An alternative would be to take the derivative of the absorption (transmission) spectrum and look for its maximum (minimum). It corresponds to the energy for which the slope is most steep. The result might not be the same as that from Tauc plot, but it wold be an assessment.

You can fit the entire stack system (usually air/film/substrate/air) using a transfer matrix programme such as OPTICAL (https://www.bo.imm.cnr.it/users/centurioni/optical.html). If your input for n&k of the film is an oscillator function that contains Eg, such as a Tauc-Lorentz or Cody-Lorentz oscillator, you can fit it. Essentially a more sophisticated version of the Tauc plot, similar to what is usually done in ellipsometry.

http://www.sciencedirect.com/science/article/pii/S0040609014000467

An alternative would be to take the derivative of the absorption (transmission) spectrum and look for its maximum (minimum). It corresponds to the energy for which the slope is most steep. The result might not be the same as that from Tauc plot, but it wold be an assessment.

A very rough, but also very fast, alternative method to the others proposed above is based on the absorption maximum wavelength. You can consider Eg [eV]=1,24 / absoprtion peak wavelength [micrometer]. You can also use it to get a rough extent of the Eg of your thin film, depending on your needs.

@Stefano: I think that only works for organic materials. It's particularly unsuited to materials that have a direct gap at higher energy than their indirect gap, e.g. Si. See link below, the max in eps2 of c-Si is 4.2eV but the band gap is 1.1eV. Similarly the maximum for a-Si is ~3eV while the band gap is typically 1.7eV.

http://www.intechopen.com/source/html/38563/media/image3.jpeg

@Manuel: yeah it's just for direct band gap materials, but not only organics.

For sure this is a very rough formula as it doesn't consider for instance the mismatch between the exciton energy and the mere energy difference Ec-Ev (or LUMO-HOMO in organics). You can check this link for the explanation:

http://books.google.co.jp/books?id=shsGEahT9wMC&pg=PA31&lpg=PA31&dq=lambda+photon%5B%CE%BCm%5D%3D1.24/E+Gap%5BeV%5D&source=bl&ots=mHEnp1Eczo&sig=ptmSp75pTplnNap4heSAP1_Wahg&hl=it&sa=X&ei=onIzVN6ILofWPdfEgZgM&ved=0CCsQ6AEwAg#v=onepage&q=lambda%20photon%5B%CE%BCm%5D%3D1.24%2FE%20Gap%5BeV%5D&f=false

For any good quality crystalline material, one should be able to observe an excitonic bandgap in absorption or transmission. In this case, the excitonic bandgap can be determined without any ambiguity. If you want to know the single particle bandgap, you need to add the exciton binding energy. See the paper of M. D. Sturge (Phys. Rev. 127, 768, 1962. This is perhaps the 1st reported excitonic absorption in a semiconductor, and has been used in many solid state/semiconductor books). In the case the excitonic bandgap is not observable, usually due to poor crystalline quality, the bandgap is not well defined. However, one can still estimate the bandgap, but the value depends what technique is used and how the fitting is done. See this short review paper discussing how different techniques (e.g., linear absorption vs. modulation spectroscopy) yield different bandgaps, and how increased disordering diminishes the excitonic bandgap, making the definition of a bandgap ambiguous (Zhang et al., phys. stat. sol. (b) 240, 396, 2003). One can then define the bandgap based on the purpose.  

May I ask Dr Zhang for a copy of this paper -Zhang et al., phys. stat. sol. (b) 240, 396, 2003?

It has been uploaded to my Researchgate account publication list in p.4. If this link below does not work, feel free to email me at [email protected]. https://www.researchgate.net/publication/229782293_Effects_of_heavy_nitrogen_doping_in_IIIV_semiconductors__How_well_does_the_conventional_wisdom_holdfor_the_dilute_nitrogen_IIIVN_alloys

Article Effects of heavy nitrogen doping in III–V semiconductors – H...

thank you

Thank you all for help!