But why would it follow the langmuir isotherm? 

And actually i have to calculate the constant by using the slope and intercept !

you can calculate the corellation coeffiecnt (R-square) from the best fit line for the two isotherms. The isotherm with greater (R-square) value is the best isotherm .

I have done that but the correlation value for freundlich is negative I could not understand the reason behind that ! While if keeping in view the n value which is 3.2 it is phsiosorption which is usually  multi layer. So should the isotherm be Freundlich?

The problem is with the experimental errors. If they are high enough, any equation can provide a good or equally bad fit.  Normally, Freundlich equation should give good fitting results, as this equation was derived empirically first for adsorption on activated carbon from water, and acetic acid is an ideal adsorbate for this.

 The Langmuir and Freundlich equations are different

approaches to describing the same phenomenon --

coverage of activated carbon pore surface with single

molecules attached to those pores. Surface coverage is of

a monolayer of solvent; there are no multiple layers of

solvent in the phenomena described by the Langmuir and

Freundlich equations; nor is volume of adsorbed solvent

directly indicated. 

The Langmuir Equation, Freundlich Equation, or the BET

Equation can be used to adequately represent experimental

data. However, only the Langmuir or Freundlich equations

are recognized as being reliable at very low solvent

concentrations in the influent gas.

As with the Langmuir Equation, the Freundlich Equation

and the BET Equation can be viewed as being linear.

**    The Langmuir Equation is linear in reciprocal

vapor concentration, and reciprocal amount of

adsorption. The slope and intercept of linear plots have

meaning.

** The Freundlich Equation is linear in the logarithm of

vapor concentration, and the logarithm of amount of

adsorption. The slope and intercept of linear plots do

not have meaning.

**   The BET Equation is linear in vapor concentration, and

a function of both amount of adsorption and vapor

concentration.

So my preference would be the Langmuir equation because it is based on an assumed mechanism (single layer adsorption) whereas the Freundlich equation is a correlation.

Freundlich model may be the right model to apply for organic chemicals with large concentration range.