Waves & Optics Reference School

Refractive Index Table

Refractive indices for common materials, with the speed of light inside each one and the critical angle for total internal reflection against air.

The refractive index n is how much slower light travels inside a material than in vacuum: n = c/v. Every value below is at the sodium D line, 589 nm, unless the note says otherwise, because an index without a wavelength is incomplete. The speed and the critical angle are calculated from the index, not tabulated separately.

Refractive indices at 589 nm, in increasing order, with the speed of light inside each medium and the critical angle for total internal reflection against air.
Material	Index, n	Speed inside, km/s	Critical angle vs air	Notes
Vacuum	1 (exact)	299,792	n/a	Exactly 1, by definition: n is the vacuum speed divided by the speed in the medium, so vacuum against itself is 1. The only value on this page that is not a measurement.
Air	1.000293	299,705	n/a	Dry air at 0 °C and 101.325 kPa. The index of a gas tracks its density, so this falls to about 1.000277 at 15 °C. Treating air as 1 costs under 0.03 per cent in a Snell’s law calculation.
Ice	1.309	229,024	49.8°	Ordinary hexagonal ice just below 0 °C. Lower than liquid water, which is why ice cubes stay visible in a glass of water rather than vanishing into it.
Methanol	1.329	225,577	48.8°	At 20 °C. Just below water, and like the other light alcohols it reads lower as it warms, so a bench measurement needs its temperature recorded.
Water	1.333	224,901	48.6°	At 20 °C, and the value nearly every optics problem means by water. It moves by roughly 0.0001 per degree Celsius, so four figures cover ordinary room temperature work.
Aqueous humour	1.336	224,396	48.5°	The fluid filling the front of the eye, close enough to water that almost none of the eye’s focusing happens as light crosses into it.
Acetone	1.359	220,598	47.4°	At 20 °C. Volatile enough that a sample left open changes composition within minutes, so a reading that drifts is the solvent evaporating rather than an instrument fault.
Ethanol	1.361	220,274	47.3°	At 20 °C, pure. The index of a water and ethanol mixture does not vary linearly between the two pure liquids, so one reading cannot fix the proportions on its own.
Cornea	1.376	217,872	46.6°	The air to cornea step is the largest in the eye and does about two thirds of its total focusing. Underwater that step nearly disappears, which is why unaided vision blurs.
Magnesium fluoride	1.378	217,556	46.5°	The standard single layer anti-reflection coating, not because it is the ideal index but because it is the lowest durable one available. It takes a crown glass surface from about 4 per cent reflection to roughly 1.3 per cent, where the ideal single layer of 1.23 would cancel it.
Calcium fluoride	1.434	209,060	44.2°	Low index and unusually low dispersion, which is why fluorite elements turn up in apochromatic camera lenses. Transparent from the ultraviolet well into the infrared.
Fused quartz	1.458	205,619	43.3°	Amorphous fused silica, not the crystal. The base material of optical fibre, where the core is doped a fraction of a per cent above the cladding to guide light.
Borosilicate glass	1.474	203,387	42.7°	Corning 7740, the Pyrex composition. Almost exactly the index of glycerol below, which is the classic index matching demonstration: a Pyrex rod in glycerol nearly disappears.
Glycerol	1.4746	203,304	42.7°	At 20 °C. The match to borosilicate glass is a coincidence of two ordinary materials rather than a designed pair. Glycerol is hygroscopic, so a sample that has taken up water reads lower.
PMMA (acrylic)	1.491	201,068	42.1°	Bends light nearly as strongly as crown glass at less than half the weight, which is what makes it the default for large moulded optics and display covers.
Benzene	1.501	199,728	41.8°	At 20 °C. High for such a light hydrocarbon because the ring’s delocalised electrons are easy to polarise, and index tracks polarisability.
Immersion oil	1.515	197,883	41.3°	Formulated to match the coverslip and the objective’s front element, so the oil replaces the air gap and that boundary stops refracting. It is what lets an objective exceed a numerical aperture of 1.
Crown glass	1.5168	197,648	41.2°	Schott BK7, the glass a physics problem means when it says glass, usually rounded to 1.52. Abbe number 64, so dispersion is low for a glass.
Sodium chloride	1.544	194,166	40.4°	Rock salt, quoted to four figures because it sits within about 0.0005 of crystalline quartz and the two cannot be separated by index alone. Used as an infrared window, but only in dry air.
Crystalline quartz	1.5443	194,128	40.4°	Birefringent, so one number is not the whole story: 1.5443 for the ordinary ray against 1.5534 for the extraordinary one, and which applies depends on polarisation.
Polycarbonate	1.585	189,144	39.1°	Higher index than acrylic and far tougher, which is why safety lenses use it. Its dispersion is high for a plastic, so thick prescriptions show colour fringes at the edge of the field.
Polystyrene	1.590	188,549	39.0°	The higher index half of most injection moulded plastic lens pairs, where it plays the flint to acrylic’s crown. More dispersive than acrylic.
Flint glass	1.620	185,057	38.1°	Schott F2. Lead or barium raises index and dispersion together: Abbe number 36 against crown glass’s 64, and cementing the two is how an achromatic doublet cancels colour.
Carbon disulfide	1.628	184,148	37.9°	At 20 °C, one of the highest index common liquids and strongly dispersive, which is why liquid prisms were once filled with it. Toxic and highly flammable.
Calcite	1.6584	180,772	37.1°	Strongly birefringent: 1.6584 ordinary against 1.4864 extraordinary. That gap is wide enough to split an image in two, which is the double image seen through a calcite crystal.
Dense flint glass	1.728	173,491	35.4°	Schott SF10, Abbe number 28, so it disperses strongly. The high index lets a lens reach a given power with gentler curves, at the cost of more colour to correct.
Sapphire	1.768	169,566	34.4°	Corundum, ordinary ray. Hard enough for watch crystals and scratch resistant windows, where the high index also means a stronger surface reflection than glass.
Cubic zirconia	2.160	138,793	27.6°	The usual diamond simulant, quoted between 2.15 and 2.18 depending on the stabiliser. Its dispersion is higher than diamond’s, so it throws more coloured fire and less white sparkle.
Diamond	2.417	124,035	24.4°	A critical angle of only 24 degrees, so light entering a cut stone bounces internally several times before it can leave. Dispersion of 0.044 spreads that light into colour on the way out.
Moissanite	2.648	113,215	22.2°	Silicon carbide, ordinary ray. Higher index than diamond and more than twice its dispersion at 0.104. It is also birefringent, which is one way to tell the two apart.
Silicon	3.480	86,147	16.7°	Quoted at 1,550 nm. Opaque across the visible band, so a sodium D line value would be meaningless. This is the telecom wavelength value that silicon photonics is built on, and the large step to air is what confines light inside a waveguide.
Germanium	4.003	74,892	14.5°	Quoted at 10,600 nm. A thermal imaging window: opaque to visible light, transparent from roughly 2 µm to 14 µm. The index is so high that an uncoated surface reflects 36 per cent at normal incidence, so these lenses are always coated.

Citing this tool

Last updated 3 October 2026. Add the date you accessed it as well, which a citation of a page that can change asks for.

Reference

ScienceQuest. (2026). Refractive Index Table [Interactive tool]. Retrieved [date] from https://sciencequest.in/tools/refractive-index-table/

BibTeX

@misc{sciencequest-refractive-index-table,
  author       = {{ScienceQuest}},
  title        = {{Refractive Index Table}},
  howpublished = {\url{https://sciencequest.in/tools/refractive-index-table/}},
  year         = {2026},
  note         = {Interactive tool, last updated 3 October 2026. Add the date you accessed it as urldate.},
}

Worked example: why a diamond sparkles

Total internal reflection starts at the critical angle, where a ray leaving a dense medium for air grazes along the surface instead of escaping. Beyond that angle none of it gets out. Diamond has one of the highest indices of any common transparent material, and that pushes the critical angle remarkably low:

Critical angle of diamond against air

sin(θc) = n_air / n_diamond
sin(θc) = 1 / 2.417
sin(θc) = 0.4137
θc = 24.4°

Air is taken as exactly 1. Its real index of 1.000293 moves the answer by less than 0.02 degrees.

About 24 degrees. Compare crown glass at 41 degrees and water at 49. A ray inside a diamond has to hit a facet within 24 degrees of the normal to leave at all, so the great majority of rays reflect, travel to another facet, and reflect again. A cutter’s job is to arrange the facets so that light bouncing around inside eventually leaves through the top, and it is that internal ricochet, not the surface, that makes a cut stone look bright. Dispersion does the rest: diamond’s index varies by 0.044 across the visible band, so each bounce spreads the light a little further into colour, which is the fire seen alongside the sparkle.

Why the index is essentially never below 1

n = c/v, and light interacts with the electrons in matter on its way through, which slows the wave down. Vacuum is the reference at exactly 1, so anything with electrons in it comes out above 1. Air only just: 1.000293, close enough that treating it as 1 changes a Snell’s law answer by under 0.03 per cent.

Indices below 1 are not forbidden by relativity, and they do occur. For X-rays the index of glass is very slightly under 1, which is why grazing incidence mirrors work at all, and engineered metamaterials can be pushed further. What is moving in those cases is the phase velocity, which carries no information; the signal still travels no faster than c. For visible light in ordinary media you will not meet one, so a computed index below 1 in a school problem means the ratio has been used upside down.

A single index needs a wavelength attached

The index of any real material depends on wavelength, and that dependence is dispersion. In ordinary glass blue light is slowed more than red, so it is refracted more strongly, which is the entire reason a prism spreads white light into a spectrum and the reason a cheap lens shows coloured fringes.

Because of that, tables adopt one wavelength as the convention: the sodium D line at 589 nm, written n_D. That is what the values here are, so they can be compared with any handbook. Two consequences worth carrying:

An index quoted with no wavelength is incomplete. For most school work the error is small, but for anything designed around a laser it is not.
Dispersion is quoted separately. Optics uses the Abbe number, where a higher number means less dispersion: crown glass is about 64 and flint glass about 36. Gemmology uses a spread instead, giving diamond 0.044 and moissanite 0.104.

An achromatic doublet is the practical use of the difference. Cement a converging crown element to a diverging flint one and the two dispersions largely cancel while some focusing power survives, which is why almost every decent lens is at least two pieces of different glass.

Getting the critical angle, and what it is for

Snell’s law is n₁ sin θ₁ = n₂ sin θ₂. Send light from the denser medium into the lighter one and the refracted ray bends away from the normal, until at some incidence it would need to sit at 90 degrees, flat along the surface. That incidence is the critical angle, θc = arcsin(n₂/n₁), and past it there is no refracted ray at all: everything reflects. The column here uses air for n₂, so it is arcsin(1/n).

This is a threshold, not a gradual effect, and it is the mechanism behind three familiar things:

Optical fibre. A core doped a fraction of a per cent above its cladding gives a critical angle just under 90 degrees, so light launched nearly along the axis stays trapped for kilometres. The guiding is between core and cladding, not glass and air.
Prisms. Crown glass turns at 41 degrees, comfortably below 45, so a 45-45-90 glass prism reflects a beam through 90 degrees with no coating at all. Porro prism binoculars and prism periscopes are built on that.
Cut stones. Diamond’s 24 degrees traps light through several internal bounces, which is what the worked example above computes.

What matters at a boundary is the ratio

A single index tells you nothing on its own. Refraction depends on the ratio of the two indices either side of the surface, so a high index material can bend light hardly at all if what it sits against is also high. Light going from crown glass at 1.517 into polystyrene at 1.590 barely deviates, while the same glass against air bends it sharply.

Two consequences follow. Total internal reflection needs the ray to start on the denser side, which is why vacuum and air have no entry in the critical angle column: there is nothing lighter for them to reflect against. And the strength of the reflection at a surface also depends on the ratio, so germanium at 4.0 against air returns 36 per cent of the light while a glass surface returns about 4 per cent. Matching the two sides suppresses that reflection, which is what an anti-reflection coating does with an intermediate layer.

Index matching makes a boundary vanish

Take the ratio all the way to 1 and the boundary stops being an optical boundary at all. Nothing refracts, nothing reflects, and the surface becomes invisible even though it is still there.

Borosilicate glass at 1.474 and glycerol at 1.4746 are close enough to show this on a bench: lower a Pyrex rod into a beaker of glycerol and the submerged part disappears. Microscopy relies on the same trick deliberately. Immersion oil is formulated to 1.515 to match the coverslip and the objective’s front element, so the air gap that would otherwise refract and reflect is replaced by a continuous medium, which is how an objective reaches a numerical aperture above 1. Fibre optic splices and the fluid used to hide cracks in gemstones work the same way.

Common mistakes

Inverting the ratio. n = c/v, so the index is greater than 1 and the speed inside is smaller than c. Getting v larger than c means the division went the wrong way round.
Using an index without its wavelength. These are 589 nm values. A design at 405 nm or 1550 nm needs data at that wavelength, not this column.
Expecting a critical angle where there is none. arcsin(n₂/n₁) only has a solution when n₁ > n₂. Going from air into glass there is no critical angle and no total internal reflection, however steep the incidence.
Assuming the boundary is always against air. The critical angle for glass against water is much larger than for glass against air. Always put the actual second medium into the ratio.
Thinking the frequency changes. Crossing into a medium the frequency is fixed by the source, so it is the wavelength that shortens by a factor of n. Colour does not change on the way into glass.
Quoting one number for a birefringent crystal. Calcite runs 1.658 one way and 1.486 the other, wide enough to split an image in two. Quartz, sapphire and moissanite are birefringent too, and the values here are the ordinary ray.
Treating a light speed as a measurement. The speed column is derived from the index, so it inherits the index’s precision. Four figures of index cannot support six figures of speed.

Common questions

Why is the refractive index never less than one?

Because it is the ratio of the speed of light in vacuum to its speed in the material, and light travels slower through matter than through vacuum. Indices below one do exist for X-rays and in some engineered materials, but for visible light through ordinary media the value is always above one. Vacuum is exactly one by definition, and air is 1.0003, close enough that most calculations treat it as one.

Why does one material have several values listed?

Because the index depends on wavelength, which is dispersion, and it is why a prism separates colours at all. Blue light is refracted more strongly than red in ordinary glass. Tabulated values are usually quoted at the sodium D line, 589 nm, so that different sources can be compared, and that is the convention followed here. For precise work at another wavelength you need the dispersion data, not a single number.

How do I get the critical angle from the index?

The critical angle is arcsin(n₂/n₁), with light travelling from the denser medium n₁ into the lighter one n₂. The column here is calculated against air, so it is arcsin(1/n). Total internal reflection happens beyond that angle, which is what makes an optical fibre work and what makes a diamond sparkle: an index of 2.42 gives a critical angle of only 24 degrees, so light entering it struggles to get back out.

Does a higher index always mean more bending?

For light entering from the same medium at the same angle, yes, a higher index bends the ray further towards the normal. But what actually matters at any boundary is the ratio of the two indices rather than either value alone. Light passing from glass at 1.5 into a plastic at 1.5 does not bend at all, and this is the principle behind index matching, where a fluid of the right index makes a boundary optically vanish.