Cuvette en quartz pour UV-Vis vs cuvette en verre : transmission, précision et caractéristiques des matériaux

Le choix d'un matériau inadapté pour la cuvette fausse les données spectroscopiques sans qu'on s'en aperçoive — et la plupart des chercheurs ne se rendent compte de l'erreur que lorsque les résultats deviennent inexplicables.

Le matériau de la cuvette n'est pas un aspect secondaire en spectroscopie UV-Vis ; il s'agit d'une variable fondamentale qui détermine directement si les mesures d'absorbance reflètent la composition chimique de l'échantillon ou s'il s'agit d'artefacts liés à l'instrument. Cet article examine les propriétés optiques, structurelles, chimiques et opérationnelles qui font du quartz le matériau de référence pour les mesures UV-Vis, tout en comparant systématiquement ses performances à celles d’autres matériaux (verre, plastique et saphir) sur l’ensemble du spectre.

Le choix du matériau en spectroscopie commence par la compréhension des exigences réelles de l’instrument vis-à-vis du récipient contenant l’échantillon. Lorsqu’un spectrophotomètre UV-Vis balaye des longueurs d’onde comprises entre 190 nm et 800 nm, chaque composant optique situé sur le trajet optique — y compris la cuvette — doit transmettre le rayonnement sans l’absorber, le diffuser ni émettre de fluorescence. Une cuvette qui perturbe le faisceau à n'importe quelle longueur d'onde comprise dans cette plage introduit une erreur systématique qui ne peut être corrigée par le seul post-traitement logiciel.

---

La transparence optique de la fenêtre détermine le choix du matériau de la cuvette

Chaque matériau optique transmet le rayonnement de manière sélective, et ce sont les limites de cette sélectivité qui déterminent si une cuvette est scientifiquement adaptée aux analyses UV-Vis ou si elle est fondamentalement incompatible avec celles-ci.

Qu'est-ce qu'une fenêtre de transmission dans les matériaux optiques ?

Une fenêtre de transmission désigne la gamme de longueurs d'onde dans laquelle un matériau laisse passer le rayonnement électromagnétique sans atténuation significative. L'atténuation résulte de trois phénomènes concurrents : l'absorption, la réflexion et la diffusion, qui contribuent chacune à la perte de signal d'une manière dépendante de la longueur d'onde.

Au niveau atomique, l’absorption se produit lorsque l’énergie du photon incident correspond à la bande interdite entre les états fondamentaux et les états excités des électrons au sein de la structure moléculaire ou cristalline du matériau. Les matériaux présentant une large bande interdite — c’est-à-dire dont l’énergie nécessaire pour exciter les électrons dépasse l’énergie des photons UV — présentent des fenêtres de transmission larges et à courte longueur d’onde. À l’inverse, les matériaux contenant des ions de métaux de transition, des structures organiques conjuguées ou des défauts structurels possèdent des états d’énergie électroniques intermédiaires qui absorbent facilement le rayonnement UV, bloquant ainsi efficacement la transmission en dessous de leurs longueurs d’onde de coupure caractéristiques.

L'implication pratique pour le choix de la cuvette est directe: tout matériau dont le seuil d'absorption se situe dans la plage spectrale de la mesure superposera son propre profil d'absorption au signal de l'échantillon, rendant ainsi impossible une correction précise de la ligne de base.

La gamme spectrale UV-Vis et ses exigences optiques élevées

La gamme spectrale UV-Vis s'étend généralement de de 190 nm à 800 nm, subdivisé en trois régions : l'ultraviolet lointain (190–280 nm), l'ultraviolet proche (280–400 nm) et le visible (400–800 nm). Chaque sous-région impose des exigences spécifiques en matière de transparence du matériau des cuvettes.

La région du visible (400–800 nm) est relativement peu contraignante ; le verre et le quartz transmettent tous deux correctement dans cette gamme, ce qui rend le choix du matériau moins critique pour les dosages colorimétriques réalisés exclusivement dans le visible. La région de l'ultraviolet proche (280–400 nm), en revanche, commence à révéler les limites du verre et de la plupart des polymères, car leurs bords d'absorption s'étendent à partir de 320 nm vers le haut. La région de l'ultraviolet lointain (190–280 nm) est la plus exigeante : Moins de trois matériaux de cuvette commercialement viables — le quartz de silice fondue, le quartz synthétique de qualité UV et le saphir — conservent une transparence suffisante en dessous de 220 nm.

La quantification des protéines à 280 nm, celle des acides nucléiques à 260 nm, l'absorption des liaisons peptidiques à 215 nm et la caractérisation des acides aminés aromatiques dans la plage de 250 à 290 nm se situent toutes dans la région de l'ultraviolet proche ou à proximité de celle-ci. Pour ces applications, qui représentent une part importante de la spectroscopie de routine en laboratoire, La transparence du matériau de la cuvette en dessous de 320 nm est une condition incontournable.

---

Comment les propriétés optiques des cuvettes en quartz améliorent les mesures UV-Vis

La transparence optique à elle seule ne suffit pas à expliquer pleinement pourquoi les mesures UV-Vis tirent un tel avantage de cuvette en quartz construction. Cet avantage s'étend de la structure moléculaire à la qualité de la surface, en passant par la précision de fabrication.

La structure moléculaire de la silice fondue et sa transparence aux UV

La silice fondue — principal matériau utilisé dans la fabrication des cuvettes en quartz pour la spectroscopie UV-Vis — est un solide amorphe composé exclusivement de dioxyde de silicium (SiO₂), organisé en un réseau aléatoire continu de tétraèdres [SiO₄] partageant leurs sommets. Contrairement au quartz cristallin, qui possède un réseau cristallin ordonné à longue portée, La silice fondue ne présente pas de régularité structurelle périodique, ce qui élimine la biréfringence et le rend optiquement isotrope quelle que soit son orientation.

La structure électronique de la liaison Si-O joue un rôle essentiel dans la transparence aux UV de la silice fondue. La bande interdite de la silice fondue de haute pureté est d'environ 8,9 eV, ce qui correspond à un début d'absorption vers 140 nm dans la région de l'ultraviolet sous vide. Les photons à 190 nm possèdent une énergie d’environ 6,5 eV — bien inférieure au seuil requis pour exciter des électrons au-delà de la bande interdite Si-O —, ce qui signifie que le rayonnement UV aux longueurs d’onde pertinentes pour la spectroscopie en laboratoire traverse la silice fondue pure sans absorption électronique. Cela contraste fortement avec les verres à plusieurs composants, dans lesquels les oxydes dopants introduisent des états électroniques situés en dessous de la bande interdite qui absorbent le rayonnement UV à des longueurs d’onde pouvant atteindre 350 nm.

La silice fondue synthétique obtenue par hydrolyse à la flamme ou par dépôt chimique en phase vapeur (CVD) au plasma à partir de SiCl₄ présente des concentrations en groupes hydroxyles (OH) inférieures à 1 ppm et des teneurs en impuretés métalliques inférieures à 20 ppb., ces deux éléments étant essentiels au maintien de la transparence aux UV. Les cristaux de quartz naturels, même après purification, conservent des traces d'impuretés qui introduisent des bandes d'absorption vers 245 nm (associées à des centres pauvres en oxygène) et 214 nm (associées à des centres E'), ce qui les rend moins performants que la silice fondue synthétique pour les applications dans l'ultraviolet profond.

Plage de transmission d'une cuvette en quartz, de l'ultraviolet profond au proche infrarouge

La silice fondue de haute pureté présente une transmission mesurable à partir d'environ de 170 nm à 2 700 nm, couvrant une gamme spectrale qu’aucun autre matériau optique d’un bon rapport qualité-prix ne peut égaler dans son intégralité. Plus précisément dans le domaine UV-Vis (190–800 nm), les valeurs de transmission d’une cuvette en silice fondue de qualité UV d’une longueur de trajet optique de 10 mm dépassent généralement 85% à 200 nm et 92% à 250 nm, les pertes étant principalement dues à la réflexion de Fresnel au niveau des deux interfaces air-verre plutôt qu’à l’absorption dans le volume.

On distingue trois catégories commerciales de cuvettes en silice fondue, qui se différencient principalement par leur teneur en groupes hydroxyles et l'absorption infrarouge qui y est associée. Silice fondue de qualité UV (faible teneur en OH, < 10 ppm OH) offre une transmission optimale en dessous de 250 nm, ce qui en fait le choix idéal pour la spectroscopie dans l'ultraviolet profond. La silice fondue de qualité standard (teneur élevée en OH, 400–1 000 ppm OH) présente une transmission UV légèrement réduite en raison des harmoniques d’absorption de l’OH proches de 245 nm, mais offre des performances acceptables pour la plupart des applications dans l’ultraviolet proche au-dessus de 220 nm. La silice fondue de qualité IR optimise la transmission dans la région de 2 000 à 3 500 nm au détriment d’une partie des performances dans l’UV.

Pour la quantification des acides nucléiques et des protéines — les deux applications UV-Vis les plus courantes dans les laboratoires de biologie —, les cuvettes en silice fondue de qualité UV, d'une longueur de trajet optique de 10 mm, présentent des valeurs d'absorbance de fond inférieures à 0,01 AU à 260 nm et 280 nm., offrant ainsi la marge de mesure nécessaire à une quantification précise sur une large plage de concentrations.

Uniformité de l'indice de réfraction et état de surface des différentes qualités de cuvettes en quartz

L'indice de réfraction de la silice fondue à 589 nm (raie D du sodium) est d'environ 1.458, avec un profil de dispersion qui varie de manière prévisible de 1,534 à 193 nm à 1,440 à 1 064 nm. L'uniformité spatiale de l'indice de réfraction sur toute la fenêtre optique de la cuvette est plus importante que sa valeur absolue : Les ébauches en silice fondue de haute qualité présentent une homogénéité de l'indice de réfraction inférieure à ±5 × 10⁻⁶, ce qui garantit que la distorsion du front d'onde provoquée par les parois de la cuvette est négligeable par rapport à la précision photométrique de l'instrument.

La qualité de la finition de surface détermine directement les pertes par diffusion au niveau des parois de la cuvette. Le polissage de qualité optique de la silice fondue permet d'obtenir des valeurs de rugosité de surface inférieures à 0.5 nm RMS (root mean square), which keeps surface scatter losses below 0.1% across the UV-Vis range. Polishing to this specification requires multi-stage lapping with progressively finer abrasives followed by pitch polishing or magnetorheological finishing — processes that are specific to optical fabrication and distinct from standard laboratory glassware production.

Two-sided polish cuvettes — where only the two faces perpendicular to the light beam are optically finished — are adequate for standard absorbance measurements. Four-sided polish cuvettes, where all four vertical faces carry the same optical finish, are required for fluorescence spectroscopy, circular dichroism (CD)¹, and optical rotation measurements where radiation enters or exits through the lateral faces.

Quartz Cuvette Optical Performance Summary

Paramètres	Silice fondue de qualité UV	Standard Fused Silica	IR-Grade Fused Silica
Transmission Range (nm)	170–2,700	190–2,700	220–3,500
Teneur en OH (ppm)	< 10	400–1,000	< 10
Transmission at 200 nm (%)	> 85	75–85	60–75
Transmission at 260 nm (%)	> 92	> 90	> 88
Refractive Index at 589 nm	1.458	1.458	1.458
Surface Roughness RMS (nm)	< 0.5	< 0.5	< 0.5
Metallic Impurities (ppb)	< 20	< 50	< 20

---

Pathlength Selection in UV-Vis Quartz Cuvettes and Its Effect on Measurement Accuracy

Beyond material transparency, the pathlength of a UV-Vis quartz cuvette is the single most consequential geometric parameter affecting both the accuracy and the linear dynamic range of absorbance measurements.

Beer-Lambert Law and the Linear Dependence on Pathlength

The Beer-Lambert law expresses the fundamental relationship between absorbance, sample concentration, and optical pathlength: A = ε × c × l, where A is absorbance (dimensionless), ε is the molar attenuation coefficient (L mol⁻¹ cm⁻¹), c is the molar concentration (mol L⁻¹), and l is the pathlength (cm). The law predicts a strictly linear relationship between absorbance and concentration at fixed pathlength, and between absorbance and pathlength at fixed concentration — but this linearity holds only within a defined range.

Deviations from Beer-Lambert linearity become significant when absorbance values exceed approximately 1.5 AU (corresponding to transmittance below 3.2%). At high absorbance, stray light within the spectrophotometer — radiation that reaches the detector without passing through the sample — constitutes a proportionally larger fraction of the detected signal, causing apparent absorbance to plateau below its true value. Additionally, at high solute concentrations, intermolecular interactions between absorbing species alter the effective molar attenuation coefficient, introducing chemical deviations from linearity. Both effects systematically underestimate true concentration, with errors reaching 5–15% at absorbance values above 2.0 AU.

Reducing pathlength is a physically precise method of restoring linearity for concentrated samples. Halving the pathlength from 10 mm to 5 mm halves the measured absorbance at constant concentration, returning measurements to the linear range without requiring sample dilution — a critical advantage when sample volumes are limited or when dilution would alter solution equilibria.

Standard Pathlengths across Quartz Cuvettes and Their Corresponding Applications

Manufacturers produce UV-Vis quartz cuvettes across a pathlength range spanning approximately trois ordres de grandeur, from 0.1 mm to 100 mm, to accommodate the full diversity of sample concentrations encountered in analytical practice.

Short pathlength cuvettes — 0.1 mm, 0.2 mm, and 0.5 mm — are used for high-concentration samples such as undiluted protein stocks, concentrated dye solutions, and pharmaceutical formulations at process concentrations. At a 0.1 mm pathlength, a sample with a molar attenuation coefficient of 10,000 L mol⁻¹ cm⁻¹ remains within the linear range up to concentrations of approximately 150 mg/mL for a typical globular protein — a range inaccessible with standard 10 mm cuvettes. The 10 mm pathlength cuvette is the universal standard for routine UV-Vis measurements, providing a practical working absorbance range of approximately 0.05–1.5 AU for most laboratory samples. Long pathlength cuvettes of 20 mm, 50 mm, and 100 mm extend sensitivity to trace-concentration samples, including environmental water samples analyzed for aromatic contaminants, ultra-dilute pharmaceutical impurity standards, and low-concentration chromophoric species in ecological research.

Micro and Sub-Micro Quartz Cuvettes for UV-Vis Analysis of Limited Volumes

Sample volume constraints, particularly prevalent in genomics, proteomics, and single-cell biology, have driven the development of micro and sub-micro quartz cuvette formats that maintain full optical performance while requiring substantially smaller volumes.

A standard 10 mm pathlength quartz cuvette requires approximately 3.0–3.5 mL of sample to fill the optical chamber above the beam path. Semi-micro cuvettes reduce this requirement to 1.4–1.7 mL by narrowing the internal chamber width while preserving the 10 mm pathlength. Micro cuvettes further reduce volume requirements to 0.6–0.7 mL, and sub-micro formats achieve usable measurements with as little as 70 µL, achieved by designing an extremely narrow internal chamber (typically 3 mm × 3 mm cross-section) with a precisely defined Z-dimension — the height of the beam center above the cuvette base — matched to the spectrophotometer's optical geometry.

The Z-dimension is a critical compatibility parameter: most bench-top UV-Vis spectrophotometers position the beam center at 8,5 mm above the cuvette base, while some instruments use 15 mm or 20 mm Z-dimensions. A mismatch between cuvette Z-dimension and instrument beam height results in the beam partially missing the sample volume, producing anomalously low absorbance readings that are indistinguishable from genuine low-concentration signals without independent verification.

Quartz Cuvette Pathlength and Volume Reference

Pathlength (mm)	Typical Sample Volume (mL)	Application primaire
0.1	0.05-0.15	Highly concentrated protein stocks, API solutions
0.5	0.10–0.30	Concentrated biological extracts
1	0.40–0.70	Intermediate concentration samples
10	3.00–3.50	Universal routine UV-Vis measurements
20	6.00–7.00	Dilute environmental samples
50	15.0–17.0	Trace contaminant analysis
100	30.0–35.0	Ultra-trace concentration measurements

---

Geometric Design Variations in Quartz Cuvettes for UV-Vis Experimental Setups

Pathlength and volume specifications address what happens along the primary optical axis, yet the geometry of the cuvette body itself introduces additional variables that affect measurement quality in specific experimental configurations.

Four-Sided Polish versus Two-Sided Polish Cuvettes

Two-sided polish cuvettes present optically flat, scratch-free surfaces exclusively on the two faces perpendicular to the excitation beam — the faces that the measurement radiation actually traverses. The remaining two lateral faces are ground to a matte finish, which is sufficient for mechanical stability and handling but optically unsuitable for transmitting radiation. Two-sided polish cuvettes are entirely adequate for standard absorbance measurements where radiation enters through one polished face and exits through the opposite polished face along a single linear optical axis.

Four-sided polish cuvettes carry the same optical finish on all four vertical faces, allowing radiation to enter or exit through any face without scatter losses. This configuration is indispensable for spectroscopie de fluorescence, where excitation radiation enters through one face and emitted fluorescence is collected at 90° through an adjacent face. It is equally required for circular dichroism (CD) spectroscopy, where the interaction geometry depends on precise polarization control through optically homogeneous surfaces, and for optical rotation measurements. The additional fabrication cost of polishing four faces rather than two — typically representing a 30–50% premium over equivalent two-sided polish specifications — is justified only when the measurement geometry requires lateral optical access.

A practical identification method: two-sided polish cuvettes typically exhibit a visible opacity or frosted appearance on their lateral walls when viewed under oblique lighting, whereas four-sided polish cuvettes appear uniformly clear from all angles.

Black-Walled Quartz Cuvettes and Stray Light Suppression in Fluorescence Measurements

Standard four-sided polish cuvettes, while optically accessible from all directions, introduce a specific artifact in fluorescence measurements: reflected excitation radiation from the internal cuvette walls reaches the emission detector, superimposing a background signal on the genuine fluorescence spectrum. This artifact is particularly severe at emission wavelengths close to the excitation wavelength — in the region of the Raman scatter peak of water and in the initial portion of the fluorescence emission band.

Black-walled quartz cuvettes address this artifact by applying an opaque black coating to the two lateral faces and the back face, leaving only the front face (excitation entry) and the 90° emission face transparent and polished. The black coating absorbs reflected and scattered excitation radiation before it can reach the emission detector, reducing stray light background by factors of 10–100× compared to standard four-sided polish cuvettes in fluorescence experiments. The quartz body of a black-walled cuvette remains UV-grade fused silica; only the external surface coating differs.

The practical consequence of using a standard clear-walled cuvette instead of a black-walled cuvette in fluorescence spectroscopy is an elevated and spectrally structured background signal that reduces sensitivity, distorts emission spectra at wavelengths within approximately 30–50 nm of the excitation wavelength, and compromises quantitative accuracy for weakly fluorescing samples.

Quartz Cuvette Design Variants and Their Applications

Type de conception	Visages polis	Stray Light Control	Application primaire
Two-sided polish	2 (front and back)	Standard	Absorbance UV-Vis
Four-sided polish	4 (all vertical)	Standard	Fluorescence, CD spectroscopy
Black-walled	2 (front + emission face)	Améliorée	Fluorescence with low background
Flow-through	2 or 4	Standard	HPLC detectors, continuous flow
Cylindrical	Continuous curve	Limitée	Specialized circular dichroism

---

Chemical Resistance Profiles of UV-Vis Quartz Cuvettes across Sample Matrices

The spectroscopic superiority of fused silica quartz is accompanied by an equally important advantage in chemical durability — one that determines the range of samples a cuvette can safely accommodate without degradation of its optical surfaces.

Acid and Organic Solvent Compatibility of Fused Silica

Fused silica's chemical resistance to acids and organic solvents derives directly from the thermodynamic stability of the Si-O bond network. The Si-O bond dissociation energy of approximately 452 kJ/mol exceeds that of most metal-oxygen bonds in multicomponent glasses, which explains why fused silica withstands reagents that readily attack conventional laboratory glassware.

Strong mineral acids — including hydrochloric acid at all concentrations, sulfuric acid up to approximately 70% concentration, nitric acid, and phosphoric acid at ambient temperature — do not measurably attack fused silica surfaces over typical laboratory exposure durations. Aqueous acidic buffers used routinely in biochemistry (citrate, acetate, phosphate, and MES buffers at pH 3–6) are similarly innocuous. Organic solvents represent an equally broad compatibility range: ethanol, methanol, isopropanol, acetone, acetonitrile, dimethyl sulfoxide (DMSO), dichloromethane, chloroform, toluene, and tetrahydrofuran (THF) are all compatible with fused silica quartz cuvettes without inducing swelling, leaching, or optical surface degradation.

A critical practical distinction separates fused silica quartz from polymeric cuvette materials: organic solvents that cause immediate PMMA or polystyrene cuvettes to craze, cloud, or dissolve — particularly chlorinated solvents, ketones, and aromatic hydrocarbons — are handled without consequence by quartz, making fused silica quartz cuvettes the only practical choice for UV-Vis measurements in non-aqueous solvent systems.

The Critical Exception of Hydrofluoric Acid and Concentrated Alkalis

Despite its broad chemical compatibility, fused silica quartz has two well-defined chemical vulnerabilities that must be observed without exception in laboratory practice.

Hydrofluoric acid (HF) reacts with SiO₂ through the stoichiometric reaction SiO₂ + 4HF → SiF₄↑ + 2H₂O, producing volatile silicon tetrafluoride² and water. This reaction proceeds at measurable rates even at HF concentrations as low as 0.1% and at room temperature, etching the polished optical surfaces of quartz cuvettes within minutes of contact. Surface etching introduces permanent scatter losses and wavefront distortion that cannot be removed by cleaning; a quartz cuvette exposed to HF must be treated as irreparably damaged and removed from service. Concentrated sodium hydroxide (NaOH > 30%) and potassium hydroxide (KOH > 20%) attack fused silica through a slower dissolution mechanism — hydroxide ions break Si-O-Si linkages hydrolytically, with dissolution rates of approximately 0.1–1 µm per hour at room temperature in concentrated alkali. Concentrated phosphoric acid at temperatures above 100°C also attacks fused silica at rates that increase sharply with temperature.

For samples requiring HF, concentrated alkali, or hot concentrated phosphoric acid, the appropriate alternative vessels are polytetrafluoroethylene (PTFE) or perfluoroalkoxy (PFA) containers, neither of which is reactive with these reagents.

Chemical Compatibility of Fused Silica Quartz Cuvettes

Reagent Class	Exemple	Compatibilité	Notes
Dilute mineral acids	HCl, H₂SO₄, HNO₃	✓ Compatible	All concentrations, ambient temperature
Solvants organiques	EtOH, Acetone, DMSO, CHCl₃	✓ Compatible	No swelling or optical degradation
Aqueous buffers (pH 3–9)	PBS, HEPES, citrate	✓ Compatible	Standard biological pH range
Dilute alkalis (< 5%)	NaOH, KOH	⚠ Caution	Slow attack; minimize exposure time
Concentrated alkalis	NaOH > 30%	✗ Incompatible	Surface dissolution within hours
Acide fluorhydrique	HF (toute concentration)	✗ Incompatible	Immediate irreversible etching
Hot concentrated H₃PO₄	> 85%, > 100 °C	✗ Incompatible	Thermal acceleration of SiO₂ attack

---

Why Glass Cuvettes Fall Short in UV-Vis Measurements Below 320 nm

Glass cuvettes remain ubiquitous in teaching laboratories and visible-range colorimetry, yet their spectroscopic limitations in the ultraviolet region are severe enough to render them scientifically unsuitable for a large proportion of analytical applications.

The Chemical Composition of Borosilicate and Soda-Lime Glass

Commercial laboratory glass cuvettes are manufactured from either borosilicate glass or soda-lime glass — two multicomponent silicate compositions that include substantial proportions of non-silica network formers and modifiers.

Verre borosilicaté (exemplified by Schott DURAN and Corning Pyrex) contains approximately 81% SiO₂, 13% B₂O₃, 4% Na₂O, and 2% Al₂O₃ by weight. Boron trioxide (B₂O₃) is incorporated to reduce the thermal expansion coefficient, improving thermal shock resistance, but boron-oxygen structural units introduce electronic transitions in the UV range absent from pure SiO₂. Verre sodocalcique contains typically 72% SiO₂, 14% Na₂O, 9% CaO, and 5% MgO — a composition optimized for workability and cost rather than optical performance. The network modifier oxides (Na₂O, CaO, MgO) disrupt the Si-O network, creating non-bridging oxygen sites that form UV-absorbing defect centers.

Trace metallic impurities in both glass types — particularly Fe²⁺ and Fe³⁺ ions present at concentrations of 50–200 ppm in standard optical glass — generate intense UV absorption bands. Fe³⁺ produces absorption bands near 225 nm and 320 nm through ligand-field d-d transitions³ and charge-transfer transitions, while Fe²⁺ contributes absorption near 200 nm. Even at sub-100 ppm concentrations across a 10 mm pathlength, these iron absorption bands produce absorbance contributions of 0.1–0.5 AU at 280 nm — magnitudes that overwhelm the signals of dilute biological samples.

The UV Absorption Cutoff of Glass and Its Spectroscopic Consequences

The ultraviolet absorption cutoff of a cuvette material is conventionally defined as the wavelength at which the material's intrinsic absorbance equals 1.0 AU for a 10 mm pathlength — the point at which only 10% of incident radiation is transmitted and signal-to-noise ratio has degraded to a level that renders quantitative measurement unreliable.

For borosilicate glass, this cutoff wavelength falls between 290 and 320 nm depending on specific glass composition and iron content. For soda-lime glass, the cutoff is typically 320–350 nm. These cutoffs mean that borosilicate glass cuvettes are unusable for measurements at 280 nm (protein quantification by aromatic absorbance), 260 nm (nucleic acid quantification), 254 nm (UV sterilization monitoring), and 214–220 nm (peptide bond absorbance and low-wavelength protein quantification). When a glass cuvette is used for a 280 nm protein assay, the glass itself contributes an absorbance of approximately 0.3–0.8 AU that varies between individual cuvettes of nominally identical specification — a batch-variable systematic error that cannot be corrected by a single blank measurement.

The spectroscopic consequences cascade: baseline drift caused by temperature-dependent glass absorbance shifts, apparent absorption peaks at wavelengths corresponding to glass absorption edges (which can be misattributed to sample chromophores), and a collapsed linear dynamic range that forces all quantification into the upper, nonlinear portion of the Beer-Lambert curve.

Autofluorescence and Scattering Artifacts in Glass at UV Wavelengths

Beyond direct absorption, glass cuvettes generate secondary optical artifacts in the UV range that further compromise measurement accuracy.

Glass autofluorescence — spontaneous emission of visible radiation following UV excitation — arises from electronic transitions within structural defect centers and organic contaminants incorporated during glass manufacturing. When borosilicate glass is irradiated at 280 nm, it emits broadband fluorescence peaking near 400–450 nm, with quantum yields that vary between glass batches. In a standard single-beam UV-Vis spectrophotometer, this fluorescence contributes to the detected signal at wavelengths where the monochromator passband overlaps with the emission spectrum, generating an apparent reduction in sample absorbance — an artifact that scales nonlinearly with excitation intensity and is absent from blank measurements made with solvent alone in a quartz cuvette.

Microscopic inclusions in glass — trapped gas bubbles, unmixed melt regions, and crystallite precipitates — act as Mie scattering centers for UV radiation. Mie scattering from spherical particles with diameters comparable to the wavelength of measurement (100–300 nm for UV radiation) generates a wavelength-dependent background that rises steeply toward shorter wavelengths, mimicking the absorption profile of colloidal particles. In practice, a glass cuvette used for measurements at 220 nm may exhibit an apparent absorbance contribution from scattering that exceeds 0.5 AU — larger than the genuine sample absorbance for many dilute biological samples.

---

Transmission Performance Compared across Quartz Glass Plastic and Sapphire Cuvettes

Selecting the appropriate cuvette material requires a systematic comparison across all four material classes available to laboratory spectroscopists — fused silica quartz, borosilicate glass, polymethylmethacrylate (PMMA), and sapphire — evaluated against the specific demands of UV-Vis measurements.

Fused silica quartz represents the broadest practical transmission window, the highest chemical compatibility, and the best optical surface stability of the four materials, at a correspondingly higher unit fabrication cost. Its transmission from 170 nm to 2,700 nm with < 0.01 AU baseline absorbance at 260 nm makes it the reference material against which all others are benchmarked.

Verre borosilicaté achieves comparable transmission to quartz above 320 nm and in the full visible range (400–800 nm), making it suitable — and cost-effective — for colorimetric assays, enzyme kinetics monitored at visible wavelengths, and any measurement that does not require UV access below 320 nm. Its UV cutoff near 290–320 nm and susceptibility to autofluorescence under UV irradiation make it inappropriate for the near-UV region.

PMMA and polystyrene plastic cuvettes are single-use, disposable formats with UV cutoff wavelengths of 300–320 nm for PMMA et 340–360 nm for polystyrene — limitations that restrict them to visible-range colorimetry. Their principal advantages are price and convenience: they eliminate cross-contamination concerns in clinical and high-throughput environments where disposable protocols are mandated. Organic solvents dissolve or craze plastic cuvettes immediately, and their non-optical-grade surfaces exhibit substantial scatter. Saphir (Al₂O₃) cuvettes transmit from approximately 145 nm to 5,500 nm with exceptional chemical resistance and mechanical hardness (Mohs 9), making them technically superior to fused silica for vacuum UV applications below 160 nm. However, sapphire's birefringence — arising from its trigonal crystal structure — complicates polarimetric measurements, and its fabrication difficulty restricts its use to specialized research applications.

Cuvette Material Transmission and Compatibility Comparison

Propriété	Fused Silica Quartz	Verre borosilicaté	Plastique PMMA	Saphir
Transmission Range (nm)	170–2,700	320–2,500	300–900	145–5,500
UV Cutoff at 1.0 AU / 10 mm (nm)	< 175	290–320	300–320	< 150
Usable below 260 nm	✓ Yes	✗ No	✗ No	✓ Yes
Résistance aux solvants organiques	Excellent	Bon	Pauvre	Excellent
Résistance HF	✗ No	✗ No	✓ Yes	✓ Yes
Autofluorescence under UV	Négligeable	Important	Modéré	Négligeable
Surface Hardness (Mohs)	7	6-7	3	9
Réutilisable	✓ Yes	✓ Yes	✗ No	✓ Yes
Birefringence	Aucun	Aucun	Aucun	Présent

---

Common Measurement Errors Traced Back to Cuvette Material or Condition

Even when the correct cuvette material has been selected, measurement errors attributable to cuvette condition — scratches, residual contamination, misorientation, and trapped bubbles — represent a persistent source of data quality problems in UV-Vis spectroscopy.

• Baseline drift and non-zero blank absorbance are the most frequently encountered cuvette-related artifacts. When a quartz cuvette that has accumulated surface contamination is used for blank measurement, the recorded baseline incorporates the contaminant absorbance as zero, causing all subsequent sample measurements to underreport true absorbance by the same amount. A protein film on a quartz cuvette optical surface can contribute 0.05–0.2 AU of apparent absorbance at 280 nm — an error sufficient to misestimate protein concentration by 10–50% in a standard Bradford or direct UV assay. By contrast, temperature-induced baseline drift arises from the sample rather than the cuvette: the refractive index of aqueous solutions changes by approximately −0.0001 per °C, shifting the Fresnel reflection losses at cuvette interfaces and producing a slow absorbance drift that distinguishes itself by its reversibility upon temperature stabilization.

• Anomalously elevated absorbance readings that do not correspond to known sample concentrations often originate from scratch-induced scatter on the cuvette optical surface rather than elevated sample absorption. A single scratch across the beam cross-section can increase apparent absorbance by 0.05–0.5 AU, depending on scratch depth and width, with the scatter contribution rising steeply at shorter wavelengths. Differentiating scratch scatter from genuine sample absorption requires measuring the apparently anomalous cuvette against a clean reference cuvette using the same blank solution; scratch scatter will remain as a persistent baseline offset whereas genuine sample absorption varies with sample concentration.

• Poor measurement reproducibility — coefficient of variation exceeding 1–2% across replicate measurements of identical samples — frequently traces to inconsistent cuvette insertion orientation. Most cuvettes are not perfectly square: wall thickness variations of ±0,01-0,05 mm between opposing faces alter the effective pathlength depending on which face is presented to the beam. Establishing a consistent insertion orientation (marked with a laboratory pen or by alignment with a manufacturer's orientation mark) typically reduces orientation-related absorbance variability to below 0.3%.

• Bubble artifacts produce sudden, large absorbance spikes — often exceeding 1.0 AU — at otherwise well-behaved wavelengths. A bubble spanning even a fraction of the beam cross-section reflects virtually all incident radiation away from the detector, simulating near-complete sample absorption. Bubbles originate from dissolved gas coming out of solution when samples are transferred to room temperature from cold storage, from turbulent sample introduction through narrow-bore pipettes, and from residual rinse solvent trapped in poorly dried cuvettes. Gentle warming to room temperature before measurement, slow sample introduction along the cuvette wall rather than directly into the beam, and thorough drying between uses reliably prevent bubble formation.

---

Verifying the Integrity of a UV-Vis Quartz Cuvette before Each Measurement Run

Establishing a brief verification routine before committing a UV-Vis quartz cuvette to quantitative measurements prevents the accumulation of uncorrected systematic errors across experimental datasets.

• Baseline transmission verification is the most informative pre-measurement check. Filling the cuvette with HPLC-grade water (or the neat solvent to be used in the experiment) and scanning from 190 nm to 350 nm against an air reference reveals both residual contamination (elevated absorbance at characteristic wavelengths) and surface scatter (elevated baseline that rises uniformly toward shorter wavelengths). A clean UV-grade fused silica quartz cuvette filled with HPLC-grade water should exhibit absorbance below 0.05 AU at 200 nm, below 0.02 AU at 230 nm, and below 0.01 AU at 260 nm against an air blank under standard spectrophotometer conditions. Deviations above these thresholds indicate either residual contamination (requiring additional cleaning) or optical surface damage (requiring cuvette replacement).

• Visual inspection under oblique illumination complements the spectrophotometric baseline check by revealing scratch patterns, cloudiness, and mechanical chips that cause scatter without necessarily producing distinctive spectral absorption features. Holding the cuvette at approximately 30° to a fluorescent tube or fiber-optic light source and examining the optical faces in transmitted light reveals scratches as bright linear streaks; cloudiness appears as diffuse glow within the glass body; mechanical chips appear as sharp-edged bright regions at the cuvette corners or edges. Any cuvette exhibiting scratches within the central 80% of the optical face — the region traversed by the spectrophotometer beam — should be removed from service for quantitative measurements.

• Matched-pair verification is required when dual-beam spectrophotometers are used with separate sample and reference cuvettes. Filling both cuvettes with identical blank solution and measuring the absorbance of one against the other across 200–400 nm quantifies their photometric equivalence. A matched pair should exhibit absorbance differences below 0.005 AU across the full wavelength range; pairs exceeding 0.5% transmittance difference at any wavelength within the measurement range should be re-matched or replaced, as the mismatch introduces a wavelength-dependent baseline error that cannot be zeroed out by a single blank measurement.

• Replacement criteria for UV-Vis quartz cuvettes are determined by optical performance rather than chronological age or number of uses. A cuvette that passes the baseline transmission test and visual inspection continues to deliver reliable measurements regardless of how long it has been in service. Conversely, a cuvette that fails the baseline test despite thorough cleaning — exhibiting persistent elevated absorbance above 0.05 AU at 260 nm in HPLC-grade solvent — has sustained permanent optical surface degradation and should be retired from quantitative UV-Vis work.

---

Conclusion

Material selection in UV-Vis cuvettes is a decision with direct consequences for data integrity across the full spectral range. Fused silica quartz stands apart from glass, plastic, and most competing materials because its molecular structure — a continuous SiO₂ network with an electronic bandgap of 8.9 eV — transmits from 170 nm to 2,700 nm without absorption, autofluorescence, or surface degradation in the presence of acids and organic solvents. Glass cuvettes fail below 320 nm due to transition metal impurities, structural defects, and multicomponent oxide compositions that introduce UV absorption, baseline drift, and fluorescence artifacts. Proper cuvette selection, matched to pathlength, geometry, and cleaning requirements, is not peripheral to UV-Vis spectroscopy — it is the physical foundation upon which every quantitative result rests.

---

FAQ

Can a glass cuvette be used for any UV-Vis measurements?
Glass cuvettes are usable for measurements conducted entirely above 320 nm — visible-range colorimetry, enzyme kinetics assays monitored at 400–800 nm, and absorbance-based turbidity measurements. They are not suitable for any measurement requiring wavelength access below 320 nm, including protein quantification at 280 nm, nucleic acid quantification at 260 nm, or any assay dependent on aromatic or peptide bond absorbance in the near-UV region.

What pathlength quartz cuvette is standard for most UV-Vis applications?
The 10 mm pathlength quartz cuvette is the universal standard because it provides a practical absorbance working range of approximately 0.05–1.5 AU for sample concentrations typical of most biological and chemical analyses, corresponds directly to the pathlength assumed in tabulated molar attenuation coefficient values (which are conventionally reported in units of L mol⁻¹ cm⁻¹, where 1 cm = 10 mm), and is compatible with the optical geometry of virtually all commercial bench-top spectrophotometers.

How often should a quartz cuvette be replaced?
Replacement frequency is determined by optical performance, not calendar time. A quartz cuvette that passes a baseline transmission test — exhibiting less than 0.05 AU at 260 nm in HPLC-grade solvent — and shows no scratches within the central optical face area may remain in service indefinitely. Replacement is indicated when persistent elevated baseline absorbance above this threshold survives thorough cleaning, confirming irreversible surface damage.

Is there any cuvette material that outperforms quartz for UV-Vis work?
Sapphire (Al₂O₃) exhibits a broader transmission window than fused silica, extending from approximately 145 nm in the vacuum UV to 5,500 nm in the mid-infrared. For laboratory UV-Vis applications confined to 190–800 nm, however, fused silica quartz performs equivalently to sapphire while avoiding sapphire's inherent birefringence — a property that complicates polarimetric and circular dichroism measurements — making UV-grade fused silica quartz the practical optimum for the vast majority of UV-Vis spectroscopic applications.

---

Références :