quenching are presented.

The photophysical processes which lead to phosphorescence are introduced in the
final section. This section deals primarily with cryogenic phosphorescence. Equations
relating phosphorescence intensity to analyte concentration are derived. Finally, the
quenching effect of molecular oxygen and the heavy atom effect in phosphorescence are

discussed.

Absorption of Light

Molecular luminescence spectroscopy is based on the emission of light by
electronically excited molecules. The excitation process involves the interaction of the
electric field associated with the exciting radiation with the loosely bound 1: or

nonbonded electrons Of the absorbing molecule. Through this interaction the electronic

5

distribution of the absorbing molecule is distorted which causes energy to be absorbed
from the electric field of the excitation beam. The energy necessary for absorption
depends on the types of bonds available in the molecule. From the standpoint Of organic
molecular spectroscopy, the most important element in the absorption of radiation in the
ultra-violet and visible region is the C=C double bond. If a molecule absorbs a photon,
an electron in the It cloud of this bond can be promoted to the n: anti-bonding orbital
(rt -) 1t*).

Because of the quantization of molecular energy levels, excitation of a molecular
species will occur only if the energy associated with the excitation radiation corresponds
to the difference in energy between the ground electronic state and an electronically
excited state of the absorbing molecule. The intensity of radiation absorbed by the
molecule (1,) can be calculated by the Beer-Lambert law:
bc) <2-1)

where I0 and It are the incident and transmitted light intensities, respectively, a is the

I,,=::Io-I,;===ro(1.6e

molar absorptivity of the molecule, and C is the molar concentration of the absorber in
pathlength b of the sample. The energy of radiation required to excite the species from
one energy level to an excited energy level (AB) is given by the Planck frequency
relation:

be
AB := hv :3 "X- (2’2)

where v is the fi'equency of the light of wavelength A, and c is the velocity of light. In

molecules, each electronic state has several associated vibrational levels (Figure 2.1). In
the ground electronic state, virtually all molecules occupy the lowest vibrational level at
room temperature. Most polyatomic molecules in the ground state have an even number
of electrons with their spins paired. The spin of the electron, s, can be equal to +1/2 or
-1/2. When the electron spins are paired, this is termed a singlet electronic state. The
lowest electronic singlet state is the So state. When a molecule absorbs radiation of

sufficient energy it can be excited vibrationally as well as electronically to energy level

Figure 2.1

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

VR
3
__ r
_111xww
“3 VR
' Isc
K _, VR
T1
A IC F' "
i, 131: P
v I

 

 

 

 

 

 

 

 

 

 

 

 

Gemeralized Jablonski diagram. A, absorbance; F, fluorescence; P,
phosphorescence; VR, vibrational relaxation; IC, internal conversion;
ISC, intersystem crossing; So, 8 , and 82 are the ground and excited
singlet states, respectively, and 1 is the excited triplet state.

Sn; the value of n refers to the elecrronic states 1, 2, etc. This transition occurs in
approximately 10"15 5. Most absorption transitions originate from the lowest vibrational
level of the ground electronic state. This ground-tO-excited singlet transition is
responsible for the ultraviolet and visible absorption spectra observed for molecules.

During the time that the molecule is in the excited state, some energy in excess of
the lowest vibrational energy level of that excited State can be rapidly dissipated by a
radiationless process known as vibrational relaxation. This excess energy is absorbed by
the solvent molecules in inelastic collisions with the excited molecule which occurs in
10'11 to 10'10 s. The excited molecule also can undergo a second type of radiationless
transition from an upper electronic state to a lower one by a process known as internal
conversion. This process can occur when a highly excited vibrational level of a lower
singlet overlaps the occupied level. Internal conversion occurs generally in 10.12 s.

If the excess energy of the excited state is not dissipated by collisions with other
molecules, the electron may return to the ground electronic state with the emission of
radiant energy. This transition, 81 —-) So, is known as fluorescence and occurs in 10'10
to 10.6 5 following excitation. The emission usually results from the lowest excited
electronic state 81. This is known as Kasha’s rule [1].

Intersystem crossing is another pathway which competes with fluorescence for
deactivation of the excited singlet state. Intersystem crossing necessitates a change in the
spin angular momentum which can occur when spin-orbital coupling is present. This
change results in parallel electron spins, thereby causing the molecule to be in a triplet
state (T n). Molecules in upper vibrational levels Of the triplet state undergo vibrational
relaxation and internal conversion to the lowest vibrational level of the lowest triplet
state, T1. These molecules can then return to the ground state by triplet-singlet
intersystem crossing (quenching) or by the emission of light. This emission of light
accompanying the T1 —> So transition is termed phosphorescence. This process is a "spin

forbidden" transition and is characterized by lifetimes on the order Of 10'4 to 10 s. The

two radiative transitions from excited energy levels, 81 and T1, to the ground state, So,

are discussed in more detail below.

Elam

Fluorescence is the radiative transition between an excited (Sn) and ground (So)
singlet state. This transition is between two levels of the same spin multiplicity and
occurs 10"10 to 10"6 s after excitation. The fluorescence most commonly observed in
solutions is called Stokes fluorescence. The photons emitted are less energetic than the
absorbed photons because of vibrational relaxation and internal conversion processes.

If thermal energy is applied to an excited state molecule, population of higher
energy vibrational levels will result. Emission at a higher energy than the absorbed
photons may then be observed. This is anti-Stokes fluorescence which is often observed
in dilute gases at elevated temperatures. Because of the large energy difference between
So and SI, essentially no fluorophores will populate 81 as a result of thermal energy.

Resonance fluorescence occurs when the energy of the emitted radiation is the
same as the absorbed energy. This type of fluorescence is not normally observed in
solutions because of intense Rayleigh scattering at the excitation wavelength. However,
it does occur in gases and crystals.

The fluorescence intensity observed (If) can be derived from the Beer-Lambert
law, (eq. 2-1):

If = 10%“ - ”(9)800. (2-3)

where f(0) represents a geometrical factor that is determined by the solid angle Of the

c-ebC

fluorescing radiation viewed by the detector, and g0») defines the efficiency of the
detector as a function of wavelength of the fluorescence incident on it. The fluorescence
quantum yield, of, is the fraction of absorbed photons converted to fluorescence
photons. This quantity is also related to the rate constants for fluorescence and

competitive deactivation processes, kf and kd, respectively, by

10

¢f= “12:51; (2'4)

The fluorescence quantum yield may approach unity in unique cases.

The exponential term in equation 2-3 can be expanded by the series

approximation to yield
Io¢f<1c- éEbC>f<e>go> = Iotfebc - (£13,312 +
my? +5fig >f<e>g0~> <2-5)
For dilute solutions where the absorbance, ebC is small, equation 2-5 reduces to
If = Io¢ff(9)g(7\)8bC (2-6)
or
If = ofKC (2-7)

where K = Iof(6)g(2.)eb. Therefore, at low analyte concentrations, where the

absorbance of the solution is less than 0.005 in a 1 cm pathlength cell, the fluorescence
intensity is linearly related to the analyte concentration to within a 1% deviation [5]. At
higher concentrations, scattering at the emission wavelength as well as the inner-filter
effect cause deviations from linearity. The inner-filter effect is caused by significant
absorption of the excitation radiation, or by significant absorption of the emitted
fluorescence when these processes occur within the volume element of illumination and

observation of the cell.

Fluorescence Lifetimes
The lifetime of the lowest excited singlet state (If) is also an important property
of fluorescing molecules. The fluorescence lifetime is defined as
tf = 1/(kf + kd) (2-8)
and is the time required for the fluorescence intensity to fall to l/e of its initial intensity.
The fluorescence intensity is related to the lifetime, If, by

If = Leona-"’1‘f ), (2-9)

11

where t = 0 is defined as the end of the excitation pulse and Ifo is the fluorescence
intensity at the time the excitation source is turned off. The fluorescence quantum yield
and If are related by

1if = 915%, r (2-10)
where rm f = llkf is the natural lifetime of the excited state. This represents the
lifetime of the fluorescing molecule if fluorescence were the sole pathway for
deactivating the lowest singlet state.

The decay of fluorescence emission is predicted to be an exponential as shown in
equation 2-9. Information that is useful in fundamental studies of energy transfer and
quenching can be provided by lifetime measurements. The luminescence lifetime can be
used to enhance selectivity or facilitate multicomponent determinations.

The primary goal of an experimental approach to lifetime determination is to
obtain data reflecring the time dependence of the decay of the luminescence. For a
sample consisting of only one fluorescing compound, the time-decay profile is described
by equation 2-9. This equation can be linearized by taking the natural logarithm of both
sides to obtain

maf) = 1110,50) - um, (2-11)
Therefore, a plot of ln(If) vs t should yield a slope of -l/tf. This is called the log slope
method and assumes that the luminescence signal decays to zero at infinite time [3]. The
contributions of the dark signal, background luminescence, and electronic offsets are
assumed to be negligible with respect to the analyte luminescence signal.

If the dark current term is not negligible, the Guggenheim method [4] may be
used to accurately linearize first-order kinetic data. In this method, the natural logarithm
of the difference in luminescence signals between pairs of data points along the decay
curve, separated by a constant time interval, are plotted against time.

ln(If’ - If) = -t/‘tf + constant (2-12)

12

where If’ is the fluorescence intensity made at time t+ A, and A is a constant time
increment. For best results, A should be two tO three lifetimes. The lifetime Of the
analyte is obtained from the negative reciprocal Of the slope.

The pulse excitation method involves monitoring the luminescence decay after the
analyte is excited with a short duration radiative pulse. If a pulsed source is not used the
radiation from a continuous source must be gated with an external modulator. The true
decay of the luminescence is Observed if the excitation pulse width is much shorter than
the luminescence lifetime [5]. Otherwise deconvolution methods must be used to obtain
the luminescence decay.

Time-correlated single-photon counting is a widely used method for lifetime
measurements. The time between the excitation pulse and the arrival of the first
fluorescence photon is measured. The time of photon emission, although random, is
weighted according to the decay kinetics of the lumiphore. Measuring the emission time
for a large number of excitation pulses enables the fluorescence decay curve to be
reconstructed. A true decay curve is obtained when the probability of Observing more
than a single luminescence photon per excitation pulse is very small. An average count
rate of 0.01 - 0.05 photons emitted per excitation pulse is used. Although this technique
has a resolution of about 0.2 ns, the low repetitive rate of the pulsed source and the large
number of counts that must be accumulated require several hours to accumulate data.

Another class of lifetime measurements is based on the use of a continuous,
sinusoidally modulated excitation source and phase sensitive detection apparatus [6].
This is called the Phase Shift Method. The finite lifetime Of the excited state causes the
modulated emission to be shifted in phase by an angle q) relative to the excitation. The
fluorescence lifetime 115 can be calculated from the phase shift and angular modulation
frequency a) (a) = 21tf where o is the linear modulation frequency)

If = m'ltanop) (2-13)

or from the angular modulation frequency and the demodulation factor m (m = cos(¢>)),

13

if = m’1~1(m‘2 - 1) (2-14)
The quantities d) and m can be measured relative to a signal from a scattering solution.
In this case, 111 is the ratio of the modulation depth of the ac waveform of the fluor0phore
to that Of the scattering solution [5]. Alternatively, (b and m can be measured relative to a

reference fluorophore of known lifetime.

Quenching Effects on Fluorescence

Quenching is the reduction of the fluorescence by competing deactivation
processes. This results from interactions between a fluorophore and another species
present in the system. Two types of quenching are recognized: static and dynamic
quenching.

Static quenching occurs when an interaction takes place between the potentially
fluorescent molecule, which is in the ground state, and the quenching species to form a
nonfluorescent complex [7]. The efficiency of this process is governed by the formation
constant of the complex as well as the concentration of the quencher. The association

constant for the formation Of the complex is given by

Ks = 1191th (2‘15)

where [FQ] is the concentration of the complex, [F] is the concentration of the non-
complexed fluorophore, and [Q] is the concentration of the quencher. Since the complex
species is nonfluorescent, the fraction of fluorescence which remains is given by
f = If/If' 0’ where If, 0 is the fluorescence intensity in the absence of quencher. The total
concentration of the fluorophore [F] o is given by
[Flo = [F] + [FQ]. (2-16)

If equation 2—16 is substituted into equation 2-15 and fluorescence intensities are used
instead of fluorophore concentrations, the result is

If, 0/If = 1 + K3[Q] (2'17)

14

It can be seen from equation 2-17 that there is a linear relationship between the ratio Of
fluorescence intensities, in the absence and presence of quencher, and the concentration
of the quencher.

Dynamic quenching occurs when the quenching species and the excited molecule
collide, and energy is transferred with no resulting radiation. This effect is summarized

in the Stern-Volmer equation:
_ - _,1_.._ _
(ii/(Po - f +kqt0-[Q] (2 18)
where kq is the bimolecular quenching constant, 90 and o are the quantum yields of
fluorescence in the absence and presence of the concentration of quencher, respectively,

and to is the lifetime of the fluorophore in the absence of the quencher. This equation

can be expressed as
If, o/If a 1 + quo[Q] = l + KD[Q]. (2-19)

where KO - kqto is the Stern-Volmer quenching constant. Quenching data are often
presented as a plot of If, o/If versus [Q] since equation 2-19 predicts a linear
relationship. In such a plot the y intercept is unity and the slope is equal to KD. A linear
Stern-Volmer plot generally indicates a single class of fluorophores, all of which are
accessible to the quencher [8].

It must be noted that the dependence of 115' 0/1 5 on [Q] is identical for both static
and dynamic quenching, equations 2-17 and 2-19. Fluorescence quenching data obtained
from intensity measurements alone cannOt determine whether static or dynamic
quenching is occurring. Fluorescence lifetime measurements are the most definitive
method for distinguishing the quenching mechanism. In static quenching the complexed
fluorophores are nonfluorescent, and the only fluorescence observed is from the non-
complexed fluorophore. Therefore, the Observed lifetime is unaffected by the quencher.
However, in dynamic quenching, the lifetime will decrease because quenching is an

additional process which depopulates the excited state without the emission Of radiation.

14

It can be seen from equation 2-17 that there is a linear relationship between the ratio of
fluorescence intensities, in the absence and presence of quencher, and the concentration
of the quencher.

Dynamic quenching occurs when the quenching species and the excited molecule
collide, and energy is transferred with no resulting radiation. This effect is summarized

in the Stern-Volmer equation:

W90 = Trig-gram (2-18)
where kq is the bimolecular quenching constant, 90 and (b are the quantum yields of
fluorescence in the absence and presence of the concentration of quencher, respectively,
and to is the lifetime of the fluorophore in the absence of the quencher. This equation
can be expressed as

If, o/If a l + kc110[Q] = 1 + KD[Q]. (2-19)
where KD :- quo is the Stern-Volmer quenching constant. Quenching data are often
presented as a plot of If, o/If versus [Q] since equation 2-19 predicts a linear
relationship. In such a plot the y intercept is unity and the slope is equal to KD. A linear
Stem-Volmer plot generally indicates a single class of fluorophores, all of which are
accessible to the quencher [8].

It must be noted that the dependence of If, o/If On [Q] is identical for both static
and dynamic quenching, equations 2-17 and 2-19. Fluorescence quenching data obtained
from intensity measurements alone cannot determine whether static or dynamic
quenching is occurring. Fluorescence lifetime measurements are the most definitive
method for distinguishing the quenching mechanism. In static quenching the complexed
fluorophores are nonfluorescent, and the only fluorescence observed is from the non-
complexed fluorophore. Therefore, the observed lifetime is unaffected by the quencher.
However, in dynamic quenching, the lifetime will decrease because quenching is an

additional process which depopulates the excited state without the emission of radiation.

15

Absorption spectra can also be used to distinguish static and dynamic quenching
of the fluorophore. Collisional quenching only affects the excited states Of the
fluorophore; therefore, no changes in the absorption spectra are predicted. However,
complexation by static quenching will result in perturbations in the absorption spectrum
of the fluorophore [8].

Dynamic and static quenching also can be distinguished by the effect of
temperature on the fluorescence intensity. Because dynamic quenching depends upon
molecular diffusion, higher temperatures will result in larger diffusion coefficients which
will decrease the observed fluorescence. In contrast, increased temperatures are likely to
decrease the stability of the fluorophore-quencher complex in static quenching. This will
then lower the value of the static quenching constant and result in an increased

fluorescence intensity being observed at higher temperatures.

W

As was mentioned earlier, phosphorescence is the radiation emitted when a
molecule in the lowest vibrational level of the first excited triplet state returns to the
ground singlet state, T1 —) So. Population of the triplet state occurs via intersystem
crossing from an excited singlet state (see Figure 2.1). Once the molecule is in the triplet
state it will undergo vibrational relaxation and internal conversion to the lowest
vibrational level. This is also consistent with Kasha’s rule.

In the ground state Of unsaturated organic compounds, two 11: electrons with
antiparallel spins occupy the orbitals of lowest energy. When a molecule is raised to an
excited energy level, an electron is promoted to an orbital that was previously
unoccupied. In this instance, the spin-angular momenta of the two electrons are no
longer restricted by the Pauli principle and their spins may be either antiparallel (singlet
state) or parallel (triplet state). At the instant of excitation, only electrons are

reorganized; the heavier nuclei keep their ground state geometry. This is a result of the

l6

Franck-Condon principle which states that electronic transitions occur without a change
in position of the nuclei.

Intersystem crossing can be made favorable when there is strong spin-orbital
coupling in which the orbital motion of the electron induces a magnetic field that
interacts with its spin magnetic moment. This interaction "flips" the spin of the
electron [9]. This effect is very weak unless the electron has a high probability density
near the nucleus. Typically only atoms with high nuclear charges (heavy atoms) exhibit
appreciable spin-orbital coupling. This heavy atom effect is partially responsible for the
population of the triplet state which leads to phOSphorescence. Because of the spin
forbidden nature of the T1 -) $1 transition, the lifetime of the triplet state can be quite
long and is typically 10'4 to 10 5. Because of this long lifetime, the triplet state is readily
quenched by collision with solvent molecules or dissolved oxygen. However, in a rigid
matrix it can survive for longer periods of time and lose energy only by the emission of
radiation.

In phosphorescence there are many nonradiative processes which compete for
deactivation of the triplet state. The probability of radiative or nonradiative transitions
can be described by their corresponding rate constants: rate constant for fluorescence, kf;
rate constant for fluorescence quenching, qu; rate constant for intersystem crossing,
k 13c; rate constant for phosphorescence, kp; and rate constant for phosphorescence

quenching, kpq. The efficiency for triplet state formation from $1 is

kia

a 2-20
kisc +kf +qu ( )

 

£1. =
Therefore, the phosphorescence quantum efficiency can be defined as
u =f - ‘Sn- -.. (241)
p T EP +kpq

01'

1" kp-
¢P kisc +kf +ltfq 12p + kpq ( )

 

17

The relatively long lifetime Of the triplet state renders it quite susceptible to the
many deactivations processes competing with phosphorescence. In the gas phase and in
solution kpq » kp, the phosphorescence yield is subsequently very low and op becomes
vanishingly small. Low temperature rigid matrices, high viscosity solvents, or solid
substrates can be used to decrease the quenching processes and consequently increase the
phosphorescence yield.

The relationship between the phosphorescence intensity and the concentration of
the phosphor is similar to that obtained for fluorescence (see eq. 2-7):

IF = K’opC (2-23)
where K’ a: Iof(0)g()u)8b. Therefore, the phosphorescence intensity is linearly related

to the concentration of the phosphor.

Phosphorescence Lifetimes
The phosphorescence lifetime is defined similarly to that of fluorescence
‘tp = 1/(ltp + kd) (2-24)
and is the time required for the phosphorescence intensity to decrease to l/e of its
original intensity following excitation. The rate constant kd represents the summation of

all nonradiative rate constants. The phOSphorescence intensity at time t is related to IP

by

Ip -.- 1p°(c"/’~'p) (2-25)
where Ip is the phosphorescence intensity at time t and IPO is the phosphorescence
intensity at t = 0.

Phosphorescence can be differentiated from fluorescence by its longer lifetime.
Fluorescence lifetimes of organic molecules are on the order of 10'10 to 10'6 5 whereas
phosphorescence lifetimes range from 10"4 s to 10 s. This difference is the basis of the
time- and phase-resolved techniques that are used to differentiate components of a

mixture exhibiting similar spectra, but different lifetimes.

18

Oxygen Quenching

The triplet nature of the ground state of molecular oxygen contributes
significantly to the collisional deactivation of phosphorescence [10]. Several theoretical
models pertaining to the effect of oxygen on phosphorescence have been proposed.
McGlynn, Azumi, and Kinoshita have treated the effect of oxygen as an intermolecular
exchange interaction [9]. A mechanism has been described by Murrell by which the
enhanced ground state to triplet state absorption intensity is borrowed from a charge-
transfer transition of the aromatic-Oz complex [11]. The concept of a "contact charge
transfer" mechanism has been developed by Tsubomura and Mulliken [12]. In this
mechanism the interaction between oxygen and the aromatic compound can occur in the
absence of definite complex formation. A theoretical study of the quenching of the
triplet state by molecular oxygen was performed by Kawaoka et a1. where it was
concluded that the most likely quenching mechanism deactivates the molecule from the
triplet state to the ground singlet state while oxygen undergoes a transition from its
ground triplet state to an excited singlet state [13].

The long triplet lifetime of phosphorescence induces an excessive oxygen
quenching effect. Although oxygen quenches fluorescence, the effect is not nearly so
dramatic as is Observed in phosphorescence. The collisional quenching mechanism can
be described by the Stern-Volmer equation for phosphorescence (see eqns. 2-18, 2-19).
For a given quenching process with a quenching rate constant kq = 108 mol L'1 3*, a
quencher concentration [Q] = 10.3 M and a fluorescent lifetime If = 10'8 s, the emitted
fluorescence intensity will decrease by only 0.1%. Under the same conditions, with a
phosphorescence lifetime of 1p 2 10'1 s, the emitted phosphorescence intensity will
suffer a decrease Of 99.9% . Therefore, for oxygen sensitive systems following the Stem-
Volmer equation, it is essential that the sample be deoxygenated prior to the

phosphorescence intensity measurement.

19

Heavy Atom Effect in Phosphorescence

It has already been noted that the heavy atom effect is the mechanism by which
the probability of the 81 —> T1 intersystem crossing is enhanced. The heavy atom effect
was first reported by Kasha where a mixture Of two colorless pure liquids,
a—chloronaphthalene and ethyl iodide, produced a yellow color with no chemical
reaction occurring [14]. The color effect was due to the development of the 81 -> T1
absorption band. He attributed this effect to collisional perturbation of spin-orbital
coupling. This phenomenon, which is known as the "external heavy atom effect," is
distinguished from the "internal heavy atom effect" in which the heavy atoms are
chemically bonded with the compound [15]. The external heavy atom effect is not as
pronounced as internal heavy atom coupling which is expected [16,17]. The internal
heavy atom effect increases the phosphorescence emission by increasing the S1 —-> T1
intersystem crossing rate and also by reducing collisional deactivation due to decreased
phosphorescence lifetimes.

The external heavy atom effect generally produces the following effects [9]: an
increase in the So -9 T1 absorption, a decrease in the phosphorescence lifetime, a
decrease in the fluorescence quantum yield, of, and an increase in the phosphorescence
quantum yield, ¢p- Although the heavy atom effect increases the rates of both T1 -) So
intersystem crossing and T1 -) So phosphorescence, the effect is usually greater on the
latter producing a net increase in ¢p [18].

Two major coupling mechanisms have been proposed to explain the external
heavy atom effect for aromatic compounds. Robinson [19] proposed an "exchange
mechanism" in which the perturbed triplet state borrows radiative intensity from a spin-
forbidden transition that is usually localized on the heavy atom. The extent of mixing of
the heavy atom triplet states depends on the extent of exchange interaction between the
triplet electron pair Of the aromatic molecule and the electrons associated with an excited

triplet state of the heavy atom. Murrell [20] and Tsubomura and Mulliken [12] proposed

20

a "charge-transfer" mechanism which involves interaction of the excited triplet states
with the charge-transfer-excited singlet state of the aromatic-heavy atom complex.

A major disadvantage of using alkyl halides to obtain the heavy atom effect is that
they begin to absorb strongly in the region from 300-350 nm. Therefore, they can absorb
a substantial fraction of the energy required to promote the formation of the T1 state

which leads to phosphorescence.

10.

ll.
12.
13.
14.
15.
16.
17.
18.
19.
20.

21

CHAPTER II REFERENCES

Kasha, M., Diss. Far. Soc., 1950, 9, 14

Wehry, E.L.; in Guilbault, G.G., Practical Fluorescence: Theory, Methods, and
Techniques; Marcel Dekker: NY, 1973, Chapter 3

Demas, J.N., Excited State Lifetime Measurements, Academic Press; NY, 1983,
p. 75

Guggenheim, E.A., Philos. Mag., 1926, 2, 538

Ingle, J.D.; Crouch, S.R., Spectrochemical Analysis, Prentice-Hall, 1988;
Chapter 12

Lakowicz, LR. Principles of Fluorescence Spectroscopy; Plenum Press: NY,
1983, Chapter 3

Schulman, SC. in Molecular Luminescence Spectroscopy: Methods and
Applications: Part 1; Elving, P.J.; Winefordner, J.D.; Kolthoff I.M., Editors,
John Wiley & Sons, NY, 1985

Lakowicz, J.R, op. cit, Chapter 9

McGlynn, S.P.; Azumi, T.; Kinoshita Molecular Spectroscopy of the Triplet State,
Prentice-Hall, Englewood Cliffs, NJ., 1969

Birks, J.B. Photophysics of Aromatic Molecules, Wiley-Interscience, London,
1970, pp. 499-504

Murrell, J.N., Mal. Phys., 1960, 3, 319

Tsubomtna, H.; Mulliken, R.S., J. Am. Chem. Soc., 1960, 82, 5966

Kawaoda, K.; Khan, A.U.; Kearns, P.R., J. Chem. Phys., 1967, 1842

Kasha, M., J. Chem. Phys., 1952, 20, 71

McClure, D., J. Chem. Phys., 1949, I 7, 905

McGlynn, S.P.; Sunseri, R.; Christodouleas, 1., J. Chem. Phys., 1962, 37, 1818
Nag-Chaudhuri, J.; Stoessel, L.; McGlynn, S.P., J. Chem. Phys., 1963, 38, 2027
Giachino, G.G.; Kearns, D.R., J. Chem. Phys., 1070, 52, 2964

Robinson, G.W., J. Chem. Phys., 1967, 46, 572

Murrell, J.N., Mol. Phys., 1960, 3, 319

CHAPTER III
ROOM-TEMPERATURE PHOSPHORESCENCE

This chapter is an introduction to room temperature phosphorescence (RTP).
The various RTP techniques to provide a controlled rrricroenvironment for the
luminphore are introduced and discussed individually. These RTP techniques are solid-
surface, sensitized, microcrystalline and colloidal, cyclodextrin-induced, and micelle-
stabilized.

Following the discussion of the RTP techniques, the methods of deoxygenation
commonly used with RTP are introduced: inert gas purging, freeze-pump-thaw, vacuum

preparation, the use of semi-permeable membranes, and chemical scavenging.

Infirfiufiigm to Rmm Tgmmrgturg Phgsgphorgsmgngg

Traditionally phosphorescence is performed in rigid media at liquid nitrogen
temperature (77 K) to minimize the effects of quenchers in solution. The lack of
widespread use of low temperature phosphorescence (LTP) can be attributed to the need
for cryogenic equipment and the problems involved with introducing the sample into the
phosphorimeter. Sample preparation involves lowering a long quartz capillary cell into
a quartz Dewar flask filled with liquid nitrogen. The rate of cooling of the sample cell
and the chemical nature and composition of the solvent system will determine whether
the cooled matrix will be a clear glass, a cracked glass, or a snow. Thus, the selection of
the proper solvent is critical; it should form a clear glass at 77 K and have a low
phosphorescence background. The analyte should be readily soluble in the solvent at

77 K. The most common solvent for nonpolar compounds is a mixture of ethanol,

22

23

isopentane, and diethyl ether in a volume proportion of 2:5:5. This solvent mixture is
commonly known as EPA [1]. The solvents should be purified to remove compounds
that phosphoresce.

The major advantages of LTP are the low detection limits (in the nanomolar
range), the linearity of calibration curves (linear over several decades of concentration),
and the high degree of selectivity. This selectivity results from the choice of excitation
and emission wavelengths and temporal discrimination which can be used to distinguish
phosphorescence from fluorescence and analyte phosphorescence from that of other
phosphorescent molecules. ’

The time required for sample preparation for LTP measurements and the
equipment required for working at cryogenic temperature has prompted interest in RTP
techniques. Perhaps the first observation of RTP was made by Millson in 1944 [2]. He
noted that various cottons, and many other fibers, phosphoresced at room temperature
when "bone dry". In 1958 Brown [3] observed RTP of naphthirrrinazole on a paper
chromatogram. The brilliant yellow phosphorescence lasted for several seconds after
the removal of the excitation radiation. This report may have been overlooked in the
literature because this observation was incidental to the main topic of his paper and was
not mentioned in the summary. In 1967 Roth [4] reported a study of the RTP of more
than 18 compounds adsorbed on several supporting media. The phosphorescence of all
phosphorescent molecules lasted for several seconds after withdrawal of the excitation
radiation. It is possible that other phosphorescent molecules were overlooked either
because of their shorter lifetimes or because the excitation wavelength was
inappropriate for the low-pressure mercury lamp used as a light source. In this paper,
Roth emphasized the value of the technique as a simple, selective, and non-destructive
detection method for paper chromatography and electrophoresis. Although Roth
appears to be the first to recognize the generality of RTP in an analytical context [5],

credit for the discovery of RTP of molecules adsorbed on a solid support is attributed to

24

Schulman and Walling in 1972 [6,7]. They reported RTP emission from salts of a wide
variety of polynuclear carboxylic or sulfonic acids, phenols, and amines adsorbed on
paper, silica, alumina and other supports. The phosphorescence was observed only with
ionic compounds. The technique is oxygen insensitive but requires thorough drying.
The emission and excitation spectra are very similar to those observed at 77 K, although
the spectra did exhibit a loss of fine structure and the lifetimes were shorter.

Recently, several different RTP methods have been discovered and developed
into analytical techniques. Of these, sOlid-surface RTP [8], micelle-stabilized RTP [9],
and cyclodextrin—induced RTP [10] all employ an interaction of the analyte with its
immediate environment (the substrate to which the analyte is adsorbed, or the micelle or
cyclodextrin cavity). This minimizes the quenching pathways which normally
deactivate the triplet state at room temperature by radiationless means. A fourth RTP
method, sensitized RTP, uses the energy transfer from a donor analyte to an acceptor
molecule which readily phosphoresces at room temperature [11]. Biacetyl and
1,4-dibromonaphthalene are two commonly used acceptor molecules. A fifth RTP
technique has been reported which employs microcrystalline or colloidal
suspensions [12,13]. These techniques are discussed individually below.

Solid-Surface RTP
In 1974 Winefordner et a1. [8] initiated a broad research effort to establish the

general use of RTP as an analytical technique. This group reported the first analytical
studies of solid-surface RTP and found that the limits of detection for various
naphthalene sulfonic acids on filter paper were competitive with other methods. The
linear dynamic ranges for many of the compounds were found to be limited because the
filter paper used as a solid substrate exhibited a phosphorescence background; this

raised the limit of detection and lowered the linear dynamic range of the compounds.

25

The number of ionic sites on a molecule is an important factor in producing
solid-surface RTP [14]. Although 2-naphthol has a polar hydroxyl group, RTP was not
observed for this compound adsorbed on paper. However, intense RTP was observed
for its ionic form, sodium 2-naphthoxide [15,16]. The ratio of RTP to LTP was found to
be less than one for all compounds tested, but the ratio was the highest for doubly
charged species in 1M N aOH [14]. Uncharged species such as sulfguanidine and
5-acetyluracil showed the lowest RTP/LTP ratios. This was interpreted to indicate that
hydrogen bonding was an active force in immobilizing the analyte [17]. This theory
was supported by the discovery that the RTP signal was proportional to the surface
concentration of -OH groups. Phosphorescence emission was drastically decreased for
sorrre compounds spotted on smooth, cherrrically inert substrates such as polyethylene
fiber, borosilicate fiber, or silanized paper. The silanization process reduces the number
of hydroxyl groups on the paper which diminishes the extent of hydrogen bonding
between the analyte and the support. The addition of sugars to the support paper to
enrich the hydroxyl concentration provided a substantial enhancement in the
phosphorescence signal and lengthened the observed lifetime [18]. This was supported
by the discovery that the lifetime increased when the sample was prepared from a basic
solution causing it to be deposited as the salt which can be more tightly bound.

Background luminescence has been one of the major drawbacks of solid-surface
RTP. Attempts to reduce the background emission from paper have been described by
Ward et al. [19]. Attempts at washing the contaminants from the paper have met with
little success [20,21]. Bateh and Winefordner [20] attempted to photobleach the lignins
or hemicelluloses believed to be the source of the background. They then concluded
that such a pretreatment improved the adsorption characteristics of the filter paper but
did not significantly reduce the phosphorescence background.

Nonuniformity in the surface of the filter paper substrate may also contribute to

systematic and random errors in the phosphorescence measurement. Batch and

26

Winefordner [20] conducted a comparison of various lots of filter paper and found that
the quality was sufficiently consistent for use in RTP applications. Lot-to-lot variation
with paper is seldom noticed since one RTP assay requires such a small amount of paper
(0.6 cm diameter circle) that it takes a very large number of measurements to use up an
entire lot.

The heavy atom effect has been found to significantly enhance the RTP signals
of polynuclear aromatic hydrocarbons [22]. The observed trend is T1+ > Ag+ > Pu2+ >
Hg“; thallium ion also produced enhancements on several spectral features of the
emission bands. The addition of 1.0 M NaI increased the RTP emission from
naphthalene sulfonates adsorbed on filter paper approximately 50-fold while decreasing
the fluorescence by about 95% [23,24]. A remarkable enhancement of solid-surface
RTP by 1.0 M NaI was observed for tryptophan (455-fold), its methyl ester (340-fold),
and for indole (370-fold) [25]. For a more comprehensive discussion of the external
heavy atom effect on solid-surface RTP the reader is referred to a recent monograph by
Vo-Dinh [26].

Other substrates used for solid-surface RTP include paper lint, ion exchange
paper, alumina, silica gel, asbestos, sodium acetate, sucrose, starch, chalk, inorganic
substrates, and polymer-salt mixtures. Since filter paper appears to be the most

favorable substrate [27], these other supports will not be discussed here.

Sgnsitized RTP

In sensitized RTP the analyte is detected indirectly via the phosphorescence of
an acceptor in liquid solution at room temperature. After excitation, the weakly (or
non-) phosphorescent analyte acts as an energy donor and transfers its triplet state
energy to an acceptor [28];

D(T1) + A(SO) —) D(So) + A('l‘1) (3-1)

27

In this equation D is the donor, A is the acceptor, and T1 and So are the triplet and
ground states, respectively. A suitable acceptor should have a high phosphorescence
quantum efficiency in liquid solutions, a triplet energy lower than that of the donor, and
a low molar absorptivity at the excitation wavelength of the donor. Commonly used
acceptors are biacetyl and 1,4—dibromonaphthalene. The role of the absorption spectrum
of the acceptor molecule in sensitized RTP has prompted interest in biacetyl as an
acceptor because of its very low molar absorptivity over a large wavelength region [28].

The intensity of sensitized RTP in liquid due to a donor-acceptor couple is given

by

P(sens) =klex(2.3eD[D]l)GISCDOtDAGPASpA (3-2)
where P(sens) is the photomultiplier output for uncorrected spectra, k is an instrumental
constant independent from the excitation and emission wavelengths, Iex is the intensity
of the excitation light source at the excitation wavelength, 8D is the molar absorptivity
of the donor D at the excitation wavelength, [D] is the donor concentration, 1 is the
optical pathlength, OISCD is the intersystem crossing efficiency of the donor, fit” is
the energy transfer efficiency, GPA is the phosphorescence efficiency of the acceptor,
and SpA is a dimensionless parameter related to the wavelength dependence of the
detector [28].

Sensitized RTP is an indirect method in which only the excitation characteristics
of the analyte play a role. Direct excitation of the acceptor must be avoided since it
would lead to a background phosphorescence. All analyte compounds are detected at
the characteristic wavelength of the acceptor molecule. Thus, effects of stray light and

fluorescence background are reduced [28].

Micr stallin and olloidal RTP
Colloidal or microcrystalline RTP was discovered during the course of

characterizing microcrystalline suspensions [29,30]. This effect was Observed initially

28

for biphenyl. These systems are produced by adding the solute as a concentrate
dissolved in a water-miscible solvent. With rapid addition a uniform dispersion is
obtained which can remain suspended for several weeks in a glass container. This
method requires minimal sample preparation; deaeration to remove oxygen is not
required.

The rnicrocrystals appear to act in an manner analogous to interactions in solid-
surface RTP in that the quenching pathways normally observed at room temperature are
minimized. The rrricrocrystals act as their own substrates and provide stability to the
triplet state. The lifetimes observed were shorter than for analogous solutions at 77 K or
for micelles [29,30]. However, the analytical utility of this phenomenon appears to be

limited as the concentration dependence of the RTP was found to be non-linear.

l x ri h n RTP

Cyclodextrins (CDs) are macrocyclic carbohydrates produced by the bacterial
degradation of starch. They are used in biochemistry research to mirrric enzymes [31].
The cyclodextrin molecules consist of six, seven, or eight glucose monomers arranged
in a torus and are denoted as or-, B-, and 'y-cyclodexuin, respectively. The CD has a
rigid conical molecular structure with a hollow interior. The inner diameters are 6 A for
ot-CD, 8 A for B—CD, and 10 A for y-CD [32]. A simplified Structural representation
ofa CD is shown in Figure 3.1.

The interior cavity of the CD, which is hydrophobic in nature, can contain one or
more highly energetic water molecules. This situation is thermodynamically
unfavorable partially because of the repulsive effect of the polarity of the water
molecule(s) with the CD interior. Thus, when less polar species are introduced into the
CD solution they will easily enter the cavity of the CD if they satisfy the size criterion

of fitting at least partially into the CD interior. Water in the interior of the CD will then

29

 

Figure 3.1. Simplified representation of a cyclodextrin molecule.

30

be expelled into the bulk phase. This will result in an inclusion complex of the less
polar species and the CD. Typically, a one-to-one complex is formed [33-35].

The CD provides favorable microenvironment polarity and decreased degrees of
freedom in molecular motion; it shields the excited state lurrriphore from interferences
from molecular oxygen and ionic species in the bulk solution. The first reports of
cyclodextrin-enhanced RTP (CD-RTP) appeared in 1982. Turro reported that CD-RTP
of l-chloro and l-bromonaphthalene was observed in deoxygenated solutions [36].
They found that the phosphorescence emission was substantially shielded from nitrite
quenching, and the lifetimes were enhanced by the addition of 10% acetonitrile. Shortly
thereafter, they observed CD-RTP of 4-bromo-l-naphthoyl in aqueous y-CD under 1
atmosphere of oxygen [37].

It was at this time that CD-RTP caught the interest of Cline Loves’s group who
evaluated it for its analytical applications [38,39]. They reported CD-RTP for several
non-heavy-atom containing polynuclear aromatic hydrocarbons and found that the
addition of an external heavy atom was found to be necessary to obtain CD-RTP. In
fluid solutions at room temperature, singlet emission is the dominant pathway for
deactivation without a mechanism for enhanced spin orbital coupling and in the
presence of oxygen. Cline Love and coworkers [38,39] obtained the external heavy
atom effect in the aqueous cyclodextrin solutions by adding 1,2-dibromoethane (DBE).
Apparently DBE is small enough to be able to fit easily into the CD cavity and
sufficiently nonpolar that it partitions readily out of the polar aqueous phase into the
CD. The proposed mechanism for CD-RTP involves the formation of a termolecular
complex between the CD, the haloalkane, and the lumiphore [38].

Although other nonionic sources of bromine may be more effective in inducing
CD-RTP, they were found not to be useful due to their limited water solubilities and/or
their sizes [38]. The addition of DBE to a solution of phenanthrene in B-CD quenched

over 90% of the fluorescence and allowed the phosphorescence to be observed. It

31

should be noted here that DBE is reported to be a known mutagen and potential
carcinogen; it should be handled with extreme care [40].

Because the CD cavity has a specific volume, it is possible to discriminate
against molecules of different sizes based on the steric requirement that the lumiphore
must fit at least partially into the CD cavity. There must also be sufficient volume
available for the inclusion of DBE to enhance the triplet emission. These criteria can be
used as a selective means of analysis based on the molecules size and shape as well as
its hydrophobicity. The volume of the CD must be such that the termolecular complex
can be formed if CD-RTP is to be observed.

The ubiquitous nature of paramagnetic oxygen must be considered in fluid
solutions where there is no rigid matrix. The lumiphore/CD inclusion complex shields
the excited state from quenchers in solution. The effect of oxygen on the triplet state of
a complexed molecule was first studied by Turro for bromonated naphthalene
derivatives [37]. Cline Love has shown that 10-40% of the RTP signal remains after the
sample is aerated [38]. Therefore, CD-RTP is at least partially immune to the

quenching effect of molecular oxygen.

Micell tabilized RTP

Surfactants are molecules having a hydrophobic group and a highly charged
hydrophylic group. The word surfactant is a syncope of surface active agents. The
hydrophobic backbone of the surfactant can vary in length from eight to twenty carbons
and the hydrophylic portion can be a partially dissociable carboxylate, a fully ionized
moiety such as anionic sulfate or cationic trimethyl ammonium plus a counterion, or an
uncharged polar species such as polyethylene glycol. At low concentrations in solution
the surfactant exists mostly as monomers, although dimers, trimers, etc. can also exist.
As the concentration is increased a specific concentration is reached, the critical micelle

concentration (CMC), at which point the monomers aggregate to form a rrricelle; for

32

sodium dodecyl sulfate (SDS), a typical sufactant used for micellar assemblies in RTP,
the CMC has been found to be 8.0 x 10.3 M [41]. Micelles are roughly Spherical in
shape and consist typically of 60 - 100 monomers. The number of monomers
incorporated into a micelle is the aggregate number, for SDS the aggregate number has
been reported to be 71 [41]. The radius of the rrricelle is approximately equal to the
length of a fully extended hydrophobic group, surrounded by an outer region containing
the hydrated hydrOphylic groups and bound water. As the concentration of the sufactant
is increased above the CMC more micelles are formed. The concentration of the

micelles in solution can be calculated by

[surfactant - CMC]

aggregate number (3'3)

[micelle] =

 

The macroscopic solution properties of micellar solutions approximate a truly
homogeneous solution; they cannot be filtered using conventional methods, and they do
not cause measurable light scattering in conventional spectrometry [42].

In aqueous solution, the hydrophobic region of the sufactant forms the interior of
the micelle with the polar group oriented outward toward the polar solvent. In organic
solvents a reverse micelle can be formed where the polar group of the surfactant forms
the interior of the micelle. These moieties are represented in Figure 3.2.

The rrricelle is an organized assembly that is in dynarrric equilibrium with the
bulk solvent phase. The dissociation rate constant for surfactant monomers is on the
order of 105 - 109 s"1 and is strongly dependent on the length of the hydrocarbon chain.
The association rate constant is nearly diffusion controlled with a rate constant on the
order of 108 - 109 M“1 s‘1 [43].

Aqueous rrricellar solutions in concentrations above their CMC can increase
dramatically the solubility of hydrophobic compounds (probe molecules) which are
normally nearly insoluble in aqueous solutions. For example, the solubility of pyrene is
increased five orders of magnitude over its normal water solubility in a 0.1 M sodium

dodecyl sulfate solution [42]. The probe is in the same phase as the micelle solution

33

Figure 3.2. A) Simplified representation of a surfactant monomer.
B) Anionic micelle in aqueous solvent.
C) Reverse rrricelle.

34

which results in a thermodynamically stable system [44]. In general, the less soluble the
probe is in water, the stronger its association with the micelle aggregate.

The micellar environment tends to shield the probe molecule located within the
assembly from reaction with species located in the bulk aqueous phase [45]. The
micelle provides sufficient protection for the triplet state of the probe molecule so that
phosphorescence can be observed at room temperature. Kalyansundaram, Grieser, and
Thomas [45] reported the first observation of RTP in micelles. They attributed the
observation of Micelle-Stabilized RTP (MS-RTP) to the protective screening of the
probe by the micelle assembly. The solutions were degassed by bubbling oxygen-free
nitrogen gas through the samples for 30 minutes. Traces levels of oxygen (<10"6 M)
were found to quench effectively the phosphorescence emission because oxygen readily
penetrates the micelle. MS—RT'P was reported for several ionic micelle systems and
phosphorescent lifetimes in the millisecond range were observed.

Kalyansundaram, Grieser, and Thomas [45] also found that in anionic micelles,
hydrophilic quenchers such as Cu2+ and '11+ are located at the surface of the micelle.
This promotes efficient quenching at the micelle-water interface. The cupric ion was
found to reduce both the lifetime and the phosphorescence yield whereas the thallous
ion decreased the lifetime while increasing the yield. The heavy-metal can act as a
counterion to the anionic micelle and be held in close proximity to the negatively
charged polar group. The micelle can therefore organize the reactants (the probe and
the thallium heavy-metal) on a molecular scale which raises the effective concentration
of the heavy-metal and increases the probability of spin orbital coupling [46].
Phosphorescence of halogenated compounds showed a less pronounced external heavy
atom effect, which indicates that the halogen atom was Operating as an internal heavy
atom. Similar effects were observed by Turro et al. [47] and Humphry-Baker et a1. [48].

Humphry-Baker et al. reported dramatic increases in the phosphorescence of pyrene and

35

naphthalene in anionic surfactant solutions with Ag+ as the counterion. This indicated
that the silver counterion facilitated spin orbital coupling by the heavy atom effect.
Almgren, Grieser, and Thomas [41] used MS-RTP to study the dynamics of
solubilization of biphenyl, naphthalene, and l-methylnaphthalene in ionic micellar
solutions. The exit rates for these compounds were found to be >5 x 104 5'1. The exit
rate of 1-bromonaphthalene from SDS was found to be 2.5 x 10‘1 s'1 and the diffusion-

controlled entrance rate was 5 - 8 x 109 M'1 s'l.

They considered this to be a
representative value for other arenes.

In 1980 Cline Love, Skrilec, and Habarta [9] demonstrated the first analytical
application of MS-RTP in aqueous solution. They found linear dynamic ranges (LDRs)
for pyrene, naphthalene and biphenyl in Na/Tl and Na/Ag mixed counterion lauryl
sulfate micelles to be over 2 decades with limits of detection (LODs) comparable to
those found in EPA at 77 K. Thallium counterions were found to give consistently
higher signals and better sensitivities than the silver counterions. This parallels the
observations found for solid-surface RTP. The RTP intensity peaked at ~30% T'l+ and
~30% Ag+, but the enhancement due to the heavy atom effect was essentially complete
by 20% metal ion concentration. Cline Love et al. used a total surfactant concentration
of 0.15 M.

The effect of temperature on RTP intensities was investigated. A modest
temperature coefficient was observed, being roughly a factor of one or two per 10’C.
This indicates that modest temperature control is required for quantitative analysis.

Cline Love’s group [49] also reported MS-RTP characteristics of functionally
substituted aromatic molecules including ketones, aldehydes, alcohols, carboxylic acids,
phenols, and amines. Once again, they observed LODs that were comparable to those
obtained at 77 K. Thallium was used as the heavy atom counterion for the surfactant

since it was determined previously that thallium is more sensitive than silver as an

MS-RTP enhancer. In general, they found that the heavy atom quenched 75 - 99% of

36

the fluorescence. Phosphorescent lifetimes were found to range from 250 its to 1 ms
for the compounds tested. The relative standard deviations (RSDs) for the lifetime
determinations were found to be better than 10% with meticulous deoxygenation.
Following this work, Cline Love et al. [50] studied the dynamics of the analyte-
micelle interaction in mixed micelles containing thallium as the counterion. The

phosphorescence lifetime ‘tp in micellar solutions can be described by

mp = k- + km, - [131%th (3-4)

where k- is the rate constant for exit of the triplet probe from the micelle, kMP is the
deactivation from within the micelle, k, is the reentry rate of the probe into the micelle,
kp is the rate of deactivation outside the micelle, and [M] is the concentration of the
micelle as calculated by equation 3-3. This equation assumes that the concentration of
internal and external quenchers are negligible and that heavy atoms are not treated
explicitly as quenchers of the excited state probe. With these assumptions it can be seen
that the observed lifetime depends on k- and kMP when the surfactant, and hence the
rrricelle concentration is kept constant.

A background correction method that relies on the reversible chemically induced
quenching of the excited triplet state by molecular oxygen has also been described by
Cline Love et al. [51]. In this method, the luminescence intensities of the deaerated and
aerated sample solutions are measured using the same instrumental parameters. The
calculated difference in the two emission intensifies then yields the signal and spectra
due to the phosphorescence alone.

Woods and Cline Love investigated heavy atom and complexation effects on
MS-RTP of pyridinic nitrogen molecules [52]. They observed greater RTP
enhancement with 0.08 M 50/50 Ag“/Na+ lauryl sulfate than with the thallium
counterion. Silver forms a strong complex through the nitrogen to keep the heavy atom

in close proximity to the phosphorescent molecule.

37

Synchronous scanning MS-RTP was demonstrated by Fernia and Cline Love for
the simultaneous determination of mixtures [53]. They found that optimum results are
obtained when the total analyte concentration is much less than 5 x 10"5 M because of
triplet transfer causing errors in the measurements. The specific threshold value
depends on the donor-acceptor system as well as the phosphorescent lifetimes of the

molecules-

mxygenation Methgs

The high efficiency of molecular oxygen as a triplet quencher necessitates
deoxygenation to observe RTP in fluid solutions. Several deoxygenation methods have
been developed. Many of these methods were developed primarily for fluorescence or
electrochemical detectors where dissolved oxygen either quenches the analyte signal or

causes a high background current.

Inert Gas Purging

Kalyanasundaram et al. [45] bubbled all solutions with nitrogen having an
oxygen concentration less than 0.5 ppm for 30 minutes prior to phosphorescence
measurements. This method saturates the solution with the inert gas so that dissolved
oxygen is removed. Argon has also been used as an inert gas to deoxygenate solutions
for polarographic analysis [54]. Although argon was found to was found to be more
efficient than nitrogen in this study, widespread use of argon is probably inhibited by its
higher cost.

One obvious result of bubbling gases through a sufactant solution is foaming,
which can cause a loss of solution due to overflow. If the flow rate of the
deoxygenating gas is reduced to minimize foaming, inordinately long deoxygenation
times are required. A gas bubbling system for use with surfactant solutions has been
described by Buell and Demas [55]. This system uses a Teflon tube running down the

center of a 60' funnel with a ground glass joint The funnel, which fits into a long

38

necked cell, collects the bubbles forced out the glass tube by the purge gas. The Teflon
tube acts as a wick to drain the fluid from ruptured bubbles back into the cell. The
oxygen concentration from an oxygen saturated solution was reduced to less than
10.5 M in about 13 minutes. This design enables higher purge gas flow rates to be used
and decreases the deoxygenation time to 8 - 10 minutes since solutions are normally

saturated with air rather than oxygen.

Vacuum Degassing
A degassing method based on evacuation has been described by Battino
et al. [56]. Although this method has not been used explicitly in RTP, it was reported to
be able to thoroughly degas 500 ml of liquid in 30 minutes or less.

Freeze-Pump-Thaw
Degassing solutions by freeze-pumpthaw cycles has been described by
Takakubo and Faure [57]. This method is not time efficient because several cycles are
required to successfully degas solutions. Solutions must also be allowed to equilibrate

at room temperature since freezing causes some partitioning of the surfactant solution.

Semipermeable Membranes

A method has been described by Reim [58] for the continuous removal of
dissolved oxygen from chromatographic effluents. The method is based on the use of
tubular semipermeable membranes. Silicone rubber was used as the membrane because
of its high oxygen permeability and chemical inertness. The membrane device was
constructed from a plastic tube shell onto which polypropylene end caps were fitted
with a double O-ring seal. One end cap contained a fitting for a vacuum pump
connection and the second contained standard male chromatographic fittings through
which a tubular silicone membrane was sealed with silicone rubber sealant. The gas
flux across the membrane is proportional to the permeability of the membrane and the

oxygen partial pressure difference across the membrane walls. It is inversely

39

proportional to the thickness of the membrane wall. Several approaches to maximize
oxygen transport across the membrane were considered. Purging the shell with an inert
gas required large volumes of gas to rapidly flush the shell. Flushing the shell with
alkaline-sulfite solution increased baseline peak-to-peak noise considerably due to the
pulsing action of the pump employed. Although this method offered the maximum
pressure difference across the membrane, it was found to have the lowest oxygen
removal of the methods tested. This was believed to be due to the slow reaction kinetics
at the outer membrane wall. (The sulfite reaction for oxygen removal is discussed
later.) Evacuation of the shell was found to be the superior method. Residence times
less than one minute were found to decrease the oxygen concentration to less than

0.1 ppm.

Chemical S cavenging Methods
Several deoxygenation methods based on chemical scavenging have been
reported [59-62]. These methods reduce the oxygen dissolved in solution to water.
These techniques are discussed below.

Chromiumfll)

Rollie, Ho, and Warner described a method based on the diffusion of oxygen
across a permeable membrane into a solution of Cr2 +. The Cr2+ is formed by the
reduction of Cr3+ by a Jones Reductor Zn(Hg). The Cr2+ then reduces the oxygen to
water. This technique is used to minimize the quenching of fluorescence in a flowing
stream [59]. This method is based on the following reactions:

Zn(Hg) + zcr3+ —> zn2+ + 2Cr2“ + Hg (3-5)
4Cr2+ + O2 + 4H3O+ —> 4Cr3+ + 6H20 (3-6)
As oxygen diffuses across the membrane it is immediately reduced to water by Cr2+
(reaction 3-6). The Cr3+ formed by this reaction is then reduced by the Jones Reductor

(reaction 3-5) to form Cr2+ which can be used in equation 36 Therefore, the

40

concentration of oxygen never reaches a level which would reduce the concentration
difference across the membrane. This causes oxygen to continue to diffuse across the
membrane until the analyte solution is oxygen free. Polyethylene tubing was used as
the membrane because silicone tubing had the tendency to expand under the pressure of
the flowing solvent. The greatest oxygen removal was obtained with a 90 minute
equilibration time. Although this method has been proven effective, the inordinately
long equilibration time precludes the usefulness of this method for routine analyses.

An automated on-line deoxygenator for luminescence was then described by
Rollie, Patonay, and Warner [60]. This system reduces sample handling to a minimum
because deoxygenation and luminescence analysis occur within a closed system. The
deoxygenator consists of a 12-inch dialyzer with a 0.2-mil very low density
polyethylene membrane. With this system samples could be deoxygenated in five
minutes. Thus, the use of a low density membrane decreased dramatically the

deoxygenation time to enable reproducible deoxygenations to be performed on-line.

Zinc Column
A zinc oxygen-scrubber column for use in liquid chromatography has been
described by MacCrehan and May [61]. This method employs a "reductor" column
packed with zinc particles to reduce the oxygen to water by the following sequence of
reactions:
Zn + O2 + 2H" —) Zn2+ + H202 (3-7)
Zn + H202 + 2H" —> Zn2+ + ZHZO (3-8)
The reductor column was found to have a finite lifetime since zinc is being oxidized in
the process. The removal of oxygen by the zinc scrubber column is reported to compare

favorably with the semipermeable membrane method discussed previously [58].

41

Sodium Sulfite

A deoxygenation technique that uses sodium sulfite as an oxygen scavenger in
micellar solutions was reported by Diaz Garcia and Sanz-Medel [62]. This method
reduces the oxygen in the micellar solution and does not rely on diffusion of oxygen
across a membrane. Sodium sulfite has been used routinely in polarographic
determinations to remove dissolved oxygen from alkaline and neutral supporting
electrolytes [63]. The reduction of oxygen occurs according to

23032‘ + 02 H 2503' (39)

Reaction 3-9 reaches a steady state when the micellar phase becomes equilibrated with
the oxygen in the headspace within the sample cell. The reaction time for the oxygen
consumption is related to the sufactant concentration. Higher surfactant concentrations
took longer to deoxygenate. The optimum sodium sulfite concentration with respect to
deoxygenation time and phosphorescence intensity was determined to be 0.010 M for
0.1 M SDS solution [62]. This method is considerably easier than any aforementioned
technique and requires no special apparatus. However, there are two immediate
disadvantages of this technique that must be mentioned. The concentration of sodium
ions from the sodium sulfite will have a counter-ion effect on micelles. This will alter
the aggregate number and change the size of the micelle. It is possible that the sulfite
ion will react with constituents in the sample: The effect this has on phosphorescence

would have to be determined experimentally.

9°."9‘MPP’

10.
ll.

12.
13.
14.

15.
16.
17.
18.
19.
20.
21.
22.
23.

42

CHAPTER III REFERENCES

Ingle, J .D.; Crouch, S.R., Spectrochemical Analysis, Prentice-Hall, 1988

Millson, H.E., The Phosphorescence of Textile Fibers and Other Substances,
Calco Tech. Bull, American Cyanamid Co., 1944, 753

Brown, D.J., J. Chem. Soc., 1958, 1974

Roth, M., J. Chromatog, 1967, 30, 276

Llyod, J.B.F.; Miller, J.N., Talanta, 1979, 26, 180

Schulman, BM; and Walling, C., Science, 1972, I78, 53
Schulman, E.M.; and Walling, C., J. Phys. Chem, 1973, 77, 902

Paynter, R.A.; Wellons, S.L.; and Winefordner, J.D., Anal. Chem, 1974, 46,
736

Cline Love, L.J.; Skrilec, M.; Habarta, J.G., Anal. Chem, 1980, 52, 754
Scypinski, S.; Cline Love, L.J., Anal. Chem, 1984, 56, 322

Donkerbroek, J .J .; Gooijer, C.; Velthorst, N.H.; Frei, R.W.; Anal. Chem, 1982,
54, 891

Weinberger, R.; Cline Love, L.J., Spectrochimica Acta, 1984, 40A, 49
Weinberger, R.; Cline Love, L.J., Appl. Spectrosc., 1985, 39, 516

Wellons, S.L.; Paynter, R.A. ;Winefordner, J.D., Spectrochim. Acta, 1974, 30A,
2133

Parker, R.T.; Freedlander, R.S.; Dunlap, R.B., Anal. Chim. Acta, 1980, 119, 189
Parker, R.T.; Freedlander, R.S.; Dunlap, R.B., Anal. Chim. Acta, 1980, 120, 1
Schulman, E.M.; Parker, R.T., J. Phys. Chem, 1977, 81, 1932

Niday, G.J.; Seybold, P.G., Anal. Chem, 1977, 50, 1577

Ward, J.L.; Batch, R.P.; Winefordner, J.D., Analyst, 1982, 107, 335

Batch, R.P.; Winefordner, J.D., Talanta, 1982, 29, 713

Ward, J.L.; Bower, E.L.Y.; Winefomdcr, J.D., Talanta, 1981, 28, 119

Bower, E.L.Y.; Winefordner, J.D., Anal. Chim. Acta., 1978, 102, 1

Seybold, P.G.; White, W., Anal. Chem, 1975, 47, 1199

24.
25.
26.

27.
28.

29.
30.
31.
32.
33.
34.

35.
36.

37.
38.
39.

41.
42.
43.

45.

46.

43

White, W.; Seybold, P.F., J. Phys. Chem, 1977, 81, 2035
Meyers, M.L.; Seybold, P.G., Anal. Chem, 1979, 51, 1609

Vo-Dinh, T., Room Temperature Phosphorimetry for Chemical Analysis; Wiley
Interscicnce: New York, 1984, Chapter 2

Ward, J.L.; Walden, G.L.; Winefordner, J.D., Talanta, 1981, 28, 201

Donkerbroek, J.J.; Gooijer, C.; Velthorst, N.H.; Frei, R.W., Anal. Chem, 1982,
54, 891

Weinberger, R.; Cline Love, L.J., Spectrochimica Acta, 1984, 40A, 49
Weinberger, R.; Cline Love, L.J., Appl. Spectrosc, 1985, 39, 516
Tabushi, 1., Acc. Chem. Res, 1982, 15 , 66

Scypinski, S.; Cline Love, L.J., Anal. Chem, 1984, 56, 322

Saenger, W., Angew. Chem, Int. Ed. Engl., 1980, 19, 344

Bender, M.L.; Komiyama, M., Cyclodextrin Chemistry; Springer-Verlag, Berlin,
1978

Cramer, F.; Saenger, W.; Spatz, H-Ch., J. Am. Chem. Soc., 1967, 89, 14

Turro, NJ .; Bolt, J.D.; Kuroda, Y.; Tabushi, I., Photochem. Photobiol., 1982,
35 , 69

Turro, N.J.; Cox, G.S.; Li, X., Photochem. Photobiol., 1983, 37, 149
Scypinski, S.; Cline Love, L.J., Anal. Chem, 1984, 56, 322
Scypinski, 8.; Cline Love, L.J., Anal. Chem, 1984, 56, 331

Second Annual Report on Carcinogens (NTP 81-43, December, 1981) pp. 112-
1 14

Almgren, M.; Grieser, F.; Thomas, J.K, J. Am. Chem. Soc., 1979, 101, 279
Cline Love, L.J.; Habarta, J.G.; Dorsey, J.G., Anal. Chem, 1984, 56, 1133A

Aniansson, E.A.G.; Wall, S.N.; Almgren, M.; Hoffman, H.; Kiclmann, I.;
Ulbricht, W.; Zana, R.; Lang, 1.; Tondre, C., J. Phys. Chem, 1976, 80, 905

Roscn, M.J, Surfactants and Interfacial Phenomena, Wilcy-Intersciencc, New
York, 1978; Chapter 4

Kalyanasundaram, K.; Grieser, E; Thomas, J K C hem. Phys. Lett., 1977, 5 I ,
501

Hurtubise, R.J., Anal. Chem, 1983, 55, 699A

47.

48.
49.
50.
51.
52.
53.
54.
55.
56.
57.
58.
59.

61.
62.
63.

44
Turro, N.J.; Kou—Chang, L.; Ming-Pea, C.; Lee, R, Photochem. Photobiol.,
1978, 27, 523
Humphry-Baker, R.; Moroi, Y.; Griitzel, M., Chem. Phys. Lett., 1978, 58, 207
Skrilec, M.; Cline Love, L.J., Anal. Chem, 1980, 52, 1559
Cline Love, L.J.; Habarta, J.G.; Skrilec, M., Anal. Chem, 1981, 53, 437
Cline Love, L.J.; Skrilec, M., Anal. Chem, 1981, 53, 1872
Woods, R.; Cline Love, L.J., Spectrochim. Acta, Part A, 1984, 40A, 643
Femia, R.A.; Cline Love, L.J., Anal. Chem, 1984, 56, 327
Lund, W.; Opheim, L.N., Anal. Chim. Acta, 1975, 79, 35
Buell, S.L.; Demas, J.N., Rev. Sci. Instrum, 1982, 53, 1298
Battino, R.; Banzhof, M.; Bogan, M.; Wilhelm, E., Anal. Chem, 1971, 43, 806
Takakubo, M.; Faure, J., Photochem. Photobiol., 1983, 38, 137
Reim,R.E., Anal. Chem, 1983, 55 , 1188
Rollie, M.E.; HO, C.-N.; Warner, I.M., Anal. Chem, 1983, 55 , 2445
Rollie, M.E.; Patonay, G.; Warner, I.M., Anal. Chem, 1987, 59, 180
MacCrehan, W.A.; May, W.E., Anal. Chem, 1984, 56, 625
Diaz Garcia, M.E.; Sanz-Medel, A., Anal. Chem, 1986, 58, 1436

Kolthoff, I.M.; LinganeJ.J., Polarography; Interscicncc, New York, 1952;
Vol. 1

CHAPTER IV
FLOW INJECTION ANALYSIS

Continuous flow analysis is a technique in which the concentration of an analyte
is measured in a continuously flowing stream of liquid or gas. Mixing and incubation of
the solution takes place as it flows through a manifold consisting of interconnected
tubing. Continuous flow analysis can be divided into two areas: segmented and non-
scgmcnted continuous flow analysis. Leonard Skeggs is accredited with the
development of the segmented technique [1]. In this technique, commonly referred to
as continuous flow analysis (CFA), the sample stream is segmented with air bubbles to
prevent excessive dispersion of the sample by the natural dispersive nature of laminar
flow. The air bubbles divide the flowing stream into a number of segments and
therefore prevent excessive dilution of the sample. They also minirrrizc carry-over and
prevent excessive dispersion of the sample. The samme is introduced into the manifold
by timed aspiration. The reaction time is determined by the length, the inner diameter,
and the flow rate of the solutions. Within each segment nearly complete mixing takes
place so that the signal obtained at the output approximates a rectangular shape as
would be expected in the ideal case of a discrete sample plug. This system was later
developed by Technicon Corporation and marketed commercially as the AutoAnalyzer.

Flow injection analysis (FIA) was developed several years later by two separate
groups. Stewart, Beecher, and Hare [2] developed FIA as an outgrowth of liquid
chromatography, while Jaromir Ruzicka and E10 Hansen [3] approached FIA as an
alternative to CFA. Ruzicka and Hansen coined the generic name for this non-

segmented continuous flow technique and are commonly given credit for inventing FIA.

45

46

Principles of Flow Injection Analysis

FLA involves the injection of a small, but precisely known, volume of sample
into a continuously) flowing stream of reagent. The sample zone then mixes with the
carrier stream on the way to the flow-through detector. The simplest flow injection
analyzer, shown in Figure 4.1, consists of a pump, that propels the carrier stream
through a polymeric tube, a method of sample injection, such as a multiport valve, and a
flow-through detector. The sample is injected as a discrete zone and then disperses and
reacts with the components of the carrier stream to form a species which is sensed by
the flow-through detector and recorded. Conditions are such that laminar flow is
predominant, which results in a parabolic velocity profile of the sample zone. This
gives rise to dispersion of the sample along the axis of the polymeric tube. This
dispersion can be controlled by judicious choice of tubing length and inner diameter,
and the type of flow cell used. The carrier stream flow-rate and sample volume also
affect the dispersion of the sample zone. A sample with an initial concentration of Co is
injected into the carrier stream. As the sample zone travels through the manifold, axial
and radial mixing occur. These combined effects result in a predominantly Gaussian-
shaped peak profile. The effect of laminar flow and molecular diffusion whish results in
the Gaussian-shaped peak is shown in Figure 4.2. Dispersion and molecular diffusion
dilute the sample so that the maximum concentration sensed at the detector is Cm“,
which is some fraction of C0. The ratio of Co to Cmax provides an empirical method by

which dispersion, D, can be measured [4]:

C° k’I-Io
D =cm = E‘H' (4'1)

For most applications, the analytical readout is based on the measurement of the peak
height, H. The conversion factors between instrumental readout and the concentration
are k’ and k'. The heights of the sample peaks before and after dispersion has taken .

Hmax

place are H0 and , respectively. Under conditions where the Lambert-Beer law is

obeyed, the two proportionality constants are equivalan Ruzicka and Hansen [4] have

47

 

 

 

 

ml / m in
‘— S
R W
Pump
Figure 4.1 Simplified representation of a flow injection system. The reagent R is

pumped through the manifold where the sample S is injected by a
sample injection valve. The product is monitored by the flow-through
detector FC. The product is then vented to waste W.

48

k

L F low Direction >

47717 1

xi 11 l

 

 

 

 

 

 

 

 

 

 

 

 

Figure 4.2 Laminar flow with molecular diffusion: (a) the sample plug, (b) under
conditions of laminar flow, (c) laminar flow with molecular diffusion,
((1) Gaussian peak profile resulting from molecular diffusion.

49

classified dispersion in FIA as limited (D = 1-3), medium (D = 3-10), and large
(D >10). The type of dispersion used depends on the application for which it is
intended. Limited dispersion will maintain sample integrity for applications where mere
transport of the sample is desired. When a chemical reaction between the sample and
the carrier must take place medium or large dispersion is desirable.

The type of manifold used is one method of controlling the dispersion of the
sample zone. Ruzicka and Hansen have reported the results of a comparison of various
types of FIA manifolds [5]. Relatively large amounts of dispersion are obtained with
straight open tubular reactors (OTRs). Coiled tubes decrease the amount of dispersion
due to secondary flow, flow that is perpendicular to the axis of the reactor. The extent
of secondary flow increases as the diameter of the coil decreases and the mean velocity
of the carrier increases. Dispersion can be minimized by using packed reactors rather
than OTRs. Although the pressure drop across a reactor packed with small diameter
particles may preclude the use of an inexpensive low-pressure pump, such as a
peristaltic pump, reactors packed with solid glass spheres with diameters slightly less
than that of the tube (60% to 80%) yield a pressure drop slightly greater than a coiled
OTR [6]. These single bead string reactors (SBSRs) decrease the dispersion by
breaking up the parabolic velocity profile.

Sample volume and manifold length also influence the dispersion obtained in
systems where the flow rate and tubing diameter are held constant. Increasing the
sample volume and decreasing the manifold length both decrease the amount of
dispersion obtained. For further information on the theory and applications of FIA the
reader is referred to the monograph by Ruzicka and Hansen [4] and a recent monograph
by Valcarcel and Luque de Castro [7].

50

Related Applicatiom of Flow In iectign Analysis

Although applications of FIA are quite numerous, several specific applications
have been used throughout the course of this research: luminescence and stopped-flow.

These techniques will now be discussed briefly.

Stopped-Flow in FIA

The stopped-flow mode can be used for two different purposes: to increase the
sensitivity of a measurement by increasing the residence time, or to measure the rate of
a reaction. An increase in the residence time will allow a reaction to increase the yield
of the measured species. The appropriate choice of the stOp period will gain reaction
time without increasing the length of the manifold which would increase the dispersion
of the sample zone due to laminar flow Therefore, dispersion will become independent
of time (with the exception of the very small contribution from molecular diffusion).

If the flow is stopped while the sample zone is within the flow cell, it is possible
to record the progress of the reaction. In this technique, the flow of the carrier must
come to a complete standstill such that the same section of the sample zone is held
reproducibly within the flow cell for measurement. This can be best achieved by an
electronic timer activated by the injection valve and connected to the pump. Injection
of the sample triggers the timer to count a delay time. Once the timer has delayed for
the predetermined time, the pump is turned off to stop the flow. After the reaction has
been monitored for the desired time, the pump is turned on to rinse the flow-cell and to
prepare for the next sample. The delay time and reaction time are chosen to suit the

particular chemistry.

Luminescence detection in FIA
Luminescence detectors monitor the radiation emitted by an excited state analyte
as it rcmrns to the ground state. In chemiluminescence the molecule is excited through

a chemical reaction. In this technique, no radiation source is required and therefore

51

filters or monochromators are not required for wavelength isolation. The detection
system consists of a radiation transducer and a readout.

In fluorescence and phosphorescence, however, the analyte is excited by
radiation of a specific wavelength, and the radiation resulting from the return to the
ground state is monitored at a specific emission wavelength. Therefore, a suitable
source of radiation is required with a wavelength isolation device to excite the analyte
molecule. The wavelength isolation device can be a filter or monochromator.
Typically, the emitted radiation is monitored at 90’ relative to the excitation beam.
With this arrangement, the flow cell must have windows oriented on adjacent sides of
the flow cell. The windows of the cell are commonly quartz or sapphire so that they
will transmit the ultra-violet radiation used for sample excitation. As with the excitation
radiation, the emitted luminescence radiation is isolated with either a filter or a
monochromator. The emitted luminescence is then detected by the radiation transducer
and monitored by the readout.

The flow cells used are typically low-pressure flow-through cells with low
volumes. Large volume flow-through cells will increase the dispersion of the sample
zone. Cells with a wide variety of volumes and pathlengths are available which can be
selected to suit the particular needs of the analysis.

To date, there have been many applications of fluorescence detection in
FIA [8-44]. A literature search of applications of RTP in flowing streams (flow
injection analysis and chromatographic systems) has resulted in few manuscripts being
found [45-49]. These applications are all based on the sensitized RTP of biacetyl. Two
methods were found that are based on the quenched RTP of biacetyl by a suitable
analyte [49,50]. None of these RTP techniques employs a direct measurement of the

phosphorescence of an analyte.

5‘1”!"

1*"

10.
ll.
12.
13.
14.

15.
16.

17.
18.
19.
20.
21.

22.

52

CHAPTER IV REFERENCES

Skeggs, L., Am. J. Clin. Pathol., 1957, 13, 451
Stewart, K.K.; Beecher, G.R.; Hare, P.E., Fed. Proc., 1974, 33, 1439
Ruzicka, J.; Hansen, E., Anal. Chim. Acta, 1975, 78, 145

Ruzicka, J.; Hansen, E., Flow Injection Analysis, John Wiley & Sons: New
York, 1981; Chapter 2

Ruzicka, J.; Hansen, E., Anal. Chim. Acta, 1984, I61, 1
Reign, I.M.; van der Linden, W.E.; Poppe, H., Anal. Chim. Acta, 1981, 123, 229

Valcarccl, M.; Luque de Castro, M.D., Flow Injection Analysis: Principles and
Applications, John Wiley & Sons: New York, 1987

Kina, K.; Shiraishi, K.; Ishibashi N., Talanta, 1978, 25, 295
Braithwaite, M.; Miller, J.N., Anal. Chim. Acta, 1979, 106, 395

Lim, C.S.; Miller, J.N.; Bridges, J.W., Anal. Chim. Acta, 1980, 114, 183
Ishibashi, N.; Kina, K.; Goto, Y., Anal. Chim. Acta, 1980, 114, 325
Imasaka, T.; Harada, T.; Ishibashi, N., Anal. Chim. Acta, 1981, 129, 195
Kelly, T.A.; Christian, G.D., Anal. Chem, 1981, 53, 2110

Playle, R.; Gleed, 1.; Jonasson, R.; Kramer, J.R., Anal. Chim. Acta, 1982, I34,
369

Kelly, T.A., Diss. Abstr.1nt., 1981, 42, 2356

Mori, K.; Imasaka, T.; Ishibashi, N.; Jin, C.; Kokubu, N., Bunseki Kagaku, 1982,
31, 103

Kelly, T.A.; Christian, G.D., Anal. Chem, 1982, 54, 1444
Rydevik, U.; Nord, L.; Ingrrran, F., Int. J. Sports Med, 1982, 3, 47
Harris, I.M., Anal. Chem, 1982, 54, 2337

Stewart, K.K.; Rosenfcld, A.G., Anal. Chem, 1982, 54, 2368

Petty, R.L.; Michel, W.C.; Snow, J.P.; Johnson, K.S., Anal. Chim. Acta, 1982,
I42, 299

Kelly, T.A.; Christian, G.D., Talanta, 1982, 29, 1109

23.

24.

25.
26.
27.

28.
29.
30.
31.

32.
33.
34.
35.
36.

37.

38.
39.
40.
41.

42.
43.

45.

53
Abdullahi, G.L.; Miller, J.N.; Sturlcy, H.N.; Bridges, J.W., Anal. Chim. Acta,
1983, I45, 109

Karlsson, J.; Jacobs, 1.; Sdein, B.; Tesch, P.; Kaiser, P.; Sahl, O.; Karlberg, B.,
Int. J. Sports Med, 1983, 4, 52

Elliott, C.B., Diss. Abstr. Int., 1983, 43, 3966
Motomizu, S.; Oshima, M.; Toei, K., Bunseki Kagaku, 1983, 32, 458

Kusube, K.; Abe, K.; Hiroshima, O.; Ishiguro, Y., Chem. Pharm Bull., 1983, 31 ,
3589

Wcicker, H., LaborPraxis, 1984, 8, 300
Wcikcr, H.; Hacgcle, H.; Komcs, B.; Werner, A., Int. J. Sports Med, 1984, 5, 47
Boza, F.L.; Luque dc Castro, M.D.; Valcarccl, M., Analyst, 1984, 109, 333

Motomizu, S.; Mikasa, H.; Oshima, M.; Toei, K., Bunseki Kagaku, 1984, 33,
1 16

Aoki, T.; Uemura, S.; Munemori, M., Bunseki Kagaku, 1984, 33 , 505
Neubert, P.; Reiff, K., Presenius Z. Anal. Chem, 1981, 305, 277

Valcarccl, M.; Luque de Castro, M.D.; Lazaro, F., Anal. Proc., 1983, 20, 313
Miller, J.N., Anal. Proc., 1983, 20, 487

Lazaro, F.; Luque dc Castro, M.D.; Valcarccl, M., Anal. C him. Acta, 1984, I65,
177

Lazaro, F.; Luque dc Castro, M.D.; Valcarccl, M., Frescnius Z. Anal. Chem,
1985, 320, 128

Aihara, M.; Aral, M.; Tomitsugu, T., Anal. Lett., 1986, 19, 1907
Masoom, M.; Worsfold, P.J., Anal. Chim. Acta, 1986, I79, 217
Hwang, 1.; Dasgupta, P.K., Anal. Chem, 1986, 58, 1521

Kelly, T., In Enzyme-Mediated Immunoassay; Ngo, T.T., Lenhoff, H.M., Eds;
Plenum: New York, 1985; pp 191-201

Maeda, M.; Tsuji, A., Analyst, 1985, 110, 665
Linares, P.; Luque de Castro, M.D.; Valcarccl, M., Anal. Chem, 1985, 57, 2101

Hult, K.; Fuchs, R. Peraica, M.; Plcstina, R. Ceovic, S., J. Appl. Toxicol, 1984,
4, 326

Donkerbroek, 1.1., Van Eikcma Hommes, N.J.R.; Gooijer, C.; Velthorst, N.H.;
Frei, R.W., Chromatographia, 1983, 15 , 218

46.

47.

48.

49.

50.

54
Donkerbroek, J .J., Van Eikcma Hommes, N.J.R.; Gooijer, C.; Velthorst, N .H.;
Frei, R.W., J. Chromatogr., 1983, 255, 581

Donkerbroek. J.J.; Veltkamp, A.C.; Praat, A.J.J.; Gooijer, C.; Frei, R.W.;
Velthorst, N.H., Appl. Spec, 1983, 37, 188

Baumann, R.A.; Gooijer, C.; Velthorst, N.H.; Frei, R.W., Anal. Chem, 1985, 5 7,
1815

Gooijer, C.; Velthorst, N.H.; Frei, R.W., TrAC - Trends Anal. Chem, 1984, 3,
259

Donkerbroek, J.J.; Veltkamp, A.C.; Gooijer, C.; Velthorst, N .H.; Frei, R.W.,
Anal. Chem, 1983, 55, 1886

CHAPTER V
LUMINESCENCE INSTRUMENTATION

This chapter describes the instrumentation that was used throughout the course
of this research. The initial studies were done with a microproceSsor-controllcd
luminescence spectrometer designed and constructed in our laboratory. Later, a
commercial instrument (Perkin-Elmer Model LS-5 Spectrofluorometer with a Model
3600 Data Station) became available. This instrument was used for most of the later
studies. Although the commercial instrument offers optimized Optics and source
correction, the home-built instrument can be more readily programmed by the user.
Data acquired on the home-built instrument can also be transmitted serially to a Digital
Equipment Corporation L81 11/23 minicomputer for graphical display, statistical
evaluation, and mass storage. This option gives the locally designed instrument more

data analysis veratility than is available with the Perkin-Elmer Data Station.

m Buil Mi r m ter- ont ll uminesc nc rom ter
The home-built luminescence spectrometer is controlled by a modular twin bus
microcomputer system designed by Newcome and Bake [1]. This microcomputer is
interfaced to a custom-designed spectrofluoromter. A mechanical chopper is included
in the design of this spectrofluorometer so that time-resolved phosphorescence
measurements can be performed. The microcomputer, luminescence spectrometer, and

interface are discussed individually below.

55

56

Twin Bus Microcomputer System

The version of the microcomputer used was constructed in this laboratory and is
based on the Intel 8088 microprocessor (Intel Corp., Santa Clara, CA). The modular
design of the microcomputer increases its flexibility and adaptability. Thus, a custom-
tailored computer system can be created from a standard set of modules. Modules used
in the rrricrocomputcr with their principal integrated circuits are listed in Table 5.1.
Some of the unique features of the microcomputer are: the programmable interrupt
controller, which expands the processors interrupt capacity and capabilities; the dual
USART, which consists of two R8232 ports that are assigned to the user’s terminal and
the L81 11/23; and the chip select, which allows addressing of other modules while
eliminating unnecessary redundancy of decoding circuits.

Two other modules are Of particular interest and deserve special mention: a
counter/timer 11, based on the AM9513 (Advanced Micro Devices, Sunnyvale, CA) and
a sample/hold multiplexer (S/H-MUX). Both of these modules, as well as the 12-bit
analog-tO—digital converter (ADC) were designed by Ratzlaff [2]. The 9513
counter/timer contains five, 16-bit, individually programmable counters. Several
applications of interest include time-Of-day clocking, real-time program independent
clocking, count up/down, polled or interrupt timing for data collection sequencing or
peripheral servicing, programmable waveform generation, and event count
accumulation.

The sample/hold multiplexer is a module containing four identical differential
input inverting sample-and-hold (S/H) circuits (SHM-IC-l, Datcl Intersil Inc.,
Mansfield, MA) which act as the first stages of the analog input circuit. With this
design, only one analog-tO—digital converter (ADC) is required. The S/H-MUX module
is included to provide interfacing to more than one experimental apparatus at a given
time and to allow multi-channel data acquisition. Each of the four S/Hs have

independent digital control, which permits flexible Operation initiated by software or

57

Table 5.1
Modules with their Principal Integrated Circuits

 

Module Principal Integrated Circuit
Central Processing Unit 8088

32k RAM/ROM 2716, 2016
Interrupt Controller 8259A-5
Dual USART 8251A
Parallel Input/Output 8255A-5
8k RAM 21 14

Chip Select a
Counter/Timer II AM9513
Dual 8-bit DAC AD558
12-bit DAC AD565
Multiplexer a

12-bit ADC AD574
SASI Interface a

 

a 74LS series T'I‘L integrated circuits

58

asynchronous events. The four S/Hs are multiplexed by a 8-channel multiplexer
(MX-808, Datcl Systems Inc., Mansfield, MA) with a buffered output to maintain
transfer accuracy. This output is used as the ADC input. For a more in-depth
discussion of the counter/timer II, S/H-MUX, and ADC modules, the reader is referred
to Ratzlaff’s thesis [2].

The microcomputer has a single 8 in. floppy disk drive to facilitate program and
data storage and to ease programming. Serial transmission of data to and from the
LSI-11/23 is made possible by the program FORTHPIP [3]. This program serves as a
means of communication between the microcomputcr’s operating environment and files
on the minicomputer. Serial transmission is a slow process and is failure prone during
heavy usage on the minicomputer. Downloading to the microcomputer from the floppy
diskette saved considerable time over the serial link to the LSI-11/23. The same is true
for the uploading process.

The microcomputer runs under the polyFORTH Operating environment
(FORTH, Inc., Hermosa Beach, CA). FORTH is a stack oriented, threaded interpreter.
It was designed to provide power, ease, and flexibility in interfacing with instruments.
A unique featrn'e of FORTH is that is does not compile, but rather links programs or
subroutines into new programs or subroutines. A FORTH programmer begins with a
subset of useful subroutines called "wo Programming entails creating new words
by linking previously defined words. Judicious choice of words should, theoretically,
make a program in FORTH read like a sentence and therefore be easy to understand.
However, it is the author’s experience that this threaded nature makes complex
programs quite difficult to decipher. FORTH excels in speed for high level
mathematical calculations because of its use of fixed point rather than floating point
math [4]. Some of the criticisms Of FORTH are its lack of file structure, and its
requirement that words be named uniquely. The version of FORTH used in this

rrricrocomputcr uses the first three characters and the length of the word as a test for

59

word uniqueness. Therefore, one must be creative in assigning names so that each word

remains exclusive.

H ome-Built Luminescence Spectrometer

A typical fluorometer consists of a radiation source to excite the sample, an
excitation and an emission wavelength selector (filters or monochromators), a radiation
transducer, a signal processor, and a readout device. A block diagram of the
microcomputer controlled spectrofluorometer is shown in Figure 5.1. The source of
radiation for this instrument is a 150 W high pressure Xe arc lamp (CA-150 Light
Source, Illumination Industries, Sunnyvale, CA). Radiation from this source is focused
onto the entrance slit of the excitation monochromator (ISA H-10, Instruments SA Inc.,
Mctuchen, NJ) by a quartz lens. The excitation and emission monochromators are
mounted at 90' around a sample chamber designed and constructed in our laboratory.
The emission beam is selected by the emission monochromator (ISA H-10). These
monochromators have a linear dispersion of 8 nm/mm and a focal length of 100 mm
with a F/3.5 aperture. The radiation transducer is a photomultiplier tube (PMT, model
R-928, Hamamatsu, Bridgewater, NJ) The PMT photocurrcnt is amplified and
converted to voltage by a commercial amplifier (Model 427 Current Amplifier, Keithley
Instruments Inc., Cleveland, OH). The current amplifier output is digitized by the ADC
module in the microprocessor. At present, the monochromators must be scanned
manually.

Temporal discrimination is provided by a rotating disk chopper which is placed
between the source and the entrance slit of the excitation monochromator. Its

construction and use are described in more detail below.

Microcomputer/Spectrometer Interface
The microcomputer controls several components of the spectrofluorometcr and

peripherals associated with experiments performed on this instrument. These include

 

PD

PRINTERh @ 6T9 Mex SC }_
i

, Mem
L51 11/23 k—‘TMICROCOMPUTER
I "r5

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

TE R M I N A L I /V l
PLO TTER 1___4
Figure 5.1 Block diagram of home-built luminescence spectrometer showing

orientation of the components. Key: S, radiation source; L, lens; C,
Optional chopper; Mex, excitation monochromator; SC, sample
chamber; PD, photodiode with amplifier circuitry; Mem, emission
monochromator; I/V, current amplifier.

61

the rotating disk chopper and a peristaltic pump and sample injection valve for flow
injection work. The chopper consists of a 10 cm diameter disk with 8 slots cut at equal
intervals. This disk is mounted to the shaft of a 12 Vdc hobby motor. The rotation
speed of the motor is controlled through one of the 9.95 Vdc full scale 8-bit digital-to-
analog converters (DACs). Since the motor requires approximately 100 mA, which is
more current than the DAC can supply, a transistor driver circuit was used to provide
the necessary current. The power transistor is powered by an external power supply
(Model EU-801-11, Heath Corp., Benton Harbor, MI). This circuit is shown in
Figure 5.2. The voltage follower with gain was included to buffer the load on the DAC
and to increase the supply voltage to the motor. This then allows slightly higher
rotation speeds (which affects the chopping frequency) to be Obtained by increasing the
maximum transistor input to 12 Vdc.

To monitor the chopping frequency, an opto-intcrrupter is placed so that the
blades of the rotating disk pass between the source and receiver. As the disk rotates, the
Opto-interruptor encodes information about the chopper status such as, is the motor
turning?, what is the position of the blade? (is the source blocked?), what is the
chopping frequency? The output of the phototransistor in the optO-interrupter is passed
through a Schmitt trigger to provide noise discrimination. This ensures a reliable T'TL
signal from the Opto-intcrrupter. With this arrangement, the optO-intcrrupter must be
"phased" so that the TH. signal reflects correctly the status of the excitation radiation
entering the sample chamber. Phasing is accomplished by placing a phototransistor
inside the sample chamber such that it is exposed to the excitation radiation. The output
of this phototransistor is also passed through a Schmitt trigger. Signals from both
phototransistors are displayed concurrently on a dual-trace oscilloscope. The position
of the Opto-intcrrupter is then adjusted physically until both traces on the oscilloscope
are in phase. If the Opto-intcrrupter is not phased correctly, the sample will or will not

be exposed to the excitation radiation as indicated by the Opto—intcrrupter signals.

62

 

 

 

 

 

10 k!)
/l\
g +15V do
From DAC
2N6045
Motor
Figure 5.2 Schematic Of transistor driver circuit used to amplify current

supplying power to chOpper motor.

63

Three examples of optodntcrruptcr/sourcc phasing are shown in Figure 5.3.
Early phasing occurs when the sample chamber HI-tO-LO transition occurs before the
Opto-interrupter status indicates this transition. Late phasing occurs when the sample
chamber HI-tO-LO transition occurs after the Opto-intcrrupter HI-tO-LO transition. The
OptO-interrupter is phased correctly when the two transitions occur concurrently.

Although this OptO-interruptcr arrangement worked satisfactorily, phasing had to
be performed prior to each experiment as motor vibrations had the tendency to disrupt
the phasing of the opto-intcrrupter. The opto-intcrrupter was discarded and a second
approach was taken to eliminate this phasing problem. An ultra-violet enhanced
photodiode (model UV-040BQ, EG & G Photon Devices, Salem, MA) was situated in
the sample chamber opposite the entrance slit. With this arrangement, the photodiode
yields a direct indication of the source chopping with respect to the sample. The circuit
used in this application is shown in Figure 5.4. The first stage amplifier is a current-to
voltage converter with gain. The second stage amplifier is an inverting voltage follower
with gain. This two stage amplifier design is necessary because of the low responsivity
of the photodiode near the excitation wavelengths used in luminescence. The
responsivity of the photodiode at 254 nm is 0.14 AM, compared to 0.62 AM at 900 nm
[4]. The output of this amplifier is then passed through a Schmitt trigger to obtain a
T'TL signal.

The microcomputer is able to monitor the status of the chopper through the
8255A parallel peripheral interface. The 8255A contains three 8-bit ports, A, B, and C.
In the mode used for basic input/output, each port can be configured individually. The
output Of the Schmitt trigger on the photodiode circuit serves as the input to the input
port CO on the 8255A. It is through bit 0 of the port C status byte that the
microcomputer can determine if the source is blocked or not. If the bit is set to 1, the

excitation radiation is viewed by the sample; conversely, if the bit is set to 0 the source

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Hi .__. _ 7
A LOI _J Jr— ] Chopper
I"; J__ J"... .1" Ce”
FmflJ—_f—_ J‘ _1 CbOPper
B _, _. q ,
1m ._. ._._l JP] Cell
C “Limos—5:3— stopper
_.1 L Lh_.l__l—L_1_ Cell
Figure 5.3 Timing diagram depicting the three types of phasing:

A, chopper signal lags source signal; B, chopper signal leads source
signal; and C, signals are in phase.

65

    

 

 

 

 

 

R, R2
._M_.
TLOB] 7407
D—fil-
PD a 4'“ >
V
T v T
—l 5 _1 5
Figure 5.4 Photodiode amplifier circuit. R1 = 1 M9, R2 = 2.2 MS), and

Rim 3 680 Q

66

is blocked. The FORTH word ?STAT checks the status of port C0 through the address
of port C, ".CPIO". (Sec code in block number 160 in Appendix A.)

The chopping frequency is adjusted by the voltage applied to the motor rotating
the masking disk. The frequency of oscillation is determined by programming the
counter/timer to count the number of HI to L0 transitions per unit time. A program,
?FREQ, was written to count the number of transitions in l s and then report the
frequency in Hertz. (See Appendix A, block number 161, 162.) This program uses two
of the counters on the 9513. Counter 1 is programmed to count the number of L0 to HI
transitions and accumulate the sum while the counter is active. This counter is active
only while counter 2 is active. Counter 2 is programmed to wait for the first L0 to H]
transition and then delay for 1 s. The delay period is adjustable to 10 s with the word
lOSEC, and back to l s with lSEC. With these programs it is possible to determine the
chopping frequency accurately.

For experiments in flow injection analysis, two other peripheral devices were
interfaced to the microcomputer: a peristaltic pump (model IP-12, Brinkmann
Instruments, Westbury, NY), and a pneumatically actuated Teflon rotary sample
injection valve (Rheodyne type 50, Anspec Co. Inc., Ann Arbor, MI). The pump is
supplied with connections to be used for remote functions through a 5-pin male plug
(Switchcraft modc105BL5M or equivalent). Speed control of the pump is accomplished
by supplying 0-5.5 Vdc to pole 3 and grounding pole 2. The variable voltage supply is
obtained by connecting a voltage divider to the second 8-bit DAC which limits the
voltage maximum to 5.5 Vdc. The pump speed is changed on the pump with a digital
setting between 0 and 99. The desired digital setting is passed to the program PUMP
(see Appendix A, block number 24) which calculates the data value for the DAC to
obtain the corresponding pump speed from the desired pump digital setting. The

relationship between the DAC output and the actual pump speed was determined

67

experimentally by measuring the voltage supplied to the pump motor for various digital
settings.

The direction of the pump is controlled through bit 0 of port B on the 8255A.
This pin is connected to pole 4 on the pump and pole 2 is connected to ground. Setting
bit B0 to 0 should cause the pump to change directions. However, the pump loaded the
8255A output and was not able to sink the current. Therefore, a 7407 HEX buffer was
added between the 8255A and the pump. The FORTH words FWRD and RVRS set the
direction of the pump to forward and reverse, respectively.

The sample injection valve is controlled through bit 0 of port A. The output of
the port is used to control a photo-isolated, zero-voltage switching relay (Magnavox
model W3PCX-1, Newark Electronics, Grand Rapids, MI). This relay controls the
120 Vac pneumatic actuator. A zero-voltage switching relay is used to minimize rf

switching noise which had the tendency to "crash" the microcomputer.

Experimental Determination of the Dominant Noise Source

This section describes experiments which were performed to deterrrrinc the
dominant noise source(s) in the home-built instrument. Knowledge of the major noise
sources could indicate modifications in the instrumental design to improve the signal-to
noise ratio. The signal-to-noise ratio (SIN) is defined as the ratio of the mean value of
the analytical signal divided by the rrns noise, S/N = E/O’E [5]. The rms noise is the
standard deviation of the analytical signal derived from 11 measurements. There are
several different types of noise. Random noise is unpredictable in its sign and
magnitude of deviation from the mean signal. The sign and magnitude of the deviation
for nonrandom noise can be correlated in time to some event.

Noise can also be classified as fundamental or nonfundamental. Fundamental
noise arises from the particle nature of light and electricity and cannot be completely
eliminated. Nonfundamental noise is due to imperfect components and instrumentation.

This noise can be eliminated through careful instrument design.

68

The noise observed in a signal is due to the summation of the fluctuations
emanating from random and nonrandom noise sources. As this section is not intended
to be a discussion of signal-to—noise ratio theory, the reader is referred to a recent text by
Crouch and Ingle for a more thorough discussion [6].

There are three limiting cases of noise: dark current noise, which is independent
of the source radiation flux, (I); shot noise, with a standard deviation (OE) proportional
to WI); and flicker noise, with GE proportional to (I).

Photons arrive randomly at the photocathode of the radiation transducer because
of the random emission of photons from the excitation source or analyte. The signal
(E) from photoanodic current (i a) of the PMT is

E = Gia = mGROt)<I> (5-1)
where G is the current-to—voltage converter gain, or is the PMT gain, and R0») is the
PMT cathodic responsivity. For the Shot noise-limited case

oE . [2eAf(1+a)mGE]l/ 2 (5-2)
where Af = 1/(4RC) is the noise bandpass of the electronics, at is the secondary
emission or multiplication noise factor, and e is the charge on an electron. The signal-
to-noise ratio determined from equations 5-1 and 5-2 reduces to

S/N = E/OE = ‘JCP/k (5-3)
where k = [2eAf(1«1-a)]1/2 and is a bandwidth constant proportional to Af [5].
Therefore, equation 5-3 indicates that the S/N increases as the square root of <1). Shot
Noise is reflected in the readout because all conversions following the photocathode
multiply the signal and noise by linear factors. Equation 5-3 indicates that a plot of
log(S/N) vs log(<b) should yield a slope of 1/2. Since flicker noise is proportional to (D,
a plot of log(S/N) vs log(¢) should yield a slope of unity. Because the dark current
noise is independent of (D, a log(S/N) vs log(F) plot will have a slope of zero. This was
tested experimentally by varying the monochromator slit width and measuring the S/N.

This experiment was performed under conditions normally used for fluorescence and

69

phosphorescence work in this research since the signal level will change the dominant
source of noise. For a continuum source, the radiant power passed by the
monochromator is proportional to the square of the slit width provided the entrance and
exit slits are of equal size [6]. In fluorescence spectroscopy, If is a linear function
of <1). Therefore, any increase in (b will increase If. The fluorescence intensity of

0.111M quinine sulfate in 0.2 M sulfuric acid was monitored at 450 nm with an
excitation wavelength of 365 nm for three different Slit widths. Data from this
experiment are shown in Table 5.2. Since the signal E is proportional to d), a plot of
log(S/N) vs log(E) is equivalent to a plot of log(S/N) vs log(<l>). Results from a linear
regression yield a slope of 0.468 with a correlation coefficient of 0.993. This is only a
6.4% deviation from the value of 0.500 expected for Shot noise. Therefore,
photocurrcnt shot noise can be assumed to be the limiting noise in this instrument under

these conditions.

 

 

 

 

Table 5.2
Effect of Radiant Power on the S/N
Slit Width Signal 0,, S/N
(mm) (ADC counts)a
2. 3600— 52 69 “"-
2. 3700 52 7 l
1. 230 9 26
1. 230 10 23
0.5 17 3 6
0.5 17 3 6

 

aaverage of 100 analog—to digital conversions

70

Perkin-Elmer Model LS-S Luminescence Spectrometer

The model LS-5 luminescence spectrometer is a microprocessor controlled
instrument with digital display. The source of radiation is 9.9 W xenon discharge lamp
which is pulsed at 60 Hz. The source pulses have a full width at half maximum of less
than 10 115. Wavelength isolation is facilitated by two F/3 monochromators. These
monochromators are Stepper motor driven and can be scanned independently. The
excitation monochromator covers the wavelength range of 230-720 run while the
emission monochromator covers 250-650 nm. The range of the emission
monochromator can be extended to 800 nm with a red sensitive PMT. Both
monochromators are capable of isolating the grating zero order. The spectral
bandpasses of the monochromators are selectable with nominal values of 3, 5, 10, and
15 nm for excitation and 3, 5, 10, and 20 nm for emission.

The LS-5 instrument includes an excitation/correction system that uses a
quantum counter (rhodamine B) in front of a reference PMT. The reference PMT is
matched to the sample PMT. Rhodamine B has a fluorescence quantum efficiency that
is nearly independent of the excitation wavelength. The ratioing system of the LS-5
corrects source spectral distribution and the wavelength dependence of the excitation
monochromator and optics.

The LS-5 is capable of gating the detector which allows phosphorescence
measurements to be made. The delay and gate times of the PMT are selectable in
increments of 10113 not to exceed a total of l3ms.

A Perkin-Elmer model 3600 data Station is connected to the LS-5 through a
R8232C interface for instrument control and external data manipulation. The data
Station also includes software for data manipulation, graphical display, and mass storage

on floppy diskette.

999°.“

71

CHAPTER V REFERENCES

Newcome, B.H.; Enke, C.G., Rev. Sci. Instrum, 1984, 55 , 2017

Ratzlaff, E.H., Ph.D. Thesis, Michigan State University, East Lansing, MI, 1982
Philip Hoffman, private communication, January, 1982

Brodie, L., Starting FORTH; Prentice-Hall: Englewood Cliffs, NJ, 1981

Malmstadt, H.V.; Enke, C.G.; Crouch, S.R., Electronics and Instrumentation for
Scientists; Benjamin/Cummings: Menlo Park, CA, 1981

Crouch, S.R.; Ingle, J.D., Spectrochemical Analysis; Prentice-Hall, 1988;
Chapter 5.

CHAPTER VI
LUMINESCENCE SPECTRA

This chapter describes an experimental approach to Obtaining RTP. First, the
purification and preparation of the reagents required are discussed. Likcnfluoresccncc,
phosphorescence arises from intersystem crossing of the excited singlet state to the
triplet state. Hence, the absorption transition leading to phosphorescence is identical to
that leading to fluorescence. This chapter discusses the fluorescence excitation and
emission spectra of selected compounds. Solutions of these compounds were then

prepared for RTP and phosphorescence Spectra obtained.

Reagents

All luminescence studies described were performed with solutions. The analyte
microenvironment for this work was provided by either rrricellar assemblies or
cyclodextrin inclusion complexes. The surfactant used for micelles was sodium dodecyl
sulfate (SDS, also known as sodium lauryl sulfate, Aldrich Chemical Co. Inc.,
Milwaukee, WI). This surfactant was 98% pure and is used without purification. The
B—cyclodextrin (CD) was also used without additional purification (Aldrich). As
discussed in Chapter III, a heavy atom is normally required to obtain phosphorescence.
For the CD solutions the heavy atom was 1,2-dibromocthane (994% Gold Label, DBE,
Aldrich) and was used as supplied. All water used in these Studies was Obtained from a
Corning Mega-Pure system. House distilled water first passes through deionizing
columns to remove organics. The water is then distilled which results in a resistance of

at least 50 k9 cm.

72

73

For micellar solutions, thallium(l) ion is the preferred heavy atom to increase the
phosphorescence yield. A high purity thallium nitrate (99.999%, Gold Label, Aldrich)
was used. To avoid unnecessarily high ionic concentration in these solutions,
thallium(I) dodecyl sulfate (TTDS) was prepared by a previously reported method [1].
This method involves mixing approximately stoichiometric quantities of SDS and
thallium(l) nitrate solutions prepared in warm water. The precipitate was allowed to
form overnight. The TlDS was then doubly recrystallized from water and dried
overnight in a vacuum oven at 40’C. Stock solutions of SDS and CD were prepared
with a concentration of 0.50 M. Working solutions with concentrations of 0.10-0.15 M
were prepared from these stock solutions. Because of the limited solubility of TlDS, a
0.10 M stock solution of this reagent was prepared. However, even at this
concentration, some precipitation occured. Therefore, the solution was warmed slightly
and mixed thoroughly prior to preparation of working solutions.

Deoxygenation of all solutions for phosphorescence measurements was
performed by the addition of sodium sulfite. Sodium sulfite was purified by double
recrystallization from cold water. The precipitate was then dried in a vacuum desiccator
for several days. Stock solutions (0.10 M) were prepared in water just prior to
preparation of working solutions to minimize the loss of sulfite because of reaction with
ubiquitous oxygen. Once working solutions were prepared, the ground glass fittings on
the sample volumeuic flasks were sealed with Parafilm laboratory film (American Can
Co., Greewich, CT) to ensure an air-tight seal.

The analytes used in this research, naphthalene, pyrene, and biphenyl, were all
doubly recrystallized from absolute ethanol and dried in a vacuum desiccator for several
days. 2-Bromonaphthalene (97%, Aldrich) was used as received. Stock solutions of
these compounds for use in micelles were prepared by dissolving the analyte in absolute
ethanol. An aliquot of this cthanolic solution was pipetted into a clean, dry volumetric

flask. The ethanol was evaporated with dry nitrogen gas. The residue was then

74

dissolved in 0.50 M SDS by ultra-sonification of the solution. The solution was then
diluted to the mark with water. The stock analyte solution was prepared with a
concentration of 1.0 mM. The surfactant concentration was either 0.50 or 0.10 M.
Although early work in solution RTP [1,2] suggested that all glassware be baked
at 1000 K prior to being used for the first time for RTP work, this was not done in these
studies. All glassware was cleaned scrupulously and rinsed with 20% nitric acid. The
glassware was then rinsed with copious amounts of water. The same volumetric flasks

were used throughout.

Fluorescencg Saga

Fluorescence excitation and emission Spectra of naphthalene,
2-bromonaphthalene, pyrene, and biphenyl, are presented in this section. The excitation
wavelength for each analyte was determined by setting the emission monochromator to
the zero order and scanning the excitation monochromator. The excitation
monochromator was then set at the wavelength yielding the greatest relative
fluorescence intensity. The emission monochromator was scanned to determine the
emission wavelength. A final excitation spectrum was then Obtained with the emission
monochromator set at the wavelength yielding the greatest relative fluorescence
intensity.

The excitation and emission spectra for naphthalene in 0.15 M SDS are Shown in
Figure 6.1. Spectra A and B were obtained on the Perkin-Elmer LS-S
Spectrofluorometer with 5 nm bandpass on both the excitation and emission
monochromators. These Spectra Show an absorbance band centered at 276 nm and a
fluorescence band at 330 nm. Spectrum C was taken on the home-built system with
16 nm bandpasses on bOth monochromators. The luminescence intensities are given in

arbitrary units. The peak at ~505 nm seen in spectrum C is believed to be due to

75

 

 

 

A
- lax
j 1923mm
:- ,‘I
; i
1 E
r ;
B . '
; ,~‘
Vnimm \a—_ A."
1:39 3‘» ‘39 m sea :5;-
lam
4000 a
E
In
S
C ’22
o -
300 350 400 450 500 550 600
xem (nm)
Figure 6.1 Excitation and fluorescence emission spectra of naphthalene.

A) excitation scan taken on Perkin-Elmer LS-S with k m = 330 nm.
B) emission scan taken on LS-S with leg = 276 nm. 5) emission
scan taken on home-built instrument Wlth Aex = 280 nm. See text for

explanation of peaks above 450 nm.

76

impurities in the quartz cell used. The peak at ~560 nm in spectrum B and C is due to
scattered source radiation in the second order.

With the same instrument parameters, the excitation and emission spectra of

2-bromonaphthalene were obtained, as shown in Figure 6.2. The excitation band is
centered at 286 nm with a fluorescence band at 316 nm.
- Figure 6.3 shows the excitation spectrum for pyrene taken with an emission
wavelength of 395 nm on both instruments. These spectra differ because the spectrum
taken on the Perkin-Elmer LS-S is corrected, while that taken on the home-built
instrument system is uncorrected. The higher resolution of Figure 6.3 B is due to the
narrower bandpasses of the monochromators. The fluorescence spectrum of pyrene in
Figure 6.3 C has an emission band at 395 nm.

Spectra of biphenyl were acquired solely on the Perkin-Elmer LS-S, as shown in
Figure 6.4. The absorption band is centered at 260 nm with the fluorescence emission at

472 nm. The bandpasses of bOth monochromators were 5 nm.

RTP 535353

The goal of this research was to develop a system for automating RTP
measurements by introducing samples via flow injection analysis (FIA). Tis requires
working with solutions. Therefore, for direct measurement of analyte phosphorescence,
only two techniques for obtaining RTP are applicable, micelle stabilized RTP
(MS-RTP) and cyclodextrin induced RTP (CD-RTP). Phosphorescence Spectra

obtained using these techniques are presented below.

MS-RTP Spectra
MS-RTP cannot be obtained by simply preparing a sample in a soap solution.
There must be a mechanism for spin-orbital coupling, and the sample solution must be
deoxygenated. Figure 6.5 is the fluorescence spectrum of naphthalene in 0.10 M SDS
and in 0.10 M SDS with 25 mM T1N03. Notice that the addition of the thallium ion

Figure 6.2

77

 

 

. 3;), I .1
‘h"”-—~_
u so ““- -----
L-' -‘ a Li?! J t v
)‘e x
i
.‘ l
\._ . ’-- \‘ci-
M\\-‘- I ’-
r -R_ f--
‘_ .
i ."', I .‘1 8i. . in —.___’.
29? 350 40$ t‘r ' J SSi a '
)Lenr
3000 -
[\4
\
E
C m
z
“4 \
z;

 

 

0 \/" \ \
V v V V 1* V V I 1 'fi
300 350 4C0 4:: 5:3 -SC 60:

Excitation and fluorescence emission spectra of 2-bromonaphthalene.
A) excitation scan from LS-S with lem = 315 nm. B) emission scan
from LS-S. C) emission scan from home-built instrument.

78

INTENSITY

 

 

 

 

1'
//
o /\'\/
260 290 320 350
)‘ex (nm)
A r” i:
L' .I' i
l ".
B i”. ’1 ‘III .7
, l f
i ‘t I",
r ‘-. ..r
p ‘x' I/P\_,
- ./"'
\‘f
"3"").0‘! _
a :e :59 l i 4 35:
18)!
it's-mas
;' r. 1' i" t
- .1 i
r ,
z' .I
a. .-
c :_ .
i. '1‘
L ,fi' ‘x______
- m"). at .2", —-.-"-.
.33 It? 2"? 35'? 2.3% t i? '
)‘em
Figure 6.3 Excitation and fluorescence emission spectra of pyrene. A) excitation

scan taken on home-built instrument with hem = 395 nm.
B) excitation scan taken on LS-S with. See text for explanation in
peak intensities. C) emission scan taken on LS-S, Xex = 335 nm.

 

79

 

 

 

 

s ‘ "
l
I"
L
I
l
[' A
If
1.
l
L f-d—u-—~.—__.
' // \.\h‘--.
"x
.‘r” x
Rafi...
:- R‘-
i q.“-
! mm. at H
:5»: 255 .8»? :éfi :9 2.5-s .le zés zée 235 3:9»:
xex

I" nax=125.no
lr
l i" ii
r x

’5 '. B

. ‘.
r-f 3

J t

I

1' i".
'1 'u

.' 1.
L i.

it
.. a)
\i'.‘ p.
II .3
P \"\ ilI i} .’
'\ .' r.
n I ”:0 o 08 \.-H- II 1' i3"
‘_ r } v T " ‘ if.”
we :35 439 439 SEN as»: 5.3;?
Aem
Figure 6.4 Excitation and fluorescence emission spectra of biphenyl taken on

LS-S. A) excitation scan with lam = 315 nm. B) emission scan with
lax = 260 nm.

4000 '1

INTENSITY

 

 

80

 

 

 

Figure 6.5

. . . . , . .
300 350 400 450 500 550 600

>‘em (nm)

Fluorescence emission of naphthalene taken on home-build
instrument with l x = 280 nm and 16 nm bandpasses. A) 0.10 M
SDS, and B) 0.1034 3133 with 25 mM TlNO3 added. See text for
explanation.

81

markedly decreases the observed fluorescence intensity. This is due to an enhanced
probability of the excited singlet-to-triplet transition. The triplet state non-radiatively
decays because of quenchers, such as molecular oxygen, in the solution. The peak
observed at 570 nm is due to scatter excitation radiation of 286 nm in the second order.
The peak at ~505 nm is believed to be due to an an impurity in the quartz cell [4].

Deoxygenation of the heavy-atom containing solution removes the quencher
inhibiting the radiative triplet-to-ground state transition. The solution was
deoxygenated by adding sodium sulfite. Figure 6.6 shows the spectra of the
deoxygenated and non-deoxygenated heavy-atom containing solutions. The
deoxygenated solution has a shoulder present at 485 nm. Also, the peak at 505 nm has
increased in intensity and shifted to 510 nm. A phosphorescence spectrum of the
deoxygenated, heavy-atom containing solution is shown in Figure 6.7. The peak at
330 nm is due to delayed fluorescence, and phosphorescence peaks are observed at 480,
510, and 560 nm. The peak at 560 nm is due partially to scattered excitation radiation in
the second order. Therefore, to observe RTP, the analyte must have a mechanism to
enhance the sin glet-to-triplet transition, and the solutions must be deoxygenated.

All solutions for MS-RTP measurements were prepared by dilution of the stock
analyte solution in SDS. The working solution contained 0.10-0.15 M 30/70 Tl/Na DS
with 10 mM sodium sulfite to reduce oxygen. The sample cell used in these
experiments was a standard 1 cm"2 fluorescence cell modified so that a rubber stopper
may be used to form an air-tight seal. This minimizes diffusion of oxygen into the
solutions once they have been deoxygenated. The deoxygenation reaction of sulfite ion
was discussed in Chapter H1.

The concentration of the surfactant influences the time for deoxygenation of the
solution [3]. Figure 6.8 shows the rise in the RTP of 2-bromonaphthalene for two
surfactant concentrations, 0.15 and 0.10 M 30/70 T1/Na DS. These plots were obtained

with the home-built instrument used in phosphorescence mode. The instrument was

82

250 -

INTENSITY

 

 

 

O T—fff'rfvrrvrvr

,r..,.--.r....l
300 350 400 450 500 550 600
>‘em (nm)
Figure 6.6 Luminescence spectra of naphthalene in 0.10 M SDS with 25 mM

TlNO3 (dotted line), and deoxygenated solution (solid line)._

83

 

 

4000 '1
E
m
z
LLJ
|._.
.2.
o --r.,-.ir1.-.-.,f. fi-,....,....,
300 350 400 450 500 550 600
Aem
Figure 6.7 Phosphorescence spectrum of naphthalene in 0.10 M SDS with

25 mM TlNO and 10 mM NaZSO3. Delay time after HI-to—LO
transition of chopper is 10 us.

 

 

 

 

1500-]
E
(I)
Z
A I5
Z
o -5..,...f,....,-...,....,....,....,....,-...,
0 51015202530354 45
TIME (min)
3000-«
E
U)
B E
I-
E
o .5“,....r.-.fi,.-..,..-.,
0 5 10 15 20 25
TIME (min)
Figure 6.8 Rise in phosphorescence intensity for two concentrations of

surfactant. A) 0.15 M 30/70 Tl/Na DS. B) 0.10 M 30/70 Tl/Na DS.
Phosphorescence intensity is in arbitrary units.

85

programmed to sample the phosphorescence intensity every 15 s for 45 min. These
studies showed that for 0.15 M surfactant solutions, approximately 23 min. should be
allowed for deoxygenation inorder to observe RTP. For 0.10 M surfactant this time is
reduced to 10 min. Although decreasing the surfactant concentration decreases the time
required for solution deoxygenation, low surfactant concentrations should be avoided
because this decreases the micelle concentration. Low micelle concentrations will
increase the number of analyte molecules located within a micelle. If two analyte
molecules are located in the same micelle, triplet-triplet annihilation or eximer
formation can become the principal deactivation mechanism at room temperature [5].
This can be a cause of curvature in MS-RTP calibration curves [1].

The MS-RTP spectra of 2-bromonaphthalene, pyrene, and biphenyl are shown in
Figures 6.9, 6.10, and 6.11, respectively. These spectra all show phosphorescence peaks
red-shifted relative to their fluorescence peaks. 2—Bromonaphthalene has RTP peaks at
485, 515 and 565 nm. This latter peak is due partially to second-order scattered
excitation radiation. Pyrene has a low intensity peak at 520 nm with two more intense

peaks at 595 and 650 nm. Characteristic RTP peaks for biphenyl are at 443 and 471 nm.

CD-RTP Spectra

The potential advantage of CD-RTP is due to its partial immunity to
phosphorescence quenching by oxygen. This would simplify greatly sample preparation
by eliminating the most difficult step. A solution of 2-bromonaphthalene in 0.01 M CD
was prepared with 0.58 M DBE. Sodium sulfite was added to deoxygenate the solution
to maximize the RTP intensity obtained. The CD-RTP spectrum of this solution is
shown in Figure 6.12. This spectrum compares favorably with the MS-RTP spectrum in
Figure 6.9. However, although a "cloudiness is the solution" was noted by Scypinski
and Cline Love [6], considerable precipitation was observed here due to the addition of

DBE. This abundant precipitate is a spectroscopists nightmare and

86

 

 

 

 

4000 T
E
(f)
A é
Z
0 . . . - , a s . , . . r a ,
450 500 550 600
)\ em (nm)
7333=125 33
r
B
- K‘N -
I \ y" -
I. .. a f ‘\
'- fl \'~.\ 1' -
(I ‘\ 3’ I
.- .II‘ \\ .-// I‘
I" V "a
If X“,
I l:
_ {I \\v.fi/A'-\
.- f \K
_ x": \H.
.J/ ‘7'“
31333.33 M~
4:33 439 5:33 533 513 533 333 5.33
lent
Figure 6.9 MS-RTP spectrum of 2-bromonaphthalene in 0.15 M 30/70 T1/Na DS
with 10 mM Na2803. A) home-built instrument, and B) Perkin-

Elmer LS-S.

87

 

 

4000 a
E
U7
2
Lu
A E
o ..-.,..Efi--f.,....,.w.fi
450 500 550 600 650 700
)‘em (nm)
I IIQX=II39J39
.3
I .
I‘ B I \‘n,
I I
If .k'
I I
. I I.
If I
P ! ".I
I‘ I
I I
- I \
I 3.
P f," \__,/‘F"\

 

I I", .

’- V/ 67's

I 31313333 ..,/I \“-~
' 4' 333 333 53:37

433 533 533
x'eH'l

Figure 6.10 MS-RTP spectrum of pyrene in 0.15 M 30/70 TI/Na DS with 10 mM
NaZSO3. A) home-built instrument, and B) Perkin-Elmer LS-S.

823

; 33x=133.33

 

 

433 433 433 4533 4:33 5:33 353 543 333 5.33 :33

em

Figure 6.11 MS-RTP spectrum of biphenyl in 0.10 M 30/70 Tl/Na DS with
10 mM Na2803.

89

 

 

I’
I III-“Iii: 1 I33 . I39
/"A'\‘ A
I" f '1‘ if \‘
h\ r" x
3 v \
P 3" \‘z
3’ ‘l‘:
..I k.
’{l -.\
If \
3 K
_ I" ‘
l( ‘\'
P I. tax, ff, \
/ \‘J’J “X
:- I \\
. {I \\
j.
I" j X“ .
/
II I ":9 . 80 M_
l T T m r f T 1 . I _ f F—
469 438 583 523 $48 568 538 I363 629 643
Aem

Figure 6.12 CD-RTP spectrum of 2-bromonaphthalene in 0.01 M B-cyclodextrin
with 0.58 M DB E.

90
precludes the use of CD-RTP in a flowing stream where the particles could plug the

tubing and flow cell. Therefore, MS-RTP was chosen as the only viable method at

present for direct analyte phosphorescence measurements in a flowing stream.

Phosnhgrgggncg Lifetimg

Lifetime data are presented here only as a means of confirming the presence of
RTP. It must be noted here that since many of the compounds used in this research
were not purified meticulously, phosphorescence lifetimes obtained in these
experiments will not necessarily compare well with values found in the literature. The
phosphorescence lifetime of a compound is dependent on many factors such as the
concentration and type of the heavy-atom species, thee micelle concentration, and the
presence of any contaminants and quenchers that may be found in the analytical
solution. Equation 34 relates the phosphorescence lifetime to the rates of radiative and
non-radiative deactivation and the micellar concentration. Since neither instrument has
the capability for thermostatting the sample chamber, variations in ambient temperature
will also influence the observed phosphorescence lifetime.

The phosphorescence lifetime of naphthalene was obtained in 0.15 M 30/70
Tl/Na DS. This solution was deoxygenated with 10 mM sodium sulfite. The FORTH
program DECAY (blocks 70-72, Appendix A) was used to obtain the data for the
lifetime determination on the home-built instrument. This program uses the
counter/timer to count a delay time in us after the HI-to-LO transition of the chopper.
At the end of this count, the ADC is sampled and the data point is stored in memory.
For each delay time, the user selects the number of ADC outputs to be averaged. The
delay time is then incremented by a predetermined amount, and the process is repeated
over a range of delay times. The data are then stored on floppy disk and transmitted

serially to the LSI-11/23 for analysis and graphical presentation.

91

The phosphorescence decay is predicted to follow a single exponential decay. A
simplex fitting routine, XYFIT, written by Wentzell [7] is used to fit the data to the
following equation:

1pc) =10(e"‘t) + B (31)
In this equation, Ip(t) is the phosphorescence intensity observed at time t, k = 1/tp, and
B is included to account for any electrical and phosphorescence background in the
signal.

The phosphorescence lifetime was also determined on the Perkin-Elmer LS-S
with the program OBEY DECAY on the Model 3600 Data Station. This program
measures the phosphorescence intensity at six user selected delay times and uses a
Guggenheim algorithm to calculate the lifetime. One immediate disadvantage of this
system is that only six delay times can be used to calculate the lifetime, whereas the
home-built instrument is not limited in this respect.

Phosphorescence decay curves of naphthalene acquired on both instruments are
shown in Figure 6.13. The phosphorescence emission is monitored at 485 nm with an
excitation wavelength of 280 nm on bOth instruments. The Perkin-Elmer DECAY
OBEY program determined the lifetime to be 2.03 ms with a correlation coefficient of
0.9996, and a standard deviation of 0.062. These data were then fit to equation 61 with
XY FIT. This determination resulted in a lifetime of 1.72 ms (IO = 818, B = 50.5) with a
standard deviation of 1.13 for 6 points.

Results obtained on the home-built instrument yield a lifetime of 1.92 ms
(10 = 2547, B = -90) with a standard deviation of .009 for 60 points. Bearing in mind
that there was no temperature control through this series of experiments, the results
from bOth instruments agree quite favorably. The phosphorescence lifetime of
naphthalene found in the literature is 0.45 ms [8]. It is interesting to note that this
lifetime is considerably shorter than observed in this laboratory. The Stem-Volmer

equation predicts that quenchers and contaminants in the solution matrix will shorten

92

 

 

 

 

 

1000 ‘I
E
m
z
A E
Z
O I I l l r i
0.00 0.50 1.00 1.50 2.00 2.50 3.00
TIME (msec)
2500 a
E
B (13
Z
LLJ
'—
Z
O T I r r T 1
0.00 0.50 1.00 1.50 2.00 2.50 3.00

TIME (msec)

Figure 6.13 Phosphorescence decay of naphthalene. Data from, A) Perkin-Elmer
LS-S, B) home-built instrument. Curve through data points represents
the best fit to data as determined by the single exponential decay of
equation 6-1.

93

the observed lifetime. Since both instruments determined the lifetime of naphthalene to
be about 2 ms, instrumental errors are not believed to be the source of this error.
Investigation of this phenomenon is reserved for future work.

The lifetime of pyrene was investigated using the home-built instrument. The
decay of the phosphorescence (Figure 6.14) was fit to equation 61. The parameters of
this fit are 10:3741, It=1.139x 10“, and B=-130. The standard deviation is
0.0056. The lifetime of pyrene from this fit is 8.78 ms. Similar results were obtained
by another researcher in our laboratory using a different experimental set-up [9]. The
reported lifetime of pyrene is 0.93 ms [8]. The Stern-Volmer equation predicts that
quenchers present in the sample solution will decrease the lifetime of the phosphore.
The lifetime observed in these experiments are all longer than reported in the literature.
This contradicts the effect predicted by the Stern-Volmer equation. The
phosphorescence lifetime on naphthalene was determined to be approximately 2 ms on
both instruments using different algorithms to calculate the lifetime. Therefore,
systematic error is not suspected to be the source of discrepancy in the observed
phosphorescence lifetimes. The explanation of this observation of increased

phosphorescence lifetimes is reserved for future work.

3000 ‘I

INTENSITY

Figure 6.14

O 1 I Y I Y I T

94

 

 

 

, . .
0.00 2.00 4.00 6.00 8.00
TIME (msec)

Phosphorescence decay of pyrene. Data from home-built instrument.
Curve represents best fit of data to single exponential decay.

95

CHAPTER VI REFERENCES

Cline Love, L.J.; Skrilec, M.; Habarta, J.G., Anal. Chem, 1980, 52, 754

Kalyanasundaram, K.; Grieser, R; Thomas, J.K., Chem. Phys. Lett., 1977, 51,
501

Diaz Garcia, M.E.; Sanz-Medel, A., Anal. Chem, 1986, 58, 1436
Parker, C.A., Photoluminescence of Solutions, Elsevier: Amsterdam, 1968

Gratzel, M.; Thomas, J. K., In Modern Fluorescence Spectroscopy, Wehry, E. L.,
Ed; Plenum: New York, 1976; Vol. 2, Chapter 4 .

Scypinski, S.; Cline Love, L.J., Anal. Chem, 1984, 56, 322

Wentzell, P., XYFIT: a FORTRAN curve fitting program using the Simplex
algorithm, Michigan State University, 1986

Skrilec, M.; Cline Love, L.J.; Anal. Chem, 1980, 52, 1559

Kim, H.-D., Ph.D. research in progress, Department of Chemistry, and the
Pesticide Research Center, Michigan State University, 1987

CHAPTER VII
ROOM TEMPERATURE PHOSPHORESCENCE
with FLOW INJECTION ANALYSIS SAMPLE INTRODUCTION

In this chapter, a system is introduced for making RTP measurements in a
flowing stream. Flow injection analysis (FIA) is used as the means of sample
introduction. The flow cell used in these studies is described first. Then, experiments
to determine the optimum carrier are presented. Different carriers were used to pump
the sample zone , and the effect of these on the peak luminescence profile was
investigated. Based on the results of this study, the carrier yielding the best peak profile
was selected. Three methods of deoxygenating flowing streams are examined. The
results from these experiments determined the deoxygenation method that was used.
Phosphorescence calibration curves obtained with the RTP/FIA configuration are then
presented.

Flow In ['31ng System

The components of a typical FIA system, a pump, polymeric tubing, sample
injection valve, flow cell, and readout, were discussed in Chapter 5. In this work, a flow
through micro fluorimeter cell (type 43, Stama Cells Inc., Atascadero, CA) with an
internal volume of approximately 220 pl, 5 mm pathlengths, and four polished
windows is used. The cell is constructed from fused Spectrosil Quartz and is suitable
for use between 170 nm and 2600 nm. The cell is reported to show no fluorescence, but
does have a strong absorption band at 2700 nm [1]. The cell has external dimensions of

12.5 mm x 7.3 mm. Therefore, a template was constructed to hold the cell square in ’

96

97

sample chambers designed for 12.5 mm square cells. This template ensures that the cell
is centered in the sample chamber. The "U-shaped" template was constructed so that all
four windows of the cell are not blocked by the template.

The effect of the smaller solid angle of the flow cell on the luminescence
intensity observed was investigated by observing the fluorescence of pyrene in 0.15 M
SDS. Identical solutions were placed in the 1 cm2 cuvette and the flow cell. Spectra of
these solutions were then obtained on both the Perkin-Elmer LS-5 and the home-built
instrument. The intensities of the major fluorescence peaks for both cells were then
compared. Data from this experiment are shown in Table 7.1 These data show that
there is an 11.1% decrease in luminescence intensity in the flow cell as compared to the
static cell. Results of a similar experiment done on the home-built instrument are shown
in Table 7.2. The luminescence intensity with the flow cell is 21.5% less than with the
static cell in the home-built system. This larger difference in intensities on the home-
built instrument is due to the optical components. Lenses were not used to focus the
excitation radiation or the emitted luminescence from the cell in the home-built
instrument. A carefully matched optical system, such as that used in the LS-S, would
decrease this difference. To obtain maximum luminescence intensity, all subsequent

work is performed on the LS-S spectrofluorimeter.

Table 7.1
Comparison of Static and Flow Cell in Perkin-Elmer LS-5a

Fluorescence Intensityb
374 nm 385 nm 395 nm

 

 

 

 

Static cell 79.70 75.60 80.85
Flow cell 72.10 66.45 71.70
% decrease: 9.8 12.1 11.3

average decrease: 11.1%

 

a

b 3 nm bandpass

arbitrary units

98

Table 7.2
Comparison of Static and Flow Cell in Home-Built Instrumental

Fluorescence Intensityb

 

 

395 nm
M”““
Flow cell 2503
% decrease: 21.5
g 8 nm bandpass

arbitrary units

Qarfl'er Study

Due to mixing of the carrier and the sample zone within the FIA manifold, the
carrier will affect the luminescence peak profile. This effect was first observed when
data were being acquired for a fluorescence calibration curve of pyrene in 0.15 M SDS.
The fluorescence intensity of pyrene was monitored at 395 nm with an excitation
wavelength of 335 nm. Water was used as the carrier. A very non-symmetric peak
profile was observed, as can be seen in Figure 7.1A. A solution of 0.15 M SDS was
then used as the carrier, and, as can be seen in Figure 7.1B, there was a marked change
in the peak profile. These latter peaks show an increase in the observed fluorescence
intensity with a shorter time to return to the baseline. Two explanations of this
* difference can be postulated: ionic strength, and surface tension effects. The difference
in the ionic strengths of the sample zone and the water carrier can result in poor mixing
of the two solutions. As the interface of these two solutions passes through the flow
cell, the convective mixing can be observed as fluctuations in the fluorescence intensity.
Mixing of the sample zone and the carrier is also affected by the surface tension of the

surfactant solution. The lower surface tension of the surfactant decreases the extent of

99

 

 

 

 

 

 

 

 

 

 

4000 1
E
(I)
Z
A 3‘3
Z
O 'I'T'IfiTTI'I'I‘T‘IfW
O 200 400 600 800 1000
TIME (s)
4000 1
E
If)
B E
F—
E
0 A] r ~ . j .
0 100 200 300 400
TIME (s)
Figure 7.1 A) fluorescence peak profile of pyrene obtained with water as the

carrier, and B) SDS carrier.

100

mixing of the sample zone and the water. The observed effect is then the same as for
ionic strength differences.

An experiment was performed in which several solutions were used as carriers:
water, and 0.15 M solutions of SDS, NaCl, NaBr, NaC2H3OZ, and LiCl. The
characteristic peak profiles obtained are shown in Figure 7.2. Although the salt
solutions do not show the anomalous peak characteristic of water, the intensities of
these peaks are much less than for SDS. The baseline of these peaks is also quite noisy.
Because the salt solutions all have the same ionic strength as the SDS, the difference in
ionic strength in the solutions is not the cause of the decreased fluorescence intensities.
However, at the carrier/sample zone interface, the mixing of the solutions can cause a
change in the micelle concentration. This can be due to a change in the aggregate
number of the micelle through counter-ion effects. Therefore, to obtain maximum

signal intensity, SDS was used as the carrier for all subsequent experiments.

Deox n ion {F1 win tr

To obtain RTP, the analyte solution must be thoroughly deoxygenated as
discussed in Chapter III. For a flowing stream there is the additional problem that the
solution must remain deoxygenated as it flows through the manifold. Three methods of
deoxygenating flowing streams were investigated. Two of these methods
(Cr(ID/ZhO-lg) reduction of Oz, and a vacuum method) are based on the diffusion of
oxygen across a gas permeable membrane. The third method is the reduction of oxygen

by reaction with sulfite. This is an "internal" method and requires no special equipment.

ChrOmium(II)/Zn(Hg) Oxygen Scavenging
The Cr(II)/Zn(Hg) deoxygenation method is based on the production of a
concentration gradient across a gas permeable membrane. Oxygen diffuses across the

membrane into a scavenging solution containing acidic Cr(II). This solution

101

4000 -

 

 

 

 

E
m
z B
LIJ
1-
.2.
A C D
E F
O ‘ I . t I ' I ' I ' 1
0 600 1200 1800 2400 3000
TIME (s)
Figure 7.2 Characteristic fluorescence peak profiles obtained with A) water,

B) 3133, C) NaCl, D) NaBr, E) NaC2H3Oz, and F) LiCl.

102

immediately reduces oxygen to water. The Cr2+ is generated by the reduction of Cr3+
in presence of amalgamated zinc:
2c:3+ + Zn(Hg) -> 2c?“ + Zn2+ + Hg (7-1)
4Cr2+ + 02 + 4H3o+ —> 40:3+ + 61-120 (7-2)
This approach was originally reported by Rollie, Ho, and Warner [2] as a means to
minimize the quenching effect of oxygen on fluorescence. It was then developed by
Rollie, Patonay, and Warner [3] for the purpose of fluorescence enhancement in flowing
streams. An oxygen concentration gradient is developed across a 0.2'mil very low
density polyethylene membrane mounted in a 12 in. dialyzer assembly.

In the studies reported here, pyrene was used to evaluate the deoxygenation
apparatus because of its long fluorescence lifetime (450 ns [4]) and sensitivity to oxygen
quenching. Pyrene solutions were prepared in cyclohexane. The efficiency of
deoxygenation was determined by calculating an enhancement factor, which is the ratio
of the luminescence intensity of the deoxygenated solution to the intensity of solutions
open to air. Enhancement factors as large as 16 have been reported for pyrene by
Warner et al. [2,3]. In our studies, a 50 cm concentric tube dialyzer was constructed
with Gore-Tex tubing (TA001, W.L. Gore and Associates, Inc., Elkton, MD) as the gas
permeable membrane, as shown in Figure 7.3. This tubing has an internal diameter of
1.0 mm and a maximum pore size of 2.0 um. The large surface area of this dialyzer
design increases the diffusion of 02 into the Cr2+ reducing solution. This is compared
to a channel dialyzer where a flat membrane separates the two solutions. Pyrene
standards were prepared in 0.15 M SDS with concentrations over the range of 1-20 11M.
All solutions were air-equilibrated prior to running the experiment. The manifold used
is shown in Figure 7.4. Samples were injected in triplicate. The residence time of the
sample zone in the dialyzer at the carrier flow rate used is approximately 25 s. The
outer tube of the dialyzer was flushed with water to obtain the fluorescence intensities

of pyrene in air—equilibrated solutions. A solution of Cr(II) was

103

Cr(II)/

CFO”) Gloss Envelope
T (5 mm o.d.)
II I

Corrier/ \ Deoxygenoted \
Sample 7 Solution /

 

 

 

 

 

 

 

 

 

 

I I

Gore-Tex T
tubing

Cr(II)

Figure 7.3 Concentric tube dialyzer used for deoxygenation with Cr(II).

104

 

 

 

SDS

 

 

 

Figure 7.4

 

 

 

 

 

 

___.
/ Deoxygenotor
T FC
Pump Sample
Waste
Cr(II)

FIA manifold used for deoxygenation with concentric tube dialyzer

105

then pumped counter-current through the outer tube to deoxygenate the sample zone as
it flowed through the dialyzer. The fluorescence intensities obtained at 395 nm
(hex = 335 nm) are listed in Table 7.3. The peak profiles are shown in Figure 7.5A and

the average peak intensities are plotted versus the pyrene concentration in Figure 7.5B.

Table 7.3

Fluorescence Enhancement Factors

 

 

 

 

 

[Pyrene] air- deoxygenateda Enhancement
(1.1M) equilibrated‘al Factor
20 1663 2142 1.29 ._
15 1405 1 802 1.28
10 975 1335 1.37
5 528 758 1.44
1 1 19 164 1.38
average 1.35
a arbitrary units

Although the enhancement factors determined in our work are considerably less
than reported in the literature [2,3], the basic principle of fluorescence enhancement by
deoxygenation was observed. Immediate observation indicates that greater
enhancement factors could be obtained with increased residence time in the dialyzer
tube. However, during an experiment with a longer sample residence time in the
dialyzer, the waste solution was observed to be tainted green. Absorption spectra of the
pyrene sample, the chromium deoxygenating solution, and the waste solution were
obtained; These spectra, Figure 7.6, confirmed the suspicion that some chromium had
passed through the membrane and was contaminating the sample solution. The increase
in absorbance observed at wavelengths lower than 300 nm for the contaminated solution
(Figure 7.6C) is not present in the pyrene sample (Figure 7.6A), but is present in the

chromium solution (Figure 7.6B). A similar effect was noted when a second concentric

106

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

1200 n
E
(f)
z
A :2
E
III
0 T fl I ' I . H I . T ' I
0 600 1200 1800 2400 3000
TIME (s)
1200-
O
O

E
m
B E
1-
Z

0 ..4.rf.-.,...-,..-.,

0.0 5.0 10.0 15.0 20.0
[Pyrene] (1.1M)
Figure 7.5 A) Fluorescence peak profiles obtained for pyrene by FIA. Sample

concentrations from left to right are 20, 15, 12, 10, 7, 5, 2, and 1 11M,
respectively. B) Calibration curve using average maximum peak
intensity obtained from A.

107

 

Figure 7.6 Absorbance spectra of A) pyrene in 0.15 M SDS, B) Chromium
deoxygenating solution, and C) pyrene waste solution showing
contamination of pyrene sample by chromium.

108

tube dialyzer was constructed. It was therefore assumed that the Gore-Tex microporous
tubing is too fragile to withstand the construction of this dialyzer or that the pore size
allows diffusion of chromium. Membranes for a channel dialyzer were not readily

available; therefore, a new deoxygenation technique was investigated.

Vacuum Deoxygenation

Vacuum deoxygenation is also based on the use of a gas permeable membrane.
The oxygen concentration gradient results from a pressure difference created by
evacuation. This technique was introduced by Reim [5] for post column deoxygenation
for liquid chromatography with reductive electrochemical detection. The
deoxygenation device constructed in our laboratory is shown in Figure 7.7. The clear
plastic shell (7.25 in. x 2 in. o.d. x 1.5 in. id) is fitted with clear plastic end caps (3 in.
x 3 in. x 0.5 in.). A 1 7/8 in. O-ring is used to form the seal between the tube and the
end caps. The device is held together with four 3/8 in. assembly screws. One end cap
contains fittings for a vacuum pump connection and a bleed hole to relieve the vacuum
in the shell. Vacuum is supplied by a Cenco Hyvac 2 vacuum pump. This pump is
capable of developing a pressure of approximately 1 mtorr. The Other end cap contains
chromatographic fittings to which the silicone tubing (0.8 mmi.d. x 2.4mmo.d. x
3 m) is sealed with silicone sealant. Silicone tubing is used because of its high oxygen
permeability and chemical inertness [6,7]. Gore-Tex microporous tubing was not used
because it is not able to withstand the high pressure differential across the membrane.

Preliminary tests were performed on the dexoygenation apparatus to determine
its ability to maintain vacuum without leaking. Water was pumped through the silicone
tubing and the pump was turned on. The apparatus was able to hold a vacuum of less
than 1 torr as determined by a mercury manometer. When a small air bubble was
introduced into the tubing, the disappearance of this bubble in the silicone tube was
monitored while in the vacuum apparatus. It was noted at this point that small air

bubbles (~16 [11) took approximately 15 min. to diffuse through the membrane and

109

Assembly Screw End COP

 

Vocuum Silicone Tubing

Bleed Shell cm;

Figure 7.7 Vacuum deoxygenation apparatus.

110

disappear. The gas flux across the membrane wall, Jp, due to a pressure difference is
given by
Jp = -K (Ap/l) (7-3)

where K is the permeability coefficient of the membrane, Ap is the difference in gas
pressure at the inner and outer membrane wall, and 1 is the membrane thickness [10].
For a given membrane, this relationship shows that if the membrane thickness is
decreased the gas flux across the membrane will increase. Unfortunately, thinner
walled membranes were not conveniently available. Likewise, equation 7-3 predicts
that as the pressure differential Ap increases the gas flux will increase. This pressure
differential is limited by the vacuum pump and the ability of the deoxygenating
apparatus to hold vacuum. The vacuum generated in our apparatus was measured to

better than 1 torr, which is the maximum vacuum capable of being developed with the

pump used. Therefore, an alternate deoxygenation technique was evaluated.

Sodium S ulfite Deoxygenation

Sodium sulfite was previously shown to be effective in the deoxygenation of
RTP solutions prior to spectral scanning (see Chapter VI). For 0.10 M SDS solutions,
sulfite was shown to require a minimum of 10 min. for RTP to be observed (Figure 6.8).
Because of the gas permeability of the polymeric tubing used in FIA, the RTP of a
deoxygenated sample can be quenched due to diffusion of oxygen through the
membrane into the sample. Nylon tubing has been reported to show the least
permeability towards oxygen [7]. Therefore, with the exception of the pump tubes and
sample injection valve connections, nylon tubing (1 mm id) was used throughout the
FIA manifold. This manifold is shown in Figure 7.8. It was designed to keep sample
dispersion to a minimum to avoid mixing of the deoxygenated sample zone with the
carrier.

A probe solution of 10 11M naphthalene in 0.10 M 30/70 Tl/Na DS with 10 mM

NaZSO3 was prepared and allowed to deoxygenate for 15 min. The home-built

_

111

30 cm Nylon
tube

 

SDS

 

 

Pump

I 69‘?

~-—~ Sample

Figure 7.8 FIA manifold used for RTP determinations.

Waste

112

instrument was used as the detector with excitation and emission wavelengths of
280 nm and 485 nm, respectively. Several injections of the solution gave no discernible
RTP signal. This indicates that oxygen diffusing through the tubing was quenching the
RTP. Therefore, for one injection, the pump was stopped while the sample zone was in
the flow cell. After approximately 6.5 min., the characteristic rise in RTP was observed,
as shown in Figure 7.9. Oxygen that had diffused into the sample solution was reduced
to water during this period enabling RTP to be observed. Further diffusion of oxygen
into the sample solution was prevented by the quartz cell which is non-permeable to
gases. This method limits the procedure to a batch-mode analysis since time must be
allowed for the sulfite to deoxygenate the sample solution. Time must also be allowed
for the oxygen diffusing into the sample through the conduit to be reduced before RTP
can be observed. In spite of these time limitations, RTP can be observed in flowing

streams.

R A II tion urv

The studies discussed above show that at present the sulfite reaction is the most
successful deoxygenation method for RTP/FIA. With the sulfite deoxygenation
technique, calibration curves were obtained for naphthalene, 2-bromonaphthalene,
biphenyl, and pyrene in 0.10 M 30/70 Tl/Na DS. Analyte solutions were prepared
containing 10 mM NaZSO3 were allowed to deoxygenate for a minimum of 10 minutes
before use. The deoxygenated sample solution (500 11L) was then injected into the
0.10 M SDS carrier. Although the carrier was air-equilibrated, it could also be
deoxygenated with sulfite. The pump was stopped when the sample zone was in the
flow cell. The phosphorescence intensity was then monitored for 10 minutes. Each
sample is injected in triplicate and averaged. For naphthalene, 2-bromonaphthalene,
and biphenyl, the characteristic rise in RTP due to the reduction of oxygen by sulfite

was observed (Figures 6.8, 7.9). However, since the oxygen in the flowing stream is

113

1500 1

 

 

TIME (min)

Figure 7.9 Rise in RTP observed for RTP/PTA system.

114

due solely to diffusion of oxygen through the manifold tubing, the oxygen concentration
in the sample will be less than in air-equilibrated solutions. Therefore, the lag time prior
to the rise in RTP should be shorter using the RTP/FIA configuration than observed in
Figure 6.8. Figure 7.10 shows the rise in RTP observed for naphthalene samples with
concentrations of 1-140 11M. The lag time until RTP is observed for all samples in
Figure 7.10 is roughly 100 s. This shows that only a small amount of oxygen diffuses
through the tubing into the sample solution. The RTP intensity rises quickly to a
steady-state. The steady-state RTP intensities are plotted versus concentration in
Figure 7.11. Results of a weighted linear regression are listed in Table 7.4. The plot
exhibits a slight "S"-shaped curvature. This may be attributed to the technique used to
prepare the solutions. Samples with concentrations 30 11M and above were diluted
from the stock solution with a 100-1000 11L digital pipette; the 1-10 11M solutions were

diluted with a 10-100 [LL digital pipette. It is possible that one or both of the pipettes
are slightly out of calibration. This may account for the small offset observed in the

plot.

Table 7.4
Results of Weighted Linear Regression for Naphthalene Calibration Curve
Number of Points: 6
Slope: 5.32
Intercept: 3.99
Chi-square: 70.5

Correlation Coefficient: 0.995
Standard Deviation of the Slope: 0.045

Similar results were obtained for 2-bromonaphthalene. The lag time is
approximately 100 s. This is consistent with the lag time observed for naphthalene.

Steady-state RTP intensities for 2-bromonaphthalene are plotted plotted versus

115

 

 

f ' ' I ' 800

3 133 :33 333 433 333

T T ' ' r ‘ 800

 

‘.

3 133 233 333 433 533 333

 

250

#4 -
.4 I —“ vwv

..__—..__——--_._ __..q_

 

 

 

>4 ./

H la

I; l

c: /

23 __.,z

E: 3 133 233 333 433 533 533
8

C: 7 Y ' l’ 1m
3 I

8 E D A WWW V
‘5 . ,,_,WWMWW ’-
: I ,.-"‘

G. I J].

U! I .‘

0 ~
J: —-;"' ,
Q- 3 133 333 333 433 333 I ~

 

r 1 r r 40

_ . _. A __f . . ~ “fin.M'1kffi‘w’fi‘WL-flwwa-IM‘M.‘

. "l ‘.Y‘.-av‘t Vd‘fi- ""

 

 

 

'
I ..,” wan?“
I .J’
I
I LA.
I A M Y T . V
3 130 .33 :33 433 533 :50
i ‘ ‘ . 20
I . “.r» 'v‘ ..V‘d“.’.f.-.lv.l"h'd fi‘le’-‘ '~‘\!'.-'yc.b¢.l '4‘ .O.! ,-'.‘.-‘-I 5....“ .- -'O‘_’\"W “M .LMxh'. D'5NH.MII'M
-. ‘7 D- ’
Ir -J 0*
l "
I ,J
I -
’l'h"'.\ul‘.h" -m‘ 1*” f
. ._ L . . . . . . : . Tl
I 1m) .13 1'1 51ng $431] SI_ I I_I g ..I 4.1

Time (s)

Figure 7.10 Rise in RTP for naphthalene observed with RTP/FIA system for
A) l40uM, B) 100uM, C) 3011M D) 1011M, E)3uM, and
F) 1 1.1M solutions. Time rs relative to injection of sample.

116

 

 

10001
i
1
A 10041
E : o
(f)
z
LLJ
I'—
é 0
CD
0
— 101
1
0.1 1.0 10.0 100.0 1000.0

log[Nophtholene] (pM)

Figure 7.11 Calibration curve for naphthalene by RTP/FIA. RTP intensity is in
arbitrary units.

117

concentration in Figure 7.12. Results of a weighted linear regression are listed in

Table 7.5. The offset observed in Figure 7.11 is also present in Figure 7.12.

Table 7.5
Results of Weighted Linear Regression for 2-Bromonaphthalene Calibration Curve
Number of Points: 6
Slope: 7.16
Intercept: 6.61
Chi-square: 75.7

Correlation Coefficient: 0.997
Standard Deviation of the Slope: 0.048

The lag times for biphenyl samples with concentrations less than 100 M are
the same as naphthalene and 2-bromonaphthalene. The 100 M biphenyl solution has
a longer lag time and does not follow the rapid rise in RTP observed in Figures 6.8, 7.9,
and 7.10. The rise appears to be exponential, as shown in Figure 7.13. If this rise in
RTP follows an exponential equation, the steady—state intensity can be predicted from

1pm =1... (1 - cxp‘kt) (7-4)
where Ip(t) is the time-dependent phosphorescence intensity, I$3 is the steady-state RTP
intensity, k is a rate constant for the rise in RTP intensity, and t is the time in seconds
relative to the start of RTP, to. This time is approximated by taking an initial slope of
the curve above 500 s. This was done to avoid the initial induction period observed at
times earlier than 500 s. The value for to obtained in this manner is 406 5. Relative to
this time, Ip(t) and t are determined from Figure 7.13 and fit to equation 7-4 using a
Simplex fitting routine [8]. The results of this determination are; I33 = 769,
k = 0.00150. The sum of the squares of the residuals for 40 data points is 6160, and the

standard error of the estimate is 13. These data are shown in Figure 7.14.

118

 

 

1000 3
.
.l
.4
A 100 1 o
C 3
('7) .
Z 4
LL] I1
P— 0
Z .
‘6
o
“ 10 -:
1 f mm", . . ”..,, .-Hm, . . .m...
0.1 1.0 10.0 100.0 1000.0

log[2—Bromonophtholene] (1.3M)

Figure 7.12 Calibration curve for 2-bromonaphthalene by RTP/FIA. RTP
intensity is in arbitrary units.

 

119

 

Figure 7.13

Rise in RTP observed for 100 “M biphenyl.

,
HHX=899. 68 .
- ,xf

3.? ,— j”

m I;

I: ..,.P‘

2 I

5 [- ‘_/'

o i ..,/-

a ‘-

0 l f

U ,r

3 l

8 ‘- If:

'5- ! f

m r /

O I ,-"

u: {f

n. |_ ,-"
y n1n=3.33 " .
'3 533 13133 1533 2333 2533

Time (s)

120

1000 e

INTENSITY

 

 

 

I
O Y I' r r I I Y i f I I 1 I T ‘l’ ‘7 V ‘r 1' l r I r

' I
0 500 1000 1500 2000 2500
TIME (s)

Figure 7.14 Fit of rise in RTP for 100 1.1M biphenyl to an exponential equation.

121

The steady-state RTP intensities for biphenyl are plotted versus concentration in

Figure 7.15. Results of a weighted linear regression are listed in Table 7.6.

Table 7.6
Results of Weighted Linear Regression for Biphenyl Calibration Curvea

Number of Points: 4
Slope: 21.67
Intercept: -6.47
Chi-square: 24.36
Correlation Coefficient: 0.998
Standard Deviation of the Slope: 0.35

 

a 100 M sample not included in regression

The calculated intensity for the 100 M sample does not fall on the best fit line and was
not included in the regression. However, the curvature observed could be due to the
inner filter effect. The common test for the inner filter effect is to dilute the sample to
observe what effect this has on the phosphorescence intensity. In this experiment,
concentrations up to 30 11M fall on a linear curve which are essentially dilutions of the
100 M sample. Therefore, the inner filter effect may be a cause of the observed
curvature.

A second cause for the curvature could be due to an interaction of the biphenyl
with some component(s) in the sample solution. The long lag time prior to the
observance of RTP for the 100 M biphenyl sample indicates that the biphenyl may
have some effect on the sulfite-oxygen reaction or that the sulfite is reacting with the
biphenyl.

Pyrene also does not exhibit the expected sharp rise in RTP. The rise in RTP for
samples with concentrations 1-200 1.1M obtained with the RTP/FIA system are shown in

Figure 7.16. For all pyrene samples, to was determined and the rise in RTP was fit to

log(lNTENSITY)

Figure 7.15

122

 

 

1000 1
1 0
.I
l
100 d:
10 -I
i
1 . H.M., ..,”... ......., ”mm,
0.1 1.0 10.0 100.0 1000.0
Iog[Biphenyl] (,uM)
Calibration curve for biphenyl by RTP/FIA. Intensity of 100 M

sample obtained from equation 7-4. RTP intensity is in arbitrary
units.

 

 

 

 

B .fd-F -
’-'P'A..-‘.
' _fi______..
1 -------- '
3 133 :33 333 433 533 333 :33 333 533
I M.,-r”- W'
! -Jf.
I -.,!" '
I ’,.d".’
I -_.,..P"’
3 233 433 333 333 1333 1233

 

 

Phosphorescence Intensrty
U
I

3 :33 433 333 333 1333 1233 1433 1333 1333

 

 

3 233 433 333 333 1333 1233 1433 1333 1333‘

If 7 1 V f Y I 1 I a 40
|

I

I

I

1

 

F ”33.-3333334

 

13".“
33'1"
'- . “Aura“..- r. 11.. A n 1-- - 1‘. A _fi_._._...~,:"fl"'v‘
3 :33 433 333 333 1333 1:33 1433 13133 1333

Time (s)

Figure 7.16 Rise in RTP for pyrene observed with RTP/FIA system for
A) 200 uM, B) 100 M, C) 30 11M, D) 10 11M, E) 3 1.1M, and
F) l 1.1M solutions. Time is relative to injection of sample.

124

equation 74. These plots are shown in Figure 7.17. Table 7.7 lists the results of the fit

to equation 7-4.

Table 7.7

Parameters for Pyrene fit to Exponential Rise in RTP

 

 

[pyrene] n 133 R3 ssna $1513ID
(uM) (x10 )
1.0 31 30 3.33 37 1.
3.0 36 54. 1.83 18.6 .7

30.0 35 372 2.24 392
100. 36 704 2.13 628

0
10.0 34 141 2.34 122 2.
3
4.
200. 35 620 3.08 3773 11.

 

a

bSum of the squares of the residuals

Standard error of the estimate

The plot of Is,s vs. pyrene concentration is shown in Figure 7.18. In this plot, the 100
and 200 M samples show a negative deviation from linearity. This is believed to be
due to the inner filter effect for reasons mentioned above. Results of the linear
regression over the range of 1—100 1.1M are listed in Table 7.8. Although the linear
correlation coefficient for pyrene indicates the largest deviation from linearity for the
four compounds tested, calibration curves are obtainable from data that exhibit a slow
rise in RTP and do not reach a steady-state intensity. In Chapter IV reproducible tinting
was mentioned as one of the characteristics of FIA. Commonly in FIA, a steady-state
peak intensity does not have to be attained because of this characteristic. The
determination discussed above is an example of how reproducible timing can be used in

kinetic determinations in FIA.

 

125

 

 

 

 

 

 

 

 

 

 

 

600 , 600
I 00,0 I
XV Md)
6’6 or!”
A /0%)6 B of
/00 V7
o-Oo do
,0
C O C
17) p 17') ff
.2. f s
2 3 :2. 3"
1d
0 fij ' V ' 1 T Y 1 I f I O l f T r ' I ,7 f V V 1
0 100 200 300 400 500 600 700 800 0 100 200 300 400 500 600 700 800
nus (s) HME (s)
300 1 150 ~
c J/ o
E / E
m W 0
Z Z /a
23 33 1p
2.: z oo
09
0’
3?
‘0
6 f
D
0 4.
O r * T * rffi ' r ' I O ' r 3 1 I r ' 1 ' i
0 '30 700 300 400 500 600 ’00 800 0 100 200 300 400 500 600 700 800
TIME (5) TIME (s)
:30 1 30 '1
°o
yeti/“7°
O
O 0
I3 I: //°
ye
6°
C ,9/ 1‘:
- 9 171
2' 11006 E Ad
00 I
u (D
,3; ./
0
00° 09}
/’ .
O
3 3
Ofi—TTTTT'TI‘I r 0' " "v‘f'r‘fi
0 100 200 300 400 500 600 700 800 0 100 200 300 400 500 500 700 300
UME (s) UME (s)
Figure 7.17 Fit of rise in RT? for pyrene samples to equation 7-4. See text for

explanation. Curves correspond to A) 200 1.1M, B) 100 1.1M,

C) 30 M, D) 10 1.1M, E) 3 1.1M, and F) 1 1.1M pyrene sample
concentrations.

126

 

 

10001
3 0 o
1
0
A 1001
E :
U) « 0
Z 4
Le « 0
Z
‘6
.9
10'1
1 . .......r -......., ........, -.mm]
0.1 1.0 10.0 100.0 1000.0

log[Pyrene] (MM)

Figure 7.18 Calibration curve for pyrene by RTP/FIA. RTP intensities are
obtained from fit of data in Figure 7.17 to equation 7-4.

127

Table 7.8
Results of Weighted Linear Regression for Pyrene Calibration Curvea

Number of Points: 5
Slope: 8.57
Intercept: 24.51
Chi-square: 44.0
Correlation Coefficient: 0.960
Standard Deviation of the Slope: 0.054

 

a 200 uM sample not included in regression

This chapter demonstrates the first known use of FIA in connection with
micelle-stabilized RTP. Results obtained for naphthalene, 2-bromonaphthalene, and
biphenyl indicate that each sample requires about 5 minutes from injection into the FIA
manifold until a steady-state RTP intensity is reached. This translates to a sampling
frequency of approximately 12 samples per hour. An instrument developed by
Vo-Dinh, Walden, and Winefordner [10] for solid-surface RTP with a continuous filter
paper device requires approximately 3 minutes per sample. This corresponds to 20
samples per hour. Although the solid-surface RTP approach exhibits a higher sampling
frequency, the technique requires much more sophisticated equipment than the FIA
approach. This increases the cost of the apparatus required and increases the likelihood
of instrument down time. The studies reported here indicate that MS-RTP can be
automated with FIA as long as efficient deoxygenation is used. The sulfite
deoxygenation used in this research has the advantage that it is "internal" and is
therefore able to eliminate any oxygen that diffuses into the FIA system; however,
sulfite may not be compatible with all compounds used for phosphorescence
measurements. This compatibility would have to be investigated for each individual

analyte used.

99’!"

9°89!“

10

128

CHAPTER VII REFERENCES

Stama Spectrophotometer Cell Catalog, Stama Cells Inc., Atascadero, CA
Rollie, M.E.; Ho, C.-N.; Warner, I.M., Anal. Chem, 1983, 55 , 2445
Rollie, M.E.; Patonay, 6.; Warner, I.M., Anal. Chem, 1987, 59, 180

Birks, J.B., Photophysics of Aromatic Molecules; Wiley-Interscicnce: New
York, 1970

Reim, R.B., Anal. Chem, 1983, 55, 1188
Yasuda, H.; Peterlin, A., J. Appl. Polym Sci., 1973, I 7, 433
Gerritse, R.G., J. Chromatogr., 1973, 77, 406

Wentzell, P., XYFIT: a FORTRAN curve fitting program using the Simplex
algorithm, Michigan State University, 1986

Vo-Dinh, T.; Walden, G.; Winefordner, J.D., Anal. Chem, 1977, 49, 1126
Vo—Dinh, T.; Walden, G.L.; Winefordner, J.D., Anal. Chem, 1977, 49, 1126

CHAPTER VIII
WHAT’S NEXT?

Throughout the course of this research, experiments performed have provided
ideas for additional research. These projects can be divided into tWO categories,
phosphorescence lifetimes, and sample deoxygenation. This chapter discusses these
projects and proposes possible methods of approaching them.

In Chapter VI, the phosphorescence lifetime of naphthalene was determined
using two different algorithms. The lifetimes calculated by these methods are
significantly longer than reported in the literature. Three factors can influence the
phosphorescence lifetime: temperature, pmity of the analyte, and the nature and
concentration of concomitants in the sample matrix. The temperature dependence of
RTP will depend on the rate of micelle formation and dissociation. Therefore, a
decrease in the sample temperature should increase the phosphorescence lifetime due to
a decrease in the association/dissociation rates of the micelles. Therefore, temperature
control will be necessary for accurate lifetime determinations

Studies performed by Cline Love et al. [1] have shown that the dynamics of
micelle-analyte interaction will also influence the phosphorescence lifetime. The
dynamics of this interaction is a complicated process. The Stern-Volmer equation
predicts that the impurities in the reagents should decrease the lifetime. Therefore,
studies should be performed to determine accurately the RTP lifetime in micellar
solutions. All reagents should be purified meticulously prior to use. In the studies

reported in this dissertation, many reagents were used as supplied. Purification removes

129

130

impurities that decrease the phosphorescence yield and lifetime. Accurate lifetimes can
then be determined in these solutions.

Deoxygenation of flowing streams attempted by diffusion of oxygen across a gas
permeable membrane was not successful in this research. The two methods
investigated, vacuum deoxygenation and chemical scavenging, both failed due to
problems with the gas permeable membrane. The vacuum deoxygenation apparatus is
believed to have failed due to the thickness of the membrane wall. Use of a thinner-
walled membrane will increase the gas flux across the membrane. This should decrease
significantly the time required for deoxygenation of the flowing stream.

The concentric tube dialyzer failed as a deoxygenation technique because
reagents (e. g. chromium) crossed the membrane into the analyte stream. Chromium
present in the analyte solution can act as a counter-ion for the micelle and be held in
close proximity to the analyte. However, the low atomic number of chromium relative
to thallium and silver indicates that it will not be as efficient at spinoorbital-coupling as
thallium and silver. Therefore, RTP was not observed in the chromium contaminated
solutions. If the dialyzer is consu'ucted with a membrane that will not allow chromium
to cross, then this technique could be useful for deoxygenating flowing streams. The
Cr(II)/Zn(Hg) method has been reported in the literature [2,3] using a channel dialyzer.
The concentric tube dialyzer was attempted in this research so that the higher surface
area of the permeable membrane would increase the flux of oxygen across the
membrane. This will decrease the deoxygenation time.

The rise in RTP observed for pyrene and higher concentrations of biphenyl
shows a slow rise in RTP. The observed concenuationdependent rise in RTP for these
compounds should be investigated to determine the cause of this observed slow rise in
RTP. It is possible that there is an interaction between the analyte and the sulfite ion.
The source of this interaction was not investigated through the course of this research.

Studies should be performed using substituted pyrene derivatives or biphenyl

131

compounds to observe the effect substitution on the parent molecule has on the rise in
RTP.

The suggested improvements to the gas permeable membrane deoxygenators
would enable RTP/FIA to be performed on-line. On-line deoxygenation would enable
RTP/FIA to be used as a detection system for HPLC effluents. This arrangement could
use the HPLC column for sample separation and clean-up with the RTP/FIA system for

post column derivatization and detection.

132

CHAPTER VIII REFERENCES

Cline Love, L.J.; Habarta, J.G.; SIG-ilec, M., Anal. Chem, 1981, 53, 437

Rollie, M.E.; Ho, C.-N.; Warner, I.M., Anal. Chem, 1983, 55 , 2445
Rollie, M.E.; Patonay, 0.; Warner, LM., Anal. Chem, 1987, 5 9, 180

CHAPTER 1x
PORTABLE DUAL BEAM FIBER OPTIC-BASED
PHOTOMETRIC COMPARATOR”

This chapter presents work that was performed under a grant provided by
Neogen Corporation. Neogen desired an insu'ument to be used as a portable and easy-
to-operate detector for an immunoassay for aflatoxins. To meet these demands, an
inexpensive and portable dual-beam photometric comparator was developed. It has an
advantage over existing single beam systems in that direct comparisons about a critical
decision point can be made with minimal error from suay radiation. A light emitting
diode (LED) was used as the light source because of its fixed wavelength, narrow
emission profile, low current requirements, and low cost. The comparator was designed
for photometric determinations where it is desirable to know if the analyte concentration
is within a certain range. The sample introduction system can accommodate microtiter
wells (as cuvettes), such as those often used in routine immunoassay procedures. The
unit has been designed to be operated by non-technically oriented persons. It can be
battery powered for field use or operated from ac power.

Introduction

The development and use of chemicals in agriculture, animal husbandry, and
human health and nutrition have increased the awareness of the importance of the

chemical environment and the effects of synthetic compounds in the flora and fauna of

 

*Submitted for publication in Analytical Chemistry.
Co-authors: Adrian P. Wade, S.R. Crouch, J.F. Holland, and Brinton M. Miller

133

134

the earth. While modem research and service-oriented laboratories of industrialized
nations are utilizing increasingly sophisticated insu'umentation, there is a growing need
for compact, rugged, inexpensive, and portable instruments which can perform
determinations in regions isolated from the conveniences of modern laboratories. This
is a consequence of the demand for instant analyses in areas remote from established
facilities. In most situations a complete quantitative determination is not required.
What is required is an indication of whether the analyte concentration in a sample is
within or outside of an acceptable range; for example, are the active ingredients of a
chemical ueatment potent enough?; is the concentration of a trace contaminant below
established levels?; is it safe to eat?; is it safe to discard? In these instances the analyte
concentration in a sample need only to be compared to a standard. For example, since
toxin and threshold limit values are defined in the Food and Drug Administration’s
Code of Federal Regulations [1] and the Federal Register [2], compliance to these
rulings often reduces the determination of these species to a binary decision based upon
a threshold level. In this work we report the development of a dual-beam fiber optic-
based photometric instrument which directly compares two solutions. The unit
compares the transmittance of a sample to that of a reference and indicates whether the
sample contains a higher, lower, or a range of approximately equivalent analyte
concentration relative to the reference. Using the reference solution as the decision
point calibration standard reduces the determination to a single measurement. The unit
is light-weight and powered by either an automobile battery (through a 12 V cigarette
lighter adapter) or by ac line voltage with an external transformer unit.

Although this instrument was designed for binary decision analyses, the
comparator basis enables it to be operated in two additional modes: i), multi-reference
bracket and ii), semiquantitative mode. These modes can be more quantitative in their

analysis than the binary comparison mode. The design of the instrument allows non-

135

technically oriented persons to perform determinations precisely with a minimum
knowledge of how the instrument operates.

The methods described here are similar in principle to isomation or null-point
methods [3-6]. True null comparison methods require, however, that the reference
concentration (or unknown) be adjusted until a null point is reached. In the techniques
described here the comparison system is utilized to give an off-null or null indication.
No attempt is made to feed this information back to adjust one of the concentrations so

that a true null point is obtained. This system thus'acts as an open loop comparator [7].

Exmrimental

Optical Arrangement.

A light emitting diode (LED) is used as the light source. The LED was used as a
light source because of its relatively narrow emission profile (Max = 660 nm), its low
power requirements and its ruggedness. The bandwidth (ca. 15 nm) is narrow compared
to the absorption bandwidths of most solution absorbers; therefore no filters are
required. This significantly reduces the cost of the apparatus. Determinations requiring
a visible wavelength other than 660 nm may be accomplished by choosing a LED that
emits at the desired wavelength.

A bifurcated fiber optic is used for light transport and significantly simplifies the
optical layout. It also minimizes the effects of stray light and drift in the source LED
intensity. Lenses to focus the light onto the photodiode were not necessary using this
design. Silicon photocells were chosen for the detectors because of their high radiant

sensitivity in the visible spectrum where the intended analyses are to be run.

Analytical Manifold.
The manifold was constructed from aluminum and is shown in Figure 9.1. The

microtiter wells used as cuvettes for the determinations (Neogen Corp. 620 Lesher

136

WELL RACK

 

Figure 9.1. Manifold of the optical comparator showing configuration of the
source, fiber optic, and photodiodes. The grooved tracks minimize
scratches in the bottom of the analysis cups.

137

Place, Lansing, MI 48912) were found to be quite susceptible to scratching. This was
found to scatter light and have deleterious effects on the results. Therefore, the
instrument was designed with recessed tracks so that the centers of the bottoms of the
wells were not in direct contact with any part of the apparatus.

The strip of wells is placed in a rack which allows indexing of individual wells.
With this mechanism, entire strips are not required for an analysis. Individual well
placement was controlled by a double ball-bearing gripping mechanism designed in our
machine shop. This provides a high degree of precision in the placement of the wells, a

feature which was found to be quite critical if repeatable comparisons were to be made.

Circuit description.

The circuit design, Figure 9.2, is based on a light balance detection circuit
suggested by Hamamatsu Corporation [8]. The components of the comparator circuit
are listed in Table 9.1 This enables the photocurrents of the sample and reference
diodes (PD1 and PD2 respectively) to be nulled easily with just one offset adjustment
and reduces the number of electronic components required. With this circuit, the output
voltage (V o) of operational amplifier 0A1 is given by

Vo = Rf X (IshZ'Ishl) (9‘1)
where Rf is the feedback resistor and 13m and Isnz are the short circuit currents of the
two photodiodes. When light of equal intensity strikes both photodiodes, V0 is zero. In
the unbalanced state V0 is determined by the difference in the short circuit currents of
the photodiodes and the value of Rf. Capacitor C1 was included to slow the response of
the circuit via a low pass RC filter with a time constant of 0.3 s.

A window comparator circuit was selected since it enables the voltage levels for
the upper and lower boundaries to be set asymmetrically when the photodiode
photocurrents are balanced. Three conditions for the sample and reference
photocurrents are then possible: sample < reference, sample as reference, and

sample > reference. These boundaries are set with the trimming potentiometers which

138

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

I > +VCC
C] (lR4
r—lH -
p0] ‘ RI ’ jzv +VCC
SW] R4 R5
902 5A Q
fig : LED}
R4
i7 4m —-x—c2
-vcc
- (if-t4
>
‘4»
PANEL METER
1 I J 6 7 0 _ CC
[gm $4 R3 R2 V

+VCC

Figure 9.2. Circuit schematic for optical comparator circuit. Light from LED4
illuminates photodiodes PD] and PD2 through the bifurcated fiber

optic.

139

Table 9.1

Components of Comparator Circuit

Component

PDl, PD2
0A1
0A2
0A3
ICl
Rf

R1
R2
R3
R4
R5
C1
C2
LEDl
LEDZ
LED3
LED4

SW1
meter
FO

Value or type

Silicon Photocell (Hamamatsu #S 1336-5BK)
TL081 operational amplifier

TL082 operational amplifier

LM31 1 comparator

exclusive NOR gate

3.3 M52

10 m 10 turn potentiometer

470 k!)

10 k9 trimming potentiometer

10 k!)

390 0

0.1 11F

20 pF

Green light emitting diode

Red light emitting diode

Yellow light emitting diode

ER-300 High brightness (660 um, Radio
Shack #276—066)

2 pole 3 position rotary switch

Acculex #DP-352

Bifurcated Fiber Optic (#EK1512: Static
Controls Corp., 8240 Goldie St., Walled
Lake, MI 48088,

140

vary the reference voltage on the LM31] comparators (OA3). The window boundaries
can also be selected to account for any uncertainty in the determination. The
sensitivities of the comparators are limited partially by their hysteresis. The accuracy of
the binary decision can be better than 0.01 absorbance units, depending on the window
boundaries and the value of Rf this enables extremely precise decisions to be performed.

A panel meter was included primarily for diagnostic purposes. The window
boundaries may be adjusted accurately via a 3-way selector switch connected to the
reference voltages. This grounds the input of the comparator and allows the selected
reference voltage to be monitored. When both boundaries are set, the switch is then
positioned so that the output of the voltage follower is monitored. Thus, V0 is not
related directly to the absorbance of the solutions and serves only as a reference for the

magnitude of the concentration difference between the reference and sample solutions.

R l nd i i

The instrument was characterized with a dye of known absorbance
characteristics. Methylene blue was chosen since it has an absorption maximum at
660 nm which is at the maximum of the source specual distribution. Before inserting
the sample and reference wells, the output of OAl was nulled with the photodiodes
using air as the reference. This was performed to minimize errors emanating from
inconsistencies in the transmittances of the wells. This error may be minimized through
judicious selection of the window boundaries.

In the null-point comparison mode a reference solution is used to determine if
the concentration of the sample solution is greater than, less than, or approximately
equivalent in concentration to the reference. Dye samples with absorbances of 0.20,
0.21, and 0.25 were compared against a reference with an absorbance of 0.21. Results
were obtained which showed equality of the two solutions with the 0.21 absorbance.

The samples with absorbances of 0.20 and 0.25 were indicated as being respectively

141

’low’ and ’high’ on the LED output. Thus the unit has a precision of better than 0.01
absorbance unit. This may be further improved if necessary by increasing the resistance
of Rf.

Although the instrument was designed to be used as a decision point comparator
we found two other modes of operation to be available: multi—reference bracket, and
semiquantitative where the panel meter is read to determine the transmittance equality
of two solutions (or two wells).

The multi~reference bracket mode is used to approximate the concentration of
the analyte when a series of reference solutions are available. Reference solutions are
selected sequentially until consecutive reference solutions are found which have lower
and higher concenu'ations than the sample. The concentration of the sample solution is
then known to be within the range of the two reference solutions.

The multi-reference mode can also be used to determine if the analyte
concenu'ation is within an acceptable range. The window boundaries may be set with
reference solutions with concentrations that define the minimum and maximum
acceptable analyte concentrations. With the instrument calibrated in this manner,
results will be obtained that indicate if the analyte concentration in the sample is within
or outside of the defined window.

When a higher correlation between reference and sample concentrations is
required or if equality of two reference solutions of apparently the same concentration
needs to be checked, the instrument can be operated in a single beam mode where the
sample channel is referenced against air. The magnitude of the panel meter reading is
monitored for both of the solutions. Although the meter reading does not directly
indicate the difference in concentration of the two solutions, the concentrations will be
comparable when the meter reading difference is minimal. Instruments such as this and
that reported elsewhere [9] will play an increasingly important role in analytical

chemisu'y, particularly where resources are limited and low cost is important.

142

One application for which this instrument has been used is in the determination
of aflatoxins. Aflatoxins are potential carcinogens produced by certain types of molds
which may be found in corn, peanuts, and other food items. They are soluble in water
and in organic solvents such as methanol and ethanol but are insoluble in oils. These
fungal toxins are stable to heat, cold, and light and do not degrade naturally unless
metabolized by enzymes such as those produced in the liver or by microorganisms.
Aflatoxins are colorless, odorless, and tasteless. The toxic nature of aflatoxins was first
reported following a large "turkey kill” in 1960 in England [10,11]. The causative agent
was aflatoxins in moldy feed. Their detection is of considerable analytical interest and
has been possible by a competitive binding immunoassay procedure (Neogen Corp.).

In this analysis, an aflatoxin derivative is conjugated with an enzyme. This
conjugate is used as the ’known antigen’. Antibodies specific to the aflatoxin are coated
to the base of a plastic micrOtiter well. The aflatoxin in a food item is extracted with a
55% (v/v) methanol-water solvent. The extract is then mixed with the enzyme-
conjugate and the resulting mixture is placed in the antibody coated well. The aflatoxin
in the sample competes with the enzyme-conjugate for active sites on the attached
antibody. Unbound aflatoxin or enzyme-conjugate is then rinsed away. An enzyme is
then added to each well which catalyzes the color change of the subsuate from colorless
to blue. As the amount of aflatoxin in the sample increases the blue color intensity
decreases. The absorbance maximum of the substrate is at 650 nm. The substrate has
approximately 90% of its absorbance at 660 nm, the principal wavelength of the LED
used in this instrument.

Results have been obtained which have demonstrated that the instrument is
capable of discerning absorbance differences of 0.01 or less. The comparator nature of

this instrument enables semi-quantitation of sample concentrations to be performed.

9999’!"

10.
11.

143

CHAPTER IX REFERENCES

Code of Federal Regulations, Food and Drug Administration, vol. 21
Federal Register, July 30, 1982, vol. 47

Ramaley, 1.; Enke, G.G., Anal. Chem, 1965, 37, 1073

Malmstadt, H.V.; Winefordner, J.D., Anal. Chim Acta, 1960, 20, 283
Malmstadt, H.V.; Chambers, W.E., Anal. Chem, 1960, 32, 225

Malmstadt, H.V.; Enke, C.G., Electronics for Scientist, W.A. Benjamin, Inc.,
Menlo Park, CA, 1962, Chapter 6

Malmstadt, H.V.; Enke, C.G.; Crouch, S.R., Electronics and Instrumentation for
Scientism, Benjamin/Cummings, Menlo Park, CA, 1981

Hamamatsu Photocell Catalog, PO. Box 6916, Bridgewater, NJ. 08807

Bauer, W.E.; Wade, A.P.; Crouch, S.R., accepted for publication in Analytical
Chemistry

Sargeant, K.; Sheridan, A.; O’Kelly, 1., Nature, 1961, 192, 1096
Nesbitt, B.F.; O’Kelly, J.; Sargeant, K.; Sheridan, A., Nature 1962, 195, 1062

APPENDIX
FORTH SOURCE CODE LISTING

144

 

 

 

 

Block Number: 9

0 ( USER DISK DIRECTORY)

1 1023T'HRUzA/D l SAMPLEA/DOSAMPLE;

2 EXIT

3

4

5

6 CHECK THE SIGNAL TO NOISE RATIO .................. 40-41
7 GET SOME XY DATA WITH AVERAGING .............. 50-51
8 MANUALLY GET A SPECTRUM - -- 55-66
9 DECAY OF RTP PEAK _ ......70-73
10 RISE OF RTP - - ----76-78 CHOPPER MOTOR
CONTROL........ - 160-165

11

12

13

14

15

Block Number: 10

0 ( DAC WORDS AND CHIP-SELECT TESTING)

é DECIMAL

3 CODE 0DAC ( n->) 0 POP 0 .ODAC MOV B NEXT 0 0DAC
4 CODE 1DAC(n->)0 POP 0 .lDAC MOV B NEXT 0 lDAC
2 CODE 2DAC ( n->) 0 POP 0 .ZDAC MOV NEX I' 0 ZDAC
7 FORTH

8

9

10

1 l

12

Block Number: 11
( PIO INITIALIZATION BLOCK)
: SCUM-HEAD ;

8 CONSTANT .AIN 0 CONSTANT .AOUT
2 CONSTANT .BIN 0 CONSTANT .BOUT
4 CONSTANT .CUPPERIN 0 CONSTANT .CUPPEROUT
1 CONSTANT .CLOWERIN 0 CONSTANT .CLOWEROUT
: AolN .AIN OR; : A-OUT .AOUT OR;
: B-IN .BIN OR ; : B-OUT .BOUT OR ;
: C-UPPERIN .CUPPERIN OR ; : C-UPPEROUT .CUPPEROUT OR ;
10 :C-LOWERIN .CLOWERIN OR; :C-LOWEROUT .CLOWEROUT OR;
11 : SET-PIO 128 ; :PIO—INIT .PIOoINlT C! ;
12
13 SET -PIO A-OUT B-OUT C-LOWERIN C-UPPEROUT PIO—INIT
14
15 FORGET SCUM-HEAD

‘OOONOsUtJiWNt-oo

 

 

145

lock Number: 12
( MULTIPLEXER)

B
0
l .
2 CODEMUX (n->) OPOP OMUXADRMOVB NEXT OMUX

3 CODE SAMPLE ( n->) 0 POP 0 HOLDER MOV B NEXT 0 SAMPLE
5

6

7

8

FORTH EXIT

USE OF THE MULTIPLEXER:
The channel to be multiplexed to the ADC input is selected
10 by n MUX where n=0—>3 as indicated on the S/H chips on the
11 multiplexer card.
12 The status of the four inputs to the S/H’s are latched by
13 executing n SAMPLE. 0 SAMPLE clears all S/H’s, 1 latches Q0,
14 2 latches Ql, and 3 latches Q0 and Q1. "0 SAMPLE” MUST be
15 executed in order to update a S/H.

Block Number: 13

0 (ADC ROUTINES) HEX

1 CODE A/D ST CONV STA B ( start 12-bit conversion)
I-ILT ( F4 C, halt processor during convert)
STCONV LDA O 0 HI XCHG B 0 ROR O ROR

0 ROR 0 ROR 0 PUSH NEXT DECIMAL

: CHECK ZSOFF BEGIN CR A/D DUP l 4094 WITHIN NOT IF BELL THEN
9 U.R ?TERMINAL END ; . ‘

:DCHK ZSOFF BEGINA/D 0511 ODOA/DOD+ LOOP 512MlCR

10 DUP l 4094 WITHIN NOT IF BELL THEN 9 UR ?TERMINAL END ;
ll EXIT

12 CODE ACQUIRE STCONV STA B BEGIN ADSTAT LDA B

13 80 #B 0 AND 08 NOT END STCONV # W MOV

l4 W)0HIMOVBWINCW)0MOVB

15 OSHROSHROSHROSHROPUSHNHTDECMAL

Block Number: 14
0 (GOOD WORDS FOR THE 9513 COUNTER/TIMER)

1
CODE CMD COMAND # W MOV 0 POP STOS B NEXT
( sends data to command regisrer)

2
3
4
5 CODE @DRG DATAREG#WMOV IISUB OOSUB W)1MOVB
6 W)0MOVB 256#2MOV 2MUL10ADD OPUSH NEXT

7

8

9

‘O

‘OOOQGUOJIWN

( fetch the lo and hi bytes from the data-register and
pushes the 16-bit number onto the stack)

10 CODE >DRG 0 POP O DATAREG MOV B 0 HI DATAREG MOV B NEXT
11 ( puts a l6-bit number into the data-regiSter as lo byte

12 then hi byte)

13

14 CODE HALT HLT NEXT FORTH ( halts the processor till interrupt)
15

 

146

Block Number: 15
(9513 COUNTERfI'IMER)

BINARY

: CLK-INITIALIZE
11111111 CMD ( Masterresetfor 9513)
01011111 CMD ( load all counters; recommended for start-up)
11100111 CMD ( select 8-bit bus)
11101000 CMD ( disable data pointer sequencing)
00010111 CMD ( go to MaSter Mode)
10 1101000010110000 >DRG
11 ( BCD div, disable pointer, 8-bit, FOUT off, /1
12 SOURCE=F1, compares disabled. TOD disabled) ;
13 :CLK-INIT CLK-INITIALIZE CR ." 9513 INITIALIZED! " BELL:

15 CLK-INITIZLIZE DECIMAL

Block Number: 16
O ( DATA ACQUISITION ROUTINE - DATA BUFFER. CONSTANTS, MISC)
3072 CONSTANT DATA-BUFFER
VARIABLE DATA DATA-BUFFER ALLOT VARIABLE #PTS VARLABLE DARK
( A place to put all the data; the number of data points)

:FRESH DATA DATA-BUFFER ERASE 0 #PTS. ' ,FRESH
(Make sure the buffer rs all clear and there are 0 #pts)

: SPY ZSOFF DATA #PTS @ BDMP;
( Sneak a peak at what’s stored in the data buffer)

11 CODE lDATUM DATA # W MOV #PTS 0 MOV 0 SHL O W ADD 0 POP
12 STOS #PTS#WMOV1#0MOV 0W)ADD NEXT(49 usec)

l3 ( Stores the datum and increments the #PTS pointer)

14

15 CODE -DARK O POP DARK O SUB 0 PUSH NEXT FORTH ( 32 usec)

Block Number: 17
0 ( SAVE THE DATA INTO A BLOCK!)

ooeqasutsuNu—o

somqa‘uhutov—

1

2 ° ".DDISK“. " DATA. FTH "

3 (The name of the FORTH disk on the 11/23)

4

5 : DDISK llSPEC 25 BLANK [’1 ”DDISK" 3 + llSPEC 25 MOVE;
6 ( Put the disk name in llSPEC)

7

8 : "DATA" .” DATA DISK” ;

90 (Title of FORTH floppy data disk)

1

11 : ?RITE EMPTY -BUFFERS O BLOCK 9 0 DO DUP [’] "DATA" 3 4» 1+
12 C@ SWAP I + C@ = NOT IF Wrong disk you clod! ok"

13 CR DROP BELL QUIT THEN LOOP DROP;
( Are you trying to write on the correcr disk?)

u—au-o
UIJB

147

Block Number: 18
0 (DATA STORAGE)
( get the date and time from the 11/2.3)

VARIABLE D:T l6 ALLOT
: GTSTR BEGIN WCHARREC UIDATA C@ DUP 62 2 END DROP ;
:TMO.11"TM"UPCRGTSTR;

:?TIME'I'MD:TDUP17+DODUP30<IFDROP

10 R> 1+ >R ELSE I C! THEN- 1+LOOP BEGIN 0= END

11 D: TC@ 45=ABORT" LOG- INBONE-HEAD! " '

12

13 VARIABLE GAIN: ?GAIN CR." Gain faCtor? " >NUMBER GAIN. ';
14 VARIABLE PMT :?PMT CR." PMT voltage? " >NUMBER PMT !;
15

Block Number: 19
0 ( STORE THE DATA IN A BLOCK AND FLUSH IT)

scooqasuusuuu

1 VARIABLE IDENT 34 ALLOT

2 : ?IDENT IDENT 36 BLANK CR ." Please enter identifier: "

3 CNT@>RIDENT36EXPECTR>CNT!;

4

5 :?#BLKS #PTS @ 32+512/MOD SWAPIF1+T'I-IEN;

6 ( finds how many FORTH blocks it takes to Store the data)

7

8 : ?FBLOCK ?#BLKS 0 BLOCK 10 + DUP @ >R +! R) UPDATE FLUSH ;
9 ( leaves Start block number for Storage on Stack & updates)

10

11 : IDENTIFY-DATA-BLOCK ( Start block # ->) ?#BLKS 0 DO

12 12‘ DUP PREV 2! DUP I + IDENTIFY UPDATE LOOP DROP ;
13 ( identifies block number with block buffer and UPDATES)

14

15 :ZAP-BUFR FIRST@ DUP 5120 + SWAP DO OI! 2 +LOOP;

Block Number: 20
0 ( STORE THE DATA IN A BLOCK AND FLUSH IT cont.)

: INSRT ( b1k# ->) EMPTY -BUFFERS ZAP-BUFR FIRST @ DUP 64 BLANK
DUP D:T SWAP 18 MOVE (date and time)
DUP 18 + IDENT SWAP 38 MOVE (identifier suing)
DUP 54 + #PTS @ SWAP ! ( the number of points)
DUP 56 + ?#BLKS SWAP ! ( the number of FORTH blocks)
DUP 58 + GAIN @ SWAP! ( the keithley gain)
DUP 60 + PMT @ SWAP ! ( the PMT voltage)
DUP 62 + DUP >R ROT SWAP ! ( the FORTH data disk block)
10 64 + DATA SWAP #PT S @ 2“ MOVE ( get the data there)
11 CR ." Data Stored in block" R> @ 4 UR SPACE;

13 : STORE ?RITE ?TIME ?IDENT ?GAIN ?PMT ?FBLOCK DUP INSRT
14 IDENTIFY-DATA-BLOCK FLUSH;

\OOOQQLlhhbJNr-t

148

Block Number: 21
0 (READ LINE 0 OF DATA BLOCK)

1
2 :WHAT(n—>) BLOCK>R

3 I 18TYPECR (getdateand time)

4 118+ 36TYPECR (gettheidentifier)

5 IS4+@."#PTS-"5U.R CR

6 156+@."#BLKS:"SU.RCR

7 158+@."GAIN="5U.R CR (getthegain)

g R>60+@."PMT ="5U.RCR (thePMTvoltage);

10 : DDMP ( n->) EMPTY -BUFFERS DUP DUP WHAT

11 BLOCK 54+DUP@>R(#ptS)2+@ (#blocks)0DO
12 DUP I + BLOCK DROP LOOP BLOCK 64 + R> BDMP ;
13 ( dumps the data along with the header)

15

Block Number: 22
O ( ELAPSED TIMING using counter group #5 of 9513)

OCTAL : SETS CMD 5450 >DRG 15 CMD 0 >DRG;
: DISARM 401 CMD ; DECIMAL

: ELAPSE -’ NOT IF DISARM

SET 80 CMD 48 CMD (get ready)

EXECUTE ( start task)

176CMD21CMD (readtime)

@DRG 166 - ( subtract overhead)
10 CR ." Elapsed time: " 5 UR ." micro-seconds "
I; ELSE DROP ." There ain’t no such word Chump! " THEN;
I
13 ( Overhead determined as time taken between starting the
14 timer and reading the register. This was done by removing
15 the EXECUTE Statement.)

Block Number: 23
0 ( SEND DATA BLOCKS TO THE 11/2.3 AS RSX FILE)

:#blks (block # ->) BLOCK 56 + @;

:4-PIP .11" FPICTI:/L B";

:/R.ll" .RSX/R "DONE;

:DEBCRDUPSU.R;

10 :UPDAT ( block# ->) 4-PIP DUP (.) UPCHAR/R DUP

11 #blkS 0 DO DUP I + DEB BLOCK SENDDAT ! UPCHU CK LOOP DROP
12 SDONE;

l3

14
15

\quathNv-o

omflasuthNr-o

149

lock Number: 24
( AUTOMATED PUMP)
( Updated for new toutlly automated pump)

( Speed control)
: PMP 2585 1000 ‘l ; : PUMP ( n->) PMP lDAC;
: RACE 255 lDAC; : IDLE 2 lDAC; O PUMP

( Direction control)
VARIABLE <BPIO>

10 CODE FWRD 1 # <BPIO> OR <BPIO> 0 MOV 0 .BPIO MOV B NEXT
11

12 CODE RVRS 255 # 0 MOV l # 0 SUB 0 <BPIO> AND <BPIO> 0 MOV
13 0 .BPIO MOV B NEXT

14

15

Block Number: 40
0 ( VARY SLITS TO OBSERVE S/N - WHAT NOISE DO WE HAVE?)

OCIAL
: 2LOAD ( N->) 02 CMD 6445 >DRG 12 CMD >DRG 102 CMD;
( Use counter 2 Since it has the interrupt line;
source is F3 @ lOkHz; N a 10,000 for 1 second)

:2ARM 42 CMD ; (arms counteritl)

: 2DISARM 202 CMD ; (disarms counter #2)
DECIMAL

10

I; :#COUNTS CR ." Acquisition rate? (ms) " >NUMBER 10 "' 2LOAD ;
1

13 VARIABLE NPNTS

14 :HOW_MANY CR ." How many points? " >NUMBER NPNTS ! ;
15 : ?OK #PT S @ ABORT" Empty data buffer first clay-head ok" ;

Block Number. 41
O ( VARY SLITS TO OBSERVE SIN - WHAT NOISE DO WE HAVE? cont.)

160 LOAD ( load the chopper control)

: GET_DARK -LIGHT CR ." Hit key when ready to get dark current"
1KEY DROP
2ARM HALTA/DONPNTS@ l-ODO HALTA/DOD+LOOP
ZDISARM NPNTS @ Ml DARK! ;

omqaubuuuom

scooqasubutow

‘OOOKJO‘MhWNr—o

10 : LET ’S_GO CR ." Hit key when ready chump " BELL lKEY DROP;

l 1

12 : S/N ?OK #COUNTS HOW_MANY GET_DARK LET’S_GO +LIGHT ZARM
13 NPNT S @ 0 DO HALT A/D -DARK lDAT'UM LOOP 2DISARM -LIGHT;
14

15

150

Block Number: 55
0 ( LOAD BLOCK FOR SPECTRUM TAKING)

0 0DAC

1
i
4 :?TYPE160LOAD5657THRU

5 CR ." Use the chopper for delays? [YIN] "
6 KEY 89 (Y) -

7 IF ( chopped) 59 61 THRU

8 ELSE ( normal) 58 LOAD

9 THEN 62 65 THRU ;

11 ?TYPE

Block Number: 56
( BETTER SPECTRUM TAKING set-up)

VARIABLE WAVE ( the emission wavelength)
VARIABLE STEP ( the Step)

VARIABLE SET (the number of ADC counts to average)
VARIABLE CURRENT ( the current ave of ADC counts)
VARIABLE RAW ( the raw ave of ADC counts)
VARIABLE VDAC ( the 0DAC value used for chopper)
VARIABLE #TGR ( the delay time in usec)

:_SAVE EM WAVE @ !DATUM CURRENT @ lDATUM;
( Stores data as wavelength. intensity pairs)
:?R.EADY CR." Hit any key when ready [Qauitr IKEY 81 (Q)
12 2 IF." QUITok" CRQUITTHEN;
1
15

Block Number: 57
0 ( BETTER SPECTRUM TAKING the queStions)

: ?SET CR ." How many ADC counts to average? " >NUMBER SET ! ;
: ?Ll CR ." Initial emission wavelength: " >NUMBER;

‘OmflGthNt-‘O

Hunt-a
NHO

: ?L2 CR ." Final emission wavelength: " >NUMBER 1+;

: ?STEP CR ." Enter the step you want to use: " >NUMBER
0 STEP ! ;
1

ll :PAIRS CR2DUP-ABS STEP@/l+DUP."Pairs="3U.R
12 ." #PTS="2"3U.R;

14 : ?PREPS ?SET ?L1 ?L2 ?STEP PAIRS ;
15

\OWQOHJOAUJNH

151

Block Number: 58
0 ( BETTER SPECTRUM TAKING more words)

0 #TGR ! 0 0DAC «LIGHT

1
g
4 :GET_EMOO SET@ODOA/DOD+LOOP
5 SET@M/DUPRAW!-DARKCURRENT!;
6 (readsandsumstheA/Deounts)

7

8
9

: DRK CR ." Tune to take the background. " ~LIGHT ?READY
10 GET__EM RAW @ DARK ! +LIGHT;
11 ( gets the dark value and Stores it in DARK)

13 : TST +LIGHT CHECK -LIGHT;

Block Number: 59
( SET -UP THE 9513 FOR DELAY AFTER H->L CHOPPER TRANSITION)
OCT AL

: TGR

2 CMD ( mode #2) 175445 >DRG

12 CMD ( load #2) #TGR @ >DRG

102 CMD ( load regiSter) 42 CMD ( arm) ;
(act 10 edge gate2 ,count falling edge, F1, reload from

load, count repetitively/binary/down, act lo TC pulse.
Delays N us after HI->LO chopper signal and sends pulse to
10 interrupt line- used to Start micro again)
11 :TRIG ( n->) #TGR. ' TGR; ( Start trigger in one word)
12 : -TGR 302 CMD ; (Stops counter #2)
DECIMAL

14:MSK IMSK C@1 OR I.MSK Cl; (mask interrupt line)
15: -MSK IMSK C@ 254 AND IMSK Cl; (reset interrupt line)

Block Number: 60
( MORE NEET STUFF FOR DELAYED DATA TAKING IN SPECTRUM)
:GET_EM 00 SET@ 0DOI~IALTAIDOD+LOOP
SET@M/DUPRAW!-DARKCURRENT!;
( read and sum and average the ADC counts)

:dly 10000 0 DO LOOP ;

: DRK CR ." Take the background!" 0 0DAC dly ~LIGHT ?READY
OOSET@ 0DOA/DOD+LOOPSET@ MIDARK!
VDAC @ 0DAC TGR ;

11 : ?TGR CR ." What trigger delay? [usec] " >NUMBER #TGR ! ;

13 : ?DAC CR ." What 0DAC value? " 15 SPACES >NUMBER VDAC ! ;

15 : ?PREPS ?TGR ?DAC ?PREPS ;

\oooqasuthr-eo

\OOONGUMhUNr—oo

152

lock Number: 61
( USE THE DELAYED AID CONVERSION TO SET RANGES)

: DLCHK 250FF TGR VDAC @ 0DAC BEGIN CR HALT AID DUP 10 U.R
l 4094 WTTHIN NOT IF BELL TI-EN
?TERMINAL END 0 0DAC dly -LIGHT ;

:DLSET ?TGR ?DAC DLCHK;

omqauhuuwow

Block Number: 62
0 ( BETTER SPECTRUM TAKING presentation - box format)

: -> ( n->) 0 DO ZCRIGHT LOOP ; (moves right it spaces)

: BX CLR-SCR
20 3 CURSOR ." WAVELENGTH AVE. ADC AVE. ADC - DARK"
20 4 CURSOR ." (nm) (counts) (counts)"
16147 03 BOX EZG

10 32 2 CURSOR 115 EMIT ZCDOWN ZCLEFT 2 ZSIDES 117 EMIT

12 45 2 CURSOR 115 EMIT ZCDOWN zcuzr'r 2 ZSIDES 117 EMIT xzo;

‘OOOQOUJO#UDNH

Block Number: 63
( BETTER SPECTRUM TAKING presentation - continued)
: L25 SZCUR 250N 11 25 CURSOR ERALIN REVERSE ." #PTS= #
SETS/PT: TGR: DARK:- " RZCUR NORMAL ;
(sets up the 25th line)

: UD25 REVERSE 16 25 CURSOR #PTS @ 4 U.R
35 25 CURSOR SET @ 5 U.R
50 25 CURSOR #TGR @ 5 U.R
64 25 CURSOR DARK @ 5 U.R NORMAL ;
( updates 25th line information)
10 : NEXTWAVE WAVE @ REVERSE . NORMAL 10 -> ." " 10 ->
11 ." " RZCUR;
12 ( reports the next wavelength in table in reverse video)

14 : REPORT RZCUR WAVE @ . 10 -> RAW @ 4 U.R 10 ->
15 CURRENT @ 4 U.R BELL; (reports the current values)

\OWQGM&UJNHC

153

Block Number: 64
0 ( BETTER SPECTRUM TAKING a1m05t there)
1

2 : SCREEN BX L25 UD25 23 6 CURSOR SZCUR;

3

4 : ?SURE #PTS @ ABORT" You’ve already got data dope! ok" ;
5

6 : GRSU GET_EM REPORT SAVE_EM UD25 ;

8 :CIRCLE#PT320>IF15CURSURDELLIN2320CURSORTHEN;

Block Number: 65
0 ( BETTER SPECTRUM TAKING the final frontier)

: SET-UP 1 8 CURSOR ?PREPS ;
: 23del l 22 CURSOR DELLIN DELLIN DELLIN ;

: SPECTRUM ?SURE CLR-SCR 0 DARK !
SET-UP ( get initial information)
CR CR DRK (get background)
SCREEN ( set-up the video screen)
SWAP DO 23del RZCU R INSLIN RZCU R

10 I WAVE ! NEXTWAVE

11 1 23 CURSOR ?READY RZCUR GRSU STEP @ +LOOP

12 1 23 CURSOR ERALIN ." DONE!!! " 23del

13 O 0DAC;

14

15

Block Number: 70
( WATCH DECAY OF RTP PEAKS)

160 LOAD ( the chOpper Stuff)
56 57 THRU (variables and core words)
59 61 THRU ( set-up 9513 and more neet words)

VARIABLE Tl ( initial delay time)
VARIABLE T2 ( fmal delay time)

SOOOQOxMJIUJNr-e

wooflmuwar-‘C

: ?Tl CR ." Enter initial delay time [usec] " >NUMBER Tl ! ;
: ?T2 CR ." Enter final delay time [usec] " >NUMBER 1+ T2 ! ;

12 :TOTPTS CRT2@ Tl @ ~STEP@I1+2"' ." #PTS =" DUP4U.R

13 2" DATA-BUFFER > IF CR ." TOO MUCH DATA FOR BUFFER! ok" BELL
14 CR QUIT THEN ;

15 : QUESTIONS ?T1 ?T2 ?STEP TOTPTS ?DAC ?SET ;

p—op—e
HO

 

 

154

Block Number: 71
0 ( WATCH DECAY OF RTP PEAKS cont.)

‘OOOQOsMhNNe-a

10

: ?SURE #P'T S @ ABORT" You’ve already got data clay-head! ok" ;

:L25 ZSOFF 250N SZCUR1125 CURSOR REVERSE ." #PTS -
#SETIPTa TGR- DARK: " NORMAL RZCUR ;

: UD25 SZCUR REVERSE 16 25 CURSOR #PTS @ 4 U.R

35 25 CURSOR SET @ 5 U.R
50 25 CURSOR #TGR @ 5 U.R
6425 CURSORDARK@ 5U.RNORMALRZCUR;

Block Number: 72
0 (WATCH DECAY OF RTP PEAKS cont.)

: REPORT ( time->) CR ." delay time:" 5 U.R ." intensityz"

CURRENT@5U.R(." STD-I" STD@5U.R)UD25;

: PREPARE ?SURE CLR-SCR O DARK ! 0 0DAC 1 10 CURSOR

QUESTIONS DRK VDAC @ 0DAC T1 @ #TGR ! TGR L25 UD25
CR CR ." Time to decay!" ?READY;

: DECAY PREPARE CR

T2@ T1@DO I#TGR!TGR UD25 GET_ EM IWAVE!
SAVE_ EM IREPORT STEP @ +LOOP CR." frequency teSting:"
FRK 0 0DAC ZSOFF dly -LIGHT;

Block Number: 76
0 ( WATCH RISE OF RTP AFTER ADDTTION OF SULFTTE- initialization)
1 160 LOAD ( chopper Stuff)

VARIABLE TTME ( how many seconds to take data)
VARLABLE INTRVL ( how often to take data: see)
VARIABLE LAG ( ignore time for mixing: sec)
VARIABLE SETS ( how many datum to average)
VARIABLE DELAY ( delay time after H->L: usec)
VARIABLE DAC ( DAC value for chopper)

:?TTME CR." Minutes to take dataz" ">NUMBER 60 " TIME '
:?INTRVL CR. Enter interval [sec]: ”>NUMBER 100 * INTRVL ! ;

:?LAG CR." Enter lag time [sec]: " >NUMBER l- LAG ! ;

: ?SETS CR ." # sets to average: " >NUMBER SETS ! ;
: ?DELAY CR ." Delay time [usec]:" " >NUMBER DELAY ! ;
: ?DAC CR. Enter dac value: ">NUMBER DAC. ';
:>#PTS TIME@ LAG@ 1+-INTRVL@100//1+2‘."#PTS-
: PREPARE ?T'IME ?INTRVL ?LAG CR >#PTS ?SETS ?DELAY ?DAC;

155

lock Number: 77
( WATCH RISE OF RTP AFTER ADDTTION OF SULFITE- CLOCK SET-UP)

OCTAL

: 1TMR 10th 007442 >DRG 11 CMD INTRVL@ >DRG 101 CMD;
:lARM 41 CMD; (uses F5 and counter #1 to count sec)

: 2T'MR 2 CMD 175445 >DRG 12 CMD DELAY @ >DRG 102 CMD ;

: 2ARM 42 CMD ; ( wait for H->L on chopper then delay)

DECIMAL

CODE T'URNOVER BEGIN COMAND#WMOV 161#0MOV OW)MOVB
10 17#0MOV OW)MOVB

11 llSUB OOSUB DATAREG#WMOV W)1MOVB

12 W)0MOVB 256#2MOV 2MUL lOADD ODEC

13 OaEND NEXTFORTH

14(readho1dreg.ofcntr1andwaitforTC)

15

Block Number: 78

( WATCH RISE OF RTP AFTER ADDTTION OF SULFITE- )

: CLOCKS lTMR lARM 2TMR ZARM ;

:GDATOOSETS@ 0DOHALTAIDOD+ LOOP SETS @ Ml-DARK;

CODE INT/lOOINTVRLOMOV 100#WMOV CWD WIDIV OPUSHNEXT

:CHT CLOCKS TTME @ 1+ INT/100 DO TURNOVER CRI DUP 5 U.R
." seconds" LAG @ > IF BELL I EDATUM
." ave." GDAT DUP lDATUM 6 U.R THEN INTRVL @ 100 / +LOOP ;

10 : GDRK -LIGHT CR ." Hit key for dark." lKEY DROP

11 oosrsrsco 0DONDOD+LOOPSETS© M/DUPSURDARKl;

12: ?RDY CR. " Hit any key when ready [Q to quit]" IKEY 81

13 21F.ok" CRQUITTHEN;

14: RTP-RISE #PTS @ ABORT” Do FRESH“ PREPARE GDRK DAC @ 0DAC
15 ?RDY CHT o 0DAC 1000 0 DO 100 0 DO LOOP LOOP -LIGHT;

quasutsuNt—ow

cooqaxutszv-eo

lock Number: 160
( NUDGE TI-E CHOPPER WI-EEL A LITTLE)
: NUDGE 150 0DAC 150 0 DO LOOP O 0DAC ;
( moves the chopper wheel a little bit)

:SPD(n->)[’]NUDGE2+!; :?SPD(->) [’]NUDGE2+@;
( changes speed) ( fetches speed)

:DLY(n->) lfi'] NUDGE8 +1; :?DLY(->) [’] NUDGE8 +@;
( changes delay) ( fetches delay)

: ?STAT .CPIO C@ 1 AND ; : WHEEL NUDGE 1000 0 DO LOOP ;
10 (checks Status of C0) (moves wheel and waits)

12 : -L1GHT ?STAT IF BEGIN WHEEL ?STAT NOT END TT-EN;
13 (makes sure light path is blocked)

14 : +LIGHT ?STAT NOT IF BEGIN WI-EEL ?STAT END TTEN ;
15 (makes sure light path is nOt blocked)

 

B
0
1
2
3
4
5
6
7
8

‘0

 

 

156

Block Number: 161
0 ( CHOPPER STUFF USING GATE #1 AS COUNTER AND #2 PERIOD TIMER)

BINARY
: CHOP-SET 11000011 CMD (disarm 1&2)
(#1: count pulses) 00000001 CMD 0000011100101000 >DRG
00001001 CMD 0 >DRG (load w/zero)
(#2: time 1 sec) 00000010 CMD 1100111000000101 >DRG
00001010 CMD 1111101000 >DRG ( 1000!);
DECIMAL

10: 1CHOP CHOP-SET 99 CMD ( load&arm) HALT ( halt while counting)
11 ( output #2 is active low TC pulse which Starts the micro)

12 161 CMD (save counts) 17 CMD ( hold reg. #1) @DRG;

13

14 :?FREQ 1CHOP CR .” Frequency 8" 5 U.R SPACE ." Hz " CR ;

15

Block Number: 162
( CHOPPER STUFF cont.)

VARIABLE FREQ-CNT 2 ALLOT ( hold the frequency counts)

: lSEC ( 1 second period) 1000 ['1 CHOP-SET“ + ! ;
: lOSEC( 10 second period) 10000 [’] CHOP-SET 44 + ! ;

: INTI-FRK CHOP-SET 0. FREQ-CNT 2! ;

:?FORMAT[’] CHOP-SET44+@ 10000=IF<###46 HOLD #8 #>
10 ELSE<##46HOLD#S#>THEN;

12: FRK INIT- FRK SZCUR
13 100DORZCURII+4UR1CHOPOFREQCNTD+l LOOP
l4 FREQ-CNT 2@ RZCUR ?FORMAT 2 SPACES TYPE. Hz" SPACE ;

‘OOOQOM&NNH

HO

Block Number: 164
( LET MICRO SET THE CHOPPING FREQUENCY)

VARIABLE TOLERANCE 1 TOLERANCE ! (default to +/- 1%)
. %TOLERANCE TOLERANCE. ';

2
3
4
5 :+LMT ( n ->) 100 TOLERANCE @ + 100 v;
6 =-LMT ( n ->) 100 TOLERANCE @ - 100 v;
7

8

9

v-‘O

: FRK-90K ( actual, desired->F) DUP -LMT SWAP +LMT WITHIN ;

157

lock Number: 165
( LET MICRO SET TIE CHOPPING FREQUENCY cont.)
VARIABLE FREQUENCY VARIABLE SUPPLY 75 SUPPLY !
: STATUS-REPORT ( freq->) 5 U.R SUPPLY @ 5 U.R 13 MT ;

: ADJUST ZDUP a E 2DROP1 ELSE > E -1 SUPPLY +1
ELSE l SUPPLY +! TEEN SUPPLY @ 0DAC 0 TIEN ;

:ALRIGHT ( ." WITHIN" TOLERANCE @ 3 U.R ." % TOLERANCE!") ;
10 : WHAT_TODO ( Sl,SZ,F->) E ADJUST ELSE ALRIGHT 2DROP 1 THEN ;

12 : SPEED-CONTROL 1CHOP ( DUP STATUS-REPORT)
13 FREQUENCY @ 2DUP FRK-?OK NOT WHAT_TODO;

15 : HZ FREQUENCY ! SUPPLY @ 0DAC BEGIN SPEED-CONTROL END ;

@mqmuhuuwow