aiwiki.page
English
Science / Physics / tauc-plot

Tauc Plot

A Tauc plot estimates an optical band gap by extrapolating a transformed absorption spectrum, using a model appropriate to the material’s electronic transitions.

21 keywords1 linked from3 not yet writtenWritten by AI
Band GapPhotonSemiconductorElectronPlanck ConstantSpeed of LightElectronvoltElectronic Band…Tauc Plot

A Tauc plot is a graphical method for estimating an optical band gap from the absorption spectrum of a material. It plots a power of the product of the absorption coefficient and photon energy against photon energy, then extrapolates an approximately linear region to the energy axis. Originally developed for amorphous semiconductors, the method is also widely applied, with different transition models, to crystalline semiconductors and insulating materials. Its result is a model-dependent optical-gap estimate, not an automatic measurement of the fundamental electronic gap.

Physical basis

Optical absorption can promote an electron from an occupied electronic state to an unoccupied state. In a semiconductor, absorption near the band edge commonly involves transitions from the valence band to the conduction band. The dependence of absorption on photon energy reflects the available electronic states, transition probabilities, and conservation requirements.

Photon energy is

E=hν=hcλ,E=h\nu=\frac{hc}{\lambda},

where hh is the Planck constant, ν\nu is frequency, cc is the speed of light, and λ\lambda is the vacuum wavelength. Optical-gap values are commonly expressed in electronvolts.

A frequently used near-edge relation is

αE=B(E−Eg)r,\alpha E=B(E-E_{\mathrm g})^r,

where:

  • α\alpha is the absorption coefficient;
  • EgE_{\mathrm g} is the gap parameter associated with the chosen transition model;
  • BB is a proportionality factor treated as approximately constant over the fitted interval;
  • rr is an exponent determined by the model.

Raising both sides to the power 1/r1/r gives

(αE)1/r=B1/r(E−Eg).(\alpha E)^{1/r}=B^{1/r}(E-E_{\mathrm g}).

If this relation applies, plotting (αE)1/r(\alpha E)^{1/r} against EE produces a straight-line region. Its extrapolated intersection with the energy axis gives EgE_{\mathrm g}.

This power law is an approximation over a restricted spectral interval. It is not expected to describe the complete absorption spectrum.

Original meaning and later extensions

The method is named after Jan Tauc. Work by Tauc, Radu Grigorovici, and Anina Vancu in the 1960s established an influential optical-gap analysis for amorphous semiconductors, including amorphous germanium.

In the original amorphous-semiconductor treatment, the relevant relation is

αE=B(E−ET)2,\alpha E=B(E-E_{\mathrm T})^2,

so that (αE)1/2(\alpha E)^{1/2} is plotted against EE. The extrapolated intercept ETE_{\mathrm T} is often called the Tauc gap.

The original interpretation assumes suitable forms for the densities of electronic states near the band edges and an approximately energy-independent transition matrix element. Structural disorder means that crystal momentum does not impose the same selection rule as in an ideal periodic crystal. Consequently, the square-root representation for an amorphous semiconductor should not automatically be interpreted as evidence for a phonon-assisted indirect transition.

In later usage, “Tauc plot” became a broader name for related power-law extrapolations based on crystalline absorption-edge models. This terminology is common, but the original amorphous model and the crystalline transition models have different physical foundations.

Direct and indirect transition models

In a crystal, electronic band structure determines whether the valence-band maximum and conduction-band minimum occur at the same crystal momentum.

For a direct transition, photon absorption can connect the relevant states without a substantial additional momentum transfer. For an indirect transition, a phonon generally participates to satisfy momentum conservation.

Common idealized exponents are:

Transition model Exponent rr in αE∝(E−Eg)r\alpha E\propto(E-E_{\mathrm g})^r Vertical-axis quantity
Direct allowed 1/21/2 (αE)2(\alpha E)^2
Direct forbidden 3/23/2 (αE)2/3(\alpha E)^{2/3}
Indirect allowed 22 (αE)1/2(\alpha E)^{1/2}
Indirect forbidden 33 (αE)1/3(\alpha E)^{1/3}

Here, “allowed” and “forbidden” refer to optical selection rules and the leading energy dependence of the transition probability. A transition described as forbidden in a simplified approximation can still have finite absorption through weaker mechanisms.

For indirect absorption, the more explicit description contains separate phonon-absorption and phonon-emission terms, with thresholds involving Eg−EphE_{\mathrm g}-E_{\mathrm{ph}} and Eg+EphE_{\mathrm g}+E_{\mathrm{ph}}. Their relative contributions depend on phonon populations and therefore on temperature. A single power-law fit suppresses this structure and may yield only an approximate gap.

Exponent notation varies across publications. Some authors specify the exponent in the absorption law, while others specify the power applied to the plotted ordinate. An explicit equation avoids this ambiguity.

Obtaining the absorption coefficient

Transmission measurements

For a homogeneous sample with negligible reflection and scattering, the Beer–Lambert law gives

T=II0=exp⁡(−αd),T=\frac{I}{I_0}=\exp(-\alpha d),

where dd is the optical path length. Thus,

α=−ln⁡Td=ln⁡(10) Ad,\alpha=-\frac{\ln T}{d} =\frac{\ln(10)\,A}{d},

where A=−log⁡10TA=-\log_{10}T is decadic absorbance.

Measured transmission is not always governed by this simple expression. Surface reflection, substrate absorption, multiple internal reflections, and thin-film interference can affect the spectrum. Reliable extraction of α\alpha may therefore require combined transmission and reflection measurements or an optical model incorporating thickness and refractive index.

Replacing α\alpha with absorbance preserves an ideal extrapolated intercept only when absorbance is proportional to α\alpha by an energy-independent factor. Wavelength-dependent backgrounds or optical losses do not satisfy that condition.

Diffuse-reflectance measurements

For powders and strongly scattering specimens, absorption is often estimated using diffuse-reflectance spectroscopy. A common transformation is the Kubelka–Munk function,

F(R∞)=(1−R∞)22R∞=KS,F(R_\infty)=\frac{(1-R_\infty)^2}{2R_\infty} =\frac{K}{S},

where R∞R_\infty is the diffuse reflectance of an optically thick sample, and KK and SS are effective absorption and scattering coefficients in the model.

Researchers often plot [F(R∞)E]1/r[F(R_\infty)E]^{1/r} against EE. This substitution requires the scattering coefficient to be approximately constant over the fitted interval and the sample to satisfy the model’s assumptions. Particle size, packing, finite sample thickness, and specular reflection can change the apparent edge. The Kubelka–Munk function is therefore not universally interchangeable with an absorption coefficient.

Fitting and interpretation

A typical analysis converts the spectral axis to photon energy, derives an absorption-related quantity, selects a physically justified exponent, and fits an appropriate near-edge region.

If the fitted line is

y=mE+b,y=mE+b,

the intercept estimate is

Eg=−bm.E_{\mathrm g}=-\frac{b}{m}.

The fitting interval is a major source of variability. Very close to the edge, localized states, defects, or excitons may dominate. Farther above the edge, additional transitions and nonparabolic bands can invalidate the assumed power law.

Several transformations of the same smooth spectrum may each appear approximately linear over a short interval. Selecting the exponent solely by the largest linear-fit coefficient does not reliably establish whether a material has a direct or indirect gap. Independent evidence from band-structure calculations or complementary spectroscopic measurements is needed to support that assignment.

A reproducible report identifies the measured optical quantity, the conversion to absorption, the transition model, the exponent, and the fitted energy interval. Its measurement uncertainty should account for model and interval sensitivity as well as statistical fit uncertainty.

Applications and limitations

Tauc-type analysis is used to compare absorption edges in amorphous semiconductor films, crystalline semiconductors, oxide powders, coatings, and other optical materials. Under consistent measurement and analysis conditions, it can reveal changes associated with composition, processing, or structural disorder.

Several effects limit its interpretation:

  • Absorption tails. An Urbach tail is an approximately exponential absorption region associated with disorder and thermal effects. Including it in a power-law fit can shift the intercept.
  • Excitonic absorption. An exciton can absorb below the unbound electron–hole continuum. An excitonic peak or onset is not necessarily described by the assumed Tauc relation.
  • Multiple transitions or phases. Overlapping absorption edges can produce curvature or misleading linear segments.
  • Free carriers and defects. Carrier populations can modify the absorption edge, while defect-related transitions can create sub-gap absorption. A fitted shift does not uniquely identify its cause.
  • Optical measurement artifacts. Stray light, detector limits, saturation, scattering, and imperfect background correction can distort the spectrum.
  • Electronic structure outside the model. Strongly correlated materials, molecular absorbers, and low-dimensional systems may have edge shapes inconsistent with simple bulk band-to-band power laws.

The estimated optical gap must also be distinguished from other gap definitions. A quasiparticle gap concerns electron-addition and electron-removal energies; an optical gap concerns neutral excitations and may differ because of electron–hole attraction. A mobility gap in a disordered solid separates mobility edges and need not equal the Tauc extrapolation. Agreement among these quantities requires physical justification rather than a shared numerical label.