Optical Materials

Every optical system begins with a material choice. The substrate determines not only how light propagates through the component — its speed, direction, and spectral content — but also how the component survives its operating environment: temperature swings, mechanical loads, chemical exposure, and laser irradiation. A lens ground to exacting tolerances from the wrong glass will underperform a modest optic made from the right one [1, 2].

1.1The Role of Substrate Materials

Optical material selection is a multi-dimensional optimization. Transmission range defines which wavelengths pass through the material without significant absorption; refractive index and dispersion govern how the material bends and separates those wavelengths; mechanical hardness affects fabrication options and coating adhesion; thermal expansion and the thermo-optic coefficient determine how stable the optical path remains as temperature changes; and chemical durability sets the component's useful lifetime in a given environment [2, 3].

No single material excels in all categories. N-BK7, the most widely used optical glass, offers excellent visible transmission, low cost, and good fabrication properties — but it absorbs strongly below 300 nm, ruling it out for ultraviolet applications. Calcium fluoride extends transmission deep into the UV and far into the infrared, but it is thermally fragile and moderately hygroscopic. Germanium transmits across the thermal infrared atmospheric windows but is completely opaque in the visible. These trade-offs make material selection one of the first and most consequential decisions in any optical design [3, 4].

1.2Historical Context

The systematic development of optical glass began in 1884 when Otto Schott, in collaboration with Ernst Abbe and Carl Zeiss, established the Schott Glass Works in Jena, Germany. Before Schott, lens designers were limited to a handful of naturally occurring glass compositions — the crown and flint types known since the 18th century. Schott's methodical exploration of glass chemistry expanded the designer's palette from a few dozen glasses to hundreds, each with precisely characterized optical and mechanical properties [1, 5].

The 20th century brought synthetic crystal growth (Czochralski, Bridgman, and Stockbarger methods), enabling large-format single crystals of calcium fluoride, sapphire, and dozens of other materials with optical qualities unachievable in naturally occurring specimens. The development of chemical vapor deposition (CVD) produced zinc selenide and zinc sulfide with the purity and homogeneity required for infrared optical systems. More recently, precision molding of chalcogenide glasses and high-volume polymer optics have expanded the material toolkit into cost-sensitive, high-volume applications [4, 7].

Optical materials fall into three broad families: glasses, crystals, and polymers. Each family possesses fundamental structural differences that determine its optical behavior, mechanical properties, and manufacturing constraints [1, 2, 4].

2.1Optical Glasses

Optical glasses are amorphous solids — they lack the long-range atomic order of crystals. This amorphous structure makes glasses inherently isotropic: their refractive index is the same in every direction, and they exhibit no birefringence (in the stress-free state). Isotropy simplifies optical design considerably because the designer need not account for crystallographic orientation [1, 5].

The two traditional categories of optical glass are crowns and flints. Crown glasses (designated “K” from the German “Kron”) exhibit lower refractive index and lower dispersion, characterized by Abbe numbers above approximately 55 (or above 50 when the refractive index exceeds 1.60). Flint glasses (designated “F”) have higher refractive index and higher dispersion, with Abbe numbers below these thresholds. This classification, formalized in the Schott glass catalogue, reflects the underlying chemistry: crown glasses are based on alkali-silicate and borosilicate compositions, while flint glasses historically contained lead oxide to increase the refractive index. Modern “N-” prefix glasses (e.g., N-BK7, N-SF11) achieve equivalent optical properties without lead or arsenic [5, 6].

Beyond the crown-flint division, optical glasses are further sub-classified by composition and position on the glass map. Common families include BK (borosilicate crown), SK (dense barium crown), BAK (barium crown), LAK (lanthanum crown), F (flint), SF (dense flint), and LASF (lanthanum dense flint). Each family occupies a characteristic region of the Abbe diagram, and designers select glass combinations from separated regions to correct chromatic aberration [1, 5, 6].

2.2Optical Crystals

Optical crystals possess long-range atomic periodicity — atoms arranged in a repeating lattice. This ordered structure produces several properties absent in glasses: birefringence (in non-cubic crystals), electro-optic and acousto-optic response, and nonlinear optical effects. Crystalline materials also tend to exhibit sharper transmission edges, higher thermal conductivity, and better laser damage resistance than glasses of comparable composition [1, 4, 8].

Crystals used in photonics fall into several chemical families. Fluoride crystals — calcium fluoride (CaF₂), magnesium fluoride (MgF₂), barium fluoride (BaF₂), and lithium fluoride (LiF) — transmit from the deep ultraviolet well into the infrared. Oxide crystals such as sapphire (Al₂O₃) and crystalline quartz (SiO₂) provide extreme hardness and broad transmission. Semiconductor crystals including silicon (Si), germanium (Ge), zinc selenide (ZnSe), and zinc sulfide (ZnS) serve the mid-wave and long-wave infrared regions where silicate glasses are opaque [4, 7, 8].

The cubic crystal system (CaF₂, ZnSe, ZnS, diamond) produces optically isotropic materials — no birefringence — making them the easiest crystals to incorporate into conventional optical designs. Non-cubic crystals such as sapphire (trigonal/rhombohedral), quartz (trigonal), and MgF₂ (tetragonal) are birefringent; their refractive index depends on the polarization direction and the orientation of the crystal axes relative to the optical beam [1, 4].

2.3Polymers and Specialty Materials

Polymer optics — made from materials such as PMMA (acrylic), polycarbonate, cyclic olefin polymer (COP), and polystyrene — dominate high-volume, low-cost applications. Injection molding produces complex aspheric and freeform shapes at a fraction of the cost of ground-and-polished glass. The trade-offs are lower scratch resistance, higher thermal expansion, limited operating temperature range, and narrower transmission windows compared to glass [3, 4].

Chalcogenide glasses — compositions based on sulfur, selenium, and tellurium with elements such as germanium and arsenic — fill a unique niche. They are amorphous (moldable like glass) but transmit across the full thermal infrared from 2 to 20 μm. This combination enables precision-molded infrared optics at significantly lower cost than diamond-turned crystalline alternatives [4, 7].

Glass-ceramics (e.g., Zerodur, Clearceram) occupy another specialty role. These materials are partially crystallized glasses with near-zero thermal expansion coefficients, making them the substrate of choice for optical mirror blanks in applications where thermal dimensional stability is paramount — telescope primary mirrors, ring laser gyroscopes, and lithographic stages [3, 4].

The refractive index — the ratio of the speed of light in vacuum to its speed in the material — is the single most important property for optical design. It governs refraction at surfaces (Snell's law), reflection losses (Fresnel equations), and the optical path length through components. Because refractive index varies with wavelength (dispersion), controlling and correcting chromatic effects is central to lens design [1, 2].

3.1Refractive Index Fundamentals

The refractive index of a material at a specific wavelength is defined as:

Where $n(\lambda)$ is the refractive index at wavelength $\lambda$ (dimensionless), $c$ is the speed of light in vacuum (2.998 × 10⁸ m/s), and $v(\lambda)$ is the phase velocity of light in the material at wavelength $\lambda$ (m/s).

For optical glasses, the principal refractive index $n_d$ is specified at the helium d-line (587.56 nm). Values range from about 1.44 (fluorite crown glasses) to above 1.95 (dense lanthanum flint glasses). Infrared materials exhibit much higher indices: silicon at approximately 3.42 and germanium at approximately 4.00 at their operating wavelengths [1, 5, 8].

At normal incidence, the fraction of light reflected at an interface between two media follows from the Fresnel equations:

Where $R$ is the power reflectance at normal incidence (dimensionless), $n_1$ is the refractive index of the incident medium, and $n_2$ is the refractive index of the transmitting medium.

For an uncoated N-BK7 surface in air ($n_1 = 1$, $n_2 = 1.517$), the reflectance per surface is about 4.2%. For germanium in air ($n_2 = 4.0$), the reflectance per surface is approximately 36% — making anti-reflection coatings essential for any practical germanium optic [1, 2, 9].

3.2The Sellmeier Equation

The wavelength dependence of refractive index in a transparent material is accurately described by the Sellmeier equation, an empirical relationship derived from a damped-oscillator model of light-matter interaction. The standard three-term form used for optical glasses is:

Where $n(\lambda)$ is the refractive index at vacuum wavelength $\lambda$, $\lambda$ is the vacuum wavelength (μm), $B_1, B_2, B_3$ are oscillator strengths (dimensionless), and $C_1, C_2, C_3$ are resonance wavelengths squared (μm²).

Each term corresponds to an absorption resonance in the material: for typical optical glasses, $B_1$ and $B_2$ represent ultraviolet electronic transitions, while $B_3$ represents an infrared lattice absorption. The Sellmeier equation is valid from the near-UV through the near-IR (approximately 365 nm to 2.3 μm for most glasses), with residual errors below 5 × 10⁻⁶ in refractive index — comparable to the homogeneity of the glass itself [5, 6, 10].

3.3The Abbe Number and the Glass Map

The Abbe number (also called the V-number or constringence) is a single-number measure of a material's dispersion across the visible spectrum:

Where $V_d$ is the Abbe number (dimensionless), $n_d$ is the refractive index at the helium d-line (587.56 nm), $n_F$ is the refractive index at the hydrogen F-line (486.13 nm), and $n_C$ is the refractive index at the hydrogen C-line (656.27 nm).

A high Abbe number indicates low dispersion; a low Abbe number indicates high dispersion. Values range from below 25 for very dense flint glasses, through about 34 for polycarbonate, to above 80 for some fluorite and phosphate crown glasses. The denominator ($n_F - n_C$) is the principal dispersion — the total refractive index change across the visible spectrum [1, 5, 6].

The Abbe diagram (or glass map) plots n_d on the vertical axis against V_d on the horizontal axis (with V_d increasing to the left, by convention). Every glass type from a manufacturer's catalog occupies a specific point on this diagram. The vertical dashed boundary separating crown from flint glasses starts at V_d = 55 for n_d < 1.60 and shifts to V_d = 50 for n_d ≥ 1.60 [5, 6].

Achromatic doublet design exploits the separation between glasses on the Abbe diagram. By combining a crown element (high V_d) with a flint element (low V_d), the designer cancels first-order chromatic aberration at two wavelengths. The greater the Abbe number separation between the two glasses, the weaker the individual element powers required — and weaker powers generally mean less spherical aberration and easier fabrication [1, 2]. For a complete treatment of achromatic doublet design, apochromats, and multi-element lens forms, see the Complex Lens Assemblies guide.

3.4Partial Dispersion and Anomalous Glasses

The Abbe number characterizes dispersion across the full visible band, but it says nothing about the shape of the dispersion curve within that band. Two glasses with identical Abbe numbers can have different dispersion profiles — one may disperse more strongly in the blue, the other more uniformly. This variation is captured by partial dispersion ratios [1, 5].

Where $P_{g,F}$ is the partial dispersion ratio (dimensionless), $n_g$ is the refractive index at the mercury g-line (435.83 nm), $n_F$ is the refractive index at 486.13 nm, and $n_C$ is the refractive index at 656.27 nm.

For most optical glasses, $P_{g,F}$ falls along a nearly linear relationship when plotted against $V_d$ — the so-called “normal line.” Glasses that deviate significantly from this normal line are called anomalous dispersion glasses. These anomalous glasses (e.g., Schott N-KZFS or FK series) are essential for designing apochromatic lenses that correct chromatic aberration at three wavelengths rather than the two achievable with a standard achromatic doublet [1, 5, 6].

The transmission range of an optical material — the spectral band over which it passes light with acceptably low absorption — is often the first filter in material selection. A material that absorbs at the working wavelength is simply unusable, regardless of its other properties [1, 3, 4].

4.1UV Materials

Applications in the ultraviolet (below 400 nm) demand materials free of electronic absorption bands at short wavelengths. Ordinary optical glasses such as N-BK7 absorb strongly below approximately 330 nm, making them unsuitable for UV work [5, 9].

UV-grade fused silica (synthetic amorphous SiO₂) is the workhorse UV material. Manufactured from silicon tetrachloride by flame hydrolysis, UV-grade fused silica transmits from approximately 185 nm to 2.1 μm with extremely low absorption in the deep UV. Its laser damage threshold is high, its thermal expansion coefficient is the lowest of any common optical material (0.55 × 10⁻⁶/K), and its refractive index homogeneity is excellent. The principal limitation is a hydroxyl (OH) absorption band near 2.73 μm that reduces IR-grade performance — specialized low-OH variants (e.g., Suprasil 300) suppress this band for applications requiring combined UV and IR transmission [4, 9].

Calcium fluoride (CaF₂) extends deeper into the UV than fused silica, transmitting from approximately 130 nm to 10 μm. This extraordinary breadth — spanning vacuum UV to thermal IR — makes CaF₂ the material of choice for excimer laser optics at 193 nm and 248 nm, where its low absorption and high damage threshold are critical. CaF₂ is cubic (no birefringence) and has a low refractive index (~1.43 at 589 nm), reducing the need for anti-reflection coatings. The trade-off is a high coefficient of thermal expansion (18.9 × 10⁻⁶/K) and moderate susceptibility to thermal shock [4, 7, 9].

Magnesium fluoride (MgF₂) pushes the UV cutoff even further, transmitting from approximately 115 nm to 7 μm. However, MgF₂ is tetragonal and therefore birefringent, requiring attention to crystal axis orientation. It is commonly used for UV windows and as a thin-film coating material (λ/4 layers for broadband anti-reflection coatings) rather than as a bulk lens substrate [4, 9].

4.2Visible and Near-IR Materials

The visible and near-infrared region (approximately 350 nm to 2.1 μm) is served by the largest selection of optical materials. Nearly all optical glasses — hundreds of catalog compositions from Schott, Ohara, Hoya, and CDGM — transmit well across this band [5, 6].

N-BK7, the most widely used optical glass in the world, typifies the category. Its useful transmission extends from approximately 330 nm to 2.1 μm, with internal transmittance exceeding 99% per centimeter across most of the visible spectrum. N-BK7 is inexpensive, widely available in many form factors, easy to fabricate using conventional grinding and polishing, and optically well-characterized with tightly controlled refractive index tolerances. For general-purpose visible-light optics — lenses, windows, prisms, beamsplitters — N-BK7 is the default starting point [5, 6, 9].

UV fused silica also performs well across the visible and near-IR, with several advantages over N-BK7: deeper UV transmission, lower thermal expansion, better index homogeneity, and higher laser damage threshold. Its higher cost and slightly lower refractive index (1.458 vs. 1.517 at 587 nm) are the primary reasons N-BK7 remains dominant for cost-sensitive visible-light applications [4, 9].

4.3Mid-Wave and Long-Wave IR Materials

Beyond approximately 2.5 μm, silicate glasses become opaque and fundamentally different materials are required. The two principal atmospheric transmission windows — the mid-wave infrared (MWIR, 3–5 μm) and long-wave infrared (LWIR, 8–12 μm) — drive the selection of infrared optical materials for thermal imaging, spectroscopy, and defense applications [4, 7].

Germanium (Ge) is the dominant LWIR material. It transmits from approximately 2 to 16 μm, with low absorption across both the MWIR and LWIR windows. Its exceptionally high refractive index (~4.0 at 10.6 μm) means that uncoated germanium surfaces reflect approximately 36% of incident light, making anti-reflection coatings essential. Germanium's transmission degrades above approximately 100°C due to thermally generated free carriers, limiting its use in high-temperature environments [4, 7, 8].

Silicon (Si) serves the MWIR window (approximately 1.2–7 μm). Its refractive index (~3.42 at 4 μm) is lower than germanium's, reducing Fresnel losses. Silicon is harder, less dense, and has higher thermal conductivity than germanium, making it suitable for high-power and weight-sensitive applications. Its 3–5 μm transmission makes it the standard substrate for MWIR imaging windows [4, 7].

Zinc selenide (ZnSe) offers the broadest infrared transmission of common materials, spanning approximately 0.5 to 20 μm. This range covers visible through LWIR, enabling visual alignment of IR optical systems — a significant practical advantage. ZnSe is the standard lens and window material for CO₂ laser systems operating at 10.6 μm. Its relatively low refractive index (~2.40 at 10.6 μm) reduces Fresnel losses compared to Ge. The primary limitation is ZnSe's softness (Knoop hardness ~120), which makes it susceptible to scratching and limits its use in harsh environments without protective coatings [4, 7, 8].

4.4Broadband and Multi-Spectral Materials

Some applications require a single material that transmits across multiple spectral regions. Two materials stand out for broadband performance [4, 7, 9].

Calcium fluoride spans from the vacuum UV (130 nm) to the thermal IR (10 μm), covering UV, visible, NIR, SWIR, and part of the MWIR in a single isotropic crystal. This breadth makes CaF₂ the most versatile broadband optical material, though its thermal fragility and moderate hardness limit its use in demanding mechanical environments [4, 7].

Sapphire (single-crystal Al₂O₃) transmits from approximately 170 nm to 5.5 μm with exceptional mechanical hardness (Mohs 9, Knoop ~1800). Its transmission covers UV, visible, NIR, SWIR, and part of the MWIR. Sapphire's hardness and thermal conductivity make it the material of choice for missile domes, high-pressure windows, and any application where optical transmission must be maintained under severe mechanical or thermal stress. The trade-off is birefringence (sapphire is uniaxial) and relatively high cost due to the difficulty of growing and polishing such a hard material [4, 7, 8].

Optical performance alone does not determine material suitability. The mechanical and thermal properties of the substrate influence fabrication methods, mounting strategies, environmental survivability, and system weight [2, 3, 4].

5.1Hardness and Scratch Resistance

Hardness determines a material's resistance to scratching during handling, cleaning, and use, and also affects the feasibility of different fabrication processes. The Knoop hardness test (HK), which measures resistance to indentation under a specified load, is the standard metric for optical materials [3, 4].

Sapphire leads common optical materials with a Knoop hardness of approximately 1800 kg/mm². Silicon (~1150) and germanium (~780) are moderately hard. Optical glasses fall in the range of 400–700 (N-BK7 ≈ 610). Infrared materials like ZnSe (~120) and ZnS (~250) are significantly softer, requiring careful handling and often benefiting from protective diamond-like carbon (DLC) or hard oxide coatings [3, 4, 7].

Hardness also correlates with coating adhesion — harder substrates generally accept thin-film coatings more readily and withstand coating-induced stress better. Soft substrates like ZnSe may require specialized coating processes or lower-stress coating designs [3]. For a comprehensive treatment of coating materials, deposition methods, and performance specifications, see the Optical Coatings guide.

5.2Thermal Expansion and Thermal Shock

The coefficient of thermal expansion (CTE) describes how a material's dimensions change with temperature. For optical systems, thermal expansion affects alignment (the spacing between elements changes), element shape (surface curvature changes), and stress at bonded interfaces where materials with different CTEs are joined [2, 3].

Where $\Delta L$ is the change in length (m), $L_0$ is the original length (m), $\alpha$ is the linear coefficient of thermal expansion (K⁻¹), and $\Delta T$ is the temperature change (K).

Fused silica's CTE of 0.55 × 10⁻⁶/K is the lowest of common optical materials, making it the preferred choice for dimensionally critical applications. N-BK7's CTE of 7.1 × 10⁻⁶/K is moderate. Calcium fluoride (18.9 × 10⁻⁶/K) and many infrared materials expand significantly more, placing greater demands on mount design for systems that experience temperature swings [3, 4].

Thermal shock resistance is related to, but distinct from, CTE. A material with high CTE but also high thermal conductivity and mechanical strength (such as sapphire) may survive rapid temperature changes better than a material with lower CTE but poor conductivity (such as CaF₂). The thermal shock resistance parameter $R_s = \sigma(1-\nu)\kappa / (\alpha E)$, where $\sigma$ is tensile strength, $\nu$ is Poisson's ratio, $\kappa$ is thermal conductivity, $\alpha$ is CTE, and $E$ is Young's modulus, provides a relative ranking [3, 4].

5.3Density and Weight

Density directly determines the mass of optical components and thus the weight of the optical assembly. For space-based, airborne, and handheld systems, weight is a primary constraint [3, 4].

Fused silica (ρ = 2.20 g/cm³) and N-BK7 (ρ = 2.51 g/cm³) are relatively light. Silicon (2.33 g/cm³) is comparable. Germanium (5.33 g/cm³) and ZnSe (5.27 g/cm³) are considerably heavier — a germanium lens weighs more than twice the equivalent glass lens, which must be accounted for in system weight budgets and mount stiffness calculations. Sapphire (3.98 g/cm³) and CaF₂ (3.18 g/cm³) fall in between [3, 4].

Long-term optical performance depends on a material's resistance to its operating environment. Moisture, chemical exposure, and high-intensity laser radiation can all degrade transmission, scatter, and surface quality over time [3, 4, 5].

6.1Hygroscopy and Moisture Sensitivity

Some optical materials absorb atmospheric moisture, leading to surface haze, loss of polish, and eventual degradation of transmission. This property — hygroscopy — is a critical concern for materials intended for use in unprotected or humid environments [3, 4].

Alkali halide crystals (NaCl, KBr, KCl) are extremely hygroscopic and require protected environments or sealed housings. These materials are used in FTIR spectroscopy as beam splitter substrates and sample windows, where their broad IR transmission is essential and controlled environments are practical [4, 7].

Among common photonics materials, calcium fluoride is mildly hygroscopic — not enough to preclude normal laboratory use, but sufficient to warrant care in long-term outdoor or high-humidity deployments. Barium fluoride is somewhat more sensitive. Fused silica, N-BK7, sapphire, silicon, germanium, and ZnSe are all effectively non-hygroscopic under normal conditions [3, 4].

6.2Chemical Resistance

Schott classifies the chemical resistance of optical glasses using standardized tests that rate resistance to acids (AR class, ISO 8424), alkalis (SR class), phosphate-containing solutions (PR class), humidity and atmospheric moisture (CR class — climatic resistance), and surface staining (FR class). Each is rated on a scale from 1 (highest resistance) to 4 or 5 (lowest). N-BK7 scores well across all categories (CR 2, FR 2, SR 2, AR 3), making it durable in most environments without specialized protective coatings [5, 6].

Fluoride crystals, while resistant to most acids, can be attacked by strong bases. ZnSe is inert to most chemicals but dissolves in strong mineral acids. Germanium oxidizes slowly in air, forming a thin GeO₂ layer that is generally benign but can scatter light at very high performance levels — anti-reflection coatings often serve double duty as oxidation barriers on germanium optics [3, 4, 7].

6.3Laser Damage Threshold

The laser-induced damage threshold (LIDT) quantifies the maximum laser fluence (J/cm²) or irradiance (W/cm²) a material can withstand without permanent damage. Damage can occur at the surface (typically at lower thresholds due to polishing defects, contamination, and sub-surface damage) or in the bulk (at higher thresholds, driven by material absorption and inclusion content) [1, 3, 4].

Fused silica has the highest laser damage threshold among common transmissive materials, typically 10–50 J/cm² at 1064 nm for 10 ns pulses (surface LIDT, depending on surface quality and coating). CaF₂ is also excellent for UV excimer laser applications. Optical glasses such as N-BK7 have moderate LIDT, adequate for most applications but limiting for high-energy pulsed lasers. Infrared materials vary widely: silicon and germanium have moderate LIDT, while ZnSe can be damaged at lower fluence levels due to its lower hardness and thermal conductivity [3, 4, 9].

This section profiles the seven most widely used optical materials in photonics, providing the specific property values, transmission characteristics, and application guidance that engineers need for practical material selection [3, 4, 5, 7, 8, 9].

7.1N-BK7 (Borosilicate Crown)

N-BK7 is a lead-free and arsenic-free borosilicate crown glass manufactured by Schott (equivalent designations: H-K9L from CDGM, S-BSL7 from Ohara). It is the most widely used optical glass in the world [5, 6, 9].

Key properties: n_d = 1.5168, V_d = 64.17, density = 2.51 g/cm³, CTE = 7.1 × 10⁻⁶/K, Knoop hardness ≈ 610 kg/mm², thermal conductivity = 1.11 W/m·K. Useful transmission: 330 nm to 2.1 μm. Internal transmittance exceeds 0.998 per 10 mm across the visible spectrum [5, 6].

Best for: general-purpose visible and NIR lenses, windows, prisms, beamsplitters, and beam-steering optics. Not suitable for UV (below 330 nm), IR (beyond 2.1 μm), or high-power pulsed laser applications requiring maximum damage threshold [5, 6, 9].

7.2Fused Silica (UV and IR Grades)

Fused silica is amorphous silicon dioxide (SiO₂) manufactured by flame hydrolysis of SiCl₄ (synthetic) or by fusion of natural quartz. Unlike crystalline quartz, fused silica is isotropic — no birefringence [4, 9].

Key properties: n_d = 1.4585, V_d = 67.8, density = 2.20 g/cm³, CTE = 0.55 × 10⁻⁶/K, Knoop hardness ≈ 500 kg/mm², thermal conductivity = 1.38 W/m·K. UV-grade transmission: 185 nm to 2.1 μm. IR-grade (low-OH) transmission extends to approximately 3.5 μm but has a UV cutoff near 260 nm [4, 9].

Fused silica's combination of deep UV transmission, low thermal expansion, high laser damage threshold, and excellent homogeneity makes it the standard for UV optics, high-energy laser optics, and precision interferometric components. It costs more than N-BK7 but is justified wherever UV performance, thermal stability, or damage resistance matters [4, 9].

7.3Calcium Fluoride (CaF₂)

CaF₂ is a cubic fluoride crystal grown using the vacuum Stockbarger technique. Its lack of birefringence and extraordinary transmission breadth make it the most versatile broadband optical material [4, 7, 9].

Key properties: n_d = 1.4339, V_d = 95.0, density = 3.18 g/cm³, CTE = 18.85 × 10⁻⁶/K, Knoop hardness ≈ 158 kg/mm², thermal conductivity = 9.71 W/m·K. Transmission: 130 nm to 10 μm [4, 7, 9].

The high Abbe number (95.0) means very low dispersion, making CaF₂ an excellent choice for achromatic and apochromatic designs. It is the standard material for excimer laser optics at 193 nm and 248 nm. The primary limitations are high thermal expansion, moderate softness, and mild hygroscopy. Deep-UV grade CaF₂ requires extreme material purity, driving costs above standard IR-grade material [4, 7, 9].

7.4Sapphire (Al₂O₃)

Sapphire is single-crystal aluminum oxide (corundum) grown by the Czochralski, Verneuil, or edge-defined film-fed growth (EFG) methods. It is the second-hardest natural material after diamond [4, 7, 8].

Key properties: n_o = 1.7681 / n_e = 1.7600 at 589 nm (birefringent, uniaxial negative), density = 3.98 g/cm³, CTE = 5.3 × 10⁻⁶/K (perpendicular to c-axis), Knoop hardness ≈ 1800 kg/mm², thermal conductivity = 46.0 W/m·K. Transmission: 170 nm to 5.5 μm [4, 7, 8].

Sapphire's extraordinary hardness, high thermal conductivity, and broad transmission make it the premium choice for windows in harsh environments — missile domes, high-pressure viewports, and scratch-resistant protective covers. The birefringence (Δn ≈ −0.008) must be managed by specifying the c-axis orientation relative to the optical path [4, 7, 8].

7.5Germanium (Ge)

Germanium is a semiconductor crystal used almost exclusively for infrared optics. It is completely opaque in the visible — there is no way to visually inspect a germanium optic for defects without IR instrumentation [4, 7, 8].

Key properties: n ≈ 4.003 at 10.6 μm, density = 5.33 g/cm³, CTE = 5.7 × 10⁻⁶/K, Knoop hardness ≈ 780 kg/mm², thermal conductivity = 59.9 W/m·K. Transmission: 2–16 μm. Useful operating temperature: below ~100°C (transmission degrades due to free carrier absorption at elevated temperatures) [4, 7, 8].

7.6Zinc Selenide (ZnSe)

ZnSe is a cubic (zinc-blende) compound semiconductor produced by chemical vapor deposition (CVD). Its broad IR transmission and moderate refractive index make it a versatile infrared material [4, 7, 8].

Key properties: n ≈ 2.403 at 10.6 μm, density = 5.27 g/cm³, CTE = 7.1 × 10⁻⁶/K, Knoop hardness ≈ 120 kg/mm², thermal conductivity = 18.0 W/m·K. Transmission: 0.5–20 μm [4, 7, 8].

ZnSe is the standard material for CO₂ laser focusing and output coupling optics. Its visible-range partial transparency (yellowish-orange appearance) allows visual alignment of IR beam paths — a significant practical advantage over germanium or silicon. The primary disadvantage is softness; ZnSe scratches easily and should be handled with care. Diamond-like carbon (DLC) coatings are often applied for environmental protection [4, 7].

7.7Silicon (Si)

Optical-grade silicon is a single-crystal semiconductor, typically float-zone or Czochralski grown and lightly doped (5–40 Ω·cm resistivity) to suppress unwanted absorption bands within the MWIR transmission window [4, 7].

Key properties: n ≈ 3.42 at 4 μm, density = 2.33 g/cm³, CTE = 2.55 × 10⁻⁶/K, Knoop hardness ≈ 1150 kg/mm², thermal conductivity = 163 W/m·K. Transmission: 1.2–7 μm [4, 7].

Silicon's high thermal conductivity (the highest of any common optical material except diamond) makes it suitable for high-power laser windows. Its low density — less than half that of germanium or ZnSe — is advantageous for weight-sensitive applications. The 3–5 μm MWIR transmission window makes silicon the standard substrate for MWIR filters and windows in thermal imaging systems [4, 7].

In crystalline materials, the atomic lattice may impose directional dependence on optical properties. When the refractive index varies with the polarization direction and propagation direction of light, the material is birefringent. Birefringence is absent in glasses (which are amorphous and isotropic) and in cubic crystals (which have sufficiently high symmetry to be optically isotropic), but it is present to varying degrees in all other crystal systems [1, 2, 4].

8.1Ordinary and Extraordinary Rays

In a uniaxial crystal (tetragonal, hexagonal, or trigonal systems), light propagating at an angle to the optic axis (c-axis) separates into two polarization components: the ordinary ray (o-ray), which obeys Snell's law normally, and the extraordinary ray (e-ray), which does not. Each experiences a different refractive index, n_o and n_e respectively [1, 2].

Where $\Delta n$ is the birefringence (dimensionless), $n_e$ is the extraordinary refractive index, and $n_o$ is the ordinary refractive index.

When Δn > 0, the crystal is positive uniaxial (e.g., quartz, MgF₂); when Δn < 0, it is negative uniaxial (e.g., sapphire, calcite). The magnitude of Δn determines the strength of the polarization effect. Calcite (Δn ≈ −0.172) has extraordinarily strong birefringence, making it the historical material of choice for polarizers (Glan-Thompson, Glan-Taylor). Sapphire (Δn ≈ −0.008) is weakly birefringent, which usually allows its use in non-polarization-sensitive applications if the c-axis is oriented parallel to the beam propagation direction [1, 2, 4].

8.2Stress-Induced Birefringence in Glasses

Although glasses are inherently isotropic, mechanical stress introduces anisotropy in the refractive index — a phenomenon called stress birefringence or photoelastic birefringence. Stress can arise from thermal gradients during cooling (residual stress), from mounting forces, or from external mechanical loads [1, 3, 5].

Where $\Delta n$ is the induced birefringence (dimensionless), $C$ is the stress-optic coefficient (Pa⁻¹ or Brewsters; 1 Brewster = 10⁻¹² Pa⁻¹), and $\sigma$ is the mechanical stress (Pa).

For N-BK7, the stress-optic coefficient is approximately 2.77 × 10⁻¹² Pa⁻¹ (2.77 Brewsters). Fine annealing — the controlled, slow cooling process applied to precision optical glass — minimizes residual stress and thus keeps stress birefringence below specified limits. Schott specifies refractive index tolerances and stress birefringence classes that tighten with more careful (and more expensive) annealing schedules [3, 5, 6].

8.3When Birefringence Matters

Birefringence is critical in applications involving polarization: interferometry, polarimetry, ellipsometry, polarization-maintaining fiber coupling, and any system using polarizing beamsplitters or waveplates. In these systems, even small amounts of uncontrolled birefringence introduce wavefront distortion between orthogonal polarization states, degrading contrast, fringe quality, or extinction ratio [1, 2].

For non-polarization-sensitive applications — standard imaging, illumination, laser focusing — low birefringence materials are preferred but not strictly required. The designer should evaluate whether the system's polarization budget can tolerate the birefringence of a crystalline substrate, or whether an amorphous (glass) alternative is more appropriate [1, 2].

Temperature changes alter not just the physical dimensions of optical materials (Section 5.2) but also their refractive index. In many precision systems, this thermo-optic effect dominates the thermal focus shift budget, exceeding the contribution of thermal expansion by an order of magnitude or more [1, 2, 3, 10].

9.1The Thermo-Optic Coefficient (dn/dT)

The thermo-optic coefficient $dn/dT$ describes the rate of change of refractive index with temperature. Values are typically quoted in units of 10⁻⁶ K⁻¹ (ppm/K). Most optical glasses have positive dn/dT, meaning the refractive index increases with temperature. Typical values at room temperature and visible wavelengths: N-BK7 ≈ +1.6 × 10⁻⁶/K, fused silica ≈ +10 × 10⁻⁶/K [5, 6, 10].

Some materials have negative dn/dT — notably calcium fluoride (approximately −10.6 × 10⁻⁶/K). This negative coefficient means that heating CaF₂ decreases its refractive index, which has both challenges (thermal focus shift in the opposite direction from thermal expansion) and opportunities (potential for athermal design by pairing positive- and negative-dn/dT materials) [4, 10].

The thermo-optic coefficient depends on both wavelength and temperature. For infrared semiconductor materials, dn/dT is dramatically larger: germanium's dn/dT is approximately +396 × 10⁻⁶/K at 10.6 μm — more than 200 times larger than N-BK7. This extreme sensitivity makes thermal management critical in germanium-based IR optical systems [4, 7, 10].

9.2Thermal Lensing

When a laser beam passes through an absorbing optical element, the absorbed energy creates a temperature gradient — hottest at the beam center, cooler at the edges. This gradient produces a corresponding refractive index gradient through the thermo-optic effect, causing the element to act as an unintended lens. This phenomenon, called thermal lensing, limits the power handling capability of transmissive optics in high-power laser systems [1, 3].

The thermal focal length of a weakly absorbing element is approximately:

Where $f_{\text{th}}$ is the thermal focal length (m), $K$ is the thermal conductivity (W/m·K), $w$ is the beam radius (1/e²) (m), $P_{\text{abs}}$ is the absorbed power (W), and $dn/dT$ is the thermo-optic coefficient (K⁻¹).

Materials with high thermal conductivity and low dn/dT produce the weakest thermal lensing. Fused silica, despite its relatively large dn/dT, benefits from very low absorption. Silicon's extremely high thermal conductivity (163 W/m·K) minimizes thermal gradients, making it a good choice for high-power MWIR laser windows [1, 3, 4].

9.3Athermal Design Principles

The total thermal focus shift of a singlet lens combines two effects: the change in surface curvature from thermal expansion, and the change in optical power from the thermo-optic effect. These are quantified by the thermal glass constant γ:

Where $\gamma$ is the thermal glass constant (K⁻¹), $dn/dT$ is the thermo-optic coefficient (K⁻¹), $n$ is the refractive index, and $\alpha$ is the coefficient of thermal expansion (K⁻¹).

Selecting the right optical material is a sequential elimination process. Starting from the broadest constraint (wavelength), each successive criterion narrows the field until a small number of candidates remain for final evaluation [2, 3, 4].

10.1Step-by-Step Selection

Step 1 — Wavelength: Determine the operating spectral range. This immediately eliminates entire material families. If the application requires 10.6 μm transmission, all oxide glasses and fused silica are excluded.

Step 2 — Transmission and absorption: Within the materials that transmit at the required wavelength, evaluate internal transmittance. Applications involving long optical path lengths or multiple elements are more sensitive to even moderate absorption.

Step 3 — Refractive index and dispersion: For lens design, select materials based on their position on the Abbe diagram. Achromatic correction requires pairs with well-separated Abbe numbers. High-index materials enable stronger surface curvatures with thinner elements.

Step 4 — Mechanical properties: Evaluate hardness (fabrication and handling), density (weight budget), and fracture toughness (environmental survival). Systems exposed to shock, vibration, or particle impingement require harder substrates.

Step 5 — Thermal properties: Assess CTE (mount design, dimensional stability), dn/dT (thermal focus shift), and thermal conductivity (power handling, thermal gradient severity). Systems with wide operating temperature ranges need materials with low CTE and well-characterized thermo-optic behavior.

Step 6 — Environmental durability: Check hygroscopy, chemical resistance, and whether protective coatings are needed. Outdoor, space, or industrial environments impose harsher requirements than laboratory settings.

Step 7 — Cost and availability: Material cost, blank availability in the required size, and fabrication difficulty (harder materials cost more to grind and polish) all affect the final selection.

10.2Material Comparison Decision Table

10.3Common Pitfalls

Ignoring thermal effects. Designers who select materials based purely on room-temperature optical properties often discover thermal focus shift, stress birefringence, or mount-induced distortion during system testing. Always evaluate dn/dT and CTE alongside refractive index and transmission.

Neglecting Fresnel losses. High-index infrared materials (Ge, Si) lose substantial light to surface reflection. Budget for AR coating performance and specify coating bandwidth requirements early in the design.

Overlooking surface durability. Soft materials like ZnSe require careful handling protocols and often protective coatings. The cost of coating and special handling should be included in the overall component budget.

Assuming all grades are equal. UV-grade fused silica is not the same as IR-grade. Standard CaF₂ may not meet excimer-laser purity requirements. Always specify the grade that matches the application, and verify that the material vendor certifies the required performance level.

Forgetting about weight. A germanium lens weighs 2.4 times more than an equivalent fused silica lens. For multi-element systems, the mass difference accumulates and can drive structural and actuator requirements in mounting hardware.