Ultraviolet–visible spectroscopy is a branch of spectroscopy concerned with the interaction of ultraviolet and visible electromagnetic radiation with matter. In its most common form, it measures how strongly a sample absorbs light at different wavelengths. The resulting spectrum reflects transitions between electronic states and provides information about chemical composition, concentration, molecular environment, and the electronic properties of materials.
Routine laboratory instruments commonly operate over approximately 190–800 nm, although many extend into the near-infrared. The visible region is conventionally placed at about 380–780 nm, with boundaries that vary by context. Measurements at shorter ultraviolet wavelengths require specialized optics and often evacuated or purged light paths. UV–visible spectroscopy is important in analytical chemistry, biochemistry, environmental analysis, and materials science because it combines relatively simple instrumentation with rapid, often nondestructive measurements.
The method ranges from single-wavelength concentration measurements to complete spectra, time-resolved experiments, and spatially resolved imaging. Its quantitative foundation is the relationship between light attenuation, optical path length, and absorber concentration. Its interpretive foundation is the connection between electronic energy levels and the wavelengths and intensities of absorption bands.
Physical basis and measured quantities
Absorption occurs when a photon transfers energy to a sample and promotes it from a lower to a higher electronic state. The photon energy is related to wavelength by
where is the Planck constant, is frequency, is the speed of light in vacuum, and is vacuum wavelength. A wavelength of 200 nm corresponds to approximately 6.20 electronvolts, whereas 800 nm corresponds to about 1.55 eV. This range is well suited to many valence-electron excitations in molecules, metal complexes, and solids.
An absorption experiment compares the radiant power transmitted through a sample, , with a reference power, . The transmittance is , and the decadic absorbance is
Absorbance is dimensionless. A value of 1 corresponds to 10% transmission; a value of 2 corresponds to 1%. “Absorbance units” are often used informally to describe instrument readings, but they are not an independent physical unit. Absorbance should also be distinguished from absorptance, the fraction of incident power actually absorbed.
In practical spectrophotometry, is usually established with a blank containing the same solvent and, where appropriate, the same reagents as the sample but not the analyte. This compensates for much of the absorption and reflection associated with the solvent and container. It does not automatically correct every optical loss. Scattering, changes in surface reflection, and differences between sample and reference cells can all contribute to measured attenuation.
A spectrum ordinarily plots absorbance against wavelength. An absorption maximum, denoted , is a convenient descriptor, but the whole band shape can be informative. Solutions often show broad bands because electronic transitions are accompanied by many closely spaced vibrational transitions and because molecules experience fluctuating solvent environments. Gas-phase spectra can show much finer structure. Consequently, UV–visible spectra usually provide less uniquely identifying structural detail than techniques that resolve numerous vibrational or magnetic-resonance features, although they can be exceptionally sensitive to changes in electronic structure.
Electronic transitions and absorption intensity
The interpretation of molecular absorption draws on quantum mechanics and molecular orbital theory. In a simplified orbital description, absorbed light promotes an electron from an occupied orbital to an unoccupied orbital. Common labels include , , , and , where denotes a nonbonding orbital and the asterisk denotes an antibonding orbital. These labels describe dominant contributions to an excitation rather than a complete many-electron account.
Transitions involving strong sigma bonds generally require high energies and frequently occur below the accessible range of routine instruments. Molecules containing double bonds, aromatic systems, or nonbonding electron pairs often absorb at longer ultraviolet wavelengths. A carbonyl group, for example, can exhibit a relatively weak band and a much stronger band. Increasing the extent of a conjugated system often reduces the energy separation of relevant orbitals, moving absorption toward longer wavelengths. This principle helps explain the visible colors of many dyes and biological pigments.
Not every energetically possible transition absorbs equally strongly. Transition probability depends on the transition dipole moment, which incorporates the initial and final electronic wave functions and the electric-dipole operator. Selection rules summarize conditions under which this quantity vanishes or becomes small. Spin selection generally favors transitions that conserve total spin angular momentum. In centrosymmetric systems, parity restrictions can make certain transitions weak. Vibrational motion, loss of symmetry, and spin–orbit coupling can relax these restrictions, so “forbidden” usually means low intensity rather than absolute impossibility.
Compounds studied in coordination chemistry show additional characteristic behavior. Transitions between predominantly metal-centered states depend on ligand geometry, metal identity, and oxidation state. They are often weaker than charge-transfer transitions, in which excitation redistributes electron density between a metal and its surrounding ligands or between different molecular components. Charge-transfer bands can be very intense and may dominate a spectrum even when metal-centered transitions are also present.
Observed band widths and shapes reflect more than orbital energies. The Franck–Condon principle describes electronic excitation as occurring much faster than nuclear rearrangement, leading to transitions among vibrational levels with different overlap probabilities. Solvent interactions, collisions, temperature, and unresolved conformations further broaden the spectrum. Absorption therefore reports both the electronic states of an isolated chemical species and the environment in which that species is measured.
Beer–Lambert law and quantitative analysis
For a homogeneous solution containing a single independently absorbing species, the Beer–Lambert law is commonly written
where is the molar absorption coefficient, is optical path length, and is molar concentration. When is expressed in centimetres and in moles per litre, is expressed in . The law follows from exponential attenuation when each increment of path length removes the same fraction of the remaining light.
Quantitative analysis usually relies on calibration with standards rather than solely on a published absorption coefficient. Standards establish the relationship between concentration and measured response under the actual solvent, wavelength, temperature, and instrumental conditions. For example, a compound with , measured in a 1 cm cell at , has a predicted absorbance of 0.10. The same solution in a 1 mm path gives 0.010, illustrating how strongly sensitivity depends on cell geometry.
For a mixture of noninteracting absorbers, individual absorbances are additive:
Measurements at several wavelengths can therefore support simultaneous determination of several components. Success depends on their spectra being sufficiently distinct. If two species have almost proportional spectra throughout the measurement range, separating their concentrations becomes mathematically ill-conditioned. Full-spectrum fitting, derivative methods, and chemometrics can exploit additional spectral information, but they cannot create chemical selectivity absent from the data.
Departures from linearity have several origins. At elevated concentrations, solute interactions and changes in refractive index may alter the absorption coefficient. More commonly, dilution changes protonation, association, or complexation equilibria, so concentration no longer tracks a single chemical species. Instrumental causes include finite spectral bandwidth, detector nonlinearity, and light reaching the detector by unintended paths. A calibration curve can appear smooth while concealing such systematic effects.
Blank subtraction, matrix matching, and dilution are central to quantitative interpretation. A calibration made in pure solvent may not remain valid in a sample containing salts, surfactants, or other absorbers. The method of standard additions can compensate for some matrix effects when added analyte behaves like the original analyte, but it does not automatically remove overlapping absorption from unrelated substances. Quantitative results therefore depend on chemical selectivity and sample handling as well as on the linearity of the optical measurement.
Instrument architecture and detection
A conventional UV–visible spectrophotometer contains a radiation source, wavelength-selection optics, a sample compartment, a detector, and electronics for signal processing. Deuterium discharge lamps provide a useful ultraviolet continuum, while tungsten–halogen lamps supply strong visible and near-infrared output. Xenon lamps, including pulsed xenon sources, can cover broad wavelength intervals with a single source. Light-emitting diodes are also used in compact instruments intended for a limited set of wavelengths.
In a scanning instrument, a monochromator selects a narrow wavelength band before or after the sample. Most modern designs use a diffraction grating, slits, and imaging optics. The spectral bandwidth is distinct from the wavelength interval between recorded data points: collecting points every 0.1 nm does not give 0.1 nm optical resolution if the transmitted band is several nanometres wide. Narrower slits improve spectral resolution but reduce radiant power, often increasing noise or requiring longer integration times.
Single-beam instruments measure a blank and sample sequentially. Double-beam instruments divide or alternate the light between sample and reference paths, allowing correction for source fluctuations and some instrumental drift. The reference does not eliminate mismatches between cuvettes or all differences in optical geometry. Some instruments monitor a reference beam without passing it through a second sample cell, providing source normalization rather than a complete sample–reference comparison.
Detectors convert light into electrical signals. Photomultiplier tubes offer high sensitivity and have long been used in scanning systems. Silicon photodiodes are robust and widely used over the ultraviolet and visible range. In a photodiode array instrument, dispersed light reaches many detector elements, allowing a broad spectral region to be recorded without mechanically scanning a monochromator. Charge-coupled and related imaging detectors are also used in spectrographs. Array acquisition is especially useful when spectra change rapidly, although exposure timing, saturation, and wavelength-dependent response still require control.
Instrument configuration determines practical performance. Baffles and filters suppress stray radiation and unwanted diffraction orders; source-switching regions can introduce baseline features if not adequately corrected. High-performance systems may use two monochromators to improve rejection of unwanted wavelengths. Detectors have finite dynamic range, and electronic gain cannot compensate indefinitely for insufficient transmitted light. Thus, wavelength range, resolution, speed, and absorbance range are linked engineering choices rather than independent specifications.
Samples, optical cells, and measurement geometries
Liquid measurements commonly use a cuvette with parallel optical windows and a nominal 10 mm path length. Fused silica is required for much of the ultraviolet region because ordinary glass absorbs strongly at shorter wavelengths. Disposable plastic cells are useful only within their specified spectral range. Optical path length, window quality, surface cleanliness, and cell orientation all affect reproducibility, especially when absorbance is small or ultraviolet wavelengths are used.
The solvent must dissolve the sample without obscuring the spectral region of interest. Every solvent has a practical ultraviolet cutoff below which its own absorption becomes excessive. Water and several high-purity organic solvents permit measurements relatively far into the ultraviolet, whereas other solvents absorb at substantially longer wavelengths. The cutoff is not a sharp universal boundary: it depends on purity, path length, and the acceptable background absorbance. Dissolved contaminants can produce features that are significant even when they are visually undetectable.
Short-path cells and variable-path instruments extend measurements to concentrated solutions. Microvolume instruments can measure microlitre-scale samples between closely spaced optical surfaces, often changing path length to keep the signal within range. Their reported absorbances may be normalized to a 1 cm path for comparison, making the distinction between measured and equivalent path length important. Microplate readers increase throughput, but the liquid depth depends on well geometry and volume. A nominally identical concentration can therefore give different absorbances at different filling volumes unless path length is corrected.
Turbid suspensions, powders, coatings, and rough surfaces require geometries beyond ordinary transmission. An integrating sphere collects light over a wide range of directions and can measure total transmission or reflection. For particulate materials, diffuse reflectance spectroscopy is often more useful than attempting to transmit light through an optically thick sample. The measured reflectance depends on both absorption and scattering, so it cannot generally be interpreted as solution absorbance.
For sufficiently thick, diffusely scattering samples, the Kubelka–Munk transformation is often used:
where is reflectance relative to an appropriate standard, and and are model absorption and scattering coefficients. The relation rests on an approximate two-flux treatment and restrictive assumptions about homogeneity and scattering. Changes in particle size or packing can change , meaning that apparent spectral differences need not reflect changes in chemical absorption alone.
Chemical structure, equilibria, and spectral interpretation
The portion of a molecule primarily responsible for a particular absorption feature is called a chromophore. Aromatic rings, conjugated carbonyl systems, azo groups, and metal–ligand assemblies are familiar examples. Substituents modify absorption through electron donation or withdrawal, changes in conjugation, steric effects, and interactions with solvent. An absorption shift to longer wavelength is termed bathochromic; a shift to shorter wavelength is hypsochromic. Increased and decreased band intensities are described as hyperchromic and hypochromic changes.
Solvent-dependent shifts, known as solvatochromism, arise because ground and excited states respond differently to their surroundings. A more polar solvent may stabilize one state more strongly than the other, changing the transition energy. Specific hydrogen bonding can be especially important for nonbonding orbitals. There is consequently no universal rule that a more polar solvent must move every band in the same direction. Interpreting a shift requires attention to the transition type and the nature of solute–solvent interactions.
UV–visible spectra are particularly useful for examining chemical equilibria that alter the chromophore. Acid–base indicators change spectrum because protonation changes their electronic structure. If only protonated and deprotonated forms contribute, measured spectra across a controlled pH range can be fitted to obtain an acid dissociation constant, provided concentrations, activities, and other equilibria are treated appropriately. Metal–ligand binding, molecular association, and redox transformations can be investigated by analogous titrations.
An isosbestic point is a wavelength at which spectra recorded during a transformation intersect. For an ideal interconversion between two species with equal molar absorption coefficients at that wavelength and constant total concentration, absorbance remains unchanged there. Such behavior is consistent with a simple interconversion but does not prove that only two species exist. Several species can share a crossing point accidentally or through constrained composition changes; instrumental or baseline effects can also obscure an otherwise expected intersection.
Structural assignments based only on UV–visible spectra are usually nonunique. Related compounds can have similar broad bands, while the same compound can change substantially with solvent, protonation, or aggregation. Comparison with infrared spectroscopy, nuclear magnetic resonance, or mass spectrometry commonly supplies the missing structural constraints. UV–visible measurements are strongest when a chemically plausible model predicts a spectral change and the experiment tests that model under controlled conditions.
Biochemical measurements and reaction kinetics
Biological macromolecules contain several useful ultraviolet chromophores. Proteins absorb near 280 nm mainly because of tryptophan and tyrosine residues, with smaller contributions from disulfide bonds. Their peptide backbone absorbs strongly in the far ultraviolet. Absorption coefficients at 280 nm can be estimated from amino-acid composition, but accuracy depends on sequence, chemical state, and measurement conditions. Proteins lacking substantial aromatic content may absorb only weakly at this wavelength despite being present at appreciable concentration.
Nucleic acids, including DNA and RNA, have strong absorption near 260 nm from their bases. Common approximate conversions for a 1 cm path are an absorbance of 1 for 50 double-stranded DNA, 40 RNA, or 33 single-stranded DNA. These are conventions rather than universal constants: sequence, strand structure, ionic conditions, and contaminants affect the result. Ratios such as provide screening information, but do not uniquely identify contaminants or establish sample suitability.
Absorption can report structural changes as well as concentration. Heating double-stranded DNA disrupts base stacking and pairing, usually increasing absorbance near 260 nm. Recording this hyperchromic change against temperature gives a melting profile. Its midpoint depends on sequence, salt conditions, strand concentration where relevant, and the definition used to extract the transition. Spectra of proteins can also respond to unfolding, although ultraviolet absorption generally provides less detailed conformational information than some complementary optical methods.
UV–visible spectroscopy is widely used to follow enzyme reactions and other processes studied in chemical kinetics. Reduced nicotinamide adenine dinucleotide, NADH, has a characteristic band near 340 nm that is weak in the oxidized form, NAD. Its commonly used molar absorption coefficient at 340 nm is approximately . When reaction stoichiometry is known, an absorbance slope can be converted into a concentration rate:
For reactions too fast to follow by manual mixing, stopped-flow instruments rapidly combine solutions and record absorbance after a short mixing interval. Experiments can resolve millisecond-scale events in suitable systems, although the effective dead time depends on instrument design and solution properties. Global analysis of spectra recorded over time can help distinguish intermediates. The inferred rate constants and mechanisms nevertheless depend on the kinetic model, mixing performance, and whether distinct intermediates have distinguishable spectra.
Environmental, industrial, and materials applications
Many substances that absorb weakly in the accessible ultraviolet–visible region are measured after conversion to colored products. In environmental and industrial analysis, this approach provides sensitivity and chemical selectivity through controlled reactions. Phosphate can be determined through phosphomolybdate chemistry, often followed by reduction to a blue product; nitrite can be measured by formation of an azo dye through the Griess reaction. Such assays measure products formed under specified conditions, so reaction time, acidity, reagent purity, and interfering substances are integral to the method.
Natural waters contain complex mixtures of dissolved organic compounds with broad ultraviolet absorption. Absorbance at 254 nm is frequently used as an indicator related to the aromatic character of dissolved organic matter. Specific ultraviolet absorbance at 254 nm, commonly termed SUVA254, normalizes absorption to dissolved organic carbon concentration under a defined unit convention. It is an operational proxy, not a direct census of aromatic molecular structures. Iron, nitrate, particles, and pH-dependent changes can complicate interpretation in particular samples.
UV–visible detection is extensively coupled to chromatography, especially high-performance liquid chromatography. A flow cell records compounds as they elute, converting absorbance changes into chromatographic peaks. Variable-wavelength detectors monitor selected wavelengths, while array detectors collect spectra across a wavelength range. Spectral comparison can support peak identification and reveal some coeluting impurities, but a spectrally uniform peak does not establish chemical purity: different compounds may have similar spectra or be present below detection limits.
In solids, optical absorption probes electronic band structure and transitions involving defects, impurities, or bound electron–hole states. Measurements near an absorption edge can estimate a band gap, but extracting a numerical value requires a suitable physical model. Plots based on Tauc-type relations depend on assumptions about transition character and the energy dependence of absorption. Scattering, excitonic features, disorder, and arbitrary selection of a fitted interval can all influence the inferred result.
Metal nanoparticles provide another important application through localized surface plasmon resonance, the collective response of conduction electrons to light. Gold nanoparticle dispersions, for example, often show a prominent visible band whose position and width depend on particle size, shape, aggregation, and surrounding medium. Conventional transmission measures extinction, which contains both absorption and scattering. This distinction becomes especially important as particle dimensions increase. Related optical measurements are used in colorant formulation, coating characterization, sensor development, and monitoring of manufacturing processes.
Calibration, uncertainty, and common artifacts
Reliable measurements require separate attention to wavelength accuracy, photometric response, spectral bandwidth, stray light, and baseline stability. Wavelength checks use materials with well-characterized spectral features, such as holmium-containing glass or solutions. Photometric checks may use certified neutral-density filters or reference solutions whose absorbance has been established under defined conditions. The reference material must be suitable for the wavelength, bandwidth, temperature, and geometry of the measurement; a nominal reference value without these conditions is incomplete.
Stray light is especially damaging at high absorbance because the intended transmitted beam has become very weak. If an unwanted signal reaches the detector, a simplified expression for apparent absorbance is
As the true absorbance increases, the measured value approaches a limiting value rather than continuing to rise. An unwanted light fraction of produces a limiting absorbance close to 3 in this simplified model, with significant error appearing before that limit. Such behavior can mimic chemical departures from the Beer–Lambert law.
Other artifacts include fingerprints, bubbles, suspended particles, solvent evaporation, sample decomposition, and mismatched blanks. A sloping baseline may arise from scattering rather than a broad molecular band. High-intensity illumination can cause photochemical changes in sensitive samples, while repeated scans may change sample temperature or composition. A negative absorbance usually means that the sample transmits more light than the chosen reference under the measurement conditions; it does not indicate negative physical absorption.
The measurement uncertainty of a concentration derived from absorbance includes more than short-term instrumental noise. Relevant contributions include standard preparation, volumetric operations, calibration fitting, cell path length, blank variability, and sample stability. At very low absorbance, small baseline errors can dominate relative uncertainty. At very high absorbance, low transmitted power and stray radiation become increasingly important. A suitable working range is therefore determined experimentally rather than by assuming that all displayed digits carry equal information.
Metrological traceability links results to appropriate references through a documented calibration chain, with uncertainties associated with each step. Repeated measurements characterize some random effects but do not expose every systematic bias. Agreement between replicate spectra, a high calibration correlation coefficient, or excellent apparent precision cannot by itself demonstrate chemical selectivity or accuracy. Independent reference materials, recovery studies, and comparisons using different measurement principles provide different kinds of evidence about method performance.
Historical development and specialized extensions
The quantitative foundations preceded modern electronic instruments. Pierre Bouguer examined attenuation of light in the eighteenth century, Johann Heinrich Lambert treated its dependence on optical path length in 1760, and August Beer described concentration-dependent absorption in 1852. The familiar Beer–Lambert relation combines these lines of work. Early absorption measurements depended on visual comparisons and photographic recording, with substantial limitations in sensitivity, speed, and objective intensity measurement.
Photoelectric detection and stable ultraviolet sources transformed the technique during the twentieth century. The Beckman DU spectrophotometer, introduced in 1941, became an influential commercial instrument for ultraviolet and visible measurements. It helped replace many laborious visual and photographic procedures with direct electrical readings. Subsequent developments in double-beam optics, automatic wavelength scanning, digital electronics, and computing made complete spectra and routine quantitative analysis increasingly accessible.
Later array detectors shifted the balance between spectral coverage and measurement speed. Instead of scanning one narrow wavelength interval after another, a dispersive array system can record many wavelengths within one acquisition. Miniaturized optics, fiber coupling, and light-emitting-diode sources enabled portable and process-mounted instruments. These developments broadened access but did not remove the underlying need to control bandwidth, scattered light, sample geometry, and reference measurements.
Specialized extensions probe information that ordinary unpolarized, steady-state absorption does not provide. Circular dichroism measures the difference in absorption between left- and right-circularly polarized light and is sensitive to chiral molecular structure. Linear dichroism measures differential absorption along different polarization directions and can reveal molecular orientation in aligned samples. These are related absorption methods, but their observables are differences between polarization-dependent signals rather than ordinary total absorbance.
Transient absorption spectroscopy uses a pump pulse to initiate excitation and a delayed probe pulse to measure the resulting change in absorption. Signals can include depletion of ground-state absorption, absorption by excited states or intermediates, and stimulated emission. Depending on the apparatus, accessible timescales range from ultrafast electronic motion to much slower chemical transformations. Interpretation requires separating overlapping contributions and accounting for the instrument response. Combining these measurements with electrochemical control, microscopy, or controlled reaction environments allows electronic changes to be connected directly with charge transfer, spatial heterogeneity, and chemical transformation.