Optical Glass

What is Optical Glass

Optical glass is an amorphous inorganic solid melted from silica or other glass-forming oxides and refined to a degree of optical perfection that ordinary glass never reaches. The defining property is not strength or appearance but the predictability of two optical constants: the refractive index, which sets how strongly the material bends light, and the dispersion, which sets how that bending varies with wavelength. A lens designer entering a glass into ray-tracing software trusts that the catalog refractive index is correct to within a few parts in ten thousand, and that every blank cut from the melt behaves the same way.

Three engineering demands separate optical glass from a sheet of float glass. First, homogeneity: the refractive index must be uniform throughout the blank, with variations held to parts per million across the clear aperture, because a slow index gradient warps the wavefront just as surely as a polishing error. Second, freedom from defects: bubbles, seeds, stones, and striae, which are thin schlieren of slightly different index left by incomplete mixing, scatter light and must be counted and limited. Third, internal transmittance: the glass must pass light efficiently across its working band, which means melting in low-iron, low-contaminant batches so the bulk does not absorb or yellow.

The history of optical glass is the history of lens making. Crown and flint glasses were distinguished by the eighteenth century, and Chester Moor Hall and John Dollond used them to build the first achromatic doublets that corrected color error. The decisive industrial step came in 1884 in Jena, Germany, when chemist Otto Schott, working with Ernst Abbe and Carl Zeiss, systematically introduced new oxides such as barium and boron to populate previously empty regions of the index-versus-dispersion map. That collaboration created the modern optical glass catalog and the company that still sets its naming conventions. Abbe in turn defined the dispersion number that bears his name and built the Abbe diagram in which every glass is a single point.

In scale, the optical glass industry is small in tonnage but pivotal in value. Annual world production is measured in thousands of tonnes, yet it underpins the entire camera, microscope, telescope, projector, machine-vision, and semiconductor-lithography supply chains. A single high-end camera zoom may contain fifteen or more elements drawn from a dozen different glass types, each chosen so its index and dispersion contribute to a corrected image. The essence of optical engineering is selecting and combining these glasses, which makes reading their datasheets a core procurement skill.

It is worth distinguishing optical glass from two neighboring materials covered elsewhere. Fused silica and fused quartz are single-component silica glasses prized for ultraviolet transmission and thermal stability but offering only one fixed index, so they cannot be color-corrected against each other. Optical crystals such as calcium fluoride, sapphire, germanium, and zinc selenide extend coverage into the deep ultraviolet and infrared where oxide glasses go opaque. Optical glass occupies the broad visible and near-infrared middle ground where the richest selection of indices and dispersions exists.

Crown, Flint and Glass Families

The first division a buyer must understand is crown versus flint. The names are historical: crown glass was once blown into crown-shaped discs, and flint glass originally used ground English flint as its silica source. In the modern catalog the distinction is precise. Crown glasses carry a type name ending in the letter K, from the German word Kron, and have an Abbe number generally above 50, meaning low dispersion. Flint glasses end in the letter F, have an Abbe number below 50, and are denser and more strongly dispersing because they contain heavy oxides. The diagonal line that splits the Abbe diagram between these two groups is the central feature of glass selection.

Within these two camps, makers prefix and suffix the names to mark sub-families by index and dispersion. The table below summarizes the principal families using representative Schott N-series types, whose published constants are widely cited. Indices are quoted at the helium d-line of 587.56 nm.

The borosilicate crown family, of which N-BK7 is the archetype, is the workhorse of optics. Its combination of moderate index near 1.517, high Abbe number near 64, excellent chemical durability, and low cost makes it the default for windows, plano elements, and the positive element of simple achromats. When a designer writes BK7 on a drawing, it usually means good optical performance is needed but no exotic correction. The eco-friendly N-BK7 replaced the older lead-and-arsenic BK7 with nearly identical constants.

The flint families sit on the high-index, low-Abbe side. Light flints and dense flints raise the index well above 1.6 and push the Abbe number down toward 25, which is exactly what is needed for the negative element of an achromat that must cancel the crown's residual color. The very dense flints historically owed their high index to large lead oxide content; modern eco-glasses substitute titanium, niobium, and tungsten oxides to reach the same index without lead or arsenic, which is why the catalog now carries the N-prefix on Schott glasses and S-prefix on Ohara glasses.

The lanthanum families are the high-value specialty glasses. By adding rare-earth lanthanum oxide, makers reach high refractive indices above 1.7 while keeping the Abbe number relatively high, a combination impossible in classical silicate glass. Lanthanum crowns and lanthanum flints enable compact, well-corrected lenses and dominate premium camera and projector designs. They are expensive, often have poorer chemical durability, and frequently need protective coatings, which Chapter 4 examines. The short flints, with anomalous partial dispersion, are the rarer pieces used to push a doublet toward apochromatic, three-color correction.

Glass Codes and the Abbe Diagram

Every optical glass can be located by two numbers: its refractive index n_d and its Abbe number V_d. These two numbers define the Abbe diagram, a scatter plot introduced by Schott in 1923 in which the Abbe number runs along the horizontal axis, conventionally with high values on the left, and the refractive index runs up the vertical axis. Crowns cluster at the upper left and lower left, flints stretch toward the upper right, and the lanthanum glasses occupy the high-index upper band. A lens designer choosing a glass is literally picking a point in this plane.

The Abbe number itself is a dispersion measure. It is defined as V_d equals (n_d minus 1) divided by (n_F minus n_C), where the three refractive indices are measured at three standard spectral lines: the helium d-line at 587.56 nm, the hydrogen F-line at 486.13 nm in the blue, and the hydrogen C-line at 656.27 nm in the red. The denominator, n_F minus n_C, is the principal dispersion, the spread of index across the visible band. A large Abbe number means a small spread and therefore low dispersion. The table below shows how the standard reference lines anchor the constants.

The six-digit glass code compresses these two key numbers into one identifier. The first three digits are the decimal part of n_d rounded to three places, and the last three digits are the Abbe number multiplied by ten and rounded. The borosilicate crown N-BK7, with n_d equal to 1.5168 and V_d equal to 64.17, becomes 517642. The dense flint N-SF6, with n_d 1.8052 and V_d 25.4, becomes 805254. Modern catalogs extend the code with a decimal suffix for density, so N-BK7 appears as 517642.251 to mark its 2.51 g per cubic centimetre. The code is a fast index for searching and comparing, but it deliberately omits transmission, thermal coefficients, and chemical durability, so it never replaces the full datasheet.

The principal dispersion alone does not fully describe a glass, because the index-versus-wavelength curve is not a straight line. Designers correcting more than two colors also consult the partial dispersion, the fraction of the total dispersion that falls between two intermediate lines. Most glasses lie along a near-straight line of normal partial dispersion; the valuable exceptions, the short flints and the fluorophosphate and fluorite-like glasses, deviate from that line and let designers correct the secondary spectrum to build apochromatic systems. This is why a catalog of 120 glasses is not redundant: the off-line glasses, though few, are indispensable for the highest correction.

Because the index depends on temperature, datasheets also publish dn over dT, the thermo-optic coefficient, typically a few parts per million per kelvin, which can be positive or negative depending on composition. In a system that must hold focus across a wide temperature range, this coefficient combines with thermal expansion to drive athermalization, and it is one more reason a glass code alone is insufficient for serious design work.

Chemical, Thermal and Mechanical Properties

Optical glasses differ enormously in how they survive grinding, polishing, coating, cleaning, and humid storage. A designer who selects purely on index and Abbe number, ignoring durability, may specify a glass that fogs on the shelf or stains during fabrication. Makers therefore publish a set of chemical resistance classes that predict surface behavior, and these classes are as load-bearing in procurement as the optical constants. The five Schott classes are summarized below.

The climatic resistance class CR predicts whether a polished surface will haze in humid air. Class 1 glasses such as N-BK7 are essentially immune and can be stored unprotected, while class 3 and 4 glasses must be coated, sealed, or kept in dry cabinets. The acid, alkali, and phosphate classes warn the optical shop which cleaning chemistry will etch the surface. The alkali class AR is defined by the time needed to remove a 0.1 micrometre layer in a caustic soda solution at pH 12 and 50 degrees Celsius, and the phosphate class PR by the same removal in a pentasodium triphosphate detergent solution, so the numbers map directly to cleaning procedures.

There is a strong correlation between optical value and chemical fragility. The most durable glasses are the ordinary borosilicate crowns; the most fragile are often the high-index lanthanum glasses and certain short flints whose exotic oxide chemistry buys optical performance at the cost of surface stability. A buyer sourcing a high-index lanthanum element should assume protective coating and humidity control are part of the process cost, not an afterthought.

Thermal and mechanical properties round out the datasheet. The coefficient of thermal expansion of common crowns and flints runs roughly 7 to 9 ppm per kelvin between 20 and 300 degrees Celsius, which matters when bonding glass to a metal mount or when the part sees thermal cycling. The transformation temperature, the point near which the glass softens enough to be molded or annealed, lies in the broad range of about 400 to 600 degrees Celsius depending on type, and it governs precision molding. Knoop hardness, near 610 for N-BK7, predicts polishing rate and scratch susceptibility. Density, ranging from about 2.5 g per cubic centimetre for borosilicate crowns to over 4 g per cubic centimetre for the densest flints and lanthanum glasses, drives the weight of large lenses and the stress in their mounts.

Internal transmittance is the last property that often forces a material change. Oxide optical glasses transmit well across the visible band but absorb in the ultraviolet, with N-BK7 cutting off below roughly 350 nm. Where deep-UV throughput is required, the design must move to UV-grade fused silica, which transmits to about 180 nm, or to crystals such as calcium fluoride for the very deep UV. In the infrared, oxide glasses absorb beyond about 2.5 micrometres, so thermal imaging optics turn to germanium, zinc selenide, and chalcogenide glasses instead.

Decoding the Glass Datasheet

A complete optical glass datasheet carries dozens of numbers, but only a handful drive a selection and quality decision: the refractive index and its tolerance, the Abbe number and its tolerance, the homogeneity grade, the stress birefringence grade, the bubble and inclusion grade, the striae grade, and the internal transmittance curve. Each of these has a defined tolerance step under the ISO 12123 raw-glass specification and the ISO 10110 series for drawing indications, and the buyer chooses the step that matches the application. The reference values for N-BK7 below show how a real datasheet reads.

The refractive index tolerance is the first step a buyer chooses. A typical catalog blank holds n_d to about plus or minus 0.0005, while precision grades tighten this to a few parts in ten thousand. ISO 12123 now defines narrow tolerance grades for both index and Abbe number, and the buyer specifies the grade that the as-built lens tolerancing demands. Over-specifying the index grade raises cost without optical benefit; under-specifying it can shift focus or color correction out of tolerance, so this is a genuine engineering trade-off rather than a default.

Homogeneity describes how constant the index is across the clear aperture of a single blank, expressed as a maximum index variation in parts per million and an associated ISO grade. Imaging optics tolerate a few parts per million, but interferometer optics and lithography elements demand grades down to fractions of a part per million, which only specially selected melts can deliver. ISO 12123 lets the drawing call out a minimum aperture over which the homogeneity grade must hold, which matters for large blanks.

Stress birefringence, bubbles and inclusions, and striae are the three defect grades now consolidated under ISO 10110 part 18. Stress birefringence is residual optical anisotropy from incomplete annealing, specified as an optical path difference in nanometres per centimetre; polarization-sensitive systems demand the lowest values. Bubbles and inclusions are counted and sized, with the grade limiting the total cross-sectional area of defects in the aperture. Striae are thin layers of slightly different index from imperfect mixing; the standard now allows specifying inspection in a second and third perpendicular direction, because a striae pattern invisible from one axis can scatter badly from another.

The internal transmittance curve, finally, tells the buyer how much light the bulk glass passes at each wavelength for a stated thickness, separate from the surface reflection losses that an anti-reflection coating addresses. A glass that looks colorless can still absorb in the violet or near-infrared, so a system working near the edge of the band must read the curve rather than trust the nominal transmission range. Reading all of these together, rather than fixating on the index alone, is what separates a robust glass specification from one that fabricates poorly or fails inspection.

Selection Decision Factors

To turn the preceding chapters into a defensible glass specification, follow the decision sequence below. Most selection mistakes come not from a single wrong number but from settling the optical design before checking durability, availability, and cost. These eight steps can serve as a fixed checklist when issuing an optical glass enquiry.

1. Optical design point: Fix the required refractive index n_d and Abbe number V_d from the lens design, then locate the candidate glasses as points on the Abbe diagram. For color-corrected systems, identify the crown-plus-flint pairing and, for apochromats, the off-line partial-dispersion glass that closes the secondary spectrum.
2. Index and Abbe tolerance grade: Choose the ISO 12123 tolerance step that the as-built tolerancing demands. Standard blanks hold n_d near plus or minus 0.0005; tighten only where the system error budget requires it, because each tighter grade adds cost.
3. Transmission band: Confirm the internal transmittance covers the working wavelengths with margin. If deep ultraviolet or mid-infrared throughput is needed, step out of oxide glass to fused silica, calcium fluoride, or infrared crystals instead.
4. Chemical durability: Read the CR, FR, SR, AR, and PR classes against the planned fabrication, coating, cleaning, and storage. A poor climatic class means protective coating and dry storage must be budgeted, not assumed away.
5. Defect and homogeneity grades: Specify homogeneity in parts per million, stress birefringence in nanometres per centimetre, plus the bubble, inclusion, and striae grades per ISO 10110, matched to whether the part is imaging, polarization-sensitive, or interferometric.
6. Anneal grade and form: Decide between standard step-down anneal and fine anneal, and between catalog form (strip, block, rod, pressing, or ready-cut blank) and a custom melt. Fine annealing tightens the index and lowers residual stress for precision and polarization optics.
7. Thermal and mechanical fit: Check the coefficient of thermal expansion against the mount, the dn over dT against athermalization needs, and the density against element weight and mount stress, especially for large or vibration-loaded assemblies.
8. Cost, availability and equivalents: Compare the Schott, Ohara, Hoya, and CDGM near-equivalents by code, weighing price, lead time, and whether a second source exists. Verify the substitute against its own datasheet, since equivalent codes are nominally but not exactly equal.

One last commonly overlooked dimension is supply continuity and serviceability. Optical glass types are occasionally discontinued as eco-regulations and demand shift, so a design built on a rare type carries obsolescence risk; checking that at least one mainstream maker keeps the type in regular production protects a long-lived product. The four global houses, Schott of Germany, Ohara and Hoya of Japan, and CDGM of China, between them list more than a hundred types each with published equivalency maps, regional stock, and custom-melt capability, which makes them the dependable starting point for a sourcing decision that must hold for the life of the optical system.