Birefringence is the optical property of a material in which two polarization modes of light propagating in the same direction experience different values of the refractive index. The modes consequently accumulate different phases and may also travel along different paths. A familiar example is the doubled image seen through a clear crystal of calcite. Birefringence does not always produce visibly separated images: it can instead change the polarization of a beam while its components remain spatially overlapped. It occurs naturally in many crystals and ordered materials, and can also be induced by mechanical stress or applied fields.
Physical basis
In an optically isotropic medium, the refractive index at a given frequency is independent of propagation direction and polarization. In an anisotropic medium, the electrical response depends on direction. For a transparent, nonmagnetic material, this response can be described by a permittivity tensor relating the electric field to the electric displacement :
where is the vacuum permittivity and is the relative permittivity tensor.
Combining this relation with Maxwell’s equations generally gives two allowed propagation modes for a specified wave-vector direction. In a transparent, non-optically-active material, these modes can usually be treated as orthogonal linear polarization states. Their different phase velocities correspond to different refractive indices:
where is the speed of light in vacuum and is the phase velocity.
For a specified propagation direction and wavelength, the birefringence is often expressed as
Some contexts use the magnitude , while others retain a sign according to an explicitly defined ordering. Birefringence is dimensionless and commonly varies with wavelength, temperature, and mechanical conditions.
A beam excites both modes only if its incident polarization has components along both of their polarization directions. Light aligned with one eigenmode need not split into two observable components.
Uniaxial and biaxial materials
The principal optical properties of a transparent anisotropic material can be represented by an index ellipsoid:
where , , and are principal refractive indices. The symmetry and relative values of these indices distinguish the main types of linear birefringence.
Uniaxial materials
A uniaxial material has two equal principal indices and one distinct principal index. Its two characteristic indices are conventionally written , the ordinary index, and , the principal extraordinary index. Its unique optic-axis direction is the direction along which the two modes have equal phase velocity.
For a general propagation direction:
- The ordinary wave has refractive index , independent of direction.
- The extraordinary wave has an effective refractive index that depends on the angle between the wave vector and the optic axis.
If that angle is , then
Thus, along the optic axis, , and there is no linear birefringence. Perpendicular to the optic axis, .
A uniaxial material is called positive when and negative when . Quartz is a familiar positive uniaxial crystal, while calcite is negative. These classifications concern linear birefringence; other optical effects can coexist.
Biaxial materials
A biaxial material has three different principal refractive indices. It has two optic-axis directions along which the two wave modes have equal refractive indices. Away from these directions, both modes generally have direction-dependent indices, so the ordinary–extraordinary distinction used for uniaxial materials does not apply in the same way.
Crystal symmetry constrains these properties. In the usual linear optical description, cubic crystals are optically isotropic, tetragonal and hexagonal crystals are uniaxial, and orthorhombic, monoclinic, and triclinic crystals are generally biaxial. Trigonal crystals are also uniaxial. Stress, defects, or externally applied fields can lower the effective optical symmetry.
Double refraction and beam walk-off
At an interface, the two modes can undergo different refraction, producing separate transmitted beams. When both beams emerge with sufficient spatial separation, an object viewed through the material appears doubled.
The geometry is subtler than applying an isotropic refraction law independently to two constant indices. For the extraordinary wave, the refractive index depends on wave-vector direction. In addition, the direction of energy flow need not coincide with the wave vector, which is normal to surfaces of constant phase. Their angular separation is called walk-off.
Consequently, a parallel-sided birefringent plate can laterally separate two polarization components even when its geometry does not produce a large difference between their emerging directions. Conversely, a plate can produce substantial phase retardation without resolving the components into separate images. Visible double refraction and polarization retardation are related manifestations of birefringence, but they are not interchangeable observations.
Phase retardation and polarization changes
For a uniform plate in which two orthogonal modes traverse the same thickness , the relative phase delay is
where is the vacuum wavelength and is the relevant mode-index difference. The quantity is the optical path difference, often called the retardation when expressed as a length.
Because the two components recombine with a changed relative phase, linearly polarized input light can become elliptically or circularly polarized.
A wave plate uses this effect deliberately:
- A quarter-wave plate introduces a phase delay of , modulo . With equal incident amplitudes along its two axes, it converts linear polarization into circular polarization.
- A half-wave plate introduces a phase delay of , modulo . It changes the orientation of linearly polarized light.
For an ideal birefringent plate between crossed linear polarizers, the transmitted intensity is
where is the intensity after the first polarizer and is the angle between a principal polarization direction of the plate and that polarizer.
This interference relation explains why a birefringent specimen becomes dark at certain orientations and bright at others. Under white light, the wavelength dependence of retardation produces interference colors. Such colors depend on thickness, orientation, and wavelength-dependent birefringence, not simply on the identity of the material.
Origins and induced birefringence
Intrinsic and structural birefringence
Intrinsic birefringence arises from the directional response of a material’s atomic or molecular structure. It is common in noncubic crystals and in materials with aligned molecular constituents.
Structural, or form, birefringence arises when an arrangement of different materials behaves as an anisotropic effective medium. Examples include aligned fibers, fine layered structures, and suitably designed subwavelength patterns. This effect can occur even when the individual constituents are optically isotropic.
Ordered biological structures, including collagen fibers and cellulose assemblies, can exhibit birefringence associated with molecular alignment and larger-scale organization.
Stress-induced birefringence
An initially isotropic solid can become birefringent under unequal stresses along different directions. This is the basis of photoelasticity.
Under suitable conditions, a stress-optic relation takes the form
where is a material-dependent stress-optic coefficient and is the relevant principal-stress difference. The exact interpretation depends on material behavior and measurement geometry.
Residual stresses can therefore make glass and plastics birefringent even though their unstressed bulk forms are approximately isotropic.
Field-induced birefringence
An applied electric field can change a material’s refractive-index tensor. In suitable noncentrosymmetric crystals, the leading change is linear in the applied field, known as the Pockels effect. A quadratic electro-optic response, called the Kerr effect, can also produce birefringence.
A transverse magnetic field can induce linear birefringence through the Cotton–Mouton effect. These field-induced responses should be distinguished from polarization rotation caused by circular birefringence.
Liquid crystals
Liquid crystals combine molecular orientational order with fluid behavior. Their birefringence can be controlled by changing molecular alignment, often through an electric field. This provides the optical basis for many liquid-crystal displays and tunable retarders.
Related optical effects
Linear birefringence involves different refractive indices for orthogonal linear polarization modes. Circular birefringence involves different indices for right- and left-circular polarization modes. Their relative phase accumulation rotates a linear polarization direction.
Optical activity in chiral media and magnetically induced Faraday rotation are associated with circular birefringence, although their symmetry and reciprocity properties differ. General anisotropic media can have elliptically polarized eigenmodes.
Birefringence is also distinct from:
- Dichroism: different absorption for different polarization states, rather than different phase velocities.
- Optical dispersion: variation of refractive index with wavelength. Birefringence itself usually exhibits dispersion.
- Scattering or depolarization: redistribution of light or loss of a well-defined polarization state, rather than coherent phase accumulation between two modes.
Several of these effects can occur together, especially in absorbing, heterogeneous, or optically active specimens.
Observation, measurement, and applications
In polarized-light microscopy, crossed polarizers reveal birefringent crystals, fibers, and other ordered structures. Rotating a specimen shows extinction positions, while compensators or calibrated retarders help determine retardation and optical orientation. In geology, these observations support mineral identification and the study of rock textures.
Quantitative measurements use polarimetry, compensators, or interferometry. An important limitation is that measured retardation usually gives the product , rather than birefringence alone. Independent thickness information is therefore often necessary. Unknown orientation, multiple layers, absorption, and unresolved spatial variation can complicate interpretation.
Technological applications include:
- Polarization control: wave plates, compensators, and tunable retarders.
- Polarization separation: birefringent prisms that separate orthogonally polarized beams.
- Optical modulation: electro-optic devices that convert an applied voltage into a polarization or intensity change.
- Nonlinear optics: birefringent phase matching, which compensates for refractive-index differences between interacting waves at different frequencies.
- Materials inspection: detection and mapping of residual stress in glass and plastics.
- Fiber optics: polarization-maintaining fibers use controlled birefringence to help preserve light launched into a selected principal mode.
Unwanted birefringence can introduce polarization errors in precision optical systems. In optical fibers, differences in the group delays of polarization modes contribute to polarization-mode dispersion. Because phase birefringence and group-delay birefringence are different quantities, a phase-retardation measurement alone does not fully characterize pulse propagation.
Historical development
In 1669, the Danish scholar Rasmus Bartholin published a study of double refraction in Iceland spar, a transparent form of calcite. Christiaan Huygens subsequently developed a wave-based explanation of ordinary and extraordinary propagation, presented in his Treatise on Light in 1690.
During the nineteenth century, work on polarization and wave optics connected double refraction with the transverse character of light. The later electromagnetic description explained birefringence through anisotropic material response, providing a common framework for crystal optics, induced optical effects, and modern polarization devices.