Manojkishor Pradhan Not sure what the issue is. If you have your plot in Excel you can highlight the points in the linear region and plot the y = mx + c line and note the equation. Setting y = 0 gives the corresponding x value.

What is your specific problem is not clear to me too. However, generally speaking your Tauc analysis must begin with plotting (Absorption x Energy)^n vs. Energy (that is, you must multiply the measured absorption by the hv abscissa as a first step, and then calculate its n-th power).

The exponent "n" is typically 2 when absorption occurs via allowed Indirict transitions and 1/2 when absorption is driven by Direct transitions. If you know which of the two is the dominating transition mode (Direct or Indirect), you can choose the proper exponent "n". Otherwise, you should perform the analysis in both modes and get a GAP value for direct transitions and another one for indirect transistions. Usualyy they differ for few tenths of eV.

Catching a good linear region to carry on the estrapolation to zero-y is not always easy, due to noise or more often to the effective overlapping of the low-energy til (the so called Urbach tail). As a rule-of-thumb, you should select a linear region around the point of the plot where the slope of a tangent is maximized, and this can not be generalized... it is up to your experience and quality of the data (and of the sample of course) to find the best fit.

Hope this can help

bye

Every simple least-square minimisation will give you a linear fit; in Excel you can use the linest function. The main probalem is how to decide the range over which to fit your data as the plots will turn into curves on either side. Most people determine the inflection point and then choose a symmtrical interval around it that gives the best R^2 value

this will you answer

https://www.youtube.com/watch?v=HM4lzO9xrCk

Fit your data in the appropriate linear range using linear least squares to a straight line equation (y =mx + b) You can determine Eopt without extrapolation to the x-axis i.e Eopt = magnitude of b divided by m (both m and b are determined from straight line fitting only without any biased extrapolation)

Example: Suppose your best fit equation is

(alpha hf)^1/2 = 11 (hf) - 18.7 then: Eopt = 18.7 /11 = 1.7 (this equation is equivalent to y = 11(x - 1.7))

You can then roughly check the result by extrapolating to the x-axis.

the explanation which I found

https://www.youtube.com/watch?v=P6Ioh6OunoU&t=392s