An Introduction to
Fluorescence
Spectroscopy
2
User Assistance
PerkinElmer Ltd
Post Office Lane
Beaconsfield
Buckinghamshire
HP9 1QA
Printed in the United Kingdom.
Notices
The information contained in this document is subject to change without notice.
PerkinElmer makes no warranty of any kind with regard to the material, including, but
not limited to, the implied warranties of merchantability and fitness for a particular
purpose.
PerkinElmer shall not be liable for errors contained herein for incidental consequential
damages in connection with furnishing, performance or use of this material.
Copyright Information
This document contains proprietary information that is protected by copyright.
All rights are reserved. No part of this publication may be reproduced in any form
whatsoever or translated into any language without the prior, written permission of
PerkinElmer, Inc.
Copyright © 2000 PerkinElmer, Inc.
Trademarks
Registered names, trademarks, etc. used in this document, even when not specifically
marked as such, are protected by law.
PerkinElmer is a registered trademark of PerkinElmer, Inc.
An Introduction to Fluorescence Spectroscopy
3
Table of Contents
Table of Contents............................................................................................ 3
Luminescence and the nature of light............................................................. 5
Fluorescence................................................................................................... 7
Phosphorescence........................................................................................... 11
Light scattering............................................................................................. 14
Rayleigh-Tyndall scattering .................................................................. 14
Raman scattering ................................................................................... 14
Quantitation .................................................................................................. 16
Sensitivity..................................................................................................... 18
Instrumental sensitivity ......................................................................... 18
Method sensitivity ................................................................................. 18
Absolute sensitivity ............................................................................... 18
Instrumentation............................................................................................. 21
General .................................................................................................. 21
Light sources ......................................................................................... 22
Wavelength selection............................................................................. 24
Detectors................................................................................................ 25
Read-out devices ................................................................................... 26
Sample holders ...................................................................................... 26
Problems of high blank values...................................................................... 28
Cuvettes................................................................................................. 28
Solvents and Reagents........................................................................... 28
Other contaminants................................................................................ 29
Working with dilute solutions ...................................................................... 30
General .................................................................................................. 30
Adsorption............................................................................................. 30
Photo-decomposition............................................................................. 31
Oxidation............................................................................................... 31
Factors affecting quantitative accuracy ........................................................ 32
Non-linearity.......................................................................................... 32
Temperature effects............................................................................... 32
pH effects............................................................................................... 33
Inner-filter effects.................................................................................. 33
Quenching.............................................................................................. 33
References .................................................................................................... 34
An Introduction to Fluorescence Spectroscopy
4
An Introduction to Fluorescence Spectroscopy
5
Luminescence and the nature of light
A hot body that emits radiation solely because of its high temperature is said to
exhibit incandescence. All other forms of light emission are called
luminescence.
When luminescence occurs, the system loses energy and if the emission is to be
continuous, some form of energy must be supplied from elsewhere. Thus the
radioluminescence emitted from a luminous clock face is supplied by high
energy particles from the radioactive material in the phosphor and the
electroluminescence of a gas discharge lamp is derived from the passage of an
electric current through an ionised gas. Other such phenomena include
chemiluminescence, derived from the energy of a chemical reaction, and this is
called bioluminescence when the reactions take place within living organisms,
for example, glow-worms and fireflies.
When the external energy supply is by means of the absorption of infrared,
visible or ultraviolet light, the emitted light is called photoluminescence and this
is the process that takes place in any fluorimetric analysis.
To account adequately for the processes of absorption and emission of light, it is
necessary to assume that radiant energy can only be absorbed in definite units, or
quanta. The energy, E, carried by any one quantum is proportional to its
frequency of oscillation, that is
E = hν = hc ergs
λ
where ν is the frequency, λ the related wavelength and h = Planck's constant
(6.624 x 10-27
ergs/seconds).
The energy of a single quantum is too small for convenience and it is usual to
talk of the energy associated with N quanta (where N = 6.023 x 1023
the number
of single molecules in a gram molecule), which is called an einstein. Thus, if in a
photochemical reaction one molecule reacts for each quantum absorbed, then the
absorption of one einstein is sufficient energy for the reaction of one gram mole.
Since the amount of energy per einstein is proportional to the frequency of the
radiation, it varies enormously over the range of the electro-magnetic spectrum,
as shown in the following table.
An Introduction to Fluorescence Spectroscopy
6
Approximate sizes of Quanta
Radiation ν(cm)
(typical
values)
Wave-
number
(µm-1
)
Size of
quantum
(electron
volts)
Size of
einstein
(kilogram
calories)
Absorption or
emission of
radiation
involves
Gamma rays 10-10
106
1.2 x 106
2.9 x 107
Nuclear
reactions
X-rays 10-8
104
1.2 x 104
2.9 x 105
Transitions of
inner atomic
electrons
Ultraviolet
Visible
10-5
4 x 10-5
8 x 10-5
101
2.5
1.25
1.2 x 10
1
3.1
1.6
2.9 x 102
7.1 x 101
3.6 x 101
Transitions of
outer atomic
electrons
Infrared 10-3
10-1
1.2 x 10-1
2.9 Molecular
vibrations
Far infrared 10-3
10-2
1.2 x 10-2
2.9 x 10-1
Molecular
rotations
Radar
Long radio
waves
101
105
10-5
10-9
1.2 x 10-5
1.2 x 10-9
2.9 x 10-4
2.9 x 10-8
Oscillation of
mobile or free
electrons
The ultraviolet and visible regions of the spectrum are of most interest in
fluorimetry and absorption in these regions causes the excitation of the
outermost electrons of the molecule. The energy associated with radiation of this
frequency is quite high, around 100 kilogram calories per einstein, and is
sometimes sufficient to break down the absorbing molecules, as for instance
with the fading of dyes by the action of sunlight.
The absorption of light results in the formation of excited molecules which can
in turn dissipate their energy by decomposition, reaction, or re-emission. The
efficiency with which these processes take place is called the quantum efficiency
and in the case of photoluminescence can be defined as:
ФE = einsteins emitted or No. of quanta emitted
einsteins absorbed No. of quanta absorbed
and never exceeds unity.
An Introduction to Fluorescence Spectroscopy
7
Fluorescence
At room temperature most molecules occupy the lowest vibrational level of the
ground electronic state, and on absorption of light they are elevated to produce
excited states. The simplified diagram below shows absorption by molecules to
produce either the first, S1, or second S2, excited state (Figure 1).
Figure 1 Transitions giving rise to absorption and fluorescence emission
spectra
(10 – 10 sec)
Vibrational
Levels of
second
excited state
Vibrational
Levels of first
excited state
Vibrational
Levels of
ground state
Rotational levels
Fluorescence
(10
-9
– 10
-6
sec)
Internal
(<10
-12
sec)
Conversion
S2
S1
Absorption(10
-15
sec)
Absorption(10
-15
sec)
An Introduction to Fluorescence Spectroscopy
8
Excitation can result in the molecule reaching any of the vibrational sub-levels
associated with each electronic state. Since the energy is absorbed as discrete
quanta, this should result in a series of distinct absorption bands. However, the
simple diagram above neglects the rotational levels associated with each
vibrational level and which normally increase the number of possible absorption
bands to such an extent that it becomes impossible to resolve individual
transitions.
Therefore, most compounds have broad absorption spectra except for those
where rotational levels are restricted (for example, planar, aromatic compounds).
Having absorbed energy and reached one of the higher vibrational levels of an
excited state, the molecule rapidly loses its excess of vibrational energy by
collision and falls to the lowest vibrational level of the excited state. In
addition, almost all molecules occupying an electronic state higher than the
second undergo internal conversion and pass from the lowest vibrational level of
the upper state to a higher vibrational level of a lower excited state which has the
same energy. From there the molecules again lose energy until the lowest
vibrational level of the first excited state is reached.
From this level, the molecule can return to any of the vibrational levels of the
ground state, emitting its energy in the form of fluorescence. If this process takes
place for all the molecules that absorbed light, then the quantum efficiency of
the solution will be a maximum, unity. If, however, any other route is followed,
the quantum efficiency will be less than one and may even be almost zero.
One transition, that from the lowest vibrational level in the ground electronic
state to the lowest vibrational level in the first excited state, the 0 - 0 transition,
is common to both the absorption and emission phenomena, whereas all other
absorption transitions require more energy than any transition in the
fluorescence emission. We can therefore expect the emission spectrum to
overlap the absorption spectrum at the wavelength corresponding to the 0 - 0
transition and the rest of the emission spectrum to be of lower energy, or longer
wavelength (Figure 2).
An Introduction to Fluorescence Spectroscopy
9
Figure 2 Idealised absorption and emission spectra
In practice, the 0-0 transitions in the absorption and emission spectra rarely
coincide exactly, the difference representing a small loss of energy by
interaction of the absorbing molecule with surrounding solvent molecules.
The absorption of energy to produce the first excited state does not perturb the
shape of the molecule greatly and this means that the distribution of vibrational
levels is very similar in both the ground and first excited states. The energy
differences between the bands in the emission spectrum will be similar to those
in the absorption spectrum and frequently the emission spectrum will be
approximate to a mirror image of the absorption spectrum. Since the emission of
fluorescence always takes place from the lowest vibrational level of the first
excited state, the shape of the emission spectrum is always the same, despite
changing the wavelength of exciting light.
A plot of emission against wavelength for any given excitation wavelength is
known as the emission spectrum. If the wavelength of the exciting light is
changed and the emission from the sample plotted against the wavelength of
exciting light, the result is known as the excitation spectrum. Furthermore, if the
intensity of exciting light is kept constant as its wavelength is changed, the plot
of emission against exciting wavelength is known as the corrected excitation
spectrum.
0-0 transition
Absorption Transmission
Intensity
λ
An Introduction to Fluorescence Spectroscopy
10
The quantum efficiency of most complex molecules is independent of the
wavelength of exciting light and the emission will be directly related to the
molecular extinction coefficient of the compound; in other words, the corrected
excitation spectrum of a substance will be the same as its absorption spectrum.
An Introduction to Fluorescence Spectroscopy
11
Phosphorescence
In the production of excited states by promotion of an electron into a higher
orbital, the direction of the spin of the electron is preserved. Since most
molecules have an even number of electrons and these are normally arranged in
pairs of opposite spin, the promotion of an electron does not disturb this parity.
However, it is possible for the spin of the promoted electron to be reversed so
that it is no longer paired and the molecule has two independent electrons of the
same spin in different orbitals. Quantum theory predicts that such a molecule can
exist in three forms of very slightly differing, but normally indistinguishable
energy, and the molecule is said to exist in a triplet state. The indirect process of
conversion from the excited state produced by absorption of energy, the singlet
state, to a triplet state, is known as intersystem crossing (Figure 3) and can occur
in many substances when the lowest vibrational level of the excited singlet state,
S1, has the same energy level as an upper vibrational level of the triplet state.
Direct transition from the ground state, usually a singlet state, for a molecule
with an even number of electrons, to an excited triplet state is theoretically
forbidden, which means that the reverse transition from triplet to ground state
will be difficult. Thus, while the transition from an excited singlet state, for
example, S1, to the ground state with the emission of fluorescence can take place
easily and within 10-9
- 10-6
seconds, the transition from an excited triplet state to
the ground state with the emission of phosphorescence requires at least
10-4
seconds and may take as long as 102
seconds. This delay was once used as
the characterisation of phosphorescence, but a more precise definition requires
that phosphorescence be derived from transitions directly from the triplet state to
the ground state.
An Introduction to Fluorescence Spectroscopy
12
Figure 3 Transition from the excited singlet state (S1) to the triplet state
(intersystem crossing)
The triplet state of a molecule has a lower energy than its associated singlet state
so that transitions back to the ground state are accompanied with the emission of
light of lower energy than from the singlet state. Therefore, we would typically
expect phosphorescence to occur at longer wavelengths than fluorescence
(Figure 4).
S0
Triplet system
Singlet system
Intersystem crossing
(10-12
– 10-4
sec)
Phosphorescence
(10-4
– 102
sec)
Absorption
Absorption
S2
S1
T2
T1
Fluorescence
An Introduction to Fluorescence Spectroscopy
13
Figure 4 Relative position of absorption, fluorescence and phosphorescence
bands of chrysene
Since molecules occupy triplet states for relatively long periods, they can easily
lose their energy by means other than phosphorescence. They are very
susceptible to collisions with solvent molecules. Solute molecules with unpaired
electrons, such as oxygen, are very effective at destroying triplet states even at
low concentrations. Therefore, phosphorescence in solution at room temperature
is a rare phenomenon, but by dispersing the sample in a rigid matrix or freezing
the solution to a low temperature, it can be observed from many compounds.
Chrysene
Wavelength (nm)
RelativeIntensity
Absorption
Fluorescence
Phosphorescence
An Introduction to Fluorescence Spectroscopy
14
Light scattering
Rayleigh-Tyndall scattering
Incident radiation is not only absorbed or transmitted by the sample but is also
scattered in all directions. The scattering takes place either from the molecules
themselves (Rayleigh scattering) or from small particles in colloidal suspension
(Tyndall scattering). Both types of scattering mean that incident radiation is
collected together with fluorescence and, unless adequate precautions are taken
to separate the fluorescence, inaccurate results will be obtained.
Rayleigh-Tyndall scattering will limit the sensitivity of an assay where
fluorescence takes place at a wavelength close to the excitation wavelength,
since total separation is impossible.
Raman scattering
During the Rayleigh scattering process, some of the incident energy can be
abstracted and converted into vibrational and rotational energy. The resulting
energy scattered is therefore of lower energy and longer wavelength than the
incident radiation. The result is a weak emission which may interfere or be
confused with the fluorescence of the sample.
Since the amount of energy abstracted is always constant, Raman bands appear
separated from the incident radiation by the same frequency difference,
irrespective of the wavelength of the exciting light, as shown in the following
table. If, therefore, the Raman band of the solvent coincides with the
fluorescence emission of the solute, separation can be achieved by changing the
excitation wavelength to a lower value; hence the Raman band will also be
lowered. Since the wavelength of fluorescence emission is independent of
exciting wavelength, the fluorescence will be separated from the Raman scatter.
Raman scatter is only important at high sensitivity since the phenomenon is
weak but must always be borne in mind. The Raman emission of some solvents
is quite complex and an emission spectrum of the solvent blank should always
be run prior to the analysis.
An Introduction to Fluorescence Spectroscopy
15
Solvent Position of Raman bands when excited at selected
wavelength
Excitation
wavelength
313 nm
Excitation
wavelength
366 nm
Excitation
wavelength
405 nm
Excitation
wavelength
436 nm
Water 350 418 469 511
Acetonitrile 340 406 457 504
Cyclohexane 344 409 458 499
Chloroform 346 411 461 502
All solvents containing hydrogen atoms linked to either carbon or oxygen show
a Raman band shifted approximately 3000 cm-1
from the exciting radiation.
An Introduction to Fluorescence Spectroscopy
16
Quantitation
The fraction of a parallel beam of light absorbed by a sample is independent of
the intensity of the incident beam and is related to the concentration of the
absorbing species by the familiar Beer-Lambert Law:Io
= e –Ecl
(1)
I
Normally written as:
log10 Io = Ecl
I
where
I = intensity of transmitted light
Io = intensity of incident light
E = molecular extinction coefficient
c = concentration in gm moles/L-1
l = pathlength of sample
and the quantity log10 Io is known as the absorbance or optical density of the
sample. I
Absorption spectra are simply a plot of absorbance against wavelength for any
sample.
The intensity of fluorescence emission of weakly absorbing, dilute samples is
related to the molecular extinction coefficient by the relation:
FI = Io (2.303 E cl) Φf (2)
An Introduction to Fluorescence Spectroscopy
17
and so the intensity of fluorescence emission is directly proportional to the
intensity of the incident radiation. These two equations explain the difference in
sensitivity between absorption and fluorimetric procedures. Sensitivity in
absorption measurements is limited by the ability of the instrument to
discriminate between the two nearly equal signals due to I and Io, and the
detection limit for even the most favorable cases rarely exceeds 10-8
moles.
Conversely, fluorimetric instruments are limited only by the intensity of exciting
light and the ability to detect low light levels, so under ideal conditions
concentrations of 10-12
moles can be measured.
Equation (2) only holds for dilute samples and it is possible to calculate the
concentration level (Cmax) above which non-linearity is observed from:
Cmax = 0.05 (3)
El
It is most important to establish the concentration at which a plot of fluorescence
emission against concentration becomes non-linear either by calculation, (3), or
by measurement of a suitable range of standards. Direct interpolation of
concentrations outside this limit will give inaccurate results.
An Introduction to Fluorescence Spectroscopy
18
Sensitivity
It is relatively easy to define the sensitivity of a UV-VIS spectrometer since
there is an absolute unit of measurement called an Absorbance value. In
fluorescence and phosphorescence spectroscopy there is no such fixed unit or
standard, hence there are a wide variety of ways with which sensitivity has been
described. Since the technique is primarily used for its inherent sensitivity and
since to a great extent the sensitivity of a fluorescence assay is dependent on the
instrument in use, some means of comparing instrument performance is clearly
needed.
Instrumental sensitivity
Instrumental sensitivity can be expressed as (a) defining the signal to noise ratio
of the instrument at a particular set of conditions or (b) the minimum detectable
quantity of a particular compound at a particular set of conditions.
Method sensitivity
Method sensitivity or limit of detection is defined in trace analysis as that
concentration which yields an instrument response, two standard deviations
above the background signal.
Absolute sensitivity
Absolute sensitivity expressed in terms of the spectral properties of the material,
for example,, quantum yield, extinction coefficient, and is better expressed in a
form that is independent of the instrument used to make the measurement.
For many years, instrument sensitivity was defined as that concentration of a
standard substance which is required to produce a signal equal to the noise at a
specified time constant. This definition is not only dependent upon the
luminescence properties of the compound and performance of the instrument but
also from the fluorescence from impurities in the solvent and sample,
fluorescence from the cuvettes and Raman scattering of the solvent.
An Introduction to Fluorescence Spectroscopy
19
Although the highest purity cuvettes can be used and the effect of the Raman
band minimized by suitable choice of solvent and/or wavelength, the limit of
detection may be totally different if different quality solute and solvent has been
used. In other words, the limit of detection is likely to be blank limited. The
usual sample often quoted is a solution of quinine sulfate in 0.1 N H2SO4. Many
manufacturers specify sensitivity in the parts per trillion range; however, it has
been shown that at those levels, non-linearity of the plot of intensity against
concentration occurs.
Because of these difficulties, an alternative approach to citing instrument
sensitivity has recently become more popular - that of measurement of the signal
to noise ratio of Raman bands of solvents such as water, alcohols or
hydrocarbons. Parker was the first to suggest that the Raman signal from a
suitable solvent provided a useful test of instrumental sensitivity. The
advantages of the test are; (a) independence of the chemistry since it measures
the properties of the solvent; (b) water, which is usually used, can be obtained
virtually free of fluorescent impurities, the principal Raman band occurring at
3380cm-1
; (c) ease of application and sensitivity; (d) versatility since it can be
applied at any wavelength between 200 and 500 nm.
The procedure involves determining the signal to noise ratio at particular slit
width (usually 5 or 10 nm), response time, with the excitation and emission
monochromators set to appropriate wavelengths to detect the Raman band. The
wavelength commands used are 350 nm excitation and 397 nm emission. The
following table lists some of the Raman peak maxima for water at particular
excitation wavelengths.
Excitation (nm) Raman emission (nm)
200 212
250 272
300 337
350 397
400 463
450 530
500 602
An Introduction to Fluorescence Spectroscopy
20
In spite of the intrinsic advantage of the Raman test, both it and the level of
detection method suffer from the disadvantage that they are measured at one set
of wavelengths. In practice, instruments from manufacturers differ in their
relative response at different wavelengths. In addition, the optical performance
tends to degrade with time, particularly below 300 nm and by different amounts
in different instruments. Having commented on these differences however, there
is no ideal method which can be used to express sensitivity to compare different
instruments. The most suitable is the Raman band test providing the information
containing the two constant spectral bandpass and scan rate.
An Introduction to Fluorescence Spectroscopy
21
Instrumentation
General
All fluorescence instruments contain three basic items: a source of light, a
sample holder and a detector. In addition, to be of analytical use, the wavelength
of incident radiation needs to be selectable and the detector signal capable of
precise manipulation and presentation (Figure 5). In simple filter fluorimeters,
the wavelengths of excited and emitted light are selected by filters which allow
measurements to be made at any pair of fixed wavelengths.
Simple fluorescence spectrometers have a means of analysing the spectral
distribution of the light emitted from the sample, the fluorescence emission
spectrum, which may be by means of either a continuously variable interference
filter or a monochromator. In more sophisticated instruments, monochromators
are provided for both the selection of exciting light and the analysis of sample
emission. Such instruments are also capable of measuring the variation of
emission intensity with exciting wavelength, the fluorescence excitation
spectrum.
In principle, the greatest sensitivity can be achieved by the use of filters, which
allow the total range of wavelengths emitted by the sample to be collected,
together with the highest intensity source possible. In practice, to realize the full
potential of the technique, only a small band of emitted wavelengths is examined
and the incident light intensity is not made excessive, to minimize the possible
photodecomposition of the sample.
An Introduction to Fluorescence Spectroscopy
22
Figure 5 Essential components of a fluorescence spectrometer
Light sources
Commonly employed sources in fluorescence spectrometry have spectral outputs
either as a continuum of energy over a wide range or as a series of discrete lines.
An example of the first type is the tungsten-halogen lamp and of the latter, a
mercury lamp. Mercury lamps are the most commonly employed line sources
and have the property that their spectral output depends upon the pressure of the
filler gas. The output from a low-pressure mercury lamp is concentrated in the
UV range, whereas the most commonly employed lamps, of medium and high
pressure, have an output covering the whole UV-visible spectrum.
Sample
Excitation
monochromator or filter
Emission
monochromator or filter
Detector
Light
source
An Introduction to Fluorescence Spectroscopy
23
Although in many cases the output from a line source will be adequate, it is rare
that an available line will exactly coincide with the optimum excitation
wavelength of the sample. It is therefore advantageous to employ a source whose
output is a continuum and the most commonly employed type is the xenon arc.
Xenon arc sources can be operated either on a continuous DC basis or
stroboscopically; the latter method offers advantages in the size and cost of
lamps. The output is essentially a continuum on which are superimposed a
number of sharp lines, allowing any wavelength throughout the UV-visible
region of the spectrum to be selected (Figure 6).
Figure 6 Emission spectrum of xenon lamp used in the PerkinElmer LS
Series
Arc lamps are inherently more unstable than discharge sources and for long term
stability a method of compensating for drift is advisable. The most satisfactory
method of doing this is to split the excitation energy so that a small portion is led
to a reference detector. The signal from this reference detector is ratioed with the
signal from the detector observing the sample. Such a ratio-recording system is
therefore independent of changes in the source intensity.
RelativeIntensity
nm
An Introduction to Fluorescence Spectroscopy
24
All sources of UV radiation will produce ozone from atmospheric oxygen,
which should be dispersed, since it is not only toxic, but also absorbs strongly in
the region below 300 nm. For this reason, most lamps will be operated in a
current of air and, if the supply fan fails, the lamp should be extinguished
immediately. Lamps must be handled with great care since fingermarks will
seriously decrease the UV output.
Wavelength selection
The simplest filter fluorimeters use fixed filters to isolate both the excited and
emitted wavelengths. To isolate one particular wavelength from a source
emitting a line spectrum, a pair of cut-off filters are all that is required. These
may be either glass filters or solutions in cuvettes. The emission filter must be
chosen so that the Rayleigh-Tyndall scattered light is obscured and the light
emitted by the sample transmitted. To avoid high blanks it may also be
necessary to filter out any Raman scatter.
Recently, interference filters having high transmission (≈ 40%) of a narrow
range (10 – 15 nm) of wavelengths have become available and it is possible to
purchase filters with a maximum transmission at any desired wavelength. UV
filters of this type, however, are expensive and of limited range. A simple filter
system is acceptable for much quantitative work, particularly where sufficient
chemistry has been carried out to eliminate interfering compounds.
However, it is useful to be able to scan the emission from the sample to check
for impurities and optimize conditions. A convenient method is to make use of a
continuous interference filter so that an emission spectrum can be recorded, at
least over the visible region of the spectrum.
A further refinement would be to use monochromators to select both the
excitation and emission wavelengths. Most modern instruments of this type
employ diffraction grating monochromators for this purpose. Such a
fluorescence spectrometer is capable of recording both excitation and emission
spectra and therefore makes full use of the analytical potential of the technique.
An Introduction to Fluorescence Spectroscopy
25
If monochromators are employed, it should be possible to change the slit width
of both the excitation and emission monochromators independently. Many
analyses will not require high resolution (essentially corresponding to high
selectivity) and greater sensitivity will be obtained with wide slit widths.
Conversely, to record the fine structure in the emission of, for example,,
polyaromatic hydrocarbons or to excite selectively one compound in the
presence of another, narrow slit widths will be necessary, and sensitivity will be
sacrificed.
Detectors
All commercial fluorescence instruments use photomultiplier tubes as detectors
and a wide variety of types are available. The material from which the
photocathode is made determines the spectral range of the photomultiplier and
generally two tubes are required to cover the complete UV-visible range. The S5
type can be used to detect fluorescence out to approximately 650 nm, but if it is
necessary to measure emission at longer wavelengths, a special red sensitive,
S20, photomultiplier should be employed.
The limit of sensitivity of a photomultiplier is normally governed by the level of
dark current (which is the signal derived from the tube with no light falling on
it). The dark current is caused by thermal activation and can usually be reduced
by cooling the photomultiplier. Another method of minimizing dark current is to
use a stroboscopic source since the ratio of dark current to fluorescence will be
very small during each high intensity flash. During the periods between flashes
when the dark current is relatively high, the photomultiplier output can be
disconnected. The overall result is that the dark current no longer becomes the
limitation to sensitivity.
The spectral response of all photomultipliers varies with wavelength, but it is
sometimes necessary to determine the actual quantum intensity of the incident
radiation and a detector insensitive to changes in wavelength is required. A
suitable quantum counter can be made from a concentrated solution of
Rhodamine 101 in ethylene glycol which has the property of emitting the same
number of quanta of light as it absorbs, but over a very wide wavelength range.
Thus, by measuring the output of the quantum counter at one wavelength, the
number of incident quanta over a wide wavelength range can be measured.
An Introduction to Fluorescence Spectroscopy
26
Read-out devices
The output from the detector is amplified and displayed on a readout device
which may be a meter or digital display. It should be possible to change the
sensitivity of the amplifier in a series of discrete steps so that samples of widely
differing concentration can be compared. A continuous sensitivity adjustment is
also useful so that the display can be made to read directly in concentration
units.
Digital displays are most legible and free from misinterpretation. Improvement
in precision is obtained by the use of integration techniques where the average
value over a period of a few seconds is displayed as an unchanging signal.
Microprocessor electronics provide outputs directly compatible with printer
systems and computers, eliminating any possibility of operator error in
transferring data.
Sample holders
The majority of fluorescence assays are carried out in solution, the final
measurement being made upon the sample contained in a cuvette or in a
flowcell. Cuvettes may be circular, square or rectangular (the latter being
uncommon), and must be constructed of a material that will transmit both the
incident and emitted light. Square cuvettes, or cells will be found to be most
precise since the parameters of pathlength and parallelism are easier to maintain
during manufacture. However, round cuvettes are suitable for many more
routine applications and have the advantage of being less expensive.
The cuvette is placed normal to the incident beam. The resulting fluorescence is
given off equally in all directions, and may be collected from either the front
surface of the cell, at right angles to the incident beam, or in-line with the
incident beam.
Some instruments will provide the option of choosing which collecting method
should be employed, a choice based upon the characteristics of the sample. A
very dilute solution will produce fluorescence equally from any point along the
path taken by the incident beam through the sample. Under these conditions, the
right-angled collection method should be used since it has the benefit of
minimizing the effect of light scattering by the solution and cell. This is the
usual measuring condition in analytical procedures.
An Introduction to Fluorescence Spectroscopy
27
Although fluorescence takes place from every point along the light path, only a
small fraction of this emission is actually collected by the instrument and
transmitted to the detector. The result is that much of the solution does not
contribute to the fluorescence emission and the same intensity will be observed
from a much smaller volume of solution contained in a microcell whose
dimensions more closely match the optical considerations of the instrument
(Figure 7).
Figure 7 Fluorescence emission from a microcell whose dimensions closely
match the optical considerations of the instrument
As the absorbance of the solution increases, the fluorescence emission becomes
progressively distorted until a point is reached where little actually penetrates the
main bulk of the solution and the fluorescence will be confined to the front
surface of the cuvette. Front surface collection will still allow measurements to
be made, although the contribution due to light scattered from the cuvette wall
will be large. Front surface collection will at least always show emission
(possibly distorted) from a fluorescent sample, whereas the fluorescence
obtained from 90° collection falls rapidly as the absorbance of the solution
increases.
It is possible to dismiss a potential sample as non-fluorescent simply because the
concentration is too high. Wherever possible, the absorbance of a completely
unknown solution should be measured before attempting a fluorescence check
and the concentration adjusted to provide a solution of absorbance <0.1 A.
Fluorescence
10 mm cell
Microcell
An Introduction to Fluorescence Spectroscopy
28
Problems of high blank values
Cuvettes
One of the advantages of fluorescence procedures compared to equivalent
absorption techniques is that routine measurements may usually be carried out in
inexpensive test-tubes rather than precision cuvettes without appreciable loss in
precision. This benefit is derived from the geometrical layout of simple
fluorimeters where only a small central area of the cuvette is actually viewed by
the detector, so that the overall dimensions of the cell are less important.
However, this statement needs careful qualification, since the use of laboratory
grade test-tubes will result in deviation arising from other sources. Variations in
cell wall width can give rise to errors by distorting the lens action of the round
cell wall from one cell to another and the native fluorescence of the material
used to manufacture the cell will often produce large blank values. Obviously,
the cell must be fabricated from a material which will transmit both the
excitation and emission wavelengths of interest and, if these are both in the
visible region, savings can be made by the use of glass or even plastic sample
containers. Only those cells recommended by the manufacturer of the instrument
should be employed and every new batch should be carefully checked to make
quite sure that their native fluorescence at the analytical wavelength chosen is
minimal. Even the highest grades of silica and glass have some fluorescence,
although it should not contribute appreciably to routine measurements.
Cells and other glassware used for fluorimetric analysis should be carefully
cleaned, preferably by boiling in 50% nitric acid followed by thorough rinsing in
distilled water.
Solvents and Reagents
It is essential to examine the quality of solvents periodically since small traces of
contaminants may be enough to produce high black values. If the solvent blank
is appreciable it can usually be reduced to a reasonable value by distillation or
washing with acid, base or water.
• Water is the most common solvent and deionized-distilled water should
always be employed.
An Introduction to Fluorescence Spectroscopy
29
• Solvents should not be stored in plastic containers since leaching of organic
additives and plasticisers will produce high blank values.
• Fluorescence derived from contaminants should not be confused with the
Raman scatter from the solvent itself.
• All other reagents used in the assay should be carefully controlled and high
quality grades are to be recommended.
• Reagent blanks should always be carried through the analytical procedure
and the actual value of the blank determined before the instrument is
re-zeroed.
• High or changed blank values should immediately cast suspicion upon the
solvents and reagents employed.
Other contaminants
Soaking of glassware in synthetic detergent solutions is gaining ground in many
laboratories as a general method of cleaning. Most commercial preparations are
strongly fluorescent. Before use, the fluorescence characteristics of a dilute
solution of the detergent should be measured to make quite sure that the
fluorescence is minimal at the chosen analytical wavelengths. Inorganic
cleansing agents are more suitable for fluorimetric work.
Stopcock grease is another common contaminant with strong native
fluorescence. The growth of micro-organisms in buffer or reagent solutions will
affect blank values by both their fluorescence and light scattering properties.
Filter paper can be another source of contamination due to residues of phenols
present from the original wood.
An Introduction to Fluorescence Spectroscopy
30
Working with dilute solutions
General
It is widely recognized that dilute solutions are not as stable as concentrated
ones. For this reason, it is common practice to store concentrated stock solutions
and make dilutions to produce working standards. It would appear that only a
small constant fraction of the molecules present are susceptible and that, as
solutions are diluted, this constant fraction becomes progressively more
significant. Final solutions are always very dilute and should never be stored for
long periods. It is good practice to carry out all procedures as quickly as is
practicably possible.
Adsorption
Loss of organic substances by adsorption onto the walls of the container
becomes particularly troublesome at the sub-microgram level. New glass
surfaces should be thoroughly cleaned in acid before use and measurements of
aromatic substances in non-polar solvents should be avoided. Often the addition
of a small quantity of a polar solvent to a non-polar one will greatly reduce
adsorption losses.
Protein precipitation produces a large, freshly prepared, adsorptive surface and
choice of precipitation agent is critical to prevent adsorption of the fluorophore
with subsequent loss. Each assay may require a specific precipitation agent and
analytical procedures should not be changed without investigation.
An Introduction to Fluorescence Spectroscopy
31
Photo-decomposition
Fluorimeters employ intense light sources to produce high sensitivity and in
some cases the level of incident light may be sufficient to decompose the sample
under investigation. A shutter is essential so that the solution is only irradiated
during the short period of measurement. The decomposition only occurs at the
point where the excitation beam is focused onto the sample and if, after closing
and then reopening the excitation shutter, the value of fluorescence rises to the
original value, photo-decomposition is almost certainly taking place. This
phenomenon is due to the diffusion of unaffected, fresh molecules into the light
path whilst the shutter is closed, with subsequent renewal of fluorescence.
Photo-decomposition may be reduced by choice of another excitation
wavelength or by reducing the excitation intensity by use of narrower slit widths
or a neutral density filter.
Oxidation
The presence of trace oxidizing agents, for example, dissolved oxygen or traces
of peroxides, can reduce fluorescence intensity.
An Introduction to Fluorescence Spectroscopy
32
Factors affecting quantitative accuracy
Non-linearity
The proportional relationship between light absorption and fluorescence
emission is only valid for cases where the absorption is small. As the
concentration of fluorophore increases, deviations occur and the plot of emission
against concentration becomes non-linear. With right-angle viewing, the
principal distortion arises from the absorption of the excited light before it can
penetrate to the heart of the cell, where the emission produced is accepted by the
detector optics.
In cases where it is necessary to work at high concentrations, for example,, in
some enzyme work using NADH, it is usually possible to increase the linear
concentration range by the use of microcuvettes, often without much loss in
overall sensitivity. Generally, deviations of this sort are negligible when dealing
with solutions of transmission greater than 95 %T at the wavelength of interest,
which can be easily checked with an absorption instrument. Other sources of
non-linearity are usually associated with the instrument's signal handling (for
example, amplifiers) and display (especially meters) systems.
Temperature effects
Changes in temperature affect the viscosity of the medium and hence the number
of collisions of the molecules of the fluorophore with solvent molecules.
Fluorescence intensity is sensitive to such changes and the fluorescence of many
certain fluorophores shows a temperature dependence. In such cases the use of
thermostatted cell holders is to be recommended. Normally it is sufficient to
work at room temperature with the proviso that any sample procedure involving
heating or cooling must also allow sufficient time for the final solution to reach
ambient before measurement.
An Introduction to Fluorescence Spectroscopy
33
pH effects
Relatively small changes in pH will sometimes radically affect the intensity and
spectral characteristics of fluorescence. Accurate pH control is essential and,
when particular buffer solutions are recommended in an assay procedure, they
should not be changed without investigation. Most phenols are fluorescent in
neutral or acidic media, but the presence of a base leads to the formation of
nonfluorescent phenate ions. 5 hydoxyindoles, for example, serotonin, show a
shift in fluorescence emission maximum from 330 nm at neutral pH to 550 nm
in strong acid without any change in the absorption spectrum.
Inner-filter effects
Fluorescence intensity will be reduced by the presence of any compound which
is capable of absorbing a portion of either the excitation or emission energy. At
high concentrations this can be caused by absorption due to the fluorophore
itself. More commonly, particularly when working with tissue or urine extracts,
it is the presence of relatively large quantities of other absorbing species that is
troublesome. The purpose of extraction procedures is usually to eliminate such
species so that the final measurement is made upon a solution essentially similar
to the standard.
Quenching
Decrease of fluorescence intensity by interaction of the excited state of the
fluorophore with its surroundings is known as quenching and is fortunately
relatively rare. Quenching is not random. Each example is indicative of a
specific chemical interaction, and the common instances are well known.
Quinine fluorescence is quenched by the presence of halide ion despite the fact
that the absorption spectrum and extinction coefficient of quinine is identical in
0.5M H2SO4 and 0.5M HCl.
An Introduction to Fluorescence Spectroscopy
34
References
1. The luminescence of biological systems, Johnson, F.H., Amer. Assoc.
Adv. Sci., Washington, D.C. (1955).
2. Fluorescence and phosphorescence analysts, Hercules, D.M., Editor
Wiley-Interscience Publishers, New York, London, Sydney (1965).
3. Fluoromestric analysis, Konstantinova-Schlezinger, M.A., Editor, Davis
Publishing Co., New York (1965).
4. Handbook of fluorescence spectra in aromatic molecules, Berlman, LB.,
Academic Press, New York, Vol. 1 (1965), Vol. II (1971).
5. Fluorometric techniques in clinical pathology, Phillips, R., et al,
Progress in Clinical Pathology, Grune and Stratton, New York (1965).
6. Biochemical fluorescence: concepts, vol. I, Chen, R.F., et al, 408
(1965).
7. Fluorescence, theory, instrumentation and practice, Edited by
Guilbault, G.G., published by Marcel Dekker, Inc., New York (1967).
8. Guide to fluorescence literature, Passwater, R.A., Plenum Press, New
York, Vol. I (1967), Vol. II (1969).
9. Fluorescence and phosphorescence of proteins and nucleic acids,
Koney, S.U., Plenum Press, New York (1967).
10. Luminescence of organic substances, Schmillen, A., et al, Hellwege
Verlag, Berlin (1967).
11. Photoluminescence of solutions, Parker, C.A., Elsevier Publishing Co.,
New York (1968).
12. Fluorescence assay in biology and medicine, Udenfriend, S., Academic
Press, Vol. I (1962), New York Vol. II (1969).
13. Theory and interpretation of fluorescence and phosphorescence, Becker
R., Wiley-Interscience (1969).
An Introduction to Fluorescence Spectroscopy
35
14. Instrumental methods of chemical analysis, 3rd Edition, Ewing, G.W.,
McGraw-Hill Book Co. (1969).
15. Fluorimetric assay of enzymes, Roth, M., Methods of Biochemical
Analyses Vol. 17, David Glick, Editor Wiley Interscience, New York
(1969).
16. Luminescence and biophysics of cells, Baremboim, G.M., et al, Plenum
Press, New York (1969).
17. Fluorescence analysis: a practical approach, While; C.E., et al, Marcel
Dekker (1970).
18. Fluorescence spectroscopy: an introduction for biology and medicine,
Pesce, A.J., et al, Marcel Dekker (1971).
19. Fluorometric analysis, Weissler, A., et al, Standard methods of
chemical analysis- Welcher, Editor D.Van Nostrand Co., Princeton, N.J.
(1970).
20. Biochemical fluorescence: concepts, vol. II, Chen, R.F., et al, M.
Dekker, New York (1970) 535 pp.
21. Luminescence spectrometry in analytical chemistry, Winefordner, J.D.,
et al. Wiley-Interscience Publishers, New York (1972).
22. Methods In pharmacology, vol. II, physical techniques, Chignell, C.F.,
Appleton-Crofts (1972).
23. Practical fluorescence: theory, methods and techniques, Guilbault,
G.G., Marcel Dekker (1973).
24. Fluorometric techniques in clinical chemistry, Elevitch, F.R., Little
Brown & Co. (1973).
25. Modern fluorescence spectroscopy, Wehry E.L., Ed; Plenum Press, New
York (1976), Vol. I 238pp.
26. Excited states of biological molecules, Birks, J.B., Ed; John Wiley,
London, (1976) 652 pp.
An Introduction to Fluorescence Spectroscopy
36
27. Modern fluorescence spectroscopy, Wehry, E.L., Ed. Plenum Press,
New York (1976), Vol. 11459 pp.
28. Physical Biochemistry-applications to biochemistry and molecular
biology, Freifelder, D., W. H. Freeman & Co. (1976).
29. Fluorescence and phosphorescence spectroscopy: physiochemical
principles and practice, Schulman, S.G., Pergamon, Elmsford, New
York (1977).
30. Molecular fluorescence spectroscopy, Guilbault, G.G., Cooper Anal.
Chem. 8, 71-205, (1977), Svehlag. Ed; Elsevier.
31. Developments in instrumentation and techniques, West, M.A., from
Photochemistry Vol. Chem. Soc. London.
32. The handbook of derivatives for chromatography, Blau, K., King, G.,
Heyden & Son Ltd. (1977).
33. Standardization & fluorescence spectrometry, Miller, G.N., Ed,
Chapman and Hall, 1981.
34. Fluorescence detection in liquid chromatography, Rhys-Williams, A.T.,
PerkinElmer Ltd.
35. Fluoroimmunoassay-a review of present methods, Rhys-Williams, A.T.,
PerkinElmer Ltd.
36. An introduction to phosphorescence spectroscopy, Rhys-Williams, A.T.,
PerkinElmer Ltd.
37. Nomenclature, symbols, units and their usage in spectrochemical
analysis - VI molecular luminescence spectroscopy, Melhuish, W.H.,
Zander, M., Pure App. Chem. 53, 1953, 1981.