Analytical Chemistry
Fluorescence Spectroscopy
Stokes shift between absorbed and emitted light — quantum yield Φ_F, lifetime τ, used in biology, sensors, single-molecule
Fluorescence spectroscopy measures the light spontaneously re-emitted by a molecule that has absorbed a higher-energy photon and relaxed nonradiatively to the lowest vibrational level of its first excited singlet state S1. Sir George Stokes first identified the wavelength shift between excitation and emission in 1852 — the Stokes shift — using a quinine sulfate solution and prism. Two key parameters define a fluorophore: the fluorescence quantum yield Φ_F (photons emitted per photon absorbed, between 0 and 1) and the fluorescence lifetime τ (typically 0.1 to 20 ns for organic fluorophores). The technique enables single-molecule detection, FRET-based distance measurement at the 1 to 10 nm scale, and live-cell imaging via genetically encoded GFP and its variants.
- Stokes shift~25 nm (fluorescein); 30 nm (GFP)
- Quantum yield Φ_F0–1; fluorescein 0.95; GFP 0.6
- Lifetime τ0.1–20 ns typical; ruthenium >100 ns
- FRET R₀5–7 nm typical; 1/r⁶ scaling
- DiscoveredGeorge Stokes, 1852
- NobelChem 2008 (GFP); 2014 (super-res)
Interactive visualization
Press play, or step through manually. The visualization is yours to drive — try it before reading on.
Watch the 60-second explainer
A condensed visual walkthrough — narrated, captioned, under a minute.
Why fluorescence matters
- Sensitivity orders of magnitude beyond UV-Vis. Fluorescence is read against a near-zero background (you measure photons emerging at 90° from a wavelength different from the excitation). Single-molecule detection is routine; bulk detection limits are 10⁻¹² to 10⁻¹⁰ M for high-Φ_F dyes — about 10⁵ times more sensitive than absorbance UV-Vis at the same wavelength.
- Multiplexing via spectral and lifetime channels. A fluorometer with four colors plus polarization gives 8 to 12 independent labels; flow cytometers routinely measure 18-color panels. FLIM imaging distinguishes fluorophores with similar emission but different τ — eGFP vs Alexa488 are co-localized only in lifetime space.
- Distance ruler at the 1 to 10 nm scale. FRET and PIFE (protein-induced fluorescence enhancement) report distances inaccessible to crystallography (which needs the molecule to crystallize) and to NMR (which requires <30 kDa molecular weight). smFRET on freely diffusing or surface-tethered molecules is now the standard biophysics tool for protein conformational dynamics on the µs to s timescale.
- Genetically encoded reporters. Fusing GFP, mCherry, miRFP, or biosensor variants (GCaMP for Ca²⁺, voltage-sensitive ASAP3 for membrane potential, ATP1.0 for ATP) lets a single transfection report on subcellular dynamics in real time. CaMP6f reports cytosolic Ca²⁺ on millisecond timescales and is the workhorse of neural-activity imaging.
- Quantitative bioassays at scale. 384- and 1536-well plate readers running fluorescence polarization, FRET, or AlphaScreen drive industrial drug screening. Roche's COBAS line uses fluorescence-quenching probes (TaqMan, FRET hydrolysis) for quantitative PCR; ~95% of all clinical PCR is fluorescence-detected.
- Sensors with chemical specificity. Fura-2 chelates Ca²⁺ with a stoichiometric 380 nm/340 nm excitation ratio; SNARF reports pH at 580/640 nm; nitric-oxide-sensitive DAF-2 fluoresces only when reacted with NO. The structural design that turns binding into an emission shift is now a mature subfield of organic synthesis.
- Forensic and document tracing. Optical brighteners (stilbenes, coumarins) added to currency and passport substrates fluoresce under UV and serve as a primary anti-counterfeit signal. Fluorescence microscopy of paper fibers identifies printer toners and inks at the trace level.
Common misconceptions
- Fluorescence intensity is linear in concentration. Only when εcl < ~0.05. Above that, inner-filter effects (excitation light absorbed before reaching the detection volume; emission re-absorbed by other fluorophores en route to the detector) bend the curve. Dilute, use shorter path cells, or apply the analytic correction F_corrected = F·10^((A_ex + A_em)/2).
- Fluorescence and phosphorescence are the same. Fluorescence is a singlet-singlet (S1→S0) transition with τ = ns. Phosphorescence is a triplet-singlet (T1→S0) spin-forbidden transition with τ = ms to s, often only visible at low temperature in glassy matrices. Phosphorescence tags (lanthanide chelates, transition-metal phosphors) gate detection 10 µs after excitation, eliminating short-lived background.
- A higher Φ_F always means a brighter probe. Brightness = ε(λ_ex) · Φ_F. A dye with ε = 5,000 M⁻¹·cm⁻¹ and Φ_F = 0.9 is dimmer than one with ε = 80,000 and Φ_F = 0.3. Compare brightness products, not Φ_F alone.
- Photobleaching is an annoyance. It is a hard limit: a single fluorophore emits only ~10⁵ to 10⁷ photons before bleaching, setting the maximum total information one molecule can provide. PALM/STORM exploits the controllable on/off of photoswitchable variants; ALEX-FRET cycles donor and acceptor excitation to disentangle photobleaching from FRET changes.
- Solvent doesn't matter. Polar solvents stabilize the excited state more than the ground state for charge-transfer fluorophores, red-shifting emission (positive solvatochromism) by 20 to 100 nm between hexane and water. Lippert-Mataga analysis quantifies the dipole-moment change on excitation. Always specify solvent when reporting Φ_F or λ_em.
- Quenching always means contamination. Quenching can be desired: TaqMan probes use FRET quenching that is relieved by polymerase cleavage, generating signal in qPCR. Stern-Volmer quenching by O₂ is the basis of optical-fiber dissolved-oxygen sensors used in fish farms and bioreactors.
Jablonski diagram and instrument geometry
The Jablonski diagram (Aleksander Jabłoński, 1933) summarizes electronic states. A fluorophore in S0 absorbs a photon to a vibrationally excited level of S1; vibrational relaxation in ~1–10 ps brings it to v=0 of S1 (Kasha's rule: emission is from v=0 of the lowest excited singlet, regardless of which state was excited); fluorescence to S0 occurs in ~1–10 ns; further vibrational relaxation in S0 in ~ps closes the cycle. Competing pathways: internal conversion S1→S0 nonradiatively, intersystem crossing S1→T1 (spin-forbidden, slow), and quenching by external molecules. The Stokes shift is the energy difference between absorption and emission maxima and originates in the two vibrational relaxations.
A spectrofluorometer puts excitation and emission monochromators at right angles to the cuvette. The 90° geometry minimizes excitation light reaching the detector; an emission cutoff filter further suppresses Rayleigh and Raman scatter. Modern instruments scan emission spectra at fixed excitation in 10 to 60 seconds, and excitation spectra at fixed emission similarly. Photon-counting photomultipliers reach single-photon sensitivity with dark-count rates below 10/s. For imaging, confocal microscopy uses a focused laser excitation and a pinhole-conjugate detection that rejects out-of-focus emission, giving optical sectioning at ~250 nm lateral and ~700 nm axial resolution. Two-photon microscopy excites with 700–1100 nm pulsed lasers and exploits the quadratic intensity dependence at the focal volume to limit excitation to a single ~femtoliter spot, enabling deep-tissue imaging up to ~1 mm.
Lifetime measurements use time-correlated single-photon counting (TCSPC). A pulsed laser fires every ~12.5 ns; a constant-fraction discriminator on the photomultiplier records the time difference between the laser pulse and the first detected photon, accumulating a histogram over ~10⁵ to 10⁹ pulses. Fitting the histogram (after deconvolving the instrument response function, typically 50 to 200 ps FWHM) recovers τ, often as a multi-exponential when multiple species contribute. FLIM applies TCSPC at every pixel of a confocal image and yields lifetime maps that distinguish fluorophores even when their spectra overlap — a key tool for FRET imaging in live cells where intensity changes are confounded by uneven concentration.
Fluorescence vs phosphorescence vs Raman vs UV-Vis
| Property | Fluorescence | Phosphorescence | Raman | UV-Vis Absorbance | Chemiluminescence | FRET |
|---|---|---|---|---|---|---|
| Origin | S1 → S0 (allowed) | T1 → S0 (forbidden) | Inelastic scattering, virtual state | S0 → Sn absorbed | Chemical reaction → S1 → emission | Donor S1 → acceptor (non-radiative) |
| Lifetime | 0.1–20 ns | 1 µs – 10 s | ~10⁻¹⁵ s (instantaneous) | Femtoseconds | seconds (limited by reaction rate) | ~donor τ / (1+E·R0⁶/r⁶) |
| Stokes shift | 20–60 nm | 50–300 nm | 10–4000 cm⁻¹ | None | Not applicable (no Ex) | Effective up to 100 nm |
| Sensitivity | ~10⁻¹² M; single-molecule | ~10⁻⁹ M (gated) | ~10⁻³ M (intrinsic) | ~10⁻⁶ M (cuvette) | ~10⁻¹⁵ mol (ATP detection) | Single-pair distances |
| Information | Concentration, environment | O₂ sensing, time-gated bg removal | Vibrational fingerprint | π-system, conjugation | Reaction rate | 1–10 nm distances |
| Best for | Imaging, tracking, sensors | Glow-in-dark, oximetry | Identity (whole spectrum) | Quantitation in linear regime | Trace ATP, blood (luminol) | Conformational dynamics |
| Typical instruments | Spectrofluorometer, confocal | Phosphorimeter, gated PMT | Raman + dispersive or FT | UV-Vis, photodiode array | Luminometer, plate reader | Smfret, FLIM, plate reader |
Famous experiments and Nobel-recognized milestones
- Stokes 1852 — origin of the Stokes shift. George Stokes used quinine sulfate in dilute sulfuric acid (Φ_F ≈ 0.55, λ_ex 350 nm, λ_em 450 nm) and a prism to show that emission was always longer in wavelength than excitation. The compound is still the gold-standard reference in tonic water; its blue glow under UV is the quantum-yield reference for spectrofluorometer calibration.
- Aequorea GFP — 2008 Chemistry Nobel. ε ≈ 55,000 M⁻¹·cm⁻¹ at 488 nm, Φ_F ≈ 0.6 for eGFP. The 2008 prize to Shimomura, Chalfie, and Tsien recognized the discovery, expression in C. elegans (1994 paper, six glowing neurons in a transparent worm), and engineering of red, yellow, cyan variants. mScarlet (2017) reaches Φ_F ≈ 0.7 in red — the brightest currently available genetically encoded label.
- FRET in protein folding — Lubert Stryer 1967. Stryer and Haugland labeled poly-L-proline with dansyl donor and naphthyl acceptor at 12 to 46 residues spacing and demonstrated the predicted r⁶ FRET dependence. The 1967 PNAS paper established FRET as a 'spectroscopic ruler.' Modern smFRET on Holliday junctions, ribosomes, and ion channels descends directly from this experiment.
- Super-resolution imaging — 2014 Chemistry Nobel. Eric Betzig (PALM, 2006), Stefan Hell (STED, 2000), and W.E. Moerner (single-molecule, 1989) shared the prize for breaking the diffraction limit. PALM/STORM achieves ~20 nm lateral resolution by stochastically activating sparse subsets of fluorophores, localizing each centroid to ~σ/√N where N ~ 10³–10⁴ photons per molecule. STED uses a doughnut-shaped depletion beam to confine emission to a sub-diffraction spot, ~50 nm in living cells.
- GCaMP for neural-activity imaging. A 2001 fusion of CaM, M13 peptide, and circularly permuted GFP (Nakai, Ohkura, Imoto) gave a Ca²⁺-sensitive fluorescence change ΔF/F ≈ 100% at saturation. GCaMP6f (2013) reaches ΔF/F > 25 per single action potential in mouse cortex; whole-brain optical recording in larval zebrafish (Ahrens et al., Nature 2013) used GCaMP3 to image ~80,000 neurons at 0.8 Hz.
Frequently asked questions
What is the Stokes shift and why does it occur?
The Stokes shift is the difference in wavelength (or energy) between the excitation maximum and emission maximum of a fluorophore. For fluorescein in water it is roughly 25 nm (494 nm absorption, 519 nm emission), for rhodamine 6G about 25 nm, and for GFP about 30 nm (488 nm excitation, 507 nm emission). The shift arises because vibrational relaxation in the excited electronic state (S1) is fast (~picoseconds) compared to fluorescence emission (~nanoseconds). The molecule absorbs a photon with energy hν_abs, instantly relaxes vibrationally to v=0 of S1 (loss to heat), then emits hν_em < hν_abs to a vibrationally excited level of S0 from which it again relaxes. The energy gap is twice the vibrational reorganization energy, named for George Stokes who described it in 1852 using quinine sulfate and a prism in his Bakerian Lecture.
What is the fluorescence quantum yield?
Φ_F is the fraction of absorbed photons that are re-emitted as fluorescence: Φ_F = k_r / (k_r + k_nr), where k_r is the radiative rate constant and k_nr lumps internal conversion, intersystem crossing to triplet, and quenching by collisions or solvent vibrations. Bright fluorophores: fluorescein in 0.1 M NaOH Φ_F ≈ 0.95, rhodamine 6G in ethanol Φ_F ≈ 0.95, GFP enhanced (eGFP) Φ_F ≈ 0.6, quinine bisulfate in 0.5 M H2SO4 Φ_F ≈ 0.55 (the long-standing reference standard). Dim fluorophores: tryptophan in water Φ_F ≈ 0.13, riboflavin Φ_F ≈ 0.27, NADH in water Φ_F ≈ 0.02. Φ_F changes with solvent polarity, viscosity, pH, temperature, and quenchers — every fluorescence assay reports it because it sets the achievable signal.
What is fluorescence lifetime and how is it measured?
The lifetime τ is the time for excited-state population to fall to 1/e of its initial value, τ = 1/(k_r + k_nr). For most organic fluorophores τ = 0.5–10 ns; tryptophan ~3 ns, fluorescein ~4 ns, GFP ~3 ns, ruthenium polypyridyl complexes 100 ns to 1 µs. Measurement uses time-correlated single-photon counting (TCSPC): a pulsed laser (~50 ps pulse, 80 MHz repetition) excites the sample, and a constant-fraction discriminator records the delay until the first emitted photon. Histogram of delays over millions of pulses gives the decay curve, fit by a single or multi-exponential. Frequency-domain alternatives modulate excitation at MHz to GHz and measure phase shift and demodulation. Lifetime is independent of fluorophore concentration and excitation intensity, so FLIM (fluorescence lifetime imaging) is robust to scattering and uneven illumination — a major advantage over intensity-only imaging.
How does FRET measure distances?
Förster resonance energy transfer (FRET) is non-radiative dipole-dipole coupling between a donor and acceptor when their separation r is less than ~10 nm and the donor emission spectrum overlaps the acceptor absorption. The transfer efficiency is E = R0⁶ / (R0⁶ + r⁶), where R0 (the Förster radius) is the distance at which E = 50%. For typical fluorophore pairs R0 = 5–7 nm: Cy3-Cy5 R0 = 5.4 nm, fluorescein-tetramethylrhodamine R0 = 5.5 nm, CFP-YFP R0 = 4.9 nm. The r⁶ dependence means FRET is most sensitive between 0.5·R0 and 1.5·R0. FRET assays measure protein-protein interactions, conformational changes (e.g. kinesin walking, polymerase fidelity), DNA folding, and lipid bilayer dynamics. The 1948 theory was developed by Theodor Förster; FRET became routine after Stryer and Haugland's 1967 protein labeling demonstration.
Why is GFP a Nobel Prize-winning tool?
Aequorea victoria green fluorescent protein (GFP) is a 27 kDa β-barrel protein whose chromophore (a p-hydroxybenzylidene-imidazolinone) forms autocatalytically from three internal amino acids (Ser-Tyr-Gly at positions 65–67) without any external cofactor. Discovered by Osamu Shimomura in 1962 in jellyfish, cloned by Douglas Prasher (1992), and shown by Martin Chalfie to fluoresce in C. elegans (1994), then developed by Roger Tsien into a palette of color variants (BFP, CFP, YFP, mCherry). Excitation at 488 nm gives emission at 507 nm with Φ_F ≈ 0.6 and ε ≈ 55,000 M⁻¹·cm⁻¹. Because GFP is genetically encoded, fusion to any protein of interest creates a non-invasive in vivo tracer — the technique that enabled live-cell, live-animal, and live-tissue imaging of protein localization, expression, and dynamics. Shimomura, Chalfie, and Tsien shared the 2008 Chemistry Nobel.
How is single-molecule fluorescence achieved?
Single fluorophores emit roughly 10⁵ to 10⁷ photons before photobleaching. To detect them you need extremely low background: confocal pinhole geometry rejects out-of-focus light, total internal reflection (TIRF) excites only the 100 nm closest to the coverslip, and zero-mode waveguides shrink the excitation volume to attoliters. Combined with avalanche-photodiode detectors at 50% quantum efficiency, single-molecule detection is now routine in flowing cells, on supported lipid bilayers, and inside living E. coli. Applications include single-pair FRET (smFRET) on freely diffusing molecules, super-resolution PALM/STORM that pushes optical resolution from 250 nm to ~20 nm, and DNA sequencing (PacBio's zero-mode waveguides watch single polymerase incorporations in real time). The 2014 Chemistry Nobel to Eric Betzig, Stefan Hell, and W.E. Moerner recognized super-resolution as the apex of single-molecule fluorescence.