FLUORESCENCE / Overview 97
See also: Clinical Analysis: Glucose. Enzymes: Immo- Osborne BG and Tyson JF (1988) Review: flow injection
bilized Enzymes. Fluorescence: Instrumentation. Food
analysis  a new technique for food and beverage anal-
and Nutritional Analysis: Overview; Soft Drinks; Alco- ysis. International Journal of Food Science and Tech-
holic Beverages. Process Analysis: Overview; Bioproc- nology 12: 541 554.
ess Analysis.
Ro
~
i%0Å„ka J and Hansen EH (1988) Flow Injection Analysis,
2nd edn. New York: Wiley.
Ro
~
i%0Å„ka J and Hansen EH (1998) Flow injection analysis 
where are we heading?. Trends in Analytical Chemistry
Further Reading
17: 69 73.
Fang Z (1993) Flow Injection Separation and Preconcen- Schügerl K (2001) Progress in monitoring, modeling and
tration. Weinheim: VCH. control of bioprocesses during the last 20 years. Journal
Hansen EH (1990) Flow injection analysis. A versatile tool of Biotechnology 85: 149 173.
for biotechnological applications. Wissenschaftliche Sosnitza P, Irtel F, Ulber R, et al. (1998) Flow injection
Zeitschrift TH Leuna-Merseburg 32: 373 387. analysis system for the supervision of industrial chro-
Karlberg B and Pacey GE (1989) Flow Injection Analysis: matographic downstream processing in biotechnology.
A Practical Guide. Amsterdam: Elsevier. Biosensors & Biolectronics 13: 1251 1255.
Keay PJ and Wang W (1997) Applications of flow injection Trojanowicz M (2000) Flow Injection Analysis: Instru-
analysis to analytical biotechnology. Trends in Biotech- mentation and Applications. Singapore: World Scientific.
nology 15: 76 81. Valcárcel M and Luque de Castro MD (1994) Flow-
Lemieux L, Puchades R, and Simard RE (1989) Applica- Through (Bio)chemical Sensors. New York: Elsevier.
tion of FIA techniques to food analysis. Lebensmittel- Valcárcel M, Luque de Castro MD, and Losada A (1987)
Wissenschaft und Technologie 22: 254 263. Flow Injection Analysis. Principles and Applications.
Miró M, Estela JM, and Cerdą V (2003) Application of Chichester: Ellis Horwood.
flowing stream techniques to water analysis. Part I. Ionic Van der Linden WE (1986) Flow injection analysis in
species: dissolved inorganic carbon, nutrients and related online process control. Analytica Chimica Acta 179:
compounds. Review. Talanta 60: 867 886. 91 101.
FLUORESCENCE
Contents
Overview
Instrumentation
Multidimensional Fluorescence Spectrometry
High-Resolution Techniques
Time-Resolved Fluorescence
Derivatization
Fluorescence Labeling
Quantitative Analysis
Clinical and Drug Applications
Environmental Applications
Food Applications
Introduction
Overview
Molecular photoluminescence processes can be clas-
sified according to the mode in which a molecule is
MEDíaz-García and R Badía-LaíÅ„o, University of
promoted to an electronically excited state and to the
Oviedo, Oviedo, Spain
type of molecular excited state (see Table 1). As the
excited molecule decays back to the ground state, or
& 2005, Elsevier Ltd. All Rights Reserved. to a lower-lying excited electronic state, light is
98 FLUORESCENCE / Overview
emitted at a characteristic wavelength. All the proc- nonradiative decay processes, such as fluorescence
esses that involve the emission of electromagnetic energy transfer, collisional quenching, or intersystem
radiation are called luminescence. crossing.
Following absorption, a number of vibrational Fluorescence spectrometry has become established
levels of the excited state are populated. Molecules in as a routine technique in many specialized applica-
these higher vibrational levels then decay to the low- tions due to its high sensitivity. Examples of the cur-
est vibrational level of the excited state (vibrational rent applications of fluorescence range from simple
relaxation). Fluorescence is the radiation released fluorimetric analysis for biomolecules, metal ions,
when a molecule, which has been promoted to an and organic compounds to identification of specific
excited singlet state by light absorption, rapidly DNA and/or RNA sequences in tissues. In addition,
relaxes from the lowest vibrational mode of the much of the current research in biochemistry, med-
electronically excited state S1 (Kasha s rule) to a icine, and molecular biology involves fluorescence
vibrational mode of the electronic ground state S0. spectroscopy of either intrinsic molecular fluores-
The energy transitions associated with photolumi- cence (from tyrosine or tryptophan residues) or
nescence can be represented on Morse energy surface exogenous fluorescent probes.
diagrams (Figure 1). In addition to fluorescence, the Besides high sensitivity, fluorescence exhibits un-
energy of the excited state may be dissipated by ique performance characteristics among which se-
lectivity, reproducibility, and temporal resolution are
included. It is relatively a simple and rapid analytical
technique, specially adequate for quantification of
Table 1 Molecular photoluminescence processes
aromatic, or highly unsaturated, organic molecules
Luminescence process
present at trace concentrations, especially in biolo-
(A) Excitation mode
gical and environmental samples. The technique can
Absorption of radiation (UV Vis) Photoluminescence
also be applied to a wide variety of organic and
Chemical reaction Chemiluminescence,
inorganic compounds via chemical labeling and
bioluminescence
derivatization procedures and can be used as the
Thermally activated ion Thermoluminescence
detection mode in flow injection analysis, chro-
recombination
Friction Triboluminescence
matography, capillary electrophoresis, and thin-
Sound waves Sonoluminescence
layer chromatography. Fluorescence output is linear
High-energy particles, radiation Radioluminescence
to sample concentration over a very broad range and
Injection of charge Electroluminescence
the technique can be used over three to six decades of
concentration without sample dilution or modifica-
(B) Excited state
First excited singlet state Fluorescence, delayed
tion of the sample cell. On the other hand, almost
fluorescence
any liquid, gaseous, or solid sample can be analyzed
Lowest triplet state Phosphorescence
by fluorescence. Sample size can be extremely small
UV Vis, Ultraviolet visible and nanoliter sample volumes may be analyzed using
Energy
Vibrational
sublevels
Excited triplet
state
Excited singlet
state
Excitation
(absorption of h 1)
Radiational deactivation
(emission of h 2)
Ground state
Internuclear distance
Figure 1 Simplified energy level diagram of a polyatomic molecule.
FLUORESCENCE / Overview 99
Sources
specialized techniques. Even, using sophisticated
techniques (e.g., flow and image cytometry), single
The source used in most commercial fluorimeters is
molecules can be studied via fluorescence, while
the 75 450 W xenon arc lamp, which covers the
multiple targets can be detected during a single assay
range of 200 1000 nm. However, for greater excita-
by using a combination of fluorophores.
tion energy at selected wavelengths, the mercury-arc
Fluorescent dyes and tailor-made fluorescence-
lamp may be used, although excitation spectra ob-
labeled compounds are growing commercially for
tained with this lamp are usually severely distorted
coupling not only to metal ions but also to biomol-
because principal lines (at 254, 302, 313, 546, 587,
ecules, as a result of the increasing demand for iden-
691, and 773 nm) are superimposed on a continuum.
tification of molecular interactions and for the
Lasers are now increasingly used due to the high op-
visualization, recognition, and quantification of mol-
tical power (number of photons per unit time). A few
ecules, which bind to target molecules like peptides,
fluorescence instruments using laser sources are com-
proteins, nucleic acids, carbohydrates, etc. Compa-
mercially available and most are intended for specific
nies have built the basic chemistries into user-friendly
applications such as analyses of uranium in the nu-
kits and systems that make labeling easy. Fluores-
clear industry and for remote measurements using
cence spectroscopy has evolved to a routinely
optical fibers. In some applications, such as frequen-
used ultra-sensitive detection technology on high-
cy domain lifetime measurements, it may be neces-
throughput screening, from a 96-well situation up to
sary to modulate the intensity of the source. In these
high-density 384- and 1536-well platforms. An
cases, some classes of laser or light-emitting diodes
integration of combinatorial chemistry and fluores-
are suitable sources.
cence spectroscopy creates a powerful tool to probe
biological targets. A consistent theme of fluorescent
Wavelength Selection
technologies is their adaptability to users needs in
Isolation of selected wavelengths is carried out with
both industry and academe.
filters, usually in conjunction with white light sourc-
es. Filters are used in inexpensive, portable fluores-
cence spectrometers and their optical quality will
Basic Instrumentation for Molecular
often determine the performance of the instrument.
Filters allow transmission of a very large number of
Fluorescence Measurement
photons from the source to the sample and from the
A block diagram of a basic fluorimeter is shown in
sample to the detector (fluorescence signal is maxi-
Figure 2. Aside from the optical components shown,
mized) but have poor wavelength selectivity. While
modern fluorimeters have dedicated computers,
filters can be used for limited applications, grating
which may control instrumental operating para-
monochromators are used in most spectrofluorime-
meters, the acquisition of spectral data and the post-
ters. For optically clear solutions, single-grating
processing of the data.
monochromators are adequate while double-grating
monochromators are ideal for turbid solutions as
stray light and scattering is reduced. The emission
wavelength selector system is generally placed at an
Excitation
monochromator,
angle of 901 with respect to the excitation axis to
M1
minimize interferences from transmitted and scat-
Sample
tered exciting light.
exc
cell
Sample Illumination
Xenon arc
lamp
The most common arrangement uses the 901 geome-
Emission
try depicted in Figure 2 rather than the 1801 geome-
monochromator,
M2 try common to absorption spectrophotometers. This
geometry is suitable for weakly absorbing solution or
em
diluted samples. For solids (samples adsorbed on
solid surfaces such as polymers, paper, etc.) and for
solutions that absorb strongly at the excitation
Photomultiplier
wavelength, a front surface geometry is preferable
Electronics
detector
and data
(see Figure 3). Fluorescence is viewed in these cases
processing
from the face of the sample on which the exciting
Figure 2 Basic component of a spectrofluorimeter. light impinges. In the front-face arrangement there is
100 FLUORESCENCE / Overview
Transparent
Information from Fluorescence
Solid sample
Diluted solution
solid sample
Measurements
All fluorescent molecules can be characterized by two
types of spectra  the excitation spectrum and the
emission spectrum. The excitation spectrum is a plot
of emitted fluorescence as a function of the excitation
wavelength at a fixed emission wavelength. The ex-
Flow-cell
1cm cuvette
citation spectrum represents the relative probability
Surface reader
that a fluorescent molecule will be excited by a given
wavelength of incident light. If the excitation energy
is constant, the fluorescence excitation spectrum is
very similar to the absorption spectrum. The photon
Optical
energy at the maximum of the excitation peak equals
path
the energy difference between the ground state (S0)
Figure 3 Sample geometry and typical cell for fluorescence and a favored vibrational level of the first excited
measurements.
state (S1) of the molecule. In some cases, the exci-
tation spectrum shows a second peak at a shorter
a risk that reflected light from the surface enters the wavelength (higher energy) that indicates transition
emission monochromator resulting in large amount of the molecule from the ground state to the second
of stray light. If the solid sample is transparent, back- excited state (S2) (Figure 4).
face illumination is also possible. A plot of the relative intensity of emitted light as a
In practice, most fluorescence measurements are function of the emission wavelength at a fixed exci-
taken in solution and 1 cm glass or fused silica cuvet- tation wavelength is termed fluorescence emission
tes with four polished windows are used. For some spectrum. The fluorescence emission is characterized
specialized applications different types of cuvettes by the transition from the lowest vibrational mode of
are commercially available. For example, flow cells
the electronically excited state (S1) to the ground
of different configurations for continuous flow spec- state. Therefore, the shape of the emission spectrum
trofluorimetric analysis have been designed. is always the same and is independent of the
wavelength of the exciting radiation.
The shape of the emission spectrum is approxi-
Detectors
mately a mirror image of the longest-wavelength
The final basic component of the fluorimeter is the excitation band. Peaks in a fluorescence excitation
detector, which is placed at the exit slit of the emis- spectrum usually correspond closely in wavelength to
sion-wavelength selection system. A key requirement absorption peaks. In theory, the maximum of the
for a detector is its ability to detect very low radi- emission peak should occur at the same wavelength
ation levels. Photomultiplier tubes (PMTs) are most than that of the excitation peak (commonly called
widely used because of their high sensitivity. In many the O O transition of absorption and emission). In
cases, PMTs are operated at low temperature to practice, the emission spectrum is always shifted to-
minimize dark counts and to push detection limits ward a longer wavelength relative to the excitation
lower. The essential principle of the PMT has re- spectrum, as shown in Figure 4. This difference be-
mained unchanged for many years and its main lim- tween the excitation and emission maxima is termed
itation is that it is a single-channel detector and some the Stokes shift. The Stokes shift represents the
important analytical information may be lost when energy lost whilst the molecule was in the excited
obtaining a spectrum (e.g., in chromatographic ap- state and, from a practical point of view, it allows the
plications). An array of detectors, as in many UV Vis excitation and emission peaks to be spectrally sep-
absorption spectrometers, would allow rapid acqui- arated and easily distinguished. When the spectra are
sition of full fluorescence spectra. Until now scanned,  scattering peaks usually appear when the
the PMT-based devices remain the most common wavelengths of excitation and emission monoch-
detectors although new classes of electronic array romators coincide. They can be eliminated if the
detectors are beginning to be employed in some excitation and emission band passes do not overlap.
fluorimeters. At present, one of the most promising The absorption spectrum is reproducible when
electronic array detectors for fluorimetry is the scanned on different instruments (provided there are
charge-coupled device (CCD), which essentially is no distortions due to inappropriate settings of
an array of semiconductor photodetectors. scan speed, band pass width, source output, etc.).
FLUORESCENCE / Overview 101
Absorption probability
or emission intensity
Excitation spectrum Emission spectrum
Scan M1, M2 set at 2 Scan M2, M1 set at 1
S0 S1 S1 S0
Stokes shift
S0 S2
exc em Wavelength (nm)
Figure 4 Typical features of fluorescence spectra.
However, fluorescence spectra (excitation and emis- presence of heteroatoms in such systems can either
sion) are less reproducible because fluorescence spec- decrease or increase fluorescence intensity, depending
tra are affected by intensity of light source and by the on the probability of intersystem crossing. Molecules
response of photomultiplier detector, both of which in which it is possible to stabilize charge by reso-
vary with the wavelength. Thus, excitation and fluo- nance are much more fluorescent than those with no
rescence spectra, obtained in the usual nonratio mode resonance structures. A typical example of this is
with single-beam instruments, are often distorted and aniline, which is highly fluorescent while anilinium
not reproducible from instrument to instrument. It is ions are not fluorescent. Finally, more rigid molecules
possible to eliminate these variations instrumentally exhibit stronger fluorescence owing to the lower
and several commercial instruments are available that probability of energy dissipation by nonradiative
provide  corrected spectra . These adjust for varia- processes such as energy transfer or by transitions
tions in the source intensity with wavelength and also between electronic states. So, while biphenyl is
correct for variations of the detector response. weakly fluorescent, fluorene is strongly fluorescent
The emission wavelength and the fluorescence in- (quantum yield of 1.0). The applicability of fluores-
tensity are determined by the structure of the mol- cence may be extended to nonfluorescent compounds
ecule. In principle, any molecule that absorbs by converting them to a fluorescent derivative. For
radiation of adequate energy could fluoresce. How- example, nonfluorescent steroids may be converted
ever, many molecules exhibit very weak fluorescence to fluorescent phenolic compounds by dehydration
and only a small fraction of molecules exhibit ana- with concentrated sulfuric acid. Antibodies may be
lytically useful fluorescence. Most intensely fluores- made to fluoresce by attaching to it a fluorescent tag
cent molecules contain highly conjugated p-electron or label (e.g., fluorescein isocyanate). Concerning
systems. For example, polycyclic compounds such as inorganic species, relatively few inorganic metal ions
aromatic hydrocarbons, vitamin K, barbiturates, (e.g., UO2 þ , Ce(III), Tl(I)) exhibit intense fluor-
2
nucleosides, and conjugated polyenes such as vita- escence. Nonfluorescent inorganic metal ions may be
min A are fluorescent. Typically, more the number of reacted with aromatic organic ligands to form fluor-
conjugated bonds in the molecule, longer is the escent chelates. Considerations about rigidity, con-
wavelength of emission observed. The presence of jugated bonds, and aromatic fused rings also apply in
heteroatoms such as O, S, and N, results in n-p these cases.
transitions that may promote a change in the spin of
the excited electron (intersystem crossing process)
and no fluorescence is observed (e.g., pyridine, pyr-
Molecular Fluorescence: Practical
role, furan, and thiophene are not fluorescent). Mol-
Considerations
ecules with fused aromatic rings have high molar
absorptivity values and, consequently, are highly flu- Conventional fluorimetric determinations are carried
orescent (e.g., naphthalene, pyrene, anthracene). The out in solution with external standardization. When
102 FLUORESCENCE / Overview
the native fluorescence of the analyte is measured, proportional to the fluorochrome concentration and
minimal sample treatment is necessary and a fluori- can be described by the following relationship:
metric analysis can be carried out in less than 10 min.
If ź 2:303 FfI0ebc
However, when fluorimetric derivatization reactions
are necessary, the analytical reaction is allowed to
where If is the observed fluorescence, Ff is the quan-
reach equilibrium before the fluorescence signal is
tum yield, I0 is the power of the incoming light, e is
measured. In any case, the fluorescence signal is re-
the molar extinction coefficient, c is the fluorophore
lated to the analyte concentration.
concentration, and b is the optical pathlength. If Ff,
Solution conditions, such as pH, viscosity, ionic
I0, e, and b remain constant, the relationship between
strength, solvents, and reagent concentrations must
the fluorescence intensity and the fluorophore con-
be carefully adjusted and controlled in order to max-
centration is linear. This linearization is valid only
imize the fluorescence signal. Background fluor-
for cases where ebco0.05 (diluted solutions). Above
escence from sample matrix or contaminants in
this value, nonlinearity occurs. At higher concentra-
solvents, reagents, and/or laboratory glassware can
tions significant absorption of the excitation beam
occur over the same wavelength range as the analyte
radiation may take place (primary absorption), thus
fluorescence (additive interference). This can be a
reducing the number of excited molecules across the
major limitation in achieving optimal detection lim-
light path. This effect causes negative deviation in a
its, especially in biological and environmental sam-
concentration response plot. Also, if the excitation
ples. Background fluorescence can be circumvented
and emission spectra overlap, significant absorption
through different ways, depending on its origin or
of the emission beam takes place at high analyte
nature. For example, the analyte can be separated
concentrations, thus causing further nonlinearity
from the interferent matrix before measurement of
(secondary absorption). The secondary absorption,
fluorescence. Time resolution may be useful when
when due to the analyte, is specifically denoted as
background fluorescence has a different decay time
self-absorption. Primary and secondary absorption
from that of the analyte. Measurement of analyte
together are called the inner filter effect.
fluorescence must be done at wavelengths at which
other sample components do not absorb or fluoresce.
Fluorescence Quenching
Fluorescent molecules are subject to intensity
variations as a function of temperature. In general,
The fluorescence signal generated by a molecule can
the frequency of collisional deactivation with solvent
be strongly altered by its environment. Apart from
molecules increases as the temperature increases.
inner filter effects, the interaction of the fluoro-
These collisions bleed energy from the excited state
chrome with its surroundings is another way in
and a fluorescence intensity decrease results. Temper-
which its fluorescence intensity can be decreased.
ature coefficients are typically 1 2% per 1C. A
This process is known as quenching. There are two
thermostated sample cell is recommended to ensure
kinds of quenching processes: (1) dynamic quenching
temperature control both in samples and in standards.
and (2) static quenching. The efficiency of the two
Photodecomposition can be a serious problem in
processes is dependent on the concentration of the
fluorescence during exposure to the exciting radia-
quencher molecules.
tion. The degree of photodegradation is proportional
Dynamic quenching or collisional quenching nor-
to the exposure time and to the incident radiant
mally refers to nonradiative energy transfer from
power. Photodegradation can be minimized through
excited species to other molecules:
the use of a longer-wavelength excitation band of the
analyte, if one is available, or through the use of
F þ Q-F þ Q
smaller excitation slit width in order to reduce the
Here, the excited analyte molecule, F , transfers ex-
incident radiant power.
citation energy to a quencher molecule, Q, causing
deactivation of F and forms an excited quencher
molecule, Q . For collisional quenching, the decrease
Analytical Techniques
in intensity often follows the well-known Stern Vol-
Steady-State Fluorescence
mer equation:
The most widely used analytical approach in fluo-
I0=I ź 1 þ K½QŠ ź1 þ kqt0½QŠ½1Š
rescence is the use of steady-state fluorescence, where
the intensity of the emitted light by the fluorochrome where I0 and I are the fluorescence intensities for the
is measured. In general terms, the intensity of emitted analyte in the absence of quencher and presence of
light, If, that is measured by the instrument is quencher at concentration [Q], respectively, K is the
FLUORESCENCE / Overview 103
Stern Volmer quenching constant, kq is the bimolec- is necessary, energy transfer can also take place over
ular quenching constant, and t0 is the unquenched large distances (up to 100 Å) without collisions. It
lifetime. The Stern Volmer quenching constant can occurs between two fluorescent molecules when one,
be estimated from a plot of (I0/I) 1 as a function of the donor, absorbs a photon, elevating an electron to
the quencher concentration. Equation [1] shows that a higher energy state and, through resonance, the
the effect of a quencher decreases as the sample is excitation of this electron is passed to another elec-
diluted because the probability of collision between tron in the second, the acceptor. This energy is then
the analyte and the quencher is minimized. Also, the re-emitted as a photon that is less energetic, and
probability of collisions between the excited fluor- therefore of longer wavelength than the photon in-
ophore and the quencher is higher as higher is the itially absorbed. The emitted radiation from the ac-
lifetime of the fluorophore. A wide variety of mol- ceptor is termed sensitized fluorescence as it is
ecules can act as collisional quenchers, among which observed without direct excitation of the acceptor.
small molecules such as oxygen, halogens, amines, For energy transfer, some basic conditions must be
and electron-deficient molecules like acrylamide are satisfied:
good quenchers for some fluorophores, particularly if
the fluorescence lifetime is greater than 10 ns. 1. the donor molecule must be a fluorophore with a
In static quenching, the quencher and the fluor- sufficiently long fluorescence lifetime;
ophore in the ground state form a stable complex. 2. the emission spectrum of the donor and the exci-
The process can be described by the following mech- tation spectrum of the acceptor must overlap par-
anism: tially; and
3. the distance between donor and acceptor must be
F þ Q-FQ Complex formation ½2Š
within a limiting range (usually 20 50 Å).
F þ hn1-F -F þ hn2 Fluorescence ½3Š
The rate of energy transfer (kT) to a specific ac-
ceptor is given by
FQ þ hn1-ðFQ Þ-FQ þ heat
kT ź t D1ðR0=RÞ6 ½5Š
Quenching by the complex ½4Š
where tD is the luminescence lifetime of the donor, R
In this case, fluorescence is only observed from the
is the distance between the donor and the acceptor,
unbound fluorophore. The decrease in fluorescence
and R0 is the critical distance (Föster distance) be-
intensity is described by eqn [1] where K must be
tween the donor acceptor pair for which the prob-
replaced by Kq, the constant of complex formation
ability of energy transfer and the deactivation of the
(Kq ź [FQ]/[F] [Q]). An example of this type of in-
donor by radiative and nonradiative processes is the
teraction is the quenching of ethidium bromide by
same. When the donor acceptor distance is equal to
caffeine.
the Föster distance, the transfer efficiency is 50%.
In contrast to dynamic quenching, static quench-
Energy transfer is widely used in biochemistry to
ing does not show an additional dependence on the
measure protein association, distances between two
lifetime of the excited state of the fluorophore. On
sites on a macromolecule, and the effects of confor-
the other hand, while formation of a complex just
mational changes on these distances. It is a powerful
reduces the concentration of the free fluorophores
method for obtaining both structural and dynamic
without affecting the lifetime of the excited mole-
information about macromolecules and macro-
cules, the lifetime of the complex is significantly
molecular complexes.
longer than the lifetime of the excited fluorophore.
So, measurement of the lifetime provides a means of
Synchronous Fluorescence
distinguishing between dynamic and static quench-
ing. Quenching studies provide direct information
A synchronous spectrum, with a constant energy
about diffusion of small molecules in solutions, in
difference (Dn), is performed by scanning the emis-
cellular tissues and in solid materials, and hence give
sion monochromator at a slightly faster rate than the
insights into solvent or solid viscosities and/or
excitation monochromator (Dl ź lem lex). Syn-
accessibilities.
chronous spectra are strongly dependent on the
wavelength offset. Ideally, only one peak is obtained
Energy Transfer
when Dn is set to the Stokes shift of the fluorophore
In contrast to fluorescence static or dynamic quench- of interest. The main purpose of synchronous
ing, for which a coupling between electronic orbitals scanning is to generate spectra having decreased
104 FLUORESCENCE / Overview
bandwidths. Synchronous scanning may decrease the emission when the sample is excited with vertically
extent of overlapping in the spectra of mixtures of polarized light. Anisotropy and polarization are both
fluorescent compounds in multicomponent fluores- expressions for the same phenomenon. Fluorescence
cence applications. polarization measurements are widely used in mo-
lecular biophysics and biochemistry for studying ro-
tational motions of electronically excited molecules,
Time-Resolved Fluorescence
to detect the binding of relatively small molecules to
In contrast to steady-state fluorimetry, time-resolved
macromolecules, to quantify protein denaturation,
fluorimetry is based on the measurement, at a fixed
internal dynamics of proteins as well as in fluoro-
wavelength, of fluorescence signal as a function of
immunoassay procedures.
time. For these measurements, the sample is exposed
to a pulse of light (pulse fluorimetry), where the pulse
Fluorescence Performance
width is short in comparison with the excited state
lifetime of the fluorescent molecule. The decay of the Characteristics
fluorescence is then recorded with a high-speed
The major advantage of fluorescence methods is sen-
detection system (nanosecond timescale). An alter-
sitivity. One can see why fluorimetric methods are so
native technique, phase-modulation fluorimetry (of-
sensitive if we compare them with absorption spec-
ten called frequency-domain fluorimetry), uses a
troscopy. In absorption methods, the absorbance of a
source that is amplitude-modulated at one or more
3
compound in the range of 1.0 10 absorbance
frequencies. Measurement of the phase or demodu-
units can be measured using a good absorption spec-
lation of the fluorescence signal can be used to gene-
trophotometer. Assuming a very efficient absorber
rate fluorescence decay times and time-resolved
with an extinction molar coefficient of eD1
fluorescence spectra. Commercial instrumentation is
1 1
105 l mol cm , we can calculate a low detection
available for these types of measurements. Time-re-
limit CDL of:
solved spectra or measurement of fluorescence decay
times reveal much of the molecular information
CDLD1:0 10 3=1 105 ź 1 10 8 mol l 1
available from fluorescence. For example, it is pos-
sible to distinguish: (1) sample constituents whose
using a 1 cm cell. On the other hand, for efficient
fluorescence spectra overlap one another; (2) between
luminescent molecules, with a good spectrofluorime-
static and dynamic quenching; and (3) the fluores-
ter in a low-background solution, detection limits in
cence of an analyte from background scattering.
fluorescence spectroscopy easily approach concen-
trations in the range of 1 10 pM. Consequently, de-
tection limits in fluorescence spectroscopy are
Fluorescence Anisotropy
generally more than three orders of magnitude bet-
Anistropy measurements are based on the photose-
ter than those achieved for the same molecules in
lective excitation of fluorophores by plane-polarized
absorption spectroscopy. For example, the best ex-
light. In an isotropic medium, the fluorophores are
perimental detection limit for efficient fluorescein, an
randomly oriented. Upon excitation with polarized
efficient fluorophore, is B0.1 pM, using laser exci-
light, those fluorophores whose absorption transition
tation and prepurified solvents. Using laser-induced
dipole is aligned parallel to the electric vector of the
fluorescence detection in capillary electrophoresis,
excitation, will be preferentially excited. If the mol-
detection limits of 0.1 100 amol (1 nl injection
ecule rotates and tumbles out of this plane during the
volume) have been reported for amino acids
excited state, light is emitted in a different plane from
(precolumn derivatization with dicarbocyanine
the excitation light. The intensity of the emitted light
fluorophore).
can be monitored in vertical and horizontal planes
Precision, at concentrations well above the detec-
and thus, fluorescence anisotropy (r) and polariza-
tion limit, is typically in the range 0.5 2% and it is
tion (P) are defined by:
limited by noise. In addition, sample treatment and
Ijj I> calibration must be considered.
r ź ½6Š
With external standardization, an equivalent anal-
Ijj þ 2I>
yte concentration in the standard and sample should
yield the same analyte fluorescence signal. The accu-
Ijj I>
P ź ½7Š
racy of the determinations is dependent on interfer-
Ijj þ I>
ences, chemical equilibrium involving the analyte,
where Ijj and I> are the fluorescence intensities of scattering, and quenching. Analyte interference due
the vertically ðjjÞ and horizontally (>) polarized to absorption, scattering, or quenching may be
FLUORESCENCE / Overview 105
eliminated by dilution if the detection limit is low. probes can be prepared chemically by preparing 50-
Also, standard additions can be useful in these cases. amino-end-labeled oligonucelotide probes on a DNA
The linear range of molecular fluorescence can be synthesizer. The use of fluorescent probes for in situ
four to six orders of magnitude for efficient fluoro- hybridization is sometimes referred to as FISH. Flu-
phores with low detection limits. As previously de- orescent DNA probe can also be prepared from tem-
scribed, nonlinearity begins to occur at analyte plate DNA by enzymatic methods (polymerases can
concentrations where inner filter effects become incorporate dUTPs, dATPs, and dCTPs that have flu-
significant. orescent dyes attached by linker arms to the nucleotide
base). An exciting new development is the use of mo-
lecular genetic methods to fuse the gene for the green
Applications fluorescent protein (intrinsic fluorophore) to other
target genes for subsequent expression in living cells.
Fluorescence spectroscopy is one of the most widely
3. Combination of fluorescence spectroscopy with
used molecular spectroscopic techniques in the fields
flow and image cytometry is a powerful tool for
of molecular biology, biophysics, and biochemistry.
diagnostic, prognostic and therapy control proce-
Some important applications have already been out-
dures in medicine as well as in the study of cellular
lined above, while describing the different fluorescent
aspects in immunology, cancer research, molecular
methods. Major classes of applications of fluorimetry
biology, and biotechnology.
include the following:
4. Multicomponent analysis using powerful fluores-
1. Determination of trace-level species, in clinical, cence lifetime discrimination approaches.
biological and environmental samples, including 5. As detection method in separation techniques,
inorganic species (metal ions, anions), organic, and especially in thin-layer chromatography, high-
biochemical compounds. Direct determination of performance liquid chromatography (HPLC) and
inorganic species through the formation of fluo- electrophoresis. In HPLC applications for non-fluo-
rescent chelates may be very sensitive. This approach rescent compounds, either precolumn or postcolumn
works well for diamagnetic metal ions as paramagne- fluorescent derivatization is used. In capillary electro-
tic ones tend to quench fluorescence through in- phoresis, the use of a laser to directly excite fluore-
tersystem crossing. Typical chelators include 8-hy- scent compounds allows a extremely high sensitivity
droxy-quinoline and derivatives, morine and related (femtomole to attomole levels of analytes can be
flavonols, benzoine, etc. Indirect methods are based detected in nanoliter volumes).
on measuring the quenching of fluorophores by 6. Automated batch and continuous flow analysis,
widely used in clinical laboratories where high sam-
inorganic species. This approach lacks sensitivity
and selectivity. Among the organic species of ple throughput is critical.
environmental concern, polycyclic aromatic hydro- 7. Studies of the microenvironment of fluorescent
carbons are the most common analytes. Direct de- probe molecules. Many environmental factors mod-
ify the fluorescent properties of a molecule, e.g.,
tection is possible for analytes of biomedical interest,
such as FAD, NAD, porphyrins, and aromatic amino solvent polarity, pH, proximity, and concentration of
acids (endogenous fluorescence). For nonfluorescent quenching species. The changes experienced by the
compounds indirect detection is used by derivatizat- fluorophore, sometimes subtle, can be used to obtain
ion using selective fluorescent probes. Most proteins information about the specific region in which the
and peptides can be directly labeled with fluoroph- fluorophore is localized.
ores via their available amine (lysine side chains) or 8. Fluorescent sensors in solution (chemosensors)
thiol (cysteine side chains) groups. Isothiocyanates, and fiberoptic sensors. Fluorescence is a particularly
such as fluorescein (FITC) and tetramethylrhodamine important technique in this field because of its high
sensitivity of detection down to a single molecule,
isothiocyanate (TRITC) are amine-reactive and are
widely used for labeling. Succinimidyl esters react subnanometer spatial resolution with submicrometer
with thiol groups and sulfonyl chlorides (e.g., dansyl visualization and submillisecond temporal resolu-
chloride) are reactive with amines and thiol groups. tion.
Fluorescent enzyme and immunoassays are widely
used to determine a variety of analytes of biological
See also: Chromatography: Overview. Fluorescence:
and clinical interest.
Instrumentation; High-Resolution Techniques; Time-Re-
2. Proteins, peptides, and nucleic acid labeling in mo-
solved Fluorescence; Derivatization; Fluorescence Label-
lecular biology research. The fluorescence modifica-
ing; Quantitative Analysis; Clinical and Drug Applications;
tion of nucleic acid molecules can be achieved in a
Environmental Applications; Food Applications. Sensors:
number of ways. For example, fluorescent DNA Photometric.
106 FLUORESCENCE / Instrumentation
Rettig W, Strehmel B, Schrader S, and Seifert H (1999)
Further Reading
Applied Fluorescence in Chemistry, Biology and Medi-
Guilbault GG (ed.) (1990) Practical Fluorescence, 2nd edn,
cine. Berlin: Springer.
revised and expanded. New York: Dekker.
Valeur B and Brochon J-C (eds.) (2001) New Trends in
Lakowicz JR (ed.) (1999) Principles of Fluorescence
Fluorescence Spectroscopy: Applications to Chemical
Spectroscopy, 2nd edn. Dordrecht: Kluwer
and Life Sciences. Springer Series on Fluorescence. Ber-
Academic.
lin: Springer.
Instrumentation
J N Miller, Loughborough University of Technology,
Sample
Loughborough, UK
& 2005, Elsevier Ltd. All Rights Reserved.
Light source Monochromator M1
Power supply
Introduction Monochromator M2
Fluorescence spectrometers capable of measuring ex-
citation and emission spectra as well as fluorescence
intensities have been commercially available for al-
most 50 years. While the functions and the overall
Power supply Detector
layout of the main components have not changed
very much over that period, the individual compo- Figure 1 Block diagram of the principal components of a flu-
orescence spectrometer.
nents have developed enormously in terms of sen-
sitivity, stability, and longevity, with the result that
modern instruments have capabilities far superior to
those 50 years ago. Moreover, many accessories,
and accessories for individual applications. Its crucial
which were either unobtainable or unusual in the
advantages are that photons transmitted by the sam-
1950s are now routinely available, allowing the
ple (the majority of the incident photons when dilute
selectivity as well as the sensitivity of fluorescence
solutions are studied) are not detected, and that
methods to be more fully exploited. Another major
Rayleigh scattered light has a minimum intensity in
change, characteristic of the last two decades, has
the 901 direction. The key optical elements of
been the universal use of personal computers to con-
the layout are the light source, the excitation
trol the instruments as well as to record and mani-
monochromator or filter (M1), the sample cell, the
pulate the data. This rapid data processing capability
emission monochromator or filter (M2), and the de-
has further enhanced the capabilities of fluorescence
tector. Separate power supplies are normally pro-
spectroscopy. This article summarizes the optical
vided for the light source and the detector, and
layout used in virtually all commercial instruments,
additional power may be needed for such items as a
and describes the main components used in modern
reference detector, sample stirrer, etc. This article
spectrometers.
deals only with steady-state fluorescence instru-
ments: systems for studying lifetimes are summa-
rized in a separate article.
Instrument Layout
The layout of the major fluorescence spectrometer
components is shown in Figure 1. The characteristic
Light Sources
feature of this arrangement is that any fluorescence
emitted by the sample is detected at 901 to the The fluorescence signal from a sample in dilute so-
incident light beam. This geometry, pioneered by lution is (if photodecomposition effects are ignored)
Stokes in the 1850s, is still used in virtually all com- proportional to the incident light intensity. Thus the
mercial and laboratory-built fluorescence instru- light source for fluorescence spectroscopy should
ments, inevitably with many additions, variations, have an intense and stable output. Other desirables